Acetoacetate is a more efficient energy-yielding substrate for human mesenchymal stem cells than glucose and generates fewer reactive oxygen species

Abstract

Stem cells have been assumed to demonstrate a reliance on anaerobic energy generation, suited to their hypoxic in vivo environment. However, we found that human mesenchymal stem cells (hMSCs) have an active oxidative metabolism with a range of substrates. More ATP was consistently produced from substrate oxidation than glycolysis by cultured hMSCs. Strong substrate preferences were shown with the ketone body, acetoacetate, being oxidised at up to 35 times the rate of glucose. ROS-generation was 45-fold lower during acetoacetate oxidation compared with glucose and substrate preference may be an adaptation to reduce oxidative stress. The UCP2 inhibitor, genipin, increased ROS production with either acetoacetate or glucose by 2-fold, indicating a role for UCP2 in suppressing ROS production. Addition of pyruvate stimulated acetoacetate oxidation and this combination increased ATP production 27-fold, compared with glucose alone, which has implications for growth medium composition. Oxygen tension during culture affected metabolism by hMSCs. Between passages 2 and 5, rates of both glycolysis and substrate-oxidation increased at least 2-fold for normoxic (20% O₂)- but not hypoxic (5% O₂)-cultured hMSCs, despite declining growth rates and no detectable signs of differentiation. Culture of the cells with 3-hydroxybutyrate abolished the increased rates of these pathways. These findings have implications for stem cell therapy, which necessarily involves in vitro culture of cells, since low passage number normoxic cultured stem cells show metabolic adaptations without detectable changes in stem-like status.

Keywords: Acetoacetate; Glycolysis; Oxidation; ROS; Stem cells.

Graphical abstract

Fig. 1

Oxidation of energy yielding substrates by hMSCs. Rates of oxidation of 5.5 mM glucose (A); 1 mM glutamine (B); 2 mM pyruvate (C) and 10 mM acetoacetate (D) are shown for cells at passage 2 (open columns) and passage 5 (solid columns). Rates of substrate oxidation were measured in 24 well plates as the production of ¹⁴CO₂ from ¹⁴C-labelled substrates. Cells were cultured as follows through passages 1–5: normoxia (20% O₂, N) or hypoxia (5% O₂, H) with 5 mM 3-HB, where indicated. n ≥ 5 for each column. *: p < 0.05, ** p < 0.01 and ***p < 0.0001 when substrate oxidation at p5 is compared with that at p2.

Fig. 2

Oxidation of acetoacetate by hMSCs: effects of additional substrates. Rates of oxidation of 10 mM acetoacetate containing a trace of 3-¹⁴C-acetoacetate were measured in the presence of unlabelled 5.5 mM glucose or 2 mM pyruvate, where marked. Oxidation was measured in 24 well plates. Cells were passage 2, cultured under normoxic conditions. *: p < 0.05 and ***: p < 0.0001 compared with acetoacetate alone.

Fig. 3

Calculated rates of ATP production by hMSCs with oxidative substrates. Rates of ATP production from individual substrates were calculated on the basis of mol ATP produced/mol substrate oxidised. Assuming full oxidation of each mole of substrate converted to CO_2., it is calculated that 1 mol of glucose will produce a mean value of 31 mol ATP; glutamine: 20; pyruvate: 12.5; acetoacetate: 20 (refer to Materials and Methods for complete explanation). For calculations of ATP production from acetoacetate plus pyruvate, separate incubations were performed with 10 mM acetoacetate labelled with ¹⁴C-acetoacetate in the presence of unlabelled 2 mM pyruvate and with 2 mM pyruvate labelled with ¹⁴C-pyruvate in the presence of unlabelled 10 mM acetoacetate. Rates of ATP production from the labelled substrate in each incubation were calculated and the results summed. Cells were passage 2, cultured under normoxic conditions.

Fig. 4

Rates of glycolysis by hMSCs at passages 2 and 5. Rates of glycolysis were measured for hMSCs at passages 2 (open columns) and 5 (solid columns) when cultured through passages 1–5 under normoxia (20% O₂, N) or hypoxia (5% O₂, H) with 5 mM 3-HB, where indicated. Glycolytic rates were measured in 24-well plates as the production of ³H₂O from 5-³H-glucose by the method outlined in the Materials and Methods section. ***: p < 0.0001 compared with normoxic cells at passage 2; *p < 0.05 compared with cells cultured under the same conditions at passage 2.

Fig. 5

Calculated rates of ATP production by hMSCs from glycolysis and from full oxidation of glucose. Rates of ATP production were calculated as previously (refer to Materials and Methods for full explanation) from the anaerobic (glycolysis) and aerobic (oxidation) pathways for glucose metabolism. Cells were cultured under normoxic conditions (20% O₂, N) or hypoxic (5% O₂, H) with 5 mM 3-HB, where indicated, through passages 1–5. ATP production from full oxidation of ¹⁴C-labelled substrates is shown in the lower column of each data point and that from glycolysis is shown in the upper column. Measurements were made at passage 2 (open columns: oxidation; light patterning: glycolysis) and passage 5 (solid columns: oxidation; heavy patterning: glycolysis) for each culture condition. n ≥ 9.

Fig. 6

Growth rates of hMSCs with passage number. HMSCs were cultured under normoxic (20% O_2, cross-hatched) or hypoxic (5% O_2, vertical lines) conditions. Cells were passaged every 48 h and cell numbers counted by haemocytometer, using Trypan Blue to assess viability. n = 4 *: P < 0.05 compared with passage 2–3; *** p < 0.0001 compared with passage 2–3.

Fig. 7

a Production of ROS by hMSCs with glucose and acetoacetate: effects of genipin. hMSCs were cultured under normoxic conditions and were passage 5 at the time of assay. For measurements in the presence of genipin, cells were pre-incubated for 24 h with 20 μM genipin. Cells were incubated for 90 min in 24 well plates in the presence of either 5.5 mM glucose or 10 mM acetoacetate before measurement of ROS. Values were calculated for 6 replicates. The first column of each pair represents production in the absence of genipin (−, open columns) and the second column, with genipin (+, solid columns). ***: P < 0.0001 for conditions with genipin compared with no addition. b Western Blot of hMSCs with anti-UCP2 antibody. Cell extracts were probed with anti-human UCP2 antibody and the probe detected using a secondary anti-rabbit antibody. Cells were cultured as follows: Lanes A and B under normoxic conditions and lanes C and D under hypoxic conditions. The presence of UCP2 protein was detected at passages 2 (lanes A and C) and 5 (lanes B and D).