Taking a Back Door to Target Myc

See allHide authors and affiliations

Science  20 Jan 2012:
Vol. 335, Issue 6066, pp. 293-294
DOI: 10.1126/science.1217819

The transcription factor Myc coordinates the expression of a vast and functionally diverse repertoire of thousands of genes that, together, are required for the orderly proliferation of somatic cells within the body. These include genes that govern processes within the cell, such as the cell division cycle, cell metabolism and biosynthesis, cell architecture, and cell survival, as well as the multitude of processes that proliferating cells need to engage in their surrounding microenvironment, such as the generation of blood vessels, tissue remodeling, and the recruitment of cells loaded with enzymes and growth factors needed to do this. Myc is functionally nonredundant and absolutely required for the efficient proliferation of normal and cancer cells. Its expression depends on growth signals in normal cells, ensuring that its growth-promoting activities are unleashed only in cells instructed to proliferate. Control of Myc expression in cancer cells is almost always compromised. Mutations that cause Myc to become hyperactived cause uncontrollable cell proliferation and tumor formation. However, Myc has proven to be an elusive target for drug development. On page 348 of this issue, Kessler et al. (1) provide insight into how Myc's oncogenic activity might be suppressed by targeting nononcogenic proteins whose functions help Myc to transform cells.

Myc is thought to act as a “driver” of cancer, but in many cancers, the Myc genes (there are three isofunctional genes) appear untainted. In these situations, Myc expression is deregulated due to its relentless induction by oncogenic mutations in upstream signaling molecules. Whether this occurs by direct or indirect mechanisms, the outcome locks cells into a continuously proliferating state.

Experiments in which a switchable transgenic form of oncogenic Myc is used to drive tumor formation in mice have shown that inactivation of transgenic Myc in such tumors triggers dramatic regression. This is mediated by a variety of mechanisms, but typically involves terminal differentiation, tumor cell death, and collapse of the tumor microenvironment (25). Further mouse studies have also demonstrated that inhibition of endogenous Myc function also elicits a therapeutic effect in diverse tumor types in which Myc is not itself mutated and where the oncogenic driver mutations lie in other signaling pathways (6, 7). Moreover, indirect pharmacological inhibition of Myc triggers tumor regression (8). These studies have strengthened Myc's candidacy as a promising cancer drug target and also intimate that Myc inhibition might be therapeutic in many or most cancer types, irrespective of the driving oncogenic mechanism. Unfortunately, we have as yet no clue how to develop drugs that inhibit Myc function.

Myc-driven cancer.

Many genes implicated in Myc-driven cancer encode proteins that are not oncogenic and may affect a specific genetic program regulated by Myc that promote tumorigenesis.


If Myc can't be targeted, what can be done? One intriguing idea is that the oncogenic mutations that drive cancers necessarily impose novel dependencies on other, collateral pathways (9, 10). For example, individual oncogenic mutations may engage growth-inhibitory tumor suppressor pathways that curb the devastating potential of rogue cells. Indeed, Myc triggers programmed cell death (apoptosis), a self-defeating propensity that must be blocked by cooperating oncogenic mechanisms before its oncogenic capacity can be manifest. Given Myc's prodigious diversity of downstream targets, it seems likely that other such acquired dependencies exist.

To find such “synthetic lethalities,” Kessler et al. employed an unbiased screening approach in which a genome-wide library of small hairpin interfering RNAs (shRNAs) was used to block expression of genes in human mammary epithelial cells engineered to express oncogenic Myc. By comparing the abundance of shRNAs in control versus Myc-expressing cells that survive, the authors identified shRNAs (and their complementary gene targets) that confer selective disadvantage only in the presence of oncogenic Myc. This approach identified several genes already implicated in survival of Myc-driven cancer cells, including MDM2, a negative regulator of the p53 tumor suppressor pathway. Unexpectedly, however, the screen also identified a specific SUMO-activating enzyme (the heterodimer SAE1/2) (see the figure). Such enzymes covalently attach small SUMO proteins to specific target proteins, thereby modifying their target's behavior. In this way, SAEs function as general purpose, context-dependent switches. Inhibition of SAE2 triggered mitotic spindle defects in Myc-expressing cells, eliciting missegregation of chromosomes. Consistent with this, breast cancer patients with high Myc-expressing tumors but low amounts of SAE1/2 fared substantially better than patients whose cancers expressed high amounts of SAE1/2. It might appear as no surprise that interference with processes necessary for successful completion of the cell division cycle provokes unrecoverable problems. However, the unbiased approach adopted by Kessler et al. is profound in its capacity to unearth components of such necessary processes irrespective of how indirectly or circuitously they are involved.

Why do Myc-driven cells die when SAE1/2 function is ablated? Although it may be a consequence of the mitotic defects that SAE1/2 inhibition elicits, there is no direct evidence for this. Oncogenic Myc reduces the threshold for triggering apoptosis and thereby sensitizes cells to many stresses, including DNA damage, growth factor or nutrient privation, matrix detachment, and hypoxia (11, 12). So it remains possible that cell death induced by a decrease in SAE1/2 arises from interference with some other, as yet unknown, SAE1/2 function. Another nagging problem is why modest expression of deregulated Myc should elicit such remarkable synthetic lethalities at all. Endogenous Myc is already expressed at appreciable levels in proliferating cells such as human mammary epithelial cells and this amount is barely doubled upon expression of ectopic Myc. Is it the increased amount of Myc that confers vulnerability or is it because ectopic Myc, unlike normal endogenous Myc, is expressed in a constitutive unregulated manner? If the former, then the synthetic lethality observed arises from some type of novel process born of excessive Myc—perhaps the engagement of novel target genes via lower-affinity Myc-binding DNA elements—and would presumably be progressively more evident in those cancer cells expressing yet greater amounts of Myc. If the latter, then the synthetic vulnerability must have its roots in the anomalous persistence of some or other Myc effector function.

In the end, the key question is whether it is best to target a collateral Myc dependency or bite the bullet and target Myc itself. Targeted therapies frequently fail because cancer cells compensate for, and evolve around, the drug-induced blockage (13). Such relapses are a tragic by-product of the fact that biological “wetwear” is inherently noisy and works reliably only because signaling networks have evolved to be robust, self-correcting, and functionally degenerate. Hence, the therapeutic durability of targeting a pharmacologically tractable synthetic lethal function like SAE1/2 will depend not only on how tight its initial interdependency with Myc is, but how easy such interdependency is to adapt or evolve around. By contrast, Myc is functionally nonredundant and essential for tumor cell survival, so the opportunities for circumventing Myc inhibition are greatly restricted and, perhaps, absent altogether. Most likely, successful, durable, and safe cancer therapies will exploit both types of approach.


View Abstract

Navigate This Article