Atypical initial cleavage patterns minimally impact rhesus macaque in vitro embryo morphokinetics and embryo outgrowth development

Abstract

Embryo morphokinetic analysis through time-lapse embryo imaging is envisioned as a method to improve selection of developmentally competent embryos. Morphokinetic analysis could be utilized to evaluate the effects of experimental manipulation on pre-implantation embryo development. The objectives of this study were to establish a normative morphokinetic database for in vitro fertilized rhesus macaque embryos and to assess the impact of atypical initial cleavage patterns on subsequent embryo development and formation of embryo outgrowths. The cleavage pattern and the timing of embryo developmental events were annotated retrospectively for unmanipulated in vitro fertilized rhesus macaque blastocysts produced over four breeding seasons. Approximately 50% of the blastocysts analyzed had an abnormal early cleavage event. The time to the initiation of embryo compaction and the time to completion of hatching was significantly delayed in blastocysts with an abnormal early cleavage event compared to blastocysts that had cleaved normally. Embryo hatching, attachment to an extracellular matrix, and growth during the implantation stage in vitro was not impacted by the initial cleavage pattern. These data establish normative morphokinetic parameters for in vitro fertilized rhesus macaque embryos and suggest that cleavage anomalies may not impact embryo implantation rates following embryo transfer.

Cleavage dimorphisms are prevalent in in vitro fertilized rhesus macaque embryos and these early cleavage anomalies minimally impact embryo morphokinetics and the events of implantation in vitro.

Graphical Abstract

Graphical Abstract.

Introduction

Embryo morphology is the primary evaluation criterion for grading embryo quality and selecting embryos for transfer to a human patient. In 2018, the live birth rate following embryo transfer in the USA ranged from 10.5% to 51.7% depending on patient age [1]. Removing an embryo from the incubator for visual assessment may only provide a snapshot of its morphology at that time. Time-lapse embryo imaging has been envisioned as a means to better select competent embryos for embryo transfer based on morphological evaluation and morphokinetics throughout the duration of culture. Morphokinetic assessment entails the annotation of the timing of developmental events including pronuclear fading, initial and early cleavage events, cellular compaction, blastocyst formation, blastocyst expansion, and initiation of herniation through the completion of blastocyst hatching. Time-lapse imaging has enabled the investigation of morphokinetic parameters that may be predictive of embryo competence, yet there remain challenges in the clinical application of this technology due to differences in selection parameters and reproducibility across laboratories [2]. Despite conflicting evidence, the use of time-lapse systems and morphokinetics to select embryos has been shown to improve live birth rates following embryo transfer in some studies [3–5].

Atypical initial cleavage morphologies were discovered through time-lapse imaging of human embryos [6–9]. Cleavage anomalies include direct cleavage of the one-cell embryo to more than two daughter cells and reverse cleavage where a cell may divide and then those cells re-fuse [2, 8–10]. Embryos with atypical cleavage patterns tend to have poorer blastocyst formation and implantation rates [8, 10]. In addition, euploid blastocysts with an atypical cleavage pattern have approximately half the live birth rate as normally cleaving euploid blastocysts [11]. Thus, embryos with cleavage anomalies might be deselected for embryo transfer.

Atypical cleavage anomalies have been observed in rhesus macaque in vitro produced embryos [12, 13], yet the developmental competence of blastocysts with cleavage anomalies remains unexplored. The objectives of the present study were to establish a normative morphokinetic database for unmanipulated in vitro fertilized rhesus macaque embryos and to assess the impact of atypical early cleavage patterns on in vitro development of embryo outgrowths. A retrospective morphokinetic analysis of rhesus macaque in vitro produced embryos that reached the blastocyst stage revealed that approximately half of the embryos had an atypical cleavage pattern. Atypically cleaving embryos had a delay in the onset of compaction, a shorter duration between compaction and the initiation of blastulation, and took longer to complete hatching once herniation was observed. Overall, the morphokinetic signatures, embryo hatching, rate of embryo attachment to the extracellular matrix and embryo growth in an in vitro implantation system were similar irrespective of initial cleavage, suggesting that cleavage anomalies do not impact in vitro embryo outgrowth developmental competence.

Methods

Animals

Rhesus macaques (Macaca mulatta) used in this study were from the colony maintained by the Wisconsin National Primate Research Center. All procedures were performed in accordance with the NIH Guide for the Care and Use of Laboratory Animals and under the approval of the University of Wisconsin College of Letters and Sciences and Vice Chancellor Office for Research and Graduate Education Centers Institutional Animal Care and Use Committee. A total of 28 females that were 4.10–14.25 years of age and weighing 4.78–10.13 kg and 6 males that were 6.83–21.50 years of age and weighing 8.80–14.44 kg were enrolled in this study to obtain gametes for in vitro fertilization (Supplementary Table 1).

Ovarian stimulation

Rhesus macaque ovarian stimulation and conventional in vitro fertilization (cIVF) methods implemented in this study have been previously described [14]. Female oocyte donors underwent ovarian hyperstimulation by administration of 15 IU recombinant human FSH (follitropin beta, manufactured for Merck Sharp & Dohme Corp, Whitehouse Station, NJ, USA) twice daily for 8–9 days followed by one injection of 1000 IU recombinant human choriogonadotropin (hCG; Ovitrelle, Merck Serono Ltd, Middlesex, UK) the next day. Laparoscopic aspiration of oocytes was performed between 30 and 32 h post-hCG injection.

In vitro fertilization

Oocytes that contained a germinal vesicle or were at metaphase I (MI) were placed in maturation medium consisting of CMRL (Thermo Fisher Scientific, cat no. 11530037) medium supplemented with 0.5 mM sodium pyruvate (Sigma-Aldrich, cat no. 2256), 2 mM l-glutamine (Sigma-Aldrich, cat no. G8541), and 20% fetal bovine serum (FBS; Peak Serum, cat no. PS-FB1). Metaphase II (MII) oocytes were placed directly into fertilization medium consisting of IVF-TL medium (Caisson Laboratories, cat no. IVL02) supplemented with 0.5 mM sodium pyruvate, 2 mM l-glutamine, and 0.1 mg/ml polyvinyl alcohol (Sigma-Aldrich, cat no. P8136). Oocytes in maturation medium were re-evaluated at ~36 h post-hCG injection and MI oocytes that progressed to MII, as evident by polar body extrusion, were transferred to fertilization medium.

Semen was collected from males 1 h prior to the oocyte retrieval and was processed as previously described [15]. Coagulum was removed following a 30-min incubation period at 37°C to allow for liquefaction. Sperm motility was estimated utilizing a computer-assisted sperm analysis system (CASA, Hamilton Thorne, Beverly, MA). Samples were washed twice in TL-hepes (Caisson Laboratories, cat no. IVL01) supplemented with 0.1 mM sodium pyruvate and 3 mg/ml bovine serum albumin (Sigma-Aldrich, cat no. A8806). A total of 20 × 10⁶ sperm were added to 2 ml of capacitation medium consisting of IVF-TL supplemented with 0.1 mM sodium pyruvate and 0.1 mg/ml polyvinyl alcohol. To facilitate capacitation, a final concentration of 1 mM caffeine (Sigma-Aldrich, cat no. C0750) and 0.5 mM db-cAMP (Sigma-Aldrich, cat no. D0627) were added to the capacitation medium 30 min prior to fertilization. At the time of fertilization, 2.5 μl sperm, 2 μl 100 mM caffeine, and 2 μl 50 mM db-CAMP were added to each 93.5-μl fertilization drop.

Embryo culture and time-lapse microscopy

Gametes were co-cultured for 16 h and then presumptive zygotes were washed and transferred to individual wells of CultureCoin dishes (Esco Medical, Denmark). Each CultureCoin contains 14 individual microwells with 25 μl culture media overlayed with 3 ml of paraffin oil (Ovoil, Vitrolife, cat no. 10029). Culture medium consisted of G-TL medium (Vitrolife, cat no. 10145) supplemented with 5% FBS. CultureCoins were then placed into a Miri Time-lapse (TL) incubator (Esco Medical) maintained at 37°C, 5% CO₂, and 5% O₂. Each well of the CultureCoin containing an individual embryo was imaged every 5 min at five different focal planes under a 20× Zeiss objective. Embryos were kept in the Miri-TL incubator for ~8–9 days or until reaching the hatched blastocyst stage.

Annotation of embryo development events

Time-lapse videos of each embryo were analyzed to determine specific time points of embryo development. Embryo morphokinetic events were analyzed using the standardized nomenclature outlined by Ciray et al. [16] for human embryos; nomenclature established by a group of embryo time-lapse users. Time points for annotation included visualization of pronuclear fading (tPNf), individual early cleavage divisions (t2–t8), first evidence of compaction (tSC), initiation of blastocyst formation (tSB), full blastocyst formation (tB), blastocyst expansion (tE), blastocyst herniation (tHN), and completion of blastocyst hatching (tHD).

Each annotation was recorded at the first indication of each development event and precise time points were obtained from the Miri Time-lapse computer software. Each video was annotated by two to three research technicians and the time stamps for each annotation were averaged. If there were >1 h discrepancies in the time stamp of the annotated event, the video was re-evaluated by the research team to reach a decision on the annotation. Initiation of compaction was the event most scrutinized and the full morula stage was not annotated since the degree of compaction varied greatly across embryos regardless of early cleavage pattern. Embryos that could not be clearly visualized at a specific time point were excluded from analysis. Blastocysts that failed to reach the full blastocyst stage and expand were removed from annotations proceeding the initiation of blastocyst formation (tSB).

The cIVF method used in this study limited visualization of some developmental events that are commonly observed in a human clinical setting. While pronuclei fading was annotated for the majority of embryos, second polar body extrusion, pronuclei fusion, and syngamy occurred prior to and during the processing of zygotes before imaging was initiated. Macaque pre-implantation embryo development is longer than that of humans [17], so while the MIRI-TL is capable of imaging for 200 h, a second time-lapse needs to be initiated to ensure there is no interruption in imaging. The initiation of a second time-lapse often occurred during blastocyst expansion creating a brief delay in imaging, thus analysis of blastocyst collapse and re-expansion was not annotated in this study.

Extended in vitro culture of hatched blastocysts

Embryos were cultured as previously described [14, 18]. In brief, blastocysts were plated on Matrigel within 12 h of hatching and cultured in 200 μl of buffalo rat liver (BRL)-conditioned medium. Media were exchanged on days 2, 6, and 10 post-plating. Prior to media exchange, embryo outgrowths were imaged using a Nikon Eclipse TE300. The diameter and area of the outgrowths were calculated using 4× images and ImageJ software where a ratio of 1 mm = 620 pixels. Embryos that grew up the side of the well were excluded from the analysis.

Statistical analysis

The proportions of abnormal and normal embryos were analyzed using the fmsb package in RStudio to perform pairwise Fisher exact tests with a post hoc Bonferroni correction for multiple comparisons. A t-test was performed to compare the mean diameter and area for embryo outgrowths using GraphPad Prism software version 9.

Results

Incidence of normal versus abnormal first cleavage events

Time-lapse image analysis was completed for 145 cIVF rhesus macaque blastocysts produced from 28 oocyte donors and 6 sires over four breeding seasons (September–May). Blastocyst demographics including embryos per season and oocyte donor ovarian stimulation event (females may undergo ovarian stimulation up to three times); oocyte donor age, sperm donor, and sperm donor ages are provided in Supplementary Table 1. Embryos were individually cultured in a MIRI-TL incubator (Figure 1A) and morphokinetic analysis of embryo developmental milestones was performed retrospectively. The annotation of developmental events with representative morphologies are illustrated in Figure 1B and C. The blastocysts served as controls for other studies in which embryos may have been removed throughout the duration of culture limiting the ability to accurately report development rates; control cleavage and blastocyst rates (~58% and 31%, respectively) for our system have been previously reported [14].

Figure 1.

Annotation of in vitro embryo development events. (A) Image of the MIRI-TL incubator with 6 chambers capable of culturing up to 14 embryos per chamber. (B) Overview of the annotated embryo developmental events. (C) Representative embryo morphologies for each annotated event. A circle encompasses the fading pronuclei in tPNf and the circle in tSB encompasses the formation of the blastocoel at the onset of blastocyst formation.

As similarly observed in human embryos, atypical or abnormal cleavage patterns were observed in macaque embryos with representative images shown in Figure 2A. Approximately 50% of the analyzed blastocysts cleaved normally from 1-cell to 2-cell (Supplementary Video 1) and ~50% of the analyzed blastocysts had an abnormal early cleavage event (Supplementary Videos 2–5; Figure 2B). Cleavage anomalies observed included (i) direct cleavage to more than two cells in the first division (DC1; 73.6%), (ii) direct cleavage in the second division (DC2; 11.1%), (iii) DC1 followed by reverse cleavage (RC) and DC2 (DC1, RC, DC2; 8.3%), and (iv) uneven division to two cells followed by DC2 (uneven 1, DC2; 6.9%). Cell fragmentation was visible during initial cleavage and throughout development; a common feature previously observed in rhesus macaque in vitro produced embryos [13]. The proportion of the total blastocysts with cell fragmentation was similar between those with a normal versus an abnormal early cleavage event (27.6% vs. 28.3%, respectively). The proportion of blastocysts with a normal versus abnormal cleavage pattern was not significantly different between breeding seasons, the number of ovarian stimulations the oocyte donor had undergone, or sperm donor (Figure 2C, D, F). Oocyte donors of age 12–15 years had a significantly higher proportion of blastocysts with an abnormal early cleavage pattern compared to oocyte donors that were less than 6 years old (Figure 2E).

Figure 2.

Incidence of normal versus abnormal initial cleavage. (A) Representative images of zygotes prior to and following the first cleavage event. (B) Proportion of blastocysts that cleaved in a normal pattern from one- to two-cell versus those with an abnormal first cleavage event (n = 145 blastocysts). (C) Distribution of the types of cleavage anomalies observed. DC1: direct cleavage I in the first division, DC2: direct cleavage in the second division, RC: reverse cleavage, uneven1: uneven size of the two cells in the first division event. Proportion of blastocysts with an abnormal initial cleavage by (D) cIVF season, (E) oocyte donor age, (F) oocyte donor’s ovarian stimulation event, and (G) sperm donor. Significance is denoted on the bar graph for donor age as determined by pairwise Fisher exact tests with a Bonferroni correction for multiple comparisons.

Morphokinetic parameters for normal versus abnormal cleaving embryos

A total of 73 blastocysts that had a normal cleavage pattern were analyzed to establish a morphokinetic timeline for normally cleaving rhesus macaque in vitro fertilized embryos. The mean time from insemination (t0) to initial cleavage (t2) was 22.15 h (Table 1). The mean time to t3, t4, t5, t6, t7, t8, tSC, tSB, tB, tE, tHN, and tHD from the time of insemination are provided in Table 1. The developmental timeframe of individually cultured and time-lapse imaged in vitro fertilized embryos analyzed in this study aligns with previous observations of in vitro [19, 20] and in vivo [21, 22] fertilized rhesus macaque embryos (Table 2).

Table 1.

Timing of developmental events of blastocysts with a normal versus abnormal initial cleavage event

Event	Normal		Abnormal		p-value
	Mean hpf ± SD	n	Mean hpf ± SD	n
tPNf	21.16 ± 3.93	41	22.29 ± 3.45	38	0.1781
t2a	22.15 ± 4.05	61	23.20 ± 5.15	64	0.2066
t3	31.53 ± 5.49	59
t4	34.21 ± 8.16	73
t5	45.66 ± 11.43	65
t6	48.93 ± 12.52	70
t7	57.07 ± 15.36	62
t8	62.65 ± 20.08	72
tSC	106.70 ± 11.18	72	111.30 ± 13.58	72	0.0280
tSB	139.30 ± 11.00	68	140.70 ± 13.58	72	0.5164
tB	151.60 ± 12.41	69	154.30 ± 14.26	63	0.2403
tE	151.70 ± 12.30	69	154.40 ± 14.27	63	0.2586
tHN	185.40 ± 14.02	53	188.40 ± 18.92	38	0.3879
tHD	193.40 ± 16.32	43	202.20 ± 20.82	36	0.0386

^aTime to first division event from 1-cell to 2-cell for normal embryos or first division for embryos with abnormal early cleavage divisions. A t-test was performed for each comparison. The bold text denotes p-values < 0.05.

Table 2.

Overview of the timing of developmental stages of in vitro and in vivo fertilized rhesus macaque embryos

		Timeline of embryo development
	Day	1	2	3	4	5	6	7	8	9
	Hours	24	48	72	96	120	144	168	192	216
In vitro	Present study with TLI	2–8 C	5–8 C	7 C − CP	CP	EA B − EX B	EA B − EX B	EX B − HN B	HN − HD B	HD B
	Ramsey and Hanna, 2019	1 C	2 C	8–12 C	12–16 C	CP	EA B	B	EX B	HD B
	Wolf et al. 1989	1 C	4–8 C	16–32 C	16 C − M	M	B	B	B	HD B
In vivo (flushed)	Seshagiri et al. 19931				8–16 C	8 C − EA B	8 C − B	B
	Goodreaux et al. 19902				M	EA B	EX B

Abbreviations: TLI, observations via time-lapse imaging; C, cell; CP, compacted embryo; M, morula; EA B, early blastocyst; B, full blastocyst; EX B, expanded blastocyst; HN B, herniated blastocyst; HD B, hatched blastocyst [1]. Day 0 is the day following the LH surge [2]. Developmental timing is reported as days post-ovulation.

The morphokinetics of blastocysts with a normal versus abnormal early cleavage pattern were compared to determine if early cleavage anomalies impacted the timing of blastocyst formation and/or embryo hatching. The timing of the initial cleavage event (t2) was similar between normal and abnormally cleaving embryos (22.15 h vs. 23.20 h, p = 0.2066; Tables 1 and 3); however, a greater, but not significant, proportion of normally versus abnormally cleaving embryos had undergone the first cleavage event prior to being cultured in the time-lapse embryo incubator (16.4% vs. 9.7%, respectively). Of note, the t2 measure is biased toward the later cleaving embryos as the early cleaving embryos were not measured. The onset of compaction (tSC) was significantly delayed in abnormally cleaving embryos (Table 1). The initiation of blastocyst formation (tSB), time to full blastocyst (tB), initiation of blastocyst expansion (tE), and initiation of blastocyst herniation (tHN) was comparable between normally and abnormally cleaving embryos, although these events tended to be delayed in abnormally cleaving embryos (Table 1). Abnormally cleaving embryos were significantly slower to complete hatching taking ~6 h longer than normally cleaving embryos (p = 0.0386, Table 1).

Table 3.

Time interval between developmental milestones of blastocysts with a normal versus abnormal early cleavage event

Interval	Mean h ± SD (n)		p-value
	Normal	Abnormal
t2–t0	22.15 ± 4.05 (61)
t3–t2	11.47 ± 5.43 (54)
t4–t3	5.13 ± 8.99 (55)
t5–t4	11.37 ± 7.85 (65)
t6–t5	4.01 ± 4.11 (61)
t7–t6	7.11 ± 8.35 (60)
t7–t8	9.52 ± 12.64 (59)
Compaction to blastulation (tSC–tSB)	33.53 ± 11.91 (64)	29.34 ± 12.14 (72)	0.0420
Blastulation (tB–tSB)	12.44 ± 7.58 (66)	14.77 ± 8.86 (63)	0.1124
Expansion (tHN–tE)	35.73 ± 10.08 (53)	38.71 ± 16.46 (38)	0.2866
Herniation (tHD–tHN)	8.47 ± 6.88 (43)	12.93 ± 11.90 (33)	0.0439

The time interval was calculated by subtracting the time to initiation of the event between two events. A t-test was performed for each comparison. The bold text denotes p-values < 0.05. The interval between early events for abnormally cleaving embryos was not calculated due to the division of an individual cell to multiple cells in one cleavage event.

The time interval between developmental events was analyzed for blastocysts arising from normal and abnormal early cleavage patterns. As shown in Table 3, the time interval between events was calculated by determining the time difference between the initiation of two events. The interval between early cleavage events was not annotated in abnormally cleaving embryos as the majority of the embryos underwent direct cleavage to multiple cells simultaneously; however, the time interval between later stages of development could be compared. The duration between the onset of compaction to the onset of blastocyst formation was significantly longer in normally cleaving embryos, although the duration of blastulation and expansion tended to be shorter for normal versus abnormally cleaving embryos (Table 3). The period of time from initial herniation from the zona pellucida to fully hatching was significantly longer for embryos with an abnormal early cleavage pattern (Table 3, p = 0.0439). Approximately 27.4% of blastocysts that had an abnormal initial cleavage pattern failed to hatch from their zona pellucida, whereas 14.3% of blastocysts with a normal initial cleavage failed to hatch (p = 0.0812, Supplementary Table 2).

Impact of initial cleavage pattern on in vitro development of embryo outgrowths

To assess the developmental impact of an abnormal early cleavage pattern on in vitro development of embryo outgrowths, hatched blastocysts were plated to Matrigel within 12 h of hatching and cultured for an additional 10 days to assess embryo attachment, survival, and growth. Representative images of embryo outgrowths on days 2, 6, and 10 of the extended embryo culture period are shown in Figure 3A. There were no differences between the proportion of blastocysts with a normal or abnormal initial cleavage pattern that attached to the extracellular matrix or survived to days 2, 6, or 10 of culture (Figure 3B and C). Embryo outgrowth diameter and area were also similar; however, the abnormal blastocysts tended to have a greater area at days 6 and 10 of the extended culture period. There were no significant differences in hatched-blastocyst developmental parameters in the extended culture period between embryos that had cell fragmentation with either normal or abnormal initial cleaving blastocysts (Supplementary Figure 1).

Figure 3.

Embryo outgrowth development of blastocysts with a normal versus abnormal initial cleavage event. (A) Representative images of embryo outgrowths at days 2, 6, and 10 of culture. The tracing in each image marks the outer edge of the outgrowth used for measurements in ImageJ. (B) Graphs of embryo growth parameters including attachment to the Matrigel matrix, survival, outgrowth diameter, and outgrowth area. (C) Metrics of data shown in graphs of panel B. A t-test was performed to compare the mean diameter and area at each time point and a Fisher exact test was used to compare the proportion of embryos that attached to the matrix and survived.

Discussion

This study presents normative morphokinetic data for in vitro fertilized rhesus macaque embryos in which abnormal initial cleavage was observed in ~50% of embryos that developed to the blastocyst stage. While the timing of the initial cleavage was similar between normal and abnormally cleaving embryos, more normally cleaving embryos had completed the first cleavage division prior to the start of imaging analysis. Normally cleaving embryos tended to initiate compaction sooner and spent a greater duration compacting prior to the onset of blastocyst formation compared to abnormally cleaving embryos. Blastocysts with an abnormal early cleavage event took longer to complete hatching, and the proportion of successfully hatched blastocysts trended toward being significantly reduced compared to normally cleaving embryos. An abnormal early cleavage pattern, however, did not impact in vitro hatched blastocyst attachment to the extracellular matrix or embryo growth through 10 days of extended culture. The present study retrospectively analyzed blastocysts that were cultured as controls from other studies focusing on experimental infection, genome editing, and/or transcriptomic analysis where the terminal endpoints did not allow for extensive characterization of the embryo outgrowths (i.e., cell number or chorionic gonadotropin secretion). While our preliminary analysis suggests that in vitro developmental competence was similar between blastocysts with differing initial cleavage patterns, future studies are needed to evaluate in vivo developmental competence between embryos with different cleavage patterns.

The developmental morphokinetics of in vitro fertilized rhesus macaque embryos in the present study aligns with other observations of in vitro fertilized and in vivo flushed embryos in which time-lapse imaging was not used. In the present study, in vitro fertilized macaque embryos initially cleaved from one to two cells at 22.15 h post-insemination, although this measure is biased toward the later cleaving embryos as the first cleavage event was not captured from some early cleaving embryos. Despite this, our t2 measure is consistent with a previous report by Burreul et al. [12] where macaque oocytes fertilized by ICSI cleaved at 22.5 h post-insemination. Across that and the current study, the timing of t2 is similar despite differences in fertilization method (ICSI vs. cIVF). In addition, our observations on the developmental timing of compaction and development to the blastocyst stage are comparable to the in vivo observations reported by Goodeaux et al. [21], although the precise timing of fertilization in vivo is unknown. The differences in the developmental timing across studies of in vitro fertilized embryos could be attributed to the reports in days versus hours (i.e., reporting in 24-h intervals from the time of fertilization) and may also be attributed to certain culture media formulations accelerating the developmental program [12, 23].

The pre-implantation developmental program is longer in rhesus macaque compared to human embryos. Human embryos reach the blastocyst stage on day 4 and implantation occurs on days 6–8, whereas the initiation of blastocyst formation occurs on days 5–6 in macaques with implantation occurring on days 8–10 [17]. The timing of early developmental events is relatively similar between human [24] and macaque embryos (present study); however, as development progresses, the later stages of development, including the time to compaction, blastulation, and hatching, occur much later in macaque embryos. The time to pronuclear breakdown occurs at ~23 h post-fertilization in human embryos compared to ~21 h in macaque, and the time to 8-cell is 55–58 h in human versus ~62 h in macaque. In human embryos, compaction, early blastocyst formation, blastocyst expansion, and completion of hatching occurs at ~86, 95–105, 105–123, and 123 h, respectively, whereas in the present study the timing of these events lags in macaques by at least 20 h and occurs at ~106, 139, 151, and 193 h, respectively. Thus, the divergence in the timing of the developmental program between humans and rhesus macaques seems to occur in the later stages of development.

Atypical cleavage errors have been previously observed in in vitro fertilized rhesus macaque embryos. Burruel et al. [12] reported that 20% of the rhesus macaque blastocysts in their study had an abnormal cleavage pattern and the authors defined criteria for scoring errors into three different types. In the present study, ~50% of the blastocysts had an abnormal initial cleavage pattern and the types of errors were similar to those described by Burruel et al. [12] with type 1 errors (direct cleavage from one cell to more than two cells) being the most commonly observed error in both studies. The proportion of blastocysts with an abnormal early cleavage pattern in the present study was consistent across four IVF seasons. A greater proportion of abnormally cleaving embryos were derived from oocyte donors 12–15 years of age. Limited sample size in the present study prevented the evaluation of maternal age effects on developmental morphokinetics. Embryos derived from human patients of advanced maternal age have similar early cleavage kinetics [25], and no differences were observed in the incidence of irregularity in the first division [26]. Despite no differences in the proportion of abnormal blastocysts across sperm donors, the majority of the males in the present study were 12 years or older. The effect of paternal age on the proportion of abnormal embryos cannot be evaluated in this study as younger males were not included in this analysis. A greater proportion of abnormal blastocysts observed in the present study could also be attributed to several differences in study design, including fertilization by gamete co-culture versus ICSI, the difference in culture systems, or ovarian stimulation regimen.

A systematic review of human embryo morphokinetic studies concluded that aneuploid embryos display delayed timing of cleavage to the 8-cell stage, 9-cell stage, time to full blastocyst formation, and time to expanded blastocyst [2]. A limitation of the present study is that embryo ploidy status was not evaluated as the majority of the embryos were utilized for other studies. Previous studies have reported that approximately 50–73% of rhesus macaque IVF-derived blastocysts contain aneuploid blastomeres, a proportion that is similar to human IVF embryos [13, 27]. A retrospective analysis of human embryos revealed that a significantly higher proportion of blastocysts with abnormal cleavage were euploid compared with blastocysts with a normal cleavage pattern (44.6% vs. 33.1%) [11]. In agreement, Desai et al. [28] reported that individual cleavage anomalies were not associated with aneuploidy; however, embryos with direct uneven cleavage or irregular chaotic division had reduced blastocyst initiation and expansion rates. Thus, cleavage anomalies are evident in both aneuploid and euploid embryos and are associated with developmental delays and reduced pre-implantation developmental potential.

Human embryos with cleavage anomalies can be morphologically similar to those that cleave normally and even meet the criteria for embryo freezing [29]. Reduced implantation and live birth rates, however, have been observed following transfer of blastocysts with atypical cleavage patterns in human patients [10, 11]. In the present study, the ability of embryos to hatch, attach to an extracellular matrix, and grow out during the extended post-hatching culture period was similar between blastocysts that had normal or atypical cleavage patterns. The extended culture period in this study was relatively short and does not fully recapitulate the uterine environment and interactions at the maternal–fetal interface, thus it is possible that the reduced developmental potential observed with abnormal human embryos was not captured within our system.

Moreover, it has been shown that different types of cleavage errors in human embryos have differing impact on developmental competencies [10]. Atypically cleaving embryos can successfully implant to establish a pregnancy as live births have occurred from transfer of human embryos with cleavage anomalies [30]. In human patients, however, deselecting those with atypical cleavage has led to improvements in live birth rates [3]. A limitation of the present study is that the small sample size hindered the ability to assess embryo outgrowth parameters relative to the type of cleavage error. Furthermore, the quality of TL images does not allow for retrospective morphological grading, although understanding the relationship between cleavage errors, blastocyst morphology, and embryo developmental competence should be a focus of future studies.

Conclusions

Time-lapse imaging of in vitro fertilized embryos allows for the visualization of developmental progression throughout the pre-implantation period. While time-lapse imaging analysis methods can be variable across laboratories [2], here we show that the proportion of abnormal embryos was consistent across four IVF seasons and different personnel. The incorporation of morphokinetic analysis enables the quantification of differences in timing of developmental events to assess the impact of experimental manipulation such as genome editing, gene knockdown, or experimental infection during this developmental time frame. Recent research has focused on developing nonhuman primate genetic models of human disease through genome editing of embryos [31, 32], yet the impact that CRISPR-Cas9 on- and off-target editing has on early embryo development remains uncertain. Time-lapse imaging could be used to deselect embryos for transfer that have had an atypical cleavage pattern, although no differences were observed in this study between blastocysts with normal or abnormal cleavage patterns in their ability to attach to an extracellular matrix and grow. The developmental competence of embryos with differing ploidy status or genome modification and/or those with different types of cleavage errors should be further investigated to better develop embryo selection criteria for improved transfer success.

Author contributions

JKS conceived and designed the study and drafted the manuscript. JKS, LNB, KMJ, HMH, KDM, BDB, and ATM analyzed the data. TGG reviewed the manuscript.

Conflict of interest

The authors have declared that no conflict of interest exists.

Data availability

Data will be made available upon request.