Metformin: treating the cause of Fanconi anemia?

In this issue of Blood, Zhang et al have uncovered that metformin, a first-line treatment for type 2 diabetes, can improve hematopoietic stem cell (HSC) function and reduce cancer risk in a mouse model of Fanconi anemia (FA).¹ 

FA is a rare genetic disease that leads to bone marrow failure and an extreme predisposition to cancer. Cells from these patients are unable to repair a particular kind of DNA damage, and this defect is thought to be the basis of the observed pathologies. At the molecular level, 21 gene products act in a common pathway to repair damage caused by DNA interstrand crosslinking agents (eg, cisplatinum). Although we understand quite a considerable amount about how these proteins cooperate to repair damaged DNA, we have very little understanding of why this should lead to bone marrow failure or why cancer predisposition is worse in certain tissues. One explanation could be that certain tissues have greater exposure to damage, but until recently, we had very little idea of the physiological sources of DNA damage that precipitate FA. A recently identified source of damage may be simple aldehydes.^2,3  Alternatively, it has been proposed that when HSCs leave their quiescent state, they accumulate DNA damage, necessitating FA-mediated repair.⁴  Building on these foundations, Zhang et al reveal that metformin may be the first agent that targets the source of DNA damage in FA.

Mice deficient in the key Fanconi protein FANCD2 were fed a diet supplemented with metformin. In FA-deficient mice, this treatment attenuated the blood cytopenias, and improved, but did not fully correct, the reduced frequency of HSCs and restored them to a quiescent state. Finally, treatment with metformin resulted in a small but significant reduction in the tumor predisposition of Fancd2^−/−p53^+/− mice. The magnitudes of these effects are relatively small but this may in part be due to the fact that the HSC loss in FA begins during embryonic development, however the metformin treatment was only initiated in adults (see figure).⁵ 

Despite this, metformin is the first example of a pharmacological intervention that both improves hematopoietic function and suppresses tumor predisposition. As the effect of metformin is restricted to Fanconi-deficient mice, it is plausible that metformin could be attenuating the source of damage that drives the FA phenotype.

The mechanism(s) of metformin’s effect remains to be fully uncovered. The authors go some way to address this by using a poly(I:C) treatment that mimics viral infection and induces a type I interferon response. This treatment has been shown to drive HSCs to cycle and cause bone marrow failure in a mouse model of FA, but there is no evidence of increased cancer predisposition.⁴  The mechanism by which the treatment with poly(I:C) causes aplastic anemia in FA is unclear. However, it has been shown that poly(I:C) treatment of wild-type mice leads to increased production of reactive oxygen species (ROS) and an accumulation of 8-oxo-dG, a base adduct of guanine caused by ROS. The increase in ROS production could be due to the increased metabolic demands of HSCs as they exit a quiescent state and begin to cycle, but it is unknown how this could lead to bone marrow failure in FA-deficient mice, as a functional Fanconi pathway is not required to repair DNA damage caused by agents that induce 8-oxo-dG (eg, H₂O₂).^6,7  It is possible that requirement for a FA pathway upon poly(I:C) treatment is because cells spend more time in S phase, the period when the FA pathway is active Zhang et al report that metformin is able to prevent the anemia and HSC loss caused by exposure to poly(I:C). Metformin is known to alter the metabolic activities of cells in a range of ways, notably by activating adenosine 5′-monophosphate–activated protein kinase.^7,8  Metformin may attenuate the metabolic response of HSCs as they enter the cell cycle.

Alternatively, the source of DNA damage that drives FA may be reactive aldehydes. It has been shown that disruption of aldehyde detoxification in Fanconi-deficient mice leads to a phenotype that is very similar to human patients with increased cancer predisposition and spontaneous bone marrow failure.^2,3  In addition, cells require a functional Fanconi pathway in order to resist the toxic effects of these aldehydes.⁹  In this report, Zhang et al propose that metformin may act by reacting with aldehydes, thereby rendering them inert. Fanconi patient cells treated with inhibitors of the enzyme that detoxifies the simplest aldehyde, formaldehyde, accumulate chromosomal breaks that metformin is capable of suppressing. They also show that aminoguanidine, structurally related to metformin, is able to increase the resistance of Fanconi patient cells to formaldehyde exposure. Taken together, this suggests that the mechanism of action of metformin may be through scavenging aldehydes. This hypothesis is attractive as reactive aldehydes are capable of driving both the cancer predisposition and the HSC loss—phenotypes that can be suppressed by metformin. It is plausible that in an analogous fashion to ROS, there may be a burst of aldehyde production when HSCs leave the quiescent state and enter the cell cycle, a hypothesis that should be tested. If true, this may mean that the HSC stress response and aldehyde toxicity are in fact a singular source of DNA damage in FA (see figure). Although it is clear that reactive aldehydes can drive the main phenotypes of FA in mice, it remains unclear how this occurs. It is supposed that aldehydes cause DNA damage that is repaired by the Fanconi pathway, but there is little or no evidence that aldehydes cause DNA damage in vivo. If DNA damage does occur, how does that lead to HSC loss? Finally, in the absence of FA repair, how is DNA mis-repaired leading to the chromosomal instability observed in FA patients?

Many questions remain about the mechanism of how metformin can suppress 2 of the key features of FA. However, it is plausible that metformin is the first example of a drug targeting the cause of FA. Given the exceptional safety record of metformin and devastating course of FA, there will be a strong case for testing this in human patients.