Increase in Histone Methylation in the Cat Germinal Vesicle Related to Acquisition of Meiotic and Developmental Competence

Authors


Author's address (for correspondence): TC Phillips, Center for Species Survival, Smithsonian Conservation Biology Institute, National Zoological Park, Washington, DC 20008, USA. E-mail: phillipst@si.edu

Contents

This study identified specific changes in histone lysine methylation patterns of the feline germinal vesicle (GV) during pre-antral-to-antral follicle transition, the latter being a key interval for competence acquisition. Oocytes from adult cats were isolated from pre-antral, early (≤0.5 mm diameter), small (0.6–1 mm) or large (1–3.5 mm) antral follicles and immuno-stained with anti-histones H3 trimethylated at lysine 9 (H3K9me3), lysine 4 (H3K4me3), lysine 27 (H3K27me3) or H3 dimethylated at lysine 79 (H3K79me2). The vast majority of oocytes (range, 72.2–85.4%; p > 0.05) contained a GV with H3K9me3 or H3K27me3, regardless of follicular stage/size. However, the proportion of GVs with H3K4me3 or H3K79me2 was higher in early antral follicles (42.6%; p < 0.05) compared with other stages (range, 12.1–15.2%). Therefore, H3K4me3 and H3K79me2 (both known to be associated with selective gene activation) appear to be reliable markers of onset of GV competence during the pre-antral-to-antral transition phase. By contrast, H3K9me3 and H3K27me3 (both known to be related to selective gene repression) seem more linked to expression patterns during the GV stage and are less useful indicators during the entire folliculogenesis interval.

Introduction

Intrafollicular oocytes in the mammalian ovary are arrested at the first meiotic prophase and contain a nucleus called the germinal vesicle (GV). Progression of the follicle from the primordial to the antral stage is a serial process that empowers the oocyte to resume meiosis after the LH surge and then support embryo development after ovulation and fertilization (Zuccotti et al. 2011). Based largely on the studies of domestic and laboratory animals, it is known that oocyte competence (ability to fertilize and develop into an embryo) is influenced primarily by epigenetic factors that control overall gene expression while significantly modifying GV chromatin configuration (Zuccotti et al. 2011). The oocyte's capacity for meiotic resumption occurs at the early antral follicular stage, whereas developmental competence is acquired during final growth of the antral follicle (as determined in the mouse, Zuccotti et al. 2002; pig, Wu et al. 2006; and cow, Lodde et al. 2007). Recent studies of the domestic cat in our laboratory have suggested that the changeover from pre-antral to the antral follicle stage is critical for permitting the GV to acquire both meiotic and developmental capabilities (Comizzoli et al. 2011). It is apparent during this transition that GV chromatin configuration changes modestly from a filamentous to a reticular pattern, whereas overall gene expression is still detected in the entire nucleus (Comizzoli et al. 2011).

It is doubtful that histone acetylation/deacetylation is a critical epigenetic factor involved in gene expression at this time because histones remain highly acetylated in cat GV oocytes (Comizzoli et al. 2009). However, histone methylations are interesting indices of acquisition since these events may occur independently of the change in GV chromatin configuration, as has been observed in the mouse (Kageyama et al. 2007; Inoue et al. 2008). Also, in contrast to what has been observed for acetylation or even phosphorylation in multiple species (i.e. mouse, pig and sheep), global histone methylation appears to be consistently expressed during oocyte growth (Gu et al. 2010). Therefore, we speculated that there may be another link among chromatin configuration, histone methylation and oocyte competence in the domestic cat.

To date, different lysine methylation residues on histone H3 have been examined in oocytes of several species (H3K4me3, mouse, Kageyama et al. 2007; H3K9me3, sheep, Hou et al. 2008; H3K79me2, mouse, Ooga et al. 2008; H3K27me3, pig, Park et al. 2009). In the mouse GV, the fluorescence signal for H3K9me3 co-localizes uniformly with the DNA signals and is believed to be involved in suppressing gene expression (Kageyama et al. 2007). H3K27me3 is another epigenetic marker that is tightly associated with selective transcriptional repression and in the pig is detectable in the nuclei of GV stage oocytes and the chromosomes of metaphase II (MII) oocytes (Park et al. 2009). However, methylation of histone H3 at K4 is known to be part of selective gene activation (Dambacher et al. 2010). Kageyama et al. (2007) have observed an increased H3K4me3 fluorescence signal coinciding with oocyte growth in the mouse. Additionally, H3K79me2 has been identified as a key activation transcriptional marker during mouse folliculogenesis (Ooga et al. 2008).

Based upon the known presence of H3K9me3, H3K27me3, H3K4me3 and H3K79m3 in these other species and the existing information on acquisition of GV competence in the cat (Comizzoli et al.2011), we conducted a study to test the hypothesis that patterns of histone lysine methylation change from the pre-antral through early antral stages of follicular development. Our objective was to identify specific modifications in the epigenetic pattern of the domestic cat GV during the key interval of competence acquisition that occurs independently of a morphological change in chromatin configuration.

Materials and Methods

Collection and classification of oocytes from diverse follicles

Ovaries from adult domestic cats were recovered after routine ovariohysterectomy at local veterinary clinics and transported in phosphate-buffered saline (PBS) at 4°C to the laboratory within 6 to 12 h of excision. Cumulus–oocyte complexes (COCs) were collected from large antral follicles (1–3.5 mm in diameter) by slicing the follicles visible at the ovarian surface with a scalpel blade in Hepes-buffered minimum essential medium (H-MEM; Gibco Laboratories, Grand Island, NY, USA.) supplemented with 1 mm pyruvate, 2 mm l-glutamine, 100 IU/ml penicillin, 0.1 mg/ml streptomycin and 4 mg/ml bovine serum albumin (pyruvate, l-glutamine, penicillin, streptomycin and bovine serum albumin; Sigma-Aldrich, St Louis, MO, USA).

Each COC was classified according to standard quality criteria (Wood and Wildt 1997) as Grade 1 (uniformly dark cytoplasm, ≥5 compact layers of cumulus cells) or Grade 2 (same as Grade 1, but with 5 cell layers). Lesser grades (partially denuded oocytes) were discarded. Remaining antral follicles (smaller and deeper within the ovarian cortex) were dissected from the ovarian tissue with a needle and forceps.

Immunostaining of the GV

The oocyte of each follicle was denuded from cumulus cells by immersion in 0.2% hyaluronidase (Sigma-Aldrich) for 30 min. Oocytes then were fixed in 4% paraformaldehyde in PBS solution (4% paraformaldehyde in PBS solution; USB Corporation, Cleveland, Ohio, USA) for 30 min at 38°C and incubated with a polyclonal anti-histone H3K9me3, H3K4me3 or H3K79me2 primary antibody [diluted 1:100 in PBS overnight at 4°C, anti-histone H3 (tri methyl K9), anti-histone H3 (tri methyl K4), anti-histone H3 (di methyl K79); Abcam, Cambridge, MA, USA], or polyclonal anti-histone H3K27me3 primary antibody [diluted 1:500 in PBS overnight at 4°C, anti-histone H3 (tri methyl K27); Millipore, Billerica, MA, USA] according to published reports (Kageyama et al. 2007; Ooga et al. 2008; Park et al. 2009) or manufacturer recommendations. A negative control group (absence of primary antibody) was used for each replicate. After three washings (15 min each) in PBS, oocytes were incubated with a secondary anti-rabbit IgG labelled with FITC (secondary anti-rabbit IgG labelled with FITC; Sigma-Aldrich) diluted 1:100 in PBS for 1 h at 38.5°C. GV chromatin then was stained with a PBS solution containing 5 mg/ml propidium iodide (propidium iodide; Sigma-Aldrich) plus 10 μg/ml Hoechst 33342 (Hoechst 33342; Sigma-Aldrich) in (Vectashield®; Vector Laboratories, Burlingame, CA, USA) mounting medium.

The diameter of every oocyte and GV, as well as presence or absence of fluorescence reflecting GV histone methylation status was measured under phase-contrast microscopy (Olympus BX41; Olympus Corporation, Melville, NY, USA) using SPOT Advance® Software 3.5.9 (SPOT Advance® Software 3.5.9; Diagnostic Instruments, Inc., Sterling Heights, MI, USA). Each oocyte and its GV then were classified as to their follicle's size based on our recent data (Comizzoli et al. 2011) for oocyte and GV diameters of pre-antral follicles average (±SD) 73 ± 4 μm and 24 ± 3 μm, respectively; early antral follicles, 94 ± 4 and 31 ± 2 μm; small antral follicles, 104 ± 4 μm and 36 ± 4 μm; and large antral follicles, 115 ± 4 μm and approximately 40 ± 3 μm.

Experimental design and statistical analysis

Denuded cat oocytes (n = 374 total in 5 replicates conducted in parallel) were allocated to an H3K9me3, H3K27me3, H3K4me3 or H3K79me2 primary antibody exposure group (including a negative control group). Percentage data were analyzed in a 2 × 2 chi-square test. Differences were considered significant at p < 0.05.

Results

The vast majority of GVs (range, 72.2–85.4%; p > 0.05) contained the suppressive methylated histones H3K9me3 or H3K27me3, with the proportions generally unrelated to follicle stage (Table 1). The one exception was that the incidence of H3K9me3 labelling was less in the large compared with all other counterpart size follicles (Table 1). In general, for this type of histone methylation, the staining was detectable throughout the entire GV (Fig. 1a–d).

Figure 1.

Oocyte germinal vesicles from early antral follicles. Chromatin was stained with propidium iodide (a, c, e, g) to visualize the DNA and immuno stained (FITC, green) for histone methylation H3K9me3 (b), H3K27me3 (d), H3K4me3 (f), H3K79me2 (h). Scale bar = 10 µm

Table 1. Incidence of histone methylation related to selective gene suppression (H3K9me3 and H3K27me3) in the germinal vesicle (GV) of oocytes collected from diverse follicular stages
Follicular origin of oocytesProportions of positive GVs (%)
nH3K9me3nH3K27me3nCombined
  1. Values with different superscripts differ (p < 0.05).

Preantral 3571.4ab1376.9a4872.9
Early antral2277.2a1586.7ab3781.1
Small antral2272.7ab2696.2b4885.4
Large antral2548.0b2993.1ab5472.2

The proportions of GVs labelled with markers of the activated genes H3K4me3 and H3K79me2 were more variable (Table 2), but generally lower than the incidence measured with the suppressive genes (Table 1). When the results for H3K4me3 and H3K79me2 were combined, the level of histone methylation ranged from only 12.1 to 15.2% (Table 2). The exception was for the higher (p < 0.05) incidence (42.6%) occurring during the early antral follicular stage (Table 2). Although not always statistically significant, this same trend was evident when examining the data separately for H3K4me3 and H3K79me2 for the early antral compared with all other follicle stages. Again, the staining process revealed methylation labelling throughout the entire GV (Fig. 1e–h).

Table 2. Incidence of histone methylation related to selective gene activation (H3K4me3 and H3K79me2) in the germinal vesicle (GV) of oocytes collected from diverse follicular stages
Follicular origin of oocytesProportions of positive GVs (%)
nH3K4me3nH3K79me2nCombined
  1. Values with different superscripts differ (p < 0.05).

Preantral316.5a1822.2ab4912.2a
Early antral3040.0b1747.1a4742.6b
Small antral1717.6ab1612.5bc3315.2a
Large antral2920.7ab293.4c5812.1a

Discussion

We observed that methylation of the histones H3K9 and H3K27 generally occurred in >70% of domestic cat GVs regardless of pre-antral through most antral follicular stages. By contrast, although there was a far lower incidence of methylation of histones H3K4 and H3K79, clearly this phenomenon was more prevalent at the early antral stage, a period known to coincide with the acquisition of meiotic and developmental competence (Comizzoli et al. 2011). We suspect that genomic regions that had histones H3 methylated at K9 and K27 were not expressed and probably corresponded to genes that were unnecessary during folliculogenesis. In contrast, the enhanced methylation of the H3K4 and K3K79 only during the early antral stage leads us to suggest that this genomic area and the respective genes may be involved in the GV's ability to resume meiosis and sustain successful embryo development. Even so, it was noteworthy that this expression occurred in <45% of the GVs at this follicle stage, thereby indicating that less than half the oocytes achieved this critical step.

Our findings in the cat revealed differences and similarities to the results of earlier studied mammalian species. For example, it has been demonstrated that DNA-methylated H3K9me3 is specifically localized in sheep GV oocytes from large antral follicular stage (Hou et al. 2008). For the cat, it was the contrary because the GV from the largest antral follicles actually had the lowest proportion of this type of methylation. Additionally, our previous studies demonstrated that global genome transcriptional activity (via a bromo-UTP assay) consistently ceased at this advanced stage of the folliculogenesis (Comizzoli et al. 2011). Therefore, it was intriguing that we observed a decline in H3K9me3 (i.e. the less selective gene suppression) at that stage in this study. This revealed some refined genomic regulation that was not detected earlier in the bromo-UTP assays and requires more investigation. Nonetheless, the other epigenetic factor related to selective gene suppression (H3K27me3) was consistently detected in the cat GV and regardless of follicular stage, similar to what has been reported for the pig oocyte (Park et al. 2009). In terms of the methylation of the H3K4, Kageyama et al. (2007) have reported increasing levels during oocyte growth in the mouse. These investigators also surmised that this progressive methylation during folliculogenesis may be involved in genome-wide alterations of chromatin configuration but not necessarily linked to transcriptional activities. Interestingly, the cat GV did not follow that pattern because the highest incidence of methylated H3K4 was only observed at the early antral follicular stage. Additionally, although H3K79me2 has been determined to be present in the mouse GV at all oocyte stages (Ooga et al. 2008), this phenomenon was different in the cat model where the highest incidence of methylation was detected at the time of early antral formation. Therefore, we assert that the methylations of H3K4 and H3K79 are two epigenetic factors that play a more fundamental and likely more important role in the cat than the mouse GV.

Our collective results highlighted the importance of the cat GV as a useful, comparative model for better understanding the critical events related to acquiring meiotic and developmental competencies. Methylations of H3K4me3 and H3K79me2 appeared to be the potential indicators of the onset of GV biological capacity at the pre-antral-to-antral transitional phase in the cat. By contrast, H3K9me3 and H3K27me3 were ineffective markers because these indices of histone expression were fully methylated in most GVs regardless of follicular stage. Further analysis of the epigenetic patterns requires further investigation because each histone methylation immuno-labelling was observed in the entire nucleus and not necessarily in specific areas of the genome.

These types of fundamental studies are important because the GVs living in all follicles, from the primordial to the large antral stage, represent an untapped source of genetic material that mostly is wasted during the animal's lifetime and then at death. Female fertility preservation will be of growing interest as advanced tools and technologies are developed for rescuing the genome from oocytes, even those existing within immature, intraovarian follicles. A key to success will be to know how to provoke the acquisition of the full competence to achieve fertilization and the formation of a viable embryo. Based on these data, it appears that the domestic cat GV, with selected histone methylation during a critical step of folliculogenesis, offers an important model for exploring epigenetic factors that will eventually permit effective and efficient preservation of the maternal genome.

Acknowledgements

We thank Drs. Brent Whitaker and Michael Cranfield (Maryland Line Animal Rescue) and Darby Thornburgh (Petworth Animal Hospital) for providing domestic cat ovaries. This project was funded by the National Center for Research Resources (R01 RR026064), a component of the National Institutes of Health (NIH) and is currently supported by the Office of Research Infrastructure Programs/Office of the Director (R01 OD 010948).

Conflicts of interest

None of the authors have any conflicts of interest to declare.

Author contribution

TCP designed and conducted the experiments, analyzed the results and wrote the manuscript. PC and DEW assisted in the experimental design, the data analysis and interpretation, and the manuscript writing.

Ancillary