Biochem Blogs

Biochemistry blog, science writing

Ferroptosis, another way cells die

Clay Clark

Clay Clark – @biochemprof I'm on ScienceSeeker-DNA

There are a number of processes that lead to the death of a cell: apoptosis, necrosis, and autophagy are the primary mechanisms – each has a distinct biochemical and morphological fingerprint. In a new paper by Brent Stockwell’s lab at Columbia University (1), the term “ferroptosis” is proposed to describe a process of cell death that doesn’t conform to these fingerprints. In this case, cell death results from an iron-dependent accumulation of lipid reactive oxygen species (ROS).

The group previously identified two structurally-unrelated small molecules, erastin (2) and RSL3 (3), that are selectively lethal to RAS mutant cell lines and named these molecules RAS-selective lethal (RSL) compounds. The underlying driving force for these experiments is that RAS family small GTPases are mutated in a large percentage of cancers (~30%), so compounds that are selectively lethal to RAS mutant cells are needed.

RSL Compounds

RSL Compounds

Stockwell’s group showed previously that erastin targeted VDAC2/3 (voltage dependent ion channels). but it was still unknown what type of cell death was activated by RSLs. Classic features of apoptosis were not observed: cytochrome c release from the mitochondria, caspase activation, chromatin fragmentation, for example. There was, however, an increase in ROS, although the source was unknown, and RSL-induced death was prevented by iron chelation or inhibition of iron uptake. In the current publication (1), the group continues their work to understand the mechanism of action (MOA) for these compounds.

Using transmission electron microscopy (TEM), the group showed that cells undergoing ferroptosis had mitochondria that appeared smaller than normal and with increased membrane density compared to cells undergoing apoptosis (treated with staurosporine) or necrosis (treated with hydrogen peroxide). They also showed that erastin-treated cells did not have bioenergetic failure (no depletion of ATP) as observed for cells undergoing necrosis. Using a modulatory profiling strategy, they showed that erastin-induced death was not affected by compounds that inhibit apoptosis or autophagy, but cells were rescued by deferoxamine (an iron chelator), antioxidants, and MEK inhibitor. Together, the data suggested that ferroptosis requires iron-dependent ROS formation over an extended time to induce cell death.

Because mitochondria displayed aberrant morphology, the group examined whether mitochondrial gene function was affected by erastin. They designed a shRNA library that targeted 1087 genes, most of which encode predicted mitochondrial proteins, combined with a high resolution array approach to identify six genes required for erastin-induced ferroptosis. The results were shown to be generalizable by silencing the genes in additional cell lines treated with erastin. Thus, the data suggest a unique genetic network is involved in ferroptosis.

Two of the genes identified, citrate synthase (CS) and acyl-CoA synthase family member 2 (ACSF2) are implicated in regulating fatty acid metabolism in the mitochondria, suggesting that a specific lipid precursor required for ferroptosis is supplied by a pathway dependent on CS and/or ACSF2. In addition, erastin-treated cells were rescued by the addition of the transaminase inhibitor aminooxyacetic acid (AOA), which prevents the conversion of glutamine to α-ketoglutarate, so the putative lipid precursor may also be dependent on glutamine metabolism.



The authors used a custom screening library of 9,517 compounds and HTS methods to identify a compound that prevents the erastin-induced increase in ROS – ferrostatin-1 (Fer-1). Analysis of the various moieties on the compound showed that Fer-1 has intrinsic antioxidant activity, suggesting that Fer-1 is a lipid ROS scavenger, and that the N-cyclohexyl group anchors the compound in biological membranes. In addition to preventing ferroptosis in cell culture, Fer-1 also prevented glutamate-induced neurotoxicity (excitotoxic cell death) in a rat organotypic hippocampal slice culture (OHSC) model, which closely resembles the hippocampus in vivo and is used for lead-compound identification and validation. Together the data for Fer-1 suggest that glutamate-induced death in OHSCs and erastin-induced death in cancer cells share a common mechanism that can be inhibited by iron chelation (or Fer-1).

amino acid uptake

Composition of system L and system x-c. Cys, cystine; NAA, neutral amino acids.

The authors used affinity purification to identify SLC7A5 as the lone protein bound by erastin. This is one of six light chains that bind to SLC3A2 to form amino acid transporters. The various subunits confer differing substrate specificity to the transporter, where the SLC7A5/SLC3A2 complex (AKA system L) transports large, neutral amino acids. Altogether, the results suggest that the binding of erastin to SLC7A5 or the SLC7A5/SLC3A2 complex interferes with cystine uptake (system x-c) and ultimately inhibits the transplasma cysteine redox shuttle. By impairing cellular antioxidant defenses, erastin treatment results in the accumulation of toxic levels of ROS.

model for ferroptosis

Model for the ferroptosis pathway. The lethal mechanism is shown in blue.

In summary, ferroptosis is characterized by an overwhelming, iron-dependent accumulation of lethal lipid ROS. In this publication the authors:

1. used an shRNA library targeting most genes encoding mitochondrial proteins and identified roles for six genes in erastin-induced ferroptosis.

2. used a high resolution arrayed approach to provide confidence that mitochondrial genes not identified are not involved.

3. Showed that in cancer cells, inhibition of system x-c-mediated cystine uptake by erastin, sulfasalazine (SAS), or glutamate may be sufficient to initiate iron-dependent ferroptosis. However, RSL3 was shown to work downstream of system x-c and triggers a similar response, demonstrating that inhibition of system x-c is not necessary.

The specific role for iron in ferroptosis remains unclear, but one or more iron-dependent enzymes may function as part of the oxidative lethal mechanism. In cases where cancer cells contain aberrantly elevated levels of iron, the cells may be predisposed to undergo ferroptotic cell death when cystine or cysteine are limiting. Thus, the results by Stockwell and his group add a new strategy to the arsenal of methods to induce cell death in cancers.


1. Dixon, S., Lemberg, K., Lamprecht, M., Skouta, R., Zaitsev, E., Gleason, C., Patel, D., Bauer, A., Cantley, A., Yang, W. & (2012). Ferroptosis: An Iron-Dependent Form of Nonapoptotic Cell Death, Cell, 149 (5) 1072. DOI: 10.1016/j.cell.2012.03.042

2. Dolma, S., Lessnick, S.L., Hahn, W.C. & Stockwell, B.R. (2003). Identification of genotype-selective antitumor agents using synthetic lethal chemical screening in engineered human tumor cells, Cancer Cell, 3 (3) 296. DOI: 10.1016/S1535-6108(03)00050-3

3. Yang, W.S. & Stockwell, B.R. (2008). Synthetic Lethal Screening Identifies Compounds Activating Iron-Dependent, Nonapoptotic Cell Death in Oncogenic-RAS-Harboring Cancer Cells, Chemistry & Biology, 15 (3) 245. DOI: 10.1016/j.chembiol.2008.02.010