News & Views | Published:

Cell biology

Metabolism meets death

A protein that controls cell death has been found in a complex with a protein involved in glucose metabolism. Is this a point of contact between these two crucial cellular processes?

All organisms, and all cells, need food to survive: it supplies energy for everything from movement to cell proliferation. But beyond this simple relationship, there is evidence that there is a more direct connection between the molecular pathways that convert food into energy within a cell and the pathways that control whether the cell lives or dies. For instance, the continued survival of mammalian cells depends on the availability of extracellular growth-factor proteins, which inhibit cell-death pathways1 — and many of these proteins affect metabolism as well. Moreover, cancer cells not only have an increased metabolic rate but are also less likely to commit suicide when damaged2. Some of the molecular links between cell death and metabolism are beginning to be fleshed out. On page 952 of this issue, Danial and colleagues3 propose another connection: they suggest that these processes are coordinately regulated through a killer protein called BAD.

Cell suicide is an everyday event in multicellular organisms and occurs by a process called apoptosis. One other hint of a relationship between apoptosis and metabolism was that, in mammalian cells, mitochondria — the compartments in which the products of the initial metabolism of glucose (glycolysis) are used to produce energy — have a key role in apoptosis. Many apoptotic stimuli cause the release of cytochrome c and other factors from mitochondria, which in turn trigger the activation of a group of proteins called caspases; these enzymes then dismantle the cell from within in an orderly manner4.

What determines whether this chain of events is activated? Proteins of the BCL-2 family are key regulators, with different members being either pro-apoptotic (such as BAD and BAX) or anti-apoptotic (such as BCL-2 itself), either promoting or preventing the release of cytochrome c respectively. BAD has been of particular interest. When cells are treated with growth or survival factors, several enzymes are activated and in turn label BAD with phosphate groups. The result is that this newly phosphorylated protein is sequestered into inactive complexes that can no longer promote cell death5. So BAD has been proposed to be a point at which 'survival' signalling pathways intersect the machinery of death.

Might it also provide a point of connection between these events and cellular metabolism? Danial and colleagues' findings3 suggest that it might (Fig. 1). The authors took the approach of purifying the protein from mouse liver mitochondria to see which other proteins it might be interacting with. They found that BAD is present in large complexes containing several proteins that are likely to be involved in regulating its phosphorylation state, namely protein kinase A (PKA), protein phosphatase 1, and a PKA-anchoring protein, WAVE-1. But it was the fifth member of the complex that provided the real surprise: it was glucokinase, a member of the hexokinase family. These enzymes are responsible for phosphorylating glucose to produce glucose 6-phosphate — the first step in several pathways of glucose metabolism, including glycolysis and the storage of excess glucose as glycogen.

Figure 1: A point of integration for glucose metabolism and cell death?

Danial et al.3 have found the pro-apoptotic protein BAD in a mitochondrially located complex with glucokinase (as well as protein kinase A, PKA; protein phosphatase 1, PP1; and WAVE-1). Quite what the purpose of the BAD–glucokinase link might be is unclear, but these proteins are known to influence apoptosis and metabolism, respectively. BAD antagonizes BCL-2, which in turn keeps the pro-apoptotic protein BAX in check. Survival signalling pathways phosphorylate BAD and prevent it from neutralizing BCL-2. BAX might interact with the voltage-dependent anion channel (VDAC), part of the channel that exchanges newly synthesized ATP for ADP7. This interaction might promote the release of cytochrome c from mitochondria11, resulting in the activation of caspases, which in turn dismantle the cell. Relatives of glucokinase inhibit the interaction10. The metabolic role of glucokinase is to phosphorylate glucose to generate glucose 6-phosphate. In the liver, this is used to generate glycogen when blood sugar levels are high.

At this point more mature readers might find themselves reaching for dusty undergraduate biochemistry textbooks, wondering what on earth a killer like BAD is doing consorting with a regulator of glucose metabolism. In fact, the purchase of up-to-date copies of Stryer and of Lehninger by laboratories working on apoptosis has been gathering pace over the past few years, for several papers have reported close links between apoptosis and glucose metabolism. In particular, the protein kinase Akt is activated by many growth and survival factors. It is also strongly anti-apoptotic, at least in part because it can phosphorylate and inactivate BAD6. And it has profound effects on cellular metabolism, stimulating glucose transport into the cell, the recruitment of hexokinases to mitochondria, and glycolysis7,8,9. Most interestingly, Akt's ability to protect cells from apoptosis has been found to depend on the availability of glucose.

Why is there this requirement for glucose? This is not yet clear, although it has been reported that Akt promotes the interaction of hexokinase with the voltage-dependent anion channel (VDAC), which is needed to pump newly synthesized ATP (the end result of glucose metabolism, and the main form of readily available stored energy in the cell) out of mitochondria7. This interaction might prevent the channel from binding to BAX10. And because hexokinase is regulated by the level of glucose metabolites, its interaction with VDAC could perhaps respond to glucose levels. BAX is thought to be the ultimate regulator of cytochrome c release, and some data suggest that its direct binding to VDAC is essential for such regulation11 — although this is controversial.

And what of BAD? Danial et al.3 report that when this protein is missing from cells, mitochondrial glucokinase activity and respiration (leading to ATP production) decrease. This hints that BAD is needed for the formation of the mitochondrially located complex that includes glucokinase, and that BAD thereby has a positive influence on metabolism.

Notably, its ability to form such a complex does not depend on its phosphorylation state. But the authors find that glucose promotes the phosphorylation of BAD, and that the glucokinase associated with phosphorylated BAD has higher activity than that associated with unphosphorylated BAD. So glucose-mediated phosphorylation of BAD might enhance glycolysis, as well as preventing cell death. That idea is supported by the authors' finding that mice that either lack BAD or express a non-phosphorylatable form of the protein show abnormal regulation of blood glucose. The defects are similar to those seen in mice with a liver-specific deletion of glucokinase, and also in people with maturity-onset type 2 diabetes of the young, a form of non-insulin-dependent diabetes mellitus in which glucokinase mutations have been found12.

So the authors propose not only that BAD influences glucose metabolism but also that it is ideally positioned to serve as a 'sentinel', responding to abnormalities in glucose metabolism by triggering apoptosis. At present, however, it is not clear that the BAD–glucokinase complex reported here has any connection to the dependence of survival signalling on glucose that has been reported by others7,8,9. Moreover, glucokinase is unique in the hexokinase family in that its expression is largely restricted to the liver and pancreatic β-cells. So it is unlikely to have a role in connecting apoptotic and metabolic pathways in all cell types. And glucokinase functions in the liver principally to drive glycogen synthesis, not glycolysis.

It will be interesting to see whether other, more broadly expressed hexokinases are found in complexes with BAD and, if so, what aspect of glucose metabolism they are involved in — or whether they simply act as glucose sensors. Danial and colleagues' report raises more questions than it answers, but it certainly provides new pointers as to where work in this area should be directed. The intimate connections between the control of cell death and metabolism are beginning to come to light, engendering renewed respect for old biochemical pathways.


  1. 1

    Raff, M. C. Nature 356, 397–400 (1992).

  2. 2

    Hanahan, D. & Weinberg, R. A. Cell 100, 57–70 (2000).

  3. 3

    Danial, N. N. et al. Nature 424, 952–956 (2003).

  4. 4

    Green, D. R. & Evan, G. I. Cancer Cell 1, 19–30 (2002).

  5. 5

    Puthalakath, H. & Strasser, A. Cell Death Differ. 9, 505–512 (2002).

  6. 6

    Datta, S. R., Brunet, A. & Greenberg, M. E. Genes Dev. 13, 2905–2927 (1999).

  7. 7

    Gottlob, K. et al. Genes Dev. 15, 1406–1418 (2001).

  8. 8

    Vander Heiden, M. G. et al. Mol. Cell. Biol. 21, 5899–5912 (2001).

  9. 9

    Plas, D. R., Talapatra, S., Edinger, A. L., Rathmell, J. C. & Thompson, C. B. J. Biol. Chem. 276, 12041–12048 (2001).

  10. 10

    Pastorini, J. G., Shulga, N. & Hoek, J. B. J. Biol. Chem. 277, 7610–7618 (2002).

  11. 11

    Scorrano, L. & Korsmeyer, S. J. Biochem. Biophys. Res. Commun. 304, 437–444 (2003).

  12. 12

    Postic, C. et al. J. Biol. Chem. 274, 305–315 (1999).

Download references

Author information

Rights and permissions

Reprints and Permissions

About this article

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.