Different pre-implantation phenotypes of bovine blastocysts produced in vitro

in Reproduction
Correspondence should be addressed to I Hue; Email: isabelle.hue@inra.fr

Embryo transfer in cattle is performed with blastocysts produced in vivo or in vitro using defined media. However, outdated systems such as those that use serum and co-culture remain of interest for research purposes. Here, we investigated the effect of additional culture time on in vitro-produced embryos. Specifically, we compared embryos that formed a blastocoel at different times after fertilisation to those that stayed in culture for up to two additional days with respect to their development in vivo after temporary transfer to oestrus-synchronised recipients. A pre-transfer set (D6, D6+1, D6+2, D7, D7+1, D8) was examined using microarray analyses and correlated with a post-transfer set that included two different days of transfer (D6-T6, D6+2-T8, D7+1-T8, D8-T8). All surviving conceptuses reached primitive-streak stages and filamentous sizes similarly to in vivo (D18) or in vitro controls (D7/T7). The recovery rate differed between D6 and D8 embryos that were immediately transferred (58 vs 25%). With an intermediate survival rate (33%), the D6 embryos with two additional days in culture produced nine times more IFN-tau (IFNT) at D18 than the D6 embryos that were immediately transferred. At the end of culture, D6 and D6+2 embryos displayed the highest number of gene expression differences. Despite a mortality of 40–60%, no signature was detectable in any of the transferred groups that would account for the embryos’ fates. Initially reputed to be beneficial in producing more blastocysts, our culture system of B2 medium plus serum and co-culture generated blastocysts that were distinct from those developed in vivo (D7).

Downloadable materials

  • Supplementary Figure 1: Principal-components analysis (PCA) of all samples (n=21: 7 groups, 3 replicates per group). In the three panels are the projections based on gene expression data from the microarray analysis on 3 main axes (1-2, 1-3, and 2-3). In red, green and pink: the D6, D7, and D8 blastocysts. In yellow, blue, and light green: the D6+1, D7+1 and D6+2 embryo groups.
  • Supplementary Table 1: Oligonucleotides used for bovine gene quantification by real time RT-qPCR. Seven pairs were designed de novo. Eight originated from previous reports: 1 (Sakurai et al. 2013), 2 (Khan et al. 2012), 3 (Bertolini et al. 2002), 4 (Nuttinck et al. 2008). (*) refers to primer sequences being adapted to the bovine NM accession numbers; (N) indicates the nucleotide that has been changed as compared to the original report.
  • Supplementary Tables 2 : Lists of enriched genes and their associated functions, in every paired comparison of the design (n=9). Provided are: i) the size of the gene set indexed in the hallmark or homemade database for each function, ii) the total number of genes enriched in each comparison (|NES|>1.8; FDR<0.05; NES: Normalised Enrichment Score, as provided by the software from the Broad Institute), iii) the number of genes similarly enriched in the three replicates (->), iv) the number of unique gene IDs per comparison, after we merged all the functions, and v) the gene IDs shared across comparisons, for each of the enriched functions. In bold, the enriched functions that were shared across all groups. In Supp. Table 2: the D6/D7/D8-forming blastocysts; Supp. Table 3a and 3b: the blastocysts that spent extra days in culture (+1/+2); Supp. Table 4: blastocysts which, overall, spent the same time in culture (7 days); Supp. Table 5a and 5b: blastocysts that were transferred in vivo at D8, plus D8 vs D6 (no NES at FDR <0.05 for the D7+1 vs D8 comparison, which is thus not shown); Supp. Table 6: the in vivo vs in vitro-produced embryos assessed with respect to metabolic sensors (NES=-1.4; FDR<0.05). Corresponding gene lists in the seven Supp. Tables from 2-bis to 6-bis are (.xls) files accessible at https://data.inra.fr/privateurl.xhtml?token=00c06d2c-f4c6-446d-bfe5-8af5e7759038.
  • Supplementary Tables 3: Lists of enriched genes and their associated functions, in every paired comparison of the design (n=9). Provided are: i) the size of the gene set indexed in the hallmark or homemade database for each function, ii) the total number of genes enriched in each comparison (|NES|>1.8; FDR<0.05; NES: Normalised Enrichment Score, as provided by the software from the Broad Institute), iii) the number of genes similarly enriched in the three replicates (->), iv) the number of unique gene IDs per comparison, after we merged all the functions, and v) the gene IDs shared across comparisons, for each of the enriched functions. In bold, the enriched functions that were shared across all groups. In Supp. Table 2: the D6/D7/D8-forming blastocysts; Supp. Table 3a and 3b: the blastocysts that spent extra days in culture (+1/+2); Supp. Table 4: blastocysts which, overall, spent the same time in culture (7 days); Supp. Table 5a and 5b: blastocysts that were transferred in vivo at D8, plus D8 vs D6 (no NES at FDR <0.05 for the D7+1 vs D8 comparison, which is thus not shown); Supp. Table 6: the in vivo vs in vitro-produced embryos assessed with respect to metabolic sensors (NES=-1.4; FDR<0.05). Corresponding gene lists in the seven Supp. Tables from 2-bis to 6-bis are (.xls) files accessible at https://data.inra.fr/privateurl.xhtml?token=00c06d2c-f4c6-446d-bfe5-8af5e7759038.
  • Supplementary Tables 4: Lists of enriched genes and their associated functions, in every paired comparison of the design (n=9). Provided are: i) the size of the gene set indexed in the hallmark or homemade database for each function, ii) the total number of genes enriched in each comparison (|NES|>1.8; FDR<0.05; NES: Normalised Enrichment Score, as provided by the software from the Broad Institute), iii) the number of genes similarly enriched in the three replicates (->), iv) the number of unique gene IDs per comparison, after we merged all the functions, and v) the gene IDs shared across comparisons, for each of the enriched functions. In bold, the enriched functions that were shared across all groups. In Supp. Table 2: the D6/D7/D8-forming blastocysts; Supp. Table 3a and 3b: the blastocysts that spent extra days in culture (+1/+2); Supp. Table 4: blastocysts which, overall, spent the same time in culture (7 days); Supp. Table 5a and 5b: blastocysts that were transferred in vivo at D8, plus D8 vs D6 (no NES at FDR <0.05 for the D7+1 vs D8 comparison, which is thus not shown); Supp. Table 6: the in vivo vs in vitro-produced embryos assessed with respect to metabolic sensors (NES=-1.4; FDR<0.05). Corresponding gene lists in the seven Supp. Tables from 2-bis to 6-bis are (.xls) files accessible at https://data.inra.fr/privateurl.xhtml?token=00c06d2c-f4c6-446d-bfe5-8af5e7759038.
  • Supplementary Tables 5: Lists of enriched genes and their associated functions, in every paired comparison of the design (n=9). Provided are: i) the size of the gene set indexed in the hallmark or homemade database for each function, ii) the total number of genes enriched in each comparison (|NES|>1.8; FDR<0.05; NES: Normalised Enrichment Score, as provided by the software from the Broad Institute), iii) the number of genes similarly enriched in the three replicates (->), iv) the number of unique gene IDs per comparison, after we merged all the functions, and v) the gene IDs shared across comparisons, for each of the enriched functions. In bold, the enriched functions that were shared across all groups. In Supp. Table 2: the D6/D7/D8-forming blastocysts; Supp. Table 3a and 3b: the blastocysts that spent extra days in culture (+1/+2); Supp. Table 4: blastocysts which, overall, spent the same time in culture (7 days); Supp. Table 5a and 5b: blastocysts that were transferred in vivo at D8, plus D8 vs D6 (no NES at FDR <0.05 for the D7+1 vs D8 comparison, which is thus not shown); Supp. Table 6: the in vivo vs in vitro-produced embryos assessed with respect to metabolic sensors (NES=-1.4; FDR<0.05). Corresponding gene lists in the seven Supp. Tables from 2-bis to 6-bis are (.xls) files accessible at https://data.inra.fr/privateurl.xhtml?token=00c06d2c-f4c6-446d-bfe5-8af5e7759038.
  • Supplementary Tables 6: Lists of enriched genes and their associated functions, in every paired comparison of the design (n=9). Provided are: i) the size of the gene set indexed in the hallmark or homemade database for each function, ii) the total number of genes enriched in each comparison (|NES|>1.8; FDR<0.05; NES: Normalised Enrichment Score, as provided by the software from the Broad Institute), iii) the number of genes similarly enriched in the three replicates (->), iv) the number of unique gene IDs per comparison, after we merged all the functions, and v) the gene IDs shared across comparisons, for each of the enriched functions. In bold, the enriched functions that were shared across all groups. In Supp. Table 2: the D6/D7/D8-forming blastocysts; Supp. Table 3a and 3b: the blastocysts that spent extra days in culture (+1/+2); Supp. Table 4: blastocysts which, overall, spent the same time in culture (7 days); Supp. Table 5a and 5b: blastocysts that were transferred in vivo at D8, plus D8 vs D6 (no NES at FDR <0.05 for the D7+1 vs D8 comparison, which is thus not shown); Supp. Table 6: the in vivo vs in vitro-produced embryos assessed with respect to metabolic sensors (NES=-1.4; FDR<0.05). Corresponding gene lists in the seven Supp. Tables from 2-bis to 6-bis are (.xls) files accessible at https://data.inra.fr/privateurl.xhtml?token=00c06d2c-f4c6-446d-bfe5-8af5e7759038.
  • Supplementary Tables 7 are also (.xls) files that are accessible at https://data.inra.fr/privateurl.xhtml?token=00c06d2c-f4c6-446d-bfe5-8af5e7759038. In Supp. Table 7: the gene lists corresponding to the shared enrichments from Figure 2 and in Supp. Table 8 the lists of the differentially expressed probes (DEP; Table 1) as well as the analyses performed on the DEGs from the in vivo vs in vitro contrast.
  • Supplementary Tables 8 are also (.xls) files that are accessible at https://data.inra.fr/privateurl.xhtml?token=00c06d2c-f4c6-446d-bfe5-8af5e7759038. In Supp. Table 7: the gene lists corresponding to the shared enrichments from Figure 2 and in Supp. Table 8 the lists of the differentially expressed probes (DEP; Table 1) as well as the analyses performed on the DEGs from the in vivo vs in vitro contrast.
  • Supplementary Tables 9
  • Supplementary Tables 10

 

    Society for Reproduction and Fertility

Article Information

Metrics

All Time Past Year Past 30 Days
Abstract Views 627 627 168
Full Text Views 107 107 10
PDF Downloads 36 36 3

Altmetrics

Related Articles

Figures

  • View in gallery

    Experimental design. After in vitro fertilisation, bovine zygotes were cultured in batches of 25 embryos in 50 µL drops, at 39°C with 5% O2, 5% CO2 and 90% N2, until blastocoels formed at D6, D7 or D8 (D6 to D8, in black). For the in vitro-extended group, blastocysts that formed at D6 or D7 were maintained in culture for one or two additional days, up to a total culture time of eight days (D7+1, D6+1, D6+2, in blue). A subset of newly formed (D6, D8) or in vitro-extended (D6+2; D7+1) blastocysts (grade 1–2) were transferred to oestrus-synchronised recipient heifers at D6 (T6) or D8 (T8) to evolve in utero up to D18 (D6-T6, D6+2-T8, D7+1-T8, D8-T8). Thin black line: in vitro development; thick dashed blue line: additional day in culture after blastocyst formation (+1 or +2); thick grey line: in utero development; grey arrows: time of in utero transfer; small (black and blue) points: time of blastocyst sampling for molecular profiling; large grey point: time of conceptus flushing for developmental evaluation and quantification of IFN-tau production.

  • View in gallery

    Common gene enrichment sets across IVP transcriptomes. GSEA identified enriched genes and functions for every paired comparison within the experimental design (9 pairs; Supplementary Tables 2, 3, 4 and 5). A search for shared genes and functions across these comparisons identified three clusters of 2–3 pairs: (D6 vs D7; D7 vs D8; D6 vs D8), (D6+2 vs D8; D6+2 vs D7+1) and (D6+1 vs D6; D7+1 vs D7). Each cluster was characterised by 2–3 shared functions, but these did not necessarily involve the identical genes in each group. The enriched functions in each subset also tended to be predominantly found in the trophectoderm cells (TE) of the embryos, though, again, not with exactly the same genes (Supplementary Table 7 and the corresponding gene lists will be provided on request). In brackets are the numbers of unique gene IDs.

  • View in gallery

    Molecular differences between in vitro and in vivo transcriptomes. Between-class analysis (BCA) on all groups of blastocysts, experimental and control (in vitro, n = 6; in vivo, n = 1). In A to C, projections based on expression values of a subset of probes that were shared among all groups, averaged on a per-group basis, along three main axes (1–3). In red, green and pink: the D6, D7 and D8 blastocysts. In yellow, blue and light green: the D6+1, D7+1 and D6+2 embryo groups. In D, the violin plots illustrate the gene scores along a normal distribution (in blue) or along each of the axes of the BCA (in red). Long distribution tails are indicative of high gene scores (or high contributions to the axis).

  • View in gallery

    Gene expression differences (n = 8) observed in microarrays (left panel) and RT-qPCRs (right panel). RT-qPCR data were normalised by qBasePLUS based on GAPDH, YWHAZ and SDHA expression. Significant differences are indicated by stars: (*) for a P value <0.05, (**) for a P value <0.01, (***) for a P value <0.001 and (****) for a P value <0.0001. PCR primers are provided in Supplementary Table 1.

  • View in gallery

    Effect of in vitro culture or culture conditions on the expression of certain DEGs as recorded by studies within the EmbryoGENE network. A meta-analysis of the entire dataset from the EmbryoGENE network was used to investigate the differential expression of certain DEGs identified herein. Presented are the adjusted P values that report the statistical relevance in studies that investigated the effects of culture conditions (SOF, co-culture in B2-BRL), origin of the oocytes (culture post ovum pick-up) and full or partial vitro vs full or partial vivo development (including the D7 control that was used herein: full vivo; Gad et al. 2012a). Among these DEGs are the eight that were selected for biological validation with qRT-PCR (in A), and the 27 that were identified as putative upstream regulators in the current dataset (in B). Note that: (i) the putative regulators encoding NANOG and STAT3 appear only once (A panel), (ii) DNMT3A was not confirmed as a DEG either internally (current RT-qPCR) or externally (meta-analysis) and (iii) the differential patterns of CDX2, DNMT3B, HDAC6, MYCN, PDX1 and TP53 were not validated beyond the current study. Adjusted P value: not validated, in grey

  • View in gallery

    Post-transfer phenotypes at D18. Bovine conceptuses were collected 10–12 days after transfer of in vitro-formed blastocysts; blastocysts were fully formed at D6–D8 or maintained in culture 1 or 2 additional day(s). In (A), recovery rates are provided. With the exception of the D7+1-T8 and D8-T8 groups, recovery rates all fell within previously reported ranges for the transfer of multiple embryos (11 per transfer, 40% (Fischer-Brown et al. 2004); 3–30 per transfer, 53% (Berg et al. 2010); 5–6 per transfer, 62% (Degrelle et al. 2012)). Conceptus sizes were recorded and conceptuses were all identified as filamentous (length range: 108–182 mm). Embryonic discs were observed and identified as gastrulating (typical HH (Hamburger and Hamilton)) stages 2–5 as defined in (Hue et al. 2001). In grey, the two control datasets: from embryos which experienced full in vivo development and those that underwent classical ET at D7 (unsorted D7 IVP embryos transferred at D7). These originated from a previous study that used the identical IVP system (culture in groups, on VERO cells, and in B2 medium plus FCS; Degrelle et al. 2012). In (B), gastrulation landmarks are indicated: streak (for primitive streak), node (for Hensen’s node), and folds (for amnion folds). In situ hybridisation with a low concentration of bovine TBXT riboprobe revealed a characteristic ‘violin’-shaped pattern in the embryonic tissues. In (C), IFN-tau (17 kDa) production was evaluated within the elongated tissues by Western blotting (3 biological samples per condition) and quantified relative to ACTB (42 kDa). The control in this case was provided by the in vivo-developed embryos.

  • View in gallery

    IFNT regulation. The expression of IFNT and its potential regulators, CDX2, ETS2 and AP-1/JUN (within the dashed lines), was analysed in the D6, D6+2 and D8 groups. RT-qPCR data were normalised by qBasePLUS based on GAPDH, YWHAZ and SDHA expression. JUN was detected and quantified at very low levels in D6 embryos, as was CDX2 in D8 embryos (lower limits of the standard curve). Significant differences are indicated by stars: (*) for a P value <0.05. PCR primers are provided in Supplementary Table 1.

PubMed

Google Scholar