Mitochondrial function and intracellular distribution is severely affected in in vitro cultured mouse embryos

Studies of mitochondrial dynamics have identified an intriguing link between energy supply balance and mitochondrial architecture. This suggests that inappropriate culture conditions might inhibit mitochondrial functions, and affect embryonic development. Therefore, this study was conducted to determine whether in vitro culture (IVC) might affect mitochondrial function, distribution, organization (by Mitotracker Green), gene expression on RNA level (by qPCR), and protein expression and localization (by western blot and immunostaining) involved in regulation of mitochondrial functions. Mitochondria in 2-cell IVC embryos were less numerous compare to IN VIVO while the localization and distribution do not differ between the groups. Mitochondria of in vivo blastocysts formed elongated network along the cells, while in IVC were fragmented, rounded, and aggregated mainly in the perinuclear region. Additionally, mitochondria of IN VIVO embryos moved back and forth along their long axis on radial tracks, while in IVC blastocysts were much less active. mtDNA copy number in IVC blastocysts (92,336.65 ± 5860.04) was significantly lower than that of IN VIVO (169,103.92 ± 16,322.41; P < 0.02) as well as lower protein expressions responsible for mitochondrial fusion was observed in IVC blastocysts. Results indicate that in vitro culture affect on perturbations in mitochondrial number and function, which is associated with decreased developmental competence of in vitro produced mouse embryos.

www.nature.com/scientificreports/ Moreover, mitochondria exhibit a unique quality maintenance function. They have numerous stages of fusion and fission. If active, mitochondria maintain a polarised membrane, and fuse with other mitochondria to transfer components and maintain or improve the function of damaged or poorly performing organells 13 . However, if the mitochondria are non-functional and their membranes are depolarised, they do not fuse with active mitochondria and are targeted for removal via mitophagy. This process prevents the mixing of damaged and high-quality mitochondria and decreases the pool of poorly performing mitochondria 13 .
In mammals, three proteins are required for the fusion process. Two mitofusins: mitofusin 1 (Mfn1) and mitofusin 2 (Mfn2), are responsible for fusion of the outer membrane, while a single dynamic family member, Opa1, is required for inner membrane fusion. It is known that Mfn2, which plays such a central role in the fusion process, is essential for embryonic development 7 . Additionally, low expression of Mfn2 attenuates blastocyst formation rate and cleavage speed in mouse zygotes and causes mitochondrial dysfunction, as confirmed by ATP and mtDNA levels, and mitochondrial membrane potential 14 .
The present study was conducted to investigate the developmental effects of in vitro culture on mitochondrial distribution, functionality, motility, membrane potential, copy number of mtDNA, gene expression of mitochondrial fusion-related proteins and their localization.

Results
Effects of in vitro culture on developmental competence of mouse embryos. The purpose of the assessment was to evaluate whether in vitro culture (IVC) affects embryo development. To do that, mouse zygotes were cultured in standard IVC medium (KSOM) till E3.5 (to the early blastocyst stage). Same-day IN VIVO developed embryos were flushed, and blastocyst rate was evaluated in both groups. Results showed that IVC blastocyst formation rate (78%, 106/136) was lower than when developed IN VIVO (98%, 100/102; P < 0.05). Moreover, microscopical observation shown that quality of the IVC blastocysts was much lower than that of the IN VIVO control group (Fig. 1A). Moreover, the blastocysts were stained with Sox2 and Hoechst to distinguish the inner cell mass (ICM) and total cell number (Fig. 1B). Results shown differences between the   (Fig. 1D). These results confirm that in vitro culture negatively affects embryos development in terms of cell number, their distribution and pluripotency genes expression.
In vitro culture effects on mitochondria distribution and their activity in mouse embryos. We next evaluated the effects of IVC on mitochondrial distribution using MitoTracker Green. Results shown that mitochondria in 2-cell IVC embryos were less numerous compare to IN VIVO, while the localization and distribution do not differ between the groups. Mitochondria were homogenously distributed in both blastomers in both groups (Fig. 2, left column). Major differences were observed at the blastocyst stage (Fig. 2, middle and right column). The proper morphology of mitochondria in cells is tubular shape. Mitochondria in IN VIVO blastocysts were numerous and formed long extended tubular networks across the cells (Fig. 2, bottom pictures). Most of the mitochondria in IVC embryos (Fig. 2, upper pictures) had a spherical or oval shape, and do not create elongated network, we classified them as "fragmented" (Fig. 2, upper pictures, central column). While the mitochondria were visible as shorter and smaller have been considered as fragmented.
We next assessed mitochondrial membrane potential in cleavage-(2-cell) and blastocyst stage IVC and IN VIVO developed mouse embryos, using JC-9 staining. We have not observed any differences between the groups in 2-cell stage embryos (data not shown), while strong differences were observed in IVC blastocyst vs IN VIVO groups. Mitochondria with low membrane potential formed aggregates, mainly in the perinuclear regions, while highly polarized mitochondria localize to the periphery of the cell (Fig. 3A). The intensity ratio of red: green hence, mitochondrial membrane potential at the blastocyst stage was significantly lower in IVC embryos when compared to IN VIVO developed blastocysts ( Fig. 3B; P < 0.05).
Thus, our results indicate that IVC causes mitochondrial dysfunction that might explain the poorer developmental potential of embryos cultured in vitro.

Analysis of mtDNA copy number in IVC and in vivo-developed embryos, and cell number in blastocyst stage embryos.
To determine if in vitro culture affected the quantity of mtDNA present in embryos at the 2-cell and blastocyst stages, assessment of mtDNA copy number was performed (Fig. 3C). We found no difference in mtDNA copy number at the 2-cell stage embryos between IVC (171,845.52 ± 15,191 This observation shows that IVC affect not only the pattern of mitochondrial distribution and activity, but also mtDNA copy number. Effect of IVC on mitochondrial dynamics and motility in mouse embryos. Mitochondria are dynamic organelles, undergoing constant migration and morphological changes. Using time-lapse confocal microscopy, we searched for aberrations in mitochondrial dynamics in association with the embryonic development environment. Most mitochondria in embryos developed in vivo were tubular (Fig. 4A), directed radially, Selected mitochondrial genes and proteins involved in mitochondrial physiology and mitophagy. The level of selecetd genes knows to play important role in mitochondral physiology (Clpp, Hspe1 and mtND2) and mitophagy (BnipL3) were measured. Among the genes tested only BnipL3 was downregulated (P = 0.0018) while Hspe1 showed a trend for lower level (P = 0.053) in IVC groups compare to control (Fig. 6A). To evaluate whether in vitro culture affects/induce the mitophagy we assessed protein expression and localisation of mitophagy proteins (BnipL3 and p62) in IVC and IN VIVO blastocysts (Fig. 6B). Immunostaining analysis did not show differences in the expession and localization of BnipL3 (left column) in analysed groups while strong differences were observed in the expression of p62 protein. Results shown very low (close to background) expression of p62 in IVC embryos. In IN VIVO blastocysts p62 was strongly expressed but manily in the trophoblast cells ( Fig. 6B; second column from left).

Discussion
In the present study, we have demonstrated that in vitro culture of mouse embryos significantly affects the functionality of the mitochondria and might be responsible for the impaired embryonic development observed here and by others 7,15,16 In vitro culture of mammalian embryos is a very important tool when studying the early stages of embryogenesis and the environment in which early embryos can develop and form blastocysts. Moreover, proper embryo culture is one of the main factors ensuring that infertility programs are successful at the human clinics 1,17 . A number of reports have shown the influence of culture medium (oxygen concentration, medium components etc.) at the in vitro culture stage on embryonic development, fetal growth, and the postnatal health of the offspring 15,16,18 . www.nature.com/scientificreports/   www.nature.com/scientificreports/ The first observation that we have made was, that number of mitochondria in IVC 2-cell embryo was lower compare to IN VIVO one (Fig. 2). This alteration would very likely affect further embryonic development 19 . It is important to point out here, that the embryo depends on the function of existing mitochondria present at ovulation till blastocyst stage 20 . As cell division begins, the total number of mitochondria within each blastomere decreases due to dilution without new mitochondrial biosynthesis 21 . Consequently, any adverse influence on mitochondrial numbers and function will negatively impact the development of the pre-and post-implantation embryo. An implication of this finding is that as long as the segregation of mitochondria is homogenous between blastomeres, each cleavage from the zygote to the blastocyst halves the number of mitochondria per cell 17 . Moreover, it has been shown that homogeneously distributed mitochondria are found in normal in vivo delivered cleaving embryos, whereas in developmentally arrested embryos they aggregate in the perinuclear/pericortical regions, likely as a result of aberrant cytoskeletal and microtubular organization 4,8,21 .
Importantly, the progression through preimplantation development can be predicted not only by mitochondrial distribution within the cell but also by their number. In fact, we have confirmed that total number of mitochondria with active membrane have decreased in IVC embryos through their development from the 2-cell stage to the blastocyst formation, confirming previous reports in mouse model 23,24 . In human and mouse cleaving embryos, mitochondria with low membrane potential tend to aggregate in the perinuclear region, as was the case with the IVC blastocysts in our study (Fig. 3A). Highly polarized mitochondria, such as in the in IN VIVO blastocysts in our study (Fig. 3A,B), localize in the periphery of the cells, without direct contact with the neighbouring cell 4 . The IVC blastocysts in our work were characterized not only by a lower total number of active mitochondria, but also by a reduced copy number of mtDNA. Each mitochondria has 2-10 copies of mitochondrial DNA. mtDNA copy number is a critical component of overall mitochondrial health and is a measure of mitochondrial DNA (mtDNA) levels per cell, while not a direct measure of mitochondrial function, is associated with mitochondrial enzyme activity and adenosine triphosphate production.
Morphology of the mitochondria, maintained by a balance between their fision and fusion, is critical for proper cellular function. Numerous studies have suggested that an imbalance in mitochondrial fusion/fission might lead to lower blastocysts quality 25,26 . It is known that Mfn2 plays a central role in the fusion process and have been reported to also play an important role in embryonic development 7 . In this work, we showed low gene expression of Mfn1 and Mfn2 and of the protein itself in IVC embryos, which might suggest an unsettled fusion process. Interestingly, we have observed that fission process was not affected in our case. mRNA and well as protein of Opa 1 level was not differing between the groups. Furthermore, the time-lapse videos revealed striking alterations in mitochondrial mobility in the IVC blastocysts cells. Most mitochondria in the in IN VIVO blastocysts cells were tubular, directed radially, and moved back and forth along their long axis on radial tracks.
Since the mitochondrial quality control plays a central role in cellular homeostasis, a couple of factors involved in mitochondrial biogenesis (Clpp, Hspe1 and mtND2) and mitophagy (BNIP3L/Nix and SQSTM1/p62) were analysed. The results showed no statistical differences in Clpp, Hspe1 and mtND2 genes while interesting observations were obtained where mitopgahy proteins were studied. In the IVC blastocyst BNIP3L/Nix mRNA level was statistically lower when compared to the IN VIVO blastocysts (Fig. 6A). Then, immunostaining analysis showed no differences in the expression of BNIP3L/Nix. Drastic differences in the expression of p62 was observed in the both analysed groups. Moreover only trophoblastic cells express p62 protein while in the ICM cells very low positive signal was detected (Fig. 6B). Besides mitochondrial biogenesis, mitochondrial elimination via the autophagosome-lysosome pathway (mitophagy), plays a crucial role in maintaining the cellular steady state of mitochondrial functions. It has been shown that SQSTM1 (p62) localization plays a role in regulating mitochondrial morphology, genome integrity and mitochondrial import of a key transcription factor 27 . In vitro environment has a big impact not only on efficiency, number, fusion/fission and homeostasis of mitochondria but also for efficient removal of poor quality mitochondria in preimplantation mouse embryos.
As already has been mentioned before, efficient mitochondrial function and turnover depends on continuous structural remodeling through fusion and fission 28 . Daughter mitochondria produced by fission can either  www.nature.com/scientificreports/ maintain intact membrane potential or become depolarized. Depolarized mitochondria may restore their membrane potential and return to normal fusion/fission equilibrium or remain as non-fusing mitochondria to be eliminated by autophagy. It has been proposed that fission acts as an autophagic checkpoint 29 . Depolarized mitochondria change their morphology to a more fragmented form. Completion of the mitophagy pathway appears to be dependent on p62 30 . Mouse zygotic genome activation (ZGA) takes place at the 2-cell stage 31 . As can be expected, this process dependents on many structural and epigenetic changes of the maternal and paternal genomes, that are reprogrammed to facilitate development of the embryo 32 . Such major reprogramming of the genome requires metabolites, such as α-ketoglutarate (α-KG) that is essential for protein and DNA demethylation, acetyl-CoA that is required for protein acetylation, ATP that is required for phosphorylation of substrates, and UDP-GlcNAc needed for glycosylation 33 . Production of each of these depends on mitochondrial enzymes that are driving the tri-carboxylic acid (TCA) cycle, and utilization of pyruvate by pyruvate-dehydrogenase. Properly functioning mitochondria are necessary from the very beginning of the existence a new organism as an energy source of extremely dynamically developing organism (polarization, compaction, cavitation, differentiation as well as implantation).
Studies on mitochondrial metabolism have identified an intriguing link between energy demand/supply balance and mitochondrial architecture, suggesting that inappropriate culture conditions might inhibit mitochondrial functions, with negative effects on embryonic development. In our work, we have confirmed this hypothesis. We have shown, to our knowledge for the first time, that a culture medium commonly used for in vitro mouse embryo culture negatively affects mitochondrial functionality and motility. These abnormalities strongly influence embryonic development and quality. Our observations suggest that studies aimed at improving the quality of IVC media and culture conditions should consider the inclusion of metabolites/factors responsible for mitochondrial activity.

Materials and methods
Unless otherwise stated, all inorganic and organic materials were purchased from Sigma.

Assessment of cell number of IVC and IN VIVO embryos. IVC and IN VIVO embryos at the blasto-
cyst stage (ten embryos per group) were fixed with 4% paraformaldehyde (PFA), permeabilized in 0.1% Triton X-100, and then stained with 5 µg/mL of Hoechst 33,342. Stained blastocysts were visualized using a confocal microscopy (Nikon Eclipse Ti-E), using NIS-Elements Confocal software (Nikon). Images were analyzed using IMARIS 6.0.1 software (Bitplane AG, UK). Nuclei were identified using the 'spot' option in the software, with an estimated diameter of 7-10 μm in the Hoechst channel. The number of nuclei identified by the software was adjusted manually.
Mitochondrial staining. Mitochondrial distribution and segregation were quantified using Mitotracker Green staining. 2-cell and blastocysts were incubated with 100 µM of Mitotracker Green FM (Invitrogen, Molecular Probes, Milan, Italy) in M2 medium for 30 min at 37.5 °C. Then, embryos were washed twice with PBS, and then counterstained with 5 µg/mL of Hoechst 33342 for 10 min. Mitochondria distribution in embryo cytoplasm appeared as small individual fluorescent dots or as aggregates. All samples were examined by confocal microscopy (Nikon A1R), using NIS-Elements Confocal software (Nikon). Images were analyzed using IMARIS 6.0.1 software (Bitplane AG, UK).
Immunofluorescence staining. The zona pellucida was removed using acidified Tyrode's solution.
Embryos were fixed in 4% paraformaldehyde in phosphate-buffered saline (PBS) with 0.1% Tween 20 and 0.01% Triton X-100 for 10 min at room temperature (RT), permeabilised in 0.55% Triton X-100 in PBS for 15 min at RT, washed 4 times (5 min each) in PBX (PBS with 0.1% Triton X-100) and then blocked in 20% fetal The ND2 region was amplified from total DNA and visualized on 1.5% agarose gel. The PCR products were purified using the QIAquick Gel Extraction Kit (Qiagen, Hilden, Germany), according to the manufacturer's instructions to generate standards for the qPCR. Based on the DNA concentration measured with a Nanodrop c2000 system (Thermo Scientific, USA), serial tenfold dilutions of a DNA with a known concentration (standards) were generated. Each standard was used as a separate template for a qPCR reaction to produce the appropriate standard curve with the Light Cycler 96 software (Roche, Switzerland). The efficiency of the reaction was analyzed. Finally, each PCR run with experimental samples contained one DNA standard of a known concentration. For each run, a standard curve was generated from tenfold serial dilutions (10 −1 to 10 −8 ). Each sample was analysed in triplicate to obtain the mean number of mtDNA copies.

Western blot. Mitochondria were extracted from the IN VIVO and IVC blastocysts using Mitochondrial
Isolation Kit (Abcam, ab110168). The equal amount of protein samples was loaded and SDS-PAGE (10% sodium dodecyl sulfate polyacrylamide gel electrophoresis) was performed. After that, the proteins were transferred to a PVDF membrane (Merck-Millipore, USA) and blocked with Tris buffered saline containing 0.05% Tween-20 (PBST) and 5% non-fat dry milk for 1 h at RT. After blocking, the membranes were incubated with corresponding antibodies [Mfn1, Mfn2, Opa1 and β-actin (sc. 1615, Santa Cruz) or H3 (ab 1791, Abcam) which were used as a loading controls)] at 1:1000 at 4 °C overnight, washed in PBST, followed by incubation with the corresponding horseradish peroxidase-conjugated secondary antibodies for 1 h at RT. Visualization of the proteins was detected with ECL chemiluminescence (IN VIVO and IVC blastocysts: n = 100 for per lane).
Mitochondrial membrane activity. JC9 is a cationic dye which binds to mitochondria and emits green fluorescence (∼525 nm) independent of mitochondrial membrane potential (MMP). In the case of mitochondrial membrane hyperpolarization, JC9 aggregates and gives off red fluorescence (∼590 nm). Therefore, the intensity ratio of red: green can be used to measure the MMP and indicates the mitochondrial activity 25 .
IN VIVO and IVC embryos were loaded with M2 medium containing JC-9 dye (final concentration: 2 μg/ml) and incubated at 37 °C for 25 min. Then, embryos were washed twice with PBS, counterstained with 5 μg/mL Hoechst 33,342, and observed under confocal microscopy (Nikon Eclipse Ti-E), using NIS-Elements Confocal software (Nikon). After incubation blastocysts were live-imaged under a 63X objective lens with a zoom factor of 1.6X. Both green and red channel images were captured. ImageJ software was used to measure the intensity of green and red fluorescent signals and the ratio of red/green fluorescent intensity was calculated.
Statistical analysis. Data were analysed using GraphPad Prism (Version 6.01, GraphPad Software, Inc, CA, USA). Statistical analyses of total cell number, mtDNA copy number, and mRNA expression rate were based on three replicates per experiment and were compared by one-way ANOVA. Differences with P < 0.05 were considered statistically significant.