Oocyte generation in adult mammalian ovaries by putative germ cells in bone marrow and peripheral blood

Abstract

It has been suggested that germline stem cells maintain oogenesis in postnatal mouse ovaries. Here we show that adult mouse ovaries rapidly generate hundreds of oocytes, despite a small premeiotic germ cell pool. In considering the possibility of an extragonadal source of germ cells, we show expression of germline markers in bone marrow (BM). Further, BM transplantation restores oocyte production in wild-type mice sterilized by chemotherapy, as well as in ataxia telangiectasia-mutated gene-deficient mice, which are otherwise incapable of making oocytes. Donor-derived oocytes are also observed in female mice following peripheral blood transplantation. Although the fertilizability and developmental competency of the BM and peripheral blood-derived oocytes remain to be established, their morphology, enclosure within follicles, and expression of germ-cell- and oocyte-specific markers collectively support that these cells are bona fide oocytes. These results identify BM as a potential source of germ cells that could sustain oocyte production in adulthood.

Figure 1.. Spontaneous Regeneration of Oocyte-Containing Follicles in Adult Female Mice

Number of nonatretic primordial (A) or total immature (primordial, primary, small preantral [B]) follicles in ovaries of adult female mice at the indicated times following doxorubicin (DXR) injection (5 mg/kg). Follicle numbers in vehicle (VEH)- versus DXR-treated mice at 2 months (2 mos.) postinjection are also shown. Values are the mean ± SEM from an analysis of four to five mice per treatment group.

Figure 2.. Ovarian-Regulated Expression of Germline Markers in BM

(A) Germline marker expression in BM of adult wild-type female mice. As a positive control for expression of germline markers, adult mouse ovary RNA was analyzed in parallel. Oct4 mRNA was low in abundance and not consistently detected in all BM samples. L7, “housekeeping” gene. Mock, mock reverse-transcribed BM RNA samples. (B and C) Analysis of MVH immunoreactivity (red, with nuclei highlighted by propidium iodide in blue) in BM of adult wild-type female mice. (D) Mouse ovary analyzed in parallel as a positive control for the immunostaining shown in (B) and (C), demonstrating a restricted expression of MVH (red) to germ cells (oocytes). (E) Real-time PCR analysis of Mvh levels in BM or PB of adult female mice during the indicated stages of the estrous cycle. The data shown represent the combined results from an analysis of three to four mice per group, with mean levels at estrus set as the reference point for comparisons to other stages of the cycle following normalization of the data against β-actin for sample loading. For mice in estrus, Mvh expression in bone marrow was detected during linear amplification in only one of the three samples analyzed. See Table S1 for additional results. (F) Number of nonatretic primordial oocyte-containing follicles in adult female mice at the indicated stages of the estrous cycle (mean ± SEM, n = 4 mice per group). (G) Quantitative PCR analysis of Mvh levels in BM of adult female mice 2 weeks following ovariectomy (OVEX) without (no treatment, No Tx) or with subsequent supplementation of estrogen (E2, 100 ng/mouse) and/or progesterone (P4, 2 mg/mouse) for the final 24 hr before BM collection. The OVEX data shown represent the combined results from an analysis of three mice per group, with mean BM Mvh levels in ovary-intact female mice at estrus set as the reference point for comparisons (BM Mvh levels in ovary-intact metestrus stage female mice are also provided for reference) following normalization of the data against β-actin to control for sample loading.

Figure 3.. Properties of Putative Germ Cells in Adult Female BM

(A) Quantitative analysis of Mvh levels in crude BM and lineage-depleted (lin⁻) BM samples without or with further fractionation by FACS based on cell surface expression of Sca-1 or c-Kit. All remaining lin⁻ cells not represented in the Sca-1⁻/c-Kit⁺ cell fraction of the were pooled and analyzed together (Other). The data shown represent the combined results from an analysis of three adult female mice, with mean Mvh levels in the crude BM sample set as the reference point for comparisons following normalization of the data against β-actin for sample loading. See Figure S5 for additional data. (B) Germline marker expression in adherent BM-derived cells following a total of three serial passages (P3) over a 6 week period in vitro. BM, freshly isolated bone marrow. β-actin, housekeeping gene. Mock, mock reverse-transcribed BM RNA samples.

Figure 4.. BMT Rescues Oocyte Production in Chemotherapy-Sterilized Female Mice

(A) Number of nonatretic immature follicles in wild-type female mice 60 days after treatment with cyclophosphamide and busulfan on day 42 postpartum without or with BMT 1 or 7 days (d) later. Values are the mean ± SEM from analysis of five mice per group, with four of the five chemoablated mice not receiving BMT completely lacking immature oocytes. (B–D) Representative ovarian histology in adult female mice treated with vehicle and no BMT ([B], control), combination chemotherapy without BMT ([C]; CF, cystic follicle), or combination chemotherapy with BMT performed 7 days later (D), as described in (A). Asterisks denote corpora lutea in (B) and (D), whereas the arrowhead in (D) points to a large healthy oocyte contained within a mature antral follicle. See Figure S6 for additional examples and results. (E and F) Ovarian histology of adult wild-type female mice 11.5 months after combination chemotherapy followed by BMT on day 42 postpartum. Follicles at various stages of maturational development are highlighted (insets).

Figure 5.. Wild-Type BM Generates Oocytes in Atm Mutant Females

(A and B) Histology of adult wild-type (A) and Atm^−/− (B) mouse ovaries. Scale bars are provided to highlight large difference in ovary size between the two genotypes. (C and D) Oocyte-containing follicles in ovaries of chemoablated Atm^−/− female mice at 11.5 months following transplantation with wild-type BM at 6 weeks of age. Follicles in hatched black boxes are highlighted (insets). (E–G) Oocyte-containing follicle in an Atm mutant mouse 2.5 months after BMT, as viewed by MVH immunofluorescence (green; nuclear DNA highlighted by TO-PRO-3 in blue).

Figure 6.. Female PB Generates Oocytes following Transplantation

(A and B) Follicles containing GFP-positive (brown) oocytes in ovaries of adult Oct4-GFP transgenic mice ([A], multiple primordial oocytes are highlighted). Scale bar, 10 µm. (C) Oocytes (arrowhead or boxed) in a wild-type ovary prior to PBCT using Oct4-GFP (TgOG2) females as donors, showing a lack of GFP signal (inset, primordial oocyte). (D–F) Primordial follicles containing GFP-positive oocytes in ovaries of wild-type female mice 28–30 hr after PBCT, using adult TgOG2 transgenic females as PB cell donors (see also Figure 7). Scale bars, 10 µm. (G and H) GFP-positive primordial oocytes in ovaries of Atm-deficient females 30 hr after PBCT using adult TgOG2 transgenic females as donors. Scale bars, 10 µm.

Figure 7.

PBCT-Derived Ovarian Follicular Cells Express Germline and Oocyte Markers Dual immunofluorescence analysis showing coexpression of GFP (green) and MVH (red) (A–F), GFP (green) and HDAC6 (red) (G–L), GFP (green) and NOBOX (red, note the nuclear localization) (M–O), or GFP (green) and GDF9 (red) (P–R) in oocytes of immature follicles within ovaries of recipient female mice 28–30 hr after transplantation with PB harvested from adult Oct4-GFP (TgOG2) transgenic females (see Figure 6 for controls). Asterisks in (P) and (Q) denote auto-fluorescent red blood cells. All cell nuclei are highlighted by TO-PRO-3 iodide staining (blue) in the merged panels. Scale bars, 10 µm.