Few phenomena in biology can today be better recognized than the special metabolic requirements of tumor cells. Not so however ten years ago, when Grahame Hardie and Dario Alessi discovered that the elusive upstream kinase that the Hardie lab had been pursuing in their research on the metabolic regulator AMPK was the tumor suppressor, LKB1, that the neighbouring Alessi lab was working on at the time. The resulting paper, published in 2003 in what was then Journal of Biology (now BMC Biology), was one of three (Curr Biol 2003, 13(22):2004-2008, Proc Natl Acad Sci USA 2004, 101(10):3329-3335) connecting these two kinases and that helped to swell of a surge of interest in the metabolism of tumor cells that was just beginning at about that time and is still growing.
To mark the tenth anniversary of the publication both of the paper and of the journal, BMC Biology invited Hardie and Alessi to write about how their discovery came about and where it has led.
The distinctive metabolic feature of tumor cells that enables them to meet the demands of unrestrained growth is the switch from oxidative generation of ATP to aerobic glycolysis – a phenomenon now well known as the Warburg effect. Operating this switch is one of the central functions of the AMP-activated protein kinase (AMPK) that has long been the focus of research in the Hardie lab. AMPK is an energy sensor that is allosterically tuned by competitive binding of ATP, ADP and AMP to sites on its g regulatory subunit (its portrait here, with AMP bound at two sites, was kindly provided by Bing Xiao and Stephen Gamblin). When phosphorylated by LKB1, AMPK responds to depletion of ATP by turning off anabolic reactions required for growth, and turning on catabolic reactions and oxidative phosphorylation – the reverse of the Warburg effect. In this light, it is not surprising that LKB1 is inactivated in some proportion of many different types of tumors.
This early on ignited the hope that metformin, an AMPK-activating drug that is already tried, tested, and very widely used to treat type 2 diabetes, might have anti-tumor potential – an idea that is supported by some evidence from cancer incidence in type 2 diabetics.
If only it were so simple. Effects of metformin on cancer in type 2 diabetics could be secondary to reduction in insulin levels, and although there is evidence for direct effects of AMPK activation on the development of tumors in mice, there is also recent evidence that tumors that become established without down-regulating LKB1 survive metformin better than those that have lost it – probably because metformin poisons the mitochondrial respiratory chain, depressing ATP levels, and cells in which AMPK can still be activated in response to the challenge do better than those in which it can’t.
In their review, Hardie and Alessi chart these twists and turns, and point to the explosion of further possibilities opened up by the discovery, since their 2003 publication, of at least one other class of kinase upstream of AMPK (the CaM kinases), and at least a dozen other downstream targets of LKB1 (AMPK-related kinases, or ARKs) – not to mention the innumerable downstream targets of AMPK; all which make half their schematic illustrations look like hedgehogs.
In another ten years, we may know what part the CaM kinases play in the processes regulated by AMPK, and what regulates the activation of the ARKs by phosphorylation by LKB1.