Heterozygosity with respect to Zfp148 causes complete loss of fetal germ cells during mouse embryogenesis.

Abstract

Zfp148 belongs to a large family of C2H2-type zinc-finger transcription factors. Zfp148 is expressed in fetal germ cells in 13.5-d-old (E13.5) mouse embryos. Germ-line transmission of mutations were not observed in chimeric Zfp148(+/-) mice, and some of these mice completely lacked spermatogonia. The number of primordial germ cells in Zfp148(+/-) tetraploid embryos was normal until E11.5, but declined from E11.5 to E13.5 and continued to decline until few germ cells were present at E18.5. This phenotype was not rescued by wild-type Sertoli or stromal cells, and is therefore a cell-autonomous phenotype. These results indicate that two functional alleles of Zfp148 are required for the normal development of fetal germ cells. Recent studies have shown that Zfp148 activates p53, which has an important role in cell-cycle regulation. Primordial germ cells stop proliferating at approximately E13.5, which correlates with induction of phosphorylation of p53 and its translocation to the nucleus. Phosphorylation of p53 is impaired in Zfp148(+/-) embryonic stem cells and in fetal germ cells from chimeric Zfp148(+/-) embryos. Thus, Zfp148 may be required for regulating p53 in the development of germ cells.

Full text

Page 1

letter
172 nature genetics • volume 33 • february 2003
Heterozygosity with respect to Zfp148 causes complete
loss of fetal germ cells during mouse embryogenesis
Akihide Takeuchi1,3, Yuji Mishina2, Osamu Miyaishi1,3, Eiji Kojima1, Tadao Hasegawa1,3& Ken-ichi Isobe1
1Department of Basic Gerontology, National Institute for Longevity Sciences, 36-3 Gengo, Morioka-cho, Obu-city, Aichi 474-8522, Japan.2Molecular
Developmental Biology Group, Laboratory of Reproductive and Developmental Toxicology, National Institute of Environmental Health Sciences/National
Institutes of Health, Research Triangle Park, North Carolina, USA.3Present addresses: Neural Stem Cell Research, Stem Cells Inc., 3155 Porter Drive, Palo
Alto, California 94304, USA (A.T.); Department of Pathology, Aichi Medical University School of Medicine, Nagakute, Aichi, Japan (O.M.);Department of
Microbiology, Nagoya University School of Medicine, Showa-ku, Nagoya, Japan (T.H.). Correspondence should be addressed to K-I. I.
(e-mail: kenisobe@nils.go.jp).
Zfp148 belongs to a large family of C2H2-type zinc-finger tran-
scription factors. Zfp148 is expressed in fetal germ cells in 13.5-
d-old (E13.5) mouse embryos. Germ-line transmission of
mutations were not observed in chimeric Zfp148+/–mice, and
some of these mice completely lacked spermatogonia. The
number of primordial germ cells in Zfp148+/–tetraploid
embryos was normal until E11.5, but declined from E11.5 to
E13.5 and continued to decline until few germ cells were pre-
sent at E18.5. This phenotype was not rescued by wild-type
Sertoli or stromal cells, and is therefore a cell-autonomous phe-
notype. These results indicate that two functional alleles of
Zfp148 are required for the normal development of fetal germ
cells. Recent studies have shown that Zfp148 activates p53,
which has an important role in cell-cycle regulation1. Primordial
germ cells stop proliferating at approximately E13.5,
which correlates with induction of phosphorylation
of p53 and its translocation to the nucleus. Phospho-
rylation of p53 is impaired in Zfp148+/–embryonic
stem cells and in fetal germ cells from chimeric
Zfp148+/–embryos. Thus, Zfp148 may be required for
regulating p53 in the development of germ cells.
Zfp148 (also known as BFCOL-1, ZBP-89, BERF1 and ZNF148)
encodes a Krüppel-type transcription factor of 89 kD with four
tandem zinc-finger DNA-binding motifs2,3. Zfp148 is thought to
promote growth arrest by stabilizing p53. Zbp-89, a rat homolog
of Zfp148, enhances p53 transcriptional activity and prevents
degradation by binding to p53 through its zinc-finger domain1.
Overexpression of Zbp-89 leads to accumulation of p53 in the
nucleus, activation of p21waf1and cell-cycle arrest. Zfp148 also
interacts with growth-arrest DNA damage-34 (Gadd34; ref. 3)
and activates the Cdkn1a promoter4. These activities suggest that
Zfp148 has an important role in regulating cellular proliferation;
it may also be important during development.
Zfp148 is highly expressed in the neural tube (A.T., Y.M., O.M.
and K.-I.I., unpublished observations) and in male and female
Published online 13 January 2003; doi:10.1038/ng1072
Fig. 1 In situ hybridization of male and female gonads of E12.5
embryos and targeted disruption of Zfp148. a–f, In situ hybridiza-
tion of male (a–c) and female (d–f) gonads with antisense (a,c,d,f)
and sense (b,e) RNA probes specific to Zfp148. Positive signal (violet
granules) was detected in the genital ridge in male and female
gonads. The positive signal was detected in the tubular structure of
the male gonads (a) but in the peripheral area of the female gonads
(d). At higher magnification, the positive signal was detected pre-
dominantly in PGCs, large cells with large nuclei (c,f). No signal was
detected with the sense probe (b,e). Scale bar = 100 µm (a,b,d,e) or
50 µm (c,f). g, Construction of the Zfp148 targeting construct. Non-
coding region is shown as a black rectangle in exon 9. Unique 5′ and
3′ probes are shown as open boxes. B, BamHI; H, HindIII; K, KpnI; S,
SpeI; DT-A, diphtheria toxin A. h, Southern-blot analysis using indi-
cated probes. ES-cell DNA digested with BamHI produces bands of
5.0 kb (wild-type) or 4.0 kb (mutated allele) with the 5′ probe, and
ES-cell DNA digested with SpeI produces bands of 8.0 kb (wild-type)
or 6.5 kb (mutated allele) band with the 3′ probe. i, Northern-blot
analysis of the targeted and non-targeted ES clones. Lanes 1–3 are
targeted clones generating infertile chimeric mice (clone 011, 013
and 014, respectively). Lanes 4–6 are control non-targeted clones
from the same screening (clone 021, 023 and 025, respectively). All
non-targeted clones showed germ-line transmission. Zfp148 mRNA
is detected as three species in Northern blots2,24,25. Expression of all
three species was 50% lower in the heterozygous mutant ES cells
than in control ES cells (52.8% for two upper bands and 45.1% for
the bottom band) when hybridized with any Zfp148 probes except
that for exon 9. Blots were hybridized with a probe for 28S riboso-
mal RNA probe to confirm equal sample loading.
a
b
c
d
e
f
g
h
i
© 2003 Nature Publishing Group http://www.nature.com/naturegenetics

Page 2

letter
nature genetics • volume 33 • february 2003
173
gonads during mouse embryogenesis (Fig. 1a–f). In situ
hybridization shows that Zfp148 mRNA is strongly expressed in
male and female gonads at E12.5 but declines from E12.5 to
E13.5 (data not shown). Zfp148is primarily expressed in primor-
dial germ cells (PGCs), which can be recognized as large cells
with large nuclei (Fig. 1c,f).
In this study, we investigated the role of Zfp148 in germ cells
during mouse development. We generated Zfp148+/–mouse
embryonic stem (ES) cells by gene targeting and injected them
into blastocysts to create chimeric Zfp148 heterozygotes. The tar-
geted Zfp148 allele had a deletion of exon 9, which corresponds
to >60% of the Zfp148 coding region (Fig. 1g,h). Five of 1,200
doubly resistant clones had the correct targeting event, and we
used 4 of these for blastocyst injection. We obtained chimeric
mice from injections with Zfp148+/–ES cells but did not observe
germ-line transmission of the mutated allele of Zfp148 in these
mice. Some chimeras with a high proportion of Zfp148+/–cells
were infertile (five of nine Zfp148+/–chimeras were infertile, and
the remaining four did not show germ-line transmission; see
Web Table A online). In contrast, normal germ-line transmission
occurred in control chimeric mice (6 of 11 control chimeras
showed germ-line transmission; see Web Table A online) using
non-targeted ES cells from the same screening (defined as con-
trol ES cells). These results suggest that chimeric mice become
infertile owing to heterozygosity with respect to Zfp148 in the
male germ line.
Northern-blot analysis showed that transcription of Zfp148
was 50% lower in Zfp148+/–ES cells than in control ES cells
(Fig. 1i). Western-blot analysis indicated that full-length 89-kDa
Zfp148 was expressed in Zfp148+/–and control ES cells, and we
found no evidence that truncated forms of Zfp148 were pro-
duced from the mutated allele (data not shown). These results
suggest that the infertility of Zfp148+/–chimeric mice is due to
the haploinsufficiency of Zfp148 and not to the expression of a
dominant negative or otherwise abnormal form of the protein.
The testes of infertile chimeric Zfp148 heterozygotes were
much smaller than those of fertile, germ-line chimeras from con-
trol ES cells (Fig. 2a). Histological analyses showed that gametes
were completely absent and that the seminiferous tubules were
populated only with Sertoli cells (Fig. 2b–d). This phenotype is
similar to the human Sertoli cell–only syndrome5,6.
Fig. 2 Typical appearance of testes in infertile Zfp148+/–chimeric mice. a, Testis in
4-mo-old infertile Zfp148+/–chimeric mice or Zfp148 wild-type chimeric mice.
Testes of infertile mutant chimeric mice (right) were markedly smaller and
weighed an average of 80% less (21.9 ± 1.8 mg; n = 6 from three chimeras) than
testes of control chimeric mice (left; 113 ± 6 mg; n = 6 from three chimeras). No
overt abnormality was detected in attached tissues. r = Bar, 10 mm. Arrow indi-
cates testes (ts). b–d, Histological sample of testis of 4-mo-old Zfp148 wild-type
control chimeric mice (c) and from Zfp148+/–chimeric mice (b,d; d is higher mag-
nification of b). In the mutant testis, no spermatogonia were visible, and
tubules were filled by Sertoli cells (b,d). Hematoxylin-positive structures
(arrow), which might be debris of Sertoli cells, were detected in some tubules
(d). r = Bar, 150 µm (b) or 50 µm (c,d).
Fig. 3 Reduction in PGCs in gonads of tetraploid Zfp148+/–chimeric embryos.
Whole-mount alkaline-phosphatase staining of E8.5 (a), E11.5 (b) and E13.5
Zfp148+/–tetraploid embryos (c) and E13.5 control tetraploid embryos (d). In
E8.5 and E11.5 tetraploid chimeric embryos, the appearance and migration of
PGCs from Zfp148+/–ES cells seemed normal (a,b). Gonads of E13.5 Zfp148+/–
tetraploid embryos had fewer alkaline phosphatase–positive cells and weaker
ladder formation (c,d). e–j, Histological analyses of tetraploid chimeric
gonads. Hematoxylin and eosin staining (e,f), immunohistochemistry with a
PGC-specific antibody 4C9 (g,h) and TUNEL staining (i,j) of E13.5 Zfp148+/–
tetraploid embryos (e,g,i) and control tetraploid embryos (f,h,j). The gonads
of Zfp148+/–tetraploid embryos had fewer PGCs by hematoxylin and eosin
staining and by immunohistochemistry with a PGC-specific 4C9 antibody and
more TUNEL-positive cells than gonads of control tetraploid embryos. r = Bar,
500 µm (a,b), 1.0 mm (c,d), 50 µm (e–h) or 150 µm (i,j).
a
b
c
d
a
b
c
d
e
f
g
h
i
j
© 2003 Nature Publishing Group http://www.nature.com/naturegenetics

Page 3

letter
174nature genetics • volume 33 • february 2003
It is possible that testes of chimeric Zfp148 heterozygotes
lacked gametes because they never formed or because the
gametes formed in early embryogenesis but were lost at a later
stage. To distinguish between these possibilities, we generated
tetraploid chimeric embryos from Zfp148+/–ES cells and exam-
ined the PGCs. Tetraploid blastocysts contribute little to
embryonic tissues but do contribute to the extra-embryonic tis-
sues; in contrast, ES cells contribute only to embryonic tissues7.
By generating tetraploid chimeric embryos with Zfp148+/–ES
cells, we could form Zfp148+/–PGCs and observe them in the
absence of PGCs from the host blastocyst (defined as Zfp148+/–
tetraploid embryos).
At E8.0–E8.5, we detected Zfp148+/–PGCs at the base of the
allantois around the hindgut pocket (Fig. 3a). PGCs appeared
normal at E11.5, increasing in number and migrating into
the genital ridges (Fig. 3b). Zfp148+/–tetraploid embryos
appeared morphologically comparable to the control
tetraploid embryos at these stages. But at E13.5, alkaline-
phosphatase activity was weaker in the gonads of Zfp148+/–
tetraploid embryos than in those of tetraploid chimeric
embryos from control ES cells (defined as control tetraploid
embryos; Fig. 3c,d).
Pro-spermatogonia are large male germ cells that differentiate
from PGCs and can be identified by staining with hematoxylin
and eosin (Fig. 3f) or 4C9 antibody (refs. 8,9; Fig. 3h). The num-
ber of pro-spermatogonia was much lower in Zfp148+/–
tetraploid embryos (Fig. 3e,g) than in control tetraploid
embryos (Fig. 3f,h), which is consistent with whole-mount alka-
line-phosphatase staining of gonads (Fig. 3c,d). TUNEL staining
showed more large positive cells (probably pro-spermatogonia)
in tubules of Zfp148+/–tetraploid embryos than in tubules of
control tetraploid embryos (Fig. 3i,j). These results strongly
suggest that haploinsufficiency of Zfp148 caused apoptotic cell
death of pro-spermatogonia.
We also injected Zfp148+/–ES cells into lacZ-positive
ROSA26 diploid blastocysts and examined the embryos at
E18.5. These chimeric embryos produced a few lacZ-positive
pro-spermatogonia derived from the host blastocyst (Fig. 4) but
no lacZ-negative pro-spermatogonia from Zfp148+/–ES cells
(Fig. 4a,d). This indicates that Zfp148+/–pro-spermatogonia
were not rescued by wild-type Sertoli cells. The Zfp148+/–pro-
spermatogonia probably entered apoptosis before E18.5.
Fig. 4 Complete loss of pro-spermatogonia in
E18.5 mutant chimeric embryos. Chimeric
embryos were generated by injection of ES cells
into ROSA26 diploid blastocysts20. ES cell lineage
can be distinguished from host cells of ROSA26
embryos by staining the embryos with X-gal, and
only blastocyst-derived cells were stained blue.
After staining of whole testes with X-gal, sections
were counterstained with nuclear fast red. The
panels show testes derived mostly from Zfp148+/–
ES cells (a, higher magnification in d), testes
derived partly from Zfp148+/–ES cells (b, higher
magnification in e) and testes derived mostly
from control ES cells (c, higher magnification in
f). When chimeric testes were derived mostly
from Zfp148+/–ES cells (a), Sertoli cells were
mostly negative for lacZ (white). In these testes,
there were small numbers of lacZ-positive pro-
spermatogonia (asterisk), but no lacZ-negative
pro-spermatogonia from Zfp148+/–ES cells were
detected (a,d), indicating that pro-spermatogo-
nia were derived only from ROSA26 host blasto-
cysts (asterisk) and that they could survive even they were surrounded by Zfp148+/–(white) Sertoli cells. In contrast, when chimeric testes were derived mostly
from host embryos, they contained many lacZ-positive pro-spermatogonia and Sertoli cells (b,e), but no lacZ-negative pro-spermatogonia from Zfp148+/–ES cells
were detected. Pro-spermatogonia derived from control ES cells developed normally in chimeric testes (arrow) with some lacZ-positive pro-spermatogonia from
host embryos (asterisk; c,f). r = Bar, 100 µm (a–c) or 50 µm (d–f).
a
c
e
b
d
f
Fig. 5 Phosphorylated p53 and cell-cycle-arrest induction in control ES cells and
its impairment in Zfp148+/–ES cells. a, MTT assay of Zfp148+/–ES cells and con-
trol ES cells after 24 h of serum starvation. Data are the mean ± s.d. of eight
samples. The same experiment was done more than three times and showed
the same pattern. Although the proliferation of control ES cells (clone 023,
closed circle) decreased with decreasing serum concentration, Zfp148+/–ES cells
(clone 011, open circle) were resistant to serum starvation. b, Western-blot
analyses of Zfp148, p53 and Ser15P-p53 in Zfp148+/–and control ES cells after
serum starvation (3% serum). Zfp148 and Ser15P-p53 were upregulated in con-
trol ES cells (clone 023, right panel), but Zfp148+/–ES cells did not upregulate
Zfp148 and Ser15P-p53 to the same extent as the control (clone 011, left
panel). No upregulation of p53 was observed in either control or Zfp148+/–ES
cells. Coomassie brilliant blue–stained SDS–PAGE samples are shown to confirm
equal sample loading.
a
b
© 2003 Nature Publishing Group http://www.nature.com/naturegenetics

Page 4

letter
nature genetics • volume 33 • february 2003
175
Recent studies1,3,4and this study suggest that Zfp148 may regu-
late cell-cycle arrest in pro-spermatogonia through p53 or p21waf1
at E13.5. For this reason, we examined the expression and phos-
phorylation of p53 in serum-starved Zfp148+/–ES cells. Previous
studies have shown that phosphorylation of p53 at Ser15 and
Ser20 is essential for p53-dependent cell-cycle arrest10,11. Prolifer-
ation was inhibited in serum-starved control ES cells (Fig. 5a); in
contrast, Zfp148+/–ES cells continued to proliferate at a higher
rate than did control cells and were less sensitive to serum starva-
tion (Fig. 5a). Expression of Zfp148 and phosphorylation of p53
at Ser15 (Ser15P-p53) increased in control but not Zfp148+/–ES
cells (Fig. 5b). This suggests that Zfp148 regulates phosphoryla-
tion of p53 at Ser15 and that a lower level of Ser15P-p53 may con-
tribute to loss of cell-cycle arrest in Zfp148+/–ES cells.
We also examined expression of Zfp148 and phosphorylation of
p53 in wild-type and Zfp148+/–fetal germ cells in vivo. We
isolated wild-type E12.5–E13.5 male gonads and Zfp148+/–E13.5
male gonads from tetraploid embryos and examined them by
immunohistochemistry (Fig. 6). In wild-type male gonads,
Zfp148 (Fig. 6b), Ser15P-p53 (Fig. 6e) and p53 (Fig. 6h) were
expressed strongly at E13.5 but weakly at E12.5 (Fig. 6a,d,g). p53
localized primarily to the cytoplasm, but Ser15P-p53 localized to
the nucleus (Fig. 6e,h). We observed similar results in female
gonads from wild-type embryos (see Web Fig. A online). In
Zfp148+/–tetraploid chimeric embryos, Zfp148 and p53 were
expressed very weakly at E13.5 (Fig. 6c,i), and Ser15P-p53 was not
detectable in germ cells positive for 4C9 antibody (Fig. 6c,f). These
results suggest that homozygosity with respect to wild-type Zfp148
is required for phosphorylation of p53 at Ser15 in E13.5 germ cells
and that haploinsufficiency of Zfp148leads to loss of Ser15P-p53.
The results presented here show that Zfp148+/–ES cells were
impaired in p53-dependent cell-cycle arrest. Despite this fact,
Zfp148+/–germ cells did not continue to proliferate in gonads. In
addition, TUNEL-positive cells were observed more frequently
in Zfp148+/–gonads than in wild-type gonads, suggesting that
deficiency in Zfp148 led to abnormal growth or differentiation,
which triggered apoptosis in fetal germ cells at approximately
E13.5. The fact that Trp53-deficient mice do not show overt
gonadal abnormalities or infertility12might be explained by the
fact that Trp53 is a member of a gene family that includes Trp73
and Trp63. Thus, partial redundancy for function of Trp53 might
compensate for loss of Trp53 during germ-cell development. But
male Trp53-hypomorphic mice experience gamete degeneration
after birth13, suggesting that a reduced level of p53 cannot be
complemented by other genes in the Trp53 family.
Unlike Trp53, Zfp148 is not a member of a family of related
genes in the mouse, so haploinsufficiency of Zfp148 cannot be
complemented by a gene with redundant function. An alternative
explanation for the difference in phenotype of Trp53-null and
Zfp148+/–mice is that heterozygosity with respect to Zfp148 might
cause altered p53 function, whereas loss of Trp53 causes complete
loss of p53 function, and that these two functional states for p53
may have different phenotypic consequences during development.
Previous studies also indicate that a mutation in Atm in humans
and mice causes complete lack of male gametes14, which may be
mediated by altered Atm-dependent phosphorylation of p53 at
Ser15 (ref. 15). In these mutants, however, germ-cell degeneration
by apoptosis occurs during early meiosis after birth16,17. These
results suggest that Zfp148 and Atm both contribute to a complex
mechanism for regulating p53, and may have different functions in
the differentiation of germ cells.
Further studies are needed to
address how Zfp148 regulates
phosphorylation of p53 and
whether impaired phosphoryla-
tion of Ser15-p53 directly leads
to loss of fetal germ cells in the
male gonad.
Fig. 6 Zfp148 and phosphorylated p53
induction in pro-spermatogonia at
E13.5 in the male gonad. Antibodies
against Zfp148 (a–c), Ser15P-p53 (d–f),
p53 (g–i) and pre-immune rabbit serum
as negative control (j–l) were used for
E12.5 (a,d,g,j) and E13.5 (b,e,h,k) male
mouse gonads of wild-type embryos
and E13.5 male gonads of Zfp148+/–
tetraploid embryos (c,f,i,l). The germ
cell–specific antibody 4C9 was also
used to detect germ cells (a–c,j–l). In
a–c, Zfp148 expression was visualized
with NBT/BCIP (blue), and the presence
of embryonic germ cells was visualized
with diaminobenzidine (brown). In d–i,
Ser15P-p53 or p53 was visualized with
NBT/BCIP (blue), and sections in g–i
were counterstained with nuclear fast
red (pink) as well. Expression of Zfp148,
Ser15P-p53 and p53 were low at E12.5
but upregulated at E13.5 in the wild-
type gonads. Expression of Zfp148 and
p53 in the Zfp148+/–gonads at E13.5
were much weaker than age-matched
wild-type gonads. Ser15P-p53 signal
was not detected in pro-spermatogo-
nia (indicated by arrowheads that cor-
respond to 4C9-positive cells in c) in
Zfp148+/–gonads (f). No signal was
detected at all after NBT/BCIP staining
with pre-immune, normal rabbit serum
(j–l). r = Bar, 25 µm.
a
bc
d
e
f
g
hi
j
kl
© 2003 Nature Publishing Group http://www.nature.com/naturegenetics

Page 5

letter
176 nature genetics • volume 33 • february 2003
In summary, this study shows that haploinsufficiency of
Zfp148, an autosomal gene, causes infertility, and that two func-
tional copies of Zfp148 are required for fetal germ cells to survive
in male mouse gonads. A putative mechanism for this effect is
suggested by the fact that Zfp148 and Ser15P-p53 are induced in
wild-type but not Zfp148+/–male germ cells. Thus, Zfp148 may
have an essential role in differentiation of fetal germ cells by stim-
ulating phosphorylation of p53 at Ser15.
Methods
Generation of Zfp148+/–ES cells and chimeric mice. We isolated
HindIII/NsiI 5′ (2.7-kb) and BamHI/KpnI 3′ (4.5-kb) fragments of Zfp148
from a 129/SvJ mouse genomic library and used them to construct a replace-
ment gene-targeting vector (Fig. 1g). We inserted a neomycin-resistance
gene with a phosphoglycerokinase promoter (PGK-neo) between the NsiI
and BamHI sites, deleting more than 60% of the coding region of Zfp148. We
added a cassette expressing diphtheria toxin A to the long homology arm of
the construct. We carried out ES cell targeting and chimera production as
described elsewhere18. We identified five correctly targeted clones and used
another three clones from this screen as controls, after confirming that both
alleles of Zfp148 were intact. We confirmed that both Zfp148+/–ES cell lines
(clone 011, 012, 013, 014) and control ES cell lines (clone 021, 023, 025) were
euploid (40 chromosomes) and carried a Y chromosome.
Production of ES cell-derived embryos.We prepared tetraploid blastocysts by
fusing 2-cell stage embryos from CD-1 females, culturing them to blastocyst
stage using a previously described protocol19and carrying out standard blasto-
cyst injection. To generate ROSA26 blastocysts20, homozygous ROSA26 males
were mated with C57BL/6J females, and embryos were injected in the usual
manner. We recovered embryos at E8.0–18.5. We carried out β-galactosidase
staining as described21. Mice used in this study were treated according to NIH
Guide for Care and Use of Laboratory Animals, with procedures approved by
the National Institute for Longevity Sciences Animal Committees.
In situ hybridization. We carried out in situ hybridization as previously
described22with modifications. Briefly, we fixed embryos with 4%
paraformaldehyde, dehydrated them, embedded them in paraffin and pre-
pared sections of 7 µm. We labeled antisense and sense RNA probes corre-
sponding to coordinates bp 1,634–2,229 of Zfp148 with digoxigenin. We visu-
alized the RNA probe with Fab fragments from antibody against digoxigenein
conjugated with alkaline phosphatase (Roche) and 5-bromo-4-chloro-3-
indolyl-phosphate (BCIP)/nitroblue tetrazolium (NBT) solutions and coun-
terstained with methyl green. For northern-blot analysis of the targeted and
non-targeted ES clones, we loaded 20 µg of total RNA per lane and quantified
radioactive signals with an image analyzer system (BAS 2500, Fuji Film).
Histological analysis. We fixed embryo and adult testes with Bouin’s or
4% paraformaldehyde, dehydrated them and embedded them in paraffin.
We prepared sections of 5–7 µm and stained them with hematoxylin and
eosin. After staining for β-galactosidase, we used nuclear fast red as a coun-
terstain. We purchased antibody against PGC, 4C9 (LEX-2, rat IgM anti-
body; refs. 8,9) from Funakoshi, visualized by staining with diaminobenzi-
dine and counterstained the sections with hematoxylin. We raised rabbit
polyclonal antibody against Zfp148 against a synthetic polypeptide accord-
ing to standard protocol. The antigenic Zfp148 polypeptide corresponded
to amino acids 452–466 of Zfp148 (EGP SKP VHS STN YDD). We pur-
chased rabbit polyclonal antibodies against p53, against Ser15P-p53 and
against Ser20P-p53 from Cell Signaling Technology, visualized with
NBT/BCIP staining and counterstained them with nuclear fast red.
Whole-mount staining of embryos for PGCs. We fixed E8.0–13.5
embryos with 4% paraformaldehyde and stained for alkaline-phosphatase
activity with Fast-Red23.
3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT)
assay and western-blot analysis. We cultured mutant Zfp148+/–and con-
trol, non-targeted ES cells on STO feeder cells in ES medium (Dulbecco’s
modified Eagle’s medium with 15% fetal bovine serum, 2 mM glutamine
and 0.1 M 2-mercaptoehanol). To prevent contamination of feeder cells, we
cultured ES cells on gelatinized plates for more than two passages in ES
medium. We used leukocyte inhibitory factor to prevent differentiation. We
seeded 1.0 × 104ES cells per well on 96-well plates for 16 h, and then trans-
ferred cells to ES medium with variable serum concentration. The MTT
assay was done 24 h after serum starvation treatment using a Cell Counting
Kit (Dojindo). For western blots, we prepared ES cells as described for the
MTT assay and plated them on gelatinized plates. We prepared nuclear
extracts at the indicated time after transfer to 3% serum. We used a 5-µg
sample for western-blot analysis. We stained SDS–PAGE gels with
Coomassie brilliant blue to confirm equal sample loading in each gel lane.
Note: Supplementary information is available on the Nature
Genetics website.
Acknowledgments
We thank E.M. Eddy, G.-Q. Zhao, Y. Matsui, P. Koopman and R. Newbold
for helpful discussion; R. Behringer and A. Bradley for use of AB-1 ES cells;
T. Castranio for advice on production of tetraploid chimeras; M. Suzuki
for technical advice on blastocyst injection; T.C., G. Scott and S. Kishigami
for critical reading of the manuscript; S. Ito for helping experiment of
chromosomal count of ES cells; and Y. Takeuchi, Y. and K. Mishina and
Chata for encouragement. This study was supported by the Fund for
Comprehensive Research on Aging and Health.
Competing interests statement
The authors declare that they have no competing financial interests.
Received 6 September; accepted 25 November 2002.
1.Bai, L. & Merchant, J.L. ZBP-89 promotes growth arrest through stabilization of
p53. Mol. Cell. Biol. 21, 4670–4683 (2001).
Hasegawa, T., Takeuchi, A., Miyaishi, O., Isobe, K. & de Crombrugghe, B. Cloning
and characterization of a transcription factor that binds to the proximal promoters
of the two mouse type I collagen genes. J. Biol. Chem. 272, 4915–4923 (1997).
Hasegawa, T., Xiao, H. & Isobe, K. Cloning of a GADD34-like gene that interacts
with the zinc-finger transcription factor which binds to the p21(WAF) promoter.
Biochem. Biophys. Res. Commun. 256, 249–254 (1999).
Bai, L. & Merchant, J.L. Transcription factor ZBP-89 cooperates with histone
acetyltransferase p300 during butylate activation of p21(waf1) transcription in
human cells. J. Biol. Chem. 275, 30725–30733 (2000).
Rosai, J. (ed.) Male reproductive system, atrophy and infertility. in Ackerman’s
Surgical Pathology 1260–1265 (Mosby, St. Louis, 1996).
Wong, T.W., Straus, F.H. & Warner, N.E. Testicular biopsy in the study of male
infertility. I. Testicular cause of infertility. Arch. Pathol. 95, 151–159 (1973).
Nagy, A. et al. Embryonic stem cells alone are able to support fetal development
in the mouse. Development 110, 815–821 (1990).
Matsui, Y., Zsebo, K. & Hogan, B.L. Derivation of pluripotential embryonic stem
cells from murine primordial germ cells in culture. Cell 70, 841–847 (1992).
Yoshinaga, K., Muramatsu, H. & Muramatsu, T. Immunohistochemical localization
of the carbohydrate antigen 4C9 in the mouse embryo: a reliable marker of
mouse primordial germ cells. Differentiation 48, 75–82 (1991).
10. Momand, J., Zambetti, G.P., Olson, D.C., George, D. & Levine, A.J. The mdm-2
oncogene product forms a complex with the p53 protein and inhibits p53-
mediated transactivation. Cell 69, 1237–1245 (1992).
11. Thut, C.J., Goodrich, J.A. & Tjian, R. Repression of p53-mediated transcription by
MDM2: a dual mechanism. Genes Dev. 11, 1974–1986 (1997).
12. Donehower, L.A. et al. Mice deficient for p53 are developmentally normal but
susceptible to spontaneous tumors. Nature 356, 215–221 (1992).
13. Rotter, V. et al. Mice with reduced levels of p53 protein exhibit the testicular giant-
cell degenerative syndrome. Proc. Natl. Acad. Sci. USA 90, 9075–9079 (1993).
14. Barlow, C. et al. Atm-deficient mice: a paradigm of ataxia telangiectasia. Cell 86,
159–171 (1996).
15. Banin, S. et al. Enhanced phosphorylation of p53 by ATM in response to DNA
damage. Science 281, 1674–1677 (1998).
16. Xu, Y. et al. Targeted disruption of ATM leads to growth retardation,
chromosomal fragmentation during meiosis, immune defects, and thymic
lymphoma. Genes Dev. 10, 2411–2422 (1996).
17. Barlow, C. et al. Atm deficiency results in severe meiotic disruption as early as
leptonema of prophase I. Development 125, 4007–4017 (1998).
18. McMahon, A.P. & Bradley, A. The Wnt-1 (int-1) proto-oncogene is required for
development of a large region of the mouse brain. Cell 62, 1073–1085 (1990).
19. Wang, Z.Q., Kiefer, F., Urbánek, P. & Wagner, E.F. Generation of completely
embryonic stem cell-derived mutant mice using tetraploid blastocyst injection.
Mech. Dev. 62, 137–145 (1997).
20. Zambrowicz, B.P. et al. Disruption of overlapping transcripts in the ROSA βgeo 26
gene trap strain leads to widespread expression of β-galactosidase in mouse
embryos and hematopoietic cells. Proc. Natl. Acad. Sci. USA 94, 3789–3794 (1997).
21. Hogan, B., Beddington, R., Costantini, F. & Lacy, E. (eds.) Techniques for visualizing
genes, gene products, and specialized cell types. in Manipulating the Mouse
Embryo 373–375 (Cold Spring Harbor Laboratory Press, Plainview, New York, 1994).
22. Takeuchi, A. et al. Microglial NO induces delayed neuronal death following acute
injury in the striatum. Eur. J. Neurosci. 10, 1613–1620 (1998).
23. Buehr, M. & McLaren, A. Isolation and culture of primordial germ cells. Methods
Enzymol. 225, 58–77 (1993).
24. Wang, Y., Kobori, J.A. & Hood, L. The htβ gene encodes a novel CACCC box-
binding protein that regulates T-cell receptor gene expression. Mol. Cell. Biol. 13,
5691–5701 (1993).
25. Taniuchi, T., Mortensen, E.R., Ferguson, A., Greenson, J. & Merchant, J.L.
Overexpression of ZBP-89, a zinc finger DNA binding protein, in gastric cancer.
Biochem. Biophys. Res. Commun. 233, 154–160 (1997).
2.
3.
4.
5.
6.
7.
8.
9.
© 2003 Nature Publishing Group http://www.nature.com/naturegenetics