Individual assessment of bovine embryo development using a homemade chamber reveals kinetic patterns of success and failure to reach blastocyst stage

ABSTRACT

Most early developmental data are lost in bovine embryo culture systems. We developed and validated a method for culture of bovine embryos in groups that allow individual assessment. An autoclavable low-cost multiembryo chamber (MEC) was prepared using a polyester mesh fixed to a glass coverslip. Embryonic development was not affected by MEC. Compared to conventional bovine culture system (oil-covered drops, control), cleavage (C, 71.2 ± 7.8%; MEC, 74.3 ± 6.0%), blastocyst rate (C, 29.9 ± 4.4%; MEC, 28.3 ± 5.0%) and blastocyst cell number (C, 94.1 ± 9.7; MEC, 92.9 ± 5.3) were similar. Caspase 3 positive cell index in blastocysts was increased in MEC group, but apoptosis rate was below 5% (C, 2.9 ± 0.5; MEC, 4.6 ± 0.6). Using MEC, we performed a retrospective analysis for ‘failure’ and ‘success’ embryos, based on their ability to reach the blastocyst stage. We detected the majority of ‘success’ embryos displayed 8 cells at 48 h post-insemination (hpi) (48.7%), but blastocysts derived from this pattern presented lower cell numbers (91.3 ± 4.2 vs. 107.9 ± 4.9) and higher apoptosis index (6.2 ± 0.6 vs. 4.4 ± 0.5) than blastocysts from 4-cell embryos at 48 hpi. Most (72.0%) embryos that were at morula stage 120 hpi reached blastocyst stage at 168 hpi. Those blastocysts presented more number of cells than blastocysts derived from embryos exhibiting 16 cells at 120 hpi (108.6 ± 4.1 vs. 83.9 ± 4.8). Combination of embryo kinetics data at 48 and 120 hpi revealed high chances of blastocyst formation for patterns: 8 cells/morula, 4 cells/morula, 8 cells/16 cells and 4 cells/16 cells. Blastocysts formed from 4-cell/morula and 8-cell/morula patterns represented 69% of all 168 hpi blastocysts. Blastocysts derived from 4 cells/16 cells displayed decreased apoptosis (3.1 ± 0.6). Our results suggest that MEC can be used for bovine embryo culture without detrimental effects on development and can help to predict blastocyst formation and quality of in vitro fertilization (IVF) embryos.

Abbreviations: BSA: bovine serum albumine; COC: cumulus–oocyte complex; FERT-TALP: Tyrode’s albumin lactate pyruvate fertilization; FBS: fetal bovine serum; IVF: in vitro fertilization; MEC: multiembryo chamber; PBS: phosphate buffered saline; SOF-AA: synthetic oviductal fluid with amino acids medium; TCM: Tissue Culture Medium

KEYWORDS:

• Embryo kinetics
• single embryo culture
• embryonic quality
• in vitro culture system

Introduction

Human embryos are usually evaluated individually, and information regarding first stages of embryo development is recorded. Studies have shown that early cleavage data, including cleavage intervals, blastomeres symmetry, cytoplasmic/nuclear fragmentation and compaction, can help to identify embryos with increased chances of implantation (reviewed by Prados et al. 2012).

Continued analysis of individual embryos would contribute to trace the history of each embryo, helping to predict its developmental competence. Among interesting assessments that affect embryo competence are the time of the first cleavage (Garcia et al. 2015; Carrocera et al. 2016) and cytoplasmic fragmentation (Prados et al. 2012). Devices designed for time-lapse image analysis of human embryos such as Embryoslide and Primo vision embryo culture dish (Vitrolife, Goteborg, Sweden) permit group culture and individual assessment, but normally are set for few embryos, and their cost discourages the use for livestock.

Bovine embryos display long and asynchronous cell cycles, and a common feature of this species is the formation of heterogeneous population of embryos in culture (Oliveira et al. 2013). Usually, bovine embryos are cultured in drops of medium covered by mineral oil, in groups of 10–20. In this system, it is not possible to evaluate embryos individually: cleavage and blastocyst rates are recorded for a group of embryos. Therefore, information regarding cleavage progression is lost.

However, individual culture negatively affects the development of mammalian embryos (Vajta et al. 2000). This is mainly attributed to embryo ‘cross talk’, mediated by growth factors, and to the depletion of embryotoxic and inhibiting components of medium (Bavister 1995). When embryos are cultured individually in small drops, they lose ‘group effects’, and ammonia concentration can increase and be harmful to embryos (Gardner and Lane 1993). Studies had indeed shown a negative impact of the individual culture for bovine embryos (Basrur et al. 2004; Nagano et al. 2013).

Based on the information above, we developed and validated a method for culture of bovine embryos in groups, allowing individual assessment. Using a homemade chamber, we demonstrate that it is possible to use early development data (48 and 120 h post-insemination, hpi) to estimate blastocyst formation rates and quality. Here, we describe interesting patterns of bovine embryo kinetics.

Results

Experimental design

Experiments performed in this study are summarized in Table 1. First, we tested whether multiembryo chamber (MEC) has a negative effect on development in comparison to conventional culture (control group, C). MEC assembling is detailed in Figure 1. Embryos were assessed for cleavage and blastocyst rate and samples were collected for apoptosis and total cell number analysis (experiments 1 and 2, groups MEC and C). Then, we performed three additional replicates to test MEC in comparison to control and determine main cleavage patterns. Embryos were assessed for cleavage stage at 48 and 120 hpi to determine their 48–120 hpi pattern. At 168 hpi, embryos were assessed for blastocyst formation and classified as ‘success’ if they formed blastocysts, or ‘failure’ if they did not. Blastocysts were collected for apoptosis and total cell number analysis (experiments 3–5, groups MEC and C). Next, we performed an additional experiment to analyze quality of embryos derived from the main cleavage patterns resulting in blastocyst found in this study. For this experiment, in order to collect an expressive number of embryos and confirm the presence of similar patterns in microdrop system, embryos were cultured in a conventional culture system in drops with 10–12 embryos until 168 hpi. At 48 hpi, embryos were separated in groups: 4 cells and 8 cells. At 120 hpi, embryos were separated in groups: 4 cells/16 cells, 4 cells/morula, 8 cells/16 cells and 8 cells/morula. Blastocysts from each pattern were collected for apoptosis and total cell number analysis (experiments 6–9, groups 4c–16c, 4c–M, 8c–16c and 8c–M).

Table 1. List of experiments performed in this study.

Experiment	Title	Oocytes (n)	Cleaved (n)	Blastocysts (n, 168 hpi)	Culture method	Embryos with 48–120 hpi registered pattern
Experiment 1	Comparison of MEC with conventional culture system (rates and quality)	27 (MEC), 14 (CC)	19 (MEC), 12 (CC)	5 (MEC), 5 (C)	MEC and conventional culture (C)	0
Experiment 2		50 (MEC), 28 (CC)	37 (MEC), 18 (CC)	5 (MEC), 3 (C)		0
Experiment 3	Comparison of MEC with conventional culture system (rates and quality) and analysis of 48–120  hpi embryo patterns (rates)	61 (MEC), 46 (CC)	55 (MEC), 39 (CC)	14(MEC), 13 (C)	MEC and conventional culture (C)	35 (9s, 26f)
Experiment 4		50 (MEC), 39 (CC)	41 (MEC), 30 (CC)	18 MEC), 7 (C)		37 (15s, 22f)
Experiment 5		135 (MEC), 59 (CC)	74 (MEC), 26 (CC)	24(MEC), 9 (C)		43 (15s, 28f)
Experiment 6	Comparison of 48–120 hpi main embryo patterns (rates and quality)	147	107	64	Conventional culture (groups of each pattern were established 120 hpi)	23 4c–16c (13s, 10f); 14 4c–M (12s, 2f); 12 8c–16c (5s, 7f); 42 8c–M(34s, 8f)
Experiment 7		153	136	52		13 4c–16c (4s, 9f); 20 4c–M (14s, 6f); 17 8c–16c (4s, f13); 47 8c–M(30s, 17f)
Experiment 8		155	121	57		15 4c–16c (6s, 9f); 27 4c–M (25s, 2f); 11 8c–16c (6s, 5f); 24 8c–M (20s, 4f)
Experiment 9		73	52	20		5 4c–16c (4s, 1f); 10 4c–M (6s, 4f); 10 8c–16c (7s, 3f); 8 8c–M (3s, 5f)

Figure 1. Preparation of multiembryo chamber (MEC). (A) The image shows pieces of polyester mesh (290 µm aperture) cut in different sizes, according to the desired amount of embryos and the round glass cover (0.13–0.16 mm/13 mm thickness). (B) Polyester mesh is attached to glass coverslip using acetic curing silicon sealant. (C) MEC is autoclaved. (D) MEC is set into four-well plates. (E) Microscopic image of MEC. (F) Embryos set into MEC for embryo culture.

Production rates and embryo quality assessment

Cleavage rates at 48 hpi (C, 71.2 ± 7.8%; MEC, 74.3 ± 6.0%) (Figure 2A) and blastocyst rates at 168 hpi (C, 29.9 ± 4.4%; MEC, 28.3 ± 5.0%) (Figure 2B) were similar between groups, suggesting MEC conditions did not affect embryonic development. Quality analysis revealed that blastocyst cell numbers were similar between groups (C, 94.1 ± 9.7; MEC, 92.9 ± 5.3) (Figure 3A), but caspase 3-positive cell index was increased (p < 0.05) in MEC group (C, 2.9 ± 0.5; MEC, 4.6 ± 0.6) (Figure 3B). Less than 5% caspase 3-positive cells were found.

Figure 2. Production rates for multiembryo chamber (MEC). Graphs show mean cleavage rate (A) and mean blastocyst rate (B) of control and MEC, obtained from five replicates (n = 509 oocytes, 186–323 per group). Bars indicate standard error.

Figure 3. Embryo quality assessment of blastocysts produced in multiembryo chamber (MEC). Images show caspase 3 (555) and Hoechst staining of control and MEC blastocysts (A). Bar indicates 20 µm. Graphs show mean cell number (B) and mean apoptosis index (C) of control and MEC embryos (n = 69 blastocysts, 24–45 per group). Bars indicate standard error. Asterisks indicate statistical difference.

Retrospective analysis of ‘success’ and ‘failure’ embryos

Mean cell number

We individually monitored the embryos cultured in MEC during early development and performed a retrospective analysis of embryos that achieved or not the blastocyst stage at 168 hpi (success, S; failure, F).

First, we searched for differences between F and S groups related to cell numbers that could help to identify S embryos at early cleavage stages (Figure 4) (S: n = 39; F: n = 76). The mean cell number was similar in both groups at 48 hpi (S: 5.4 ± 0.4 and F: 4.3 ± 0.3), but at 120 hpi, cell number was increased (p < 0.05) in S embryos (S: 24.7 ± 2.2 and F: 10.1 ± 1.2).

Figure 4. Approximate cell number analysis of ‘success’ (embryos that formed blastocysts) and ‘failure’ (embryos that did not form blastocysts) embryos at 48 and 120 hpi. Graph shows mean cell number and standard error for each group at 48 and 120 hpi evaluation. Data from 115 cleaved embryos, 39–76 per group. Asterisks indicate statistical difference.

Embryo kinetics and blastocyst formation

Then, we calculated the rates of blastocyst formation for the main embryonic stages found at 48 and 120 hpi (Figure 5). This analysis reflected the probability of success or failure for each embryo stage.

Figure 5. Chance of success of bovine embryos evaluated at 48 and 120 hpi. (A) Graph shows percentage of embryos that form blastocysts for different stages seen at 48 hpi (data from 29 2-cell, 42 4-cell and 39 8-cell embryos. Five embryos were still zygotes at 48 hpi). (B) Graphical summary of 48 hpi analysis, indicating embryos at two-cell stage have a high chance of failure and embryos at 4- and 8-cell stage have a high chance of success at this time point. (C) Graph shows percentage of embryos that form blastocysts for different stages seen at 120 hpi (data from 36 8 cells, 26 16 cells and 25 morula. Twenty-eight embryos had less than 8 cells or were degenerated and were not considered for this analysis). (D) Graphical summary of 120 hpi analysis, indicating embryos at 8-cell stage have a high chance of failure and embryos at 16-cell and morula stage have a high chance of success at this timepoint. Asterisks indicate statistical difference.

In 48 hpi evaluation, 8-cell embryos had an increased (p < 0.05) chance of becoming blastocysts (48.7%) compared to 2-cell embryos (20.7%) (Figure 5A,B). In 120 hpi evaluation, we detected that morulae (72%) and 16 cells (53.8%) had an increased (p < 0.05) chance of becoming blastocysts compared to 8-cell embryos (19.4%) (Figure 5C,D).

After we assessed embryo pattern at 48 and 120 hpi to investigate if a specific pattern would be more successful than others would, the highest percentage of successful embryos was 8c–M (8 cells at 48 hpi and morula at 120 hpi) (Figure 6A). Patterns 4c–M, 4c–16c and 8c–16c were also common in S group. In 8c–M, 4c–M and 4c–16c patterns, success was more frequent (p < 0.05) than failure. Patterns 4c–8c, 4c–4c and 2c–2c rarely reached blastocyst stage – failure was more frequent than success (p < 0.05). The two main patterns of success and failure are shown in Figure 6B.

Figure 6. Cleavage pattern of ‘success’ and ‘failure’ embryos. (A) Graph shows percentage of main cleavage patterns in failure and success group (n = 14 8c–M, 11 8c–16c, 14 8c–8c, 6 4c–M, 7 4c–16c, 16 4c–8c, 13 4c–4c, 13 2c–2c. 21 embryos from other patterns are not presented in the graph due to their low frequency). Asterisks indicate statistical difference comparing F and S groups for each pattern. (B) Graphical summary indicating critical patterns of cleavage combining 48 and 120 hpi data for failure (red) and success (green) embryos.

Embryos that displayed 8 cells at 48 and 120 hpi were not predominant in failure group (13.2% F embryos and 10.3% S embryos were 8c–8c). However, considering 8c–8c and 4c–8c together, this corresponded to the highest percentage of arrested embryos (32.9%).

In Figure 7, results are presented as the percentage of success for each pattern commonly observed. The highest rate of success was found in 4c–M (100%) followed by 8c–M (71.4%), 4c–16c (71.4%) and 8c–16c (45.5%). Only 28.6% 8c–8c embryos formed blastocysts, followed by 6.3% 4c–8c embryos

Figure 7. Probability of success of main cleavage patterns combining 48 and 120 hpi evaluations. Graph shows percentage of blastocyst formation in main cleavage patterns (n = 6 4c–M, 14 8c–M, 7 4c–16c, 11 8c–16c, 14 8c–8c, 16 4c–8c).

In experiments 5–9, 4c–M, 8c–M, 4c–16c and 8c–16c patterns were followed outside the MEC, using conventional culture system. In Figure 8A, we present the percentage of each pattern among cleaved embryos. The highest frequency is 8c–M (29.0% of cleaved) and the lowest is 8c–16c (12.0%). In Figure 8B, we show the percentage of blastocyst formation in each pattern. We found higher (p < 0.05) percentages of blastocysts in 4c–M (80.3%) and 8c–M (71.9%), compared to 4c–16c (48.2%) and 8c–16c (44.0%). In Figure 8C, we present the percentage of each pattern among day 7 (168 hpi) blastocysts. The highest (p < 0.05) frequency was 8c–M (45.1%), followed by 4c–M (29.5%). Patterns 4c–16c (14.0%) and 8c–16c (11.4%) had lower (p < 0.05) frequencies.

Figure 8. Performance of main success patterns (4c–16c, 4c–M, 8c–16c and 8c–M). (A) Graph shows the percentage of each pattern among cleaved embryos. (B) Graph shows percentage of blastocyst formation for each pattern. (C) Graph shows percentage of each pattern among day 7 blastocysts. (A, B, C) Different letters indicate statistical difference. N = 56 4c–16c, 71 4c–M, 50 8c–16c, 121 8c–M; 118 embryos (28%) displayed other patterns and are not presented in the graph.

In Table 2, we presented regression analysis for potential predictor variables for blastocyst conversion rate of cleaved embryos. The number of blastomeres at 48 hpi (2, 4 or 8 cells) was a significant (p < 0.05) predictor for blastocyst formation. Rates of 4 cells at 48 hpi, 8 cells at 48 hpi, 16 cells at 120 hpi and morula at 120 hpi were not significant predictor for blastocyst conversion among replicates.

Table 2. Linear regression analysis among potential predictor variables for blastocyst conversion rate of cleaved embryos in bovine.

Potential predictor variables	R^2	Slope	P
Number of blastomeres at 48 hpi*	0.4377	5.2690	0.0072
4-cell rate at 48 hpi	0.1518	0.6604	0.5167
8-cell rate at 48 hpi	0.0961	0.2868	0.6116
16-cell rate at 120 hpi	0.4614	1.4220	0.2072
Morula rate at 120 hpi	0.4970	0.6037	0.1836

*The number of blastomeres at 48 hpi was categorized as 2 cells, 4 cells and 8 cells.

Embryo kinetics and blastocyst quality

Assessment of embryo quality revealed that blastocysts derived from 8c–16c pattern had lower (p < 0.05) number of cells than 4c–M and 8c–M patterns (Figure 9A). Apoptosis was decreased (p < 0.05) in blastocysts from 4c–16c pattern compared to 8c–M pattern (Figure 9B).

Figure 9. Influence of cleavage pattern on embryo quality. Images show caspase 3 (555) and Hoechst staining of blastocysts derived from 4c–16c, 4c–mo, 8c–16c and 8c–mo patterns (A). Bar indicate 20 µm. (B, C) Graphs show blastocyst cell number (B) and apoptosis index (C) of the main success patterns (n = 116 blastocysts, 21–38 per group). Different letters indicate statistical difference. (D, E) Graphs show blastocyst cell number (D) and apoptosis index (E) based on 48 hpi pattern (4c or 8c, n = 116 blastocysts, 57–59 per group). (F, G) Graphs show blastocyst cell number (F) and apoptosis index (G) based on 120 hpi pattern (16c or M, n = 116 blastocysts, 43–73 per group). (D, E, F, G) Asterisks indicate statistical difference.

Speed of development at 48 hpi also affected blastocyst quality. We found higher (p < 0.05) cell numbers and increased (p < 0.05) apoptosis in embryos with 8 cells at 48 hpi (Figure 9C,D). Considering embryonic stage at 120 hpi, we found increased (p < 0.05) cell numbers in blastocysts derived from morulae than those derived from 16-cell embryos, and no difference in apoptosis was observed between groups (Figure 9E,F).

Discussion

Embryo culture using MEC

In this work, we describe a simple adaptation to culture bovine embryos in groups in 4-well plates that allows to assess development of single embryos: a homemade chamber (MEC), which was tested and compared to conventional culture of bovine embryos (microdrops).

MEC is prepared using a polyester mesh, routinely used for imaging purposes in mouse embryos. Polyester mesh was previously used for bovine embryos (Matoba et al. 2010; Somfai et al. 2010), but our method allows sterilization and package of the chamber. MEC can be a powerful tool for scientific purposes, since information can be collected systematically for each embryo. This can be helpful to address experimental issues, especially for early embryo development, associating patterns with outcomes. This chamber can be used without oil, which may be an advantage for some experiments testing molecules diluted in oil-affinity substrates. For commercial purposes, the use of MEC can be beneficial for low-recovery donors, since zygotes can be grouped with others to increase embryo density. Also, MEC can be used to improve embryo development, culturing zygotes from high conversion donors as ‘helper embryos’ (Deb et al. 2011) with other zygotes, and to predict rates and embryo quality based on speed of development (reviewed by Gutierrez-Adan et al. 2015).

Our results show that embryo development (cleavage rate, blastocyst rate and blastocyst cell number) is not affected by MEC. However, an increased apoptosis index was detected in blastocysts developed in the chamber, measured by caspase 3, an apoptosis downstream effector enzyme. Apoptosis is a physiological event of embryo development and prevents damaged cells from contributing to the formation of an individual, thus functioning as a survival mechanism (Paula-Lopes and Hansen 2002). MEC resulted in less than 5% apoptosis rate, which is below usual rates found for bovine blastocysts (Wang et al. 2013; Sudano et al. 2014). Therefore, even though apoptosis rate was higher in MEC, the low rates combined to normal cell numbers suggest no detrimental effect on embryo development.

In MEC, distance between embryos ranges from 140 to 778 µm. This is usually more than the distance in conventional system, where most embryos are close to each other. In commercial micro-well group culture dishes, designed for individual culture as Primo vision 9-well dishes, distance between wells is 100–130 µm, and the size of a micro-well is 350 µm in diameter (Lehner et al. 2017). Therefore, depending on embryo position, distance between embryos can range from 100 to 590 µm (considering a 120 µm embryo), slightly less than MEC. Close proximity between embryos is important to allow paracrine signaling and embryo-to-embryo communication (Wydooghe et al. 2017), but MEC conditions were not detrimental for development.

In contrast, Gopichandran and Leese (2006) found, using an attachment adhesive, that bovine embryos had increased development when kept at a 165-µm distance from each other; higher or lower distances decreased blastocyst rates, and no blastocyst was found when 540 µm distance was used. There are two crucial differences between Gopichandran and Leese study and ours. The first is fetal bovine serum (FBS) supplementation, which could not be used by the authors because of their aim, and is supplemented in our study. Serum provides a wide array of nutrients, including growth factors and cytokines, and can possibly attenuate lower embryo-to-embryo communication. The second point is when distance between embryos is increased, embryo density is decreased. Lehner et al. (2017) compared embryo development in high density of embryos (7–9 embryos/25 µL media), moderate density (5–6 embryos/25 µL media) and low density (2–4 embryos/25 µL media). Their results suggest that a moderate density has beneficial effects for human embryos (5–6 embryos/25 µL, 5–4.2 µL media for each embryo). In our study, we calculated the media volume as a proportion (5–6 µL per cultured embryo), and the same proportion was used in microdrops.

Individual embryo assessment

Using MEC, we evaluated embryos individually at 48 and 120 hpi and assessed the effect of speed of development and cleavage patterns on blastocyst formation. This evaluation is simple and can be performed in combination with media exchange in most bovine in vitro fertilization (IVF) laboratories, especially because no time-lapse equipment is required.

At 48 hpi, 2-cell embryos had the lowest probability of blastocyst formation, and the number of blastomeres was a significant predictor for blastocyst conversion. Several studies show that the timing of the first cleavage can affect bovine embryo development, and embryos with 2 cells in the first evaluation (40–48 hpi) have a lower chance of forming blastocysts (Garcia et al. 2015; Carrocera et al. 2016). Chromosomal defects could be delaying the first cleavage (Lonergan et al. 1999; Sugimura et al. 2012) and causing this 2-cell block. In addition, slow embryos activate different pathways for embryo survival (Gutiérrez-Adán et al. 2004), and their metabolites related to cellular signaling, energy and lipid metabolism and antioxidants differ from fast embryos (Silva et al. 2016), indeed suggesting a distinct phenotype. The embryos that reached blastocyst stage at 168 hpi were 2c–8c (1), 2c–16c (3) and 2c–M (2).

Although no statistical difference was detected for blastocyst formation rate between 4 and 8 cells (48 hpi evaluation), we found a difference in cell number and apoptosis later. Blastocysts derived from 8 cells at 48 hpi presented decreased cell number and increased apoptosis, suggesting being a fast embryo at 48 hpi is detrimental for embryo quality, corroborating with recent findings. Although the time of the first cleavage is decisive in the rate of blastocyst formation in bovine, studies demonstrated that intermediate speed developing embryos might form better quality blastocysts (reviewed by Gutierrez-Adan et al. 2015). Fast mouse embryos present abnormal expression of imprinted genes (Market Velker et al. 2012), and Milazzotto et al. (2016) showed that slow dividing blastocysts present a gene expression pattern closer to in vivo embryos than fast blastocysts do. In addition, no difference in hatching ability was found between slow and fast embryos (Garcia et al. 2015), suggesting that although they form blastocysts in lower abundance, those blastocysts are healthy. It would be very interesting to study pregnancy rates of blastocysts derived from 4- or 8-cell embryos at 48 hpi, aiming to address if this cell number reduction and apoptosis increase would affect pregnancy establishment and embryonic losses.

The speed of development can also be affected by other factors, such as gamete influence. We observed in the previous study that bull can alter developmental kinetics, suggesting a paternal effect on early embryo development even before major embryonic genome activation (Oliveira et al. 2016). Xue et al. (2013) demonstrated stage-specific monoallelic expression patterns for a significant proportion of polymorphic gene transcripts, suggesting maternal and paternal genetics influence early embryo development in many pathways including cell cycle. Indeed, in bovine, the minor wave of embryonic transcriptional activation during 2-cell stage is involved in essential signaling and metabolic pathway and many transcription factors take part in the initial embryo cleavage (Zuo et al. 2016).

Regarding 120 hpi evaluation, we detected a marked difference in cell number of embryos that do or do not form blastocysts. We also detected more cell numbers for blastocysts derived from morulae than from 16 cells, suggesting cell number at 120 hpi is indeed a great indicator of success in blastocyst formation and increased blastocyst cell number. The main embryonic stages that resulted in blastocysts were 8 cells, 16 cells and morula.

At this time point, 8-cell embryos had only 19.4% chance of forming a blastocyst. The 8-cell block is a crucial phenomenon in bovine embryology, characterized by the arrest of embryos that cannot achieve the major activation of the embryonic genome and follow development (Meirelles et al. 2004). This is mainly attributed to a deficit in maternal stock that should hinder embryonic survival until this stage (Lonergan et al. 2003) and to difficulties to bypass repressive chromatin marks of the embryonic genome (Betts and King 2001). In our conditions, most 8-cell blocked embryos appeared at 120 hpi, corresponding to embryos that were 4 cells at 48 hpi and 8 cells at 120 hpi. Thus, the presence of an 8-cell embryo at 48 hpi and 8 cells again at 120 hpi was not an indicative of blockade in our conditions, since 28.6% of this phenotype formed blastocysts. However, it is important to note that blastocysts derived from 8c–16c pattern, the closest to 8c–8c assessed in this study, presented an inferior amount of cells compared to other successful patterns and should probably be avoided for embryo transfer purposes.

It is possible that these success 8c–8c embryos were close to 16-cell cleavage at 120 hpi evaluation and formed blastocysts with small numbers of cells, as cavitation would happen only 48 h after the last evaluation (120 hpi). These small blastocysts are commonly observed in female embryos, especially those suffering from marked cell death (Oliveira et al. 2016).

Finally, we associated the cleavage patterns in 48 and 120 hpi in order to identify successful profiles. The patterns 8c–M, 4c–M, 8c–16 and 4c–16 were the most frequent in success group, and their blastocyst rates remained similar in MEC and conventional culture experiments. Our data suggest blastocysts derived from 8c–M and 4c–M are the most abundant in bovine embryology, corresponding to 97 + 63 out of 232, together 69% of all blastocysts. This is important to keep in mind since other possibly favorable patterns would represent only a few embryos. Both patterns presented increased number of cells, which was evident when we combined them and analyzed against 4c–16c and 8c–16c. Pattern 4c–16c resulted in higher total cell number and decreased apoptosis, suggesting a favorable pattern, but represented only 32 out of 232 blastocysts, 14% of all blastocysts. We identified an interesting predominant profile among the failure groups, 4c–8c, which resulted in arrest in 15 out of 16 occurrences.

We conclude that the developed MEC is suitable for the culture of bovine embryos, and individual embryo assessment at media exchange time points can be used as a noninvasively predictor for IVF. Our results suggest that the presence of 4 and 8 cells at 48 hpi and 16 cells and morula at 120 hpi is indicative of high probability of blastocyst formation, and blastocysts derived from embryos with 4 cells at 48 hpi have increased quality.

Materials and methods

All experimental procedures followed ethical guidelines for animal experimentation and were approved by local committee (CEUA EGL 23/2015 protocol). Reagents were purchased from Sigma unless otherwise specified.

Embryo production (IVM and IVF)

Embryos were produced using oocytes from slaughterhouse ovaries, transported to the laboratory in a thermal bottle containing 0.9% NaCl physiological solution (35–36°C) and processed up to 4 h after slaughter. The follicles were manually punctured with a 20-mL syringe and 18 G needle and follicular fluid was transferred for conical tubes. Cumulus–oocyte complexes (COCs) were selected according to morphological classification (compact nonatretic cumulus with more than two layers of cumulus cells and homogeneous cytoplasm). In vitro maturation was performed in 100 µL drops of Tissue Culture Medium (TCM) 199 medium supplemented with FBS, 1.0 µg/mL follicle stimulating hormone (Folltropin™, Bioniche Animal Health, Belleville, Canada), 50 µg/mL hCG (Profasi™, Serono, Sao Paulo, Brazil), 1.0 µg/mL estradiol, 16 µg/mL sodium pyruvate and 83.4 µg/mL amikacin and covered with mineral oil for 24 h, in an incubator (38.5°C, 5% CO₂ in atmospheric air and high humidity). IVF was performed after maturation. Frozen straw of bovine semen was thawed at 36°C for 30 s. Semen was transferred to a 45/90 Percoll gradient and centrifuged at 3600× g for 7 min. After supernatant removal, the pellet was suspended with 1 mL of Tyrode's albumin lactate pyruvate fertilization (FERT-TALP) medium supplemented with 0.6% bovine serum albumine (BSA), 10 µg/mL heparin, 18 µM penicillamine, 10 µM hypotaurine and 1.8 µM epinephrine and centrifuged at 520× g for 5 min. Coculture of oocytes with spermatozoa was conducted in 50 µL drops of FERT-TALP medium for 18 h (38.5°C, 5% CO₂ in atmospheric air and high humidity). In vitro maturation and fertilization were performed in 60 mm tissue culture dishes (Corning™, New York, US). Presumptive zygotes were used for the experiments (n = 1037).

Preparation of MEC

Mesh squares were cut according to the number of embryos (3 × 3 mm for approximately 10 embryos, 8 × 8 mm for 50 or more embryos) (Figure 1A). In this experiment, 8 × 8 mm meshes were used. We used a polyester mesh with 290 µm aperture produced by PlastOk® (Birkenhead, UK). Using a fine-tipped forceps to hold the mesh, a thin layer of acetic curing silicon sealant was applied with a needle or other fine-point instrument the borders. The excess of silicon was removed. For 8 × 8 meshes, a drop of silicon was applied in the middle of the mesh. The mesh was attached to a round glass cover, model 13 mm, thickness 0.13–0.16 mm (Figure 1B). After dried, the chambers were washed in phosphate buffered saline (PBS) for residue removal. PBS was exchanged after 12 h. On the next day, the PBS was completely removed, and the chambers were dried in a laminar flow hood. They were then autoclaved for 15 min at 120°C (Figure 1C).

Embryo culture

Presumptive zygotes were submitted to vigorous pipetting in hyaluronidase solution for cumulus cells removal and randomly distributed between groups. A 4-well 10 mm deep, 15 mm diameter polystyrene plate (Nunc™-Thermo Fisher Scientific, Waltham, US) was used. Synthetic oviductal fluid with amino acids medium (SOF-AA) supplemented with 1.5% FBS and 6 mg/mL BSA was used for embryo culture. Control (C) embryos were cultured in conventional culture system, in wells containing 10–12 presumptive zygotes per 60 µL drop of medium under mineral oil. MEC was filled up with medium (Figure 1D,E) and air bubbles were removed with needles. After that 5 µL medium was added per cultured embryo. If medium volume was greater than 150 µL, it is not necessary to cover the wells of the plate with mineral oil, but for smaller volumes, oil coverage is recommended. After assembly, the presumptive zygotes were placed in each compartment of the chamber, created by the overlap of the polyester wires (Figure 1F).

Control and MEC embryos were kept in the same 4-well plates, in separate wells. Distilled water (2 mL) was used to fill the interior of the 4-well dish to reduce medium evaporation.

The plates were not handled with sudden movements to prevent embryo rearrangements. Medium was half replaced at 48 and 120 hpi, by slowly withdrawn and addition.

Experiments 6–9 were performed in conventional system, and embryos were rearranged according to their classification at 48 hpi (4 cells and 8 cells) and rearranged again at 120 hpi (4c–16c, 4c–M, 8c–16c, 8c–M) so that same pattern was cultured in each drop.

Embryos were kept for 7 days (168 h) at 38.5°C, 5% CO₂ in atmospheric air and maximum humidity.

Evaluation of cell number and caspase 3

Blastocysts (168 hpi) were fixed in 4% paraformaldehyde for 30 min at 37°C and stored at 4°C in PBS. Immunofluorescence for caspase 3 and nuclei staining with Hoechst was performed as described in Oliveira et al. (2016). Briefly, embryos were permeabilized in 0.5% Triton X-100 solution for 30 min at room temperature and washed three times in PBS supplemented with 0.2% Tween-20 (PBS-T) for 10 min. Blocking was performed for 30–60 min in 3% BSA at room temperature. Incubation with primary antibody (anti-caspase 3, rabbit, 1: 750) was done at 4°C, for 12 h. The embryos were then washed three times in PBS-T and incubated with secondary antibody (Donkey anti-rabbit IgG conjugated to Alexa Fluor 555, 1: 400) for 30–120 min at room temperature. Nuclei were stained with Hoechst 33,342 for 20 min and embryos were examined under fluorescence microscopy (Alexa Fluor 555: wavelengths Ex 556–Em 573 nm; Hoechst 33,342: wavelengths Ex 352–Em 461 nm). Images were taken for each embryo (n = 185) and cells were counted using ImageJ (NIH, USA) software for total cell number assessment.

Embryonic development assessments

Cleavage rate was calculated as the percentage of presumptive zygotes that cleaved at 48 hpi, and blastocyst rate was calculated as the percentage of cleaved embryos that develop in blastocysts at 168 hpi.

For individual analysis in MEC, each embryo was monitored and evaluated according to the number of cells in three time points: 48, 120 and 168 hpi. For embryo quality assessment of main cleavage patterns (experiments 6–9), embryos were classified and cultured separately as 4 cells or 8 cells (groups of 10–12 embryos) 48 hpi and classified and separated in 16 cells or morulae 120 hpi, when four groups were formed (4c–16c, 4c–M, 8c–16c, 8c–M). Since after 8-cell stage, it was not possible to count cell number of live bovine embryos accurately in our system, for cell number analysis, embryos with more than 8 cells (9–16) were considered with 14.3 cells, and morulae were considered with 39.2 cells. This numbers were previously established for our conditions based on Hoechst staining of fixed embryos and confocal imaging (data not shown).

Embryos individually assessed were classified at 168 hpi as ‘success’, if they reached blastocyst stage, or ‘failure’ if they do not. These groups were used for retrospective analysis (prediction of blastocyst formation).

Statistical analysis

Cleavage and blastocyst rates were compared between groups using Fisher’s exact test. The t-test was used to compare the blastocysts mean cell number and Mann–Whitney test to compare apoptosis index between control and MEC groups, 4c versus 8c 48 hpi and 8c versus M 120 hpi. The comparison of blastocyst quality among 4c–16c, 4c–M, 8c–16c and 8c–M was performed using ANOVA and Tukey for mean number of cells and Kruskal–Wallis and Dunn for apoptosis. Analysis of ‘success’ and ‘failure’ groups regarding approximate cell number was performed using Mann–Whitney test. Percentage of success or failure for each embryonic stage, and cleavage pattern percentages, frequency of embryo patterns was compared using Fisher’s exact test. Linear regression was used to estimate effects of potential predictor variables for blastocyst conversion. A 5% significance level was considered for all analysis.

Authors’ contributions

Conceived and designed the study: CSO, RVS, LSAC; Performed the experiments: BAFB, CASM, PMSR, GRL; Analyzed the data: BAFB, CASM, CSO; Wrote the manuscript: CSO, LSAC.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by the Empresa Brasileira de Pesquisa Agropecuária: [Grant Number 01.13.06.001.05.01.003]; Fundação Carlos Chagas Filho de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ): [Grant Number 290.007/2016] and Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq): [Grant Number 118798/2015-4].