RESEARCH ARTICLE
Molecular Reproduction & Development 78:33–47 (2011)
Fer Tyrosine Kinase Is Required For Germinal Vesicle
Breakdown and Meiosis-I In Mouse Oocytes
1221LYNDA K. McGINNIS,* XIAOMAN HONG,LANE K. CHRISTENSON,AND WILLIAM H. KINSEY
1 Department of Anatomy and Cell Biology, University of Kansas Medical School, Kansas City, Kansas 2 Department of Molecular and Integrative Physiology, University of Kansas Medical School, Kansas City, Kansas
SUMMARY
The control of microtubule and actin-mediated events that direct the physical arrangement and separation of chromosomes during meiosis is critical since failure to maintain chromosome organization can lead to germ cell aneuploidy. Our previous studies demonstrated a role for FYN tyrosine kinase in chromosome and spindle organization and in cortical polarity of the mature mammalian oocyte. In addition to Fyn, mammalian oocytes express the protein tyrosine kinase Fer at high levels relative to other tissues. The objective of the present study was to determine the function of this kinase in the oocyte. Feline encephalitis virus (FES)-related kinase (FER) protein was uniformly distributed in the ooplasm of small oocytes, but became concentrated in the germinal vesicle (GV) during oocyte growth. After germinal vesicle breakdown (GVBD), FER associated with the metaphase-I (MI) and metaphase-II * Corresponding author: (MII) spindles. Suppression of Fer expression by siRNA knockdown in GV stage Department of Anatomy and Cell Biology oocytes did not prevent activation of cyclin dependent kinase 1 activity or chromo- University of Kansas some condensation during in vitro maturation, but did arrest oocytes prior to GVBD or Medical Center during MI. The resultant phenotype displayed condensed chromosomes trapped in 3901 Rainbow Blvd. mail stop 3038 the GV, or condensed chromosomes poorly arranged in a metaphase plate but with an Kansas City, KS 66160 USA. underdeveloped spindle microtubule structure or chromosomes compacted into a E-mail: lmcginnis2@kumc.edu tight sphere. The results demonstrate that FER kinase plays a critical role in oocyte meiotic spindle microtubule dynamics and may have an additional function in GVBD.
Mol. Reprod. Dev. 78: 33–47, 2011. ß 2010 Wiley-Liss, Inc. Published online 10 December 2010 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/mrd.21264 Received 6 October 2010; Accepted 15 November 2010
INTRODUCTION understood. The precise control of the microtubule and
actin-mediated events that direct the physical arrangement Maturation of the mammalian oocyte requires synchro- and separation of chromosomes during meiosis is critical nous progression and integration of many signaling path- since failure to maintain chromosome organization can lead ways. Historically, focus has been directed toward maturation promoting factor (MPF), mitogen-activated pro- tein kinases (MAPK), and other primary drivers that control Abbreviations: CDK1, cyclin-dependent kinase; FCH, FER-CIP1 homology meiosis (Masui and Markert, 1971; Su et al., 2002b; Peng domain; FER, FES-related kinase; FHM, ，ushing and holding medium; et al., 2007). Meiosis in oocytes also involves additional FES, feline encephalitis virus; GV, germinal vesicle; GVBD, germinal vesicle breakdown; hCG, human chorionic gonadotropin; IVM, in vitro maturation; oocyte-speci？c mechanisms involved in sister chromatid KSOM-MAT, KSOM medium with supplements for oocyte maturation; MAPK, cohesion (Hodges et al., 2005), chromosome condensation mitogen-activated protein kinase; MI, metaphase-I; MII, metaphase-II; MPF, (Swain and Smith, 2007), and spindle formation (Lindeman maturation promoting factor; PMSG, pregnant mare serum gonadotropin; PTK, and Pelegri, 2009) which are just now beginning to be protein tyrosine kinase; SH2, SRC homology domain.
ß 2010 WILEY-LISS, INC.
Molecular Reproduction & Development CGINNIS ET AL. M
to germ cell aneuploidy, a major risk to fertility especially in et al., 2003; Senis et al., 2003; Shapovalova et al., 2007) ageing women (Hunt and Hassold, 2008). Recent studies and its contribution remains an important question. have demonstrated that oocytes are highly specialized for Efforts to study FER following targeted gene knockout protein tyrosine kinase (PTK) signaling, with localized PTK have not been successful to date as a FES knockout activity occurring in the vicinity of spindle poles of proved to be embryonic lethal (Hackenmiller et al., 2000; metaphase-II (MII) oocytes as well as in the cortex of Hackenmiller and Simon, 2002). Successful generation of fertilized eggs (McGinnis et al., 2007; McGinnis and Alber- kinase dead mutant FER and FES mice, however, suggests tini, 2010). Studies revealed that the SRC family kinase that FER and FES perform some critical functions unrelated FYN plays an important role in maintaining meiotic spindle to kinase activity. Kinase dead FER or FES mutant females organization (Kinsey et al., 2003; Meng et al., 2006; McGin- exhibit reduced fertility (Senis et al., 2003) and double nis et al., 2007, 2009; Luo et al., 2009). Suppression of FYN mutant crosses exhibited reduced fertility and early repro- in oocytes by chemical inhibition, siRNA knockdown, or ductive senescence. Expression array data indicate that gene knockout led to the disorganization of metaphase-I oocytes express Fer at unusually high levels compared to (MI) and MII spindles that frequently correlated with meiotic other tissues, while Fes is barely detectable, suggesting arrest (McGinnis et al., 2009). In addition, Fyn-suppressed that Fer may play an important role in oocyte development. oocytes exhibited defects in cortical actin organization and The objective of the present study was to determine the polarity that correlated with reduced developmental com- function of FER in oocytes, and the results demonstrate that petence (Meng et al., 2006; McGinnis et al., 2007; this kinase plays an important role in meiosis. Experiments Tomashov-Matar et al., 2008; Luo et al., 2009). The above designed to test the function of this kinase during oocyte ？ndings highlighted the importance of identifying the control maturation indicated that the kinase is required for germinal mechanisms that direct spindle function in oocytes because vesicle breakdown (GVBD) and separately, for assembly of their importance to oocyte quality. The objective of the and function of the meiotic spindle. Suppression of Fer present study was to determine if other PTKs are critical to caused meiotic arrest, suggesting that FER activation may meiosis in oocytes. be a critical element of oocyte maturation and quality. One candidate family of protein kinases that may play a
role in meiotic spindle function is the Fer/Fes family. Fer and
Fes are the only known members of a distinct family of non- RESULTS receptor tyrosine kinases (Smithgall et al., 1998). Feline Expression of Fer Kinase in Oocytes encephalitis virus (FES)-related kinase (FER)-like proteins
Fer kinase is expressed at a low level in most cell types have been identi？ed in a diverse range of species including
humans and other mammals, Drosophila, C. elegans, birds, (Greer, 2002), but analysis of expression array data and marine sponges (Feldman et al., 1986; Pawson et al., (Novartis BioGPS, http://biogps.gnf.org; Su et al., 2002a) 1989; Katzen et al., 1991; Paulson et al., 1997; Cetkovic and http://amazonia.montp.inserm.fr (Assou et al., 2006, et al., 1998; Putzke et al., 2005). The FER proteins consist 2007; Wood et al., 2007) indicated the Fer transcripts are primarily of four domains: FER-CIP1 Homology (FCH), highly expressed in mouse (black bar Fig. 1A) and in human three-coiled coils, SRC homology domain (SH2), and (not shown). Quantitative RT-PCR analysis con？rmed ele-
C-terminal kinase domains. The kinase region includes an vated Fer expression in mouse oocytes (Fig. 1B). Fer atypical nuclear localization signal that requires the pres- mRNA was easily detectable in MII oocytes (Fig. 1B), while ence of the coiled coil and SH2 domains for regulation of Fes kinase mRNA was barely detectable above back- nuclear exit (Ben-Dor et al., 1999). This PTK family reg- ground (not shown). This con？rmed the expression array
ulates numerous cellular processes including cytoskeletal data and suggests that the oocyte is specialized for use of organization, cell adhesion, vesicle transport, and intracel- Fer and not Fes. Western blot analysis of whole ovarian lular signaling cascades (Greer, 2002). The close sequence tissue and oocytes demonstrated that both ovary (not homology between Fer and Fes suggests that these ki- shown) and oocytes (GV and MII) express FER protein nases have similar and sometimes redundant biological (Fig. 1C). The FER antibody detected a single band at functions (Smithgall et al., 1998; Greer, 2002). Several 94 kDa, the predicted molecular weight of full length FER. studies have described Fer (also called FerT2 in mouse) FES protein was detected in whole ovary, but was not expression in male germ cell maturation (Hazan et al., 1993; detectable in oocytes (not shown). The relative amount of Chen et al., 2003; Kierszenbaum et al., 2008), where a FER protein in each oocyte was determined by the ratio of truncated form lacking the N-terminal FCH and coiled coil FER to GAPDH within each set of oocytes. Although there domains participates in formation of the maturing sperm appeared to be a slight decrease in FER during maturation heads (Letwin et al., 1988; Pawson et al., 1989; Hazan et al., from GV to MII, this change was not statistically signi？cant
1993; Kierszenbaum et al., 2008). FER has been shown to (Fig. 1D; P > 0.05; four replicates). Immunohistochemistry associate with spindle microtubules in somatic cells, and indicated that FER was consistently concentrated in oo- can phosphorylate and promote elongation of microtubules cytes and associated granulosa cells of ovarian follicles in vitro (Kogata et al., 2003; Lee, 2005; Shapovalova et al., (Fig. 1E arrows), and was less abundant in stromal com- 2007). In spite of the strong evidence obtained in vitro, a ponents of the ovary and in the ovarian epithelium. Control requirement for FER in spindle function has not been sections (secondary antibody only) showed no labeling con？rmed in intact cells or gene mutant models (Kogata (data not shown). Detailed immunofluorescence analysis
34 Mol Reprod Dev 78: 33–47 (2011)
FER KINASE IN OOCYTE MEIOTIC MATURATION
Figure 1. Fer kinase is expressed in eggs and ovaries. Query of a mouse expression array database (BioGPS) showed that Fer kinase mRNA was
expressed at very high levels in oocytes and zygotes while the closely related Fes kinase was not highly expressed in these cells (A). This was
con？rmed by qRT-PCR ampli？cation of Fer transcripts from total RNA puri？ed from different tissues presented in Panel B relative to the ampli？cation of 18s mRNA. Western blot analysis of GV and ovulated MII stage oocytes (25 oocytes per lane) probed with anti-FER antibody
detected a single protein band at 94 kDa (C). The original blot with anti-FER was stripped and reblotted for GAPDH. Metamorph software was
used to determine a ratio of FER:GAPDH. This relative concentration of FER was not signi？cantly different between GV and MII oocytes (D) The anti-FER antibody was also used to detect FER kinase in paraf？n sections of ovaries from young female mice followed by alexa-488 coupled secondary antibody (green) and with Hoechst 33258 (red) to detect DNA. FER was present at a low level in most cells of the ovary (E) and was strongly expressed in follicles and oocytes (E arrows).
demonstrated that FER kinase undergoes subcellular pregnant mare serum gonadotropin (PMSG)-primed redistribution during oocyte growth. Small oocytes en- females in medium containing cilostamide to maintain closed in primordial and early primary follicles exhibited meiotic arrest at the GV stage, then washed free of the an even distribution of FER between the ooplasm and GV drug, and allowed to mature. Confocal immuno，uores-
(Fig. 2A,B). However, in growing oocytes of secondary and cence analysis demonstrated that FER kinase was concen- early antral follicles, FER was concentrated within the GV trated within the GV prior to acquisition of meiotic (Fig. 2C*,D*). competence (nucleolus not surrounded by chromatin; NSN,
Fig. 3A) and after acquiring meiotic competence (nucleolus
surrounded by chromatin; SN, Fig. 3B). The FER-speci？c Redistribution of FER During Meiotic Maturation fluorescence in the GV quickly declined once the oocytes In order to study the pattern of expression of FER kinase were washed free of cilostamide, and maturation was during oocyte maturation, oocytes were collected from allowed to proceed (Fig. 3C). During this period, FER was
35 Mol Reprod Dev 78: 33–47 (2011)
Molecular Reproduction & Development CGINNIS ET AL. M
to localize regionally along the chromatin, FER and DNA rarely overlapped (co-localized). Figure 3 also demon- strates the characteristic relationship between FER, chromatin, and peri-nuclear actin. FER was speci？cally excluded from patches of peri-nuclear f-actin (detected with phalloidin) surrounding the GV (Fig. 3C00 ). Once GVBD did occur, the pattern of FER kinase locali- zation changed again. As the chromosomes began to associate with the forming spindle 1–2 hr after release from meiotic arrest, FER-speci？c fluorescence was concentrat- ed adjacent to or surrounding the condensed chromatin in a manner similar to that observed when the chromosomes
were still retained within the GV (Fig. 4A00). In addition, a ,Bring of FER protein surrounded the ？lamentous actin ？bers
that encircled the forming spindle (Fig. 4A0000 ). At MI ,B (Fig. 4C) and MII (not shown), FER kinase was strongly
associated with the spindle itself and appeared to be
distributed throughout the ooplasm with regular granularity.
FER Kinase Is Required for Meiotic Maturation In order to determine if FER kinase played a signi？cant role in oocyte maturation and quality, Fer expression was suppressed by injection of siRNA complimentary to Fer during maturation in vitro. GV stage oocytes recovered from large follicles were injected with different concentra- tions (0.05–0.30 mM) of siRNA speci？c for the Fer sequence or with a scrambled control siRNA (see Materials and Methods Section). Western blot analysis of four experimen- tal groups was performed to demonstrate the effectiveness of the knockdown procedure (Fig. 5). The Fer siRNA (0.30 mM) reduced FER protein by 40% in GV oocytes that were arrested for 7 hr following siRNA injections (Fig. 5B, P < 0.05). Morphological analysis of the injected oocytes following 17 hr of maturation demonstrated that knockdown of FER strongly inhibited meiotic maturation while the control siRNA had no effect, as seen in Table I. Injection of Fer siRNA at all concentrations caused a delay in the rate of GVBD. Treatment with lower dosages (0.05–0.10 mM) of Fer siRNA caused a concentration-de- pendent meiotic arrest during MI, but injection of 0.30 mM Fer siRNA also caused arrest prior to GVBD. In order to demonstrate the speci？city of the Fer siRNA knockdown methodology, we tested the ability of exogenous recombinant FER kinase protein to rescue the Figure 2. FER kinase localizes to the nucleus of growing oocytes. FER phenotype caused by siRNA suppression of Fer mRNA kinase was present in the nucleus and cytoplasm of oocytes and levels. In this experiment, Fer siRNA (0.1 mM) was granulosa cells from primordial and small primary follicles (A,B). co-injected with active recombinant FER kinase (0.133 or However, as oocytes grow, FER becomes localized speci？cally within 0.200 mg/ml). The addition of active FER protein produced a the GV (C,D). (A-D) confocal images of immunohistochemically dose-dependant recovery in the rate of GVBD by 2 hr of in labeled tissue sections of adult mouse ovary (2 mm thick). (A’-D’) vitro maturation (IVM; Fig. 6A) and in the percentage of the same sections as (A-D) but showing FER kinase only. The location
of the GV in each oocyte is marked with an * (Green ? FER; oocytes that matured to MII (Fig. 6B). MII oocytes appeared
normal with spindles similar to those in control oocytes red ? DNA. Bar ? 20 mm).
indicating that exogenous Fer kinase overcame the defect more easily detected in the ooplasm, suggesting that some caused by siRNA knockdown of endogenous FER. In order of the FER protein was released from the GV and dispersed. to rule out the possibility that exogenous Fer kinase caused In addition, co-labeling with DNA indicated that FER protein an independent stimulation of maturation, we injected within the GV seemed to surround condensing chromatin as oocytes with exogenous, active FER protein alone and
compared their phenotypes to buffer control injected oo- maturation progressed (Fig. 3A’–C’). While FER was seen
36 Mol Reprod Dev 78: 33–47 (2011)
FER KINASE IN OOCYTE MEIOTIC MATURATION
Figure 3. FER kinase is localized within the GV during meiotic arrest. Oocytes were collected from PMSG primed ovaries into FHM þ cilostamide
to maintain GV arrest. Incompetent NSN oocytes exhibited FER sequestered primarily within the GV (A). Competent SN oocytes also contain high levels of FER kinase located primarily within the GV (B), although they also had an increased level of FER in the cytoplasm as compared to the NSN oocytes. By 30 min of maturation, GV levels of FER were further reduced (C). The FER within the GV localize near to, but not overlaying, chromatin
(Green ? FER; red ? f-actin; white ?DNA. Bar ? 10 mm). A–C are confocal sections (2-mm thick) taken through the center of the chromatin and showing the entire oocyte. A0 –C0 enlargements of the nuclear region within the same oocytes. A00–C00 are the same images as A, B, and C but with red (actin) included to better show the outline of the oocytes and the actin matrix surrounding the chromatin (A00 and B00).
cytes. The exogenous, active FER had no signi？cant effect characteristic morphology in which chromosome conden- on the percentage of oocytes that matured to GVBD at 1 hr sation had occurred without arrangement of the chromo-
somes on the spindle. Normal GV stage oocytes exhibit (84% n ? 31: 89% n ? 22, respectively) nor on the percent-
age that reached MII (74%:88%, respectively) indicating chromatin that is uniformly distributed within the GV, with a that exogenous FER kinase did not produce an unregulated thin strand surrounding the nucleolus (Fig. 7A). Approxi- stimulation of maturation beyond that accomplished by the mately 61% of the oocytes that arrested at the GV stage as a endogenous pool of native kinase. result of Fer siRNA injection exhibited condensed chroma-
tids lying within the GV near the nuclear envelope and tightly Immunofluorescence analysis revealed that the oocytes
arrested as a result of Fer-speci？c siRNA injection had a enclosing the nucleolus (Fig. 7B). In those cases where
37 Mol Reprod Dev 78: 33–47 (2011)
Molecular Reproduction & Development MCGINNIS ET AL.
Figure 4. FER kinase associated with the forming meiotic spindle during prometaphase and metaphase-I. During maturation, FER kinase
localized to the region surrounding condensing chromosomes and the forming spindle (A,B thin arrow). FER was also concentrated in the region immediately outside of the actin matrix that surrounds the chromatin (A,B thick arrow). By 7 hr of culture, the oocytes had reached metaphase-I, with FER localized speci？cally to the spindle and diffusely throughout the cytoplasm (C arrow). (A-D) confocal images of immunohistochemically labeled whole mounted oocytes (2 mm thick). (A’-D’) magni？cation of chromatin and spindle region of the oocytes in (A-D). (A’’ -D’’ ) the same
oocytes with actin (red) label added. (Green ? FER; red ? f-actin; white ? DNA. Bar ? 10 mm).
knockdown samples were labeled with an antibody to the GVBD was not blocked by 0.30 mM Fer siRNA injection,
cyclin-dependent kinase 1 (CDK1) target proteins (MPM2 73% of the oocytes exhibited chromosomes condensed into
a compact sphere of chromatids with a hollow core rather epitope). The MPM2 antibody was designed to recognize a than arranged on a spindle (Fig. 7C). These chromatin subset of proteins that are ser/thr phosphorylated by active structures were usually located at the oocyte cortex where CDK1, and is commonly used as an indicator of mitotic and the MI spindle would normally migrate; however, no spindle meiotic cell cycle activation (Vandre et al., 1984; Centonze was observed. and Borisy, 1990; Messinger and Albertini, 1991; In order to determine whether or not the knockdown Westendorf et al., 1994; do Carmo Avides et al., 2001; Lee of FER kinase prevented activation of the cell cycle et al., 2006; Ito et al., 2008; McGinnis and Albertini, 2010). machinery involved in meiosis resumption, control and Fer The images presented in Figure 8 demonstrate that Fer
38 Mol Reprod Dev 78: 33–47 (2011)
FER KINASE IN OOCYTE MEIOTIC MATURATION
Figure 5. Western blot analysis shows a reduction in FER protein in Figure 6. Rescue of the Fer siRNA phenotype by co-injection with siRNA injected eggs. Groups of 30 oocytes were injected with 0.3 mM active recombinant FER kinase. GV stage oocytes were injected either of Fer-siRNA or a scrambled control siRNA, then allowed to mature in with Fer siRNA (0.1 mM) alone or co-injected with siRNA and active vitro as described in Materials and Methods Section. Oocytes were recombinant FER protein (0.133 or 0.200 mg/ml). Oocytes were held processed for Western blot and probed for FER protein and GAPDH arrested at the GV stage for 12 hr, then washed and allow to mature for (A), as described previously. Fer siRNA injection reduced the 18 hr. The percentage of each group (N) of oocytes that successfully relative amount of Fer (B) in oocytes by 40% (P ? 0.05; n ? 4 completed GVBD (A) or proceeded to MII (B) is indicated in the Y-axis. replicates). Error bars indicate standard deviation of the mean. N ? number of
replicates; n ? total number of oocytes per treatment. Data were analyzed by ANOVA with SIDAK post hoc comparisons. Bars with different subscripts (a, b, or c) are signi？cantly different at
P < 0.05).
TABLE I. Fer siRNA Inhibition Meiotic Maturation in a Dose-Dependent Manner Treatment MII Concentration (mM) n 1 hr GV 3 hr GV 17 hr GV MI þ MII
bcaaaNon-injected 0 5 0 01008236–41 bcaabaControl siRNA 0.30 19 4 0 95 76 42–90 ababbFer siRNA 0.05 81 44 2 78 35 39–50 aabacbcFer siRNA 0.10 98 60 21 54 16 abd38–45 acFer siRNA 0.30 100 93 54 31 2 44–53
Data presented as percentages of oocytes with a visible GV at 1, 3, or 17 hr of maturation or reached at least metaphase-I (MI þ MII) and those that matured all the way to metaphase-II (MII) at 17 hr with 3–4 replicates per treatment. a,b,cValues in columns with different superscripts are signi？cantly different at P < 0.05.
Mol Reprod Dev 78: 33–47 (2011) 39
Molecular Reproduction & Development CGINNIS ET AL. M
Figure 8. Fer siRNA oocytes resume meiosis, but fail to complete Figure 7. Fer knockdown caused meiotic arrest at GV or MI. Normal
GV stage oocytes have diffuse chromatin throughout the nucleus and GVBD or to assemble an MI spindle. Oocytes injected with Fer siRNA surrounding the nucleolus (A). Knock-down of Fer kinase with siRNA were labeled with MPM2 antibody to detect activation of CDK1 and initiation of meiotic maturation. GV stage control oocytes have no produced two primary phenotypes: one arrested at GV/GVBD and the MPM2 positive epitopes in the cytoplasm and only faint labeling other at MI. Those Fer siRNA-injected oocytes that were arrested at within the GV (A). Control oocytes that resumed meiosis have high GVBD exhibited condensed chromatids lying close to the nuclear levels of MPM2 epitopes in the cytoplasm, especially in a sub-cortical envelope and other chromatids surrounding the nucleolus (B). Those region and associated with the metaphase spindle (B) demonstrating Fer siRNA-injected oocytes that were arrested at MI often contained activation of CDK1. Fer siRNA injected oocytes that arrested at GVBD compacted chromatids surrounding a spherical center without DNA. contained high levels of MPM2 epitopes within the GV and a region of This sphere of chromatids migrated to the cell cortex as would be the sub-cortex (C). Fer siRNA oocytes with condensed chromosome expected of a normal MI spindle (C) (Red ? f-actin; white ? DNA). spheres contained high levels of MPM2 epitopes throughout the A–C are 2 mm sections taken through the center of the chromatin; cytoplasm and especially near the cortex (D) similar to the MII stage A0–C0 are enlargements of the GV/chromatin of the same oocytes. control oocyte (B) (Green ? MPM2, red ? f-actin, white ? DNA; C0and D0 are the same oocytes as seen in C and D but without the kinase knockdown did not prevent CDK1 activation. Normal actin channel (red) to allow clearer observation of the MPM2 signal). control GV arrested oocytes, in which CDK1 is inactive,
show no MPM2 epitopes in the cytoplasm and only faint
labeling within the GV (Fig. 8A). Fer knockdown oocytes with condensed spheres of chromatin, but no spindle, that were arrested at the GV stage exhibited accumulation exhibited intense MPM2 accumulation in the cytoplasm, of MPM2 epitope within the GV and near the cell cortex, indicating that CDK1 was strongly activated in Fer siRNA- indicating that CDK1 was activated in this compartment injected oocytes (Fig. 8D).
(Fig. 8C). Control oocytes that have initiated maturation In order to determine if the unusual morphology of MI exhibit high levels of MPM2 epitopes throughout the cyto- chromosomes that resulted from Fer-knockdown might plasm with speci？c concentration in the subcortical region be associated with a failure of spindle formation, these (Fig. 8B). Those Fer knockdown oocytes that were arrested oocytes were stained with anti-tubulin to visualize spindle
40 Mol Reprod Dev 78: 33–47 (2011)
FER KINASE IN OOCYTE MEIOTIC MATURATION
Figure 9. FER depletion affects spindle formation. Oocytes injected with Fer siRNA (0.1 mM) were matured for 17 hr, then ？xed, labeled for b-tubulin, and imaged by confocal microscopy as described in the Materials and Methods Section. Variation was seen in the presence and/or formation of microtubules that correlated with chromatin con？gurations. Oocytes arrested in MI with abnormal chromosomes had few or no detectable microtubules (A–C). The few oocytes that matured to MII exhibited apparently
normal spindles (D) similar to non-injected control oocytes (E) (Green? b-tubulin; white ? DNA; (A-E) beta-tubulin only; (A’-E’) DNA only; (A’’ -E’’ ) combined images.
41 Mol Reprod Dev 78: 33–47 (2011)
Molecular Reproduction & Development CGINNIS ET AL. M
microtubules (Fig. 9). The Fer-knockdown oocytes arrested their associated matrix. The subcellular localization of FER with condensed chromatids, exhibited very few micro- in the oocyte has parallels in somatic cells. For example, tubules that were poorly organized and short in length FER kinase contains an atypical nuclear localization signal (Fig. 9A,C) compared with controls (Fig. 9E). Those within the kinase region (Ben-Dor et al., 1999) and the Fer-knockdown oocytes arrested with a condensed sphere closely related FES kinase was reported to translocate from of chromosomes exhibited signi？cant tubulin staining in the the nucleus of interphase somatic cells and to the cytoplasm central cavity, but no evidence of organized microtubules during prometaphase (Carlson et al., 2005). In addition, (Fig. 9B). The few Fer-knockdown oocytes with chromo- FER tagged with green fluorescent protein (GFP)-FER and somes aligned on a metaphase plate did exhibit a fairly GFP-FES have been localized to spindle microtubules in normal spindle, as detected with the anti-tubulin antibody mitotic cells, suggesting that this kinase family is capable (Fig. 9D), suggesting that if Fer was not suppressed suf？- of a speci？c interaction with microtubules (Kogata et al., ciently, the spindle could form normally and remain 2003; Laurent et al., 2004). Structure function studies have organized. implicated the FCH domain in binding to microtubules In summary, knockdown of Fer expression at the while the coiled coil regions function in dimerization GV stage did not block CDK1 activation or chromosome and are required for nuclear exit (Kim and Wong, 1995; condensation; however, it had two major effects on the Smithgall et al., 1998; Ben-Dor et al., 1999). ability of the oocyte to mature. The ？rst effect was to Functionally, a role for FER/FES in microtubule dynam- block GVBD resulting in an intact GV often containing ics and spindle assembly have been proposed (Kogata condensed chromosomes. The second effect was apparent et al., 2003; Laurent et al., 2004; Shapovalova et al., as a novel phenotype in which the distinct chromatids 2007) and there is strong evidence for this in vitro. Func- remained disorganized or grouped into a sphere, while the tional analysis of FER has been hampered in many somatic MI spindle failed to form or was unstable and dissociated cells by the tendency of FES and possibly SRC to compen- after forming. sate for the loss of FER. For example, Fer and Fes mutant
mice expressing kinase dead FER or FES do not exhibit
overt defects in microtubule function (Craig et al., 2001;
Senis et al., 2003). Double mutant females, however, have DISCUSSION fertility defects of unknown origin with reduced numbers of Oocyte maturation is a critical step in preparing the pups per litter and early reproductive senescence in addi- oocyte for fertilization, and defects in this process have tion to other compromised health issues (Senis et al., 2003). profound implications for oocyte quality and developmental Knockdown of FER within the fully grown oocyte immedi-
ately before maturation caused a dramatic inhibition of competence of the zygote. Mammalian oocytes are
meiosis (current studies). When an mRNA is depleted arrested at prophase of the ？rst meiotic division until
the signal for ovulation is received. Once released from immediately prior to induction of maturation, oocytes prob- the suppressive signals produced by follicle cells, the ably are unable to compensate for this sudden loss of an oocyte resumes meiosis, which involves a complex series essential mRNA (Fourcroy, 1982; Bachvarova, 1985; of intracellular events that are directed by protein kinases. In Murchison et al., 2007). However, whole animal mutants addition to the cell cycle kinases, which act as the primary survive because they can compensate for loss of FER control for cell cycle resumption, functional studies using kinase activity by substituting one or many of the other protein kinase inhibitors suggest that PTKs perform impor- similarly functioning tyrosine kinases.
tant steps during the complex intracellular events that are To determine if FER/FES may play a direct role in the critical elements of meiosis (Zheng et al., 2007). For exam- oocyte, we queried online expression array databases, and ple, FYN is required for maintenance of spindle organization found that Fer kinase was expressed at higher levels in the
oocyte than any other tissues reported for both human and during chromosome segregation (McGinnis et al., 2009),
mouse (Novartis BioGPS, http://biogps.gnf.org; Su et al., and for maintaining cortical actin polarity once MII has been
reached (Luo et al., 2009). Little is known about other PTKs 2002a and http://amazonia.montp.inserm.fr; Assou et al., that may perform critical steps during oocyte maturation. 2006, 2007; Wood et al., 2007). We con？rmed this result by
The FER tyrosine kinase is highly expressed in oocytes measuring mRNA and protein levels for FER/FES in mouse
ovary and oocytes; oocytes expressed high levels of Fer relative to other cell types and therefore was a likely candi-
date for oocyte-speci？c functions. The results presented mRNA and protein, while Fes was not detectable in oocytes here demonstrate that FER is expressed in much of the by qRT-PCR or Western blot. This suggested that the ovary, but was particularly enriched in oocytes and the oocyte might rely heavily on FER kinase for some aspect surrounding granulosa cells. FER was primarily concentrat- of oocyte function. This is supported by our observation that ed within the GV of growing and fully grown oocytes. At this knockdown of Fer in maturing oocytes had a very strong stage, the level of FER kinase was more than 1- to 2-fold inhibitory effect on maturation. The siRNA injections higher in the GV than in the surrounding cytoplasm. Once reduced FER protein in GV arrested oocytes by 40%. To
ensure that the effects seen from Fer siRNA were not meiosis arrest was terminated following release from the
follicle, the level of FER in the GV declined and the level in caused by a non-speci？c, off target response, we con-
ducted rescue experiments. This entailed the co-injection the ooplasm increased. Once the meiotic spindle formed,
of active FER kinase with or without the Fer siRNA and FER associated closely with the spindle microtubules or
42 Mol Reprod Dev 78: 33–47 (2011)