Animal ethics

Female (21–23 days) and male (6–8 weeks old) CBA × C57BL/6 first filial (F1) generation (CBAF1) as well as female (6–8 weeks old) Swiss mice were obtained from Laboratory Animal Services (LAS University of Adelaide, SA, Australia) and maintained on a 12 h light:12 h dark cycle with rodent chow and water provided ad libitum. All experiments were approved by the University of Adelaide's Animal Ethics Committee and were conducted in accordance with the Australian Code of Practice for the Care and Use of Animals for Scientific Purposes.

Media

All gamete and embryo culture took place in media overlaid with paraffin oil (Merck Group, Darmstadt, Germany) in a humidified atmosphere of 5% O₂, 6% CO₂ with a balance of N₂ at 37°C unless stated otherwise. All media were preequilibrated for at least 4 h prior to use. All handling procedures were performed on microscopes fitted with warming stages calibrated to maintain the media in dishes at 37°C.

The base medium used for handling and culture was phenol red-free Minimum Essential Medium Eagle (MEM-E). For collection and imaging of immature COCs, ovaries were collected in handling medium comprised of MEM-E supplemented with 6 mM NaHCO₃, 50 mg/L gentamicin sulfate, 10 mM HEPES, 2 mM glutamax (Gibco by Life Technologies, CA, USA), and 3 mg/mL fatty acid-free bovine serum albumin (BSA, MP Biomedicals, Albumin NZ, Auckland, NZ). The media was filtered before adding 1 × 10^–2 µM estradiol [43, 44] and 50 µM 3-isobutyl-1-methylxanthine [45]. The handling medium for ovary collection and handling of COCs prior to IVM was filtered MEM-E supplemented with 6 mM NaHCO₃, 50 mg/L gentamicin sulfate, 10 mM HEPES, 2 mM glutamax, and 5% (v/v) fetal calf serum (FCS) [46]. The culture medium for IVM was MEM-E supplemented with 26 mM NaHCO₃, 50 mg/L gentamicin sulfate, 2 mM glutamax, 5% (v/v) FCS, 50 mIU/mL of recombinant human follicle-stimulating hormone (Los Angeles Biomedical Research Institute, LA, USA) and 3 ng/mL epidermal growth factor, hereafter referred to as “IVM medium” [9, 46, 47]. Research Wash Medium, Research Fertilization Medium and Research Cleave Medium (ART Lab Solutions, SA, Australia) supplemented with 4 mg/mL of low fatty acid BSA were used for IVF and embryo culture, respectively.

Collection of immature cumulus oocyte complexes

Mice were injected subcutaneously with 5 IU equine chorionic gonadotrophin (eCG, Braeside, VIC, Australia). At 44 h post-eCG, mice were culled by cervical dislocation and ovaries collected in warmed handling medium. The COCs were isolated from ovaries by puncturing follicles using a 29-gauge × ¹/₂ in. insulin syringe with needle (Terumo Australia Pty Ltd., NSW, Australia). Isolated immature COCs were used for IVM or for imaging using two-channel laser scanning confocal microscopy to measure the optical redox ratio.

In vitro maturation

Immature COCs were cultured in groups of 20 in drops of 100 µL IVM medium overlaid with paraffin oil. Mature COCs (12 h post IVM) were used for imaging using two-channel laser scanning confocal microscopy to measure the optical redox ratio. A separate cohort of COCs were matured in the absence or presence of etomoxir (100 µM; 16 h) [9]. Maturation of COCs occurred within 20% O₂, 6% CO₂ with a balance of N₂ at 37°C. Following IVM, 12 h COCs were treated with metabolic inhibitors/uncoupler and imaged using two-channel scanning confocal microscopy. COCs matured for 16 h were either imaged (two-channel scanning confocal microscopy or hyperspectral microscopy) or fertilized in vitro.

In vitro fertilization and embryo culture

One hour prior to IVF, male mice with proven fertility were culled by cervical dislocation with the epididymis and vas deferens collected in Research Wash Medium. Spermatozoa were released from the vas deferens and caudal region of the epididymis by blunt dissection in 1 mL of Research Fertilization Medium and allowed to capacitate for 1 h in a humidified atmosphere of 5% O₂, 6% CO₂ with a balance of N₂ at 37°C. Mature COCs were then co-cultured with capacitated spermatozoa (35 000 sperm/mL) for 4 h at 37°C in a humidified atmosphere of 5% O₂, 6% CO₂ with a balance of N₂ [48, 49]. The resulting presumptive zygotes were cultured in groups of 10–12 in 20 µL drops of Research Cleave Medium overlaid with paraffin oil in a humidified atmosphere of 5% O₂, 6% CO₂ with a balance of N₂ at 37°C. Embryos were assessed for on-time morphological development on day 2 (2-cell; 24 h post-IVF) and day 5 (blastocyst-stage; 96 h post-IVF). The rate of development to the 2-cell and blastocyst stages was calculated from the initial number of oocytes. Two-cell-stage embryos were identified by the presence of two regular blastomeres of equal size, while blastocysts were identified by the presence of a blastocoel cavity ≥ two-thirds the size of the embryo; or expanded; or hatching.

Use of metabolic inhibitors

Mitochondrial inhibitors (oligomycin, carbonyl-4-phenylhydrazone (FCCP) and Rotenone/antimycin A (Rot/AA)) used for this study were from the Seahorse XF Cell Mito Stress Test Kit (Agilent Technology, CA, USA). Inhibitors were dissolved as per manufacturer's instructions in either MEM-E (for experiments where metabolic cofactors were measured by optical imaging) or Seahorse XF DMEM medium (measurement of oxygen consumption rate (OCR) by extracellular flux assay; Agilent Technology, CA, USA) and stored in –80°C. One hour prior to imaging or extracellular flux analysis, inhibitors were diluted to required concentrations with pre-warmed MEM-E or DMEM.

The optimal dose for each mitochondrial inhibitor/uncoupler was determined. Doses chosen for analysis ( Supplementary Figure SS1 (supp_fig_assessment_of_oocyte_quality_ioac145.docx)) were based on manufacturers' instructions and prior literature [41]. The final concentrations used for this study were 2.0 µM oligomycin; 1.0 µM FCCP, and 2.5 µM Rot/AA, as these elicited an appropriate response in oxygen consumption in COCs ( Supplementary Figure SS1 (supp_fig_assessment_of_oocyte_quality_ioac145.docx)).

Measurement of oxygen consumption rate in immature cumulus oocyte complexes using extracellular flux analysis

The base medium used for extracellular flux analysis was Seahorse XF DMEM medium, supplemented with 1 mM pyruvate (Agilent Technology, CA, USA), 2 mM glutamine (Agilent Technology, CA, USA) and 10 mM glucose (Agilent Technology, CA, USA). Immature COCs were isolated in pre-warmed handling medium as described above and placed into wells at a density of 20 COCs/well. One hour prior to the assay, COCs were washed three times in Seahorse XF DMEM before being replaced with fresh Seahorse XF DMEM and cultured for another hour in a non-CO₂ gassed, humidified incubator at 37°C.

The Seahorse Bioscience XF analyzer and Mito Stress Test Kit (Agilent Technology, CA, USA) were used according to the manufacturer's instructions. The sensor containing fluxpak (Agilent Technology, CA, USA) was hydrated and incubated overnight at 37°C in a non-CO₂ gassed humidified incubator. The sensor containing fluxpak was calibrated for approximately 15 min as per manufacturer guidelines. Upon completion, the pre-warmed cell plate containing immature COCs was loaded into the machine. The OCR was analyzed using a protocol involving a 12 min equilibration period and alternating between a 3 min measurement period and a 3 min re-equilibration period. During the measurement period, the sensor containing the probe was lowered down, creating an airtight 2.3 µL microenvironment. The output of extracellular flux analysis was given as OCR in pmol/min/well.

In a separate cohort, the basal OCR was measured in immature COCs cultured in either Seahorse XF DMEM or MEM-E to ensure that the base medium did not affect oxygen consumption ( Supplementary Figure SS2 (supp_fig_assessment_of_oocyte_quality_ioac145.docx)). This was performed as the goal of the study was to use optical imaging of autofluorescent metabolic co-factors to assess oocyte developmental competence and the need to culture COCs in standard IVM medium (MEM-E).

Measurement of metabolic co-factors NAD(P)H and FAD using two-channel laser confocal microscopy

Immature or mature (12 h) COCs were imaged. Images were acquired at baseline (basal), followed by sequential exposure to oligomycin, FCCP and Rot/AA (15 min between each treatment). Following IVM in the absence or presence of etomoxir (16 h), mature COCs were imaged. Imaging of COCs occurred within a 35 mm glass-bottom dish (Ibidi, Martinsried, Planegg, Germany) in 2 µL of Research Wash Medium, overlaid with paraffin oil.

The autof luorescence intensity indicative of co-enzymes NAD(P)H and FAD content was recorded on an Olympus Fluoview FV10i confocal microscope (Olympus, Tokyo, Japan). Cells were excited at a wavelength of 405 nm (emission detection bandwidth: 420–450 nm) for NAD(P)H (referred to as the NAD(P)H channel), and excited at a wavelength of 473 nm (emission detection bandwidth: 490–590 nm) for FAD (referred to as the FAD channel) as described in ref [35]. Image acquisition occurred at 60× magnification, numerical aperture equal to NA = 1.4, with a single z-plane chosen for each COC where the oocyte diameter was largest. Imaging parameters were kept constant between replicate experiments. Fluorescence intensity was measured using ImageJ software (National Institute of Health). For oocyte intensity measurements, a region of interest was created that encompassed the entire oocyte for each COC. Quantification of intensity within the cumulus cell compartment was performed by creating two regions interest of equal size and placing these on opposing sides of, and adjacent to, the oocyte. A mean of these two regions of interest was calculated for each COC. The optical redox ratio was calculated using the intensity of the FAD channel divided by the sum of the intensity of NAD(P)H and FAD channels (FAD / (NAD(P)H + FAD), which ref lects the activity of the mitochondrial electron transport chain and therefore indicates the dynamic changes of cellular metabolism [30].

Hyperspectral autofluorescence and brightfield imaging

In this work, the hyperspectral microscopy system (Quantitative Pty Ltd, Mount Victoria, NSW, Australia) was built by adapting a standard epifluorescence microscope (Nikon Eclipse TiE, 40× objective, NA = 1.3), fitted with a multi-LED light source (Prizmatix Ltd, Givat-Shmuel, Israel). These low-power LEDs provided 56 spectral channels: 21 excitation wavelength ranges and 3 emission wavelength filters, covering excitation wavelengths from 348 to 649 nm and emission wavelengths from 450 to 715 nm (for details of spectral channels, see  Supplementary Table SS1 (supp_table_assessment_of_oocyte_quality_ioac145.docx)). The fluorescence of native endogenous fluorophores found within cells was captured by a 40× objective imaging onto a digital camera C1140, OCRA Flash 4.0 (Hamamatsu, Shizuoka, Japan) using all 56 spectral channels. Image acquisition times of up to 3 s per channel were used, with multiple averaging (typically 1–3 times) to optimize image quality and minimize any potential photo-damage to the cells in each channel.

Mature COCs were imaged on a microscope slide and mounted using a 0.12 mm Secure Seal spacer (Molecular Probes, Invitrogen). Blastocysts were imaged on glass bottom confocal dishes (Ibidi, Martinsried, Planegg, Germany) containing 2 µL of Research Wash Medium overlaid with paraffin oil. Hyperspectral microscopy images were taken by adjusting the input light beam to specifically focus on the equatorial plane of the oocyte for COCs (i.e., the widest diameter) or the inner cell mass for individual blastocysts.

Analysis of hyperspectral microscopy data

To avoid errors in quantification, images were digitally processed first using the custom-made GUI_Preprocess v3.12 software [50]. This is required to remove image artifacts and noise, such as background fluorescence, Poisson's noise, dead or saturated pixels, and illumination curvature across the field of view, as described in detail in previous work [51–53]. At the beginning of each experiment, two calibration images were captured using the hyperspectral system: a “background” reference image of a culture dish with medium only, and another with calibration fluid only. The “background” reference image was included and subtracted from all images to remove any background signals. The microscope system was calibrated with a mixture of 50 µM NADH and 10 µM riboflavins whose spectrum spans all spectral channels and the concentration used sufficient for visualization of their autofluorescence. The excitation and emission spectra of this calibration fluid were measured using a spectrometer (FLS1000 Photoluminescence Spectrometer) and imaged on the hyperspectral microscope across all spectral channels.

Based on the known spectral properties of NAD(P)H and FAD [32, 54], we selected channels 1 and 2 (NAD(P)H channels) and channels 24 and 25 (FAD channels) ( Supplementary Table SS1 (supp_table_assessment_of_oocyte_quality_ioac145.docx)) to capture their autofluorescence. The COCs and blastocyst-stage embryos were manually segmented using the brightfield image to create a region of interest [49]. For COCs, the region of interest was as described above for confocal microscopy. For blastocyst-stage embryos, the region of interest was manually drawn around the inner cell mass. The intensity of NAD(P)H and FAD was quantified using custom-made GUI_Preprocess v3.12 software [50].

Embryo transfer and postnatal outcomes

Mature COCs (16 h) were divided into two groups: those imaged using the hyperspectral microscope and those that were not. Both groups of COCs were then fertilized in vitro and allowed to develop to the blastocyst-stage. Embryos were assessed for on-time morphological development on day 2 (2-cell; 24 h post-IVF) and day 5 (blastocyst-stage; 96 h post-IVF). Resultant blastocyst-stage embryos were then vitrified. Embryo vitrification and warming were performed as previously described [49]. On the day of transfer, blastocyst-stage embryos were warmed 2 h prior to embryo transfer to allow sufficient time for recovery. Blastocyst-stage embryos were transferred into the uterine horns of pseudopregnant Swiss mice 2.5 days post-coitum. Embryo transfers were performed on mice under anesthesia with 1.5% isoflurane as previously described [49]. Eight to twelve morphologically normal, expanded blastocysts were transferred to each uterine horn. Mice were monitored daily. The number of pups for each recipient was recorded on delivery. Pregnancy rate was calculated from the number of pregnant recipients over the total number of pseudopregnant females. Live birth rate was calculated from the number of live pups over the number of embryos transferred (non-pregnant mice were excluded from this analysis). At post-natal day 21, offspring were weaned, accessed for gross facial deformities and weight recorded.

Statistical analyses

Wave software (Agilent Technology, CA, USA) was used to determine OCR in pmol/min/well. Statistical analyses were carried out using GraphPad Prism Version 9 for Windows (GraphPad Holdings LLC, CA, USA) except for weight of offspring data where Statistical Package for Social Science (SPSS) version 28.0.1.0 software was used. Data were subjected to normality testing using the D'Agostino-Pearson Omnibus normality test prior to statistical analysis. Normally distributed data were analyzed either by an unpaired Student t-test or an ordinary one-way analysis of variance (ANOVA) with Holm-Šídák post-hoc test. For data that did not follow a normal distribution a Mann–Whitney test or Kruskal-Wallis test with Dunn post-hoc test were used. Continuous data are presented as mean ± SEM. Weight of offspring at weaning is presented as estimated marginal mean ± SEM and was analyzed by linear mixed model with litter size as a covariate. Categorical data are described as percentages and compared using the Fisher exact test. Details of statistical tests are stated in the figure legends. Statistical significance was set at a P-value <0.05.

Optical redox ratio detects dynamic changes in immature COCs that correlate with oxygen consumption rate

We determined whether the optical redox ratio detects dynamic metabolic changes within immature COCs in response to mitochondrial inhibitors/uncoupler. We evaluated metabolic changes within the oocyte and cumulus cell compartments separately (Figure 1A). Compared to basal levels, there was a reduction in the intensity of NAD(P)H within the oocyte in response to oligomycin, although this did not reach statistical significance (Figure 1B). There was a significant reduction in FAD within the oocyte following the addition of oligomycin compared to basal levels (Figure 1C). The changes in NAD(P)H and FAD in response to oligomycin yielded no impact on the optical redox ratio of oocytes compared to basal levels (Figure 1D). Treatment with FCCP resulted in a significant increase in NAD(P)H, FAD, and optical redox ratio compared to levels observed in the presence of oligomycin (Figure 1B–D, respectively). Following addition of Rot/AA, there was no change in the levels of NAD(P)H but a significant decrease in FAD and optical redox ratio compared to levels seen in the presence of FCCP (Figure 1B–D, respectively).

In cumulus cells, the addition of oligomycin caused no changes in NAD(P)H, FAD, or the optical redox ratio compared to basal levels (Figure 1E–G, respectively). The addition of FCCP caused no change in the intensity of NAD(P)H but did result in a significant increase in FAD and the optical redox ratio compared to oligomycin (Figure 1E–G, respectively). Subsequent addition of Rot/AA yielded no change in the levels of NAD(P)H compared to FCCP but was significantly higher compared to basal levels (Figure 1E). For both FAD and the optical redox ratio, Rot/AA caused a significant decrease compared to FCCP levels (Figure 1F and G, respectively). Excitingly, when comparing the optical redox ratio with OCR, we observed that changes in optical redox ratio were consistent with OCR for the oocyte for all mitochondrial inhibitors/uncoupler (Figure 1H, black dashed line vs red line). The optical redox ratio for cumulus cells was consistent with changes in OCR with the exception of the response to oligomycin (Figure 1H; black solid line vs red line).

Optical imaging of metabolic cofactors in mature COCs detect metabolic changes associated with oocyte developmental competence

The above data demonstrated that label-free optical imaging of metabolic co-factors and subsequent calculation of the optical redox ratio were robust measures of metabolism in the immature oocyte. We next determined whether this approach was equivalently accurate for the mature COC (expanded, meiotically mature). In the oocyte, the addition of oligomycin yielded no impact on the intensity of NAD(P)H compared to basal levels (Figure 2A). Compared to basal levels, there was a decrease in FAD and optical redox ratio, although this was not statistically significant (Figure 2B and C, respectively). The addition of FCCP resulted in a significant decrease in the NAD(P)H and increase in the FAD compared to levels observed in the presence of oligomycin (Figure 2A and B, respectively). The changes in NAD(P)H and FAD in response to FCCP led to a significant increase in optical redox ratio compared to oligomycin (Figure 2C). Following addition of Rot/AA, there was no change in the intensity of NAD(P)H (Figure 2A) but there was a significant decrease of FAD and optical redox ratio compared to FCCP (Figure 2B and C, respectively).

In cumulus cells, there was no impact on NAD(P)H, FAD signal or optical redox ratio in response to oligomycin compared to basal levels (Figure 2D–F). The addition of FCCP caused no change in the level of NAD(P)H (Figure 2D) but resulted in a significant increase in FAD and the optical redox ratio compared to oligomycin (Figure 2E and F, respectively). Similar to FCCP, Rot/AA had no impact on the level of NAD(P)H (Figure 2D) but led to a significant decrease in the intensity of FAD and optical redox ratio (Figure 2E and F, respectively).

Toward determining whether this or similar optical approaches could detect metabolic changes associated with oocyte developmental potential, we utilized a well-described model of poor oocyte quality [9, 56]. COCs were matured in vitro in the absence or presence of etomoxir, which is known to inhibit fatty acid oxidation and result in decreased oocyte developmental potential [9]. The intensity of metabolic cofactors NAD(P)H and FAD was quantified in the oocyte and cumulus cells separately (Figure 2G–L). Compared to control, etomoxir treatment during IVM significantly reduced the intensity of NAD(P)H (Figure 2G and J) and FAD (Figure 2H and K) in both the oocyte (Figure 2G and H) and cumulus cells (Figure 2J and K). The presence of etomoxir during IVM did not alter the optical redox ratio for oocytes (Figure 2I) or cumulus cells (Figure 2L). Additionally, we confirmed that maturation in the presence of etomoxir negatively affects oocyte developmental potential. While maturation in the presence of etomoxir did not affect fertilization rate (control: 90.2 ± 2.5% vs etomoxir: 72.9 ± 8.8%, data not shown), it did result in significantly fewer embryos reaching the blastocyst stage of development (Figure 2M).

Hyperspectral microscopy detects metabolic changes associated with oocyte quality that persist in resultant blastocysts

The use of laser-scanning confocal microscopy as a clinical measure of oocyte quality is hampered by the high laser energy dose required for imaging, likely resulting in photodamage [29]. Consequently, we next investigated whether hyperspectral microscopy could detect analogous changes in autof luorescence in the COC. Hyperspectral microscopy was chosen due to its 100-fold lower energy dose requirement compared to laser-scanning confocal imaging [35, 57]. Channels 1 and 2 were used to quantify the intensity of NAD(P)H, whereas Channels 24 and 25 were used to capture FAD (as described in Materials and Methods). Using the same model of poor oocyte quality, hyperspectral microscopy showed similar changes to NAD(P)H and FAD in both the oocyte and cumulus cells in response to etomoxir. Compared to control, there was a significant decrease in fluorescence in both the oocyte and cumulus cells following IVM in the presence of etomoxir in Channel 1 (Figure 3A and E, respectively) but no difference was seen in Channel 2 (Figure 3B and F, respectively). For both FAD channels (Channels 24 and 25) there was a significant decrease in autofluorescence in the oocyte and cumulus cells following IVM in the presence of etomoxir compared to control (oocyte: Figure 3C and D; cumulus cells: Figure 3G and H).

Figure 1.

Optical imaging of autofluorescent metabolic cofactors in immature COCs reflects changes in oxygen consumption rate. Metabolism was measured in immature COCs in response to oligomycin (oligo: 2.0 µM), carbonyl cyanide-4-(trifluoromethoxy) phenylhydrazone (FCCP: 1 µM) and Rotenone/antimycin A (Rot/AA: 2.5 µM). The metabolic response to these inhibitors of the electron transport chain was measured by either laser scanning confocal microscopy (intracellular NAD(P)H and FAD) or extracellular flux analysis (oxygen consumption rate). Representative images of NAD(P)H and FAD autofluorescence are shown in (A). The intensity of NAD(P)H (B and E) and FAD (C and F) was quantified for the oocyte (B and C) and cumulus cells (E and F). The optical redox ratio (FAD / [NAD(P)H + FAD]) was calculated for the oocyte (D) and cumulus cells (G) as an indicator of overall metabolic activity. The optical redox ratio for the oocyte and cumulus cell compartments was compared with the oxygen consumption rate for whole COCs (H). Data are presented as mean ± SEM; optical redox ratio: n = 12 COCs per drug treatment, four independent experimental replicates; OCR: n = 16 wells (20 COCs/well; OCR was normalized using the number of COCs per well and presented as pmol/min/COC), four independent experimental replicates. Data were analyzed either by a Kruskal-Wallis with Dunn multiple comparison test (B and D) or a one-way ANOVA with Holm-Šídák multiple comparison test (C and E–G). * P < 0.05, * * P < 0.01, * * * P < 0.001, * * * * P < 0.0001. Scale bar = 80 µm.

Figure 2.

Optical imaging of metabolic cofactors in mature COCs is associated with oocyte quality. Metabolism was measured in mature COCs in response to oligomycin (oligo: 2.0 µM), carbonyl cyanide-4-(trifluoromethoxy) phenylhydrazone (FCCP: 1 µM) and Rotenone/antimycin A (Rot/AA: 2.5 µM). In a separate cohort, COCs were matured in vitro in the absence or presence of etomoxir, an inhibitor of fatty acid metabolism (β-oxidation). The metabolic response to inhibitors/uncoupler (A–F) or etomoxir (G–L) was measured by laser scanning confocal microscopy (intensity of NAD(P)H and FAD autofluorescence). The intensity of NAD(P)H (A, D, G, and J) and FAD (B, E, H, and K) was quantified in the oocyte (A, B, G, and H) and cumulus cells (D, E, J, and K). The optical redox ratio (FAD / [NAD(P)H + FAD]) was calculated as indicator of metabolic activity in oocytes (C and I) and cumulus cells (F and L). The effect of inhibiting fatty acid metabolism during IVM on oocyte developmental competence was assessed by subsequent development to the blastocyst stage (M; blastocyst-rate calculated from starting oocyte number). Data were analyzed by either a Kruskal-Wallis with Dunn multiple comparison test (B) or a one-way ANOVA with Holm-Šídák multiple comparison test (A and C–F) or two-tailed unpaired Student t-test (G and M) or Mann–Whitney test (H–L). Data presented as mean ± SEM, three independent experimental replicates; n = 12 COCs per treatment (A–F); n = 33 for control COCs, n = 32 for etomoxir-treated COCs (G–L); n = 174 and 138 embryos developed from control and etomoxir-treated COCs, respectively (M). *P <0.05, **P <0.01, ***P <0.001, ****P <0.0001.

Figure 3.

Hyperspectral microscopy detects metabolic changes associated with oocyte quality. COCs were matured in vitro in the absence or presence of etomoxir. Etomoxir is a known inhibitor of fatty acid metabolism (β-oxidation) and negatively affects oocyte developmental competence (see Figure 2M). Matured COCs were imaged using hyperspectral microscopy. Autofluorescence intensity was quantified for the oocyte (A–D) and cumulus cells (E–H) in hyperspectral channels that match the spectral properties of NAD(P)H: Channel 1 (A and E) and Channel 2 (B and F); and FAD: Channel 24 (C and G) and Channel 25 (D and H). Data were analyzed by a two-tailed unpaired Student t-test (A–C and H) or Mann–Whitney test (D–G). Data presented as mean ± SEM, three independent experimental replicates; n = 12 for control COCs, n = 16 for etomoxir-treated COCs. *P < 0.05, * *P < 0.01.

Figure 4.

Altered metabolism during oocyte maturation persists in resultant blastocyst-stage embryos. COCs were matured in vitro in the absence or presence of etomoxir, an inhibitor of fatty acid metabolism (β-oxidation), which negatively affects oocyte developmental competence (see Figure 2M). Following in vitro maturation, COCs were fertilized in vitro and developed to the blastocyst stage in the absence of etomoxir. Embryos were imaged using hyperspectral microscopy. Autofluorescence intensity within the inner cell mass was quantified for hyperspectral channels that matched the spectral properties of NAD(P)H: Channel 1 (A) and Channel 2 (B); and FAD: Channel 24 (C) and Channel 25 (D). Data were analyzed by either a two-tailed unpaired Student t-test (C and D) or Mann–Whitney test (A and B). Data presented as mean ± SEM, three independent experimental replicates; n = 24 for control and n = 29 for blastocysts developed from control and etomoxir-treated COCs, respectively. **P <0.01, ***P <0.001.

To determine whether metabolic changes detected in etomoxir-treated oocytes persisted in resultant embryos, COCs matured in the absence or presence of etomoxir were fertilized in vitro and allowed to develop to the blastocyst stage. The intensity of NAD(P)H (Channels 1 and 2; Figure 4A and B) and FAD (Channels 24 and 25; Figure 4C and D) was significantly lower in the fetal cell lineage (inner cell mass) of blastocyst-stage embryos that developed from etomoxir-treated COCs compared to those matured in control conditions.

Safety of hyperspectral imaging

To assess whether photodamage had occurred in response to imaging, untreated mature COCs were either not imaged (non-imaged) or imaged using hyperspectral autof luorescence microscopy (imaged). Both groups of COCs were then fertilized in vitro and allowed to develop to the blastocyst-stage. Imaging of COCs did not affect fertilization rate (non-imaged: 90.15 ± 1.71% vs imaged: 84.49 ± 2.76%, data not shown) or subsequent development to the blastocyst stage (non-imaged: 63.47 ± 5.18% vs imaged: 61.87 ± 4.78%, data not shown). We next examined whether hyperspectral imaging of the COC altered the potential of the subsequent embryo to implant and result in the birth of a healthy live off-spring. Imaged COCs resulted in the birth of live pups (Figure 5A). There were no significant differences in pregnancy or live birth rate (P > 0.05;  Supplementary Table SS2 (supp_table_assessment_of_oocyte_quality_ioac145.docx)). Imaged versus non-imaged COCs resulted in offspring with similar, and not significantly different, weights at weaning (Figure 5B;  Supplementary Table SS3 (supp_table_assessment_of_oocyte_quality_ioac145.docx)). Similarly, there was no difference in weight at weaning according to sex ( Supplementary Figure SS3 (supp_fig_assessment_of_oocyte_quality_ioac145.docx)). No gross facial deformities were noted across treatment groups.