Rational Design and Synthesis of Novel Acadesine-Like Modulators of AMPK

The discovery of the numerous ties that bind AMPK to cancer has a long history, which started as long ago as 1923. In that year, German biochemist Otto Warburg postulated that a dysregulation of metabolism was a characteristic of cancerous cells. Normal cells in the presence of oxygen follow the normal metabolism pathway, from glucose to pyruvate and then to ATP through mitochondrial oxidation. Instead, tumor cells convert pyruvate to lactate, as if they were working in an anaerobic fashion and restrict mitochondrial oxidation. This observation has come to be known as the “Warburg effect” or aerobic glycolysis [6]. Even though this suggested a possible implication of AMPK, given its role in restoring energy homeostasis and mitochondrial oxidation, it wasn’t until its identification as a substrate to LKB1 that it sparked interest as a new candidate for cancer treatment. LKB1 being known as the tumor suppressor gene on human chromosome 19p13 responsible for the inheritable cancer disorder known as Peutz-Jeghers Syndrome and for being frequently mutated in various cancer types [32]. Upon further investigation of the LKB1-AMPK link it has become apparent that this pathway is crucial in the development of non-small-cell lung cancer (NSCLC), colorectal cancer (CRC), liver cancer and increasingly in other cancers like prostate, cervical and melanoma.

1. Among lung cancers, the NSCLC type is the most prevalent and in about 30-50% of the cases an LKB1 mutation is implicated, more frequently in smokers. This is thought to cause the inability to activate AMPK leading to unregulated cell proliferation. Trials with AMPK activators have shown a better prognosis in patients.

2. CRC is one of the least treatable cancers and its insurgence has a strong inflammatory and metabolic component that suggests a possible use of AMPK activators, and a better prognosis was associated with AMPK activators in a specific subset.

3. AMPK represents an obvious target in liver cancer given its central role in the organ’s function. Loss of LKB1 and AMPK expression is linked to poor prognosis in hepatocellular carcinoma (HCC) and this pathway seems to directly influence HCC cell proliferation.

4. In many other cancers as well the LKB1-AMPK pathway appears to be involved though evidence of better prognosis through AMPK activation is currently lacking. What is known is that LKB1 is somatically mutated in 20% of cervical carcinomas, in 90% of primary breast cancers AMPK activity appears to be diminished, in melanoma it regulates MITF, a protein associated with its progression and in Chronic Myeloid Leukemia, the over expression of mTOR suggests AMPK activators may provide a therapeutic advantage [33].

The evidence of AMPK implication with tumorigenesis has led to the study of the mechanisms that underlie this connection. The first observation is that, given the role of AMPK as an energy regulator whose activation inhibits all anabolic pathways thus stunting cell proliferation, it is obviously at odds with cancerous cells that have an elevated energy demand due to the continuous cell growth and division. Once activated, AMPK restores energy homeostasis by activating catabolic pathways that generate ATP efficiently rather than energy-consuming processes such as biosynthesis and cell-cycle progression. Therefore in tumor cells there is a selective pressure towards the down-regulation of AMPK [6]. More in detail the LKB1-AMPK pathway also inhibits mTOR, a known regulator of cell growth, proliferation, motility and of protein transcription and synthesis. As we have seen, AMPK inhibits mTORC1 through the phosphorylation of TSC2 and by doing so it also controls autophagy which is repressed by mTORC1 and is a key regulatory mechanism to block cell proliferation. AMPK is also linked to p53, a tumor suppressor that responds to DNA damage and cellular stress by inducing cell cycle arrest or senescence. In fact, it is inactivated in nearly half of all human cancers. AMPK phosphorylates p53, stabilizes it under conditions of metabolic stress, and promotes its transcriptional activity. In the healthy cell this regulation is enacted to ensure that in low-energy conditions the cell wastes no energy in cell growth and to survive in conditions of nutrient deprivation. Lipid metabolism in general and ACC in particular are linked to tumor progression in several cancers, and ACC inhibition has been found to induce apoptosis in these [33]. In fact, an increased rate of de novo fatty acid synthesis is a direct consequence of a tumor’s shift towards glycolytic metabolism. When the cancer cell takes up glucose it is converted to ATP and pyruvate. The latter is then converted to acetyl CoA which can either be converted through Krebs cycle to other ATP or to malonyl-CoA especially under anaerobic conditions. This passage is enacted by ACC mediated carboxylation which provides malonyl-CoA to the FAS which condenses acetyl-CoA and malonyl-CoA to form saturated long chain fatty acids. These fatty acids are transformed to phospholipids, triglycerides, and cholesterol esters and used to sustain the rapid proliferation of the cancer cell, primarily to form the lipid bilayer. Activation of AMPK causes inhibition of ACC and arrests cells between metaphase and telophase [6].

Another role of AMPK that suggests it as a target for cancer treatment is once again its role in downplaying inflammation, a condition with more than one link to tumor initiation and progression. Taken together these observations would make a compelling case for AMPK activation for tumor suppression but AMPK’s role in cancer treatment is not without controversy. It has been pointed out [34] that current AMPK activators and inhibitors are far from being specific, thus many of the results obtained are not AMPK-dependent and AMPK activation might actually occur as a defense mechanism against xenobiotics. In fact, certain AMPK-null cancer cells have shown an increased vulnerability to activators of AMPK such as acadesine and metformin. Moreover in vitro studies have shown that AMPK can even support cancer cell proliferation and can actively cooperate with oncogenes to ensure their survival. The argument for a more complex view of AMPK’s role in tumorigenesis makes many valid points:

1. Although LKB1 is a tumor suppressor it is also one whose genes are rarely frequently mutated in cancers other than lung adenocarcinoma and uterine cancer, and is very often retained suggesting it may enhance survival of cancer cells. Also worth noting is that LKB1 has at least other 13 non-AMPK targets, so the attribution of the anti-cancer benefits is tricky. At the same time the other major activator of AMPK, CAMMKβ is not a tumor suppressor but is over expressed in gastric cancer, and in colon and prostate cancer and when it activated AMPK it caused metabolic reprogramming, migration and invasion. Even so it is sometimes deleted in nerve sheath tumors and melanomas. So too AMPK and its subunits are mutated in some tumors, for example α2 in NSCLC; amplified in others, α1 in 15% of lung squamous cell cancer; and deleted in others still, β1 in 13% of nerve sheath tumors.

2. In lipogenesis although it is true that AMPK inhibits FA de novo synthesis via ACC1 inhibition, and inhibition of the FA synthesizing gene (FASN) has been shown to cause apoptosis in cancer cells, but it remains unclear if this is due to increased malonyl-CoA or lack of FAs. Lung, glioma and breast cancer are known to be resistant to FASN inhibition and ACC1 inhibition and in general they are not particularly amplified in human tumors. In addition, to target lipogenesis ACC1 might make a better target than AMPK because it would avoid the protumor functions of AMPK activation such as enhanced glycolysis, macromolecule import and autophagy.

3. Even the inhibition of mTOR is less simple than it might appear, as there are many AMPK-independent activation and inhibition mechanisms of mTOR even in the normal cell. In the cancer cell alternative splicing of TSC2, observed in prostate cancers, leads to constitutive mTOR activation even delinking it from AMPK. Furthermore, cancer cells have been known to switch from CAP-dependent translation, which requires an active mTOR, to an internal ribosome entry site (IRES)-dependent translation

The complexity of AMPK relation to cancer, the opportunities it might hold for its treatment if activated or inhibited, call for more research in this sector to better understand the subtleties of its action.