Selective utilization of glucose metabolism guides mammalian gastrulation

Abstract

The prevailing dogma for morphological patterning in developing organisms argues that the combined inputs of transcription factor networks and signalling morphogens alone generate spatially and temporally distinct expression patterns. However, metabolism has also emerged as a critical developmental regulator^1-10, independent of its functions in energy production and growth. The mechanistic role of nutrient utilization in instructing cellular programmes to shape the in vivo developing mammalian embryo remains unknown. Here we reveal two spatially resolved, cell-type- and stage-specific waves of glucose metabolism during mammalian gastrulation by using single-cell-resolution quantitative imaging of developing mouse embryos, stem cell models and embryo-derived tissue explants. We identify that the first spatiotemporal wave of glucose metabolism occurs through the hexosamine biosynthetic pathway to drive fate acquisition in the epiblast, and the second wave uses glycolysis to guide mesoderm migration and lateral expansion. Furthermore, we demonstrate that glucose exerts its influence on these developmental processes through cellular signalling pathways, with distinct mechanisms connecting glucose with the ERK activity in each wave. Our findings underscore that-in synergy with genetic mechanisms and morphogenic gradients-compartmentalized cellular metabolism is integral in guiding cell fate and specialized functions during development. This study challenges the view of the generic and housekeeping nature of cellular metabolism, offering valuable insights into its roles in various developmental contexts.

Fig. 1. A wave of glucose activity precedes mouse gastrulation.

a, Schematic of the gastrulation stage mouse embryo. A, anterior; P, posterior; D, distal; Pr, proximal; L, left; R, right. b, Top and middle, orthogonal transverse sections of in vivo and ex vivo mouse embryos showing compartmentalized 2-NBDG uptake and GLUT1 expression in epiblast and lateral mesodermal wings (orange-dashed regions). Heat map intensity colours via LUT in Fiji. Bottom, illustration of the glucose-uptake ‘wave’ phenotype throughout progression through the epiblast-to-mesoderm transition. Images are representative of 15 independent experiments. Epi, epiblast; meso, mesoderm; PS, primitive streak; VE, visceral endoderm. Scale bars, 40 µm. c, Maximum (Max.) projection images of a 2-NBDG incubated mid-streak embryo, revealing a gradient of glucose uptake in the epiblast (pink asterisks) and migratory mesoderm (white asterisks). Heat map intensity colours via LUT in Fiji. Scale bar, 40 µm. d, Single z-sections of in utero-dissected gastrula stage embryos stained for GLUT1 and SNAI1 (top; n = 53 embryos: early streak, n = 8; mid streak, n = 26; late streak, n = 19). Middle, gastrula stage embryos following 2 h of ex vivo incubation with 2-NBDG (n = 27 embryos: early streak, n = 5; mid streak, n = 13; late streak, n = 9). Epiblast is indicated by dashed white lines; primitive streak is indicated by dashed magenta lines. Heat map intensity colours via LUT in Fiji. Bottom, schematic of the glucose-uptake wave phenotype throughout the progressive stages of mouse gastrulation. ES, early streak; MS, mid streak; LS, late streak. Scale bars, 40 µm. e, Mean angle of the range of observable GLUT1 and GLUT3 (GLUT) expression (Fig. 1b,d,f) in the epiblast, plotted against the mean of the distal length of the primitive streak (Methods, ‘Image analysis’). As development progresses, the region of epiblast GLUT expression extends anteriorly. f, In utero-dissected mid-stage embryo stained for GLUT3 (n = 15 embryos) (also see Extended Data Fig. 2c). Heat map intensity colours via LUT in Fiji. Scale bar, 20 µm. g, Live multiphoton imaging of endogenous NAD(P)H in TCF/LEF:H2B–GFP (which marks the primitive streak) reporter embryos, as a readout of glucose activity (n = 12 embryos: early streak, n = 7; mid streak, n = 5). Epiblast is indicated by white dashed lines; primitive streak is indicated by green dashed lines. Scale bars, 40 µm.

Fig. 2. Glucose metabolism via HBP is necessary for mesoderm fate acquisition and maintenance.

a, Schematic of the three branches of glucose metabolism: HBP, glycolysis pathway and pentose phosphate pathway (PPP). GFAT, glutamine fructose-6-phosphate amidotransferase; PFK, phosphofructo-1-kinase; TCA, tricarboxylic acid. b, Representative z-sections of embryos in the indicated conditions following 12 h ex vivo culture. Inhibition of glucose metabolism (2-DG + BrPA) and HBP (Azaserine) impairs primitive streak elongation (dashed cyan lines) and reduces expression of T-box transcription factor T (also known as Brachyury) (cyan). Bottom, quantification of streak distal elongation lengths in embryos treated as follows: pre-culture control (n = 7), control (after 12 h of culture, n = 44), glucose metabolism inhibition (2-DG and BrPA; n = 16), 2-DG and BrPA with galactose rescue (n = 8), glycolysis inhibition (YZ9; n = 9), glycolysis inhibition (shikonin; n = 9), HBP inhibition (azaserine; n = 21), no glucose, no glutamine (nutrient sparse) plus galactose (n = 6), pentose phosphate pathway inhibition (6-AN; n = 4), ATP synthase inhibition (oligomycin; n = 8) and lactate dehydrogenase inhibition (galloflavin; n = 6). Data are mean ± s.e.m. Tukey’s multiple comparison test following ordinary one-way ANOVA; P values are shown. Scale bars, 40 µm. c, GLUT1 (heat map intensity colours via LUT in Fiji) expression in epiblast co-localizes to regions of intact basement membrane identified via laminin staining (n = 28 embryos). Scale bars, 40 µm. d, Quantification of basement membrane (BM) distal breakdown length (yellow arrows), identified from laminin expression. Pre-culture control (n = 7), control (after 12 h of culture; n = 42), glucose metabolism inhibition (2-DG and BrPA; n = 16), 2-DG and BrPA with GlcNAc rescue (n = 14), glycolysis inhibition (YZ9; n = 12), glycolysis inhibition (shikonin; n = 3), HBP inhibition (azaserine; n = 11), azaserine with GlcNAc rescue (n = 7), pentose phosphate pathway inhibition (6-AN; n = 4), ATP synthase inhibition (oligomycin; n = 6), and lactate dehydrogenase inhibition (galloflavin; n = 6). Data are mean ± s.e.m. Tukey’s multiple comparison test following ordinary one-way ANOVA; P values are shown. Scale bars, 40 µm. e, Representative outcome of in vitro EMT assay with DQ gelatin applied on epiblast-like stem cells (EpiSCs) at differentiation stage (day 2 (Extended Data Fig. 5a)). Bottom, the number of DQ⁺ clusters identified per imaging field (n = 25 fields quantified over 2 independent experimental replicates). Dunnett’s multiple comparison test following ordinary one-way ANOVA; P values are shown. Scale bars, 100 µm. Source Data

Fig. 3. Glucose metabolism modulates ERK signalling in the epiblast.

a, Western blot showing total ERK1/2 and phosphorylated (p-)ERK(T202/Y204) in EpiSCs after 12 h of the indicated treatment. n = 2 experimental replicates. For gel source data, see Supplementary Fig. 1. b, Left, schematic of tetraploid complementation assay, which generates embryos in which all cells in the embryo proper are derived from ERK-KTR^mClover 2N mouse embryonic stem cells. Right, live imaging of ERK-KTR^mClover signal in a representative embryo generated from a tetraploid complementation assay. Scale bars, 40 µm. c, Representative gastrulas are shown for ERK-KTR^mClover signal under control (n = 10), 2-DG and BrPA (n = 8), NaClO₃ (n = 4), azaserine (n = 8) and azaserine with GlcNAc rescue (n = 8) treatments. n = 3 independent experiments. Scale bars, 20 µm. Also see Extended Data Fig. 8d,e. d, Nuclear:cytoplasmic ratio of ERK-KTR^mClover in EpiSCs treated as indicated. Data are mean ± s.e.m. Control (n = 310 cells quantified) was significantly different from all other conditions (PD0325901 (PD), n = 188 cells; 2-DG, n = 166 cells; azaserine, n = 193 cells; NaClO₃, n = 120 cells; PD0325901 + FGF2, FGF4 and FGF8 (FGFs), n = 201 cells; 2-DG + FGFs, n = 178 cells; azaserine + FGFs, n = 186 cells; NaClO₃ + FGFs, n = 132 cells). Dunnett’s multiple comparison test following ordinary one-way ANOVA; P values are shown. e,f, Representative images (e) and nuclear:cytoplasmic ratio of mClover signal (f) of ERK-KTR^mClover embryonic stem cells differentiated to mesoderm under control (n = 75 cells), azaserine, 2-DG or PD0325901 (n = 50 cells each) conditions. Black horizontal line represents the mean. Dunnett’s multiple comparison test following ordinary one-way ANOVA; P values are shown. All comparisons shown are to control. Scale bars, 40 µm. g, Directed-differentiated mesoderm cells in nutrition-sparse medium (free of glucose, pyruvate and glutamine) and rescue conditions (selective reintroduction of the excluded nutrients) for 6 h. Black horizontal line represents the mean. Dunnett’s multiple comparison test following ordinary one-way ANOVA; P values are shown. Control, n = 83 cells; nutrient depleted, n = 41 cells; glucose only, n = 70 cells; glutamine only, n = 66 cells; pyruvate only, n = 47 cells. Source Data

Fig. 4. Glucose metabolism is necessary for proper mesoderm migration.

a, Schematic of mesoderm explant isolation for live imaging of migration dynamics. LS, late streak; OB, no bud. b, Control, PD0325901 or YZ9-treated mesoderm explants after 4 h of live imaging. Red asterisks mark daughter cells that divided since imaging started. TCF/LEF:H2B–GFP signal (cyan) was used for nuclear segmentation and tracking. Scale bars, 40 µm. Plots display migration metrics from AIVIA cell-tracking software across two (PD0325901) or three (all other groups) independent experimental replicates. Cell displacement and mean velocity data points represent unique cell tracks. Data are mean ± s.e.m. Dunnett’s multiple comparison test following ordinary one-way ANOVA; P values are shown. All comparisons are to control (control, n = 104 cells for displacement; n = 105 cells for velocity; 2-DG and BrPA, n = 41 cells; PD0325901, n = 85 cells for displacement, n = 87 cells for velocity; YZ9, n = 20 cells for displacement, n = 61 cells for velocity; azaserine, n = 33 cells for displacement, n = 39 cells for velocity). c, Invadapodia assay of mesoderm explants with 2-DG and BrPA, YZ9 and PD0325901 treatments. Yellow arrowheads show examples of individual punctae demonstrating substrate degradation. Scale bars, 100 µm. Data are mean ± s.e.m. Dunnett’s multiple comparison test following ordinary one-way ANOVA; P values are shown. Control, n = 20 fields; 2-DG and BrPA, n = 7 fields; PD0325901, n = 35 fields; YZ9, n = 38 fields; azaserine, n = 26 fields; 3 independent experiments for all groups. d, ECAD expression in mesoderm explants for treatment groups in c (heat map intensity colours used for better visualization). Three independent experiments for all groups. Scale bars, 100 µm. e, KEGG pathway enrichment analysis of differentially expressed genes (DEGs) in PD0325901- and YZ9-treated mesoderm explants showing specifically downregulated terms. Benjamini-adjusted P values from two-sided Fisher’s exact tests are shown. f, Summary illustration of the two distinct waves of glucose metabolic activity that selectively control cellular processes in mouse gastrulation. AVE, anterior visceral endoderm. Source Data

Extended Data Fig. 1. Mouse gastrulation is preceded by compartmentalised glucose activity.

(a-c) Visualizing glucose uptake in early-streak (a) and mid-to-late-streak (b, c) mouse gastrulas via live-imaging of 2-NBDG (fluorescent glucose analog). Illustrations demonstrate developmental stage and plane of imaging. 15 independent experiments. (a) 2-NBDG signal is high in the transitionary epiblast on the posterior side of the embryo distal to the primitive streak. (b, c) 2-NBDG remains high in the transitionary epiblast distal to the primitive streak (purple cells in illustration, wave 1), is low in the primitive streak itself (grey cells in illustration), and is high in the lateral mesoderm that has exited the primitive streak as mesenchyme (yellow cells in illustration). Scale bars, 20 µm.

Extended Data Fig. 2. Characterisation of compartmentalised glucose activity in mouse embryos during gastrulation.

(a-d) Representative z-sections of GLUT1 (a, d) and GLUT3 (b, c) protein expressions (heatmap intensity colours via Fiji’s LUT) in relation to cell types of interest in mid- (MS) or late-streak (LS) stage mouse gastrulas. Zoom insets of the epiblast and primitive streak border shown on right. A, Anterior; P, Posterior. (a (n = 21 embryos), b (n = 7 embryos), c (n = 8 embryos), d (n = 14 embryos). Scale bars represent 40 µm. (e) Representative z-section of NAD(P)H and 2-NBDG in MS stage mouse gastrula. Zoom insets of the epiblast and primitive streak border shown on the right. Scale bars represent 40 µm. (f) Top: Schematic of mouse embryo transverse planes showing laser capture microdissection transcriptome spatial coordinates of Bottom: corn plots of primitive streak gene T, glycolytic genes Slc2a1, Gpi1, Pfkp, Ldhb, and HBP genes Ogt and Gnpnat1 generated by querying genes of interest from the online e-gastrulation Geo-seq database (Peng et al.). A, Anterior; P, Posterior; AE, Endoderm Anterior; EP, Endoderm Posterior; M, Mesoderm; L, Left; R, Right.

Extended Data Fig. 3. Glucose metabolism via HBP controls mesoderm fate acquisition.

(a) Left: Schematic of glucose metabolism and its three branches: light-grey = Hexosamine Biosynthetic Pathway (HBP); pink = core/late-stage glycolysis; green = Pentose Phosphate Pathway (PPP). Red text indicates chemical inhibitors and their metabolic enzyme targets in blue. (b) Representative z-sections showing the expression pattern of SOX2, T/BRA, and LEF1 proteins in embryos treated with 2-DG+BrPA for 18 h ex vivo and (c) their Theiler developmental outcomes: Control (n = 17), 2-DG+BrPA (n = 24) across 6 experimental replicates; early-streak (ES), mid-streak (MS), late-streak (LS), no bud (OB), early bud (EB), late-bud (LB), early head fold (EHF). Scale bars represent 40 µm. Plots show mean ± SEM. On average, ~53% of control embryos develop to OB stage gastrulation, while ~73% of 2-DG+BrPA treated embryos only develop to MS stage gastrulation. (d) Dose response curves demonstrating efficacy of the inhibitors 2-DG (left; pre-culture, n = 7; control, n = 42; 0.5 mM, n = 5; 1 mM, n = 4; 2 mM, n = 7; 3 mM, n = 4; 5 mM, n = 5 embryos) and Azaserine (right; pre-culture, n = 7; control, n = 42; 1 μM, n = 3; 3 μM, n = 4; 5 μM, n = 12; 7 μM, n = 4; 10 μM, n = 3 embryos). Plots show mean and SD. Tukey’s multiple comparison tests following ordinary one-way ANOVAs. ****P < 0.0001 whenever indicated. Pre-culture vs. 1 mM 2-DG **P = 0.0066, pre-culture vs. 5 mM 2-DG **P = 0.0036, 1 mM 2-DG vs. 2 mM 2-DG **P = 0.0016, 1 mM 2-DG vs. 3 mM 2-DG *P = 0.0173, 2 mM 2-DG vs. 5 mM 2-DG *P = 0.0152, 3 mM 2-DG vs. 5 mM 2-DG *P = 0.0217. Pre-culture vs. 1 μM Azaserine *P = 0.0522, pre-culture vs. 3 μM Azaserine *P-0.0435, control vs. 5uM Azaserine ***P = 0.0006, 1 μM Azaserine vs. 10 μM Azaserine ***P = 0.0001, 5 μM Azaserine vs. 7 μM Azaserine ***P = 0.0001, 5 μM Azaserine vs. 10 μM Azaserine ***P = 0.0009. Please note that the “pre-culture” condition refers to the embryos at the beginning of the experiment, when they already exhibit some level of primitive streak formation. Drug concentrations that overall compromised embryo viability are considered toxic. Concentrations used in this study (boxed in red) altered primitive streak elongation without compromising embryo viability. (e) Representative images show mouse gastrula developmental outcomes following 12 hr metabolic inhibitor treatment. Also see Fig. 2b. A, Anterior; P, Posterior. Scale bars represent 40 µm. (f) Mouse gastrulas cultured in nutrition-sparse media (free of glucose, pyruvate, and glutamine), and rescue conditions (selective reintroduction of the excluded nutrients) to functionally validate the specific effects of chemical inhibitors. A, Anterior; P, Posterior. Scale bars represent 40 µm. Graph on the left shows mean ± SEM of SOX2 expression intensity (normalised to DAPI). Dunnett’s multiple comparison tests following ordinary one-way ANOVAs. Control (n = 4 embryos) compared to no nutrients (n = 3 embryos), glucose (n = 5 embryos), pyruvate (n = 5 embryos), glutamine (n = 4 embryos), and glucose + glutamine (n = 5 embryos for anterior epiblast, n = 3 embryos for PS + mesoderm). Graph on the right shows PS distal elongation % (Control, n = 7 embryos; no nutrient, n = 6 embryos; glucose, n = 8 embryos; pyruvate, n = 9 embryos; glutamine, n = 7 embryos). Plot shows mean + SD. (g) Representative z-sections showing similar PHOSPHO-HISTONE H3 localizations (cyan) in 2-DG-BrPA and Azaserine treated embryos compared to control embryos at the MS stage of gastrulation. Scale bars represent 40 µm. Plots show mean + SD. Ordinary one-way ANOVA. P-values shown. Source Data

Extended Data Fig. 4. Stem cell models shows the direct impact of glucose metabolism on mesoderm fate acquisition.

(a) Representative images of mouse gastruloids treated with 2-DG and YZ9. 2-DG treated gastruloids fail to elongate and do not express T. YZ9 treated gastruloids elongate and have T expression comparable to control. (b) Quantitative RT-PCR demonstrates significantly reduced T and Etv4 in 2-DG treated gastruloids. Plots show mean + SD. Dunnett’s multiple comparisons test following one-way ANOVA. P-values shown. N = 3 replicates. (c) Representative images from mXEN cells treated with 2-DG+BrPA, Azaserine, YZ9, Galloflavin, and 6AN for 24 h and treated with Oligomycin for 6 h, all shown to maintain expression of endoderm marker GATA6 (red). N = 3 independent experimental replicates examined for each condition. (d) Representative images from live imaging of early-streak stage gastrulas incubated with dyes for active mitochondria (MitoTracker, top) and mitochondrial membrane potential (TMRM, bottom). Zoomed-in panels visualizing the epiblast-VE boundary. Heaptmap intensity colours via Fiji’s LUT. N ≥ 3 embryos examined for each dye. Source Data

Extended Data Fig. 5. Mesoderm directed-differentiation reveals the direct impact of glucose metabolism on mesoderm fate acquisition and maintenance.

(a) Top Left: Schematic of mesoderm directed-differentiation steps from mESCs. Top Right: Brightfield image of starting mESC population and control cells fixed on Day (D)4 stained for T (red). Bottom: Representative images from D1-D4 2-DG+BrPA, Azaserine, PD, and SU-treated groups show an absence of T/BRA (red). Scale bars represent 200 µm. Boxes in bottom-left corner are 2X magnified from each image. Graphs show the number of T/BRA-expressing cells in (b), and number of pyknotic cells, indicating cell death, in (c), both relative to total cell number for control treatment conditions: Control (n = 12 from 6 independent experiments), 2-DG+BrPA (n = 3 per treatment length from 2 independent experiments), Azaserine (n = 5 per treatment length from 2 independent experiments), YZ9 (n = 3 per treatment length from 2 independent experiments), PD (n = 3 per treatment length from 2 independent experiments), and SU (n = 4 for D1-D4 treatment, n = 3 for D3-D4 treatment from 2 independent experiments). Graphs show mean ± SEM. Dunnett’s multiple comparison test following ordinary one-way ANOVA. Figure shows P-values. (d) qPCR analyses of directed-differentiation experiments, querying transcript changes in mesodermal pluripotency genes (treatment from D1-4 or D3-4) across 3 independent experimental replicates (4 replicates for all pluripotency genes in D1-4 Control and Azaserine, 4 replicates for Sox2 and Klf2 in D1-4 2-DG+BrPA and D1-4 PD, 2 replicates for Pdgfra in D3-4 Azaserine). Plots show mean ± SEM. Dunnett’s multiple comparisons tests following one-way ANOVAs. ****P < 0.0001. Day 1-4 compared to control: Mesp1 2-DG *P = 0.0309, Mesp1 Azaserine *P = 0.0103, Rex1 PD *P = 0.0326, Klf2 PD *P = 0.0395. Day 3-4 inhibition compared to control: Pdgfra1 PD **P = 0.0052, Sox2 PD *P = 0.0354, Klf2 PD ***P = 0.0003. Source Data

Extended Data Fig. 6. Epiblast glucose metabolism is associated with the epithelial-to-mesenchymal transition (EMT) program.

(a) Representative montage of sagittal z-sections through the MS stage embryo shows epiblast GLUT1 (heatmap intensity colours via Fiji’s LUT) expression co-localizing to regions of intact basement membrane, as identified via LAMININ (magenta) immunostainings (n = 28 embryos). A, Anterior; P, Posterior. Scale bar represents 100 µm. (b) Anterior and posterior epiblast (Epi) insets from the same embryo stained for GLUT1 and MMP14 (heatmap intensity colours via Fiji’s LUT for both protein expressions). In the posterior region, GLUT1 expression co-localises to regions of MMP14 high activity (n = 3 embryos), particularly in Epi cells bordering the primitive streak (PS) (white asterisks). Scale bars represent 40 µm. (c) Z-section of a posterior MS stage embryo showing GLUT1 and GLUT3 co-localizations to CADHERINs, especially within epiblast and PS cell neighbours (ECAD, white asterisk). Heatmap intensity colours via Fiji’s LUT. Scale bars represent 40 µm. P-values shown. (d) Principal curves of EMT, glucose metabolism, and Krebs Cycle genes over pseudo-time (Epiblast to Primitive Streak to Nascent Mesoderm states) in gastrulating in vivo mouse embryos.

Extended Data Fig. 7. Epiblast glucose metabolism is associated with the epithelial-to-mesenchymal transition (EMT) program.

(a) Representative outcome of a PD-treated in vitro EMT assay with DQ Gelatin (magenta). Quantification shows the number of DQ+ clusters identified in each imaging field (n = 25 fields quantified over 2 independent experimental replicates). Two-tailed parametric t-test. **P = 0.0011. (b) qPCR results of directed-differentiation experiments, show a downregulation of key EMT transcripts upon 2-DG+BrPA treatment. Plots show mean + SD. Two-tailed parametric t-tests. N = 3 experimental replicates for all groups. (c) Left: Single z-sections of sagittal, frontal, and transverse views of control mouse embryo’s epiblast (green-dotted line) show that GLUT1 expression (heatmap intensity colours via Fiji’s LUT) is anterior and distal to dpERK expression (green). Plot profile shows expression intensities in the epiblast across the anterior-posterior axis of the orthogonal image above. Right: 2-DG+BrPA treatment abrogates dpERK activity in cells within the epiblast (green-dotted line): Control (n = 6 embryos), 2-DG+BrPA (n = 5 embryos). Graph shows mean expression intensities in the epiblasts of each embryo. Scale bars represent 40 µm. Plots show mean + SD. Two-tailed parametric t-test. P-value shown. (d) Illustration shows experimental design. Embryo phenotypes after ERK (PD) and HBP inhibition (2DG+BrPA or Azaserine) for 12 h in culture. T/BRA marks the primitive streak. (e) Mouse gastrulas cultured with 2-DG+BrPA or PD for 18 h ex vivo show equivalent developmental delay: Control (n = 21), 2-DG+BrPA (n = 24), PD (n = 16). Plots show mean ± SEM. Source Data

Extended Data Fig. 8. HBP mediates ERK activity in the epiblast during mesodermal transition.

(a) Left: Tetraploid complementation assay generates embryos where cells of the embryo proper are only derived from ERK-KTR^mClover 2 N mESCs. Scale bar represents 40 µm. Right: Nuclear and cytoplasmic manual segmentations of an ES stage embryo (generated by tetraploid complementation) were used to quantify nuclear-to-cytoplasmic ratios (N:C) of ERK-activity, quantified in order of anterior (A) to posterior (P) location in the epiblast. Scale bar represents 40 µm. (b) Quantifications of ERK activity by calculating N:C ratio in the epiblast tissue. As development proceeds, ERK activity becomes graded across the anterior-posterior axis. Plots show epiblast N:C ratio in representative ES (R² = 0.24) and MS (R² = 0.40). (c) Multi-photon live imaging of NAD(P)H confirms that glucose metabolism activity occurs anterior to ERK activation in the epiblast (nuclear-excluded regions). n = 3 embryos. Heatmap intensity colours via Fiji’s LUT. (d) Representative images of ERK-KTR embryos collected at E6.5 and cultured for 7hrs with SU or PD show nearly uniform cytoplasmic and nuclear reporter levels in ERK-inhibited conditions. (e) 2-DG+BrPA, NaClO3, Azaserine, Azaserine+GlcNAc rescue treatments of representative MS gastrula stage embryos decrease the strength of the nuclear-excluded phenotype of ERK activity in posterior epiblast regions. Heatmap intensity colours used for better visualization of ERK-KTR nuclear exclusion. (f) SYNDECAN1 (SDC1) (upper panel) and 10E4 (lower panel) expressions within the embryonic region of mouse gastrulas after 12 h culture with indicated conditions. N = minimum of 6 embryos per condition. (g) Confirmation of EpiLSC state with upregulated OTX2 protein expression in culture. Naïve mESC pluripotency condition is indicated with 2iLif, which lacks OTX2 expression (n = 2 independent experiments). (h) Mid-embryo sagittal z-sections show PS elongation (dotted line, T/BRA expression) under indicated conditions for 12 h culture. Graph shows comparison of PS distal elongation lengths in embryos grown under indicated conditions displaying mean ± SEM. Dunnett’s multiple comparison test following ordinary one-way ANOVA. Figure shows P-values. Azaserine (n = 13), NaClO₃ (n = 9), Azaserine+FGFs (n = 7) each compared to Control (n = 9) across 3 independent experiments. Source Data

Extended Data Fig. 9. Metabolic response within the committed mesoderm.

(a) Transverse section of E6.75 mouse embryo labelled for GLUT3 and NCAD+ lateral mesoderm (n = 10 embryos) across 3 independent experiments. (b) Representative sagittal sections of mouse gastrulas treated with inhibitors (independent experiments: Azaserine (n = 11), YZ9 (n = 10), SU (n = 4) or PD (n = 13)) for 12 h ex vivo. A, Anterior; P, Posterior. Scale bar represents 40 µm. (c) Embryo illustration depicts spatiotemporal mesodermal migration upon exiting the primitive streak (PS) (d) Quantification shows NAD(P)H signal intensity for control (n = 12), YZ9 (n = 18) and 2DG+BrPA (n = 12) treated embryos. Graph shows mean as black bar. Dunnett’s multiple comparison test following ordinary one-way ANOVA. Figure shows P-values. Data points indicate different regions measured within epiblast, with each embryo colour coded. (e) PS surface areas calculated from mid-embryo sagittal z-sections show a significant size increase in PD (n = 18 embryos) and YZ9 (n = 8 embryos) treated groups compared to control (n = 35 embryos). Plots show mean + SD. Dunnett’s multiple comparison test following ordinary one-way ANOVA. Figure shows P-values. (f) Representative single z-sections show that PS surface areas (marked by T/BRA, in magenta) are expanded in PD & YZ9 treated embryos (magenta-dotted regions), with high PHOSPHO-HISTONE H3 (cyan) expression in those regions: Control (n = 7), PD (n = 6), YZ9 (n = 7). Scale bars represent 40 µm. (g) Transverse sections confirm expanded PS phenotype and PHOSPHO-HISTONE H3 localizations in the PS (red arrows). Scale bars represent 40 µm. Quantification shows Phospho-Histone H3 counts in the PS (indicated via morphology or T expression) are significantly increased in PD (n = 7) and YZ9 (n = 7) treated embryos compared to control (n = 8) and Azaserine (n = 3) treated. No significant differences are seen in SU (n = 7) and PD + YZ9 (n = 7) treated embryos. Plots show mean ± SEM. Tukey’s multiple comparison test following ordinary one-way ANOVA. Figure shows P-values. (h) Representative image from PD + YZ9 dual-treated embryo. (i) Late streak stage (E7.5) embryos with 2NBDG uptake after 4 h control and PD treatment. Source Data

Extended Data Fig. 10. Glycolytic pathway is necessary for proper mesoderm migration.

(a) Representative surface view images from control, PD and YZ9 treated embryos stained for lateral mesoderm marker SNAI1 to visualise migration distance. (b) Surface area changes of mesoderm explants following 17 hr ex situ culture show that only 2-DG-BrPA treatment (n = 3 explants) results in a significant size decrease compared to other groups (Control, n = 4; PD, n = 3; YZ9, n = 3; Azaserine, n = 3 explants). Plots show mean ± SEM. Two-tailed parametric t-test. Dunnett’s multiple comparison test following ordinary one-way ANOVA. Figure shows P-values. (c) AIVIA “cell tracking” software allows for automated segmentations of live-imaging videos to track the migration rate (graph on the left) and velocity (graph on the right) of TCF/LEF-GFP mesodermal explants across different treatment groups. Individual data points represent unique cell tracks (Control, n = 104; 2-DG+BrPA, n = 41; PD, n = 85; YZ9, n = 60; Azaserine, n = 39 cells). 3 independent experiments. Plots show mean ± SEM. Dunnett’s multiple comparison test following ordinary one-way ANOVA. Figure shows P-values. (d) NCAD expression (heatmap intensity colours via Fiji’s LUT) in mesoderm explants is retained across different treatment groups. 3 independent experiments for all groups. Scale bar represents 40 µm. (e) LEF1 (magenta) and EOMES (cyan) expression in mesoderm explants are consistently expressed across treatment groups (Azaserine, 2-DG-BrPA, YZ9, PD) after 16 h. 3 independent experiments for all groups. Scale bar represents 40 µm. (f) GO biological process enrichment of downregulated genes in mesoderm explants after PD or YZ9 treatments. P-values from Fisher’s exact tests on the X-axis. Source Data