A hexa-species transcriptome atlas of mammalian embryogenesis delineates metabolic regulation across three different implantation modes

ARTICLE https://doi.org/10.1038/s41467-022-30194-x OPEN A hexa-species transcriptome atlas of mammalian embryogenesis delineates metabolic regulation across three different implantation modes 1234567890():,; Anna Malkowska 1,2, Christopher Penfold 1,3,4, Sophie Bergmann1,3,4 & Thorsten E. Boroviak 1,3,4 ✉ Mammalian embryogenesis relies on glycolysis and oxidative phosphorylation to balance the generation of biomass with energy production. However, the dynamics of metabolic regulation in the postimplantation embryo in vivo have remained elusive due to the inaccessibility of the implanted conceptus for biochemical studies. To address this issue, we compiled single-cell embryo profiling data in six mammalian species and determined their metabolic dynamics through glycolysis and oxidative phosphorylation associated gene expression. Strikingly, we identify a conserved switch from bivalent respiration in the late blastocyst towards a glycolytic metabolism in early gastrulation stages across species, which is independent of embryo implantation. Extraembryonic lineages followed the dynamics of the embryonic lineage, except visceral endoderm. Finally, we demonstrate that in vitro primate embryo culture substantially impacts metabolic gene regulation by comparison to in vivo samples. Our work reveals a conserved metabolic programme despite different implantation modes and highlights the need to optimise postimplantation embryo culture protocols. 1 Department of Physiology, Development and Neuroscience, University of Cambridge, Downing Site, Cambridge CB2 3EG, UK. 2 Wellcome Trust/Cancer Research UK Gurdon Institute, Henry Wellcome Building of Cancer and Developmental Biology, Cambridge, UK. 3 Centre for Trophoblast Research, University of Cambridge, Downing Site, Cambridge CB2 3EG, UK. 4 Wellcome Trust – Medical Research Council Stem Cell Institute, University of Cambridge, Jeffrey Cheah Biomedical Centre, Puddicombe Way, Cambridge CB2 0AW, UK. ✉email: NATURE COMMUNICATIONS | (2022)13:3407 | https://doi.org/10.1038/s41467-022-30194-x | www.nature.com/naturecommunications 1 ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-022-30194-x T processes of glycolysis and oxidative phosphorylation (OxPhos) (Fig. 1a). Glycolysis converts the 6-carbon molecule glucose (sugar) into two 3-carbon molecules of pyruvate to generate 2 ATPs. In anaerobic conditions, pyruvate is predominantly reduced to lactate; however, in the presence of oxygen, pyruvate can be additionally further converted to water and CO2 by he generation of complex living organisms requires coordinated large-scale construction of biomass. Similar to most cellular functions, the anabolic synthesis of biomolecules is thermodynamically unfavourable and thus coupled to adenosine triphosphate (ATP) hydrolysis as a source of free energy1,2. In mammalian cells, ATP production relies on the a b Glycolytic metabolism glucose HK PFK G6P PPP Eccentric implantation nucleotides amino acids R5P Gastrulation E6.5 Zygote 4 cells 8 cells Early Blastocyst Late blastocyst Egg cylinder E5.5 E4.5 E3.5 PGK Epi ICM Hyp ENO PK pyruvate acetyl-CoA CO2 + NADH PHD FADH2 complex LDHA ExEct VE Epi TE Gast Tb Epi c Krebs cycle lactate Implantation Blastocyst Cleavage Zygote ETC IV III ATP synthase II O2 Oxygen consumption I H 2O Pyruvate consumption Oxidative metabolism Glucose consumption e d OxPhos Glycolysis E3.5 ICM Glycolysis module Investment phase Cytochrome C Oxidase ATPase synthase (complex 4) NADH:ubiquinone oxidoreductase (complex 1) i .5 Ep .5 E6 .5 H yp VE E6 E4 5 5. Em E .5 E6 Ep i Ep i M .5 E4 .5 IC ll ce i E3 Ep Zy 4 c 8 ell ce ll E3 .5 IC E4 M .5 Ep i yp VE .5 .5 .5 H Em i Ep .5 E5 g UbiquinolCytochrome Succinate-CoQ Reductase Reductase (complex 3) (complex 2) OxPhos module Zy 0 −50 PC1 (11.22%) Zy 4 c 8 ell ce ll E3 .5 IC E4 M .5 Ep i f −100 E6 −150 0.0 .5 Zy E6 −100 E5 −50 0.5 ll E6.5 Epi 8 E5.5 Epi 4 cell ce E6.5 VE 8 cell E4 PC2 (7.19%) E4.5 Epi Score Deng et al., 2014 Mohammed et al., 2017 Zy 4 cell 8 cell E3.5 ICM E4.5 Epi E5.5 Epi E6.5 Epi E4.5 Hyp E6.5 VE E4.5 Hyp 0 1.0 4 50 Hk1 Hk2 Hk3 Gck Gpi1 Pfkl Pfkm Pfkp Aldoa Aldob Aldoc Tpi1 Gapdh Gapdhs Pgk1 Eno1 Eno2 Eno3 Pklr Pkm Adpgk Bpgm Foxk1 Foxk2 Hif1a Ldha Lin28a Lin28b Myc Pdk1 Pdk3 Pgam1 Pgam2 Pgm1 Pgm2l1 Pygl Expression −2 h i Hif1a −1 0 1 2 Ldha 5 2.0 4 Ep i Ep i .5 E6 Ep i .5 E5 E4 .5 M IC ll ce 8 E3 .5 ce 4 i Ep i Ep 5 .5 E6 E5 . M IC 5 E3 . E4 .5 ce 8 ll ce Ep i 0 ll 1 0.0 ll 2 0.5 Zy 2 3 Zy mRNA levels 1.0 4 mRNA levels 1.5 NATURE COMMUNICATIONS | (2022)13:3407 | https://doi.org/10.1038/s41467-022-30194-x | www.nature.com/naturecommunications NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-022-30194-x ARTICLE Fig. 1 Metabolic module score analysis from mouse single-cell RNA-seq datasets. a Glycolysis and OxPhos pathways. G6P glucose 6-phosphate, R5P ribose-5-phosphate, PPP phosphate pentose pathway, ETC electron transport chain. b Mouse embryonic development. TE trophectoderm, Hyp hypoblast, Epi epiblast, ExEct extraembryonic ectoderm, VE visceral endoderm, Gast gastrulating cells, Zy zygote. c Substrate consumption in mouse embryos based on18–23. d Principal component analysis (PCA) of mouse embryo samples45,46; proportion of variance in parentheses. e Boxplot graph of scores of glycolysis and OxPhos modules in cell clusters representing embryonic tissues in the mouse dataset shown in (d). N numbers for each group are as follows: Zy = 4, 4 cell = 14, 8 cell = 47, E3.5 = 99, E4.5 Epi = 28, E5.5 = 261, E6.5 Em = 205. The boxplots are defined by the 25th and 75th percentiles, with the centre as the median. The minima and maxima extend to the largest value until 1.5 of the interquartile range (IQR) and the smallest value at most 1.5 of IQR, respectively. f, g Heatmaps for scaled gene expression of the mouse dataset in (d) for the OxPhos and glycolysis module. h, i Scaled expression of Hif1a and Ldha for samples shown in (d). OxPhos3,4. OxPhos refers to the step-wise reduction of oxygen to water via the electron transport chain in the mitochondrial membranes, generating approximately 33 ATP molecules, as defined by empirical measurements3,5. Interestingly, fast proliferating cells, including cancer, often favour the apparently less efficient glycolysis pathway even under aerobic conditions (Warburg effect)1,6. The most likely explanation for the Warburg effect is that high glycolytic flux generates a multitude of anabolic precursors to synthesise biomass, while still producing sufficient ATP for cell homoeostasis2,4. Metabolites from glycolysis feed into the pentose phosphate pathway (PPP), which generates ribose-5-phosphate for nucleotides and NADPH-reducing power for the nucleotide and lipid biosynthesis (Fig. 1a)7. Similar to other rapidly proliferating cells, pluripotent cells in the ea (...truncated)