Fine morphological assessment of quality of human mature oocytes after slow freezing or vitrification with a closed device: a comparative analysis

This project has been approved by our internal review board and the Italian Ministry of Health. The study involves patients who decided to donate their supernumerary eggs to research between January 2012 and January 2013. The age of women ranged from 27 to 32 years old, and their infertility was due to male, tubal factors or idiopathic issues. Patients with endometriosis or other conditions, that could influence oocyte quality, were excluded from the study. Ovarian stimulation was induced with a long protocol using GnRH agonist and rFSH, according to previous clinical report [28]. HCG (10,000 IU) was injected 36 hours before retrieval. Oocytes were cultured in fertilization media for at least 2 hours (Cook IVF, Brisbane, Australia) before cumulus/corona cells complete removal; this was performed enzymatically (hyaluronidase 20–40 IU/ml) and mechanically (using fine pipettes Flexi pipette Cook). Mature oocytes displaying a 1st polar body (PB) with a clear cytoplasm were assigned to either the fresh control or study groups. Fresh control oocytes (CO) were fixed after a period of 3–4 hours following retrieval. Cryopreservation was performed within 3–8 hours after retrieval according to the laboratory workload. Slow frozen/thawed (SFO) and vitrified/warmed oocytes (VO) were cultured for a further hour before fixation.

Oocytes were cryopreserved using a slow-freezing protocol adapted from the one originally described for embryos [29]. Eggs were equilibrated in one step homemade solution containing 1.5 mol/L 1,2-propanediol (PrOH) supplemented with 20% of synthetic serum supplement (SSS) (Irvine Scientific, Santa Ana, CA, USA) for 10 minutes then transferred for 5 minutes into a loading solution containing 1.5 mol/L PrOH +0.2 mol/L sucrose +20% SSS. Afterwards the oocytes were loaded into plastic straws (Paillettes Crystal 133 mm; CryoBioSystem, Paris, France) and placed into an automated Kryo 10 series III biological freezer (Planer Kryo 10/1,7 GB; Planer, Surrey, UK). All the procedures were carried out at room temperature (around 25°C). Once the loaded straws were placed in the machine the temperature was gradually lowered from 20°C to -7°C at a rate of -2° C/min. Manual seeding was induced during a 10 minutes holding ramp at -7°C. The temperature was then decreased to -30° C at a rate of -0.3°C/min and finally rapidly to -150° C at a rate of -50°C/min. The straws were then plunged into liquid nitrogen and stored for later use. For the thawing procedures the straws were air-warmed for 30 seconds and then placed in a 30°C water bath for 40 seconds. The cryoprotectant was removed by stepwise dilution of PrOH at room temperature. The thawing solutions were homemade and contained:

Survived oocytes were finally placed in culture in Cleavage medium (Cook) at 37°C and 5% CO₂ and re-checked 1 hour later to confirm their good status prior to fixation.

Oocytes were vitrified using commercial kits (Irvine Scientific CA) and closed vitrification devices (CryoTip). The oocytes were washed in a drop of hepes-buffered medium (gamete buffer Cook) and, subsequently in three drops of equilibration solution containing ethylene glycol (EG) (7.5% vol/vol) and dimethylsulphoxide (DMSO) (7.5% vol/vol). After 8 minutes, eggs were transferred into a vitrification solution drop containing 15% vol/vol EG, 15% vol/vol DMSO, and 0.5 mol/L sucrose, for a total of 20 seconds before being loaded into a Cryotip and sealed properly at both ends. The device was directly plunged into liquid nitrogen and stored.

Oocytes were rapidly warmed by transferring the Cryotip directly from liquid nitrogen to a 37°C water bath for 3 seconds. The device was cut at the end and the oocyte/s released in a thawing solution (1.0 mol/L sucrose) for a minute then moved to a dilution media (0.5 mol/L sucrose) for 4 minutes and finally washed twice in a washing solution (6 minutes). Survived oocytes were cultured in Cleavage medium (Cook) before the fixation.

Fortyseven supernumerary human MII oocytes (12 CO, 20 SFO and 15 VO) were donated by consenting patients (N = 12), and included in this study. In detail: 3 patients donated 12 CO (4 oocytes from each patient), 5 patients donated 20 SFO (2, 3, 4, 5, 6 oocytes respectively from each patient), 4 patients donated 15 VO (1, 3, 6, 5 oocytes respectively from each patient).

Only oocytes that appeared of good quality when observed by PCM after thawing or warming were selected for EM evaluation. They should have: 1. a rounded, regular shape; 2. a clear, moderately granular cytoplasm; 3. a narrow perivitelline space (PVS) with the 1st PB and, 4. an intact, colorless ZP [30].

Oocytes were fixed and processed for LM and TEM analysis as follows. Oocyte fixation was performed in 1.5% glutaraldehyde (SIC, Rome, Italy) in PBS solution. After fixation for 2–5 days at 4°C, the samples were rinsed in PBS, post-fixed with 1% osmium tetroxide (Agar Scientific, Stansted, UK) in PBS, and rinsed again in PBS. Oocytes were then embedded in small blocks of 1% agar of about 5×5×1 mm in size, dehydrated in ascending series of ethanol (Carlo Erba Reagenti, Milan, Italy), immersed in propylene oxide (BDH Italia, Milan, Italy) for solvent substitution, embedded in epoxy resin (Electron Microscopy Sciences, Hatfield, PA, USA) and sectioned by a Reichert-Jung Ultracut E ultramicrotome. Semithin sections (1 μm thick) were stained with toluidine blue, examined by LM (Zeiss Axioskop) and photographed using a digital camera (Leica DFC230). Ultrathin sections (60–80 nm) were cut with a diamond knife, mounted on copper grids and contrasted with saturated uranyl acetate followed by lead citrate (SIC, Rome, Italy). They were examined and photographed using a Zeiss EM 10 and a Philips TEM CM100 Electron Microscopes operating at 80KV. Images were acquired using a GATAN CCD camera.

According to Nottola et al. [23], as reviewed by Khalili et al. [26], the following parameters have been evaluated by LM and TEM and taken into consideration for the qualitative morphological assessment of the ultrastructural preservation of oocytes: main characteristics (including shape and dimensions), ZP texture, PVS appearance, integrity of the oolemma, microtopography, type and quality of the organelles, presence and extent of ooplasmic vacuolization.

The presence and characteristics of the 1st PB in the PVS and the arrangement of the MII spindle were not systematically assessed, due to their detection only in sections laying on appropriate planes.

The presence of vacuoles ≥1 μm was evaluated at the LM level on at least 3 equatorial sections per oocyte (distance between the sections: 3–4 μm), and values were expressed in number of vacuoles per 100 μm² of the oocyte area. The evaluation of cortical granule (CG) density was performed through collection of TEM microphotographs of the entire surface profiles at a magnification of 6300X on 3 equatorial sections per oocyte. Images were further enlarged on the PC screen, in order to easily recognize and count CG. Values were expressed in number of CG per 10 μm of the oocyte linear surface profile [23].

Authors presented statistical data as a mean value ± standard deviation (SD); P value and significance were evaluated using Student’s t test (http://​www.​graphpad.​com/​quickcalcs, last access: September 2, 2014). The cut off for significance was P < 0.05.

LM and TEM techniques allowed analyzing and comparing size, shape and organelle distribution in CO, SFO and VO. CO, SFO and VO were all generally rounded, 90–100 μm in diameter, with normal ooplasm showing uniform distribution of organelles. All CO, SFO and VO showed an intact ZP, separated by a narrow PVS from the oolemma, continuous and provided with microvilli (Figure 1a, b, c, d, e and f).

Only in favorable sections, the 1st PB was detected by LM in the PVS (which appeared wider in this region) (Figure 1a). The MII spindle (or a portion of this) was also visible in the ooplasm of a number of CO (N = 4), SFO (N = 6) and VO (N = 5), assuming a peripheral position (Figure 1b). By TEM, the 1st PB contained condensed chromatin, mitochondria, residual microtubules and scattered CG. The MII spindle consisted of chromosomes, with a dense granulo-fibrillar microstructure, and associated microtubules converging at each pole (data not shown).

Using TEM, the most numerous organelles found in all the CO, SFO and VO cultured for 3–4 hours, consisted of aggregates of anastomosing tubuli of smooth endoplasmic reticulum (SER) surrounded by mitochondria (M-SER aggregates). The diameter of the tubular network of the M-SER aggregates varied from 1 to 5 μm (Figure 2a and b). Small vesicles, 0.3 – 0.5 μm in diameter, containing a slighlty electrondense material were associated with mitochondria forming the so-called mitochondria-vesicle (MV) complexes (Figure 2b). No overt qualitative differences in the fine structural morphology of M-SER aggregates and MV complexes came out from the comparison among CO, SFO and VO after 3–4 hours of culture. M-SER aggregates appeared instead partially replaced by numerous, large MV complexes, up to 2.5 μm in vesicle diameter, when SFO and VO extended culture for a prolonged period of time (8–9 hours) (Figure 2c and d). Mitochondria, either associated with membranes or isolated, revealed a normal fine structure in all the samples observed (CO, SFO and VO). They were rounded or oval in profile, with a diameter varying from 0.5 to 0.8 μm and a few peripheral arch-like or transverse cristae and contained a moderately electrondense matrix (Figure 2a, b and d; Figure 3c).

LM showed very rare vacuolization of the ooplasm in CO (Figure 1a). TEM confirmed the very sporadic incidence of vacuolization in these oocytes (Figure 1d). A slight to moderate vacuolization was found in the cytoplasm of cryopreserved oocytes, more pronounced in SFO than in VO (Figure 1b and e; Figure 3a, b and c). Vacuoles, varying in diameter from 0.2 to about 2.5 μm, were apparently empty (Figure 3a,b and c) or contained cell debris (Figure 3a and b). Vacuole membrane was often discontinuous (Figure 3a, b and c). Sometimes, secondary lysosomes and/or multivesicular bodies (MVBs) associated with vacuoles were present (Figure 3c). Otherwise, it was usually detected a normal pattern of organelles in the cytoplasmic areas adjacent to vacuoles (Figure 3b and c).

The morphometric analysis performed on fresh and cryopreserved oocytes belonging to both study groups (i.e. subjected to slow freezing or vitrification with a closed system) revealed that the difference in vacuole number between CO and SFO was statistically significant. In particular, the mean number ± SD of vacuoles ≥1 μm in diameter per 100 μm² was 0.40 ± 0.12 and 2.09 ± 0.85 in CO and SFO, respectively (P < 0.0001) (Table 1). A statistically significant difference in vacuole number between CO and VO was also present. In fact, the mean number ± SD of vacuoles ≥1 μm in diameter per 100 μm² was 0.40 ± 0.12 and 0.85 ± 0.37 in CO and VO, respectively (P = 0.0005) (Table 1). However, we found a significant difference in vacuole number also between SFO and VO (P < 0.0001), confirming the presence of more vacuoles in SFO than in VO (Table 1).

In fresh oocytes (CO) the cortex was rich in CG aligned in a continuous array, at the periphery of oocyte, under the oolemma. CG were rounded and showed a diameter varying from 300 to 400 nm (Figure 4a). CG formed instead a discontinuous layer and represented less in both SFO and VO in comparison with CO (Figure 4b and c). The mean number ± SD of CG per 10 μm was 8.10 ± 1.59, 2.95 ± 0.96, and 2.40 ± 0.84 in CO, SFO and VO, respectively. These differences between fresh and cryopreserved groups were highly significant (CO and SFO: P < 0.0001; CO and VO: P < 0.0001). On the contrary, we did not find any significant difference in amount of CG between SFO and VO (P = 0.086) (Table 2).

All CO, SFO and VO appeared normal in shape, dimensions and texture of the ooplasm under LM and low magnification TEM examination. This features are superimposable to those described in our previous reports on human mature oocytes subjected to different protocols of slow cooling [21, 22, 25] or vitrification [23, 42], reinforcing the concept that current cryopreservation protocols, including slow freezing and cryopreservation with a closed device, do not produce major alterations in the oocyte microdomains.

In our study, when observable in sections laying on appropriate planes, the MII spindle appeared regularly shaped and properly positioned not only in the ooplasm of CO but also in that of SFO and VO, confirming that meiotic spindle alterations eventually occurring during cryostorage (both slow freezing and vitrification) can be reversed on rewarming [43]. However, due to specific technical requirements and to the actual difficulty to fully visualize the spindle scaffolding and associated chromosomes in TEM ultrathin sections, we consider other morphological ultrastructural tools, such as confocal laser microscopy, more suitable than TEM for studying spindle morphodynamics during cryopreservation [21].

The finding in both SFO and VO of a good preservation of mitochondrial structure, as in our previous studies [21‐23, 25, 42] further emphasizes that both slow freezing and vitrification with a closed device do not produce marked and diffuse alterations in the oocyte microstructure. This does not confirm what reported by other Authors, who described alterations in mitochondrial matrix and cristae in a percentage of oocytes cryopreserved with both slow cooling and vitrification [19, 44]. Due to the many differences among protocols and technique of sampling, it is not possible to hypothesize here an explanation, but this aspect deserves further evaluation.

We found that M-SER aggregates and small MV complexes were the most common organelles found in all the oocytes cultured for 3–4 hours, both fresh and cryopreserved, with either slow freezing or vitrification.

Small MV complexes, isolated SER tubules and/or SER networks are present in human immature, GV-stage oocytes [42, 45]. Only mature, MII oocytes show fully organized M-SER aggregates of various sizes. These aggregates, according to the biochemical needs of the mature, healthy oocyte, may regulate calcium levels and mitochondrial ATP production, thus potentially contributing to the modulation of calcium-dependent signal transduction pathways at fertilization [46]. M-SER aggregates may have a role in production of substances useful at fertilization and/or in rapid neoformation of nuclear and cytoplasmic membranes during early embryogenesis [47‐49].

We did not find evident qualitative differences in the fine structural morphology of M-SER aggregates and MV complexes in both SFO and VO, when compared to CO, after 3–4 hours of culture. In other reports, a percentage of human oocytes, vitrified with open devices, showed small and slender M-SER aggregates [23]. Differently, previous studies on oocytes subjected to slow freezing and treated with PrOH did not evidence any ultrastructural change of these aggregates in respect to fresh controls [21, 25].

In the present study, we also considered the amount of M-SER aggregates, evidencing that they were partially replaced by larger MV complexes in the oocytes maintained in culture for a prolonged period of time (8–9 hours), either SFO or VO. This particular feature, led us to hypothesize that prolonged culture may induce an intracellular membrane “recycling”, causing a discrete enlargement of the MV complexes to which also contribute a considerable SER membrane reassembly. The presence of numerous and large MV complexes was previously described in MII oocytes kept in vitro for 24 (oocytes aged in vitro [50]) or 48 hours (oocytes insemined but not fertilized [47]). Large MV complexes are also present in GV oocytes that have reached MII stage after 24 hour-culture (in vitro matured oocytes [27]). Therefore, changes of MV complexes and SER do not seem related only to the maturative stage of the oocyte at the beginning of the culture, but mainly to the culture period. In the present study, we originally demonstrated that abnormally large MV complexes may form very early during culture, being already present 8–9 hours after oocyte retrieval. The role of these structures is still unknown. We may speculate that they may be associated to impairment of calcium homeostasis, thus reducing the oocyte competence for fertilization after a prolonged culture.

In our study, a slight to moderate vacuolization was present in the cytoplasm of SFO. Only a slight vacuolization was instead present in VO. Vacuoles were almost completely absent in CO.

Vacuolization is an important dysmorphism constantly detected in human mature oocytes treated with different protocols of slow freezing [21, 22, 25, 44, 51]. This is explainable as a non-specific response of the oocyte to thermal, osmotic and/or chemical injuries that may occur during cryopreservation. We observed an association of vacuoles with lysosomes and MVBs that are types of organelles rarely present in human healthy mature oocytes and of degenerative significance [44]. The appearance of vacuoles in human oocytes is likely in relation with a reduced oocyte fertilizability and with an impairment of embryo development [52]. The opinions about the presence and extent of vacuolization in human VO are controversial. In fact, some Authors reported an evident vacuolization in oocytes vitrified with both closed (Cryotip) and open (Cryotop) devices [19, 42, 44]. Vacuolization, however, was more severe when a closed device was applied. This may be due to the Cryotip thermosealing and/or to the longer time for closed-vitrification oocyte discharge into the warming solution [19]. In our previous studies, we instead referred a virtual absence of vacuolization in oocytes vitrified using different open devices (Cryoleaf and Cryoloop) [23]. Taken together, all these data suggest the occurrence of a possible cryodevice-dependant vacuolization, that may be also associated to the skill of the operator and even to the sample processing for EM. The different morphometric approaches in vacuole counting and measuring should be also taken into consideration [42]. Our results on SFO well correlate with the data of the recent literature on oocyte cryopreservation, confirming the association between slow freezing and occurrence of vacuolization in the oocyte. We also found only a slight vacuolization in all the VO, significantly lesser than that observed in SFO. Thus our morphometric results, differently from previous qualitative observations, failed in demonstrating a severe incidence of vacuolization in oocytes vitrified with a closed device.

We found that amount and density of CG appeared abnormally reduced in both SFO and VO, in comparison with CO.

Usually, CG are regularly present, stratified in one to three rows, in the subplasmalemmal areas of mature oocytes of most mammalian species including humans [47]. They are Golgi-derived, membrane-bounded organelles formed during the early stages of oocyte growth and contain glycosaminoglycans, proteases, acid phosphatases and peroxidases [53]. At fertilization, CG content is suddenly and massively exocytosed by the activated oocyte in the PVS (‘cortical reaction’), leading to a hardening of the inner face of the ZP (‘zona reaction’) and to a consequent inhibition of eventual penetration of supernumerary spermatozoa into the oocyte (polyspermy) [54, 55]. Calcium transients at fertilization seem involved in triggering CG fusion with the oolemma, resulting in the release of their contents into the PVS [56]. Our findings, are comparable to previous studies on slow freezing [21, 22, 25, 44, 51] and vitrification [19, 23, 42, 44], and suggest that a precocious oocyte activation may occur during cryopreservation, irrespective of the protocol applied (slow freezing or vitrification, even with a closed device). This activation leads to a premature CG exocytosis and a consequent diffuse hardening of the ZP, thus greatly impairing oocyte competence to fertilization.