Fanconi anemia, put to sleep

In this issue of Blood,Okamoto et al describe how reduced levels of Schlafen protein SLFN11 can rescue the DNA damage sensitivity in a Fanconi anemia (FA) cell line. They propose SLFN11 as a potential target for treatment of FA.¹ 

FA is the most common inherited bone marrow failure syndrome and is also associated with an elevated predisposition to cancer. All of the phenotypes of FA are associated with unrepaired DNA interstrand crosslink (ICL) damage. Thus, FA cells are hypersensitive to endogenous byproducts of metabolism such as formaldehyde²  or chemical agents such as cisplatin that act to covalently crosslink the 2 strands of DNA. ICLs are a potent barrier to DNA replication because the strands cannot be separated for duplication. This leads to stalled replication forks (see figure). In FA cells, something else happens; not only is replication blocked by the ICL, but newly synthesized DNA adjacent to the replication fork becomes actively degraded.³  For this reason, it is thought that a major cellular function of FA gene products is fork protection.⁴ 

Although the exact mechanics of fork protection are enigmatic, they center on 2 processes. One process is the monoubiquitination-induced clamping of FANCD2:FANCI at DNA adjacent to the fork.⁵  The other process is RAD51-, BRCA1-, and BRCA2-dependent homologous recombination–mediated stabilization of the recently synthesized DNA strands.⁶  One or both of these processes are absent in all cases of FA,⁴  meaning that stalled forks are no longer protected. DNA damage then accumulates, and the cells arrest in G₂ phase because of incomplete DNA replication.

But what exactly do the stalled forks need protection from? The main bad guys are nucleases, of which there are 2 kinds: endonucleases such as MRE11 and MUS81 that cut internally to DNA, and exonucleases such as DNA2 or EXO1 that chew up DNA from exposed ends. This is a problem because nuclease-mediated degradation leads to permanent loss of genetic information (which can drive either cell death or cancer-causing mutations, depending on the context of the genes affected). Importantly, short interfering RNAs or chemical inhibition of MRE11 or DNA2 nucleases can significantly reduce the degradation of stalled forks and even ameliorate the crosslinker sensitivity of FA cells.³  In their newly published work, Okamoto et al have shown that at unprotected forks, these nucleases manage to get access to the unprotected DNA with the help of SLFN11.

The Schlafen family of genes, named after the German word meaning “to sleep,” were originally identified as factors that suppress the growth of thymocytes. Of these, SLFN11 is particularly highly expressed in blood cell lineages and plays a role in native immune, interferon, and T-cell responses during normal development (reviewed in Murai et al⁷ ). In an unbiased screen of the NCI-60 cancer set, SLFN11 is also the gene whose mutation is most highly correlated with resistance to DNA-damaging chemotherapies.⁸  Okamoto et al used a series of genetic and single-molecule experiments to show that suppression of SLFN11 expression leads to resistance of FANCD2- or FANCA-deficient fibroblasts to cisplatin, mitomycin C, or formaldehyde. They also discovered that SLFN11 deletion reversed the characteristic G₂ arrest and chromosome instability of FA cells. By using single-molecule chromatin fiber techniques, wild-type levels of fork protection in FANCD2^−/−SLFN11^−/−HAP1 cells was observed. This indicated that SLFN11 normally lets the bad guys in at stalled forks in FANCD2^−/− cells. But how does it do this?

All 13 Schlafen proteins contain a core slfn box, but at least half also contain a putative DNA unwinding domain of the superfamily 1 helicases. The adenosine triphosphate (ATP)-dependent unwinding of DNA and RNA structures at DNA damage sites is a critical role of many such helicases. In a final set of experiments, the Okamoto team demonstrated the necessity of a functional ATP-binding site for SLFN11 to facilitate mitomycin sensitivity in FA. This points to a helicase-based mechanism, which, through further experiments, was shown to involve suppression of recombination-associated factors such as RAD51 at stalled forks. A different study showed by various techniques that SLFN11 ATPase activity is also required to open chromatin in the vicinity of stalled forks,⁹  which may expose the DNA to nuclease attack. In support of this hypothesis, DNA2 or MRE11 inhibition suppresses the fork degradation of FANCD2^−/− cells that overexpress SLFN11. Thus, SLFN11 is a major factor that promotes nuclease attack on unprotected replication forks in FA cells, ultimately causing their death.

So, SLFN11 drives the death of FA cells. This might make it a target for inhibition in suppressing the phenotypes of FA. But unfortunately, this has not proved to be beneficial with other genetic suppression mechanisms observed in vitro in FA. For example, suppressing nonhomologous end-joining factors can rescue the mitomycin C sensitivity of FA cells in culture, but leads to lethality in mice.¹⁰  It remains to be determined whether SLFN11 deletion or inhibition could be similarly problematic, but given that SLFN11 is inactivated in many tumor types and at least 50% of cancer cell lines,^7,9  it is potentially a tumor suppressor. Inactivating tumor suppressors is fraught with danger. Future work should include systematic evaluation of mouse and tumor growth models in addition to examination of blood differentiation rescue phenotypes in a double-mutant FA and SLFN11 background.

FA is a complex disorder that has influenced our understanding of how cancer cells respond to interstrand crosslinking chemotherapies.⁴  SLFN11 seems to be another important part of the puzzle. The essential function of SLFN11 in inducing the death of FA cells (and probably most cancer cells, too) means it is likely to be the subject of significant further investigation for treatment of cancer and FA.