MIWI2 Is Essential for Spermatogenesis
and Repression of Transposons
in the Mouse Male Germline
Michelle A. Carmell,1,7Ange ´lique Girard,1,6Henk J.G. van de Kant,2,3Deborah Bourc’his,4Timothy H. Bestor,5
Dirk G. de Rooij,2,3and Gregory J. Hannon1,*
1Cold Spring Harbor Laboratory, Howard Hughes Medical Institute, Watson School of Biological Sciences, 1 Bungtown Road,
Cold Spring Harbor, NY 11724, USA
2Department of Endocrinology, Faculty of Biology, Utrecht University
3Department of Cell Biology, University Medical Center Utrecht
3584 CH Utrecht, The Netherlands
4INSERM U741, Institut Jacques Monod, 2 Place Jussieu, 75251 Paris, CEDEX 05, France
5Department of Genetics and Development, College of Physicians and Surgeons, Columbia University, New York,
NY 10032, USA
6Ecole Nationale Supe ´rieure des Mines de Paris, 60 boulevard Saint-Michel, 75272 Paris, France
7Present address: Howard Hughes Medical Institute, Whitehead Institute for Biomedical Research, Massachusetts Institute of
Technology, 9 Cambridge Center, Cambridge, MA 02142, USA.
Small RNAs associate with Argonaute proteins
and serve as sequence-specific guides for
regulation of mRNA stability, productive trans-
lation, chromatin organization, and genome
structure. In animals, the Argonaute superfam-
ily segregates into two clades. The Argonaute
clade acts in RNAi and in microRNA-mediated
gene regulation in partnership with 21–22 nt
RNAs. The Piwi clade, and their 26–30 nt piRNA
partners, have yet to be assigned definitive
functions. In mice, two Piwi-family members
in spermatogenesis. Here, we examine the ef-
fects of disrupting the gene encoding the third
family member, MIWI2. Miwi2-deficient mice
display a meiotic-progression defect in early
prophase of meiosis I and a marked and pro-
gressive loss of germ cells with age. These
phenotypes may be linked to an inappropriate
activation of transposable elements detected
in Miwi2 mutants. Our observations suggest
a conserved function for Piwi-clade proteins in
the control of transposons in the germline.
complex, and are defined by the presence of two domains,
PAZ and Piwi. Phylogenetic analysis of PAZ- and Piwi-
containing proteins in animals suggests that they form two
selection of silencing targets. The second clade is more
similar to Drosophila PIWI. Like Argonautes, Piwi proteins
have been implicated in gene-silencing events, both tran-
scriptional and posttranscriptional.
Piwi-clade proteins have been best studied in the fly,
which possesses three such proteins: PIWI, AUBERGINE,
and AGO3. Until recently, evidence for the involvement of
Piwi proteins in gene silencing was mainly genetic. The
first biochemical insight into the biological role of Piwi-
family proteins was the observation that both PIWI and
AUBERGINE exist in complexes with repeat-associated
siRNAs (rasiRNAs) (Saito et al., 2006; Vagin et al., 2006).
RasiRNAs were first described in Drosophila as 24–
26 nt, small RNAs corresponding to repetitive elements,
including transposons (Aravin et al., 2001, 2003). The in-
teraction between Piwi proteins and rasiRNAs dovetails
nicely with the observation that, in Drosophila, both piwi
and aubergine are important for the silencing of repetitive
Mutations in Piwi-family genes cause defects in germ-
line development in multiple organisms. For example, in
flies, piwi is necessary for self-renewing divisions of germ-
line stem cells in both males and females (Cox et al., 1998;
Lin and Spradling, 1997). Mutations in aubergine cause
male sterility and maternal effect lethality (Schmidt et al.,
to silence the repetitive stellate locus. Mutant testes also
suffer from meiotic nondisjunction of sex chromosomes
and autosomes (Schmidt et al., 1999). A recent study indi-
cates that the sterility observed in female flies bearing mu-
tations in Piwi-family proteins is also likely to result, at
least in part, from the deleterious effects of transposon
activation (Brennecke et al., 2007).
As is seen in other organisms, the expression of the
three murine Piwi proteins, MIWI (PIWIL1), MILI (PIWIL2),
Developmental Cell 12, 503–514, April 2007 ª2007 Elsevier Inc. 503
and MIWI2 (PIWIL4), is largely germline restricted (Kura-
mochi-Miyagawa et al., 2001; Sasaki et al., 2003). Thus
far, MIWI and MILI have been characterized in some de-
tail, with mice bearing targeted mutations in either Miwi
(Deng and Lin, 2002) or Mili (Kuramochi-Miyagawa et al.,
2004) being male sterile. Although both MIWI and MILI
are involved in regulation of spermatogenesis, loss of
either protein produces distinct defects that are themati-
cally different from those seen upon mutation of Drosoph-
ila piwi. Based upon their expression patterns and the re-
ported phenotypes of mutants lacking each protein, the
most parsimonious model is that both MIWI and MILI per-
form roles essential for the meiotic process. So far, no
mammalian Piwi protein has a demonstrated role in
stem cell maintenance as proposed for Drosophila PIWI.
This raised the possibility that any role for mammalian
Piwi proteins in stem cell maintenance might reside in
the third family member, MIWI2.
Despite the presence of conserved RNA-binding motifs
and an expectation that mammalian Piwi proteins might
be involved in RNA-induced silencing mechanisms, no in-
teraction was described for these proteins with siRNAs or
miRNAs. Recently, we and others identified small RNA
binding partners for Piwi proteins in the male germline,
designated as piRNAs (Piwi-interacting RNAs) (Aravin
et al., 2006; Girard et al., 2006; Grivna et al., 2006; Lau
et al., 2006; Watanabe et al., 2006). piRNAs show distinc-
tive localization patterns in the genome. They arepredom-
inantly grouped into 20–90 kb genomic regions, wherein
numerous small RNAs are produced from only one geno-
mic strand. Most piRNAs match the genome at unique
sites, and less than 20% match repetitive elements.
piRNAs become abundant in germ cells around the
pachytene stage of prophase of meiosis I, but they may
be present at lower levels during earlier stages. Unlike
microRNAs, individual piRNAs are not conserved. Thus,
it is difficult to intuit biological function based on the
sequence content of piRNA populations.
To investigate the role of MIWI2 in gametogenesis, we
disrupted the gene encoding this third mouse Piwi-family
member. We find that Miwi2 mutants have two discrete
defects in spermatogenesis. The first is a specific meiotic
block in prophase of meiosis I that exhibits distinctive
morphological features. This is followed by a progressive
loss of germ cells from the seminiferous tubules. These
phenotypes, and the fact that Miwi2 is expressed both in
germline and somatic compartments, highlight similarities
between MIWI2 and Drosophila PIWI. In this regard, we
find that disruption of Miwi2 also interferes with transpo-
son silencing in the male germline.
Miwi2 Mutants Have Defects in Spermatogenesis
We used an insertional mutagenesis strategy to disrupt
the Miwi2 gene(Figures 1Aand 1B). Theallele that wecre-
ated contains a 10 kb segment of vector sequence follow-
genomic region encompassing exons 9–12 is duplicated.
This is predicted to insert multiple in-frame stop codons
and to produce a nonfunctional allele. When primers
downstream of the insertion are used, quantitative
RT-PCR indicates that Miwi2 transcripts are essentially
undetectable in homozygous mutant animals at 10 days
postpartum (dpp), before mutants phenotypically diverge
from wild-type (Figure S1; see the Supplemental Data
available with this article online). This is precisely what
ing on the predicted mRNA containing numerous prema-
ture stop codons. However, all of the coding capacity of
Miwi2 still exists in the mutant genome, and splicing
around the insertion could conceivably produce a func-
tional Miwi2 transcript. Using RT-PCR primers that flank
the duplicated exons, we could not detect any wild-type
sert with confidence that our allele produces, at the very
least, a severe hypomorph and is likely a null allele.
Mice heterozygous for the Miwi2 mutant allele grew to
adulthood, were fertile, and appeared phenotypically nor-
mal. Upon intercrossing, it became obvious that male
mice homozygous for a mutant allele of Miwi2 were infer-
tile, although they exhibited normal sexual behavior.
Homozygous females, however, were fertile and had no
obvious defects. Males and females of both sexes were
of normal size and weight and had the expected life span.
Initial examination of testes of adult Miwi2 mutants re-
vealed a very obvious and severe phenotype. Although
all other reproductive organs were of normal size and ap-
pearance, Miwi2 mutant testes were substantially smaller
than their wild-type or heterozygous counterparts (Fig-
were indistinguishable both morphologically (not shown)
and histologically (Figures 2A and 2B). However, cellular
defects became apparent a few days later as germ cells
proceeded through the first round of spermatogenesis.
Mouse spermatogenesis is a highly regular process that
takes ?35 days to complete (de Rooij and Grootegoed,
1998). Spermatogonia, a very small percentage of which
are stem cells, line the periphery of the seminiferous
tubule and divide mitotically to maintain the stem cell pop-
ulation throughout the lifetime of the animal. These divi-
sions also give rise to differentiating cells that undergo
several rounds of mitotic division before entering meiosis.
Meiotic cells, or spermatocytes, advance through meiotic
prophase I, which can be separated into five phases. In
leptotene (phase 1), duplicated chromosomes begin to
condense. More extensive pairing and the formation of
synaptonemal complexes occur in zygotene (phase 2)
and are completed in pachytene (phase 3), when crossing
over occurs. Homologs begin to separate in diplotene
(phase 4), and chromosomes move apart in diakinesis
(phase 5). Prophase I is followed by two meiotic divisions
that eventually generate haploid products. The immediate
product of meiosis is the round spermatid, which will
mature and elongate until being released into the lumen
of the tubule.
MIWI2 Is Essential for Spermatogenesis
504 Developmental Cell 12, 503–514, April 2007 ª2007 Elsevier Inc.
At the stage when tubules of wild-type siblings con-
tained germ cells at the zygotene and pachytene phases
of meiosis I, germ cells in the mutant became noticeably
atypical (Figures 2C–2F). Two abnormal nuclear morphol-
ogies were observed in mutant spermatocytes. In ?80%
of abnormal spermatocytes, the nuclei were very con-
densed and stained intensely with hematoxylin and DAPI
(Figures 2D and 2F, arrow). The remaining 20% of abnor-
mal nuclei were extremely large and had an ‘‘exploded’’
morphology with apparently scattered chromatin (Figures
2D and 2F, arrowhead). The two types of abnormal nuclei
appear simultaneously. Therefore, it is unlikely that the
same cell transitions from one nuclear morphology to
the other. Mutant spermatocytes never proceeded further
into, or completed, meiosis I. Consequently, histological
examination also revealed that mutant testes contained
no postmeiotic cell types such as haploid spermatids or
mature sperm. Instead, mutant testes degenerated with
age (Figures 2G and 2H).
Meiotic Defects in Miwi2 Mutants
To examine the apparent meiotic defect more closely, we
tracked the progress of synapsis by using spermatocyte
spreads (Figure 3). When spreads were prepared from
mutant testes, the vast majority of spermatocytes
(>95%) were in the leptotene stage (Figure 3D). At this
stage, Scp3, a component of the axial element of the
synaptonemal complex, becomes associated with the
two sister chromatids of each homolog (Lammers et al.,
1994; Moens et al., 1987). Only a few percent of mutant
spermatocytes reached zygotene, when longer paired
and unpaired axial elements are observed (Figures 3E
and 3F). Normal pachytene spermatocytes (e.g., Fig-
ure 3C) with fully condensed, paired chromosomes were
never observed in mutant animals.
Phosphorylated histone H2AX (g-H2AX) marks the sites
of Spo11-induced DNA double-strand breaks that oc-
cur during leptotene (Celeste et al., 2002; Fernandez-
Capetillo et al., 2003; Hamer et al., 2003; Mahadevaiah
et al., 2001). In wild-type cells, double-strand breaks
were repaired normally, and most of the g-H2AX signal
disappeared as cells entered pachytene (Figure 4A, P =
pachytene). In Miwi2 mutant spermatocytes, g-H2AX
staining appeared normal during the leptotene stage
(compare Figures 4A and 4B, L = leptotene). However,
concomitant with the change in morphology to highly
condensed nuclei, mutant spermatocytes appeared to
spermatocyte) as compared to wild-type zygotene cells
(Figure 4C, Z = zytogene). The persistence and strength
of the g-H2AX staining may indicate the presence of
unrepaired double-strand breaks and/or widespread
Figure 1. Generation of a Mutant Miwi2 Allele
(A) Theinsertionaldisruptionstrategy forMiwi2isshown schematically.Theinsertionduplicates exons9–12.Approximately10kbofvector sequence
is also inserted into the gene.
The targeted allele gives two signals, both distinct from wild-type, because the probe is within the duplicated region.
(C) RT-PCR was used to amplify wild-type Miwi2 transcripts in testes of 14-day-old animals.
(D) Gross appearance of wild-type and mutant testes at 6 weeks of age.
Developmental Cell 12, 503–514, April 2007 ª2007 Elsevier Inc. 505
MIWI2 Is Essential for Spermatogenesis
asynapsis, as the cells failed to progress successfully to
pachytene. Similar patterns have been observed previ-
ously, as mutants defective in synapsis or double-strand
break repair fail to eliminate g-H2AX from bulk chromatin
During male meiotic prophase, the incorporation of the
X and Y chromosomes into the sex or XY body correlates
with their transcriptional silencing. By pachytene stage,
a second wave of g-H2AX accumulates in the sex body
in association with the unsynapsed axial cores of the sex
chromosomes (de Vries et al., 2005; Turner et al., 2005)
(Figures 4A and 4E). When using standard histological
staining, the ‘‘exploded’’ nuclei in Miwi2 mutants often
contained structures that look remarkably like sex bodies
(Solari, 1974) (Figure 4F); however, these fail to stain with
g-H2AX despite its appearance on the scattered chroma-
tin (Figure 4F). At this time, it is unknown whether these
structures contain the sex chromosomes or whether other
proteins known to populate the sex body are present. This
structure may also be a nuclear organelle, such as the nu-
cleolus, that is not normally as prominent at this stage.
Nevertheless, we consistently fail to observe a g-H2AX fo-
formed sex body.
As Miwi2 mutant animals aged, they exhibited dramati-
cally increased levels of apoptosis in the seminiferous tu-
bules as compared to wild-type (Figure 5). A fluorescent
TUNEL assay revealed that, while a section through
a wild-type testis (Figure 5A) showed few or no apoptotic
cells (green), a large fraction of tubules in the mutant had
manydying cells (Figure 5B). Thesedevelopmental abnor-
malities arose during prophase of meiosis I. Although oc-
casional TUNEL-positive spermatocytes were present in
many tubule sections, larger groups of apoptotic sper-
matocytes were found in epithelial stage IV, characterized
by the presence of mitotic intermediate spermatogonia
and early B spermatogonia (Figure 5C). The apoptosis of
spermatocytes in stage IV resulted in the absence of sper-
matocytes in later stages, except for a few that entered
apoptosis a little more slowly and disappeared in stages
V–VII. While the apoptosis of virtually all spermatocytes
in stage IV has been observed in many mutants defective
Figure 2. Histological Analysis Reveals
Abnormalities in Miwi2 Mutants
(A–H) Hematoxylin and eosin staining of (A, C,
E, and G) wild-type and (B, D, F, and H) mutant
testes at indicated ages. (A)–(F) depict the first
wave of spermatogenesis at (A and B) 10 days
post partum (dpp) and (C–F) 15 dpp. Adult
testes from 3-month-old animals are depicted
in (G) (wild-type) and (H) (mutant). The arrow
and arrowhead indicate two types of abnormal
mutant spermatocytes (see text).
506 Developmental Cell 12, 503–514, April 2007 ª2007 Elsevier Inc.
MIWI2 Is Essential for Spermatogenesis
in meiotic genes (Barchi et al., 2005; de Rooij and de Boer,
2003), the Miwi2 mutation elicits a unique spermatocyte
behavior, as they either condense or enlarge long before
they reach epithelial stage IV and apoptose (Figure 5D).
In light of these results, we concluded that the seemingly
more intense g-H2AX staining of mutant spermatocytes
(Figure 4D) was not due to the creation of double-strand
breaks upon induction of apoptosis, as the observed
tubules had not yet reached stage IV.
Miwi2 Mutants Deplete Germ Cell Lineages
As mutant animals aged, their seminiferous tubules be-
came increasingly vacuolar (Figure 2H). Staining with
germ cell nuclear antigen (GCNA), which is expressed in
all germ cells, indicated that Miwi2 mutants exhibited
a marked decrease in the number of germ cells with age
(Figures 6A–6D). Before the onset of meiosis, the number
of germ cells was indistinguishable from that in wild-type
(Figures 2A and 2B). However, with age, mutant tubules
contained fewer spermatogonia and abnormal spermato-
cytes. Tubules lacking germ cells and containing only Ser-
toli cells began appearing as early as 3 months of age. As
the animals aged, Sertoli-cell-only tubules increased in
number and became predominant. The Sertoli cells that
populate these germ cell-less tubules appeared histolog-
Spermatogenic failure and germ cell loss can result
from defects in germ cells or in their somatic environment
(Brinster, 2002). In addition to being expressed in premei-
otic germ cells, Miwi2 is expressed at significant levels in
c-kit mutant testes (W/Wv) that are virtually germ cell free
line (Figure S1; M. Griswold, personal communication).
Thus, we sought to determine whether the defects ob-
served in Miwi2 mutant testes reflect a cell-autonomous
defect in the germ cells themselves or whether MIWI2
plays a critical role in somatic support cells.
To address this question, we transplanted wild-type
germ cells into Miwi2 mutant testes to assess the integrity
of the mutant soma. Recipient animals reconstituted
complete spermatogenesis in a subset of tubules, with
successful completion of both meiotic divisions and pro-
duction of mature sperm (Figure 6E, lower). These sper-
matogenic tubules existed side by side with noncolonized
tubules that displayed the characteristic Miwi2 mutant
phenotype (Figure 6E, upper). Although our conclusions
must be tempered by the remote possibility that the mu-
tant soma could harbor a level of Miwi2 that escapes de-
tection by RT-PCR, these studies strongly suggest that
Miwi2 mutant soma can successfully support germ cells
and lead to the conclusion that wild-type levels of Miwi2
expression in the germ cells themselves is necessary
and sufficient to support meiosis and spermiogenesis.
A Role for MIWI2 in Transposon Control
Two lines of circumstantial evidence point to a potential
role for mammalian Piwi proteins in transposon control.
First, in Drosophila, Piwi proteins have a demonstrated
role in the control of transposons (Aravin et al., 2001,
2004; Kalmykova et al., 2005; Saito et al., 2006; Sarot
Figure 3. Mutant Spermatocytes Arrest before the Pachytene Stage of Meiosis I
(A–E) Spermatocyte spreads from (A–C) heterozygous and (D and E) mutant animals in leptotene, zygotene, and pachytene stages, as indicated.
(F) Percentage of nuclei in each stage.
Developmental Cell 12, 503–514, April 2007 ª2007 Elsevier Inc. 507
MIWI2 Is Essential for Spermatogenesis
et al., 2004; Savitsky et al., 2006; Vagin et al., 2004, 2006).
onic defects that result in female sterility through a phe-
nomenon called hybrid dysgenesis. This is characterized
by a depletion of germline stem cells, abnormal oogene-
sis, and defects in oocyte organization. Second, a link
between the inappropriate expression of certain repetitive
elements and meiotic arrest has previously been demon-
strated in mammals. In particular, animals bearing muta-
tions in a catalytically defective member of the DNA
Figure 5. Mutant Testes Exhibit High
Levels of Apoptotic Cell Death
(A and B) TUNEL assay on 3-month-old (A)
wild-type and (B) mutant animals. Apoptotic
cells are labeled with FITC (green), and DAPI
stained nuclei are blue.
(C and D) Colorimetric TUNEL assay with pos-
itive nuclei staining brown. (C) 3 month old an-
imal with apoptosis in epithelial stage IV, and
(D) a 15 dpp tubule that has not yet reached
Figure 4. Mutant Spermatocytes Show
Abnormalities in a Marker for Double-
(A–F) Phosphorylated histone H2AX (g-H2AX)
staining (brown) of (A, C, and E) wild-type and
(B, D, and F) mutant seminiferous tubules of
14-day-old animals. Tubules containing (A
and B) leptotene, (C) zygotene, and (A and E)
pachytene cells are shown. Leptotene sper-
matocytes are indicated by L, zygotene by Z,
pachytene by P, and abnormal spermatocytes
by AS. The sex body in the wild-type pachy-
tene spermatocyte is indicated with an arrow
in (E). A similar structure in the mutant that fails
to stain is indicated with an arrow in (F).
508 Developmental Cell 12, 503–514, April 2007 ª2007 Elsevier Inc.
MIWI2 Is Essential for Spermatogenesis
posons in the male germline, resulting in abnormal and
abundant expression from several transposon families
(Bourc’his and Bestor, 2004; Hata et al., 2006; Webster
et al., 2005). This phenomenon is correlated with a meiotic
arrest prior to pachytene as well as germ cell loss. We
therefore considered that the germ cell loss and prevalent
apoptosis that we observe in Miwi2 mutants might corre-
late with transposon activation.
To investigate whether Miwi2 mutation affected expres-
sion from normally silent transposons, we used in situ hy-
bridization. When using this method, long interspersed el-
(Figure 7A). However, in Miwi2 mutants, a strong signal
can be seen with probes that detect sense-oriented
LINE-1 transcripts (Figure 7B). Similar approaches were
also used to monitor expression of intracisternal A particle
(IAP) elements that belong to the most active class of LTR
retrotransposons in the mouse. Sense strand IAP tran-
scripts were undetectable by in situ hybridization in wild-
type animals, while they were readily detectible in Miwi2
were detected exclusively in germ lineages, with no ap-
parent activation in Sertoli or interstitial cells of the testes.
Results from in situ analyses were supported and
extended by quantitative RT-PCR (Figure 7E). A 7- to
12-fold increase in LINE-1 expression was detected in
the mutants relative to heterozygous animals when
primers directed to the 50UTR and ORF2 were used. Sim-
ilar results were obtained with strand-specific RT-PCR
measuring only sense-orientation LINE-1 transcripts (not
shown). IAP elements were activated more modestly. Ele-
vated expression of these elements was detected only in
the testes, and not in the kidneys, of mutant animals
Figure 6. Miwi2 Mutant Testes Progres-
sively Lose Germ Cells
nuclear antigen (GCNA) (brown) in (A) 15-day-,
(B) 3-month-, (C) 6-month-, and (D) 20-month-
old animals. Sertoli-cell-only tubules are indi-
cated with an asterisk.
(E) A Miwi2 mutant testis that has been trans-
planted with wild-type germ cells is shown. A
noncolonized tubule displaying the Miwi2 mu-
tant phenotype is at the top. Abnormal mutant
spermatocytes are indicated by AS. The lower
tubule has been colonized by wild-type germ
cells. Normal spermatocytes and spermatids
are indicated. Zygotene nuclei are indicated
by Z, haploid round spermatids are indicated
by RS, and elongating spermatids are indi-
cated by ES.
Developmental Cell 12, 503–514, April 2007 ª2007 Elsevier Inc. 509
MIWI2 Is Essential for Spermatogenesis
(data not shown). To ensure that the observed effects
were not a secondary consequence of meiotic arrest, we
analyzed testes from meiosis defective-1 (Mei1) mutant
animals, which display a meiotic arrest phenotype similar
to Miwi2 mutants, and failed to observe increased trans-
poson expression (Figure 7E).
Transposable elements are thought to be maintained in
a silent state by DNA methylation and packaging into het-
erochromatin. We investigated the methyation status of
LINE-1 in the Miwi2 mutants by Southern blot analysis af-
ter digestion with a methylation-sensitive enzyme, HpaII,
and found that LINE-1 elements become demethylated
in Miwi2 mutants as compared to wild-type and heterozy-
gous animals (Figure 7F). Demethylation was detected
specifically in DNA prepared from the testes and not
from the tail. Thus, compromising Miwi2 can affect the
methylation of repetitive elements specifically inthe germ-
line. For comparison, we assayed LINE-1 methylation in
testes from several mutants that show a meiotic arrest
similarto Miwi2 mutants(FigureS2).None of thesemutant
animals show LINE-1 demethylation.
We used bisulfite sequencing to examine methylation of
the first 150 bp of the 50UTR of a specific copy of L1Md-
A2. In heterozygous animals, this region is almost com-
pletely methylated, with 95% of all CpGs modified
(Figure 7G). In the mutant, only 60% of CpGs are methyl-
ated overall, with two distinct populations of PCR prod-
ucts being apparent. These are represented at the
extremes by 34% of the clones that are completely unme-
thylated, and 46% that retain full methylation (Figure 7G;
Figure 7. Miwi2 Mutants Derepress and
Demethylate Transposable Elements
(A–D) In situ hybridization of testes of the indi-
cated genotypes with probes recognizing the
sense strands of (A and B) LINE-1 and (C and
D) IAP elements.
(E) Quantitative RT-PCR analysis of transpos-
able elements in 14-day-old animals. Data are
represented as mean ± SEM.
(F) DNA isolated from the tail or testes of wild-
type, heterozygous, and Miwi2 mutant animals
was digested with either methylation-insensi-
tive (MspI, M) or methylation-sensitive (HpaII,
H) restriction enzymes. Southern blot analysis
of these DNAs was conducted, and mem-
branes were probed with a fragment of the
LINE-1 50UTR. Asterisks indicate the expected
band sizes. The probe recognizes four bands
of 156 bp generated by HpaII sites in the
50UTR,and aband of1206bpthatisgenerated
by one HpaII site in the 50UTR and one site in
the coding sequence.
(G) Lollipop representation of the sequences
obtained after bisulfite treatment of Miwi2
+/? and ?/? testis DNA. The first 150 bp of
a specific L1 element were selectively ampli-
fied and analyzed for the presence of methyl-
ated CpGs. Methylated and unmethylated
CpGs are represented as filled and empty lolli-
pops, respectively. Out of 75 sequences ob-
tained for each genotype, 20 randomly chosen
plete set can be found in Figure S3.
510 Developmental Cell 12, 503–514, April 2007 ª2007 Elsevier Inc.
MIWI2 Is Essential for Spermatogenesis
Figure S3). Based on our Southern blot and quantitative
RT-PCR analyses that show normal methylation and
transposon repression in somatic tissues, we suggest
that these two populations are likely derived from germ
cells (unmethylated) and somatic cells (methylated).
Successful expansion by selfish genetic elements can
only occur if increased copy numbers can be transmitted
to the next generation. Consistent with this notion, LINE
and IAP elements are known to be active almost ex-
clusively in the germline (Branciforte and Martin, 1994;
LINE-1 transcripts, and the ORF1 protein that they en-
matocytes in pubertal mouse testes (Branciforte and Mar-
tin, 1994). In the adult male, truncated transcripts and
ORF1 protein are present in somatic cells and haploid
germ cells (Branciforte and Martin, 1994; Trelogan and
Martin, 1995). ORF1 protein is also present in oocytes
and steroidogenic cells in the female germline (Branciforte
the deleterious and cumulative effects of unregulated
evolutionary pressure to evolve effective transposon con-
trol strategies in the germline. Our data indicate thatmam-
malian Piwi proteins form at least part of such a defense
In Drosophila, Piwi proteins are reported to have both
cell autonomous and nonautonomous roles in maintaining
the integrity of the germline (Cox et al., 2000). In particular,
integrity of the germline stem cell niche. In mammals,Miwi
and Mili mutants arrest spermatogenesis at different
stages, but neither is reported to lose germ cells, as might
be expected if, like PIWI, either protein had a role in stem
cell maintenance. Here, we show that disruption of Miwi2
creates two distinct phenotypes in the male germline of
mice. First, Miwi2 mutant germ cells that enter prophase
of meiosis I arrest prior to the pachytene stage. Second,
Miwi2 mutants progressively lose germ cells and accumu-
late tubules that contain only somatic Sertoli cells. The lat-
ter observation suggests that MIWI2 may conserve some
of the stem cell maintenance functions played by PIWI in
for Piwi proteins in stem cell maintenance in flies is due
to their role in regulating gene expression, or whether
the phenotypes of Piwi-family mutations can be solely
explained by loss of transposon control.
proteins play a prominent and essential role in transposon
control (Aravin et al., 2001, 2004; Kalmykova et al., 2005;
Sarot et al., 2004; Savitsky et al., 2006; Vagin et al., 2004).
One consequence of disrupting transposon suppression
in flies is the appearance of DNA damage, as evidenced
by the accumulation of phosphorylated histone H2AX
(Belgnaoui et al., 2006; Gasior et al., 2006) A key role for
DNA-damage pathways in the ultimate output of Piwi-
family mutations, production of defective oocytes, is indi-
cated by the fact that mutation of key DNA-damage sens-
ing pathways can at least partially suppress the effects of
transposon activation (Klattenhoff etal., 2007). Our results
As in flies, Miwi2 mutations also result in accumulation of
DNA damage, as indicated by g-H2AX accumulation. The
relationship between the molecular phenotypes of Piwi-
family mutations in flies and mice, particularly whether ac-
tivation of DNA-damage response pathways plays a role
in the meiotic defects observed in Miwi2 mutants,remains
to be determined.
Drosophila Piwi proteins interact with small RNAs of
?24–26 nucleotides in length (Aravin et al., 2001; Saito
et al., 2006; Vagin et al., 2006). These are highly enriched
for sequences that target repetitive elements and are
therefore called rasiRNAs (repeat-associated siRNAs)
(Aravin et al., 2003; Saito et al., 2006). In contrast, mam-
malian Piwi-family proteins, MIWI and MILI, bind to an
?26–30 nucleotide class of small RNAs known as piRNAs
(Piwi-interacting RNAs) (Aravin et al., 2006; Girard et al.,
2006; Grivna et al., 2006; Lau et al., 2006; Watanabe
et al., 2006). A large proportion of piRNAs are only compli-
mentary to the loci from which they came, leading to the
getsof MILI andMIWI RNPs. Resultspresented herepoint
to a role for piRNAs in transposon control in mammals
Unexpectedly, we have found that the rasiRNA system
in flies shows many characteristics in common with the
piRNA system in mammals (Brennecke et al., 2007).
Piwi-interacting RNAs in Drosophila are derived from dis-
crete genomic loci. At least some of these loci show the
profound strand asymmetry that characterizes mamma-
lian piRNA loci. These observations begin to unify Piwi
protein functions in disparate organisms. However, future
work will be required to understand how the meiotic
piRNA loci, which are depleted of repeats, relate function-
ally to the piRNA loci in flies that act as master controllers
of transposon activity.
Silencing of mammalian transposons depends on their
methylation status (Bourc’his and Bestor, 2004). Ge-
nomes of primordial germ cells undergo demethylation
followed by de novo remethylation in prospermatogonia,
a nondividing cell type that exists only in the perinatal pe-
riod. How the patterns of methylation are determined in
developing germ cells is not understood. In Arabidopsis,
it is well established that the RNAi machinery can use
small RNAs to direct genomic methylation, though the
precise biochemical mechanism underlying these events
remains unclear (Matzke and Birchler, 2005). In plants,
ARGONAUTE4, a member of the Argonaute rather than
the Piwi subfamily, binds to 24 nt, small RNAs and mainly
directs asymmetric cytosine methylation (CpNpG and
CpHpH). However, such asymmetric methylation is rare
or absent in mammalian genomes. Here, we provide
Developmental Cell 12, 503–514, April 2007 ª2007 Elsevier Inc. 511
MIWI2 Is Essential for Spermatogenesis
evidence that loss of MIWI2 function affects the methyla-
tion status of LINE-1 elements. MIWI2 complexes, which
we presume are directed to their targets by associated
piRNAs, might help to establish genomic methylation pat-
terns on repetitive elements during germ cell develop-
ment. It is also possible that removal of MIWI2 interferes
with the maintenance of genomic methylation patterns
that normally occurs in dividing spermatagonia. A detailed
analysis of patterns of Miwi2 expression and identification
of piRNAs that interact with MIWI2 during germ cell devel-
opment will be needed to distinguish roles for this protein
complex in de novo versus maintenance methylation.
Gene Targeting and Mice
The Miwi2 targeting construct was obtained by screening of the
lambda phage 30HPRT library described by Zheng et al. (1999) that
is now the basis of the MICER system (Adams et al., 2004). The resul-
tant targeting construct, containing exons 9–12 of Miwi2, was electro-
porated into AB2.2 mouse embryonic stem (ES) cells. Targeted clones
age chimeras, four of which were able to pass the allele through the
germline. Results presented herein were obtained from mice with
a mixed 129/B6 background. In general, younger animals were back-
crossed to B6 4–6 generations, and older animals were backcrossed
less. Mouse genotyping was performed by Southern blot analysisafter
digestion of genomic DNA with AccI. The 332 bp probe was amplified
from genomic DNA with primers described in Table S1.
Testes were collected and fixed in Bouin’s fixative at 4?C overnight,
then dehydrated to 70% ethanol. After embedding in paraffin, 8 mm
sections were made by using a microtome. For routine histology, sec-
tions were stained with hematoxylin and eosin. For routine histology
and subsequent staining, at least three animals of each age and geno-
type were examined.
Slides were rehydrated and treated with 3% hydrogen peroxide for 10
min.Blocking wascarriedoutin5%goat serum,1%BSAinPBS for10
min. Slides were incubated overnight at 4?C with primary antibody as
follows. Antibody to g-H2AX (Upstate) was used at 1:150 in 1% BSA in
PBS. GCNA (a gift of G. Enders) was used neat. Detection was per-
formed by using the Vector ABC kit according to the manufacturer’s
of PBS. Slides were counterstained with Mayer hematoxylin, mounted
with Histomount mounting media, and coverslipped.
For immunocytological analysis of synaptonemal complex forma-
tion, surface spreading of spermatocytes was performed as described
by Matsuda et al. (1992). Spreads were hybridized with goat anti-Scp3
(gift of T. Ashley) at 1:400 dilution. Approximately 200 nuclei from each
of three animals were counted, for a total of 600 nuclei of each geno-
type. Spreads were conducted on animals at 16 dpp.
Slides containing Bouin’s-fixed testes sections were rehydrated and
microwaved for 5 min in 10 mM Citrate buffer (pH 6.0). After incubation
in 3% hydrogen peroxide, slides were incubated with 0.3 U/microliter
deoxynucleotidal terminal transferase(Amersham)and 6.66 mM biotin-
16-dUTP (Roche) for 1 hr at 37?C. After washing in 300 mM NaCl, 30
mM NaCitrate in MilliQ water for 15 min at room temperature, slides
were blocked in 2% BSA in PBS for 10 min. Slides were incubated in
a 1:20 dilution of ExtrAvidine peroxidase (Sigma) in 1% BSA in PBS
for 30 min at 37?C. Detection was achieved by using diaminobenzi-
dine. Slides were counterstained withMayer hematoxylin, dehydrated,
and mounted. Fluorescent TUNEL assay was conducted by using the
Roche In Situ Cell Death Detection kit according to the manufacturer’s
Germ Cell Transplants
Transplants were carried out as described by Buaas et al. (2004). Do-
nor cells wereharvested from the transgenic mouseline C57BL/6.129-
TgR(Rosa26)26S (Jackson Laboratory). Donor cells were transplanted
depleted due to the mutation, or into W/Wvmice that have no endog-
enous spermatogenesis asacontrol (JacksonLaboratory,WBB6F1/J-
KitW/KitWv). Recipient testes were analyzed with standard histological
methods to identify areas of colonization by donor cells. One out of 10
Miwi2 mutant recipients and 2 out of 5 W/Wvwere successfully colo-
RT-PCR and QPCR
TotalRNA wasextracted from mouse tissuesbyusingTrizol according
to the manufacturer’s recommendations. cDNA was synthesized by
with random hexamers. QPCR was carried out by using Sybr Green
PCR Master Mix (Applied Biosystems) on a Biorad Chromo 4 Real
Time system. Two animals of each genotype were examined, with
the exception of Mei1, for which we had only one specimen. Assays
were done in triplicate. Miwi2 animals were 14 days old, and Mei1 an-
imals were 21 days old. Primers Miwi2-F and Miwi2-R are downstream
of the duplicated exons and cannot distinguish between wild-type and
mutant transcript. Primers Miwi2-exon7F and Miwi2-exon14R flank
the duplicated exons in the mutant transcript and therefore assay for
only the wild-type transcript. The wild-type transcript produces
a band of 1006 bp, while the mutant would yield a larger product
due to the duplication of exons 9–12. Primers are listed in Table S1.
In Situ Hybridization
In situ hybridization was done as described by Bourc’his and Bestor
(2004). The 50LTR IAP probe was as described by Walsh et al.
(1998), and the LINE-1 50UTR probe is complementary to a type A
LINE-1 element (GenBank accession number: M13002, nucleotides
515–1,628) (Bourc’his and Bestor, 2004).
Methylation Southern Blot Analysis
Southern blot analysis to assay for methylation was done as described
by Bourc’his and Bestor (2004). The same LINE-1 50UTR probe was
used as for in situ hybridization, except a gel-purified fragment was
random prime labeled by using the Rediprime II kit (Amersham). DNA
from testis and tail were digested with the methylation-sensitive en-
zyme HpaII and its methylation-insensitive isoschizomer, MspI.
Bisulfite DNA Sequencing
DNA from Miwi2 +/? and ?/? testes was bisulfite treated and purified
by using the EZ DNA Methylation Gold kit (Zymo Research). Primers
MethylL1-F and MethylL1-R were designed to specifically amplify
one occurrence of L1Md-A2 located on chromosome X. The PCR
and analyzed by using BiQ-Analyzer (Bock et al., 2005). Primers and
the sequence of the amplified region are given in Table S1.
methylation controls, the entire bisulfite DNA-sequencing data set, and
primer sequences and are available at http://www.developmentalcell.
We thank Takeshi Soda and Pedro Aponte (Utrecht University) for in-
struction in germ cell transplantation. We also thank Terry Ashley
and Adelle Hack (Yale) for instruction on spermatocyte spreads,
512 Developmental Cell 12, 503–514, April 2007 ª2007 Elsevier Inc.
MIWI2 Is Essential for Spermatogenesis
antibodies, and helpful discussions. We thank Peter de Boer (Univer-
Harbor Laboratory [CSHL]) for microscopy assistance and helpful dis-
cussions, and George Enders (University of Kansas) for GCNA anti-
body. We thank Ignasi Roig and Scott Keeney (Memorial Sloan Ketter-
ingCancerCenter)fortissuesfrom Atmand Dmc1mice,and XinLiand
John Schimenti (Cornell) for tissue from Mei1 mice. We thank Lisa
Bianco, Jodi Coblentz, Lea Manzella, and Gula Nourjanova (CSHL)
for animal assistance and histology. We thank Mike Griswold and Liz-
hong Yang (Washington State University) for access to unpublished
data. We thank members of the Hannon laboratory for comments on
the manuscript. A.G. is a Florence Gould Fellow of the Watson School
of Biological Sciences. G.J.H. is an investigator of the Howard Hughes
Medical Institute. This workwas supported bygrants from theNational
Institutes of Health (G.J.H).
Received: November 8, 2006
Revised: February 28, 2007
Accepted: March 1, 2007
Published online: March 28, 2007
Adams, D.J., Biggs, P.J., Cox, T., Davies, R., van der Weyden, L.,
Jonkers, J., Smith, J., Plumb, B., Taylor, R., Nishijima, I., et al.
(2004). Mutagenic insertion and chromosome engineering resource
(MICER). Nat. Genet. 36, 867–871.
Aravin, A., Gaidatzis, D., Pfeffer, S., Lagos-Quintana, M., Landgraf, P.,
Iovino, N., Morris, P., Brownstein, M.J., Kuramochi-Miyagawa, S.,
Nakano, T., et al. (2006). A novel class of small RNAs bind to MILI pro-
tein in mouse testes. Nature 442, 203–207.
Aravin, A.A., Naumova, N.M., Tulin, A.V., Vagin, V.V., Rozovsky, Y.M.,
and Gvozdev, V.A. (2001). Double-stranded RNA-mediated silencing
of genomic tandem repeats and transposable elements in the
D. melanogaster germline. Curr. Biol. 11, 1017–1027.
Aravin, A.A., Lagos-Quintana, M., Yalcin, A., Zavolan, M., Marks, D.,
Snyder, B., Gaasterland, T., Meyer, J., and Tuschl, T. (2003). The small
RNA profile during Drosophila melanogaster development. Dev. Cell 5,
Aravin, A.A., Klenov, M.S., Vagin, V.V., Bantignies, F., Cavalli, G., and
Gvozdev, V.A. (2004). Dissection of a natural RNA silencing process in
the Drosophila melanogaster germ line. Mol. Cell. Biol. 24, 6742–6750.
Barchi, M., Mahadevaiah, S., Di Giacomo, M., Baudat, F., de Rooij,
D.G., Burgoyne, P.S., Jasin, M., and Keeney, S. (2005). Surveillance
of different recombination defects in mouse spermatocytes yields dis-
tinct responses despite elimination at an identical developmental
stage. Mol. Cell. Biol. 25, 7203–7215.
Belgnaoui, S.M., Gosden, R.G., Semmes, O.J., and Haoudi, A. (2006).
Human LINE-1 retrotransposon induces DNA damage and apoptosis
in cancer cells. Cancer Cell Int. 6, 13.
Bock, C., Reither, S., Mikeska, T., Paulsen, M., Walter, J., and Lenga-
uer, T. (2005). BiQ Analyzer: visualization and quality control for DNA
methylation data from bisulfite sequencing. Bioinformatics 21, 4067–
Bourc’his, D., and Bestor, T.H. (2004). Meiotic catastrophe and retro-
transposon reactivation in male germ cells lacking Dnmt3L. Nature
Branciforte, D., and Martin, S.L. (1994). Developmental and cell type
specificity of LINE-1 expression in mouse testis: implications for trans-
position. Mol. Cell. Biol. 14, 2584–2592.
Brennecke, J., Aravin, A.A., Stark, A., Dus, M., Kellis, M., Sachidanan-
dam, R., and Hannon, G.J. (2007). Discrete small RNA-generating loci
Published online March 8, 2007. 10.1016/j.cell.2007.01.043.
Brinster, R.L. (2002). Germline stem cell transplantation and transgen-
esis. Science 296, 2174–2176.
Buaas, F.W., Kirsh, A.L., Sharma, M., McLean, D.J., Morris, J.L., Gris-
wold, M.D., de Rooij, D.G., and Braun, R.E. (2004). Plzf is required in
adult male germ cells for stem cell self-renewal. Nat. Genet. 36, 647–
Celeste, A., Petersen, S., Romanienko, P.J., Fernandez-Capetillo, O.,
Chen, H.T., Sedelnikova, O.A., Reina-San-Martin, B., Coppola, V.,
Meffre, E., Difilippantonio, M.J., et al. (2002). Genomic instability in
mice lacking histone H2AX. Science 296, 922–927.
Cox, D.N., Chao, A., Baker, J., Chang, L., Qiao, D., and Lin, H. (1998).
A novel class of evolutionarily conserved genes defined by piwi are
essential for stem cell self-renewal. Genes Dev. 12, 3715–3727.
Cox, D.N., Chao, A., and Lin, H. (2000). piwi encodes a nucleoplasmic
factor whose activity modulates the number and division rate of germ-
line stem cells. Development 127, 503–514.
de Rooij, D.G., and de Boer, P. (2003). Specific arrests of spermato-
genesis in genetically modified and mutant mice. Cytogenet. Genome
Res. 103, 267–276.
de Rooij, D.G., and Grootegoed, J.A. (1998). Spermatogonial stem
cells. Curr. Opin. Cell Biol. 10, 694–701.
deVries, F.A., deBoer, E.,vanden Bosch, M.,Baarends,W.M., Ooms,
M., Yuan, L., Liu, J.G., van Zeeland, A.A., Heyting, C., and Pastink, A.
(2005). Mouse Sycp1 functions in synaptonemal complex assembly,
meiotic recombination, and XY body formation. Genes Dev. 19,
Deng, W., and Lin, H. (2002). miwi, a murine homolog of piwi, encodes
a cytoplasmic protein essential for spermatogenesis. Dev. Cell 2, 819–
Dupressoir, A., and Heidmann, T. (1996). Germ line-specific expres-
sion of intracisternal A-particle retrotransposons in transgenic mice.
Mol. Cell. Biol. 16, 4495–4503.
Fernandez-Capetillo, O., Mahadevaiah, S.K., Celeste, A., Roma-
nienko, P.J., Camerini-Otero, R.D., Bonner, W.M., Manova, K., Bur-
goyne, P., and Nussenzweig, A. (2003). H2AX is required for chromatin
remodeling and inactivation of sex chromosomes in male mouse
meiosis. Dev. Cell 4, 497–508.
Gasior, S.L., Wakeman, T.P., Xu, B., and Deininger, P.L. (2006). The
human LINE-1 retrotransposon creates DNA double-strand breaks.
J. Mol. Biol. 357, 1383–1393.
Girard, A., Sachidanandam, R., Hannon, G.J., and Carmell, M.A.
(2006). A germline-specific class of small RNAs binds mammalian
Piwi proteins. Nature 442, 199–202.
Grivna, S.T., Beyret, E., Wang, Z., and Lin, H. (2006). A novel class of
Hamer, G., Roepers-Gajadien, H.L., van Duyn-Goedhart, A., Gade-
man, I.S., Kal, H.B., van Buul, P.P., and de Rooij, D.G. (2003). DNA
double-strand breaks and g-H2AX signaling in the testis. Biol. Reprod.
Hata, K., Kusumi, M., Yokomine, T., Li, E., and Sasaki, H. (2006). Mei-
otic and epigenetic aberrations in Dnmt3L-deficient male germ cells.
Mol. Reprod. Dev. 73, 116–122.
Kalmykova, A.I., Klenov, M.S., and Gvozdev, V.A. (2005). Argonaute
protein PIWI controls mobilization of retrotransposons in the Drosoph-
ila male germline. Nucleic Acids Res. 33, 2052–2059.
Klattenhoff, C., Bratu, D.P., McGinnis-Schultz, N., Koppetsch, B.S.,
Cook, H.A., and Theurkauf, W.E. (2007). Drosophila rasiRNA pathway
mutations disrupt embryonic axis specification through activation of
an ATR/Chk2 DNA damage response. Dev. Cell 12, 45–55.
Kuramochi-Miyagawa, S., Kimura, T., Yomogida, K., Kuroiwa, A., Ta-
dokoro, Y., Fujita, Y., Sato, M., Matsuda, Y., and Nakano, T. (2001).
Two mouse piwi-related genes: miwi and mili. Mech. Dev. 108, 121–
Kuramochi-Miyagawa, S., Kimura, T., Ijiri, T.W., Isobe, T., Asada, N.,
Fujita, Y., Ikawa, M., Iwai, N., Okabe, M., Deng, W., et al. (2004). Mili,
Developmental Cell 12, 503–514, April 2007 ª2007 Elsevier Inc. 513
MIWI2 Is Essential for Spermatogenesis
a mammalian member of piwi family gene, is essential for spermato- Download full-text
genesis. Development 131, 839–849.
Lammers, J.H., Offenberg,H.H., van Aalderen, M., Vink, A.C., Dietrich,
A.J., and Heyting, C. (1994). The gene encoding a major component of
the lateral elements of synaptonemal complexes of the rat is related to
X-linked lymphocyte-regulated genes. Mol. Cell. Biol. 14, 1137–1146.
Lau, N.C., Seto, A.G., Kim, J., Kuramochi-Miyagawa, S., Nakano, T.,
Bartel, D.P., and Kingston, R.E. (2006). Characterization of the piRNA
complex from rat testes. Science 313, 363–367.
Lin, H., and Spradling, A.C. (1997). A novel group of pumilio mutations
affects theasymmetric divisionof germline stemcells intheDrosophila
ovary. Development 124, 2463–2476.
Mahadevaiah, S.K., Turner, J.M., Baudat, F., Rogakou, E.P., de Boer,
P., Blanco-Rodriguez, J., Jasin, M., Keeney, S., Bonner, W.M., and
Burgoyne, P.S. (2001). Recombinational DNA double-strand breaks
in mice precede synapsis. Nat. Genet. 27, 271–276.
Matsuda, Y., Moens, P.B., and Chapman, V.M. (1992). Deficiency of X
and Y chromosomal pairing at meiotic prophase in spermatocytes of
sterile interspecifichybrids between laboratory mice (Musdomesticus)
and Mus spretus. Chromosoma 101, 483–492.
Matzke, M.A., and Birchler, J.A. (2005). RNAi-mediated pathways in
the nucleus. Nat. Rev. Genet. 6, 24–35.
Moens, P.B., Heyting, C., Dietrich, A.J., van Raamsdonk, W., and
Chen, Q. (1987). Synaptonemal complex antigen location and conser-
vation. J. Cell Biol. 105, 93–103.
Saito, K., Nishida, K.M., Mori, T., Kawamura, Y., Miyoshi, K., Nagami,
T., Siomi, H., and Siomi, M.C. (2006). Specific association of Piwi with
rasiRNAs derived from retrotransposon and heterochromatic regions
in the Drosophila genome. Genes Dev. 20, 2214–2222.
Sarot,E., Payen-Groschene, G., Bucheton, A., and Pelisson, A. (2004).
Evidence for a piwi-dependent RNA silencing of the gypsy endoge-
nous retrovirus by the Drosophila melanogaster flamenco gene.
Genetics 166, 1313–1321.
Sasaki, T., Shiohama, A., Minoshima, S., and Shimizu, N. (2003). Iden-
tification of eight members of the Argonaute family in the human ge-
nome small star, filled. Genomics 82, 323–330.
Savitsky, M., Kwon, D., Georgiev, P., Kalmykova, A., and Gvozdev, V.
(2006). Telomere elongation is under the control of the RNAi-based
mechanism in the Drosophila germline. Genes Dev. 20, 345–354.
Schmidt, A., Palumbo, G., Bozzetti, M.P., Tritto, P., Pimpinelli, S., and
Schafer, U. (1999). Genetic and molecular characterization of sting,
a gene involved in crystal formation and meiotic drive in the male
germ line of Drosophila melanogaster. Genetics 151, 749–760.
Silvers, W.K. (1979). The Coat Colors of Mice (New York: Springer
Solari, A.J. (1974). The behavior of the XY pair in mammals. Int. Rev.
Cytol. 38, 273–317.
Trelogan, S.A., and Martin, S.L. (1995). Tightly regulated, developmen-
tally specific expression of the first open reading frame from LINE-1
during mouse embryogenesis. Proc. Natl. Acad. Sci. USA 92, 1520–
Turner, J.M., Mahadevaiah, S.K., Fernandez-Capetillo, O., Nussenz-
synapsed meiotic chromosomes in the mouse. Nat. Genet. 37, 41–47.
Vagin, V.V., Klenov, M.S., Kalmykova, A.I., Stolyarenko, A.D., Kotelni-
kov, R.N., and Gvozdev, V.A. (2004). The RNA interference proteins
and vasa locus are involved in the silencing of retrotransposons in
the female germline of Drosophila melanogaster. RNA Biol. 1, 54–58.
Vagin, V.V., Sigova, A., Li, C., Seitz, H., Gvozdev, V., and Zamore, P.D.
(2006). A distinct small RNA pathway silences selfish genetic elements
in the germline. Science 313, 320–324.
endogenous retroviruses is constrained by cytosine methylation. Nat.
Genet. 20, 116–117.
Wang, H., and Hoog, C. (2006). Structural damage to meiotic chromo-
somes impairsDNA recombination and checkpointcontrol inmamma-
lian oocytes. J. Cell Biol. 173, 485–495.
Watanabe, T., Takeda, A., Tsukiyama, T., Mise, K., Okuno, T., Sasaki,
H., Minami, N., and Imai, H. (2006). Identification and characterization
of two novel classes of small RNAs in the mouse germline: retrotrans-
poson-derived siRNAs in oocytes and germline small RNAs in testes.
Genes Dev. 20, 1732–1743.
Webster, K.E., O’Bryan, M.K., Fletcher, S., Crewther, P.E., Aapola, U.,
Craig, J., Harrison, D.K., Aung, H., Phutikanit, N., Lyle, R., et al. (2005).
Meiotic and epigenetic defects inDnmt3L-knockout mouse spermato-
genesis. Proc. Natl. Acad. Sci. USA 102, 4068–4073.
Xu, X., Aprelikova, O., Moens, P., Deng, C.X., and Furth, P.A. (2003).
Impaired meiotic DNA-damage repair and lack of crossing-over during
spermatogenesis in BRCA1 full-length isoform deficient mice. Devel-
opment 130, 2001–2012.
Zheng, B., Mills, A.A., and Bradley, A. (1999). A system for rapid gen-
eration of coat color-tagged knockouts and defined chromosomal
rearrangements in mice. Nucleic Acids Res. 27, 2354–2360.
514 Developmental Cell 12, 503–514, April 2007 ª2007 Elsevier Inc.
MIWI2 Is Essential for Spermatogenesis