At conception, female mice fed an obesogenic diet had a 43% increase in body weight, 2.5 fold increase in fat pad and a significantly higher concentration of serum fatty acids (p<0.05) compared with chow fed controls (Table 1). Although maternal serum leptin concentrations were similar in obese and lean mice, the concentration of leptin in the oviductal fluid was significantly elevated in obese females in comparison with lean females (p<0.05).

Live cell dynamic fluorescence imaging was employed to study effects of maternal obesity on mitochondrial function in oocytes and embryos. This technique together with a range of targeted fluorescent probes permitted comprehensive evaluation of mitochondrial function with simultaneous measurement of multiple mitochondrial variables in a single oocyte and embryo. The common vital mitochondrial membrane–specific dyes; MitoTrackers and JC1 have been extensively used to study mitochondrial dynamics and function in oocytes and embryos. However, prolonged excitation of cells loaded with MitoTrackers may impair mitochondrial function [20] whereas the JC1 dye appears sensitive to factors other than inner mitochondrial membrane potential (Δψm) [21] and may inhibit mitochondrial complex 1 [22]. We therefore chose to measure Δψm in single oocytes and embryos using a low toxicity potentiometric fluorescent dye-tetramethyl rhodamine methyl ester (TMRM; [23]).

Maternal diet-induced obesity led to a dramatic increase in Δψm in oocytes and zygotes. The intensity of mitochondrial localised TMRM fluorescence in oocytes and zygotes of obese females increased by 147% (p<0.01) and 74% (p<0.01), respectively, compared with oocytes and zygotes of lean females (Figure 1A and 1B). Eggs from obese mice also had a different pattern of mitochondrial distribution as visualised by the distribution of TMRM staining (Figure 1C and 1D). Mitochondria were distributed evenly throughout the ooplasm of the eggs from lean females. In contrast, mitochondria in eggs from obese females had a discontinuous distribution with high density clusters localised to the cortical ooplasm and surrounding the nucleus. Since the organisation of mitochondria within cells is important for the signalling events associated with fertilisation [10], this abnormal localisation of mitochondria may be detrimental for the preimplantation embryo. Indeed, a similar aggregated mitochondrial structure was identified in preimplantation embryos undergoing arrest [24].

An increase in Δψm may have a number of origins including a higher supply of substrates and increased mitochondrial respiratory activity or inhibition of ATP synthase activity [25]. To differentiate between different mechanisms we analysed the redox state in oocytes and zygotes, measured as the autofluorescence signal derived from NAD(P)H or FAD²⁺ [23]. Confocal laser-scanning microscopy (CLSM) revealed that the NAD(P)H in oocytes and zygotes from obese mice was more oxidised than in eggs from lean mice (oocytes - 36%±3 vs 55%±4, p<0.05; zygotes – 27%±2 vs 86%±5, p<0.05 on a scale that runs from 0% for full oxidation to 100% for full reduction) (Figure 2A). Imaging FAD²⁺ autofluorescence also revealed that flavoproteins were more oxidised in oocytes and zygotes from obese females compared to eggs from lean females (oocytes – 74%±6 vs 41%±3, p<0.05, zygotes- 56%±4 vs 40%±4, p<0.05: note that for this signal, 100% is fully oxidised and 0%, fully reduced) (Figure 2B). These changes indicate increased oxidation of the pyridine nucleotide and flavoprotein pools [23] and suggest a shift in the intracellular redox status towards oxidation in the eggs from obese mice.

Studies in mitochondria respiratory chain function in vivo [26] and in vitro [11], [23] have shown that the redox state of the pyridine nucleotide and flavoprotein pools reflects the balance between the rate of reduction by substrate utilization and the rate of oxidation by mitochondrial respiration. The shift of the redox balance towards a net reduced state occurs as a consequence of up-regulated substrate processing and inhibition of respiration, whereas an increase in mitochondrial respiratory rate favours the shift of the redox potential towards the oxidised state. Therefore it is conceivable that oxidised state of the pyridine nucleotide and flavoprotein pools in oocytes and zygotes from obese females may be attributable to increased mitochondrial respiratory chain activity.

In order to evaluate the level of oxidative stress and antioxidant defence in oocytes and zygotes we measured rates of intracellular ROS generation using dihydroethidium (HEt), a non-fluorescent derivative of ethidium which is oxidised to a fluorescent product by superoxide. In oocytes and zygotes of obese mice the rate of ROS production was significantly increased by 2.1 (p<0.05) and 1.6 (p<0.05) fold respectively compared to the eggs of lean mice (Figures 3A and 3B). We measured levels of the antioxidant glutathione (GSH) using the monochlorobimane (MCB) which forms a fluorescent adduct following an enzyme catalysed reaction with GSH. In parallel with the increase in the oxidative load, GSH was depleted in oocytes and zygotes from obese females when compared with eggs from lean females (oocytes – 80%±7 vs 100%±6, p<0.05; zygotes- 68%±4 vs 93%±6, p< 0.05) (Figures 3C and 3D).

Alterations in the rates of intracellular ROS generation are associated with changes in mitochondrial abundance and mtDNA copy number. Oxidative stress damages bases as well as causing single or double-strand breaks in mtDNA which are mutagenic and can inhibit mtDNA replication [27]. However, excessive ROS generation has been associated with an increase in mtDNA copy number in aging tissues as a result of a feedback response which compensates for defective mitochondria bearing impaired respiratory chain or mutated mtDNA [28].

To test the hypothesis that maternal obesity-associated oxidative stress in oocytes and zygotes may affect mitochondrial biogenesis we measured mtDNA copy number and expression of key genes involved in the regulation of the replication and transcription of the mitochondrial genome. mtDNA copy number was significantly increased in oocytes from obese compared to lean mice (Figure 4A), and the expression of nuclear genes encoding mtDNA transcription factors - mtTFAM and NRF1 - was also elevated in oocytes from obese females suggesting upregulation of mitochondrial biogenesis (Figure 4B and 4C). Interestingly, mtDNA copy number, TFAM and NRF 1 expression were not altered in zygotes from obese mice.

No significant differences were detected in the number of zygotes between lean and obese females. However, the ability of zygotes to develop to the blastocyst stage was reduced in obese mice (Table 2). Thus, the number of obese females who failed to produce blastocysts 4 days after mating was higher in comparison with controls (5 vs 2). In three of 5 obese mice no blastocyst were found in either uterus and oviduct whereas in two obese females a small number of arrested fragmented embryos at various stages of development was recovered from the oviduct. In the remaining 6 obese (54%) females blastocyst development per mouse was similar to that in lean females. Differential nuclear labeling did not reveal any differences in the number of cells within trophectoderm and inner cell mass lineages of blastocysts from obese and lean mice (Table 2).

This study was conducted in accordance with the UK Home Office Animal (Scientific Procedures) Act 1986. Six week-old virgin females C57BL/6J mice (Charles River Laboratories, UK) were fed either a standard chow diet (7% simple sugars, 3% fat, 50% polysaccharide, 15% protein [w/w] RM1, Special Dietary Services, n = 10) or highly palatable obesogenic diet (10% simple sugars, 20% animal lard, 28% polysaccharide, 23% protein [w/w], Special Dietary Services, n = 10) supplemented with sweetened condensed milk and micronutrient mineral mix (AIN93G, Special Dietary Services) [19]. After 6 weeks of diet, all females were induced to superovulate by consecutive ip injections of 10 IU pregnant mare's serum (Dunlops) and 10 IU human chorionic gonadotrophin (hCG, Dunlops). Prior to embryo collection mice were culled by cervical dislocation. Blood samples were taken via cardiac puncture. Abdominal and inguinal fat pads and body weights were recorded. Twelve hours post hCG fully grown oocytes were collected by puncturing pre-ovulatory follicles with sterile needles and treated with hyaluronidase (0.5 mg/ml) to remove surrounding cumulus cells. Zygotes and blastocysts were collected after successful mating with males at 24 and 84 h after post-hCG as described [15].

Oviduct fluid was collected as described in [50] with minor modifications. Briefly, pregnant female mice (day 4) were culled and oviducts were ligatured at uterotubal junction with 6-0 black suture silk. Oviducts with attached ovaries were excised, rinsed in saline and placed under mineral oil on a watch glass. Ovaries were separated and the thread was removed at the cut end of the uterus under the microscope. A curved blunt-end metal capillary connected to a 1-ml syringe was inserted into infundibulum and oviducts were carefully flushed with 50 µl of flushing medium (PBS supplemented with 0.3% polyvinylpyrrolidone). Flushing from both oviducts was collected and placed into a sterilized 0.5-ml microcentrifuge tube and centrifuged at 10,000×g for 10 min to remove cellular debris, and the supernatant was aspirated into another sterilized 0.5-ml microcentrifuge tube and stored at −80°C until assayed for concentrations of leptin and glucose.

Mitochondria of living oocytes and zygotes were imaged using a Zeiss 510 uv-Vis CLSM META and a range of targeted fluorescent probes (Molecular Probes) as described previously [23], [51]. Some measurements were made using a cooled CCD camera (Orca ER).

The distribution of active mitochondria and Δψ_m were analysed in eggs incubated with tetramtehyl rhodamine methyl ester (TMRM; 25 nM) in M2 medium at 37°C for 30 min. The TMRM is a fluorescent lipophilic cation and accumulates into mitochondria in response to the negative mitochondrial membrane potential. TMRM was excited using the 543 nm laser line and fluorescence measured using a 560 long-pass filter.

Measurements of NAD(P)H and FAD²⁺ autofluorescence intensity in oocytes and zygotes were used to estimate the mitochondrial redox potential. The reduced forms of pyridine nucleotides (NAD(P)H) are excited by ultraviolet light at 351 nm excitation line of the CLSM and measured between 435–485 nm. The oxidised form (NAD(P)⁺) is non-fluorescent. In contrast to NAD(P)H, it is the reduced form of flavoproteins (FADH₂) that is non-fluorescent. Fluorescence of the oxidised form of flavoproteins (FAD²⁺) was excited at 458 nm and emitted fluorescence was collected throughout the 505–550 nm bandpass filter.

The resting redox state was defined as a function of the maximally oxidised and maximally reduced signals which were obtained by adding FCCP (1 µM) to drive the signals to maximal oxidation followed by addition of 1 mM NaCN which drives the signals to a maximally reduced state. The fluorescence signals are then normalised between 100 (maximal reduction to NAD(P)H and maximal oxidation to FAD²⁺) and 0 (maximal oxidation to NAD(P)⁺ and maximal reduction to FADH₂), giving a value which is a measure of the resting relative redox state [23].

For measurement of cytosolic ROS production, HEt (2 µM) was added to M2 medium and remained present throughout the experiment (15 min). This is a non-fluorescent derivative of the red fluorescent ethidium, and so an increase in red fluorescence (excited at 543 nm and measured at >560 nm) gives a measure of the rate of oxidation of the dye and therefore of the rate of ROS generation.

In order to measure GSH, cells were incubated with 50 µM MCB in M2 medium at 37° for 40 min, or until a steady state had been reached before images were acquired. Non-fluorescent MCB undergoes a reaction with glutathione catalysed by glutathione-s-transferase to yield a fluorescent adduct which therefore gives a measure of GSH content [52]. MCB fluorescence was excited at 351 nm and measured at 430–480 nm.

Fluorescent images were obtained from at least from 3 oocytes/zygotes from 5 females per diet. Image analysis, differentiation and exponential curve fitting were performed using Origin 8 software (OriginLab Corporation).

Total DNA was extracted from groups of two oocytes and zygotes (n = 8 females/group) as described [53]. A mtDNA content was measured by qPCR on Thermal Cycler Corbett Rotorgene ™ 6000 (Corbett research) using QuantiFast SYBR Green PCR Kit (Qiagen) and primers corresponding to 16S ribosomal gene. Sample copy number was determined with mtDNA standard curves using Rotorgene 6000 software.

Poly(A)+ RNA was isolated from snap-frozen oocytes (n = 20 oocytes/8 females/group) and zygotes (n = 10zygotes/8 females/group) using magnetic oligo(dT) beads (Dynabeads mRNA DIRECT Kit). cDNA synthesis was performed by random hexamer priming and the Transcriptor First Strand cDNA Synthesis Kit (Roche). qPCR took place on a Chromo4 thermocycler (MJ) using the Precision SybrGreen Master Mix (Primerdesign). Each assay was performed in duplicates using intron-spanning primers (Operon Biotechnologies GmbH). Stability of housekeeping genes was validated using geNorm application and H2afz mRNA was selected for normalization. Relative mRNA abundance was determined using the comparative deltaCt method. A list of genes investigated and Primer sequences are shown in Table 3.

The number of cells per blastocyst and their distribution between the inner cell mass and the trophectoderm were counted by differential fluorochrome nuclear labelling [54].

Fasted glucose, fatty acids and triglycerides concentrations were assessed using autoanalyser (LX20, Beckman Coulter) and the assay kits (Glucose; UV-hexokinase; nr. GLU 1442640; triglycerides; enzymatic GPO method; nr. TG 445850; total cholesterol; enzymatic method; nr. CHOL 467825) as described in [19]. Concentrations of leptin were measured by a sandwich ELISA using paired leptin antibodies (Duoset, R&D Systems Ltd). Serum and oviduct fluid samples from the mice were diluted 1∶20 for the assay and leptin concentration of samples was calculated from a standard curve constructed with mouse leptin standards.

All results are expressed as mean ± SEM. Data were analysed by Student's t-test after testing for normal distribution using Graphpad Prism v. 2.01 (Graphpad Software, USA). A value of p<0.05 was considered significant.