407
Development 128, 407-416 (2001)
Printed in Great Britain © The Company of Biologists Limited 2001
DEV3307
Function of basonuclin in increasing transcription of the ribosomal RNA
genes during mouse oogenesis
Qinjie Tian1,*, Gregory S. Kopf2, Raymond S. Brown3 and Hung Tseng1,‡
1Department of Dermatology, University of Pennsylvania School of Medicine, Philadelphia, PA 19104, USA
2Center for Research on Reproduction and Women’s Health, University of Pennsylvania School of Medicine,
Philadelphia,
PA 19104, USA
3Department for Molecular and Structural Biology, University of Aarhus, DK-8000 Arhus C, Denmark
*Present address: Department of Obstetrics and Gynecology, Peking Union Medical College Hospital, Beijing, China
‡Author for correspondence (e-mail: [email protected])
Accepted 20 November 2000; published on WWW 11 January 2001
SUMMARY
Active protein synthesis during early oogenesis requires
accelerated transcription of ribosomal RNA genes
(rDNAs). In response to this demand, rDNAs are amplified
more than 1000-fold early in Xenopus oogenesis. Here, we
report evidence that rDNA is not amplified in mouse
oocytes, but these cells may instead employ the zinc-finger
protein basonuclin, a putative rDNA transcription factor,
to enhance rRNA synthesis. This conclusion is based on
observations that basonuclin is localized in the nucleolus in
the mouse oocyte early in its growth phase, when rRNA
INTRODUCTION
Developing oocytes accumulate a large quantity of proteins,
ribosomes and other cellular components to support early
embryonic development. In Xenopus, a mature oocyte contains
approx. 100,000 mitochondria (compared with the 70 or so on
a sperm), 100,000 DNA polymerases and 200,000 ribosomes
(Laskey, 1974). To amass this storage, Xenopus oocytes
produce, at their peak, 34 ng of protein per hour. Because of
this burden of protein synthesis, one of the highly transcribed
genes in Xenopus oocytes is the 45S ribosomal rRNA gene
(rDNA), a key factor in determining ribosome production.
Xenopus possess approx. 450 copies of rDNA per haploid
genome and in the early meiotic oocytes (4C), there are
approximately 1800 copies of this gene. This large number of
copies, however, is apparently not adequate to supply a
sufficient amount of rRNA. During the early pachytene stage,
these rDNAs are amplified to nearly two million copies (Brown
and Dawid, 1968; Perkowska et al., 1968). It was estimated
that without this amplification, it would take a frog oocyte
~1.7×105 days to accumulate the same amount of the rRNA
that it normally accumulates in 65 days (Perkowska et al.,
1968). Mammalian oocytes differ from those of the amphibians
in many ways, the most apparent being in their size. In the
mouse, the diameter of a primordial oocyte is approx. 15 to 20
µm, which increases rapidly during the growth phase to 80 to
85 µm in fully grown oocytes. However, this is much smaller
transcription is highly active; and that the binding sites of
basonuclin zinc fingers on the human and mouse rDNA
promoters are homologous. In a co-transfection assay,
basonuclin can elevate transcription from an rDNA
promoter, and its zinc-finger domain can inhibit RNA
polymerase I transcription, as detected by a run-on assay,
in growing mouse oocytes.
Key words: rDNA amplification, RNA polymerase I, Upstream
binding factor, Oocytes, Nucleolus, Mouse
than that of the mature Xenopus oocyte, which reaches 1 mm
and is 4000 times larger in volume than the mouse oocyte. One
of the probable reasons for the smaller size of the mouse oocyte
is that fewer embryonic cell cycles are supported by maternal
contribution in mouse than that in amphibians (Schultz, 1993)
and that no yolk proteins are accumulated. In the mouse
embryo, rRNA transcription is activated at six- to eight-cell
stage (Kopecny et al., 1995), which may render the storage of
huge quantities of ribosomes unnecessary. This, however, does
not imply that mouse oocytes do not accumulate proteins and
ribosomes. On the contrary, this accumulation is rapid during
the growth phase of the developing oocyte and nears
completion (95%) within 2 weeks, when the oocyte reaches 6070% of its final size (Kaplan et al., 1982). During this period,
protein synthesis climbs to a peak of 0.175 pg per hour per pl
of active cytoplasm, which is comparable with the 0.34 pg/hr/pl
of the Xenopus oocyte (Schultz et al., 1978). Accompanying
this protein accumulation is a steady increase in RNA,
consistent with a doubling of the activities of RNA polymerase
I and II (pol I and II) during this period (Moore and LinternMoore, 1978). The majority of these RNAs are rRNA (6070%), whose rate of synthesis remains constant at 0.01 pg per
minute over the first 9 days of oocyte growth and accelerates
to 0.015 pg/min in the next 5 days, until the cessation of
synthesis at 15 days of growth (Kaplan et al., 1982). As a
reflection of this rate of rRNA accumulation, the nucleolus
enlarges from 2 µm to almost 10 µm in diameter and there is
408
Q. Tian and others
a fourfold increase of ribosomes in the oocyte. Based on this
measurement of the rate of rRNA synthesis and the 4000-fold
difference in volume between mouse and Xenopus oocytes, it
has been proposed that rDNA amplification in the mouse
oocyte at the scale observed in the Xenopus oocyte is
unnecessary (Kaplan et al., 1982). However, little is known
regarding the mechanism that may account for the increase in
pol I activity.
Transcriptional regulation of rDNA is one of the first and the
best studied gene regulatory mechanisms. In mammals, two
conserved cis-elements within the promoter of rDNA have
been identified, the upstream promoter element (UPE or UCE)
and the core promoter element (CPE). In humans, the basic
transcription machinery for rDNA contains, in addition to pol
I, at least two transcription factors, the upstream binding factor
(UBF) and the SL-1 complex. The latter comprises several
proteins, including the TATA-binding protein (TBP), which is
also found in transcriptional complex of pol II and III. Pol I
transcription factors are targets of regulatory mechanisms such
as phosphorylation (Voit et al., 1995), and of regulatory
proteins such as SV40 large T-antigen (Zhai et al., 1997), Kuassociated proteins (Kuhn et al., 1993; Hoff et al., 1994; Niu
and Jacob, 1994), p53 (Budde and Grummt, 1999) and Rb
proteins (Cavanaugh et al., 1995b; Voit et al., 1997).
Basonuclin, a zinc-finger protein of 120 kD first discovered
in human keratinocytes, is also highly expressed in the germ
cells of testis and ovary (Tseng and Green, 1992; Tseng and
Green, 1994; Yang et al., 1997; Mahoney et al., 1998). In
human keratinocytes, basonuclin was detected in the nucleolus
of interphase cells and at the rDNA loci on mitotic
chromosomes (Tseng et al., 1999). The zinc fingers of
basonuclin were shown to bind to the upstream control element
(UCE) of the promoter of human rDNA (Iushi and Green 1999,
Tseng et al., 1999). These observations suggest that basonuclin
may be a transcriptional regulator for the rDNA. Here, we
tested the hypothesis that basonuclin enhances rDNA
transcription in the mouse oocyte. This hypothesis has several
predictions: (1) that basonuclin is localised in oocyte nucleolus
and associated with pol I transcription both temporally and
spatially; (2) that its interaction with rDNA promoter is
evolutionarily conserved; (3) that it can increase rDNA
transcription; (4) and that perturbation of its function will cause
reduction in oocyte rRNA synthesis.
MATERIALS AND METHODS
Cells
To compare the copy number of rDNA, oocytes were collected from
4- to 5-week-old F1 hybrid (CF1) mice (Charles River Laboratories,
Wilmington, MA) by needle puncture of isolated ovaries followed by
mechanical removal of the attached cumulus cells with a fine-bore
glass micropipet (Fig. 1A). For immunocytochemistry and the run-on
assay, oocytes of 10- to 12-day-old CF1 mice (Charles River) were
collected similarly into M2 media in the presence or absence of 100
µg/ml dbcAMP, which had no effect on the results presented in this
study. Primary mouse fibroblasts (C57BL/6J) were cultured in
DMEM (Life Science Technology) with 10% calf serum.
Extraction of DNA from oocytes and somatic tissues
Three hundred oocytes, which contain 3.6 ng of genomic DNA, were
collected for each PCR reaction. We employed two complementary
methods of obtaining oocyte genomic DNA. The first is a modification
of the method described by Dawid for frog oocytes (Dawid, 1965).
Mouse oocytes were gently homogenized in 20 volumes of a solution
containing 0.07 M Tris-HCl, pH 7.7, 3 mM EDTA and 2% SDS. The
homogenate was extracted with an equal volume of phenol for 15
minutes. Sodium acetate was added to the aqueous phase to a
concentration of 0.3 M and the nucleic acid was precipitated with 2.5
volumes of ethanol. This method produces better quality DNA but
may risk DNA loss during purification. To prevent DNA loss, oocytes
were put into a 200-µl PCR tube and genomic DNA was released by
rapid freeze-thaw cycles twice (on dry ice and 37°C water bath) and
used directly for PCR. Somatic DNA was extracted as described
previously (Tseng and Green, 1988) and the DNA concentration
determined by spectrophotometry.
PCR
PCR was performed with primer pairs mrDP5 (5′-AGAAGCCCTCTCTGTCCCTG) and mrDP3 (5′-GCGTGGCTCGGGGACAGCTTC) for the promoter region (−286 to +81) of the mouse rDNA
and Mid1-5 (5′ TCGCCGTGCCTACCATGGTG), and Mid1-3 (5′CGGGCCTGCTTTGAACACTC) for a region in the coding sequence
(+4385 to +4843). Cycle parameters were 94°C for 5 minutes, 20
cycles of 94°C for 1 minute, 63°C for 1 minute and 72°C for 1.5
minutes, and finally 72°C for 7 minutes. One of the primers in each
pair was labeled with T4 polynucleotide kinase (New England
Biolabs, Beverly, MA). The amplified fragment was analysed on a 1%
agarose gel and the autoradiogram of the gel was scanned using a
STORM-80 PhosphorImager and the image quantified using an
ImageQuant program (IQMac V1.2). To reduce cross contamination
of samples due to spillage during gel loading, the PCR samples of
oocytes were always separated from those of the somatic tissue by
one or more gel lanes.
Transcription run-on assay
Oocytes and fibroblasts were incubated at room temperature for 5
minutes in a permeabilization buffer containing 0.4% and 0.04% of
Triton X-100, respectively. A run-on buffer (50 mM Tris-HCl, pH 7.4,
100 mM KCl, 5 mM MgCl2, 0.5 mM EGTA, 25 mM S-adenosyl-Lmethonine) containing 0.5 mM ATP, CTP, GTP and 0.2 mM Br-UTP
was added to the cells and incubated for 1 hour at 37°C. In some cases,
anti-basonuclin or anti-UBF antibodies were included in the run-on
reaction and these antibodies were visualized along with BrU at the
end of the run-on. When indicated, actinomycin D (1 µg/ml) or αamanitin (100 µg/ml) were added to the run-on buffer. To demonstrate
that the BrU was incorporated into RNA, an RNase digestion (RNase
A, 400 µg/ml in PBS) was performed before immunodetection. At the
end of the run-on, the cells were washed twice in PBS containing
5 units/ml RNasin (Promega) and fixed immediately in 2%
paraformaldehyde. The incorporated BrU was detected with a
monoclonal anti-BrdU antibody (1:400) and visualized by a goat-antimouse secondary antibody conjugated with FITC, as described above.
Recombinant basonuclin (B10B) and Krüppel (4015) zinc fingers
were expressed and purified as previously described (Tseng et al.,
1999). Briefly, appropriate cDNA sequences were cloned into the
expression vector pRSET (Invitrogen) and expressed in Escherichia
coli strain BL21. The recombinant proteins were then purified to near
homogeneity on a Biorex-70 column (BioRad), stored in 500 mM
NaCl and 50% v/v glycerol at 350 µg/ml, and diluted 1:10 to 1:100
into run-on buffer. All blocking experiments were performed in the
presence of α-amanitin (100 µg/ml).
Immunocytochemistry
Immunocytochemistry on cryosectioned ovary was performed as
described in Mahoney et al. (Mahoney et al., 1998). Isolated oocytes
were washed in PBS and a permeabilization buffer (20 mM Tris-HCl,
pH 7.4, 5 mM MgCl2, 0.5 mM EGTA, 0.5 mM PMSF) briefly and
incubated at room temperature for 5 minutes in the permeabilization
Basonuclin in mouse oogenesis
buffer containing 0.4% Triton X-100 (Sigma) followed by three
washes in the permeabilization buffer without Triton X-100. Affinitypurified anti-mouse basonuclin antiserum (MBP-767) at a dilution of
1:20 or an autoantibody against UBF (anti-NOR 90) (Chan et al.,
1991) at a dilution of 1:200 was diluted in the run-on buffer (see
below). Anti-nucleolus antibody (ANA) (Binding Site) was used
undiluted according to the manufacturer’s instructions. Permeabilized
oocytes were incubated with antibodies at 37°C for 1 hour, followed
by fixation for 20 minutes with 2% paraformaldehyde in PBS and by
permeabilization in 0.2% Triton X-100 in PBS for 15 minutes. After
incubation with 2% BSA in PBS for 60 minutes, the oocytes were
incubated for 1 hour at room temperature with either Texas Redlabeled, goat-anti-rabbit secondary antibodies (Molecular Probes) (for
basonuclin) or with FITC-labeled, goat-anti-human secondary
antibodies (Biosource)(for UBF). DNA was visualized with 1 µg/ml
DAPI (Sigma).
DNase I footprinting
DNase I footprinting was performed exactly as described previously
(Tseng et al., 1999). A DNA fragment (259 bp) containing the
promoter region of mouse rRNA gene was prepared by PCR with
primer pair (mrp-5 5′-CAGTTCTCCGGGTTGTCAGGG; mrp-3
5′-AACAGCCTTAAATCGAAAGGG). End-labeling of one of the
primers with [32P] was achieved by T4 polynucleotide kinase.
Transfection and primer extension assay
COS-7 cells (ATCC) were cultured in DMEM with 20% fetal calf
serum to 90% confluence in 55 mm dishes. A total of 6 µg of DNA,
of which 3 µg is the reporter and 3 µg the expression construct, was
transfected with lipofectamine (Life Science Technology) according
to the manufacturer’s instructions. Transfection efficiency was 60%,
as monitored by a green fluorescence protein (GFP) expression vector
(pGreen Lantern-1, Life Science Technology). All cDNA constructs
were checked by in vitro translation and their expression by
immunocytochemistry. Primer extension was performed as described
by Zhai et al. (Zhai et al., 1997). RNA was extracted using Trizol
Reagent (Life Science Technology) from COS cells 48 hours posttransfection. Total RNA (40 µg) was incubated with a [32P]-labeled
primer (HuCAT-PE, 5′-AGAGTTGAGAGGGTACGTACG) in 15 µl
409
of 150 mM KCl, 10 mM Tris-HCl (pH 8.3) and 1 mM EDTA in an
Eppendorf tube submerged in a 500 ml water bath, which was allowed
to cool from 95 to 40°C at room temperature. For primer extension,
30 µl of solution containing 30 mM Tris-HCl, pH 8.3, 15 mM MgCl2,
8 mM DTT, 0.23 mg/ml actinomycin D, 0.22 mM 4 dNTP and 5 units
of AMV reverse transcriptase (Promega) was added to the
primer/RNA mix and incubated at 42°C for 1 hour.
RESULTS
The 45S ribosomal RNA genes (rDNA) are not
amplified in the mouse oocyte
We compared the copy number of rDNA in the mouse oocyte
to that of the somatic tissues by a PCR method with [32P]labeled primers. Oocyte genomic DNA was isolated from a
population in which most of the oocytes were in the growth
phase or prior to germinal vesicle breakdown (Fig. 1A). The
rDNA amplification in Xenopus oocytes takes place during the
pachytene stage of meiosis prior to the onset of rRNA
synthesis. By analogy, any rDNA amplification that would
benefit rRNA synthesis during the growth phase of mouse
oocytes should take place before this growth period. By
selecting oocytes that were in the growth phase or after, we
ensured that such rDNA amplification had already taken place
and was therefore detectable. Care was taken to remove the
granulosa (cumulus) cells that surround the oocytes. These
cells are of somatic origin and, although the copy number of
their rDNA has not been investigated, they are not likely to
have been amplified. Granulosa cell (2C to 4C, depending on
the stage of cell cycle) contamination in our oocyte (4C)
preparations was estimated to be approx. 10% in total cell
number (Fig. 1A, arrows) and less than 10% in genomic DNA.
As it is not feasible to obtain a large number of oocytes from
mouse to employ the conventional method to determine DNA
concentration, we relied on the generally accepted estimate that
Fig. 1. A comparison of the copy number of rDNA in the
mouse oocyte and somatic tissues. (A) Oocyte genomic
DNA was obtained from a mixed population of oocytes
isolated from 4-week old mice. A sample of this population
is shown. Most oocytes contained a recognizable germinal
vesicle (n), indicative of prophase of meiosis I.
Contaminating cumulus cells are indicated by arrows.
(B) Linear input/output response of the PCR; the data were
derived from quantification of autoradiography of the gel
shown in C. The promoter region of the somatic rDNA was
PCR-amplified with a radiolabeled primer. A tenfold serial
dilution of template DNA was used to produce the
concentration response. The asterisk shows the
quantification of the PCR fragment amplified from 3 pg of
pCR-mrDNAP, which contains a similar amount of rDNA
to 3.5 ng of genomic DNA. (C) An autoradiograph of the
gel from which the amplified somatic rDNA promoter
sequence was subjected to electrophoresis. Lanes 1-3,
genomic DNA of 3.5 ng, 35 ng and 350 ng were used as
template. Lane 4, 3 pg of pCR-mrDNAP was used as
template. (D-G) Copy numbers of two regions of the rDNA
were examined, the promoter (D,E) and a region in the
middle of the coding sequence (F,G). Oocyte DNA was
obtained by two methods, the method of Dawid (Dawid, 1965) (D,F), and the method of freeze-thaw and direct amplification (E-G). In D-G,
the template DNAs were lane 1, oocyte DNA; and lanes 2 and 3, somatic DNA (3.5 ng and 35 ng, respectively). No lane 3 is present in G,
owing to sample loss.
410
Q. Tian and others
Fig. 2. Intranuclear distribution of basonuclin in growing oocytes.
Cryosectioned ovary of a 12 day-old mouse was stained with an antibasonuclin antibody (MBP-767) (red) and Hoechst 33258 dye (blue).
Shown is the nucleus of a primary oocyte. The Hoechst stain revealed the
condensed heterochromatin patches (arrows), which surround the lightly
stained nucleoli (n) (A). Basonuclin appears to be evenly distributed
between the nucleoli and the nucleoplasm, regions of more intense
basonuclin stain are shown by arrowheads (B). (C) A and B combined.
each oocyte contains 12 pg of genomic DNA. Somatic genomic
DNA was extracted from mouse liver, kidney, lung and
intestine.
Using a mouse rDNA promoter fragment (−300 to +150,
transcription starts at +1) as a control, we first established
conditions whereby the quantity of the amplified fragment was
linearly proportional to the quantity of the template over a
range of two orders of magnitude (3.5 to 350 ng) (Fig. 1B).
Under these conditions, we determined that the minimal
quantity of genomic DNA required for the detection of rDNA
is 3.5 ng (Fig. 1C). This amount of genomic DNA should
contain approximately 0.3 pg of the target rDNA fragment of
446 bp, since the copy number of rDNA is 400-500 per haploid
mouse genome of 2.7×109 bp. We tested this estimate by
amplifying serial dilutions of the same target fragment that was
cloned in pCR-II (pCR-mrDP), in which the rDNA promoter
sequence is approx. 10% of the mass of the entire plasmid
(446 bp insert and 3948 bp vector). When 0.3 pg of the cloned
rDNA promoter sequence was used as template, the amplified
fragment (Fig. 1C, lane 4) contained radioactivity similar to
that from 3.5 ng of genomic DNA (Fig. 1C, lane 1), which,
according to our estimate, also contains 0.3 pg of the rDNA
promoter fragment. This agreement demonstrates that under
our PCR conditions, the quantity of rDNA in the mouse
genomic DNA could be measured correctly (Fig. 1B).
We examined two regions of the 45S rRNA gene: the
promoter region and a region in the middle of the coding
sequence. We found that at comparable concentrations, the
oocyte DNA produced a similar quantity of amplified rDNA to
that obtained from somatic cell DNA (Fig. 1D-G). This was
true for both regions of the rDNA tested as well as for oocyte
DNA prepared by two different methods. Therefore, no rDNA
amplification could be detected in mouse oocytes.
Basonuclin distribution during the growth phase of
oocyte development
We reported previously that basonuclin is present in the
nucleus of developing mouse oocytes (Mahoney et al., 1998).
Here, we investigated its nuclear distribution during the first
meiosis by immunocytochemistry on cryosectioned mouse
ovaries. In the 10- to 15-day-old postnatal ovary, basonuclin
was found exclusively in the nucleus of primordial, primary
and secondary oocytes. At this age, the heterochromatin
patches can be identified by their intense DAPI fluorescence
(Fig. 2A, arrows). These heterochromatin patches surround the
nucleolus of oocytes that are competent to resume meiosis and
can serve as landmarks for this subnuclear structure (Mattson
and Albertini, 1990). The immunostaining of basonuclin
appeared uniform within the nucleus of growing oocytes.
An exception was that basonuclin fluorescence in the
heterochromatin regions was weaker than the rest of the
nucleus (Fig. 2B). Occasionally, small regions with stronger
basonuclin fluorescence were seen within the nucleoplasm
(Fig. 2B, arrow), but the underlying subnuclear structure could
not be identified.
Nucleolar basonuclin is associated with active pol I
transcription
To investigate pol I transcription in mouse oocytes, we
established a run-on assay for pol I transcription in Triton
X100-permeabilized primary oocytes (Masson et al., 1996),
when transcriptional activity is at its peak. Pol I transcription
was detected by using BrU as one of the substrates, and the
incorporated BrU was visualized by immunocytochemistry.
The nucleolar location of the incorporated BrU was
demonstrated by its juxtaposition with the heterochromatin
patches (arrows, Fig. 3B) and by their co-localization with an
anti-nucleolus antibody (ANA) (not shown). The interpretation
that the incorporated BrU reflects pol I transcription was
supported by its sensitivity to actinomycin D (Fig. 3C) and
RNase (not shown), and resistance to α-amanitin (Fig. 3D). Pol
I transcription could be detected readily within two hours of
oocyte permeabilization (Q. T. and H. T., unpublished
observations) and beyond that the BrU incorporation
diminished rapidly. Because it was estimated that the
Fig. 3. Pol I transcription in permeabilized mouse oocytes in their
growth phase. Immature oocytes were isolated from 10- to 12-dayold mice and permeabilized with Triton X-100. Run-on reaction was
performed as described by Masson et al. (Masson et al., 1986) with
BrUTP as one of the nucleotide precursors. (A) BrU incorporation
into an oocyte after 1 hour run-on reaction. (B) In another oocyte,
nucleolar localization of the incorporated BrU (green) is shown by its
juxtaposition with heterochromatin patches (blue, arrows). Note that
the periphery of the nucleolus contains more label. (C) Run-on
reaction in the presence of 1 µg/ml actinomycin D, no incorporated
label was observed. (D) Run-on reaction in the presence of 100
µg/ml α-amanitin. Incorporated label is indicated by an arrow. In BD, the image of DNA stain (DAPI) is superimposed on the BrU stain.
Basonuclin in mouse oogenesis
completion of one round of transcription of the rRNA gene in
HeLa cells requires less than 5 minutes (Penman et al., 1966),
it is very likely that the incorporated BrU was the result
of multiple rounds of transcription initiation after
permeabilization. No significant BrU incorporation in the
nucleoplasm was detected, suggesting that pol II activity was
suppressed under our experimental conditions.
To investigate the distribution of basonuclin and UBF in the
permeabilized oocytes, we included the anti-basonuclin
(M767) and the anti-UBF (anti-Nor 90) antibodies in the runon buffer for the duration of the assay. The antibodies had little
effect on the BrU incorporation into the nucleolus (Fig. 4G).
Anti-UBF antibody detected a protein localized exclusively
in the nucleolus surrounded by the intact heterochromatin
patches (Fig. 4B,E). The unperturbed heterochromatin and
UBF distribution suggests that the basic nucleolar organization
was maintained, consistent with the ability of the Triton-treated
oocytes to continue pol I transcription. In Triton X-100-treated
oocytes, however, nucleolar basonuclin became more
prominent and the fluorescence intensity of nucleoplasmassociated basonuclin was reduced notably but still visible
(compare Fig. 4C,H with Fig. 2B). The nucleoplasmassociated basonuclin was therefore more easily extracted by
Triton X-100 than was the nucleolar-associated basonuclin.
The Triton-permeabilization method thus allows us to
distinguish cytochemically the nucleolar basonuclin from that
in the nucleoplasm and, more importantly, to
demonstrate the presence of basonuclin in nucleoli
undergoing active pol I transcription.
Basonuclin-binding sites on the human
and mouse rDNA promoters are similar
We and others have shown that recombinant
proteins containing the N-terminal pair of zinc
fingers of human basonuclin interact with the UCE
of the rDNA promoter (Iuchi and Green, 1999,
Tseng et al., 1999). Our recombinant protein,
designated B10B, binds specifically to three sites
within the promoter region of a human rDNA
(Tseng et al., 1999). The N-terminal pair of
basonuclin zinc fingers is evolutionarily highly
conserved. The amino acid sequence of B10B is
identical to that of the corresponding pair of zinc
fingers of mouse basonuclin, despite nucleotide
substitutions of about 8% (Matsuzaki et al., 1997).
The mouse rDNA promoter contains cis-elements
(the core and the UCE) that are functionally
equivalent to the corresponding elements in the
human promoter (Grummt, 1999). However, there is
very little overall DNA sequence homology
between the mouse and human rDNA promoters.
Therefore, it is of interest to compare the binding
sites of B10B on the human rDNA promoter with
those of the mouse. We examined a DNA fragment
of the mouse rDNA promoter (−188 to +71) by
DNase I footprinting (Fig. 5). This fragment
contains the cis-elements: the transcription
initiation site, the UCE, the core and the terminator.
B10B protein binds to three binding sites on the
mouse rDNA promoter, whose relative positions to
the cis-elements are very similar to those on the
411
human rDNA promoter (Fig. 6). Binding site A overlaps with
the 3′ half of the UCE. Binding site B is positioned in between
the UCE and the core. Binding site C, which localizes in the
transcribed region just beyond the transcription initiation site,
is divided into two parts, designated C and C′ (Figs 5, 6).
The B10B-binding sites on the mouse and human rDNA
promoters are not only similar in their location, but also in their
affinity for B10B. Binding site A, which was fully protected
from DNase I digestion at the lowest B10B concentration (14
µg/ml), is the high-affinity site in both human and mouse
promoters. At the highest protein concentration (56 µg/ml),
both sites extended toward the 5′ region (Fig. 5). In both
species, binding sites B and C have lower affinity for B10B;
they were fully protected only at the highest protein
concentration.
Based on the DNase I footprints of B10B on the human
rDNA promoter, we reported a consensus of the target
sequence of B10B: 5′-(G/C)G(C/T)G(A/T)C (Tseng et al.,
1999), which was also noted by Iuchi and Green (Iuchi and
Green, 1999). The footprints of B10B on the mouse rDNA
promoter have a similar consensus: 5′- (A/G/C)GTG(G/A/T)C.
By combining these results, we propose a new consensus target
of B10B (5′-(A/G/C)G(C/T)G(G/A/T)C), in which the G at the
second and the fourth positions, and the C at the sixth position
are invariable. We conclude that B10B-binding sites on the
mouse and human rDNA promoters are homologous.
Fig. 4. Nucleolar localization of basonuclin in permeabilized mouse oocytes
under conditions that favor transcription by pol I but not pol II. (A-F) A Triton X100-permeabilized, immunocytochemically stained, immature oocyte. (A) Phase
microscopy. (B-F) Fluorescence microscopy for UBF (green, B), basonuclin (red,
C), DNA (blue, D), UBF and DNA (E), and UBF and basonuclin (F).
(G-I) A similarly permeabilized oocyte viewed for BrU incorporation
(G), basonuclin (H), and BrU and basonuclin (I).
412
Q. Tian and others
was made in the pFBwt to remove the C-terminal two thirds
of the basonuclin coding sequence, including the nuclear
localization signal (pFBd). This basonuclin deletion mutant
lost its ability to elevate CAT RNA level (Fig. 7, lanes 5 and
6).
Fig. 5. The DNase I footprints of the N-terminal zinc fingers of
human basonuclin (B10B) on the mouse rDNA promoter. Both the
noncoding (left) and the coding (right) strands are shown. The start
site and the direction of transcription are shown by an arrow. The two
cis-elements, the core and the UCE are marked by brackets. The
number on top of each lane indicates the relative amount of the B10B
protein used (lane 1, 14 µg /ml; lane 2, 28 µg /ml; lane 4, 56 µg /ml).
The lane marked ‘G’ contains the sequencing reaction that serves as
a molecular weight marker. Regions protected from DNase I
digestion are indicated by black bars, which are labeled A, B, C and
C′, corresponding to the respective binding sites. The sites of
enhanced DNase cleavage are shown by arrowheads.
Basonuclin enhances transcription from an rDNA
promoter
To test the effect of basonuclin on the transcription from the
rDNA promoter, we co-transfected into COS-7 cells a reporter
plasmid (prHuCAT), which contains the human rDNA
promoter sequence +88 to −483 (Zhai et al., 1997), with an
expression construct of the human basonuclin cDNA (pFBwt)
transcriptionally driven by the cytomegalovirus promoter of
pcDNA3.1. In the control, the reporter plasmid was cotransfected with pcDNA3.1. Transcription from the rDNA
promoter was monitored by primer extension. Correct pol I
initiation from the rDNA promoter yields a primer extension
product of 123 bases (Zhai et al., 1997). We detected an
approximately twofold increase in the steady-state CAT RNA
in the presence of basonuclin compared with the control (Fig.
7). To verify that this stimulatory effect on transcription from
the rDNA promoter was dependent on basonuclin, a deletion
The zinc finger domain of basonuclin inhibits pol I
transcription in mouse oocytes
We tested the ability of B10B to interfere with pol I
transcription in mouse oocytes. This is because B10B binds to
the mouse rDNA promoter and is small enough (12 kDa) to
diffuse freely into the nucleolus of the permeabilized oocytes.
When B10B was added at the onset of the run-on reaction, BrU
incorporation into the nucleolus was abolished completely
(Fig. 8A). This effect was observed at a B10B concentration
of 35 µg/ml and 11 µg/ml, but not at 1 µg/ml, revealing a
concentration dependency. At 35 µg/ml, a recombinant protein
containing a pair of C2H2 zinc fingers derived from the
Drosophila Krüppel protein (Stanojevic et al., 1989), failed to
produce the same effect (Fig. 8B). B10B binds to a GC-rich
sequence 5′-(G/C)G(C/T)G(A/T)C (Iuchi and Green, 1999;
Tseng et al., 1999; see Figs 5, 6), whereas the Krüppel protein
binds to an AT-rich sequence 5′-AAGGGTTAA (Pankratz et
al., 1989; Treisman and Desplan, 1989), suggesting that the
inhibitory effect of the basonuclin zinc fingers is related to their
DNA-binding sequence specificity. To investigate if the effect
of B10B is dependent on the presence of basonuclin, we
repeated the experiment in mouse fibroblasts, which express
UBF but not basonuclin. Mouse fibroblasts treated with Triton
X-100 showed the same ability to incorporate BrU into their
nucleoli and similar permeability to immunoglobulin (not
shown). But when B10B was added at the concentration that
blocked pol I transcription in oocytes (of 35 µg/ml and 11
µg/ml), BrU incorporation was still detected, but at a reduced
level (Fig. 8C) compared with the oocytes exposed to the zinc
fingers of Krüppel (Fig. 8D).
DISCUSSION
Previous studies have suggested that, unlike amphibians, rDNA
amplification may not be employed to enhance rDNA
transcription during mouse oocyte development (Kaplan et al.,
1982). However, to our knowledge, direct measurement of
rDNA copy number in the mouse oocyte by PCR has not been
reported. Here, we have compared the copy number of rDNA
in mouse oocytes and somatic tissues. Under our PCR
conditions, which could easily detect an amplification of rDNA
of less than tenfold, we failed to detect any amplification of
oocyte rDNA at the promoter and the coding region of the
gene. Our run-on assay also failed to label any ‘mini-nucleoli’,
which are present in their hundreds in Xenopus oocytes (Brown
and Dawid, 1968). These observations therefore support the
notion that no rDNA amplification occurs prior to the growth
phase of mouse oocyte development. Mouse, and perhaps
mammals in general, may rely on alternative mechanisms to
enhance rDNA transcription during oocyte development.
The intranuclear distribution of basonuclin is intriguing. In
the growing oocyte, its uniform distribution between the
nucleoplasm and nucleolus clearly differs from the exclusive
nucleolar distribution of UBF, a bona fide pol I transcription
Basonuclin in mouse oogenesis
413
Fig. 6. A comparison of B10B-binding sites on human and mouse rDNA promoter. The DNA sequences of the promoter region of the human
(top) and the mouse (bottom) rDNA are aligned according to the transcription start site (arrow). The cis-elements are shown in gray. The
footprints on the noncoding and the coding strands are indicated, respectively, by black bars above and below the DNA sequence. The
corresponding binding sites on human and mouse rDNA promoters are indicated by A, B and C. Note the similarity of the binding sites in
relation to the cis-elements on each promoter.
factor. This also differs from the intranuclear distribution of
basonuclin in the basal keratinocytes of human plantar
epidermis (Tseng and Green, 1994) and in the neonatal mouse
spermatogonia (Mahoney et al., 1988), in which basonuclin
appears to concentrate within certain subnuclear structures,
presumably nucleoli. In cultured human keratinocytes,
although basonuclin was detected in the nucleolus, its presence
was predominantly nucleoplasmic (Tseng et al., 1999). Here,
by combining a run-on assay and immunocytochemistry, we
show clearly that in the permeabilized oocytes, basonuclin is
present in the nucleolus where active pol I transcription can be
simultaneously detected.
We and others have shown that the recombinant basonuclin
zinc fingers bind to the promoter region of the human rDNA.
The locations of the binding sites detected by our group and
by Iuchi and Green agree well and both groups described
Fig. 7. Basonuclin enhances transcription from an rDNA promoter in a co-transfection assay. (A) The reporter construct, pHrCAT. The position
of the three binding sites of B10B are indicated by hatched bars and labeled by capital italics. Transcription initiation site within the rDNA
promoter is shown by an arrow. The location of the primer (HuCAT-PE) used in the primer extension assay is indicated by a black bar. (B) The
basonuclin expression constructs. The coding region of the wild-type human basonuclin cDNA clone was placed under the transcriptional
control of the cytomegalovirus promoter (CMV Pr.) in the pcDNA3.1 vector (pFBwt) and a C-terminal deletion mutant (pFBd). The extent of
the sequence of B10B is indicated by a black bar below the wild-type cDNA. (C) Reporter plasmid prHuCAT was co-transfected into COS-7
cells with the empty expression vector pcDNA3 (lane 1 and 2), with basonuclin expression plasmid pFBwt (lane 3 and 4) and with pFBd.
Transcription from the reporter was monitored by primer extension. Correct initiation from the rRNA transcription start site yields a primer
extension product of 123 bases. Lane M, a molecular weight marker; numbers on the left are size in bases. The value below each lane indicates
the relative amount of radioactivity in the primer extension product, with lane 1 arbitrarily defined as 1.00. (D) A northern analysis of β-actin
serves as a control of the amount of RNA used in the primer extension assay. nls, nuclear localization signal; Ser, serine stripe; Z12, zinc fingers
1 and 2.
414
Q. Tian and others
Fig. 8. Basonuclin zinc fingers (B10B) block pol I transcription in
mouse oocytes. (A) Run-on reaction in the presence of B10B at 11
µg/ml, no BrU incorporation was detected (arrowheads). (B) Run-on
in the presence of the zinc fingers of Krüppel at 35 µg/ml; BrU was
incorporated (arrows). Three oocytes are shown in both A and B.
(C,D) Run-on reaction in Triton X-100-treated mouse fibroblasts, in
the presence of B10B at 11 µg/ml (C, weaker but clear incorporation
of BrU was detected, arrowheads) and in the presence of the zinc
fingers of Krüppel at 35 µg/ml (D, the incorporation of BrU was not
affected, arrows).
similar target sequences for the basonuclin zinc fingers (Tseng
et al., 1999, Iuchi and Green, 1999). These binding sites could
be related to rDNA transcription or they could be coincidental.
Since mouse and human rDNA promoters share very little
overall DNA sequence homology, if the binding sites in the
human promoter are coincidental, they would not occur at the
same location in the mouse promoter. Our result demonstrates
that basonuclin-binding sites on the human and mouse rDNA
promoter are remarkably similar in location as well as in
organization, which strongly disputes the coincidental nature
of the binding sites. Moreover, the evolutionary conservation
of this homology between human and mouse suggests that
there is functional interaction of basonuclin and rDNA
promoter.
The tissue and cellular distribution of basonuclin suggests
that it is a positive regulator of rRNA transcription. Basonuclin
is expressed in tissues and cells that are highly proliferative,
e.g. in basal epidermal and hair follicular keratinocytes and in
spermatogonia of the seminiferous tubule, or in cells with
active protein synthesis, e.g. in the developing oocyte. Here,
we have shown that basonuclin enhances transcription from an
rDNA promoter in a co-transfection assay in COS cells. We
detected a small but clear positive effect, which is dependent
on the integrity of the basonuclin-coding sequence. It is not
clear, however, whether the magnitude of this positive effect in
COS cells reflects what might be occuring in the oocyte,
because of the apparent difference between the two cell types
(i.e. since the action of basonuclin is likely to be cell-type
specific, other oocyte factors may be required for the full effect
of basonuclin in the oocyte); neither do we know the exact
amount of increase in rRNA transcription that a growing mouse
oocyte requires (e.g. a doubling of the rate of synthesis may be
all that is required). The co-transfection experiment was
performed in COS cells because they do not express basonuclin
– any effect of the protein can therefore be detected. It is
difficult to use the run-on assay to measure directly the
magnitude of the effect of exogenous basonuclin (e.g.
introduced by microinjection) in the oocyte because the rRNA
transcription may already be at a maximum in the presence of
the endogenous basonuclin. In any case, the result of the cotransfection experiment is in agreement with the dominantnegative effect of basonuclin zinc fingers (B10B) on oocyte
rRNA transcription, which also suggests that basonuclin is a
positive regulator. The dominant-negative effect of B10B on
pol I transcription may occur through two mechanisms,
namely, its interference with basonuclin function, and/or with
other components of the pol I transcriptional apparatus, such
as UBF, whose binding site overlaps that of basonuclin-binding
site A. Similarly, this effect could be exerted at the step of
transcription initiation or elongation. The differential effect of
B10B on pol I transcription of oocytes and fibroblasts suggests
that the larger inhibitory effect of B10B in oocytes cannot be
explained by the transcriptional process common to both cell
types.
Our results show that basonuclin differs from UBF, a
dedicated pol I transcription factor, in at least two aspects.
First, basonuclin has a restricted tissue and cell distribution
(Tseng and Green, 1994; Yang et al., 1997; Mahoney et al.,
1998). UBF, on the other hand, has a much wider distribution
and is thought to be ubiquitous. Second, the intranuclear
distribution of UBF is exclusively in the nucleolus, whereas
basonuclin is found both in the nucleoplasm and nucleolus.
The nucleoplasmic localization of basonuclin is consistent
with its ability to interact with promoters other than that of
rDNA (Tseng et al., 1999). These observations suggest that
basonuclin may have target genes other than rDNA. In this
context, it is of interest to note that the polymerase I-andtranscript-release factor (PTRF), first identified as a poI Iassociated factor (Jansa et al., 1998; Jansa and Grummt, 1999),
was recently described to have a restricted tissue distribution
and is thought to involve also in the transcriptional regulation
of two type-I collagen genes in mouse fibroblasts (Hasegawa
et al., 2000).
Another issue regarding the relationship of basonuclin and
UBF is how they interact on the rDNA promoter. We showed
that they co-exist in the oocyte nucleolus (Fig. 4) and on the
mitotic chromosomes of keratinocytes (Tseng et al., 1999).
Moreover, the DNase I footprints of basonuclin zinc fingers
and UBF overlap on the rDNA promoter (Iushi and Green,
1999; Tseng et al., 1999). These observations raise the
question of whether basonuclin replaces UBF in oocytes
and keratinocytes, or whether they work cooperatively. Any
speculation about their interaction must also take into account
that the binding of UBF to DNA is not entirely sequence
specific, but may also depend on the secondary structures
of the DNA (McStay et al., 1991; O’Mahony et al., 1992;
Kuhn et al., 1994). Since this question may hold a key to
understanding how rRNA transcription is regulated by a cell
type-specific factor, further investigation is warranted.
Because of the importance of rRNA synthesis to cellular
Basonuclin in mouse oogenesis
function, it is tightly regulated during the cell cycle, cellular
proliferation and differentiation. It has been shown that the
mammalian basal pol I transcriptional regulators are targets of
phosphorylation, which activates their ability to assist pol I
transcription (Voit et al., 1995). Pol I transcription can be
downregulated by the tumor suppressors p53 (Budde and
Grummt, 1999) and Rb protein (Cavanaugh et al., 1995a; Voit
et al., 1997), and by DNA-activated protein kinase (Kuhn et
al., 1995), or it can be target of viral oncogenes such as SV40
large T antigen (Zhai et al., 1997; Zhai and Comai, 1999) and
Ku-related proteins (Hoff et al., 1994), which activate rRNA
transcription. Because the basal pol I transcription factors are
thought to be ubiquitous, these regulatory mechanisms should
be functional in all cell types. The existence of basonuclin
suggests that in addition to the above described mechanisms,
some cell types may employ their specific transcription factors
to regulate rRNA transcription.
A unique problem faced by oocytes is their large cytoplasmto-nucleus volume ratio, which increases dramatically during
the growth phase. It is known that the ribosome density per
volume of cytoplasm in developing mouse oocyte is
comparable with that in somatic cells, which have a much
smaller cytoplasm-to-nucleus volume ratio. This means that at
transcriptional level, oocyte must use a similar number of
genes (4C of oocyte versus 2C at G1 and 4C at G2 of the
somatic cells) to produce ribosomes to fill a much larger
volume of cytoplasm. In the Xenopus oocyte, one solution is
rDNA amplification. In the smaller mouse oocyte, where the
demand for ribosomes may not be as large as in the Xenopus
oocyte, the problem remains. Here, we have shown that no
rDNA amplification is detected in the mouse oocyte, instead,
these cells may use basonuclin to enhance rRNA synthesis. If
this is indeed the case, then inhibition or targeted disruption of
the function of basonuclin should interrupt mouse oocyte
development.
We thank Edward K. L. Chan and Ingrid Grummt for the
autoantibodies against UBF, Lucio Comai for the prHuCAT reporter
construct and the primer extension protocol, Frank Rauscher III for
the recombinant Krüppel zinc fingers, Carmen Williams for mouse
oocytes, and Richard Schultz for critical reading of the manuscript.
This work was supported by grants from NIH (AG14456 to H. T. and
HD22732 and 06274 to G. S. K.).
REFERENCES
Brown, D. D. and Dawid, I. B. (1968). Specific gene amplification in oocytes.
Oocyte nuclei contain extrachromosomal replicas of the genes for ribosomal
RNA. Science 160, 272-280.
Budde, A. and Grummt, I. (1999). p53 represses ribosomal gene
transcription. Oncogene 18, 1119-1124.
Cavanaugh, A. H., Hempel, W. M., Taylor, L. J., Rogalsky, V., Todorov, G.
and Rothblum, L. I. (1995a). Activity of RNA polymerase I transcription
factor UBF blocked by Rb gene product. Nature 374, 177-180.
Chan, E. K., Imai, H., Hamel, J. C. and Tan, E. M. (1991). Human
autoantibody to RNA polymerase I transcription factor hUBF. Molecular
identity of nucleolus organizer region autoantigen NOR-90 and
ribosomal RNA transcription upstream binding factor. J. Exp. Med. 174,
1239-1244.
Dawid, I. B. (1965). Deoxyribonucleic acid in amphibian eggs. J. Mol. Biol.
12, 581-599.
Grummt, I. (1999). Regulation of mammalian ribosomal gene transcription
by RNA polymerase I. Prog. Nucleic Acid Res. Mol. Biol. 62, 109-154.
Hasegawa, T., Takeuchi, A., Miyaishi, O., Xiao, H., Mao, J. and Isobe, K.
415
(2000). PTRF (polymerase I and transcript-release factor) is tissue-specific
and interacts with the BFCOL1 (binding factor of a type-I collagen
promoter) zinc-finger transcription factor which binds to the two mouse
type-I collagen gene promoters. Biochem. J 347, 55-59.
Hoff, C. M., Ghosh, A. K., Prabhakar, B. S. and Jacob, S. T. (1994).
Enhancer 1 binding factor, a Ku-related protein, is a positive regulator of
RNA polymerase I transcription initiation. Proc. Natl. Acad. Sci. USA 91,
762-766.
Iuchi, S. and Green, H. (1999). Basonuclin, a zinc finger protein of
keratinocytes and reproductive germ cells, binds to the rRNA gene promoter.
Proc. Natl. Acad. Sci. USA 96, 9628-9632.
Jansa, P. and Grummt, I. (1999). Mechanism of transcription termination:
PTRF interacts with the largest subunit of RNA polymerase I and dissociates
paused transcription complexes from yeast and mouse. Mol. Gen. Genet.
262, 508-514.
Jansa, P., Mason, S. W., Hoffmann-Rohrer, U. and Grummt, I. (1998).
Cloning and functional characterization of PTRF, a novel protein which
induces dissociation of paused ternary transcription complexes. EMBO J.
17, 2855-2864.
Kaplan, G., Abreu, S. L. and Bachvarova, R. (1982). rRNA accumulation
and protein synthetic patterns in growing mouse oocytes. J. Exp Zoo 220,
220-226.
Kopecny, V., Landa, V. and Pavlok, A. (1995). Localization of nucleic acids
in the nucleoli of oocytes and early embryos of mouse and hamster: an
autoradiographic study. Mol. Reprod. Dev. 41, 449-458.
Kuhn, A., Stefanovsky, V. and Grummt, I. (1993). The nucleolar
transcription activator UBF relieves Ku antigen-mediated repression of
mouse ribosomal gene transcription. Nucleic Acids Res. 21, 2057-2063.
Kuhn, A., Voit, R., Stefanovsky, V., Evers, R., Bianchi, M. and Grummt,
I. (1994). Functional differences between the two splice variants of the
nucleolar transcription factor UBF: the second HMG box determines
specificity of DNA binding and transcriptional activity. EMBO J. 13, 416424.
Kuhn, A., Gottlieb, T. M., Jackson, S. P. and Grummt, I. (1995). DNAdependent protein kinase: a potent inhibitor of transcription by RNA
polymerase I. Genes Dev. 9, 193-203.
Laskey, R. A. (1974). Biochemical processes in early development. In
Companion to Biochemistry. Vol. 2 (ed. A. T. Bull, J. R. Lagnado, J. O.
Thomas and K. F. Tipton), pp. 137-160. London: Longman.
Mahoney, M. G., Tang, W., Xiang, M. M., Moss, S. B., Gerton, G. L.,
Stanley, J. R. and Tseng, H. (1998). Translocation of the zinc finger protein
basonuclin from the mouse germ cell nucleus to the midpiece of the
spermatozoon during spermiogenesis. Biol. Reprod. 59, 388-394.
Masson, C., Fomproix, N., Szöllösi, M. S., Debey, P. and HernandezVerdun, D. (1996). Conditions favoring RNA polymerase I transcription in
permeabilized cells. Exp. Cell Res. 226, 114-125.
Matsuzaki, K., Iuchi, S. and Green, H. (1997). Conservation of human and
mouse basonuclins as a guide to important features of the protein. Gene 195,
87-92.
Mattson, B. A. and Albertini, D. F. (1990). Oogenesis: Chromatin and
microtubule dynamics during meiotic prophase. Mol. Reprod. Dev. 25, 374383
McStay, B., Frazier, M. W. and Reeder, R. H. (1991). xUBF contains a novel
dimerization domain essential for RNA polymerase I transcription. Genes
Dev. 5, 1957-1968.
Moore, G. P. M. and Lintern-Moore, S. (1978). Transcription of the mouse
oocyte genome. Biol. Reprod. 18, 865-870.
Niu, H. and Jacob, S. T. (1994). Enhancer 1 binding factor (E1BF), a Kurelated protein, is a growth-regulated RNA polymerase I transcription factor:
association of a repressor activity with purified E1BF from serum-deprived
cells. Proc. Natl. Acad. Sci. USA 91, 9101-9105.
O’Mahony, D. J., Smith, S. D., Xie, W. and Rothblum, L. I. (1992). Analysis
of the phosphorylation, DNA-binding and dimerization properties of the
RNA polymerase I transcription factors UBF1 and UBF2. Nucleic Acids
Res. 20, 1301-1308.
Pankratz, M. J., Hoch, M., Seifert, E. and Jackle, H. (1989). Krüppel
requirement for knirps enhancement reflects overlapping gap gene activities
in the Drosophila embryo. Nature 341, 337-340.
Penman, S., Smith, I., Holtzman, E. and Greenberg, H. (1966). RNA
metabolism in the HeLa cell nucleus and nucleolus. Natl. Cancer Inst.
Monogr. 23, 489-512.
Perkowska, E., Macgregor, H. C. and Birnstiel, M. L. (1968). Gene
amplification in the oocyte nucleus of mutant and wild-type Xenopus laevis.
Nature 217, 649-650.
416
Q. Tian and others
Schultz, R. M. (1993). Regulation of zygotic gene activation in the mouse.
BioEssays 15, 531-538.
Schultz, R. M., LaMarca, M. J. and Wassarman, P. M. (1978). Absolute
rates of protein synthesis during meiotic maturation of mammalian oocytes
in vitro. Proc. Natl. Acad. Sci. USA 75, 4160-4164.
Stanojevic, D., Hoey, T. and Levine, M. (1989). Sequence-specific DNAbinding activities of the gap proteins encoded by hunchback and Krüppel in
Drosophila. Nature 341, 331-335.
Sternlight, A. L. and Schultz, R. M. (1981). Biochemical studies of
mammalian oogenesis: kinetics of accumulation of total and poly(A)containing RNA during growth of the mouse oocyte. J. Exp. Zool. 215, 191200
Treisman, J. and Desplan, C. (1989). The products of the Drosophila gap
genes hunchback and Kruppel bind to the hunchback promoters. Nature 341,
335-337.
Tseng, H., Beigel, J. A. and Brown, R. S. (1999). Basonuclin is associated
with the ribosomal RNA genes on human keratinocyte mitotic
chromosomes. J. Cell Sci. 112, 3039-3047.
Tseng, H. and Green, H. (1988). Remodeling of the involucrin gene during
primate evolution. Cell 54, 491-496
Tseng, H. and Green, H. (1992). Basonuclin: a keratinocyte protein
with multiple paired zinc fingers. Proc. Natl. Acad. Sci. USA 89, 1031110315.
Tseng, H. and Green, H. (1994). Association of basonuclin with ability of
keratinocytes to multiply and with absence of terminal differentiation. J.
Cell Biol. 126, 495-506.
Voit, R., Kuhn, A., Sander, E. E. and Grummt, I. (1995). Activation of
mammalian ribosomal gene transcription requires phosphorylation of the
nucleolar transcription factor UBF. Nucleic Acids Res. 23, 2593-2599.
Voit, R., Schafer, K. and Grummt, I. (1997). Mechanism of repression of
RNA polymerase I transcription by the retinoblastoma protein. Mol. Cell.
Biol. 17, 4230-4237.
Yang, Z. H., Gallicano, G. I., Yu, Q.-C. and Fuchs, E. (1997). An unexpected
localization of basonuclin in the centrosome, mitochondria and acrosome of
developing spermatids. J. Cell Biol. 137, 657-670.
Zhai, W., Tuan, J. and Comai, L. (1997). SV40 large T antigen binds to the
TBP-TAFI complex SL-1 and coactivates ribosomal RNA transcription.
Genes Dev. 11, 1605-1617.
Zhai, W. and Comai, L. (1999). A kinase activity associated with simian virus
40 T antigen phosphorylates upstream binding factor (UBF) and promotes
formation of a stable initiation complex between UBF and SL1. Mol. Cell.
Biol. 19, 2791-2802.