A microscopic and biochemical study of fragmentation phenotypes in stage-appropriate human embryos

Jonathan Van Blerkom1,2,3, Patrick Davis1,2 and Samuel Alexander2

1 Department of Molecular, Cellular and Developmental Biology, University of Colorado, Boulder, CO 80309 and 2 Colorado Reproductive Endocrinology, Rose Medical Center, Denver, CO 80220, USA


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
The occurrence of a pleiomorphic population of cytoplasmic fragments is a common characteristic of early human embryos fertilized in vitro. Here, temporal, spatial, fine structural, and biochemical aspects of fragmentation were examined in fragmented monospermic and dispermic pronuclear to early cleavage stages human embryos classified as stage-appropriate during the first 3.5 days of culture. The morphodynamics of certain common patterns of fragmentation and the movement and composition of fragments were analysed by time-lapse video, mitochondrial fluorescent probes, and transmission electron microscopy. Plasma membrane and nuclear DNA integrity were assessed by annexin V staining, and terminal deoxynucleotidyl transferase-mediated dUTP nick end-labelling (TUNEL) and single-cell alkaline gel electrophoresis (`comet') assays respectively. Developmental competence for affected embryos was related to outcome after embryo transfer. The results demonstrate that certain common forms of spontaneous fragmentation affecting early human embryos are not lethal, and that clusters of apparent fragments are often transient structures, which disappear by resorption or lysis. The findings suggest that the occurrence and fate of fragments characteristic of these phenotypes may be related to oncosis-like processes associated with transient and focal ATP deficiencies in blastomeres and mitochondrial deficiencies or absence in extracellular fragments.

Key words: `comet' assay/electron microscopy/embryo fragmentation/time-lapse video microscopy/TUNEL


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Fragmentation is a common feature of the early development of embryos, and its occurrence often affects some proportion of embryos in cohorts produced by IVF. Within the same culture dish, some embryos progress with little or no fragment formation while others exhibit patterns of fragmentation of varying degrees of severity. An estimation of the number and volume of fragments in cleavage stage embryos is used in virtually all current schemes designed to assess relative developmental competence (Giorgetto et al., 1995Go; Hoover et al., 1995Go; Morgan et al., 1995Go; Alikani et al., 1999Go; Gerris et al., 1999Go). Recently, we (Antczak and Van Blerkom, 1999Go) and others (Alikani et al., 1999Go) have suggested that specific temporal or spatial patterns of fragmentation may be more closely related to competence than to the occurrence of fragmentation per se. Typically, the presence of fragments is noted during the initial inspection for fertilization and during the cleavage stages, when morphological evaluations are made to select embryos for uterine transfer or cryopreservation. However, certain forms of fragmentation are so extensive that few, if any, intact blastomeres are detected. Where some apparently intact blastomeres are visible amongst a mass of fragments, such embryos are usually de-selected for transfer and judged inappropriate for cryopreservation. In these instances, an apoptotic process that may be associated with the activation of programmed cell death pathways has been implicated in embryo demise (Jurisicova et al., 1996Go; Levy et al., 1998Go).

Other common fragmentation phenotypes exhibited by early human embryos are the occurrence of multiple clusters of fragments, which arise during early cleavage and persist as the embryo continues to divide, or which seem to `disappear' during culture resulting in apparently normal-appearing embryos at transfer. These specific patterns of fragmentation are perhaps the most difficult to evaluate for competence because it is unclear whether they indicate an underlying defect in the early embryo that could have potentially adverse consequences during later stages of embryo development (Antczak and Van Blerkom, 1999Go). In clinical IVF, it can be difficult to distinguish between fragmentation that is acute or progressive if assessments of embryo development are infrequent. Here, temporal, spatial, and morphodynamic aspects of these fragmentation patterns were analysed from the pronuclear to the 12-cell stage by means of time-lapse video and TEM. Alterations in the composition of the plasma membrane and the integrity of the DNA which could signify premorbid alterations at the blastomere level were assessed by annexin V staining and TUNEL (terminal deoxynucleotidyl transferase-mediated dUTP nick end-labelling) or single-cell alkaline gel electrophoresis (`comet') assays respectively.

The results provide new insight into the morphodynamics of fragmentation and for these embryos, demonstrate that fragments are not stationary, that many apparent fragments are actually bulbous/bleb-like elaborations of the plasma membrane and subjacent cytoplasm, rather than discrete extracellular structures, and that the disappearance of `fragments' during culture may occur by resorption or lysis. Competence for these embryos is suggested by the absence of detectable DNA damage or plasma membrane alterations and successful outcome after transfer. The absence of mitochondria in most of the fragments detected in these embryos is discussed with respect to a non-apoptotic aetiology.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Time-lapse and transmission electron microscopic analyses
Ovarian stimulation, conventional insemination in vitro, and routine embryo culture (modified MediCult Universal IVF medium) followed protocols described previously (Van Blerkom et al., 1995Go). Dispermic and fragmented monospermic embryos were examined with patient consent. Dispermic embryos were characterized by the presence of three distinct pronuclei in oocytes that abstricted a second polar body. Previous immunohistochemical analysis of this type of tripronucleate oocyte demonstrated the presence of two sperm tails within the cytoplasm (Van Blerkom et al., 1995Go). Embryos were cultured in 1 ml of medium in Falcon IVF organ culture dishes maintained in individual, humidified culture chambers (Billups-Rothenberg; Forma Scientific, Marion, OH, USA) in an atmosphere of 90% N2, 5.5% CO2, 4.5% O2 (Van Blerkom, 1993Go). According to this protocol, stage-appropriate embryos were usually at the 8–10-cell stage on day 3, and at the 10-12-cell stage on day 3.5. For time-lapse analysis, embryos were cultured from the pronuclear stage under oil in 250 µl microdroplets using the {Delta}T Culture Dish System (Bioptechs, Inc., Butler, PA, USA) at 37°C (Van Blerkom et al., 1995Go). {Delta}T dishes were glued onto a 100 mm plastic culture dish in which a portion of the plastic was removed in order for the glass bottom of the {Delta}T dish to directly face the objective lens. This unit was sealed with a rubber gasket and placed in a light-tight incubation chamber (Nikon NP-2 incubator) attached to an inverted microscope with ambient temperature of 37°C. A constant atmosphere of 90% N2, 5.5% CO2, 4.5% O2 was maintained by means of a stream of gas delivered into the plastic culture dish from a tank of pre-mixed gas after passage through a water-filled chamber maintained at 39°C. To ensure positional stability, the glass surface of the {Delta}T dish was treated with poly-L-lysine, to which the embryos remained affixed during culture.

Images were captured on an optical disc recorder at intervals ranging from 30 s to 10 min, depending upon the stage of development. For some analyses, embryo recordings were made under constant, low-light illumination using image intensification afforded by high-resolution video cameras. In these instances, simultaneous optical disc (images captured at >=5 min) and videotape recordings (images captured at 1 s intervals, with a time-lapse videocam recorder) were made in order to time precisely the occurrence of specific morphodynamic events. Protocols for time-lapse recording and standard procedures for electron microscopy and analysis of serial thin sections have been described previously (Van Blerkom et al., 1995Go).

Detection of DNA breaks
TUNEL analysis
Fragmented embryos with intact blastomeres were prepared for TUNEL analysis by two different methods: as whole mounts (Takase et al., 1995Go; Van Blerkom and Davis, 1998Go), or as air-dried preparations on microscope slides (Jurisicova et al., 1996Go; Perez et al., 1999Go). Both procedures used reageants supplied in an In Situ Cell Detection Kit with fluorescein-tagged nucleotides (Roche Molecular Biochemicals, Indianapolis, IN, USA). TUNEL analysis used zona-intact, zona-thinned, and zona-denuded specimens (Antczak and Van Blerkom, 1999Go). Briefly, samples were fixed in a freshly prepared 3.7% formaldehyde solution in phosphate-buffered saline (PBS, pH 7.35) for 45 min and washed for >12 h in PBS. Zona-free embryos were washed in PBS containing 4% poyvinylpyrrolidone. Samples were incubated in TUNEL reagents for 1 h at 37°C according to the recommendations of the manufacturer and the findings of our previous studies. Embryos were treated with Slow-Fade Lite (Molecular Probes, Eugene, OR, USA) and examined first by conventional epifluorescence microscopy with ultraviolet illumination and then by scanning laser confocal microscopy (SLCM) to obtain serial optical sections through entire specimens. All TUNEL-negative specimens were treated with DNase II (Sigma Chemical Co., St Louis, MO, USA) and restained with TUNEL reageants as described previously (Van Blerkom and Davis, 1998Go).

For air-dried samples, embryo preparation for TUNEL involved partial zona thinning with acidic Tyrode's (15–25 s) in order to retain fragments. After denudation, specimens were immediately fixed for 30–45 min in a freshly prepared 3.7% formaldehyde solution (pH 7.2) during which the zona remnant was removed by microneedle dissection. Fixed embryos were washed in PBS, deposited by micropipette in <=2 µl of fluid onto treated slides (Superfrost Plus; Fisher Scientific, Pittsburgh, PA, USA) and air-dried. Prior to deposition, a 10 mm square box was delineated with a Pap Pen (Sigma Chem. Co.) in order to confine solutions and reageants to a specific location. Slides were incubated at 65°C for 4h, stored overnight at 4°C, reheated for 30 min at 65°C, then rehydrated through a graded series of ethanol, and gently washed in multiple changes of high-purity, cell culture-grade water. Slides were incubated for 30 min at 37°C in PBS containing proteinase K at a concentration of 10 µg/ml (Roche Molecular Biochemicals). After enzymatic treatment, slides were washed in PBS containing 3% bovine serum albumin (BSA) at room temperature, excess fluid was withdrawn by micropipette, and the samples immediately exposed to TUNEL reagents as described above. TUNEL staining was performed at 37°C for 1.5 h followed by numerous rinses in PBS. Slides were placed in Slow Fade Lite equilibration buffer for 30 min (Molecular Probes, Eugene, OR, USA), excess buffer removed with a micropipette, and a small drop of Slow Fade was placed over the specimen and covered with a glass coverslip. Specimens were examined by conventional epifluorescence microscopy and by SLCM. After examination, the coverslip was gently removed and the slides washed in HEPES-buffered saline (pH 7.4), stained with DAPI for 15 min, and rinsed in HEPES-buffered saline for 30 min. Excess buffer was removed and replaced with 70% glycerol, and the specimens re-examined to detect chromosomal and nuclear DNA. All TUNEL-negative specimens were rehydrated, washed in PBS, and exposed to either DNase or to PBS containing 0.5 mol/l H2O2 for 15-30 min. These slides were processed and restained with TUNEL reagents as described above.

`Comet' analysis
The following describes in detail the standard protocol we devised for single cell alkaline gel electrophoresis (`comet' assay) which was modified from Singh and Stephens (1997). After partial thinning of the zona pellucida, embryos were washed in HEPES-buffered human tubal fluid medium (Quinns's HTF) and deposited onto a treated microscope slide in a 2 µl volume (Comet Slides, Trevigen, Inc., Gaithersburg, MD, USA). Most of the fluid surrounding the specimen was carefully and rapidly removed by a mouth-operated micropipette drawn to a fine taper. A 3 µl drop of a 0.4% solution of agarose (high-resolution blend 3:1 agarose; Amrecso, Solon, OH, USA) prepared in calcium- and magnesium-free PBS (pH 7.4) kept at 43°C was deposited over the embryos and gently extended outward with a micropipette whose tip had been fire-polished to form a small bead. Glass microneedles were used to disrupt mechanically the remnant zona, allowing blastomeres and fragments to be completely encapsulated by agarose within a confined area. The residual zona had no effect on the `comet' assay. After solidification, the agarose layer was immediately surrounded by a 25 µl ring of agarose kept at 45°C. With a beaded micropipette, the molten agarose was rapidly extended inward to contact the agrose-entrapped specimen, and when joined, formed a uniformly thin agarose microgel. During the above operations, slides were observed at x20 magnification on a Nikon Diaphot inverted microscope maintained at 39°C by means of a heated chamber attached to the microscope (Nikon, NP-2 incubated chamber). The entire process, from the initial deposition of the specimen to formation of the final agarose thin layer, was performed in <60 s.

After formation of microgels, the following steps were done under reduced ambient illumination or in complete darkness: agrose microgels were placed at 4°C on a refrigerated cooling block (Millipore, Bedford, MA, USA) for 10 min and then immersed for 45 min in a lysis solution (2.5 mol/l NaCl, 100 mmol/l disodium EDTA, 10 mmol/l Tris base, 1% sodium lauryl sarcosinate, 0.1% Triton X-100, pH 10) maintained at 4°C in glass Petri dishes on the same cooled platform. For some preparations, microgels were treated with proteinase K (1 mg/ml) in detergent-free lysis solution for up to 2 h at 37°C, and then immersed in 100 ml of alkali buffer (300 mmol/l NaOH, 0.1% 8-hydroxyquinoline, 2% dimethylsulphoxide, 100 mmol/l EDTA, pH 10) for 1 h at room temperature. Microgels were immersed in TBE (Tris-borate-EDTA buffer) for 5 min and positioned in a horizontal slab gel electrophoresis apparatus, equidistant from the electrodes. Electrophoresis was performed in total darkness for exactly 12 min at 200 mA in 1 litre of pre-chilled TBE. After electrophoresis, slides were neutralized in 100 ml of 0.4 mol/l Tris, pH 7.4, excess fluid drained, and the microgels stained with SYBR Green (Molecular Probes), over-layered with a coverslip, and examined immediately by conventional epifluorescence in the fluorescein isothiocynate (FITC) channel. Slides were dehydrated in ethanol, air-dried and stored in desiccant-containing chambers at 5°C.

Annexin V staining and exposure of embryos to hypotonic solutions or hydrogen peroxide
The integrity of the plasma membrane was assessed in living specimens by the presence or absence of cytoplasmic and nuclear fluorescence after staining with propidium iodide (Van Blerkom and Davis, 1998Go), or by cytoplasmic incorporation of Trypan Blue (0.2% in culture medium, 5 min exposure). Phosphatidylserine translocation from the inner to outer leaflet of the plasma membrane was determined by staining living embryos with FITC-tagged annexin V (AV; Clontech, Palo Alto, CA, USA) as previously described (Van Blerkom and Davis, 1998Go). Controls for the specificity of the AV, TUNEL and `comet' assays used unfragmented dispermic embryos exposed to medium containing hydrogen peroxide (500 µmol/l, 30 min) or to hypotonic conditions (distilled water or a 0.5% sodium citrate), treatments known to cause rapid loss of plasma membrane and DNA integrity by cell lysis and oxidation, respectively. Additional controls included normal 8-12-cell cryopreserved embryos that showed lethal cellular damage and lysis after thawing and rehydration.

Fluorescent microscopic determinations of mitochondrial distributions
Embryos where the occurrence of fragments could be timed by time-lapse microscopy during the pronuclear and early cleavage stages were stained with one of the following mitochondria-specific fluorescent probes (Molecular Probes) as described previously (Van Blerkom et al., 1995Go, 1998Go): nonyl-acridine orange (NAO), rhodamine 123 (R123). The distribution of mitochondria between fragments was determined by scanning laser confocal microscopy (SLCM) analysis of living embryos, after which embryos were returned to culture.

Statistical analysis
{chi}2-analysis was used to determine whether differences in clinical pregnancy rates were significant between patients in whom transfers involved only embryos which remained intact during culture, or only embryos which exhibited the fragmentation patterns and behaviours which were studied.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Time-lapse analysis
The development of 91 embryos (50 monospermic, 41 dispermic) that showed fragmentation during the pronuclear or early cleavage stages but which were stage-appropriate and largely normal in appearance on days 3.0–3.5 was documented by time-lapse video microscopy. The findings indicate two basic patterns of fragment formation, often occurring in the same embryo: definitive fragmentation and pseudo-fragmentation. Definitive fragmentation was characterized by the elaboration of fragments that were clearly detached from a blastomere(s). In contrast, pseudo-fragmentation, as defined here, was a transient event in which fragments were no longer detectable during subsequent development.

Fragment formation at the pronuclear stage
Figure 1A–HGo show a representative pattern of early fragmentation observed in monospermic (Figure 1AGo, n = 26) pronuclear oocytes that, on day 3, appeared intact and normal. In these instances, fragment formation was detected at syngamy (asterisks, Figure 1BGo) and during the first cell division (Figure 1C–EGo). For the embryo shown, images were captured at 1 min intervals over a 20 h period from the pronuclear (Figure 1AGo) to the 4-cell stage (Figure 1HGo, 39 h). However, by the 4-cell stage, virtually all fragments had disappeared resulting in a morphologically normal classification. Eighty-seven per cent (23/26) of these embryos progressed to the 12-cell stage by day 3.5. The apparent disappearance of clusters of `fragments' was a common feature of these embryos. For example, the cluster of fragments indicated by an asterisk in Figure 1GGo appeared during a 15 min interval between 33 h 56 min and 34 h 11 min, but was undetectable at 34 h 26 min. After the 4-cell stage (Figure 1HGo), fragments were undetectable by light microscopy on either the exposed surface or the interior of the embryo (as observed in optical sections) as cleavage progressed.



View larger version (211K):
[in this window]
[in a new window]
 
Figure 1. Representative time-lapse images of human embryos, which exhibited different temporal and spatial patterns of fragmentation during early preimplantation development. Times of observation after insemination are indicated in hours (h) and minutes (m).

 
Fragment formation during early cleavage
Sixty-five embryos (24 monospermic, 41 dispermic) examined by time-lapse microscopy developed fragments during early cleavage but were classified as normal by the 8-cell stage. Analysis of high-resolution images indicated that the location, size and apparent number of fragments changed during cleavage. For example, at 22 h 20 m after insemination, the syngamic egg shown in Figure 1IGo was virtually devoid of extracellular structures. However, 1 h later at the 2-cell stage, numerous, small fragments were localized to one hemisphere of each blastomere (arrow, Figure 1J–LGo). Time-lapse showed that fragments, both in clusters or as individual structures, often moved in concert with the subjacent blastomeres (compare images in 1N and O, taken at a 15 min interval), demonstrating that the relative position(s) of blastomeres and associated fragments was not static. Fragments detected on the apical surface of one blastomere appeared to move internally and often seemed to disappear, only to reappear on the exposed surface of the same blastomere some 75–120 min later. The rotation of a fragment cluster during a 15 and 60 min interval is indicated by arrows in Figure 1N–O and J–LGo, respectively. The embryo shown in Figure 1I–PGo is representative of a pattern of fragmentation in which virtually all detectable fragments present during early cleavage seemed to disappear from the apical surface of blastomeres by the 8-cell stage. However, for 38 embryos (eight monospermic, 30 dispermic), some of the fragments that had developed during the 2–4-cell stage were still detectable in small clusters or individually on apical surfaces on days 3.0 and 3.5, at which time these embryos had progressed to the 8-12-cell stage (e.g. Figure 1Q–TGo). The composition and fine structure of fragments elaborated during the pronuclear and early cleavage stages for the embryos we examined by time-lapse microscopy are described below.

Apparent fragment resorption and lysis
The rapidity with which fragments seemed to disappear in time-lapse images (within minutes) suggested that fragment loss may involve lysis or resorption. For cleavage stage embryos, evidence for lysis of small apical fragments (e.g. arrows, Figure 1J, MGo) which seemed to burst during intervals of image-capture was suggestive but not definitive. In contrast, clear evidence for lysis was obtained with larger fragments. For example, fragments (e.g. white arrows, Figure 1U, VGo; white asterisks, Figure 1W–YGo) which began to swell and display a transparent cytoplasm seemed to disappear in a `burst-like' fashion. Membranous remnants of these fragments were detectable in the adjacent perivitelline space (arrows, Figure 1W, XGo). The arrows in Figure 1YGo indicate the portion of the fragment from which cytoplasm was observed in the process of being discharged. During a subsequent 12 h culture period, other fragments in this representative embryo underwent swelling and apparent lysis. Time-lapse images provided suggestive evidence for fragment resorption, as these entities appeared to merge with the plasma membrane. A possible cellular basis for resorption was indicated by TEM findings described below.

Transmission electron microscopy
Based on the timing and behaviour of fragments observed by time-lapse, the following types of embryos (as shown in Figure 1Go) were examined by TEM: (i) monospermic embryos which appeared to contain `pseudo-fragments' (n = 8), (ii) monospermic embryos in which fragment lysis was suggested by changes in cytoplasmic texture and swelling (n = 9), and (iii) monospermic (n = 6) and dispermic (n = 36) embryos in which fragments detected during early cleavage were largely undetectable at the 8-12-cell stage.

The TEM images of fragmented embryos shown in Figures 2 and 3AGoGo were selected because they demonstrate fine structural characteristics common to all embryos which were examined. Spherical elements that occurred in columns were the most common and consistent type of fragment observed at the pronuclear and early cleavage stages. Typically, these structures exhibited a uniform cytoplasmic texture with virtually no organelles. Other fragments appeared lysed (L, Figure 2B, EGo) or were likely candidates for lysis (i.e. prelytic, Lp, Figure 3AGo). All lysed fragments were enclosed by a discontinuous plasma membrane (arrows, Figure 2FGo). Figure 2AGo is derived from an embryo similar to the one presented in Figure 1CGo, and shows fine structural characteristics of a typical cluster of fragments that arose during the pronuclear stage. Analysis of serial sections indicated virtually no mitochondria or other cytoplasmic components present in these structures. When organelles were present in columns of fragments, they were mostly confined to the fragment directly associated with the blastomere plasma membrane, such as fragment 3 in Figure 2DGo. Cytoplasmic continuity between fragments was another consistent TEM finding. In the selected images shown in Figures 2AGo (pronuclear stage), 2B (4-cell), 2D, (2-cell), 2E (8-cell) and 3A (12-cell), serial section analysis demonstrated cytoplasmic continuity between fragments (e.g. 1–3 in Figure 2A, DGo), as well as between columns of interconnected fragments and the underlying blastomere (e.g. fragment 3, Figure 2DGo; asterisks, Figure 2EGo; arrows, Figure 3AGo). For example, the two images shown in Figure 2CGo correspond to the region indicated by an arrow in the fragment labelled with a black asterisk in Figure 2BGo. Cytoplasmic continuity between columns of fragments and the underlying blastomere(s) observed by serial section analysis is also indicated by asterisks in Figure 2EGo and by a series of arrows in Figure 3AGo. Such complexes and columns of interconnected fragments may represent `pseudo-fragments' and continuity with the underlying cell could provide a basis for resorption. The occurrence of detached fragments, including those relatively few which contained cytoplasmic components such as mitochondria (M, Figure 2E, 3AGoGo), may result from breakage of cytoplasmic bridges during blastomere movement and rotation.



View larger version (177K):
[in this window]
[in a new window]
 
Figure 2. Transmission electron micrographs of fragmented human embryos at the pronuclear through mid-cleavage stages of development. PB2 = second polar body; M = mitochondria. A, x2000; B, x2000; C, x12 000; D, x2000; E, x2000; F, x20 000; G, x2500; H, x2500; I, x17 000.

 


View larger version (171K):
[in this window]
[in a new window]
 
Figure 3. (A) Representative transmission electron micrograph of a 12-cell dispermic embryo which showed apical clusters of fragments at the 2–4-cell stage but which appeared fragment-free at the 8-cell stage. (B-H) Light (B, D) and conventional epifluorescent images of embryos stained with DAPI (E) or annexin V (C, F, G, H). (I-K, M, O, P, S) Fully compiled scanning laser confocal microscopy images of terminal deoxynucleotidyl transferase-mediated dUTP nick end-labelled human oocytes and embryos (L, N, Q, R). PB1 = first polar body; MII = metaphase II chromosomes; N = nucleus; M = mitochondria. A, x2500.

 
Time-lapse observations demonstrated that fragment clusters detected on the apical surface of blastomeres changed position in concert with the movement and rotation of blastomeres during the cleavage stages. For TEM, of particular interest were the fate and location of fragments in embryos, which, by time-lapse, appeared fragmented at the 2–4-cell stage (day 1.5 and 2) but were largely fragment-free and stage-appropriate at the 8–12-cell stage (day 3.0 and 3.5). Figures 2E and 3AGoGo are TEM images of two such embryos, derived from monospermic and dispermic fertilizations, respectively, and are shown because fragments were confined to the interior of the embryo (n = 18); for five of these embryos, the presence of fragments between opposed blastomeres could potentially be developmentally significant. The internal location of fragments was undetectable by routine light microscopy. Lysed, `prelytic', and intact, organelle-free fragments interposed between blastomeres seemed to prevent the close opposition of adjacent plasma membranes permissive for the formation of occluding-type junctions, such as indicated by large arrowheads in Figure 3AGo). However, serial section analysis demonstrated that these separated zones were focal and that close contact between opposed plasma membranes existed in other regions of these blastomeres.

Serial section reconstructions of late cleavage-stage embryos indicated that associations between intact fragments involved membrane specializations (large arrow Figure 2EGo, and at higher magnification in 2GGo) similar in appearance to the occluding junctions between blastomeres in the same embryo(s). For fragments, this type of tight association could provide a mechanism for coordinated movement with an underlying blastomere(s) as detected by time-lapse or their apparent disappearance from apical surfaces if internalized, as indicated by TEM.

Analysis of plasma membrane and DNA integrity
Annexin V staining
Prior to fixation for TEM, TUNEL or `comet' analysis, embryos (n = 23) were stained with AV where time-lapse indicated (i) disappearance of fragments by the 8–12-cell stage (e.g. Figures 1P, 3BGoGo), or (ii) where fragments persisted through cleavage in otherwise stage-appropriate embryos (e.g. Figure 1TGo; arrow, 3DGo). None of these embryos exhibited blastomere-associated AV fluorescence on days 3 or 3.5 (Figure 3CGo). In contrast, where fragments that first appeared at the 2-cell stage were still evident during subsequent culture, AV fluorescence was detected in some fragments. The arrow in Figure 3DGo indicates a cluster of fragments in a 6-cell embryo (DAPI epifluorescent image showing four nuclei, Figure 3EGo) that first appeared at the 2-cell stage. One fragment in this cluster was clearly AV positive (arrow, Figure 3FGo). As a control for specificity, AV-negative embryos and embryos with an occasional positive fragment (arrow, Figure 3GGo) were restained after fixation in formaldehyde, or exposure to either hydrogen peroxide or hypotonic conditions. In all cases, AV fluorescence was associated with the plasma membrane of blastomeres and retained fragments. The epifluorescent image shown in Figure 3HGo is the same embryo as in Figure 3GGo after fixation and restaining with AV.

TUNEL assay
TUNEL analysis was performed on 41 stage-appropriate (8–12-cell) day 3.0–3.5 embryos where fragment clusters that arose during early cleavage were either no longer apparent (n = 14) or were still detectable to different extents (n = 27). Prior to TUNEL analysis, embryos were stained with Trypan Blue to confirm blastomere integrity. Embryos were prepared either as whole mounts (n = 19) or as air-dried specimens (n = 22). Equivalent sensitivity and specificity of the two methods was confirmed in other studies that used intact and zona-free unfertilized oocytes and dispermic embryos (J.Van Blerkom, unpublished observations). For example, the epifluorescent image of an unpenetrated oocyte prepared as a whole mount in Figure 3IGo shows TUNEL fluorescence localized to the first polar body. The oocyte shown in Figure 3JGo was an air-dried preparation that exhibited a TUNEL-positive polar body. After treatment with DNase, metaphase II chromosomal TUNEL fluorescence was detected in all oocytes (e.g. Figure 3KGo). None of the embryos prepared as whole mounts (Figure 3L, MGo) or air-dried (Figure 3N, OGo) showed nuclear TUNEL staining, and after DNase treatment, all nuclei were TUNEL positive (Figure 3PGo, the same embryo as in O; Figure 3LGo is the same embryo shown in Figure 1TGo). If present, the only TUNEL fluorescence detected in embryos was associated with the second polar body (arrow, Figure 3OGo).

As an additional control for the specificity of this assay, unfragmented 2–8-cell dispermic embryos (n = 12) were exposed to hypotonic conditions and fixed for TUNEL analysis at specific times after blastomere swelling and Trypan Blue incorporation were detected. Figure 3Q–RGo are representative images of a 4-cell dispermic embryo with mononucleated blastomeres as it appeared prior to (Figure 3QGo) and after 12 min (Figure 3SGo) in hypotonic conditions (Figure 3RGo). Nuclear TUNEL fluorescence in this embryo as observed in a fully compiled SLCM image is shown in Figure 3SGo.

`Comet' assay
For `comet' analysis, stage-appropriate embryos (monospermic, n = 6; dispermic, n = 17) were examined in which fragments that occurred during the 2–4-cell stage were still detectable at the 8–12-cell stage (e.g. Figure 1Q–TGo). The analysis was confined to these embryos because, in current clinical practice, the presence of fragments of this type could be considered indicative of reduced competence. Eighty-two per cent (172/207) of the blastomeres in 23 cleavage stage embryos prepared for `comet' were detected in the agarose minigels. Under the conditions and duration of electrophoresis used, none of the 172 nuclei displayed tails indicative of DNA fragmentation. A representative image of three fluorescent nuclei from the 8-cell embryo shown in Figure 4AGo is presented in Figure 4BGo. In contrast, `comet' tails (Figure 4CGo) were observed in all second polar bodies that were Trypan Blue positive in the living state (arrow Figure 4AGo) and could be identified in the agarose minigels prior to electrophoresis owing to significant dye incorporation.



View larger version (159K):
[in this window]
[in a new window]
 
Figure 4. Single-cell alkaline gel electrophoresis (`comet') analysis of DNA (B, C, F, I, J, K) in fragmented human embryos (A, D, G). Trypan Blue incorporation is noted by an arrow in A and asterisks in E. (H) scanning laser confocal microscopy image of the embryo shown in G after and terminal deoxynucleotidyl transferase-mediated dUTP nick end-labelling. (L-O) Representative images of apparently fragmented human embryos (L and N are the same embryo at the pronuclear and 4-cell stages respectively) stained with mitochondria-specific probes rhodamine 123 (M) or nonyl-acridine orange (O).

 
To confirm the ability of the `comet' assay to detect DNA cleavage under the conditions used, the following types of embryos were examined: (i) highly fragmented, dispermic embryos (n = 7, Figure 4DGo, `comet', 4FGo) in which significant Trypan Blue incorporation was observed several days after cleavage arrested (asterisks, Figure 4EGo); (ii) unfragmented monospermic 8–16-cell embryos (n = 13) that were normal and stage-appropriate prior to cryopreservation, but after thawing and rehydration, underwent cellular disruption and lysis (Figure 4GGo, `comet' patterns from similar embryos shown in 4IGo). These damaged embryos always contained TUNEL-positive nuclei (arrow, Figure 4HGo); (iii) intact 4–8-cell dispermic embryos exposed to hypotonic conditions (n = 3; Figure 4JGo) or hydrogen peroxide (n = 5; Figure 4KGo). In all instances, `comet' tails of varying thickness and length were detected.

Mitochondrial content of fragments
Eighteen embryos (four monospermic, 14 dispermic) examined by time-lapse developed fragment clusters at the pronuclear or early cleavage stages and were apparently fragment-free or contained apical clusters at the 8–12-cell stage as described above. Embryos were stained with R123 within 4 h after fragments were initially detected at the pronuclear or 2-cell stage (e.g. Figure 1QGo), and after SLCM analysis, were returned to culture and time-lapse observations continued. The pronuclear oocyte shown in Figure 4LGo is presented because of the apparent severity of fragmentation observed at the light microscopic level. However, the fully-compiled SLCM R123-fluorescent image of this embryo (Figure 4MGo) is representative of all 18 embryos that developed fragments during early development but which appeared largely normal on day 3 or 3.5. The absence of detectable mitochondrial fluorescence was characteristic of most of the fragments in these embryos (asterisks, Figure 4MGo). The relatively few fragments with detectable mitochondrial fluorescence are indicated by an arrow in Figure 4MGo. After SLCM examination, time-lapse analysis showed fragment resorption or lysis as described above, and if present during later development, only small clusters of fragments remained. For example, the embryo presented in Figure 4NGo (day 2) is the same embryo shown in Figure 4LGo and Figure 4OGo is a fully-compiled SLCM image after staining with NAO. The absence of fluorescence in most of the residual fragments (arrow, Figure 4NGo) was a consistent finding for progressive embryos which exhibited this pattern of fragmentation during early development.

Outcome after embryo transfer
In parallel with time lapse and biochemical analyses, the timing and pattern of fragmentation was systematically assessed in embryos destined for transfer or cryopreservation, by sequential observations (4–8 h) of individual embryos during 3 days of culture. Seventy-two day 3 embryos transfers to women between the ages of 28 and 37 years involved two 8–10-cell embryos with the pattern of fragmentation shown in Figure 1Go: (i) fragment clusters present during early cleavage were no longer detectable at transfer (type 1, n = 23) and (ii) some of the fragments that arose at the pronuclear or early cleavage stages were still detectable at transfer (type 2, n = 49). The clinical pregnancy rate for patients with type 1 and type 2 embryos was 61% (14/23) and 55% (27/49) respectively. Clinical pregnancy rates for a comparable patient population (n = 75) after the transfer of two normal-appearing (unfragmented during culture) 8–10-cell embryos was 67% (50/75). Although embryos that exhibited these two types of fragmentation patterns appeared to have an implantation rate lower than their unfragmented counterparts, these differences were not statistically significant.


    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
While fragmentation during the early preimplantation stages is a common phenomenon in embryos produced in clinical IVF, the causes of this cellular behaviour are unknown and its impact on developmental competence can be unclear. What is apparent, however, is that fragmentation is largely unpredictable and occurs in different media formulations and embryo culture systems (e.g. conventional and co-culture). Within cohorts of fertilized oocytes maintained in the same culture dish, fragmentation can affect none, some, or all embryos, demonstrating that this phenomenon is both embryo- and patient-specific. Empirically derived estimates of fragment number and size/volume are used clinically to select cleavage stage embryos for transfer or cryopreservation, and it has been generally assumed that embryos with higher fragmentation scores are less likely to implant and develop progressively than are their relatively unfragmented counterparts (Giorgetto et al., 1995Go). However, outcome with fragmented cleavage stage embryos indicates that when intact blastomeres are detected, fragmentation may not be an accurate measure of competence (Hoover et al., 1995Go; Alikani et al., 1999Go; Antczak and Van Blerkom, 1999Go; Gerris et al., 1999Go), suggesting that different patterns of fragmentation may have different aetiologies and effects on embryo competence.

Here, analytical findings and outcome results are reported with embryos that exhibited certain patterns of fragmentation not associated with early cleavage arrest. We focused on these specific phenotypes because they are both common in clinical IVF and their occurrence can be problematic in the assessment of competence. Time-lapse imaging from the pronuclear to the 10–12-cell stage provided new insights into the fragmentation phenomenon that cannot be obtained or appreciated when embryos are examined infrequently. Within cohorts, fragments formed in some embryos during the pronuclear or early cleavage stages but were undetectable at later stages, resulting in morphologically normal, stage-appropriate embryos. For other embryos, some of the fragment clusters that had occurred during early cleavage were still detectable on apical surfaces of blastomeres during late cleavage. Time-lapse images indicated that fragments seemed to be resorbed into the underlying blastomere(s), while others disappeared abruptly, leaving behind membranous ghosts and debris. Often, these dynamic activities occurred in the same embryo.

Static TEM images of dynamic activities in embryos offer a possible explanation for apparent fragment resorption and lysis observed in the living state. In the types of embryos that were examined, TEM showed that fragments often occurred in columns, a finding that confirms an arrangement we have previously observed by SLCM (Antczak and Van Blerkom, 1999Go). The cytoplasm between fragments in columns and clusters, and between these fragments and the underlying blastomere(s), was often continuous, suggesting a means by which resorption could occur. The apparent absence of mitochondria in most detached fragments could lead to lysis, if depleted ATP concentrations are not replenished (see below). The apparent movement of fragments was a common and consistent activity observed by time-lapse with clusters of fragments changing location in concert with the movement of the underlying cell. Fragments at the apical surface of one or more blastomeres moved inward as cell(s) rotated giving the embryo a largely fragment-free appearance. During cleavage, fragments seemed to reappear in some of these embryos. Although it was not possible to determine definitively whether fragments internalized at one time were the same ones that appeared somewhat later, analysis of time-lapse images, which included notations of number, size and relative position, strongly suggested the same cohorts re-emerged into view if the fragment-containing region of the affected blastomere(s) returned to an apical location. TEM images indicating that clusters of fragments were tightly associated by occluding-type junctions provide a possible structural basis for movement in concert with the corresponding blastomere(s).

The fine structural images of 8–12-cell embryos were selected because they showed intact fragments, lysed fragments, and likely candidates for lysis (termed prelytic) interposed between adjacent blastomeres. It has been suggested that if positioned between blastomeres, fragments could perturb the normal close association of opposed blastomere plasma membranes required for compaction, such that their mechanical extraction during early cleavage could correct potential defects in morulation and improve outcome after transfer (Alikani et al., 1999Go). Routine light microscopy could not determine whether the types of fragments we observed by TEM were present between blastomeres at the 8–12-cell stage. Consequently, it is unknown whether, if present in some embryos, they could represent an irreversible, focal barrier to compaction. One experimental approach to this question currently under investigation involves labelling fragments from one embryo with permanent fluorescent probes followed by deposition in intact recipient (dispermic) embryos and tracking of the donated fragments by sequential SLCM observation. However, intact embryos and embryos which exhibited the fragmentation patterns reported here had comparable outcomes. This finding indicates that the occurrence of these fragmentation phenotypes during early development is not necessarily developmentally lethal.

Activation of programmed cell death (PCD) pathways leading to blastomere apoptosis has been suggested to be a proximate cause of fragmentation in cleavage stage human embryos, especially when significant fragmentation is accompanied by cleavage arrest (Jurisicova et al., 1996Go) or is so extensive that no intact blastomeres are detectable (Yang et al., 1998Go). For intact 10-cell embryos, the detection of intense AV staining of the plasma membranes suggested to Levy et al. (1998) that early processes associated with PCD may be initiated in normal-appearing embryos. Therefore, even if the embryos described here appear normal on day 3, the possibility remains that lethal alterations in plasma membrane composition and DNA structure that could contribute to subsequent developmental failure may affect some blastomeres. The clinical relevance of this possibility is indicated by the detection of TUNEL-positive nuclei indicative of PCD and apoptosis (Hardy, 1999Go) in cultured morula and blastocyst stage human embryos. Here, day 3 and day 3.5 embryos were examined in which fragment clusters were classified as transient or persistent for evidence of impending cell death indicated by changes at the plasma membrane (AV staining) and DNA concentrations (TUNEL and `comet' assays).

The TUNEL assay is a widely used histochemical marker that detects DNA breakage associated with apoptosis. However, TUNEL staining does not reliably discriminate between cell death that results from necrosis, autolysis and apoptosis (Kressel and Groscurth, 1994Go; Charriaut-Marlangue and Ben-Ari, 1995; Grasl-Kraupp et al., 1995Go; Frankfurt et al., 1996Go). In addition, normal nuclei undergoing active transcription can also be TUNEL positive (Kockx et al., 1998Go). Consequently, the `comet' assay was also used, a method reported to be capable of identifying patterns of early DNA cleavage associated with different aetiologies of cell death. The `comet' assay is a single cell analysis of DNA integrity that is based upon the ability of denatured, cleaved DNA to migrate from the nucleus in an agrose gel when an electric current is applied. After electrophoresis, DNA is resolved with fluorescent stains and, if compromised, the fluorescent patterns resemble a comet, where tail characteristics such as shape and length can be used to determine the type of cell death (necrosis, lysis, apoptosis) and extent of DNA damage (Singh et al., 1988Go; Singh and Stephen, 1997). In contrast, intact, supercoiled DNA remains within the confines of the nucleus and appears as a highly fluorescent `head' with no detectable tail. The `comet' assay has been recently used to detect DNA cleavage induced in cultured hamster embryos by hydrogen peroxide treatment, ultraviolet irradiation, or prolonged exposure to visible light (Takahashi et al., 1999Go). Translocation or `membrane flipping' of phosphatidylserine (PS) from the inner to outer leaflets of the plasma membrane is an early and apparently characteristic indication of apoptosis (Martin et al., 1995Go). Externalization of this moiety can be identified after staining living cells with tagged AV. However, owing to loss of plasma membrane integrity, cells undergoing necrosis or lysis will also stain positively because they permit entrance of AV and labelling of PS on the inner leaflet of the plasma membrane. Collectively, these analytical methods should detect subtle changes in plasma membrane and DNA integrity in the classes of human embryos that were examined.

The results of the current study show no indication of nuclear DNA cleavage detectable by `comet' and TUNEL assays, or loss of plasma membrane integrity detectable by AV staining. TUNEL-negative embryos were all positive after DNase treatment, and both positive TUNEL fluorescence and `comets' were evident in second polar bodies, and in all embryos where DNA cleavage occurred as a consequence of cell lysis or was induced experimentally. Embryos, that were AV negative in the living state were all AV positive after fixation or exposure to hydrogen peroxide or hypotonic conditions. This may explain the intense plasma membrane staining observed by Levy et al. (1998) for intact 10-cell embryos, which were AV stained after fixation. AV-positive structures detected in the present study may be lysed fragments with a discontinuous plasma membrane, as indicated in TEM images.

The absence of detectable alterations at the plasma membrane and DNA concentrations, combined with time-lapse, TEM findings and outcome, indicate that the patterns of fragmentation described here are not characteristic of apoptosis or indicative of PCD. While the underlying cause(s) of the fragmentation patterns described here are unknown, it is intriguing that these activities resemble some of the events associated with another form of cell death, oncosis. The characteristics of oncosis in general and how they differ from apoptosis in particular have been described in detail by Majno and Joris (1995). When deprived of oxygen (ischaemic conditions), certain cell types elaborate plasma membrane blebs devoid of organelles. Over a period of hours, these blebs progressively swell by fluid uptake as unreplenished ATP reserves diminish with increased permeability resulting from failure of plasma membrane-associated ionic pumps. At the terminal stage of oncosis, DNA fragmentation occurs and when fluid accumulation reaches lethal amounts, the affected cell undergoes lysis. As observed in culture, oncotic blebs are continuous evaginations of the plasma membrane and, if they become detached from the affected cell, they `float away' (Majno and Joris, 1995Go). However, if normal conditions are restored sufficiently early, resorption of blebs and return of cell function and viability can occur.

Correlated time-lapse and electron microscopic analyses provide a possible aetiology for these fragmentation patterns and for the absence of adverse effects on competence. Most of the intact fragments examined by serial thin section reconstruction contained few, if any, detectable mitochondria. This observation was confirmed in living embryos with fluorescent probes that detected active mitochondria, or all mitochondria, regardless of activity, although some fluorescent fragments were always observed. It is suggested that the longevity of detached fragments may be related to mitochondrial complement and their relative contribution to ATP concentrations by oxidative phosphorylation. Swelling and lysis may result if unreplenished ATP stores decline below a critical threshold required to maintain plasma membrane ionic pumps and cytoplasmic integrity.

The fragmentation patterns observed by time-lapse microscopy and the TEM findings showed swelling of detached fragments prior to lysis and the occurrence of columns and clusters of organelle-free elaborations of the plasma membrane in which cytoplasmic continuity with the affected cell persisted. Taken together, these features seem to resemble the initial stages of an oncotic-like process. However, an oncotic-like process would seem to be an unlikely explanation for the fragmentation phenotypes described here because culture does not occur under conditions of severe oxygen deprivation and not all embryos are affected. One possible association with an ATP-driven oncosis-like process may be related to mitochondrial distributions in pronuclear oocytes and early blastomeres, where relatively large regions of cortical cytoplasm can be permanently or transiently devoid of mitochondria (Van Blerkom, 2000Go; Van Blerkom et al., 2000Go). For example, disproportionate mitochondrial segregation at the first cleavage division can result in some human blastomeres where the inherited complement of these organelles may be unable to contribute ATP at concentrations sufficient to maintain normal cell function and integrity in cortical regions (Van Blerkom et al., 2000Go). For the fragmentation patterns described, we are investigating whether mitochondria-depleted regions of pronuclear oocytes and blastomeres exist, and whether they experience a corresponding depletion of ATP that could have focal effects on plasma membrane function associated with the elaboration of mitochondria-free elements similar to nascent oncotic blebs. According to this notion, restoration of an appropriate mitochondrial density could result in resorption where cytoplasmic continuity with the underlying blastomere was maintained, or lysis of detached structures when ATP concentrations can no longer maintain membrane integrity.


    Acknowledgements
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
We thank Deborah Campbell, Sandra Hahn, Sharon Ross, and Jane Sinclair for the clinical contributions they provided at different times during the course of this study.


    Notes
 
3 To whom correspondence should be addressed at: Department of Molecular, Cellular and Developmental Biology, University of Colorado, Boulder, CO 80309, USA. E-mail: vanblerk{at}spot.colorado.edu Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Alikani, M., Cohen, J., Tomkin, G. et al. (1999) Human embryo fragmentation in vitro and its implications for pregnancy and implantation. Fertil. Steril., 71, 836–842.[ISI][Medline]

Antczak, M. and Van Blerkom, J. (1999) Temporal and spatial aspects of fragmentation in early human embryos: possible effects on developmental competence and association with the differential elimination of regulatory proteins from polarized domains. Hum. Charriaut-Marlangue, C. and Ben-Ari, Y. (1995) A cautionary note on the use of the TUNEL stain to determine apoptosis. Neuro. Rep., 7, 61–64.

Frankfurt, O., Robb, J., Sugarbaker, E. and Villa, L. (1996) Monoclonal antibody to single-stranded DNA is a specific and sensitive cellular marker of apoptosis. Exp. Cell Res., 226, 387–397.[ISI][Medline]

Gerris, J., De Neubourg, D., Magngelschots, K. et al. (1999) Prevention of twin pregnancy after in-vitro fertilization or intracytoplasmic sperm injection based on strict embryo criteria: a prospective randomized clinical trial. Hum. Reprod., 14, 2581–2587.[Abstract/Free Full Text]

Giorgetto, C., Terrou, P., Auquier, P. et al. (1995) Embryo score to predict implantation after in-vitro fertilization: based on 957 single embryo transfers. Hum. Reprod., 10, 2427–2431.[Abstract]

Grasl-Kraupp, B., Ruttkay-Nedecky, B., Koudelka, H. et al. (1995) In-situ detection of fragmented DNA (TUNEL Assay) fails to discriminate among apoptosis, necrosis, and autolytic cell death: a cautionary note. Hepatology, 21, 1465–1468.[ISI][Medline]

Hardy, K. (1999) Apoptosis in the human embryo. Rev. Reprod., 4, 125–134.[Abstract/Free Full Text]

Hoover, L., Baker, A., Check, J. et al. (1995) Evaluation of a new embryo-grading system to predict pregnancy rates following in vitro fertilization. Gynecol. Obstet. Invest., 40, 151–157.[ISI][Medline]

Jurisicova, A., Varmuza, S. and Casper, R. (1996) Programmed cell death and human embryo fragmentation. Mol. Hum. Reprod., 2, 93–98.[Abstract]

Kressel, M. and Groscurth, P. (1994) Distinction of apoptotic and necrotic cell death by in situ labelling of fragmented DNA. Cell Tissue Res., 278, 549–556.[ISI][Medline]

Kockx, M., Muhring, J., Knaapen, M. and de Meyer, G. (1998) RNA synthesis and splicing interferes with DNA in situ end labeling techniques used to detect apoptosis. Am. J. Pathol., 152, 885–888.[Abstract]

Levy, R., Benchaib, M., Cordonier, H. et al. (1998) Annexin V labelling and terminal transferase-mediated DNA end labelling (TUNEL) assay in human arrested embryos. Mol. Hum. Reprod., 4, 775–786.[Abstract]

Majno, G. and Joris, I. (1995) Apoptosis, oncosis, and necrosis: An overview of cell death. Am. J. Pathol., 146, 3–15.[Abstract]

Martin, S., Reutelingsperger, C., McGahon, A. et al. (1995) Early redistribution of plasma membrane phosphatidylserine is a general feature of apoptosis regardless of the initiating stimulus: Inhibition by overexpression of Bcl-2 and Abl. J. Exp. Med., 182, 1545–1556.[Abstract]

Morgan, K., Wiemer, K., Steurwald, N. et al. (1995) Use of videocinematography to assess morphological qualities of conventionally cultured and cocultured embryos. Hum. Reprod., 10, 2371–2376.[Abstract]

Perez, G., Tao-X-J. and Tilly, J. (1999) Fragmentation and death (a.k.a. apoptosis) of ovulated oocytes. Mol. Hum. Reprod., 5, 414–420.[Abstract/Free Full Text]

Singh, N. and Stephens, R. (1997) Microgel electrophoresis: Sensitivity, mechanisms, and DNA electrostretching. Mutat. Res., 383, 167–175.[ISI][Medline]

Singh, N., McCoy, M., Tice, R. and Schneider, E. (1988) A simple technique for quantitation of low levels of DNA damage in individual cells. Exp. Cell Res., 175, 184-191.[ISI][Medline]

Takahashi, M., Saka, N., Takahashi, H. et al. (1999) Assessment of DNA damage in individual hamster embryos by comet assay. Journal?, 00, 000-000.

Takase, K., Ishikawa, M. and Hoshiai, H. (1995) Apoptosis in the degeneration process of unfertilized mouse ova. Tohoku J. Exp. Med., 175, 69–76.[ISI][Medline]

Van Blerkom, J. (1993) Development of human embryos to the hatched blastocyst stage in the presence or absence of a monolayer of Vero cells. Hum. Reprod., 8, 1525–1539.[Abstract]

Van Blerkom, J. (2000) Intrafollicular influences on human oocyte developmental competence: perifollicular vascularity, oocyte metabolism, and mitochondrial function. Hum. Reprod., 15 (Suppl. 2), 133–188.

Van Blerkom, J. and Davis, P. (1998) DNA strand breaks and phosphatidylserine redistribution in newly ovulated and cultured mouse and human oocytes: occurrence and relationship to apoptosis. Hum. Reprod., 13, 1317–1324.[Abstract]

Van Blerkom, J., Davis, P., Merriam, J. and Sinclair, J. (1995) Nuclear and cytoplasmic dynamics of sperm penetration, pronuclear formation and microtubule organization during fertilization and early preimplantation development in the human. Hum Reprod. Update, 1, 429–461.[Abstract]

Van Blerkom, J., Sinclair, J. and Davis, P. (1998) Mitochondrial transfer between oocytes: potential applications of mitochondrial donation and the issue of heteroplasmy. Hum. Reprod., 13, 2857–2868.[Abstract/Free Full Text]

Van Blerkom, J., Davis, P. and Alexander, S. (2000) Differential mitochondrial inheritance between blastomeres in cleavage stage human embryos: determination at the pronuclear stage and relationship to microtubular organization, ATP content and developmental competence. Hum. Reprod., 15, 2621–2633.[Abstract/Free Full Text]

Yang, H., Hwang, K., Kwon, H. et al. (1998) Detection of reactive oxygen species (ROS) and apoptosis in human fragmented embryos. Hum. Reprod., 13, 998-1002.[Abstract]

Submitted on August 10, 2000; accepted on January 15, 2001.