The mammalian oocyte undergoes two rounds of asymmetric cell divisions during meiotic maturation and fertilization. Acentric spindle positioning and cortical polarity are two major factors involved in asymmetric cell division, both of which are thought to depend on the dynamic interaction between myosin II and actin filaments. Myosin light chain kinase (MLCK), encoded by the Mylk1 gene, could directly phosphorylate and activate myosin II. To determine whether MLCK was required for oocyte asymmetric division, we specifically disrupted the Mylk1 gene in oocytes by Cre-loxP conditional knockout system. We found that Mylk1 mutant female mice showed severe subfertility. Unexpectedly, contrary to previously reported in vitro findings, our data showed that oocyte meiotic maturation including spindle organization, polarity establishment, homologous chromosomes separation, and polar body extrusion were not affected in Mylk1fl/fl;GCre females. Follicular development, ovulation, and early embryonic development up to compact morula occurred normally in Mylk1fl/fl;GCre females, but deletion of MLCK caused delayed morula-to-blastocyst transition. More than a third of embryos were at morula stage at 3.5 Days Postcoitum in vivo. The delayed embryos could develop further to early blastocyst stage in vitro on Day 4 when most control embryos reached expanded blastocysts. Our findings provide evidence that MLCK is linked to timely blastocyst formation, though it is dispensable for oocyte meiotic maturation.
INTRODUCTION
Mammalian oocyte maturation requires two successive rounds of highly asymmetric cell divisions to produce a functional haploid egg. This meiotic program involves several key events critical for asymmetric division in oocyte maturation, including spindle acentric positioning and the establishment of cortical polarity [1]. In meiosis I, following nuclear envelope breakdown, the spindle is initially assembled around the chromosomes at the center of the oocyte. Then the assembled spindle migrates to the subcortical area and becomes localized to the oocyte surface [2], inducing formation of a thickened F-actin cap surrounded by a myosin II ring above the MI spindle [3]. After first polar body extrusion, the meiosis II spindle becomes organized beneath the cortex and maintains its asymmetric position while awaiting fertilization, and a similar myosin II ring is formed and continuously maintained during MII arrest [4, 5]. After fertilization, the MII spindle rotates to be perpendicular to the cortex to facilitate the second polar body extrusion [6]. Studies suggest that dynamic actin filaments [7, 8], which are nucleated by Formin-2 and Spire1/Spire2 [9] in meiosis I and by the Arp2/3 complex [10] in meiosis II, contribute to the spindle's acentric positioning. A question remains concerning how the actin filaments generate the force to promote spindle relocation. Studies show that myosin II, which is enriched at the spindle poles, could pull the actin filaments and generate a force to drive the spindle migration in oocytes [11]; accordingly, myosin II is a key regulator of asymmetric division in oocytes.
Myosin light chain kinase (MLCK) is a Ca2+/calmodulin-dependent, actin and myosin binding, Ser/Thr protein kinase [12]. In smooth muscle, phosphorylation of the 20-kDa regulatory light chain (RLC) of myosin by MLCK is a well-documented event, contributing to the initiation of contraction [13, 14]; furthermore, MLCK could directly phosphorylate myosin II RLC at Thr18/Ser19 to control its assembly and activity in nonmuscle cells [1516–17]. Being a critical activator, MLCK participates in various cytoskeleton-related biological processes in nonmuscle cells, including cell motility, contraction, and shape change [18], but there is a lack of research, especially in vivo, on the roles of MLCK in female meiosis and early embryo development.
The purpose of the present study was to investigate the roles of MLCK in folliculogenesis, oocyte maturation, and early embryonic development by conditional gene knockout technology [19, 20] because conventional deletion of MLCK resulted in embryonic or perinatal lethality [21]. In this study, we found that MLCK mutant female mice showed decreased fertility. To investigate the causes of this subfertility, we explored the folliculogenesis, ovulation, oocyte polarity, polar body extrusion, fertilization, and early embryonic development of Mylk1mutant females. The results indicated that Mylk1 mutant females showed delayed blastocyst development with normal follicular development, oocyte maturation, and fertilization, which might explain the reduced litter size.
MATERIALS AND METHODS
Mice
To obtain Mylk1fl/fl;GCre+ females, we crossed Gdf9-Cre (C57BL6 background) [22] with previously reported Mylk1fl/fl mice [14], and the resulting offspring were intercrossed or mated with Mylk1fl/fl mice to generate Mylk1fl/fl;GCre+ female mice (C57BL6 and 129 mixed background). Unless otherwise specified, the Mylk1fl/fl female mice were used as the control group. DNA extraction from mouse tails was used to genotype the Mylk1fl and Mylk1Δ alleles. The primer pair for Mylk1fl allele was 1) 5′-TAGTGCGAGTGTCACTGTTG-3′ and 2) 5′-TGACTGGAAAAGGAGCCA-3′. The primer pair for Mylk1Δ allele was 3) 5′-TAGTGCGAGTGTCACTGTTG-3′ and 4) 5′-CCCCATGATTTGCCTCTAGT-3′. All the animal operations conformed to the guidelines by the Animal Research Committee principles of the Institute of Zoology, Chinese Academy of Sciences. All the mice were housed in a temperature-controlled room with a 12D:12L cycle.
Antibodies
Antibodies used in our experiments were available from the following companies: rabbit monoclonal anti-MLCK antibody (Abcam); rabbit polyclonal anti-phospho-myosin light chain 2 (pMLC2) (Ser19) antibody (CST); mouse monoclonal anti-β-actin antibody (Santa Cruz); goat polyclonal anti-Oct4 antibody (Santa Cruz); rabbit polyclonal anti-Bub3 antibody (Santa Cruz); rabbit monoclonal anti-α-tubulin antibody (CST); fluorescein isothiocyanate (FITC)-conjugated rabbit anti-goat immunoglobulin G (IgG) and FITC-conjugated goat anti-rabbit IgG (Zhongshan Golden Bridge Biotechnology); and Cy5-conjugated goat anti-rabbit IgG (Jackson ImmunoResearch Laboratory).
Western Blot Analysis
For Western blot analysis, 150 germinal vesicle (GV) oocytes were collected in 2× SDS sample buffer and boiled for 5 min at 100°C. Western blot analysis was performed as described previously [23]. Briefly, the separated protein in SDS PAGE were electrically transferred to a polyvinylidene fluoride membrane, and then the membrane was blocked in 50 mM Tris, 150 mM NaCl, and 0.1% Tween (TBST) containing 5% skimmed milk for 2 h, followed by incubation overnight at 4°C with rabbit monoclonal anti-MLCK antibody (1:1000) and mouse monoclonal anti-β-actin antibody (1:1000). After washing three times in TBST, 10 min each time, the membranes were incubated for 1 h at 37°C with peroxidase-conjugated anti-rabbit IgG (1:1000) and peroxidase conjugated anti-mouse IgG (1:1000), respectively. Finally, the membranes were washed three times in TBST and processed for detection with Bio-Rad ChemiDoc XRS+.
Semiquantitative RT-PCR
Total RNA of oocytes was extracted by RNeasy Micro Kit (Qiagen) and reverse transcribed with cDNA synthesis kit (Invitrogen). Peptidylprolyl isomerase A (Ppia) was used as the control gene to correct the cDNA level of the samples. The primers used were as follows: 1) Mylk1: 5′-GATGAAGTGGAAGTGTCCGA-3′ and 5′-CCAGAACCATGACAATGTTG-3′; and 2) Ppia: 5′-CGCGTCTCCTTCGAGCTGTTTG-3′ and 5′-TGTAAAGTCACCACCCTGGCACAT-3′.
Oocyte and Embryo Collection and In Vitro Culture
In our experiments, 6- to 8-wk-old Mylk1fl/fl;GCre+, Mylk1fl/fl, and wild-type females were used. Forty-eight hours after injection of eCG (10 international units [IU]), mice were used to collect GV oocytes or were then administrated hCG (10 IU) to collect MII eggs. To collect preimplantation embryos, the Mylk1fl/fl;GCre+ females, Mylk1fl/fl females, or superovulated wild-type females were caged with 8-wk-old C57BL/6J wild-type males. Noontime of the day when a vaginal plug was observed was considered as 0.5 Days Postcoitum (dpc). Zygotes were collected from the ampullar region of the oviduct at 0900 h on the day when the vaginal plug was found. Embryos were cultured in vitro to the morula stage in KSOM medium (Millipore) in a humidified 5% CO2 incubator at 37°C. Blastocysts were flushed from the uterus at 3.5 dpc and then cultured in vitro to 4.0 dpc in KSOM medium.
Immunofluorescence
Immunofluorescent staining was performed according to previously published methods [24]. Briefly, MI oocytes, MII eggs, morula, and blastocysts were fixed for 30 min at room temperature in 2% formaldehyde supplemented with 100 mM HEPES, 50 mM ethylene glycol tetraacetic acid, 10 mM MgSO4, 0.2% Triton X-100 (pH 7, titrated with KOH). Then they were treated with PBS containing 0.1% Triton X-100 overnight at 4°C and incubated with antibodies (MLCK, 1:50; pMLC2, 1:100; α-tubulin, 1:1000; Oct4, 1:500) in PBS, 0.1% Triton X-100, and 3% bovine serum albumin (BSA) overnight at 4°C. DNA was stained with propidium iodide (PI) or 4′,6-diamidino-2-phenylindole (DAPI) for 10 min. The oocytes or embryos were mounted on glass slides and examined with a laser scanning confocal microscope (Zeiss 780 META) and Perkin Elmer Ultra-VIEW VOX confocal Imaging System, respectively.
Fertility Analysis
Eight-week-old C57BL/6J wild-type males were mated with 6-wk-old Mylk1 mutant females and control siblings. Two females were caged with one male. A total of six Mylk1 mutant females and six control females were used in this experiment. During a 6-mo period, the cages were monitored daily, and the number and size of the litters were recorded.
Hematoxylin and Eosin Staining and Quantification of Ovarian Follicles
Ovaries were dissected from 8-wk-old mutant and control mice immediately after euthanasia and fixed in 4% formaldehyde overnight, dehydrated with a graded ethanol series, and embedded in paraffin. Paraffin-embedded ovaries were cut into sections of 8-μm thickness and mounted on glass slides. After adequately drying at 48°C overnight, sections were deparaffinized in xylene, hydrated by a graded alcohol series, and stained with hematoxylin and eosin for histological analyses. Ovarian primordial follicles and activated follicles were counted in all the sections of an ovary. Quantification of ovarian follicles was performed as previously reported [25]. In each section, only follicles that contained oocytes with clearly visible nuclei were counted.
Chromosome Spread
Chromosome spread was performed as described previously [26]. Briefly, the zona pellucida of MII eggs was removed by acid Tyrode solution (Sigma). After washing two to three times in M2 medium (Sigma), the oocytes were transferred to glass slides dipped in a solution of 1% paraformaldehyde in distilled H2O (pH 9.2) containing 0.15% Triton X-100 and 3 mM dithiothreitol. After the slides were dry, they were blocked with 1% BSA for 1 h at room temperature and incubated in anti-Bub3 antibody overnight at 4°C, washed three times and incubated with FITC-conjugated goat anti-rabbit IgG antibody (1:100) for 2 h at room temperature. DNA was stained with PI for 10 min. A laser scanning confocal microscope (Zeiss 710 META) was used for microscopy and image analysis.
RESULTS
Oocyte-Specific Deletion of the Mylk1 Gene
To obtain oocyte-specific Mylk1 mutant mice, we crossed Mylk1fl mice in which exons 23–25 were flanked by two loxP sites with transgenic mice expressing Gdf-9 promotor-driven Cre recombinase (Supplemental Fig. S1, A–C; all Supplemental Data are available online at www.biolreprod.org). In Gdf-9 Cre mice, Cre was expressed from primordial to later follicular stages. PCR genotyping of the pups of Mylk1fl/fl;GCre+ females mated with wild-type males indicated that recombination occurred in floxed oocytes (Supplemental Fig. S1D). Immunofluorescent analysis of MII oocytes from Mylk1fl/fl;GCre+ mice showed loss of MLCK localization in spindle poles, which indicated the deletion of MLCK protein (Fig. 1A). Immunoblotting and semiquantitative RT-PCR analysis further confirmed that the expression of MLCK in oocytes from Mylk1fl/fl;GCre+ mice was efficiently deleted at the mRNA and protein levels (Fig. 1, B and C). However, analysis of MLCK-deleted oocytes showed that pMLC2 (Ser19) localized in one or both spindle poles, which revealed that myosin II was activated in the absence of MLCK (Fig. 2).
FIG. 1
Verification of loss of Mylk1 in mouse oocytes. A) Immunofluorescence detection of MLCK loss in MII oocytes of Mylk1fl/fl;GCre+ mice. The superovulated MII eggs were stained for MLCK (white), and DNA was counterstained with DAPI (red). Boxed regions are magnified in the second images (red box). Colored arrowheads indicate the location of MLCK. Bar = 10 μm. B) Western blot detection of MLCK protein levels in oocytes fromMylk1fl/fl and Mylk1fl/fl;GCre+ mice. C) Deletion of Mylk1 in mutant and control oocytes was determined by semiquantitative RT-PCR. The 319 bp bands correspond to the Mylk1, and the 150 bp bands correspond to the Ppia.
FIG. 2
The presence of pMLC2 (Ser19) at one or two spindle poles at the oocytes of Mylk1fl/fl and Mylk1fl/fl;GCre+ females, respectively. MI and MII oocytes were collected and stained for pMLC2 (white), and DNA was counterstained with DAPI (red). Boxed regions are magnified in the second images (red box). Colored arrowheads indicate the location of pMLC2. Bar = 10 μm.
Mylk1 Mutant Females Displayed Subfertility, but Follicular Development Was Not Affected
To study the effect of oocyte-specific deletion of MLCK on fertility, we conducted a breeding assay. Wild-type males were mated with 6-wk-old Mylk1 mutant and control females. The results indicated that Mylk1 mutant females were significantly subfertile (Fig. 3A), generating on average 4.9 offspring per female compared with 10.3 offspring for control females (Fig. 3B). The subfertility could be due to ovarian dysfunction, resulting in functional oocyte loss. However, a comparison of ovarian morphology and the number of primordial and activated follicles in Mylk1 mutant and control females showed no significant difference. The results indicated that reduced fertility was not caused by follicular abnormality (Fig. 3, C and D).
FIG. 3
Mylk1fl/fl;GCre+ females displayed subfertility. A) Comparison of the accumulative number of pups per Mylk1fl/fl female and Mylk1fl/fl;GCre+ female. B) Comparison of the average number of pups per Mylk1fl/fl female and Mylk1fl/fl;GCre+ female (n = 6 for both groups). **Significant difference, P < 0.01. C) Histological images of ovaries of Mylk1fl/fl and Mylk1fl/fl;GCre+ females. Bar = 500 μm. D) Quantification of ovarian follicles in Mylk1fl/fl and Mylk1fl/fl;GCre+ females. Numbers of primordial and activated follicles per ovary were counted. Error bars denote SEM of three experiments.
MLCK Deletion Had No Effect on Meiotic Spindle Organization, Chromosome Alignment, and First Polar Body Extrusion
To determine whether the subfertility was caused by oocyte maturation abnormality, we collected MII eggs from Mylk1 mutant and control mice after injection of eCG (48 h) and hCG (13 h). The results revealed that Mylk1 mutant females could ovulate normally, and there was no significant difference in the number of superovulated oocytes from each group (Fig. 4A). Immunofluorescent results showed that the spindle was correctly positioned beneath the cortex and showed normal morphology (Fig. 4B). To evaluate whether the homologous chromosomes were segregated correctly, we performed chromosome spread on superovulated eggs. Centromeric protein Bub3 was stained to assist chromosome counting. Our results indicated that the oocytes had 20 pairs of sister chromatids and they were closely connected at centromeres (Fig. 4C), suggesting that homologous chromosomes were segregated accurately. These findings showed that MLCK-deficient oocytes could undergo normal first meiotic division to produce fully mature eggs.
FIG. 4
Evaluation of MII egg quantity and quality of Mylk1fl/fl females and Mylk1fl/fl;GCre+ females. A) Left: microscopic images of MII eggs from mutant and control females. Right: number of MII eggs recovered from Mylk1fl/fl females and Mylk1fl/fl;GCre+ females. Error bar denotes SEM of five experiments. Original magnification ×20. B) Left: confocal microscopic images of spindle morphology and chromosome alignment of control and MLCK-deficient eggs. Eggs were stained with α-tubulin antibody (green) and PI (red) to show spindle morphology and chromosome alignment, respectively. Right: the spindle status was assessed by its shape. Error bars denote SEM of three experiments. Original magnification ×400. C) Left: chromosome spread of control and MLCK-deleted eggs. Eggs were stained with Bub3 antibody (green) and PI (red) to assist chromosome counting. Right: ratio of MII eggs with euploidy. Error bars denote SEM of three experiments.
Fertilization and Embryonic Development up to the Morula Stage Were Not Affected by MLCK Deletion
To analyze if the subfertility was caused by fertilization or early embryonic development failure, we crossed Mylk1 mutant and control females with wild-type males. The zygotes were collected and cultured in vitro. The results revealed that MLCK-deficient oocytes could be fertilized successfully and form two pronuclei (Fig. 5A). Further culture showed that the zygotes could develop normally to the 2-cell (Fig. 5B) and morula stage (Fig. 5C). These data showed that MLCK-deficient eggs were able to undergo normal second polar body extrusion, fertilization, and development up to the compact morula stage.
FIG. 5
MLCK-deficient eggs could develop normally to compact morula stages in vitro. A) Left: MLCK-deficient eggs could naturally fertilize in vivo. Right: numbers of zygotes. B) Left: the fertilized eggs could develop normally to 2-cell stage embryos. Right: average rate of 2-cell development of MLCK-deficient females and wild-type females. C) Left: Mylk1 mutant zygotes could develop normally to compact morula stage. Right: average rate of compact morula development of Mylk1 mutant and wild-type females. The data are presented as the mean ± SEM of at least three experiments. Original magnification ×20.
MLCK Absence Delayed the Development from the Morula Stage to the Blastocyst Stage
When the 3.5 dpc embryos were flushed from the uterus, we found that an average of 36.6% embryos from Mylk1 mutant females (n = 5) were in the morula stage compared with only 6.4% morula stage embryos from control females (n = 6). In vitro culture showed that the delayed morula could develop to early blastocyst stage at Day 4, when most control blastocysts reached expanded blastocysts (Fig. 6A). For the embryos that developed normally to the blastocyst stage, the distribution of cells in trophectoderm and the inner cell mass, total cell number, and Oct4 expression pattern showed no abnormalities (Fig. 6B). Further immunofluorescent staining showed that MLCK was located in the nucleolus of blastomeres at the morula stage. However, no special signal was found in blastocyst (Fig. 7). These results suggested that MLCK might be involved in timely morula-to-blastocyst transition, which could be the reason for subfertility of Mylk1 mutant females.
FIG. 6
Delayed morula-to-blastocyst transition of mutant embryos. A) Left: microscopic images of 3.5 and 4 dpc embryos from Mylk1fl/fl and Mylk1fl/fl;GCre + females. Eight-week-old Mylk1 mutant and control female siblings were mated with wild-type males. Plugged females were sacrificed at 3.5 dpc. Embryos were collected by uterine flushing and then cultured in vitro to 4 dpc. Right: blastocyst rates were scored based on 3.5 and 4 dpc embryos from Mylk1fl/fl and Mylk1fl/fl;GCre+ females, respectively. B) Left: confocal microscopic images of blastocysts from Mylk1fl/fl and Mylk1fl/fl;GCre+ females. The collected blastocysts were stained for Oct4 (green). DNA was counterstained with DAPI (blue), and morphology was determined by differential interference contrast (DIC) microscopy.< Right: the average number of total cell in blastocysts from Mylk1fl/fl and Mylk1fl/fl;GCre+ females, respectively. The data are presented as the mean ± SEM of at least three experiments. Bar = 10 μm.
DISCUSSION
In this study, we investigated the roles of MLCK in oocyte meiotic maturation and preimplantation embryonic development by the Cre-LoxP system. Compared to the Mylk1fl/fl females, Mylk1fl/fl;GCre+ females exhibited a sharp reduction in litter size. Quantification of the ovarian follicles in Mylk1fl/fl and Mylk1fl/fl;GCre+ females indicated that follicular development was not affected in Mylk1fl/fl;GCre+ females. Analysis of superovulated eggs showed that the MLCK-depleted eggs could form normal spindles and undergo the separation of homologous chromosomes properly. Furthermore, MLCK-deleted oocytes could fertilize and develop normally to the compact morula stage. However, more than a third of the embryos showed delayed blastocyst development. Although these delayed embryos could develop to early blastocysts on Day 4, they may miss the implantation window. Our further experiments showed that MLCK was expressed in nucleolus of blastomeres at compact morula stages, which might be involved in morula-to-blastocyst transition. These findings implied that MLCK might be a maternal-effect gene that was transcribed during oogenesis and degraded in later embryonic development stages until the activation of the embryonic Mylk1 gene. We proposed that delayed morula-to-blastocyst transition might be the cause of the subfertility.
Several previous in vitro studies in which inhibitors were used indicated that MLCK might participate in mammalian oocyte maturation by controlling myosin II activity. Specifically, ML-7, an inhibitor of MLCK, can efficiently block polar body extrusion in vitro by slowing down spindle relocation during first meiosis [11, 27, 28]. Furthermore, both blebbistatin, an inhibitor of myosin II, and ML-7 block cortical cap protrusion [28]. In addition, ML-7 inhibits second polar body extrusion in a dose-dependent manner, and it reduces cortical granule exocytosis [5, 29]. Another study, however, indicates that neither ML-7 nor blebbistatin can disrupt chromosome movement to the cortex although blebbistatin completely inhibit polar body extrusion [30]. The results of these chemical inhibitor experiments are inconsistent and cannot exclude nonspecific or toxic effects of the inhibitors.
Our in vivo experiment results found that MLCK was located in spindle poles in MII eggs. Strikingly, deletion of MLCK in oocytes caused the loss of its localization, but pMLC2 was still present at the spindle poles. There could be several reasons for this phenotype. One reason is that MLCK deletion might be compensated by other potent myosin II activators. Another reason is that MLCK might not be required for myosin II activation, and the phenotypes displayed in ML-7-treated oocytes might not be caused by MLCK inhibition, but rather by nonspecific inhibition of other proteins or inhibitor toxicity. In summary, oocyte-specific deletion of MLCK does not affect oocyte maturation, fertilization, and early embryo development up to the morula stage, but delays morula-to-blastocyst transition, which may be the reason for reduced fertility.
ACKNOWLEDGMENT
We appreciate Shi-Wen Li, Li-Juan Wang and Hua Qin for their technical assistance. The Mylk1fl/fl mice are obtained from the Model Animal Research Center of Nanjing University, Nanjing, China.
