Spatial Characterization of Bioenergetics and Metabolism of Primordial to Preovulatory Follicles in Whole Ex Vivo Murine Ovary

Abstract

Previous work characterizing ovarian bioenergetics has defined follicular metabolism by measuring metabolic by-products in culture media. However, culture conditions perturb the native state of the follicle, and these methods do not distinguish between metabolism occurring within oocytes or granulosa cells. We applied the phasor approach to fluorescence lifetime imaging microscopy (phasor FLIM) at 740-nm two-photon excitation to examine the spatial distribution of free and protein-bound nicotinamide adenine dinucleotide hydride (NADH) during primordial through preovulatory stages of follicular development in fresh ex vivo murine neonatal and gonadotropin stimulated prepubertal ovaries. We obtained subcellular resolution phasor FLIM images of primordial through primary follicles and quantified the free/bound NADH ratio (relative NADH/NAD+) separately for oocyte nucleus and oocyte cytoplasm. We found that dynamic changes in oocyte nucleus free/bound NADH paralleled the developmental maturation of primordial to primary follicles. Immunohistochemistry of NAD+-dependent deacetylase SIRTUIN 1 (SIRT1) in neonatal ovary revealed that increasing SIRT1 expression in oocyte nuclei was inversely related to decreasing free/bound NADH during the primordial to primary follicle transition. We characterized oocyte metabolism at these early stages to be NADH producing (glycolysis/Krebs). We extended the results of prior studies to show that cumulus and mural granulosa cell metabolism in secondary through preovulatory follicles is mainly NADH producing (glycolysis/Krebs cycle), while oocyte metabolism is mainly NADH consuming (oxidative phosphorylation). Taken together, our data characterize dynamic changes in free/bound NADH and SIRT1 expression during early follicular development and confirm results from previous studies defining antral and preovulatory follicle metabolism in culture.

Keywords: FLIM; Krebs cycle; NADH; cumulus cells; folliculogenesis; glycolysis; granulosa cells; oocyte; oxidative phosphorylation; phasor approach to fluorescence lifetime imaging microscopy; sirtuin 1.

FIG. 1

Apparent oocytes in 740-nm two-photon excitation intensity images colocalize with oocyte-specific GFP fluorescence. A) Intensity (left) and FLIM images (right) of TGOG2 mouse ovaries with germ-line-specific GFP expression; 880-nm two-photon excitation specific for GFP and 740-nm two-photon excitation specific for NADH; 58-μm field of view. B) Phasor FLIM lifetime pixel distribution of corresponding images represented in 2D histograms. Color bar on the 740-nm NADH phasor 2D histogram represents the color code of free/bound NADH in the FLIM image. C) Intensity histogram of image pixels for GFP at 880-nm two-photon excitation (upper) and NADH at 740-nm two-photon excitation (lower), where X-axis is intensity and Y-axis is pixels. A low-level threshold (red vertical line on left of histogram) was applied to 880-nm GFP images to eliminate background, and no threshold was applied to 740-nm NADH images.

FIG. 2

Spatial NADH intensity and lifetime dynamics over the course of early stages of folliculogenesis. A) 740-nm two-photon excitation intensity images and lifetime FLIM images of primordial follicles, transitional follicles, and primary follicles imaged within PND 4–7 neonatal ovary. Minimal threshold applied to FLIM map reference images to highlight only follicle structures. Field of view from left to right: 60 μm, 30 μm, 38 μm, 38 μm, 46 μm, 60 μm, and 90 μm. B) Depiction of oocyte nucleus mask applied to unaltered/unthresholded images to quantify average free/bound NADH pixels per oocyte nucleus and resulting quantitation of mean ± SEM free/bound NADH for primordial, transitional, and primary follicle nuclei (P < 0.001, effect of follicle type by GEE; *P < 0.05 vs. primordial) transformed from C. C) Average free/bound NADH pixels per oocyte nucleus plotted as g versus s along the free/bound NADH axis; N = 22 follicles across nine animals. D) Depiction of oocyte cytoplasm mask applied to quantify average free/bound NADH and resulting mean ± SEM free/bound NADH pixels for primordial, transitional, and primary follicle cytoplasm (P = 0.020, effect of follicle type by GEE; †P < 0.05 vs. primary) transformed from E. E) Average free/bound NADH per oocyte cytoplasm plotted as g versus s along the free/bound NADH axis; N = 22 follicles across seven animals.

FIG. 3

SIRT1 immunostaining over the course of early stages of folliculogenesis in PND 7–9 neonatal ovary. A) ×20 magnification image of PND 7 neonatal ovary anti-SIRT1 immunostaining. B) Images of primordial follicles with oocyte nuclei immunonegative for SIRT1. C) Images of primordial follicles immunopositive for SIRT1. D) Images of transitional follicles with oocyte nuclei immunopositive for SIRT1. E) Images of primary follicles with oocyte nuclei strongly immunopositive for SIRT1. F) Proposed model for SIRT1 activity and oocyte nucleus NADH intensity depletion during the course of primordial follicle awakening. Cells are colored to indicate free/bound NADH ratio according to the color code in Figure 1C.

FIG. 4

Spatial NADH intensity and lifetime dynamics over the course of secondary through preovulatory stages of folliculogenesis. Phasor FLIM images of secondary (green closed arrows), antral (yellow closed arrows), and preovulatory (yellow open arrows) follicles imaged within ovaries dissected from females after eCG or eCG+hCG stimulation starting on PND 23, as described in Materials and Methods; 925-μm field of view for all images. Minimal thresholding of low-intensity pixels applied to clarify follicle structures and remove background pixels in follicle antrum.

FIG. 5

Free/bound NADH in oocytes and granulosa cell subpopulations over the course of secondary through preovulatory stages of folliculogenesis. Measurements were made 48 h post-eCG or 7–9 h post-hCG administration. A) Mean ± SEM free/bound NADH pixels for PND 23 secondary granulosa cells, antral cumulus granulosa cells, and preovulatory cumulus granulosa (P < 0.001, effect of group; *P < 0.05 for indicated intergroup comparisons). B) Mean ± SEM free/bound NADH pixels for PND 23 secondary granulosa cells, antral mural granulosa cells, and preovulatory mural granulosa (P < 0.001, effect of group; *P < 0.05 for indicated intergroup comparisons). C) Mean ± SEM free/bound NADH in mural compared to cumulus granulosa cells for antral and preovulatory follicles (P < 0.001, effects of GC type, follicle group, and GC type × follicle group interaction; *P < 0.05 compared to cumulus at same stage). D) Mean ± SEM free/bound NADH for PND 23 secondary, antral, and preovulatory follicle oocytes transformed from E (P < 0.05, effect of group; *P < 0.05 for indicated intergroup comparisons). E) Average phasor for individual secondary, antral, and preovulatory follicle oocytes plotted in g versus s; N = 57 follicles across nine animals. F) Model of free/bound NADH dynamics during primordial to preovulatory folliculogenesis in the context of the ovarian developmental niche. Cells are colored according to the color code in Figure 1C, with declining free/bound NADH from red to fuchsia to purple to blue to green. See text for details.

FIG. 6

Comparison of free/bound NADH ratio in the oocytes of follicles of different developmental stages in prepubertal gonadotropin-stimulated ovaries and neonatal ovaries. Oocyte free/bound NADH ratio data from Figures 2 and 5 are organized according to the type of follicle and condition of the measurement (neonatal ovary for primordial, transitional, and primary follicles and eCG or eCG+hCG stimulated prepubertal ovaries for secondary, antral, and preovulatory follicles). Free to bound NADH ratio (A) and position of the average of the data (B) in the phasor plot.