Iron-sulfur cluster deficiency can be sensed by IRP2 and regulates iron homeostasis and sensitivity to ferroptosis independent of IRP1 and FBXL5

Abstract

Intracellular iron levels are strictly regulated to support homeostasis and avoid iron-mediated ROS production. Loss of iron-sulfur cluster (ISC) synthesis can increase iron loading and promote cell death by ferroptosis. Iron-responsive element-binding proteins IRP1 and IRP2 posttranscriptionally regulate iron homeostasis. IRP1 binding to target mRNAs is competitively regulated by ISC occupancy. However, IRP2 is principally thought to be regulated at the protein level via E3 ubiquitin ligase FBXL5-mediated degradation. Here, we show that ISC synthesis suppression can activate IRP2 and promote ferroptosis sensitivity via a previously unidentified mechanism. At tissue-level O₂ concentrations, ISC deficiency enhances IRP2 binding to target mRNAs independent of IRP1, FBXL5, and changes in IRP2 protein level. Deletion of both IRP1 and IRP2 abolishes the iron-starvation response, preventing its activation by ISC synthesis inhibition. These findings will inform strategies to manipulate ferroptosis sensitivity and help illuminate the mechanism underlying ISC biosynthesis disorders, such as Friedreich's ataxia.

Fig. 1. Suppression of ISC synthesis activates the iron-starvation response and sensitizes cells to ferroptosis.

(A) Relative viability of indicated cell lines expressing an shRNA targeting NFS1 (shNFS1, +) relative to a nontargeting shRNA (shGFP, −) and treated with erastin (0, 2.5, 5, and 10 μM; 2 days), 3% O₂. (B) Schematic of iron-starvation response and ferroptosis sensitivity. (C) Immunoblots from lysates derived from cell lines as in (A) for indicated proteins. ORF, open reading frame; GSH, glutathione. (D) Relative viability of indicated cell lines expressing doxycycline (dox)–inducible ISCU-DN treated with 100 nM dox (1 day) and then treated with 100 nM dox, 1 μM ferrostatin, and erastin (0, 2.5, 5, and 10 μM; 2 days) as indicated. (E) Immunoblots from lysates derived from indicated cell lines expressing dox-inducible ISCU-DN treated with 100 nM dox (1 day). (F) Relative viability of MDA-MB-231 cells expressing shGFP shNFS1 or an shRNA targeting TFR1 (shTFR1), treated with erastin (0, 2.5, 5, and 10 μM; 2 days). (G) Immunoblots of lysates derived from MDA-MB-231 cell line expressing indicated shRNAs for indicated proteins. *P < 0.01; error bars are SEM.

Fig. 2. IRP1 knockout cells activate the iron-starvation response upon inhibition of ISC synthesis and become sensitive to ferroptosis.

(A) Relative viability of parental (WT) and IRP1 knockout MDA-MB-231 cell line expressing a nontargeting shRNA (shGFP, −) or NFS1 (shNFS1, +), treated with erastin (0, 2.5, 5, and 10 μM; 2 days). (B) Labile iron pool of parental and IRP1 knockout MDA-MB-231 cell line expressing shGFP or shNFS1. (C) Immunoblots for indicated proteins from lysates derived from parental (WT) and IRP1 knockout MDA-MB-231 cell line expressing shGFP or shNFS1, treated with 50 μM Deferoxamine (DFO) and 35 μM iron nitrate as indicated (1 day). (D) Relative abundance of TFR1 and FTH1 mRNA in parental and IRP1 knockout MDA-MB-231 cell line expressing shGFP or shNFS1. (E) Relative viability of WT and IRP1 knockout 293 human embryonic kidney (HEK) cell line expressing dox-inducible ISCU-DN treated with 100 nM dox (1 day) and then treated with 100 nM dox, 1 μM ferrostatin (Fer.), and erastin (Eras.; 0, 2.5, 5, and 10 μM; 2 days) as indicated. (F) Immunoblots for indicated proteins of lysates derived from parental and IRP1 knockout HEK-293 cell line expressing dox-inducible ISCU-DN treated with dox (0, 10, and 100 μM; 1 day) as indicated. (G) Immunoblots for indicated proteins of lysates derived from parental and IRP1 knockout MDA-MB-231 cells expressing GFP reporter with 5′UTR IRE sequence from FTH1 mRNA and shRFP or shNFS1. *P < 0.01; error bars are SEM.

Fig. 3. Suppression of ISC synthesis increases IRP2 stability independent of IRP1 in atmospheric O₂ levels.

(A and B) Immunoblots for indicated proteins of lysates derived from parental (WT) (A) and IRP1 knockout (B) MDA-MB-231 cells expressing a nontargeting shRNA (shGFP, −) or shRNA targeting NFS1 (shNFS1, +) incubated in 3% or 21% O₂. (C and D) Immunoblots for indicated proteins from lysates derived from IRP1/IRP2 double-knockout MDA-MB-231 expressing dox-repressible IRP2 and shGFP or shNFS1, treated with 500 nM dox for 0, 1, 2, or 3 days in 21% (C) or 3% O₂ (D).

Fig. 4. FBXL5 binding is not required for regulation of IRP2 by ISCs.

(A) Immunoprecipitation and immunoblots for indicated proteins of lysates derived from parental HEK-293 cell line expressing dox-inducible ISCU-DN transfected with empty vector (EV), FLAG-tagged FBXL5 (WT), or FBXL5 in which ISC-coordinating cysteine resides were mutated to alanine (ISCmut). Two days after transfection, cells are treated with 100 nM dox (1 day), and coimmunoprecipitation is performed using antibody against FLAG. (B) Immunoblots for indicated proteins of lysates derived from IRP1/2 double-knockout MDA-MB-231 cell line expressing dox-repressible WT IRP2 and ubiquitously expressed IRP2 R779E mutant, treated with 100 nM dox for 2 days and then treated with 100 nM dox, 50 μM DFO, or 35 μM iron nitrate (1 day) as indicated. (C) Immunoblots for indicated proteins of lysates derived from IRP1/2 double-knockout MDA-MB-231 cell line expressing dox-repressible WT IRP2 and ubiquitously expressed IRP2 R779E mutant, treated with 100 nM dox for 2 days and infected with shGFP or shNFS1.

Fig. 5. Suppression of ISC synthesis increases IRP2-mRNA binding independent of altered IRP2 stability at tissue O₂ levels.

(A and B) EMSA using ³²P-labeled IRE probe (A), or immunoblot for indicated proteins (B), of lysates derived from parental and IRP1 knockout HEK-293 cell line expressing dox-inducible ISCU-DN, treated with 100 nM dox (1 day), and incubated in 21% O₂, lysis in 1% O₂. (C and D) EMSA using ³²P-labeled IRE probe (C) and immunoblot for indicated proteins (D) of lysates derived from parental and IRP1 knockout MDA-MB-231 cell line expressing shGFP or shNFS1, incubated in 3% O₂, lysis in 1% O₂.

Fig. 6. IRP2 is sufficient to sense ISC deficiency and sensitize cells to ferroptosis.

(A) Proliferation (relative cell number) of IRP1/2 double-knockout MDA-MB-231 cells grown in media supplemented with 35 μM iron nitrate and 50 μM Trolox or upon withdrawal of supplements at day 0. (B) Immunoblots for indicated proteins of lysates derived from IRP1 knockout and IRP1/2 double-knockout MDA-MB-231 cell line treated with 50 μM DFO and 35 μM iron nitrate (1 day) as indicated. (B and C) Immunoblots for indicated proteins of lysates derived from parental (WT) and IRP1/2 double-knockout MDA-MB-231 cell line expressing a nontargeting shRNA (shGFP, −) or shRNA targeting NFS1 (shNFS1, +). (D) Relative abundance of TFR1 and FTH1 mRNA in parental, IRP1 knockout, and IRP1/2 double-knockout MDA-MB-231 cell line expressing shGFP or shNFS1. (E and F) Immunoblots for indicated proteins of lysates derived from IRP1/IRP2 double-knockout MDA-MB-231 cells expressing dox-repressible IRP2 treated with 100 nM dox for 2 days, and then treated with 100 nM dox, 50 μM DFO, or 35 μM iron nitrate (1 day), or infected with shGFP or shNFS1. After drug selection for shRNA transduction, cells are treated with 100 nM dox and 35uM iron nitrate for 1 day. (G) Relative viability of cells as in (E) and (F). After drug selection for shRNA transduction, cells were treated with 100 nM dox, 1 μM ferrostatin, or erastin (0, 2.5, 5, and 10 μM) for 2 days as indicated.

Fig. 7. IRP1/2 double-knockout cells expressing TFR1 can proliferate without iron supplementation and are sensitive to ferroptosis, but cannot sense ISC deficiency and enhance ferroptosis sensitivity.

(A) Proliferation (relative cell number) of MDA-MB-231 cells expressing dox-repressible IRP2, TFR1 open reading frame (TFR1 without IRE) with 100 nM dox as indicated. (B) Immunoblots of IRP1/IRP2 double-knockout MDA-MB-231 cells expressing dox-repressible IRP2, TFR1 open reading frame, and nontargeting shRNA (shGFP, −) or shRNA targeting NFS1 (shNFS1, +), treated with 100 nM dox. (C) Relative viability of cells as in (B), treated with 100 nM dox, 1 μM ferrostatin, or erastin (0, 2.5, 5, and 10 μM) for 2 days as indicated. (D) Immunoblots of MDA-MB-231 cells expressing dox-repressible IRP2, TFR1 open reading frame, and shGFP or shRNAs targeting FXN (shFXN, 1 or 2), treated with 100 nM dox. (E) Relative viability of cells as in (D), treated with 100 nM dox, 50 μM Trolox, or erastin (0, 2.5, 5, and 10 μM) for 2 days as indicated. Asterisks indicate significance comparing shGFP to shNFS1 or shFXN for each group. P < 0.05, error bars are SEM.

Fig. 8. IRP2 regulation by ISC deficiency.

Model for IRP2 activation in ISC deficiency and subsequent ferroptosis sensitivity. At atmospheric O₂ levels, the FBXL5 ISC is oxidized, and IRP2 is ubiquitinated and degraded. Partial inhibition of ISC synthesis by shNFS1 or a dominant negative ISCU (mISCU) abrogates the FBXL5:IRP2 interaction, leading to increased IRP2 stability and IRP2 binding to target mRNAs. At physiologic O₂ levels, the FBXL5 ISC is reduced, and degradation of IRP2 is limited. Partial inhibition of ISC synthesis does not further alter IRP2 stability but still releases IRP2 from FBXL5. ISC inhibition enhances IRP2 binding to target mRNAs independent of FBXL5 binding. IRP2 activation increases intracellular Fe⁺² pool and sensitize cells to ferroptosis. Deletion of both IRP1 and IRP2 is sufficient to inhibit the iron accumulation and subsequent ferroptosis sensitivity.