Maternally recruited Aurora C kinase is more stable
than Aurora B to support mouse oocyte maturation
and early development
Karen Schindler1,2, Olga Davydenko2, Brianna Fram, Michael A. Lampson3, and Richard M. Schultz3
Department of Biology, University of Pennsylvania, Philadelphia, PA 19104
Edited* by John J. Eppig, The Jackson Laboratory, Bar Harbor, ME, and approved June 14, 2012 (received for review December 14, 2011)
Aurora kinases are highly conserved, essential regulators of cell
division. Two Aurora kinase isoforms, A and B (AURKA and
AURKB), are expressed ubiquitously in mammals, whereas a third
isoform, Aurora C (AURKC), is largely restricted to germ cells.
Because AURKC is very similar to AURKB, based on sequence and
functional analyses, why germ cells express AURKC is unclear. We
report that Aurkc−/−females are subfertile, and that AURKB func-
tion declines as development progresses based on increasing se-
verity of cytokinesis failure and arrested embryonic development.
Furthermore, we find that neither Aurkb nor Aurkc is expressed
after the one-cell stage, and that AURKC is more stable during
maturation than AURKB using fluorescently tagged reporter
proteins. In addition, Aurkc mRNA is recruited during maturation.
Because maturation occurs in the absence of transcription, post-
transcriptional regulation of Aurkc mRNA, coupled with the
greater stability of AURKC protein, provides a means to ensure
sufficient Aurora kinase activity, despite loss of AURKB, to support
both meiotic and early embryonic cell divisions. These findings
suggest a model for the presence of AURKC in oocytes: that
AURKC compensates for loss of AURKB through differences in
both message recruitment and protein stability.
Aurora kinases in mammals: Aurora kinases A and B (AURKA
or -B) are ubiquitously expressed and their functions have been
extensively studied, whereas AURKC is largely limited to germ
cells (1–3); many human cancer cell lines express AURKC (4)
and some somatic tissues express AURKC at low levels (5–7). It
is not clear, however, why germ cells require a third AURK.
Because isoforms can have different functions, it is tempting to
speculate that AURKC exists because its mitotic counterparts
simply cannot execute unique features of meiosis.
One unique feature of meiosis is the generation of haploid
gametes from diploid precursor cells by a reductional chromo-
some segregation during meiosis I (MI) followed by an equa-
tional division at meiosis II (MII) without an intervening round
of DNA replication. In oocytes, another unique feature is that
meiosis is not a continuous process because there is a growth
period during a prolonged arrest at prophase I, followed by a cell
division cycle during oocyte maturation, and a second arrest at
metaphase of MII, until fertilization, which triggers completion
of MII. Furthermore, proteins in the oocyte must support the first
mitotic cell cycles of the embryo before zygotic genome activation.
Despite these obvious differences, several observations suggest
that AURKC may not have a specialized function. AURKB and
AURKC are highly similar in sequence (61% identical), and
AURKC can functionally compensate for loss of AURKB when
ectopically expressed in somatic cells (8, 9). Furthermore, em-
bryos that lack AURKB can develop to but not beyond the
blastocyst, as long as AURKC is present, consistent with the idea
that AURKB and -C have similar functions (10).
Given the sequence similarity and apparent redundant func-
tion, it is unclear why germ cells have a third AURK. Male mice
lacking AURKC are subfertile because of postmeiotic defects,
urora kinases are highly conserved cell-cycle regulators with
essential roles in chromosome segregation. There are three
including abnormally condensed chromatin and abnormally
shaped acrosomes, but females were not examined (11). Muta-
tions in human AURKC cause meiotic arrest and formation of
tetraploid sperm (12), suggesting an essential role in cytokinesis
in male meiosis. Experiments in mouse oocytes using a chemical
inhibitor of AURKB (ZM447439) do not address the function of
AURKB because AURKC is also inhibited (13–17). Strategies
using dominant-negative versions of AURKC are also difficult to
interpret, because the mutant may also compete with AURKB
(18). Overexpression studies have similar limitations because
both kinases interact with inner centromere protein (INCENP)
and these studies did not report expression levels of AURKB
versus AURKC (19). Therefore, no experiments to date have
directly addressed why oocytes contain a third AURK.
A hint as to the need for oocytes to express AURKC comes
from comparisons of AURKB and AURKC sequences. AURKB
contains N-terminal destruction motifs that AURKC lacks. In
mitotic cell cycles, these motifs regulate AURKB destruction by
the anaphase-promoting complex/cyclosome (APC/C) at cytoki-
nesis, before G1 of the following cell cycle (20, 21). The CDH1
(FZR1 in mouse) regulator of the APC/C binds the KEN box
(amino acids 4–9 in mouse) and both CDH1 and CDC20 bind the
A-box (amino acids 26–29 in mouse). AURKB also contains four
putative D-boxes, which AURKC also contains, the functions of
which in regulating its stability are unclear (20, 21). Because there
are two rounds of chromosome segregation without an interven-
ing cell cycle in meiosis, if AURKB is degraded after MI (as it is
during mitosis), there may be no opportunity to regenerate addi-
tional AURKB to support MII. Based on these sequence com-
parisons, we hypothesized that oocytes contain a third AURK
because of differential regulation of AURKC protein levels rela-
tive to AURKB.
We demonstrate that female mice lacking AURKC are sub-
fertile because of phenotypes that begin with mild chromosome
misalignment causing arrest at MI and increase in severity during
embryogenesis with cytokinesis failure. The progression of these
phenotypes indicates a gradual loss of AURK activity during
oocyte maturation and early development. Consistent with this
model, we find that AURKB protein is less stable than AURKC
during meiotic maturation. Moreover, we find that the Aurkc
Author contributions: K.S., O.D., M.A.L., and R.M.S. designed research; K.S., O.D., and B.F.
performed research; K.S., O.D., M.A.L., and R.M.S. analyzed data; and K.S., M.A.L., and
R.M.S. wrote the paper.
The authors declare no conflict of interest.
*This Direct Submission article had a prearranged editor.
1Present address: Department of Genetics, Rutgers, State University of New Jersey, Piscat-
away, NJ 08854.
2K.S. and O.D. contributed equally to this work.
3To whom correspondence may be addressed. E-mail: firstname.lastname@example.org or
See Author Summary on page 13150 (volume 109, number 33).
This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.
| Published online July 9, 2012
message is recruited for translation during oocyte maturation to
ensure sufficient AURK activity during meiosis and embryonic
development. Expression of AURKB in Aurkc−/−oocytes and
embryos rescues the meiotic and cytokinesis defects, respectively,
consistent with the hypothesis that AURKB can compensate for
the loss of AURKC. Taking these data together, we propose that
AURKC is an example of an isoform of a cell-cycle regulator that
is recruited during meiotic maturation to support meiosis, fertil-
ization and early embryonic cell division.
AURKC Is Required for Normal Oocyte Maturation and Embryo
Development. AURKC expression is largely limited to germ
cells, yet a role during oocyte meiotic maturation is not clear. We
rederived cryopreserved Aurkc+/−embryos to generate mice
lacking Aurkc (11) to determine the requirement of AURKC in
female gametes. Oocytes from knockout mice do not express
detectable Aurkc message or AURKC protein; the amount of
Aurkb mRNA is unaffected in knockout mice (Fig. S1). The
number of full-grown oocytes from superovulated, sexually ma-
ture female Aurkc−/−mice (34.6 ± 3.8, n = 10) was not statisti-
cally different from the numbers from control littermates (44.67
± 6.2, n = 9). Nevertheless, our breeding trials revealed that they
were subfertile, averaging two fewer pups per litter (Fig. 1A), but
with a similar number of days between litters compared with
control littermates (Fig. S2).
To determine if AURKC is required for oocyte meiotic mat-
uration, we first matured oocytes to metaphase of MI and, after
fixation and staining, examined chromosome alignment. Al-
though the overall percentage of MI oocytes from knockout mice
with chromosome misalignment was not strikingly different from
wild-type controls, we observed a greater variation from mouse-
to-mouse in oocytes from knockout females (Fig. 1B). Next, we
monitored the time of polar body emission and found that sig-
nificantly more oocytes from AURKC knockout mice arrested in
MI, as indicated by failure to extrude a polar body (Fig. 1C).
Furthermore, the oocytes that failed to extrude a polar body did
not undergo cytokinesis, whereas those oocytes that did extrude
a polar body were delayed by 1 h entering anaphase I compared
with controls (Fig. 1D). These oocytes were then fixed and an-
alyzed by immunocytochemistry, and a significant portion of
those from Aurkc−/−mice displayed abnormal chromosome
alignment regardless of whether they arrested at MI or pro-
gressed to MII (Fig. 1 D–F). We did not observe, however, any
increase in aneuploidy incidence (1 of 35 aneuploid eggs in
knockouts and 0 of 12 aneuploid eggs in controls). This finding is
consistent with oocytes containing misaligned chromosomes ar-
resting at metaphase of MI.
it is likely that AURKC also functions during early embryonic cell
divisions. To test this hypothesis, we isolated one-cell embryos
from wild-type, heterozygous, or knockout females that were
mated to wild-type males and allowed the embryos to develop in
vitro. Significantly fewer embryos from Aurkc−/−females cleaved
numbers of one-cell embryos attempted and failed to complete
membrane ruffling or blebbing (Fig. S3). Cytokinesis failure is
a well-established phenotype of AURKB inhibition or mutation
in MI vs. the one-cell stage in oocytes and embryos from Aurkc−/−
the one-cell stage, indicating that AURKB activity declines as
development progresses (Fig. 2D).
over an 8-mo breeding trial is shown. (B) GV-intact oocytes were isolated from mice of the indicated genotype and matured in vitro for 8 h before fixation at
MI. Percent of oocytes with misaligned chromosomes was plotted for each mouse analyzed. (C–F) GV-intact oocytes were isolated from mice of the indicated
genotype and matured in vitro to determine incidence (B) and timing (C) of polar body extrusion (PBE). Cells were fixed when controls (WT) had reached MII
and processed for immunocytochemistry to detect chromosomes and spindles. The percentage of oocytes that contained abnormal chromosome config-
urations at either MI or MII was determined (E), and representative images are shown (F). Graphs represent mean (± SEM) from at least 30 oocytes from three
independent experiments. (Scale bars, 5 μm.) One-way ANOVA was used to analyze the data in B–D. *P < 0.05, **P < 0.01, ***P < 0.001; WT, wild-type; HET,
heterozygous; KO, knockout.
Loss of AURKC leads to meiotic abnormalities. (A) Results of fertility trials. The average number of pups born to females of the indicated genotype
| www.pnas.org/cgi/doi/10.1073/pnas.1120517109 Schindler et al.
Consistent with a gradual decline in AURKB activity, de-
velopment of embryos from Aurkc−/−mothers became pro-
gressively worse with extended culture periods. Sixty-four hours
after mating, ∼35% of the control embryos developed to the
eight-cell embryonic stage compared with only ∼15% of the
embryos from the knockout mothers (Fig. 2E). Some of the latter
embryos were fragmented, with many arrested at the two-cell
stage. At 113 h, when nearly 75% of control embryos developed
to the blastocyst stage, only ∼25% of embryos from the knockout
mothers were blastocysts (Fig. 2F). Those embryos from the
knockout mothers that did develop to the blastocyst stage
appeared indistinguishable from control embryos. At this time
point, most of the embryos that had lagged behind in de-
velopment had fragmented. These findings indicate that the
subfertility of Aurkc−/−females was because of perturbations that
begin during meiotic maturation (chromosome misalignment
and cell-cycle arrest) and become more pronounced during
embryonic development (failure in cytokinesis).
AURKC Is More Stable than AURKB. Mice lacking AURKB develop
up to the blastocyst stage of embryonic development because
AURKC is sufficient for supporting the early mitotic divisions
(10). In contrast, AURKB function declines between MI and the
one-cell stage, as indicated by increasing cytokinesis failure (Fig.
2D), and AURKB does not support development to the blasto-
cyst stage (Fig. 2 E and F). AURKB is unstable during mitosis
because it is targeted for ubiquitin-mediated proteolysis after M
phase during each round of the cell cycle (20, 21). AURKC,
however, lacks destruction motifs that AURKB contains, sug-
gesting that it could be more stable. Moreover, quantitative real-
time PCR demonstrated that although oocytes (meiotically in-
competent and full-grown oocytes) and eggs (MII) contained
both Aurkb and Aurkc mRNA, as previously described (13, 14,
18), these messages declined by the two-cell stage (Fig. 2 G and
H). Taken together, these data support the model that the
amount of AURKB declines during meiosis and early embry-
onic mitosis and cannot be resynthesized until the blastocyst
stage, whereas the amount of AURKC is maintained during
To test whether AURKB and AURKC are differentially sta-
ble, we used a live-cell imaging assay. We comicroinjected
Aurkb-mCherry and Aurkc-Gfp cRNAs into germinal vesicle
(GV) oocytes, then inhibited translation by adding cyclohexi-
mide, and measured mCherry and GFP intensities over time as
reporters for the abundance of AURKB and -C. Because long-
term incubation in cycloheximide inhibits major meiotic cell-cy-
embryos were isolated and cultured in vitro. (A) The percentage of one-cell embryos that cleaved to the two-cell embryonic stage. (B–D) Embryos were
imaged live by DIC every 5–7 min. In B, the percentage of embryos with abnormal cytokinesis was plotted for each mouse. C contains representative images of
a wild-type embryo undergoing normal cytokinesis (Upper) and a KO embryo failing cytokinesis (Lower). The time stamp is h:min after hCG injection. (Scale
bar, 5 μm.) (D) The percentage of normal cytokinesis events from Fig. 1C (MI) and panel B (1C) were compared. (E and F) Embryo development was monitored
at the indicated times after hCG and mating. These experiments were conducted four times with at least two mice per genotype each time, and the data are
expressed as the mean ± SEM. (G and H) mRNA levels were measured by quantitative RT-PCR at the indicated stages. Data were normalized against a probe
that detects exogenously added Gfp message. Mean (± SEM) from three independent experiments are shown. One-way ANOVA was used to analyze the data.
*P < 0.05; **P < 0.01; ***P < 0.001. BL, blastocyst; frag, fragmented; GV, full grown GV-intact oocyte; Inc, meiotically incompetent oocyte; Mor, morula; 1C,
one-cell embryo; 2C, two-cell embryo; 4C, four-cell embryo; 8C, eight-cell embryo.
Embryonic development is compromised in Aurkc−/−mice. (A–F) Female mice of the indicated genotype were mated to wild-type males and one-cell
Schindler et al.PNAS
| Published online July 9, 2012
cle transitions (26, 27), we monitored the destruction of AURKB
and C during three shorter intervals: from GV to MI (GV-MI),
from MI to MII (MI-MII), and in activated eggs that are exiting
the MII arrest. Although both AURKC-GFP and AURKB-
mCherry are progressively degraded (Fig. 3 A–C), AURKC-GFP
is more stable than AURKB-mCherry in each interval. Overall,
∼30% of AURKC-GFP was degraded between the GV stage and
the two-cell embryonic stage compared with ∼70% of AURKB-
mCherry (Fig. 3D). To confirm that degradation was not specific
to the fluorophore, we compared AURKB-GFP to AURKB-
mCherry and found identical degradation kinetics (Fig. S4A).
As a photo-bleaching control, nondegradable cyclin B-GFP
fluorescence was constant under identical conditions (Fig.
S4B). Taken together, these results show that AURKB and -C
gradually decline during meiotic maturation but AURKC is
There are conflicting reports regarding whether oocytes contain
AURKB. Some studies have reported localization of endogenous
AURKB and exogenous AURKB-GFP to centromeres and mid-
bodies in mouse oocytes (17), whereas others have failed to do
so (18). This discrepancy may be because of differences in the
abundance of AURKB in the different mouse strains used cou-
pled with poor reagents for detection. Aurkb and Aurkc mRNA
are detectable by quantitative real-time PCR, and immunocyto-
chemistry on maturing oocytes and developing embryos showed
that AURKC localized to metaphase chromatin, and was un-
detectable by the blastocyst stage (Fig. S5). Even though Aurkc
message was virtually absent in the early embryonic stages (Fig. 2
G and H), the protein still remained in the embryo (Fig. S5). We
failed to reliably detect AURKB protein with commercially
available reagents, but given the presence of Aurkb message and
that Aurkc−/−mice are viable, it is likely that oocytes and early
embryos contain AURKB protein. Furthermore, immunocyto-
chemistry of human oocytes also demonstrates that both AURKB
and AURKC are present (28).
AURKB Degradation Does Not Depend on Its N Terminus.Because the
N terminus of AURKB contains destruction motifs (Fig. 4A) that
indicated stage. In each graph, the first time point is 1 h after cycloheximide addition, and fluorescent images were obtained at the indicated times. Below are
representative images from each time course. Data represent mean (± SEM) from at least 30 oocytes from two independent experiments. (Scale bars, 5 μm.)
(D) Merge of the data from A–C.
AURKC is more stable than AURKB during meiotic maturation. (A–C) GV-intact oocytes were coinjected with the indicated cRNAs and matured to the
| www.pnas.org/cgi/doi/10.1073/pnas.1120517109Schindler et al.
are recognized by the APC/C during mitosis, we asked if
AURKB and AURKC degradation is proteasome-dependent.
Treatment with the proteasome inhibitor MG132 reduced deg-
radation of both AURKs (Fig. 4B and Fig. S6). To test whether
the N-terminal motifs in AURKB could explain the difference in
kinetics of proteasome-dependent destruction between AURKB
and -C, we tested the effects of mutating the KEN and A-boxes
and deleting the first 93 amino acids of AURKB-GFP. In both
cases the degradation kinetics were not affected (Fig. 4 C and D
and Fig. S7 A and B). Furthermore, a chimeric version of
AURKC that contained the first 90 amino acids of AURKB was
degraded with similar kinetics as wild-type AURKC (Fig. S7C).
Taken together, these data strongly suggest that the difference in
degradation kinetics between AURKB and C, although protea-
some-dependent, cannot be explained by the destruction motifs
in AURKB’s N terminus.
Aurkc Is a Maternally Recruited Message. AURKC protein appears
to increase between MI and MII when immunofluorescent sig-
nals are compared (Fig. S5). Because oocyte meiotic maturation
takes place in the absence of transcription, protein levels are
regulated by mRNA recruitment and mRNA degradation; re-
cruitment is regulated by cytoplasmic polyadenylation element
(CPE) and DazL binding sequences in 3′UTRs (29, 30). Aurkc
contains a conserved CPE (UUUUAU) in its 3′ UTR that is in
close proximity (11 nucleotides) to its hexanucleotide polyade-
nylation sequence (HEX) (Fig. 5A) and three putative DAZL
binding sites (U2–10C/GU2–10), suggesting that the increase in
AURKC protein is caused by recruitment of Aurkc mRNA
during meiotic maturation. The Aurkb 3′ UTR, however, also
contains a CPE adjacent to the HEX and a second CPE further
5′ to the HEX. To test whether Aurkb and Aurkc mRNAs are
recruited during meiotic maturation, we fused the 3′ UTRs of
these kinases to a firefly luciferase (Luc) reporter (Fig. 5A).
Following injection and assay for luciferase activity, we found
that the 3′ UTR of Aurkc recruited Luc mRNA ∼10-fold fol-
lowing maturation (Fig. 5B), compared with ∼threefold for the
Aurkb 3′ UTR. Recruitment of Luc mRNA significantly depen-
ded upon the CPE in Aurkc because it was reduced to ∼2.5-fold
when two nucleotides of the Aurkc CPE had been mutated (Fig.
5). These data indicate that AURKC protein levels increase
during meiotic maturation because its message is recruited for
translation in a CPE-dependent manner.
AURKB Can Compensate for Loss of AURKC in Vivo. Mitotic cells
depleted for AURKB are rescued by ectopic expression of
AURKC, suggesting that they can carry out the same functions
(8, 9). Our data indicate that AURKB protein levels decline
during oocyte maturation and embryonic development (Figs. 1–
4). To determine if AURKB is functionally equivalent to
AURKC and if increasing the amount of AURK activity in
knockout oocytes and embryos can compensate for loss of
AURKC, we matured oocytes from Aurkc−/−mice that were
microinjected with cRNA encoding mCherry, Aurkc-mCherry,
or Aurkb-mCherry; oocytes from wild-type mice injected with
mCherry served as controls. Expression of either Aurkc- or
Aurkb-mCherry rescued both the MI arrest phenotype in
Aurkc−/−oocytes (Fig. 6A) and the cytokinesis phenotype in
one-cell embryos from Aurkc−/−mothers (Fig. 6C). Further-
more, AURKB was no longer restricted to centromeres at MI
in Aurkc−/−oocytes, as it was in wild-type cells, but it adopted
a chromatin localization that is typical of AURKC (Fig. 6B).
Similarly, AURKB-GFP mimicked AURKC centromere lo-
calization in Aurkc−/−eggs, whereas AURKB was not detected
at centromeres in wild-type MII eggs (Fig. 6B). Our findings
that AURKB rescued the Aurkc−/−polar body extrusion and
one-cell cytokinesis defects and adopted the subcellular local-
ization of AURKC indicate that AURKB can compensate for
the loss of AURKC during meiosis and support the model that
AURKB activity declines during meiotic maturation and embryonic
Here, we report that female mice lacking Aurkc are subfertile.
Oocytes from knockout mice have a higher incidence of chro-
mosome misalignment and arrest at MI, and one-cell embryos
often fail in cytokinesis (Figs. 1 and 2). These phenotypes are
hallmarks of cells that lack Aurora kinase activity (22, 24, 25)
and worsen as development continues (Fig. 2D). Interestingly,
the phenotypic severity (both in vivo and in vitro) varies between
mice (Figs. 1F and 2B, and Fig. S8B). These differences could be
explained by the degree of compensation by AURKB. In some
representation of AURKB and AURKC. KEN, A-, and D-boxes are indicated,
and the conserved regions are shaded in light gray. (B–D) GV-intact oocytes
were coinjected with the indicated cRNAs and matured to the indicated
stage. Cycloheximide was added 1 h before the first time point and fluo-
rescent images were obtained at the indicated times. Data represent mean
(± SEM) from at least 30 oocytes from two independent experiments. MG132
was added as indicated to inhibit the proteasome (B).
AURKB stability does not depend upon its N terminus. (A) Schematic
Schindler et al. PNAS
| Published online July 9, 2012
cases AURKB levels may fall below a threshold necessary to
support maturation or embryo development, and therefore
viability is compromised. It is unlikely that the six additional
pseudogene copies of Aurkc (31) or putative N-terminal splice
variants (6) compensate for loss of AURKB because we did not
detect their presence in oocytes from knockout mice (Fig. S1A)
using a Taqman probe that is 97–100% identical in sequence to
these regions, nor did we detect AURKC protein by immu-
nocytochemistry with an antibody that recognizes the C ter-
minus (Fig. S1C). Our in vitro maturation and development
data demonstrate that the requirement for AURKC in vitro
appears more severe than its requirement in vivo (compare
Figs. 1 B–F and 2 A–F with Fig. 1A). This difference may be
because of hormonal priming for the in vitro experiments
versus natural matings in the in vivo fertility trials. Alterna-
tively, the amount of AURKB may be less following culture in
vitro, and therefore it is unable to compensate as well as in
vivo. When we isolated in vivo developed blastocysts from
hormonally primed mice, we found fewer in the knockout
females compared with wild-type, favoring the first model (Fig.
S8). Unfortunately, testing the second model is technically not
feasible because of limitations in our ability to quantify the
amount of AURKB protein.
Our findings that AURKB and AURKC stability and re-
cruitment differ during meiosis (Figs. 3, 4C, and 5B) provide an
explanation for the existence of AURKC in oocytes. AURKB is
a critical regulator of mitotic cell division, but it is not stable
during meiotic maturation, and it is not transcribed again until
the blastocyst stage (Figs. 2G and 3). In contrast, AURKC
protein likely increases during meiotic maturation, due to its
greater stability and CPE-mediated recruitment of Aurkc mRNA
(Fig. 5B and Fig. S4). Although AURKB degradation depends
upon the proteasome and the N terminus of AURKB contains
destruction elements that AURKC lacks, the N terminus does
not appear to regulate AURKB stability during meiosis as it does
in mitosis (Fig. 4). Because both exogenous proteins gradually
decline during meiosis, both may be regulated via their C-ter-
minal D-boxes (Fig. 4A), but AURKB degradation is more pro-
nounced (Fig. 3D).
The discovery of these regulatory mechanisms, combined with
evidence that embryos can develop to the blastocyst stage in the
absence of AURKB (10), suggests that oocytes express a third
AURK to compensate for the inherent instability of AURKB.
Notably, expression of a nondegradable AURKB in somatic cells
leads to aneuploidy, indicating that its degradation is important
for mitotic cell cycles (20). Consistent with others studies, our
data support the hypothesis that AURKB and AURKC have
similar functions (8–10). There may be subtle functional differ-
ences between AURKB and -C, however, that are not revealed
in our assays, and oocytes lacking both AURKB and AURKC
will be a powerful tool for answering this question.
tion of the Firefly luciferase (Luc) fusions to the Aurkb and Aurkc 3′ UTRs.
Putative CPE and HEX motifs are underlined and the number of nucleotides
between the motifs are indicated in the gray boxes. In Aurkc, the two
nucleotides in the CPE that were mutated are in bold. Note that the lengths
are not to scale. (B) GV-intact oocytes were coinjected with Luc fused to the
indicated 3′ UTR and Renilla luciferase as an injection volume control. Lu-
minescence was measured and quantified as the fold-difference in MII
compared with GV. Data represent mean (± SEM) from 10 oocytes from two
independent experiments. mut, mutated.
Aurkc is a maternally recruited message. (A) Schematic representa-
mice rescues their defects. (A and C) GV-intact oocytes or one-cell embryos
from a single mouse of the indicated genotype were subdivided and
microinjected with the indicated cRNAs and matured to MII to determine
incidence of PBE or developed to the two-cell stage to monitor cytokinesis.
These experiments were repeated three times with two to four KO mice per
experiment. (B) Wild-type or Aurkc KO oocytes were microinjected with
Aurkb-Gfp cRNA, matured to MI and MII, and imaged live. Shown are rep-
resentative images. (Scale bar, 5 μm.) One-way ANOVA was used to analyze
the data. *P < 0.05; mCh, mCherry.
Ectopic expression of AURKB in oocytes and embryos from Aurkc KO
| www.pnas.org/cgi/doi/10.1073/pnas.1120517109Schindler et al.
AURKC is an example of a cell-cycle regulator isoform that is
recruited during meiotic maturation to support meiosis, fertil-
ization, and early embryonic cell division. Other examples in-
clude WEE1B kinase and IP3receptor, which contribute to exit
from MII (32, 33), and CDC6 and ORC6L, which allow DNA
replication following fertilization (34, 35). We propose that these
recruited mRNAs comprise a strategy to switch from a program
of oocyte growth without cell division, during the prolonged cell-
cycle arrest, to meiotic and mitotic divisions without growth (i.e.,
early embryonic cleavage stages).
Materials and Methods
Oocyte and Embryo Collection, Culture, and Microinjection. For experiments in
Figs. 1–4, fully grown GV-intact oocytes from equine CG-primed (44–48 h
before collection) 6-wk-old female CF-1 mice (Harlan) (Figs. 1–4) or from the
indicated Aurkc genotype (Figs. 5–6) were obtained as previously described
(35). Meiotic resumption was inhibited by addition of 2.5 μM milrinone
(Sigma) to the collection, culture, or microinjection medium (36). Oocytes
were cultured in Chatot, Ziomek, and Bavister (CZB) medium in an atmo-
sphere of 5% CO2in air at 37 °C and were collected and microinjected in
MEM/PVP (polyvinylpyrrolidone, 3 g/L) (37). To inhibit the proteasome,
MG132 (Sigma) was added to the culture medium to a final concentration of
20 μM. To activate MII eggs, the eggs were placed in Ca2+/Mg2+-free CZB plus
10 mM SrCl2and 5 mg/mL cytochalasin B (Sigma C2743) 18 h after matura-
tion. After 3 h, the eggs were washed free of SrCl2. Oocytes were micro-
injected with 250 ng/μL of Aurkb/c cRNAs, as previously described (35).
Injected oocytes were held for 2–12 h before induction of maturation. For
maturation experiments, oocytes were washed and cultured in milrinone-
free CZB medium.
In Figs. 2 and 6, primed mice of the indicated Aurkc genotype (Fig. 6) were
administered human CG (hCG) and mated to B6D2F1/J males (Jackson Labo-
ratories). In Fig. 6, one-cell embryos were isolated 20 h post-hCG from the
ampullaeandcultured inKSOM+aminoacids (Millipore) inan atmosphereof
as previously described and immediately frozen (38). All animal experiments
were approved by the institutional animal use and care committee and were
consistent with the National Institutes of Health (NIH) guidelines.
Generation and Genotyping of Aurkc−/−Mice. Generation of Aurkc−/−was
described previously (11). Lexicon Pharmaceuticals transferred cryopreserved
Aurkc+/−embryos into pseudopregnant females and the resulting pups were
sent to the University of Pennsylvania for further breeding. For genotyping,
the copy number of Neo was quantified by real-time PCR per the manu-
facturer’s protocol. Briefly, tails were digested in 400 μL of lysis buffer (125
mM NaCl, 40 mM Tris, pH 7.5, 50 mM EDTA, pH 8, 1% (vol/vol) sarkosyl, 5
mM DTT, and 50 μM spermidine) with 6 μL of Proteinase K (Sigma #P4850)
for 2 h at 65 °C. After dilution of 1:30 in water, the lysates were boiled for 5
min to denature Proteinase K. Two miroliters of the diluted DNA was added
to each reaction. Primers to detect Neo (F: 5′ CTCCTGCCGAGAAAGTATCCA-
3′; R: GGTCGAATGGGCAGGTAG-3′) were used at a final concentration of
300 nM and primers to detect Csk (for sample normalization) (F 5′-CTGGC-
CATCCGGTACAGAAT-3′; R 5′-TGCAGAAGGGAAGGTCTTGCT-3′) were used at
a final concentration of 100 nM. The TAMRA-quenched Neo probe (ABI) was
conjugated to 6-fluorescein amidite and used at a final concentration of 100
nM and the TAMRA-quenched Csk probe was conjugated to VIC and used at
a final concentration of 100 nM. The comparative Ctmethod was used to
calculate the Neo copy number.
Cloning, Mutagenesis, and in Vitro Synthesis of cRNA. Generation of H2B-
mCherry and nondegradable cyclin B-GFP were described previously (39, 40).
To generate Aurkb and Aurkc-GFP and -mCherry, mouse Aurk sequences
were amplified and cloned into pIVT-GFP or -mCherry (41). Truncated
AURKB and chimeric AURKB/C constructs were also generated by PCR and
cloned into pIVT-GFP. The Aurkb-KEN/A box mutant was generated by site-
directed mutagenesis using the QuikChange Multisite Mutagenesis (Agilent
Technologies) kit per the manufacturer’s instructions. The KEN box (amino
acids 4–9) was changed to AAN and the A-box (amino acids 26–29; QRVL)
was changed to QAVA. The 3′ UTRs of Aurkb and Aurkc were PCR-amplified
from full-length clones obtained from Open Biosystems and cloned into pIVT
containing firefly luciferase. Renilla luciferase was generated as described
previously (34). The CPE (TTTTAT) in the 3′ UTR of Aurkc was mutated to
TTGGAT using QuikChange.
Gfp- and mCherry-containing constructs. EcoRI digestion was used for lu-
ciferase-containing constructs. After purification of the digests (Qiagen PCR
Cleanup) complementary RNAs (cRNAs) were generated using the mMes-
sage mMachine T7 kit (Ambion) according to the manufacturer’s instruc-
tions. The cRNA was purified using the RNAeasy kit (Qiagen) and eluted in
30 μL of RNase-free water.
Real-Time PCR. Fifty oocytes or embryos at the indicated stage were isolated
from CF-1 mice and frozen before processing. After thawing on ice, 2 ng of
Gfp mRNA was added to each sample. Next, total RNA from the mixtures
were purified using the PicoPure RNA isolation kit (Arcturus) per the man-
ufacturer’s protocol. After purification, cDNA was generated by reverse
transcription using random hexamers and SuperScript II, as previously de-
scribed (35). Taqman probes specific for Aurkb (Mm01718146_g1) and Aurkc
(Mm03039428_g1) (Applied Biosystems) were used for gene expression de-
tection and the comparative Ctmethod was used to determine the differ-
ence in expression levels between stages. Data were acquired using an ABI
Prism 7000 (Applied Biosystems).
Luciferase Assay. The cRNAs of firefly luciferase fused to the 3′ UTR of Aurkb
or Aurkc (200 ng/μL) and Renilla luciferase (25 ng/μL) were coinjected into
GV-intact oocytes from CF-1 mice, as described above, and incubated in
vitro for at least 2 h in milrinone-containing medium. After incubation,
one-half of the injected oocytes were matured to MII for 17 h. The other
half was maintained at the GV stage. After washing in PBS + PVP, single
oocytes were collected and lysed in Passive Lysis Buffer (12 μL/oocyte) for
15 min at room temperature with shaking followed by incubation on ice
until processing with the Dual Luciferase reporter assay system (Promega).
The manufacturer’s instructions were followed except that 10 μL of sample
and 50 μL of Luciferase Assay Reagent II and of Stop and Glo Reagent
were used. Signal intensities were obtained using a Monolight 2010
luminometer (Analytical Luminescence Laboratory). Background fluo-
rescence was subtracted by measuring signals in uninjected oocytes and
firefly luciferase activities were normalized to that of Renilla luciferase
and expressed as the fold-recruitment in MII eggs compared with
Immunocytochemistry. Following meiotic maturation or embryo develop-
ment to the indicated stage, oocytes and embryos were fixed in PBS
containing 3.7% paraformaldehyde for 1 h at room temperature and
transferred to blocking buffer [PBS + 0.3% (wt/vol) BSA + 0.01% (vol/vol)
Tween-20] for storage overnight at 4 °C. After 15 min of permeabilization
in PBS containing 0.1% (vol/vol) Triton X-100 and 0.3% (wt/vol) BSA, cells
were incubated in rabbit anti-AURKC antibody (Bethyl A400-023A; epitope
BL1217) at 1:30 or a polyclonal anti–β-tubulin antibody conjugated to
Alexa 488 (Cell Signaling 3623) at 1:75 in blocking buffer for 1 h at room
temperature in a humidified chamber. After washing, secondary Alexa
Fluor 594 anti-rabbit antibody (Invitrogen A11012) was diluted 1:200 in
blocking solution and the samples incubated for 1 h at room temperature.
After a final wash in blocking buffer containing Sytox Green (Invitrogen
S7020; 1:5,000), cells processed to detect AURKC were mounted in Vec-
taShield (Vector Laboratories). Cells processed to detect the spindle were
mounted in VectaShield containing 3 μg/mL propidium iodide. Ploidy
analysis was acquired as described previously (40, 42). Images were col-
lected with a spinning disk confocal using a 100× 1.4 NA oil immersion
objective and processed using ImageJ software (NIH), as previously de-
Live-Cell Imaging. To monitor destruction of the exogenous AURKs, GV
oocytes were comicroinjected with the indicated cRNAs. Following injection,
oocytes were maintained at the GV stage for 16–17 h (for the GV-MI time
course), 10–11 h (for the MI-MII time course), or 1 h (for the MII activation
time course) before maturation. To inhibit translation, oocytes were placed
in 10 ng/μL cycloheximide (Sigma C7698) 1 h before live imaging. Similar,
but not as robust as effect, was observed when cycloheximide was omitted.
For live imaging, each oocyte or activated egg was transferred into an in-
dividual drop of CZB medium or Ca2+/Mg2+-free CZB containing SrCl2and
cytochalasin B, respectively. All drops contained cycloheximide and were
under oil in a FluoroDish (World Precision Instruments). Differential in-
terference contrast (DIC), GFP, and mCherry image acquisition was started
at 0 (GV-MI), 6 (MI-MII), and 18 h (MII activation) after maturation on
a Leica DM6000 microscope with a 40× 1.25 NA oil immersion objective and
a charge-coupled device camera (Orca-AG; Hamamatsu Photonics) con-
trolled by Metamorph Software. The microscope stage was heated to 37 °C
and 5% CO2was maintained using a microenvironment chamber (PeCon)
and an airstream incubator (ASI 400, Nevtek). Images of individual cells
Schindler et al.PNAS
| Published online July 9, 2012
were acquired every 30–60 min during the indicated intervals. For each Download full-text
oocyte, fluorescence was calculated at each time point as a fraction of the
fluorescence at the first time point.
To monitor meiotic timing and timing of one-cell cleavage, oocytes and
embryos from the indicated genotypes were matured in drops of CZB or
KSOM + AA, respectively, under oil in a FluoroDish. DIC images were ac-
quired every 10 min on the microscope described above using a 20× 0.4 NA
objective to monitor the timing of GVBD and polar body extrusion or every
5 min to capture cytokinesis in one-cell embryos.
Fertility Trials. Wild-type and knockout Aurkc female mice of sexual maturity
(6 wk) were continually mated to wild-type B6D2 (Jackson Laboratories)
male mice of known fertility for 8 mo. Cages were checked daily for the
presence and number of pups.
Statistical Analysis. One-way ANOVA and Student’s t test, as indicated in
figure legends, were used to evaluate the differences between groups using
Prism software (GraphPad Software). P < 0.05 was considered significant.
ACKNOWLEDGMENTS. The authors thank Fabian Cardenas and Danielle
Young for assistance with mouse genotyping, Kristy Shuda for generation of
the Aurora kinase isoforms-GFP constructs, and Paula Stein and Sarah
Kimmins for helpful discussions. This work was supported by National
Institutes of Health Grants HD 055822 (to K.S.) and HD 058730 (to R.M.S.
and M.A.L.), and a Searle Scholar Award (to M.A.L.).
1. Gopalan G, Chan CS, Donovan PJ (1997) A novel mammalian, mitotic spindle-
associated kinase is related to yeast and fly chromosome segregation regulators. J Cell
2. Yanai A, Arama E, Kilfin G, Motro B (1997) ayk1, a novel mammalian gene related to
Drosophila aurora centrosome separation kinase, is specifically expressed during
meiosis. Oncogene 14:2943–2950.
3. Tseng TC, Chen SH, Hsu YP, Tang TK (1998) Protein kinase profile of sperm and eggs:
Cloning and characterization of two novel testis-specific protein kinases (AIE1,
AIE2) related to yeast and fly chromosome segregation regulators. DNA Cell Biol
4. Baldini E, et al. (2011) Aurora kinases are expressed in medullary thyroid carcinoma
(MTC) and their inhibition suppresses in vitro growth and tumorigenicity of the MTC
derived cell line TT. BMC Cancer 11:411.
5. Price DM, Kanyo R, Steinberg N, Chik CL, Ho AK (2009) Nocturnal activation of aurora
C in rat pineal gland: Its role in the norepinephrine-induced phosphorylation of
histone H3 and gene expression. Endocrinology 150:2334–2341.
6. Yan X, et al. (2005) Cloning and characterization of a novel human Aurora C splicing
variant. Biochem Biophys Res Commun 328:353–361.
7. Yan X, et al. (2005) Aurora C is directly associated with Survivin and required for
cytokinesis. Genes Cells 10:617–626.
8. Slattery SD, Mancini MA, Brinkley BR, Hall RM (2009) Aurora-C kinase supports mitotic
progression in the absence of Aurora-B. Cell Cycle 8:2984–2994.
9. Sasai K, et al. (2004) Aurora-C kinase is a novel chromosomal passenger protein that
can complement Aurora-B kinase function in mitotic cells. Cell Motil Cytoskeleton
10. Fernández-Miranda G, et al. (2011) Genetic disruption of aurora B uncovers an
essential role for aurora C during early mammalian development. Development
11. Kimmins S, et al. (2007) Differential functions of the Aurora-B and Aurora-C kinases in
mammalian spermatogenesis. Mol Endocrinol 21:726–739.
12. Ben Khelifa M, et al. (2011) A new AURKC mutation causing macrozoospermia:
Implications for human spermatogenesis and clinical diagnosis. Mol Hum Reprod
13. Shuda K, Schindler K, Ma J, Schultz RM, Donovan PJ (2009) Aurora kinase B modulates
chromosome alignment in mouse oocytes. Mol Reprod Dev 76:1094–1105.
14. Swain JE, Ding J, Wu J, Smith GD (2008) Regulation of spindle and chromatin
dynamics during early and late stages of oocyte maturation by aurora kinases. Mol
Hum Reprod 14:291–299.
15. Lane SI, Chang HY, Jennings PC, Jones KT (2010) The Aurora kinase inhibitor
ZM447439 accelerates first meiosis in mouse oocytes by overriding the spindle
assembly checkpoint. Reproduction 140:521–530.
16. Gautschi O, et al. (2008) Aurora kinases as anticancer drug targets. Clin Cancer Res
17. Vogt E, Kipp A, Eichenlaub-Ritter U (2009) Aurora kinase B, epigenetic state of
centromeric heterochromatin and chiasma resolution in oocytes. Reprod Biomed
18. Yang KT, et al. (2010) Aurora-C kinase deficiency causes cytokinesis failure in meiosis I
and production of large polyploid oocytes in mice. Mol Biol Cell 21:2371–2383.
19. Sharif B, et al. (2010) The chromosome passenger complex is required for fidelity
of chromosome transmission and cytokinesis in meiosis of mouse oocytes. J Cell Sci
20. Nguyen HG, Chinnappan D, Urano T, Ravid K (2005) Mechanism of Aurora-B
degradation and its dependency on intact KEN and A-boxes: Identification of an
aneuploidy-promoting property. Mol Cell Biol 25:4977–4992.
21. Stewart S, Fang G (2005) Destruction box-dependent degradation of aurora B is
mediated by the anaphase-promoting complex/cyclosome and Cdh1. Cancer Res
22. Yabe T, et al. (2009) The maternal-effect gene cellular island encodes aurora B kinase
and is essential for furrow formation in the early zebrafish embryo. PLoS Genet
23. Severson AF, Hamill DR, Carter JC, Schumacher J, Bowerman B (2000) The aurora-
related kinase AIR-2 recruits ZEN-4/CeMKLP1 to the mitotic spindle at metaphase and
is required for cytokinesis. Curr Biol 10:1162–1171.
24. Giet R, Glover DM (2001) Drosophila aurora B kinase is required for histone H3
phosphorylation and condensin recruitment during chromosome condensation and
to organize the central spindle during cytokinesis. J Cell Biol 152:669–682.
25. Hauf S, et al. (2003) The small molecule Hesperadin reveals a role for Aurora B in
correcting kinetochore-microtubule attachment and in maintaining the spindle
assembly checkpoint. J Cell Biol 161:281–294.
26. Golbus MS, Stein MP (1976) Qualitative patterns of protein synthesis in the mouse
oocyte. J Exp Zool 198:337–342.
27. Downs SM (1990) Protein synthesis inhibitors prevent both spontaneous and
hormone-dependent maturation of isolated mouse oocytes. Mol Reprod Dev
28. Avo Santos M, et al. (2011) A role for Aurora C in the chromosomal passenger
complex during human preimplantation embryo development. Hum Reprod
29. Richter JD (2007) CPEB: A life in translation. Trends Biochem Sci 32:279–285.
30. Chen J, et al. (2011) Genome-wide analysis of translation reveals a critical role for
deleted in azoospermia-like (Dazl) at the oocyte-to-zygote transition. Genes Dev
31. Hu HM, Chuang CK, Lee MJ, Tseng TC, Tang TK (2000) Genomic organization,
expression, and chromosome localization of a third aurora-related kinase gene, Aie1.
DNA Cell Biol 19:679–688.
32. Xu Z, Williams CJ, Kopf GS, Schultz RM (2003) Maturation-associated increase in IP3
receptor type 1: Role in conferring increased IP3 sensitivity and Ca2+ oscillatory
behavior in mouse eggs. Dev Biol 254:163–171.
33. Oh JS, Susor A, Conti M (2011) Protein tyrosine kinase Wee1B is essential for
metaphase II exit in mouse oocytes. Science 332:462–465.
34. Murai S, Stein P, Buffone MG, Yamashita S, Schultz RM (2010) Recruitment of Orc6l,
a dormant maternal mRNA in mouse oocytes, is essential for DNA replication in 1-cell
embryos. Dev Biol 341:205–212.
35. Anger M, Stein P, Schultz RM (2005) CDC6 requirement for spindle formation during
maturation of mouse oocytes. Biol Reprod 72:188–194.
36. Tsafriri A, Chun SY, Zhang R, Hsueh AJ, Conti M (1996) Oocyte maturation involves
compartmentalization and opposing changes of cAMP levels in follicular somatic and
germ cells: Studies using selective phosphodiesterase inhibitors. Dev Biol 178:393–402.
37. Chatot CL, Ziomek CA, Bavister BD, Lewis JL, Torres I (1989) An improved culture
medium supports development of random-bred 1-cell mouse embryos in vitro.
J Reprod Fertil 86:679–688.
38. Ma P, Schultz RM (2008) Histone deacetylase 1 (HDAC1) regulates histone acetylation,
development, and gene expression in preimplantation mouse embryos. Dev Biol
39. Schindler K, Schultz RM (2009) CDC14B acts through FZR1 (CDH1) to prevent meiotic
maturation of mouse oocytes. Biol Reprod 80:795–803.
40. Duncan FE, Chiang T, Schultz RM, Lampson MA (2009) Evidence that a defective
spindle assembly checkpoint is not the primary cause of maternal age-associated
aneuploidy in mouse eggs. Biol Reprod 81:768–776.
41. Igarashi H, Knott JG, Schultz RM, Williams CJ (2007) Alterations of PLCbeta1 in
mouse eggs change calcium oscillatory behavior following fertilization. Dev Biol
42. Stein P, Schindler K (2011) Mouse oocyte microinjection, maturation and ploidy
assessment. J Vis Exp 53:e2851.
43. Chiang T, Duncan FE, Schindler K, Schultz RM, Lampson MA (2010) Evidence that
weakened centromere cohesion is a leading cause of age-related aneuploidy in
oocytes. Curr Biol 20:1522–1528.
| www.pnas.org/cgi/doi/10.1073/pnas.1120517109Schindler et al.