Fat metabolism is important in oocyte development (Kim et al. 2001; Sturmey et al. 2009). Our main objective was to develop a straightforward method that allows us to describe fat distribution and quantify fat accumulation in bovine oocytes throughout folliculogenesis. Fat accumulation dynamics can be studied by making microscopically repeated cross-sections or Z-stacks through oocytes of different developmental stages. The most important limiting factor in oocyte 3D visualisation studies is the size of the oocyte (between 110 and 120 μm in diameter), making it impossible to visualise it as a whole. If visualisation of the bottommost hemisphere is desirable, overlying Z-stacks will disperse the excitation and emission light, creating artefacts that will distort the image and trouble the results (indicated by former unpublished results). Therefore, we investigated the use of an alternative visualisation protocol staining the oocytes with Nile Red as an intracellular, triglyceride-specific (when emission is captured at 590 nm), fluorescent dye (Greenspan et al. 1985; Leroy et al. 2005), this combined with 2-photon excitation technology. In total, ovaries from 10 cows were collected at slaughter. Ovaries were pooled per cow. Only cows with apparent follicular activity on both ovaries were selected. Antral follicles with a diameter of <3 mm, between 3 and 6 mm, and >6 mm were aspirated. Apart from oocytes collected from each of these 3 follicle classes, preantral follicles were harvested from ovarian cortex tissue through mechanical isolation and enzymatic digestion by collagenase type IA. In total, 3 oocytes per follicle class were collected for each pair of ovaries. They were fixed in 2% glutaraldehyde and 2% formaldehyde and stained with Nile Red (1 μg mL–1). After mounting, images were acquired with a Zeiss LSM 510 reversed 2-photon microscope (Carl Zeiss GmbH, Jena, Germany) using an excitation wavelength of 807 nm, with an emission spectrum covered between 580 and 591 nm. Our results show that even with biphoton excitation, it remains very difficult to visualise fat droplet distribution in >100-μm oocytes. However, accurate images can be obtained of the upper hemisphere of the oocyte. These images can be used in future research on the dynamics of the distribution and the accumulation of lipid. Preliminary descriptive results clearly show that the relative amount of lipid droplets is lower and their size is smaller in oocytes from preantral follicles compared with antral counterparts. We can preliminary conclude that fat accumulation and the aggregation in fat droplets might take place preceding or even during antral development. We can visualise preantral oocytes as a whole in contrast to antral oocytes where only the upper hemisphere is visible without distortions. Image analysis software is currently applied to allow for a more quantitative interpretation.
D. De Rijck.
The Old and the New: Can Physics-Informed Deep-Learning Replace Traditional Linear Solvers?