Intracellular pH controls Wnt downstream of glycolysis in amniote embryos

Abstract

Formation of the body of vertebrate embryos proceeds sequentially by posterior addition of tissues from the tail bud. Cells of the tail bud and posterior Presomitic Mesoderm (PSM), which control posterior elongation (1), exhibit a high level of aerobic glycolysis which is reminiscent of the metabolic status of cancer cells experiencing Warburg effect (2, 3). Glycolytic activity downstream of FGF controls Wnt signaling in the tail bud (3). In the Neuro-Mesodermal precursors (NMP) of the tail bud (4), Wnt signaling promotes the mesodermal fate required for sustained axial elongation, at the expense of the neural fate (3, 5). How glycolysis regulates Wnt signaling in the tail bud is currently unknown. Here we used chicken embryos and human tail bud-like cells differentiated in vitro from induced Pluripotent Stem (iPS) cells to show that these cells exhibit an inverted pH gradient, with extracellular pH (pHe) lower than intracellular pH (pHi), as observed in cancer cells (6). Our data suggest that glycolysis increases extrusion of lactate coupled to protons via the monocarboxylate (MCT) symporters. This contributes to elevate the pHi in these cells, creating a favorable chemical environment for non-enzymatic β-catenin acetylation downstream of Wnt signaling. As acetylated β-catenin promotes mesodermal rather than neural fate (7), this ultimately leads to activation of mesodermal transcriptional Wnt targets and specification of the paraxial mesoderm in tail bud precursors. Our work supports the notion that some tumor cells reactivate a developmental metabolic program.

In differentiated adult cells, the intracellular pH (pHi) is around 7.2, whereas the pHe is around 7.4. In most cancer cells, the regulation of pHe and pHi is significantly different from normal cells (8): whereas pHe is lower (6.7-7.1) than in normal adult cells, pHi is higher (>7.4) (8). To examine whether PSM and tail bud cells also exhibit an inverted pHe-pHi gradient, we electroporated the ratiometric pH sensor pHluorin (9, 10) in the anterior primitive streak region (which contains the PSM precursors) of stage 4HH chicken embryos (11). One day after electroporation, embryos were incubated in buffers of different pH (pH5.5, pH6.5, pH7.5) with the protonophores Nigericin and Valinomycin to equilibrate the pHi to the buffer pH (10). The 488/405 nm ratio in the tail bud and PSM decreased as the buffer pH increased (Extended data Figure 1a–b). We next measured the 488/405 nm ratio in electroporated embryos without protonophores. We observed a posterior to anterior gradient of pHluorin 488/405 nm signal ratio in most wild type embryos (n=11/13) (Fig. 1a–b). Tail-bud cells show a lower 488/405 nm ratio (higher pHi) whereas anterior PSM cells show a higher ratio (lower pHi) (Fig. 1a–b). We also microdissected entire PSMs from 2-day old chicken embryos and incubated them with the ratiometric pH-sensitive dye BCECF. Variations of the fluorescence ratio along the PSM confirmed the existence of a pHi gradient (Extended data Figure 1c–d). Thus, anterior PSM cells have higher intra-cytoplasmic acidity compared to posterior cells.

Figure 1: A glycolysis-dependent pHi gradient in the tail bud.

(a) (Left) Tail bud (TB) and presomitic mesoderm (PSM) region in 2-day chicken embryo; pHi gradient (pink). (Right) Corresponding region electroporated with pHluorin (n=13). Yellow: acidic regions. A: anterior. P: posterior.

(b) 488/405 nm ratio in cells in a representative embryo (n=3).

(c) Average 488/405 nm ratio of PSM region in control (Ctl) (n=8) and 2DG-treated (n=8) embryos. Two-sided unpaired t-test, *p=0.01.

(d) Normalized Average 488/405 nm ratio in control (n=8) and 2DG-treated (n=8) embryos. Red lines: Mean (±SD) of different embryos. Two-sided paired t-tests, * p=0.02; n.s. p=0.69.

(e) Axis elongation in 2-day embryos in vitro: (Top) pH7.2, 8.3mM (n=6), 0.83mM (n=3) and 0 mM Glucose (n=6). (Bottom) pH7.2 (n=6), pH6.0 (n=5) and pH5.3 (n=8). Two-way Anova followed by Tukey’s multiple comparison test. ***p<0.0005.

(f) Average 488/405 nm ratio of PSM region of 2-day embryos at pH7.2 with 8.3 mM (n=6), 0.83 mM (n=5) and 0 mM Glucose. Two-sided unpaired t-test. *p=0.01; **p=0.001; n.s. p>0.05.

(g) Normalized Average 488/405 nm ratio. Embryos cultured at pH7.2: 8.3 mM (n=6), 0.83 mM (n=5) and 0 mM Glucose (n=6). Compared average ratios +/− SD (ends of the red line) of pooled samples. Two-sided paired t-test. *p=0.002. n.s. p>0.05.

(h) Average 488/405 nm fluorescence ratio of 2-day chicken embryos at pH5.3: n=4, pH6.0: n=4, pH7.2: n=4, pH7.6: n=4; (two independent experiments). Two-sided unpaired t-test. *p<0.01; **p<0.001; n.s. p>0.05.

(i) MSGN1, SAX1, SOX2 and AXIN2 expression in the posterior region of 2-day chicken embryos at different pH (n=4 for each gene). Data normalized with pH7.2 samples. Tukey’s multiple comparison test. * p<0.05, ***p<0.001.

(j) 2-day chicken embryos hybridized with AXIN2 (pH5.3: n=6, pH7.2: n=6, pH7.6: n=4) and SOX2 (pH5.3: n=6, pH7.2: n=4, pH7.6: n=5) probes.

(b-h) Mean +/− SD. (a, b, j) Ventral views, anterior to the top. Bar, 100 μm. Exact p values can be found in the corresponding Source Data files.

Treating embryos electroporated with pHluorin in the PSM with the glycolysis inhibitor 2DG ex ovo increased the overall 488/405 nm ratio of the tail bud and posterior PSM (Fig. 1c), suggesting a decrease in pHi. A majority of electroporated embryos grown on 2DG plates show no anterior-posterior pHi gradient compared to control embryos (Fig. 1d). Culturing 2-day old chicken embryos in agar plates containing only PBS and 8.3 mM glucose at pH7.2 can sustain normal development (3) (Fig. 1e). Decreasing the glucose concentration to 0.83 mM and 0 mM resulted in a dose-dependent slowing down of axis elongation and a decrease in pHi as observed with 2DG treatment (3) (Fig. 1e–f, Extended data Fig. 2a–b). Shallower and more variable pHi gradients were evidenced in embryos cultured in low glucose conditions (Fig. 1g). Thus, decreasing glycolytic activity leads to an increase of intracytoplasmic acidity in PSM cells.

When embryos electroporated with pHluorin in minimal medium buffered at different pH were cultured for 3 hours, the signal ratio in the posterior PSM increased when the pH of the medium decreased. Thus, exposing embryos to a lower pHe leads to a corresponding acidification of the pHi in the PSM (12–14) (Fig. 1h). When embryos were cultured at pH6.0 or pH5.3, axis elongation was significantly slowed down and rapidly stalled as when glycolysis is inhibited (Fig. 1e, Extended data Fig. 2c–d) (3). Remarkably, embryos exhibiting elongation arrest after exposition to a low pH buffer resumed normal development when switched back to control medium (Extended data Fig. 2e). This developmental arrest resembles the reversible growth arrest observed in cancer cells exposed to a lower pHi (6, 12, 15).

Blocking glycolysis with 2DG in vivo results in most NMPs to differentiate to a neural SOX2/SAX1-positive fate, ultimately leading to elongation arrest (3). Strikingly, culturing embryos in a low pH buffer decreases expression of the Wnt target AXIN2 while high pH results in stronger expression (Fig. 1i–j). A similar behavior is observed for MSGN1, a PSM-specific Wnt target (Fig. 1i–j, Extended data Figure 1e) (16). In contrast, the neural genes SOX2 or SAX1 showed an opposite behavior, being upregulated at low pH (Fig. 1i–j, Extended data Figure 1f). Thus, higher pHi favors canonical Wnt signaling, promoting the paraxial mesoderm fate from NMPs, while lower pHi inhibits Wnt, promoting the neural fate.

Lactate and protons are extruded from the cell by the lactate/H+ symporters MCT1/4 which regulate pHi in highly glycolytic tissues (13). MCT1 (SLC16A1) was expressed in a graded fashion in the tail bud and posterior PSM, paralleling the pHi gradients (Extended data Figure 3a). MCT4 (SLC16A3) was also expressed in the chicken and mouse PSM (2, 15). Day 2 chicken embryos treated with the MCT inhibitor α-cyano-4-hydroxycinnamate (CNCn) exhibit increased intracellular lactate and downregulate AXIN2 and T while upregulating SOX2 and SAX1 (Extended data Figure 3b–d). Thus, inhibition of MCT function mimics glycolysis inhibition or decreasing pHi. This suggests a mechanism whereby elevated glycolysis leads to increased lactate and proton extrusion via MCT transporters thus increasing pHi in tail bud cells.

Treatment of human iPS cells with the GSK3β inhibitor CHIR99021 (Chir) with the BMP inhibitor LDN193189 (LDN) (CL medium), induces cells to differentiate to a SOX2-BRACHYURY positive NMP-like fate after day 1 (17, 18). After day 2, more than 95% of the cells differentiate to a MSGN1-positive posterior PSM fate (17–19). We observed a significant decrease of the Extracellular Acidification Rate (ECAR) and the glycolytic ATP production (Fig. 2a–b) during in vitro differentiation. In the differentiating iPS cultures, nuclear β-catenin localization (which is generally associated to Wnt activation (20)) peaked at day 2 (Extended data Figure 4a–b) (19). Time-lapse imaging of a human iPS reporter line expressing a β-catenin-GFP fusion differentiating toward the PSM fate confirmed the transient nuclear localization of β-catenin at day 1-2 (Fig. 2c). This was correlated with the expression of the PSM Wnt targets AXIN2, TBX6 and MSGN1, which peaked around day 2-3 of differentiation (Extended data Figure 4c–e) (21–23). Thus, human PSM cells exhibit an increased Wnt response during the early phase of PSM maturation in vitro, as observed in the posterior PSM in vivo (24, 25). We next examined the pHi of differentiating human iPS cells using BCECF (10). We observed a higher pHi (>7.5) at day 2 progressively decreasing in sync with Wnt signaling after day 3 (Fig. 2d). Thus, the decrease of glycolytic and Wnt activity parallels a progressive decrease of pHi in human PSM cells differentiating in vitro.

Figure 2: pHi and glycolysis decrease during in vitro differentiation of human iPS-derived PSM cells.

(a) Extra Cellular Acidification Rate (ECAR) and (b) Glycolytic ATP production rate in iPSCs and Day (D) 1-5 differentiated iPSCs cells. n=10.

(c) Snapshots from time-lapse imaging of eGFP-tagged β-catenin in differentiating iPSCs. (n=3 independent experiments). Scale bar = 50 μm.

(d) pHi analysis in human iPS cells differentiating to the PSM fate in vitro (n=3). One-way ANOVA followed by Tukey’s multiple comparisons test. **p < 0.01.

(e) Analysis of the pHi using BCECF dye in day 1-2 human iPS cells differentiated to the PSM fate in vitro and cultured for 3 h in CL culture medium at different pH (n=6 independent experiments). One-way ANOVA with Tukey’s multiple comparisons test. *p =0.023; ****p=7.26.e-6; ns p=0.999.

(f) pHi analysis in human iPS cells cultured for one day in CL medium followed by 24 h in CL medium with 2DG (n=4 for each condition). Unpaired two-tailed t-test: control vs 5mM 2DG, *P=0.01.

(g) SOX1 and AXIN2 expression in day 2 human PSM-like cells cultured for 3 h in CL medium at different pH (AXIN2 (Control, Acid, Alkali), n=5; SOX1 (Control, Acid); n=5, SOX1 (Alkali); n=6). Two-way ANOVA followed by Tukey’s multiple comparisons test. *p < 0.05; **p< 0.01.

(h-i) SOX1 and AXIN2 expression in human iPS cells cultured for one day in CL medium followed by a 24 h 2DG treatment (h) or in glucose-free CL medium (i) at normal or alkaline pH (n=3 experiments for each condition). Two-way ANOVA followed by Tukey’s multiple comparisons test. *p < 0.05; **p< 0.01.

(Control: pH7.0-7.2), Acid: pH6.3-6.5; Alkali: pH7.5-7.8) (a, d, f-h) Mean +/− SD. Exact p values can be found in the corresponding Source Data files.

Culturing iPS cells in CL medium at different pH could predictably change the pHi of cells within physiological range (Fig. 2e). Decreasing the pHi by culturing cells in acidic conditions at a stage equivalent to the NMP stage increased the expression of SOX1 and decreased the expression of AXIN2 (Fig. 2g). A similar phenotype was observed with CNCn (Extended data Figure 3e). In contrast, cells exposed to basic conditions (pH7.5 to 8) showed a reverse phenotype (Fig. 2g). Therefore, higher pHi promotes the paraxial mesoderm fate from NMP-like cells differentiating in vitro whereas lower pHi favors the neural fate.

Reducing glycolytic activity by treating iPS cells differentiating in vitro with 2DG decreases the pHi of cells and increases SOX1 expression (Fig. 2f, h). This effect can be significantly rescued by incubating cells cultured in 2DG-containing medium at a higher pH (Fig. 2h). In these conditions, SOX1 expression was decreased while AXIN2 was increased (Fig. 2h). A similar rescue was observed in glucose free-medium (Fig. 2i). Thus, our in vivo and in vitro data suggest that pHi controls Wnt signaling downstream of glycolysis in NMP-like cells.

The dynamic regulation of Wnt signaling in differentiating iPS cells is somewhat unexpected as it occurs despite the constant presence of the Wnt activator Chir in the culture medium. Chir inhibits the kinase GSK3-β which phosphorylates β-catenin and targets it for degradation (20, 26). In whole extracts from cultures of human iPS-derived PSM cells, we observed a stable expression of β-catenin and its non-phosphorylated (active) form between day 1 to day 4 (Fig. 3a). The down-regulation of nuclear β-catenin and Wnt targets observed in spite of Chir-treatment suggests that events downstream of β-catenin stabilization are dynamically regulated in PSM cells. One such candidate is K49 β-catenin acetylation, which acts as a switch controlling mesodermal vs neural gene activation in mouse embryonic stem cells (7). We observed a dynamic expression of K49 acetyl β-catenin peaking at day 1-3 in vitro (Fig. 3a), coincident with the peak of nuclear β-catenin (Fig. 3b). Switching differentiating iPS cells to acidic conditions for 3 hours at day 2 led to a decrease of K49 acetylated β-catenin whereas exposure to alkaline conditions increased acetylated β-catenin levels (Fig. 3c). In human PSM cells, K49 acetylated β-catenin expression decreased in absence of glucose (Fig. 3d). In all these conditions, levels of non-phosphorylated and total β-catenin remained stable (Fig. 3c–d). Stimulating global acetylation by treating differentiating cells with Sodium acetate, led to a K49 β-catenin acetylation increase while β-catenin and its non-phosphorylated form remained stable (Extended data Figure 5a). Acetate treatment resulted in upregulation of MSGN1 and decrease of SOX1 expression (Extended data Figure 5b). Thus, our data suggest that K49 β-catenin acetylation plays a role in the regulation of Wnt signaling during PSM differentiation in vitro.

Figure 3: Regulation of β-catenin acetylation by glycolysis and pHi.

(a) Western blot analysis of whole cell extracts prepared from hiPS cells differentiated to PSM in vitro in CL medium, using anti-acetylated K49 β-catenin, anti-active β-catenin, anti-actin and anti-β-catenin antibodies (n=3).

(b) Western blot analysis using anti-acetylated K49 β-catenin, anti-laminB, anti-GAPDH and anti-β-catenin antibodies of cytoplasmic (Cyt) and nuclear (Nucl) extracts prepared from hiPS cells differentiating to PSM in vitro in CL medium (n=3).

(c-d) Western blot analysis using anti-acetylated K49 β-catenin, anti-active β-catenin, anti-actin and anti-β-catenin antibodies. (c) Whole cell extracts of day 2 hiPS cells differentiated to PSM in vitro in CL medium and cultured at different pH for 3h (n=4). (d) Whole cell extracts of day 2 hiPS cells differentiated to PSM in CL medium in vitro and cultured for 6 h in 2DG and glucose-free (Glu(−)) CL medium (n=3).

(e) Immunoprecipitations with an anti β-catenin antibody of extracts of 2-day chicken embryos cultured at different pH. Western blot analysis using anti-acetylated lysine, anti-acetylated K49 β-catenin and anti-β-catenin antibodies. IgG: control immunoprecipitation. (n=3).

(f) Immunoprecipitations of extracts of 2-day chicken embryos cultured at different glucose concentrations with an anti β-catenin antibody. Western blot analysis using anti-acetylated lysine, anti-acetylated K49 β-catenin and anti-β-catenin antibodies. IgG: control immunoprecipitation. (n=4).

(g) Western blot analysis using anti-acetylated lysine antibody (top) and anti-β-catenin antibodies (bottom). Recombinant β-catenin protein was incubated with 500 μM or 10 mM acetyl-CoA sodium salt in PBS at different pH conditions for 3 h at 37C. (n=3).

(h) Western blot analysis using anti-acetylated lysine antibody (top) and anti-β-catenin antibodies (bottom). Recombinant β-catenin protein was incubated with 10 mM acetyl-CoA sodium salt in PBS at pH6.5~pH7.5 for 3 h at 37C. (n=3).

For gel source data, see Supplementary Figure 1

β-catenin is also acetylated in the PSM of chicken embryos (Fig. 3e). Exposure of 2-day chicken embryos to low pH (pH5.3 or 6.0) but not to high pH (pH7.6) decreased acetylated β-catenin levels in the PSM (Fig. 3e). In contrast, levels of active non-phosphorylated and of total β-catenin were stable when embryos were exposed to buffers of different pH (Fig. 3e, Extended data Fig. 6a). Thus, β-catenin acetylation is affected by pH changes as observed in vitro. Reduction of acetylated and K49 acetyl but not total or non-phosphorylated β-catenin levels was also observed when reducing glucose concentration or treating with CNCn in embryos cultured in minimal medium (Fig. 3f, Extended data Figure 3f, 6b). Conversely, treatment of 2-day chicken embryos with Sodium acetate led to an upregulation of K49 β-catenin acetylation (Extended data Figure 5c). This also resulted in a significant upregulation of AXIN2 and MSGN1 (Extended data Figure 5d–e). Together these data show that β-catenin acetylation levels in the chicken embryo PSM can be modulated by pHi, glucose or acetate independently of total and non-phosphorylated β-catenin levels.

The pHi can control the protonation of specific histidines in proteins acting as pH sensors, leading to changes in the protein properties (8, 27). While acetylation of protein substrates such as histones is largely regulated by HATs and HDACs enzymes, non-enzymatic acetylation of proteins has also been demonstrated in many different cell types (28). This chemical addition of acetyl-residues to cellular proteins can be regulated by the pHi of cells (29). Thus, the high pHi observed in PSM cells downstream of glycolysis might provide favorable chemical conditions for non-enzymatic β-catenin acetylation. To test this possibility, we incubated recombinant β-catenin protein with acetyl-CoA sodium salt in vitro at pH5.0, pH6.3, pH7.2, and pH7.6 for 3 hours at 37C. A clear dose-dependent increase of β-catenin acetylation was detected when the pH was raised (Fig. 3g). Even small pH increments within a physiological range (pH6.5 to pH 7.5) led to dose-dependent increase of β-catenin acetylation (Fig. 3h, Extended data Figure 6c).

Decreasing the pHi in tumor cells was proposed as a therapeutic strategy to treat cancer as it leads to growth arrest largely due to mTOR inhibition (6, 12). In Wnt-addicted tumor cells, lowering the pHi was recently shown to inhibit Wnt signaling, as reported here for differentiating tail bud cells (30). Thus, our findings further emphasize the tight similarity between the developing tail bud cells and cancer cells that exhibit high Warburg metabolism resulting in high pHi and low pHe (6).

METHODS

Chicken embryo culture

All animal experiments were performed in accordance to all relevant guidelines and regulations. The office for protection from Research Risks (OPRR) has interpreted “live vertebrate animal” to apply to avians (e.g., chick embryos) only after hatching. All of the studies proposed in this project only concern early developmental stages (prior to 5 days of incubation), therefore no IACUC approved protocol is required. Fertilized chicken eggs were obtained from commercial sources. Eggs were incubated at 38 °C in a humidified incubator, and embryos were staged according to Hamburger and Hamilton (HH) (11). We cultured chicken embryos mainly from stage 9HH at 37°C on a ring of whatman paper on agar plates as described in the EC culture protocol (31). Chemically-defined plates (3.5 mm petridish X25) were produced by combining an agarose solution (0.15 g agarose melted by heating in 25 ml ddH2O (MilliQ)), and 25ml 2XDPBS solution with 8.3mM, 0.83mM, and 0mM glucose. For chemically-defined medium, a 2X PBS solution (pH5.3, pH6.0, pH7.2, pH7.6) was prepared by combining 50 ml 10XDPBS (Sigma; D1283) and 450ml of ddH2O, adding (0g: pH5.3, 0.15g: pH6.0, 0.6g: pH7.2, 1.5g: pH7.6) sodium bicarbonate (Sigma, S5761) and 2.5 ml Penicillin-Streptomycin solution (Gibco (10,000 U/mL)). Embryos were prepared on a ring paper as for EC culture and incubated on the agarose plates in 2 ml of chemically-defined medium. For drug treatments, 2mM 2DG (2-Deoxy-D-glucose; Sigma), 5mM CNCn (4-Chloro-α-cyanocinnamic acid; Sigma) and 10mM sodium acetate (Sigma) were mixed to Chemically-defined plate (8.3mM glucose PBS-plate, pH7.2).

Time-lapse microscopy and axis elongation measurements

Stage 9HH chicken embryos were cultured ventral side up on a microscope stage using a custom built time-lapse station (1). We used a computer controlled, wide-field (10× objective) epifluorescent microscope (Leica DMR) workstation, equipped with a motorized stage and cooled digital camera (QImaging Retiga 1300i), to acquire 12-bit grayscale intensity images (492 × 652 pixels). For each embryo, several images corresponding to different focal planes and different fields were captured at each single time-point (frame). The acquisition rate used was 10 frames per hour (6 min between frames). To quantify axis elongation length, the last formed somite at the beginning of the time-lapse experiment was taken as a reference point, and the position of the Hensen’s Node with respect to this somite was tracked as a function of time using the manual tracking plug-in in Image J (32).

Whole mount in-situ hybridization

Stage 9HH embryos were cultured at 38 °C in chemically-defined medium. After 10 h of incubation, embryos were fixed in 4% paraformaldehyde (PFA). Whole mount in situ hybridization was carried out as described (33). Probes for AXIN2 (34), CMESPO (Mesogenin1)(35), SOX2 and SAX1 (36) have been described. Probes for MCT1 (SLC16A1) were generated from chicken embryo cDNA by PCR using published sequences.

Plasmid preparation and electroporation for chicken embryo

The ratiometric pH sensor pHluorin (9) was used to generate the expression vector pCAGG-pHluorin-IRES2-Td-Tomato. Full length pHluorin sequence was sub-cloned in pENTR-1A vector (Invitrogen), and inserted into a pCAGGS-IRES2-tdt-RFA destination vector using the Gateway system (Invitrogen). Chicken embryos ranging from stage 6HH to stage 7HH were prepared for EC culture. A DNA solution (1.0-5.0 μg/μl) was microinjected in the space between the vitelline membrane and the epiblast at the anterior primitive streak level which contains the precursors of the paraxial mesoderm. In vitro electroporations were carried out with five successive square pulses of 8V for 50ms, keeping 4mm distance between anode and cathode using Petri dish type electrodes (CUY701P2, Nepa Gene, Japan) and a CUY21 electroporator (Nepa Gene, Japan).

Intra-cytoplasmic pH measurement for chicken embryo

In order to measure the intra-cytoplasmic pH (pHi) in PSM and tail bud cells, stage 4-5 HH chicken embryos were electroporated with the pCAGGS-pHluorin-IRES2-Td-Tomato construct which contains the ratiometric pH biosensor pHluorin (9) and cultured until stage 12 HH at 38 °C. Embryos strongly expressing the constructs were subsequently selected based on Td-Tomato expression in the PSM and tail bud. We first tested whether pHluorin can accurately report on pHi differences when electroporated in the chicken embryo. Embryos were electroporated at the anterior primitive streak level with the pHluorin construct to target tail bud and PSM cells as described previously (37) and they were reincubated until they reached stage 12 HH. Embryos were then incubated in different buffers (pH5.5, pH6.5, pH7.5) from the pH calibration buffer kit (molecular probe) with 10 μM of the protonophores Nigericin and Valinomycin at 37 °C for 30 minutes. Exposure to the protonophores allows the cells’ pHi to equilibrate to the buffer pH. Then embryos were mounted on MatTek glass-bottom dishes soaked in pH calibration buffer (pH5.5, pH6.5, pH7.5). Images were captured using a laser scanning confocal microscope (TCS SP5; Leica or LSM780; Zeiss) at 37 °C in humidified atmosphere. The protonophore treatment completely abolished the gradient of pHluorin 488/405 signal ratio and the average emission signal ratio under 488/405nm excitation decreased as pH increased (Extended data Fig. 1), indicating that the pHluorin reporter electroporated in vivo can report on intra-cytoplasmic pH changes. However, because it is impossible to perform the different steps required for calibration in the same embryo, different electroporated embryos were used for each pH value and for experimental measurements. Therefore, the absolute value of the pH could not be defined and only relative pH differences are discussed. Thus, for all in vivo measurements, we only compared the 405 to 488 nm fluorescence ratios.

To measure the pHi in vivo, embryos were electroporated at stage 4-5HH with the pHluorin construct and cultured in EC culture until stage 9HH and then cultured for 10 hours with and without 2DG in EC cultures or in chemically-defined conditions before mounting. Next, embryos were mounted on MatTek glass-bottom slides on a thin albumin/agar gel in the same culture conditions. Images were captured as described above. After image capture, 3-channel z-stacks (.lif) of individual embryos were exported using Fiji into single channel single plane images (.png) for import into GoFigure2 software (www.gofigure2.org, (38)). Spherical segmentations of 9 μm radius were generated manually on the Td-Tomato channel image to encircle single cells on the raw images. Intensities of the 405nm-excited and 488nm-excited pHluorin signals were automatically calculated within the segmentations by GoFigure2. The measurements were exported for further analysis in custom made Matlab routines (Mathworks). To compare pH between different embryos, segmentations from individual embryos measured in the same day experiment were averaged to obtain a global ratio, followed by unpaired t-tests. To compare pH between different regions of individual embryos, segmentations were first normalized to the global ratio to minimize the contribution of embryo to embryo variation of overall signal intensity, followed by paired t-tests. In Figure 1b, average values are calculated every 30 segmented cells along the AP axis. Normalized Average 488/405 nm ratio in anterior and posterior PSM in control and experimental embryos were calculated from the average of 40 posterior-most and 40 anterior-most cells. Black lines connect anterior and posterior ends of the same embryo in Figure 1d and g. To calculate the average 488/405 nm fluorescence ratio of 2-day chicken embryos shown in Fig. 1f, h, in each sample, average ratio (black dot) was obtained by averaging all cells (n between 50-250/embryo). Samples were then pooled for comparing the average ratio. Ratio of fluorescence intensity is shown with Fire color using image J.

Alternatively, BCECF-AM (ThermoFisher Scientific #B1150) was used to analyze the intracellular pH in PSM explants. PSM were dissected in PBS with Ca²⁺, Mg²⁺ (Sigma) using collagenase type IV to loosen embryonic tissue. After dissection explants were incubated for 1 hour with 20 μM BCECF diluted in culture medium (DMEM/F12, 10% Fetal Bovine Serum, 1% Penicillin Streptomycin) at 37°C and 5% CO₂. Then, explants were washed with PBS with Ca²⁺, Mg²⁺. BCECF fluorescence was measured with a confocal microscope (LSM780, Zeiss) at two excitation wavelengths (405 nm and 488 nm), both collected at a collection window of 515-555nm. Statistical significance was assessed by two-sided paired t-test.

Maintenance and differentiation of human iPS cells

Human stem cell work was approved by Partners Human Research Committee (Protocol Number 2017P000438/PHS). We complied with all relevant ethical regulations. Written informed consents from the donors of the iPS cells were obtained at the time of sample collection. The cell line NCRM1 (RUCDR, Rutgers University) was used for most human iPS cell experiments. For Fig. 3B, the previously described MSGN1-RepV reporter line was used (39). We also obtained the mono-allelic mEGFP-tagged CTNNB1 WTC iPSC line from the Allen Cell Collection at the Coriell Institute (Cat. No. AICS-0058-067). Authentication was unnecessary due to the unique morphology of human iPS, as well as their unique differentiation potential. All cell lines tested negative for mycoplasma contamination. Human iPS cells were cultured on Matrigel (BD Biosciences)-coated dishes in mTeSR media (StemCells Technologies). Differentiation was performed as described in (18). Briefly, cells were plated at a density of 30,000-35,000 cells/cm2 in mTeSR supplemented with 10μM Y-27362 dihydrochloride (Rocki; Tocris Bioscience #1254). The next day (day 0 of differentiation), the medium was changed to DMEM/F12 GlutaMAX (gibco # 10565-018) supplemented with Insulin-Transferrin-Selenium (ITS, Gibco), 3 μM Chir99021 (Tocris #4423) and 0.5 μM LDN193189 (Stemgent #04-0074) (CL medium). At differentiation day 3, 20 ng/ml FGF (PeproTech #450-33) was added for an additional 3 days. After 6 days of differentiation, cells were changed to DMEM medium (gibco #11965-092) supplemented with 10 ng/ml HGF (Pepro Tech #315-23), 2 ng/ml IGF-1 (Pepro Tech #250-19), 20 ng/ml FGF and 0.5 μM LDN193189. After differentiation day8, cells were cultured with 15% KSR in DMEM supplemented with 2 ng/ml IGF-1 until day10.

Culture medium preparation

To examine the effect of medium pH, cells were incubated in media buffered at different pH for 3 hours at differentiation day2. To adjust the medium pH, HCl and Sodium bicarbonate (Sigma #S5761) were used and the pH of the medium was measured using a pH meter (Mettler toledo) or pH indicator papers (pH6.0-8.1) (GE Healthcare Life Sciences #2629-990). Prior to experiments, the medium pH was equilibrated by incubation in 5% CO2 at 37°C for 24 h.

To examine the effect of glucose-free condition, cells were incubated in glucose (−) medium for 6h at differentiation day 2 or for 24h at differentiation day 1. 5mM 2DG (Sigma #D8375) was used in 5mM D-(+)-Glucose (Sigma # G7021) containing DMEM medium (#D9807-02) to inhibit glycolysis. Culture in CL medium with 2 or 10 mM Sodium acetate (Sigma #S2889) for 24 h at differentiation day1 was used to increase the acetylation level.

Quantitative RT-PCR

RNA was extracted using Trizol (Invitrogen #15596-018), followed by precipitation with chloroform (Sigma #288306) and Ethanol (Sigma #459836). RT-PCR was performed using 1μg total RNA using SuperScriptIII Reverse Transcriptase (Invitrogen #18080-051). qPCR assays were run on a Bio-Rad CFX384 thermocycler using the iTaq Universal SYBR Green kit (Bio-Rad cat. no. 1725124). Beta-Actin was used as an internal control.

The following primers were used for PCR of human genes. ACTIN (Forward; ccaaccgcgagaagatga) (Reverse; ccagaggcgtacagggatag), MSGN1 (Forward; ctgggactggaaggacagg) (Reverse; acagctggacagggagaaga), TBX6 (Forward; aagtaccaaccccgcataca) (Reverse; taggctgtcacggagatgaa), AXIN2 (Forward; ggagtgcgttcatggtttct) (Reverse; tgcatgtgtcaatggtaggg), SOX1 (Forward; ggaatgggaggacaggattt) (Reverse; acttttatttctcggcccgt).

The following primers were used for PCR of chicken genes. BETA-ACTIN (Gg_ACTB_1_SG QuantiTect Primer Assay), AXIN2 (Gg_AXIN2_1_SG QuantiTect Primer Assay), T (Forward: cgaggagatcacagctttaaaaatt, Reverse : tcatttctttcctttgcgtcaa), MSGN1 (Forward :aaagccagtgagagggagaa, Reverse: ggtgcacttgagggtctgta), SOX2 (Forward :gcagagaaaagggaaaaagga, Reverse: tttcctagggaggggtatgaa), SAX1 (Forward:cagctttcacctacgagcag, Reverse: tggaaccagatcttcacctg).

Measurement of intracellular pH in human iPS cells

The BCECF-AM pH sensitive dye (ThermoFisher Scientific #B1150) was used to analyze the intracellular pH in human iPS cells according to the manufacturer’s instructions. Human iPS cells were differentiated to day 1 in a 96 well plate format, then incubated for 3hr in differentiation media calibrated at different pH values (described above in “Maintenance and differentiation of human iPS cells” and “Culture medium preparation”). After 3hr incubation cells were washed with HBSS (gibco #14025-092), to eliminate any exogenous source of nonspecific esterases that could prematurely cleave the lipophilic blocking groups of the BCECF-AM interfering with permeabilization of the dye, cells were then incubated in 1μM BCECF-AM ester solution (diluted in HBSS) for 20 minutes at 37°C. Next, cells were washed in HBSS, to remove non-incorporated BCECF-AM ester that remained in the media, and incubated at 37°C for 10 min in fresh differentiation media calibrated at their corresponding pH values (matching the pH of the initial 3h incubation step). To measure fluorescence of the dye, we replaced the media with calibration buffers (ThermoFisher Scientific P35379, adjusted by HCl or NaOH if necessary) to the corresponding pH values of the differentiation media (matching the pH of the initial 3h incubation step), and immediately (less than 5 minutes after the last media change; replacing differentiation media by calibration buffer) measured the ratio of emission intensity at 535 nm, by exciting the dye at 490 nm and at its isosbestic point of 440 nm using a GloMax-Multi Detection System (Promega). To determine the absolute pH values, we generated a calibration curve by incubating, in parallel, cells in wells with calibration buffer at 5 different pH values (ThermoFisher Scientific P35379, 4.5-8.5, last value adjusted by NaOH) with the addition of the protonophores Nigericin and Valinomycin (each at10μM) at 37°C for 15minutes after BCECF-AM ester incubation. As can be seen on the calibration curve shown in Extended data Figure 7, significant variation is observed for individual pH measurements, thus precluding conclusions on exact pH values. Fluorescent acquisition for calibration wells was done as described above.

Lactate assay

To detect lactate production by the differentiated iPS cells, we used a lactate assay kit (Bio Vision #K607-100) according to the manufacturer’s instructions. Measurement of O.D. at 560nm was performed using GloMax-Multi Detection System (Promega). Values were normalized by total protein amount in each well. Lactate detection in chicken tail bud was carried out as described (3).

ECAR and glycolytic ATP production measurement

NCRM1 iPSCs were differentiated on consecutive days to obtain day 1-5 populations on the day of the seahorse assay. iPSCs and day1-5 differentiated cells were dissociated using TryplE Express (Gibco cat. no. 12605036) and reseeded onto Matrigel-coated Seahorse XF96 Cell Culture Microplates (Agilent cat. no. 101085-004) at a density of 50,000 cells per well in Seahorse XF DMEM assay medium (Agilent cat. no. 103575-100). After allowing cells to attach for 2 hours, the assay medium was refreshed and the Seahorse XF Real-Time ATP Rate Assay (Agilent cat. no. 103592-100) was carried out according to the manufacturer’s instructions. All samples were run in 10 replicates in a Seahorse XFe96 Analyzer and the data were analyzed in Wave and Microsoft Excel using macros provided by the manufacturer.

Immunocytochemistry

Cells were washed with PBS and fixed in 4% PFA /PBS for 20 min at room temperature. Fixed cells were washed with PBS, then blocked and permeabilized with 3% FBS and 0.1% Triton X-100/PBS at room temperature for 30 min. Cells were then incubated with an anti-β-catenin antibody (1/500, BD #610153) and a chicken anti-GFP antibody (1/800, invitrogen #ab13970) diluted in PBS containing 3% Fetal Bovine Serum (FBS) and 0.1% Triton X-100/PBS at 4°C overnight, and next with secondary antibodies conjugated with Alexa 488 and Alexa 594 (1:500, Invitrogen #A11039, #A11037) for 30 min at room temperature. Images were captured using a laser scanning confocal microscope (LSM780, Zeiss)

Quantification of β-catenin nuclear localization

Nuclear localization of β-catenin was measured in cultures immunostained with the anti-β-catenin antibody using the Fiji software. First, we splitted the channels and smoothened the Hoechst staining images using the “Gaussian Blur” filter. Then, we binarized the image, cut the border of adjacent cells using “watershed” and eliminated the β-catenin expression in outermost of nucleus area using the “erode” filter. We next manually drew the contours of the nucleus and used the “analyze particles” tool to measure the signal intensity of β-catenin in each nucleus.

Time-lapse Imaging of β-catenin-GFP fusion iPSC line

iPS cells harboring a β-catenin-GFP fusion were differentiated as described above in 24 well glass-bottom plates and imaged on a Zeiss LSM 780 point-scanning confocal inverted microscope fitted with a large temperature incubation chamber and a CO₂ module. An Argon laser at 488 nm and 2% power was used to excite the eGFP through a 20X Plan Apo (N.A. 0.8) objective. Images were acquired every hour from the onset of differentiation until day 3 (72 hours). Representative images corresponding to 0, 12, 24, 36, 48, 60 and 72 hours were selected for illustration. Images were subjected to background subtraction and gaussian blurring in ImageJ for improved quality.

Western Blotting

Cells were collected with lysis buffer (50 mM Tris-HCl pH 8.0, 100 mM NaCl, 5 mM MgCl2, 0.5% Triton-X100, proteinase inhibitor cocktail (#78443S), 250 U/ml Benzonase) and incubated on ice for 30 min. After addition of SDS (1% final), protein concentrations were measured using a protein assay kit (BIO-RAD). The protein solution (10μg-20μg) was boiled in sample buffer and then ran on 10% or 12.5% SDS-PAGE. After transfer from the gel to polyvinylidene fluoride membrane (Millipore), the membrane was immersed in buffer containing 5% skim milk at room temperature for 1h. Membranes were then incubated with the primary antibody (diluted in 5% skim milk) overnight at 4°C. The next day, membranes were washed with PBT (PBS, 0.1% Tween 20) and incubated with Goat anti-Rabbit IgG secondary antibody, Horseradish peroxidase (HRP) conjugate (1/1000-1/10000; Invitrogen #31460) or Goat anti-Mouse IgG secondary antibody, and HRP conjugate (1/1000-1/10000; Invitrogen #31430) for 1h at room temperature. Immunoreactive bands were visualized with ECL Blotting Reagents (GE Healthcare #RPN2109) or SuperSignal™ West Pico PLUS Chemiluminescent Substrate (Thermo Scientific #34577) and detected using a Kodak X-OMAT 200A processor. After detecting the bands, membranes were washed with PBT and incubated in stripping buffer (2% SDS, 62.5mM Tris-HCl pH6.8, 0.7% 2-Mercaptoethanol) for 30min at 50°C to remove the antibody. Subsequent stainings were performed after removing the antibody. We used an Anti Acetyl-β-catenin (Lys49) antibody (1/1000; Cell Signaling Technology, #9534S), Anti-Active-β-catenin antibody (1/1000; Millipore #05-665), anti-β-catenin antibody (1/1000; BD #610153), anti-actin antibody (1/5000; Millipore #MAB1501), anti laminB1 antibody (1/5000; abcam #ab16048) and anti GAPDH antibody (1/1000; abcam #ab125247)

For chicken experiments, stage 9 HH chicken embryos were incubated at 37°C in different culture conditions (different pH or glucose concentrations) as described above. After 8 h of incubation, the posterior end of embryos was dissected and pooled (3 embryos for each sample). Samples were prepared, and Western blots were performed as for human iPS experiments.

Immunoprecipitation

2-day chicken embryos were cultured on agar plates for 8h as described above and lysed using the lysis buffer of the Immuno-precipitation kit (abcam: ab206996). Total protein concentration was adjusted to 100μg per sample. Each sample was immuno-precipitated using 1μg anti-β-catenin antibody (BD #610153) or 1 μg Normal Mouse IgG (ab188776) overnight at 4°C according to the manufacturer’s instructions. After 3 washes using the kit’s washing buffer, samples were diluted into 2X sampling buffer, then Western blotting was performed as described above using Anti Acetyl-β-catenin (Lys49) antibody (1/1000; Cell Signaling Technology, #9534S), Anti-acetyl Lysine antibody (1/1000; cell signaling #9441), anti-β-catenin antibody (1/1000; BD #610153).

In vitro acetylation of β-catenin proteins

Recombinant human β-catenin protein (20 μg; abcam ab63175) was incubated with 500μM or 10mM acetyl-CoA sodium salt (Sigma; A2056) in PBS solutions (final volume 20ul) at different pH conditions for 3 hours at 37°C. PBS solutions were prepared from 10XDPBS (Sigma; D1283) adding different amounts of sodium bicarbonate (Sigma, S5761). After incubation, 10μl of each sample were run on a 10% SDS-PAGE gel. Western blot was performed as described above using Anti-acetyl Lysine antibody (1/1000; cell signaling #9441), and anti-β-catenin antibody (1/1000; BD #610153).

Nuclear fraction

Nuclear and Cytoplasmic Extractions were performed using NE-PER™ Nuclear and Cytoplasmic Extraction kit (Thermo Fisher Scientific #78833) in accordance with the manufacturer’s instructions. After extraction, proteins were analyzed by Western blotting as described above.

Statistical analysis

Statistical analyses were performed with Prism 7.0 software (GraphPad). p values lower than 0.05 were considered to be significant. Statistical methods used in the analysis were described in figure legends.

Extended Data

Extended data Figure 1: Analysis of the intracellular pH (pHi) in the chicken embryo in vivo.

(a) Ratiometric live expression of pHluorin (488/405 nm) detected in the posterior domain of electroporated embryos exposed to different pH buffers and Nigericin and Valinomycin (n=7). Fluorescence intensity is shown by pseudocolor image (Fire color) using image J. Yellow signal indicates lower pH. Ventral view, Anterior to the left. Scale bar :100 μm.

(b) Each dot represents the average 488/405 nm signal ratio of ~300 single cells segmented in one embryo (n=2 for pH 5.5, n=3 for pH 6.5, n=2 for pH 7.5). Embryos were treated for 20 min in different pH buffers with the protonophores Nigericin and Valinomycin, before live imaging.

(c) Micro-dissected posterior PSM incubated with 20 μM BCECF (n=7). Left panel: fluorescent intensities for excitation at 405 nm (blue) and 488 nm (green). Right panel: 488/405 nm ratio. Scale: 100 μm.

(d) Fluorescence 488/405 nm ratios along the PSM. Each colored line corresponds to an explant (n=7 from two independent experiments). med: medial PSM; pos: posterior region. Two-sided paired t-test. **P=0.009

(e-f) Whole-mount in situ hybridization of 2-day chicken embryos cultured at different pH and hybridized with MSGN1 (pH5.3: n=4, pH7.2: n=3, pH7.6: n=3) (e) and SAX1 (pH5.3: n=6, pH7.2: n=4, pH7.6: n=5) (f). Ventral view, anterior to the top. Scale bar: 100 μm.

Extended data Figure 2: Lowering the pHi can reversibly slow down embryo elongation.

(a-b) Snapshots of 2-day chicken embryos cultured in minimal medium at pH 7.2 with 8.3 mM (n=6) glucose or 0.83 mM glucose (n=3).

(c-d) Snapshots of 2-day chicken embryos cultured in minimal medium with 8.3 mM glucose in acidic conditions (pH6.0: n=6, pH5.3: n=6)

(e) Snapshots of a 2-day chicken embryo first cultured in minimal medium with 8.3 mM glucose at pH5.3 showing the arrest of elongation after 9 hours and returned to control medium after 10.5 hours showing the rescue of elongation (n=6).

Bright field micrographs of the posterior region of chicken embryos taken at 1.5 hour intervals. Somites formed at the last time point are indicated by asterisks on the right. Ventral views, anterior to the top. Scale bar: 100 μm.

Extended data Figure 3: Inhibiting lactate transporters down-regulates Wnt signaling.

(a) Whole-mount in situ hybridization of a 2-day chicken embryo hybridized with MCT1 (n=4). Scale bar: 100 μm.

(b) Comparison of lactate amounts in cellular extracts of the posterior region of 2-day chicken embryos cultured for 10 h in chemically defined medium with or without 5 mM CNCn (n=3). Mean+/− SD. Two-sided unpaired t-test. P=0.0292. *p<0.05.

(c) Whole-mount in situ hybridization of 2-day chicken embryos cultured with 0 (CTL) or 5 mM CNCn and hybridized with AXIN2 (control n=8, 5mM CNCn n=7). Scale bar: 100 μm.

(d) qPCR analysis of MSGN1, SAX1, SOX2, T and AXIN2 expression in the posterior region of 2-day chicken embryos cultured with or without 5 mM CNCn (n=3 for each gene). Data were normalized by control samples. Mean +/− SD. Two-sided unpaired t-test. AXIN2, p= 0.0197, T, p= 0.0270, SOX2, p= 0.0458. * p<0.05.

(e) Comparison of AXIN2 and SOX1 mRNA expression in day 2 human iPS cells differentiated in vitro and cultured for 24h in CL medium containing 5mM CNCn or vehicle control (DMSO). n=3 biological replicates. Mean +/− SD. Two-way ANOVA followed by Tukey’s multiple comparisons test: ***p= 0.0004; **p= 0.0067. n=3.

(f) Western blot analysis using anti-acetylated K49 β-catenin, anti-active β-catenin, anti-actin and anti-β-catenin antibodies of whole cell extracts of 2-day chicken embryos cultured in chemically defined medium with 0 or 5mM CNCn for 10h. (n=3).

For gel source data, see Supplementary Figure 1

Extended data Figure 4: kinetics of Wnt/β-catenin signaling during human iPS differentiation to PSM.

(a) Immunohistochemistry showing the dynamic expression of β-catenin and Venus (YFP) proteins in human MSGN1-Venus iPS reporter cells differentiated to the PSM fate in vitro (n=3). Hoechst labeling of the nuclei is shown in blue. Scale bar: 30 μm

(b) Quantification of the intensity of nuclear localization of β-catenin shown in (a) using Fiji. Mean +/− SD, (Day 0, D1, D4, n=3; D2, D3, n=4). One-way ANOVA followed by Tukey’s multiple comparisons test: Day1 versus Day2, P=0.0411, Day2 versus Day3, P=0.0038, Day2 versus Day4, P=0.0009. *p < 0.05; **p< 0.01; ***p< 0.001.

(c-e) qPCR analysis comparing the expression level of MSGN1 (c), TBX6 (d) and AXIN2 (e) of human iPS cells differentiating to the PSM fate in vitro. Values were normalized by the results of differentiation at day 0. Mean +/− SD (n=3). One-way ANOVA followed by Tukey’s multiple comparisons test: TBX6; Day0 versus Day2, P<0.0001, Day0 versus Day3, P<0.0001, Day3 versus Day4, P<0.0001, Day4 versus Day10, P=0.0466, AXIN2; Day0 versus Day1, P=0.0006, Day0 versus Day3, P<0.0001, Day3 versus Day4, P=0.0003. *p < 0.05; ***p< 0.001; ****p< 0.0001.

Extended data Figure 5: Sodium Acetate treatment increases Wnt/β-catenin signaling in vivo and in vitro.

(a) Western blot analysis using anti-acetylated K49 β-catenin, anti-active β-catenin, anti-actin and anti-β-catenin antibodies. Whole cell extracts of day 2 hiPS cells differentiated to PSM in vitro in CL medium and treated with Sodium Acetate (SA) for 24 h (n=3).

(b) qPCR analysis of SOX1 and MSGN1 mRNA expression in day 2 hiPS cells differentiated to PSM in vitro and treated with Sodium acetate (SA) in CL medium for 24 hours. Mean +/− SD (n=3). Two-way ANOVA followed by Tukey’s multiple comparisons test: MSGN1; Control versus 10mM SA, P=0.003. **p< 0.01.

(c) Western blot analysis using anti-acetylated K49 β-catenin, anti-active β-catenin, anti-actin and anti-β-catenin antibodies of whole cell extracts of 2-day chicken embryos cultured in chemically defined medium with 0 or 10mM Sodium acetate (SA) for 10 h. (n=3).

(d) Whole-mount in situ hybridization of 2-day chicken embryos cultured with 0 or 10 mM Sodium Acetate (SA) and hybridized with AXIN2 (control n=5, 10 mM Sodium acetate n=7). Scale bar: 100 μm.

(e) qPCR analysis of AXIN2, MSGN1, SOX2 and SAX1 expression in the posterior region of 2-day chicken embryos cultured with 0 or 10 mM Sodium Acetate. Data were normalized by control samples. Mean +/− SD (n=4). Two-sided unpaired t-test. AXIN2, p= 0.0070, MSGN1, p= 0.0298. *p<0.01, **p<0.001.

For gel source data, see Supplementary Figure 1

Extended data Figure 6: β-catenin acetylation depends on intracellular pH and glycolytic activity.

(a-b) Western blot analysis using anti-acetylated K49 β-catenin, anti-active β-catenin, anti-actin and anti-β-catenin antibodies. Extracts of 2-day chicken embryos cultured in minimal medium at pH7.2 with various glucose concentrations (n=4 per condition) (a) or in minimal medium with 8.3 mM glucose at various pH (n=3 per condition) (b).

(c) Quantification of acetylated lysine and β-catenin intensity in Fig. 3h using Fiji. Acetylation rate is calculated from the slope of graph. n=2 independent experiments. Linear approximation, mean +/− SD of slope.

For gel source data, see Supplementary Figure 1

Extended data Figure 7: calibration curve used for quantifying pHi variations as a function of pHe in differentiating human iPS cells.

Calibration curve obtained for the pH measurements in differentiated iPS cells in vitro using BCECF as described in Methods. N=6 independent experiments.

DATA AVAILABILITY STATEMENT:

All data generated or analyzed during this study are included in this published article (and its supplementary information files).