FACS-sorted putative oogonial stem cells from the ovary are neither DDX4-positive nor germ cells

Abstract

Whether the adult mammalian ovary contains oogonial stem cells (OSCs) is controversial. They have been isolated by a live-cell sorting method using the germ cell marker DDX4, which has previously been assumed to be cytoplasmic, not surface-bound. Furthermore their stem cell and germ cell characteristics remain disputed. Here we show that although OSC-like cells can be isolated from the ovary using an antibody to DDX4, there is no good in silico modelling to support the existence of a surface-bound DDX4. Furthermore these cells when isolated were not expressing DDX4, and did not initially possess germline identity. Despite these unremarkable beginnings, they acquired some pre-meiotic markers in culture, including DDX4, but critically never expressed oocyte-specific markers, and furthermore were not immortal but died after a few months. Our results suggest that freshly isolated OSCs are not germ stem cells, and are not being isolated by their DDX4 expression. However it may be that culture induces some pre-meiotic markers. In summary the present study offers weight to the dogma that the adult ovary is populated by a fixed number of oocytes and that adult de novo production is a rare or insignificant event.

Figure 1. DDX4 is a germline cytoplasmic marker in oocytes.

(A) DDX4 comparison in human and mouse. The DDX4^C25 antibody used for the isolation of the OSCs was raised against the C-terminus (in red) of H.s DDX4. In M.m this sequence shares 92% identity. (B,C) DDX4^C25 immunostaining in permeabilized (B) and non-permeabilized (C) fully-grown oocytes and ovarian somatic cells (a mixture of stroma and granulosa cells). Staining was only observed in the cytoplasm of permeabilized oocytes. It total we analysed 70 oocytes and 250 somatic cells taken from 14 ovaries. Scale bar: 20 μm.

Figure 2. Evaluation of fresh and cultured DDX4^C25-positive cells.

(A) FACS sorting of DDX4^C25-positive cells. Cells were firstly separated from debris; then, addition of DAPI for dead-live exclusion and detection of DDX4^C25-Alexa Fluor 633 allowed to isolate a small population of live cells correspondent to the putative OSCs (n = 10). (B) Confocal immunocytochemistry for germline markers (DDX4, PRDM1, DPPA3, IFITM3, DAZL) on freshly sorted DDX4^C25-positive cells, some of which were permeabilized before labelling. Chromatin was stained with DAPI. Scale bar: 20 μm. (C) Gene expression of freshly sorted and 2-months-old cultured DDX4^C25-positive cells (the ‘OSCs’), ovary, testis, or fibroblasts for Prdm1, Dppa3, Ifitm3, Ddx4, Dazl, Pou5f1, Stra8, Nobox and Zp3. Representative of 7 independent runs.

Figure 3. Evaluation of germline-specific markers in the oviduct.

(A) Adult oviduct sections permeabilized and immunostained for DDX4, PRDM1, IFITM3, DPPA3 and DAZL. Chromatin stained with DAPI. E = epithelial cells; M = muscular mucosa; L = lumen (asterisks mark site of insert). Scale bar: 50 μm (inserts 10 μm). (B) Gene expression analysis of fresh sections of oviduct for Prdm1, Dppa3, Ifitm3, Ddx4, Dazl, Nobox and Zp3. Representative of 3 independent runs.

Figure 4. DDX4^C25-positive FACS-sorted cells cannot be detected by another C-terminal DDX4³⁵¹ antibody.

(A) Schematic analysis of DDX4 in FACS-sorted cells. DDX4^C25-positive FACS-sorted cells, which recognize an external green triangle (DDX4 if really being sorted for DDX4) are then fixed, permeabilized and immunostained using DDX4³⁵¹ antibody. (B) Confocal immunohistochemistry using the DDX4³⁵¹ antibody on permeabilized FACS-sorted DDX4^C25-positive (the ‘OSCs’) and negative cells, oviduct and whole ovary. Chromatin stained with DAPI. E = epithelial cells; M = muscular mucosa; L = lumen. Asterisks mark site of insert. Scale bar: 50 μm (inserts 10 μm).