Metabolic control of ferroptosis in cancer

Michael P. Murphy 
Medical Research Council Mitochondrial Biology Unit, University of Cambridge, Cambridge Biomedical Campus, Cambridge CB2 0XY, UK.
e-mail: mpm@mrc-mbu.cam.ac.uk 

Abstract
Ferroptosis is a regulated non-apoptotic form of cell death and its functional role in tumorigenesis remains elusive. A study now shows that the tumour suppressor BAP1 enhances ferroptosis by modulating expression of the cystine transporter SLC7A11, leading to improved control of tumour growth. 

Iron is an essential element for cell survival and its transport and storage are tightly regulated by iron-binding proteins 1. However, free iron readily catalyses oxidative damage 2. This catalytic process occurs through the formation of ferrous iron (Fe2+), which generates highly reactive free radicals that cause peroxidation of unsaturated fatty acids in membrane lipids leading to ferroptosis, a regulated form of cell death 3,4. However, the function and regulation of ferroptosis in tumorigenesis remains unknown. 

In this issue of Nature Cell Biology, Zhang et al demonstrate that ferroptosis is metabolically regulated by BAP1 through the cystine transporter SLC7A11, and thus at least partly accounts for the tumour-suppressing function of BAP1 5. BAP1 (BRCA1-associated protein-1) is a known tumour suppressor that functions as a deubiquinitase (DUB) and removes ubiquitin (Ub) from histone 2A (H2A), thereby regulating gene expression 6,7. Mutations to BAP1 lead to various types of cancers such as renal cell carcinoma and mesothelioma 7. The authors now show that BAP1 mutations enable tumour cells to evade ferroptosis and continue to form malignant lesions. 

	Zhang et al started by identifying genes that are transcriptionally regulated by BAP1-dependent H2A-Uq occupancies, using an elegant combination of transcriptomics, epigenomic analyses and cancer genomics. In the UMRC6 cell line that has low levels of BAP1, they expressed an active or inactive, mutated form of BAP1, followed by ChIP-Seq to compare the genome-wide occupancies of H2A-Uq. This revealed that BAP1 decreased H2A-Uq occupancies for more than 5,000 genes. In parallel, they used RNA-seq to identify genes whose expression was changed in response to BAP1 activity. Combining the two data sets led to a manageable set of genes that fitted the expected pattern of lower expression when BAP1 deubiquitinated the associated H2A. To pinpoint the candidate genes associated with cancer, they analysed BAP1 levels, gene expression profiles and patient survival in a TCGA Kidney Clear Cell Carcinoma (KIRC) data set compared with that of normal cells, and identify SLC7A11, which was prominently downregulated in response to BAP1 expression. SLC7A11 was also among the top hits identified in the RNAseq and ChIP seq analysis of UMRC6 cells. This finding is in line with a recent study reporting that p53 activates ferroptosis by suppressing the expression of SLC7A11 8, implicating SLC7A11 as a promising candidate gene responsible for tumor suppression by BAP1. The authors further showed that BAP1 expression was inversely correlated with that of SLC7A11 in KIRC and other types of cancer. In addition, re-expression of active BAP1, but not of its inactive, mutant form, in BAP1-deficient cells decreased H2A-Uq occupancy of the SLC7A11 promoter and gene region, and reduced SLC7A11 expression. Together these data indicate that BAP1 acts as a tumour suppressor by decreasing SLC7A11 expression to enhance cell death. 

SLC7A11 is a cystine-glutamate exchanger which takes up cystine into the cell 9. Cystine is critical for glutathione (GSH) synthesis, a major component of the cell’s antioxidant defences. Crucially, downregulating expression of SLC7A11, or inhibiting its activity with Erastin, leads to cell death by ferroptosis 3, 4. Ferroptosis occurs when free iron induces lipid peroxidation, and hence inhibition of glutathione peroxidase 4 (Gpx4), which degrades the lipid hydroperoxides generated during lipid peroxidation, greatly enhances ferroptosis 3, 4, 10. Lipid peroxidation also generates a range of reactive aldehyde by-products that cause damage by modifying amines on protein and DNA, and GSH inactivates these damaging species catalysed by glutathione-S-transferases (GSTs)11. Ferroptosis can also be prevented by lipophilic antioxidants such as Vitamin E and by iron chelators 3, 4. 

The link between SLC7A11 and ferroptosis led the authors to propose that BAP1 acts as a tumor suppressor by downregulating the expression of SLC7A11 in transformed cells. Indeed, BAP1 overexpression in UMRC6 cells led to suppressed cystine uptake, decreased GSH levels and increased ferroptosis induced by erastin treatment. These effects were abrogated by restored SLC7A11 expression. In addition, SLC7A11 re-expression in BAP1-expressing UMRC6 cells partially restored xenograft tumour development. Mutations that inactivate BAP1 enabled cancer cells to maintain their supply of GSH, fend off death by ferroptosis, and form tumours. Supporting the idea that ferroptosis is a normal mode of cell death in preventing cancer, the authors observed condensed mitochondria by electron microscopy in tumours caused by BAP1 mutations, suggesting that these cells die by ferroptosis 12. 

	These intriguing findings have a number of implications. One is that we can add ferroptosis to the armoury cells use to prevent cancer arising. It will be interesting to see how widespread ferroptosis is compared to apoptosis in preventing cancer initiation. This further raises the question of how ferroptosis is regulated. Is suppression of cystine supply to the cell all that is required to set off cell death by ferroptosis? If so, it implies that many cells are at risk of undergoing ferroptosis. However, it is more likely that induction of ferroptosis requires several steps, potentially including enhanced iron uptake or release from internal stores. It is also conceivable that ferroptosis is constrained by multiple mechanisms, similar to apoptosis 13. This work also suggests new therapeutic approaches. An obvious strategy is to inhibit cystine uptake to starve transformed cells of cysteine, as demonstrated in vitro with the SLC7A11inihibitor erastin in the study. In addition, inhibition of Gpx4, for example with RSL310, is also likely to induce ferroptosis in cancer cells. Other approaches worth considering as combination therapies include elevating oxidative stress by enhancing free radical generation, or depleting antioxidant defences, for example by sequestering GSH.

	Zhang et al have used an elegant heuristic approach to explore the function of tumour suppresser BAP1, and in the process have uncovered a role for ferroptosis in suppressing cancer that is enhanced by attenuating cystine uptake. These findings expand our knowledge on cellular mechanisms that prevent cancer and provide further insights into the process of free radical production and the role of oxidative stress in the cell. Finally, this work suggests that ferroptosis not only arises as part of the pathology of tissue injury, but that in some circumstances it may be part of a physiological process 3,4. 

References
1.	Anderson, C. P. et al. Biochim. Biophys. Acta 1823, 1468-1483 (2012).
2.	Halliwell, B.H. & Gutteridge, J.M.C. Free radicals in Biology and Medicine, Edn. 5th. (Oxford University Press, Oxford; 2015).
3.	Stockwell, B.R. et al. Cell 171, 273-285 (2017).
4.	Yang, W.S. & Stockwell, B.R. Trends Cell Biol. 26, 165-176 (2016).
5.	Zhang et al Nat. Cell Biol. XX, x-y (2018).
6.	Rai, K., Pilarski, R., Cebulla, C.M. & Abdel-Rahman, M.H. Clin. Genet. 89, 285-294 (2016).
7.	Jensen, D.E. et al. Oncogene 16, 1097-1112 (1998).
8.	Jiang, L. et al. Nature 520, 57-62 (2015).
9.	Koppula, P., Zhang, Y., Zhuang, L. & Gan, B. Cancer Commun. 38, 12 (2018).
10.	Yang, W.S. et al. Cell 156, 317-331 (2014).
11.	Yang, Y. et al. J. Biol. Chem. 276, 19220-19230 (2001).
12.	Dolma, S., Lessnick, S.L., Hahn, W.C. & Stockwell, B.R. Cancer Cell 3, 285-296 (2003).
13. 	Portt, L. et al. Biochim. Biophys. Acta 1813, 238-259 (2011).


Competing interests

The author declares no competing interests.

Figure 1 | BAP1 enhances ferroptosis and suppresses tumorigenesis through SLC7A11. 

BAP1 acts as a deubiquinitase (DUB) that removes ubiquitin (Ub) from histone 2A (H2A). This epigenetic change decreases expression of the SLC7A11 gene which encodes a cystine/glutamate carrier, and this in turn leads to lower levels of GSH. Consequently, lipid peroxidation induced by iron is no longer prevented by the activity of GPX4, resulting in death of cancer cells. Mutations in BAP1 block this process and enable the cells to evade ferroptosis and ultimately form tumours. BAP1, BRCA1-associated protein-1; DUB, deubiquinitase; GSH, glutathione; Gpx4, glutathione peroxidase 4; H2A, histone 2A; Ub, ubiquitin.  


4