Cell lines and culture conditions
HEK293T, U-2 OS, NCI-H460, NCI-H441 and NCI-H1703 cells were obtained from the University of California, Berkeley (UC Berkeley) Cell Culture Facility. NCI-H520 cells were purchased from the American Type Culture Collection (HTB-182). HEK293T and U-2 OS cells were cultured in DMEM containing 4.5 g L−1 glucose and L-glutamine without sodium pyruvate (Corning, 10-013-CMR). NCI-H460, NCI-H520, NCI-H441 and NCI-H1703 cells were cultured in RPMI-1640 medium with L-glutamine (Gibco, 21875034).
To directly compare the impact of vitamin B2 on FSP1 levels in U-2 OS, cells were cultured in riboflavin (vitamin B2)-deficient medium (Gibco, 10372019) supplemented with 1× GlutaMax (Gibco, 35050079).
All media were supplemented with 10% FBS (Gemini Bio Products) and all cells were maintained at 37 °C in 5% CO2. Penicillin–streptomycin (Life Technologies, 15140122) was added to the growth medium for CRISPR–Cas9 screens and FACS. All cell lines were tested for Mycoplasma.
Generation of CRISPR–Cas9 genome-edited cell lines
The endogenous, dual-fluorescent FSP1GFP-P2A-BFP knock-in reporter cell line was generated by cotransfection of U-2 OS cells with the donor plasmid pUC57 (described below) and px330 (a gift from F. Zhang; Addgene, plasmid 42230), encoding FSP1 sgRNA guide 1 at a 2:1 w/w ratio using X-tremeGENE HP. Then, 6 h after transfection, cells were treated and maintained in medium containing 1 µM SCR7 for 1 week. GFP-P2A-BFP knock-in reporter cells were enriched by six rounds of FACS and individual clones were isolated using limiting dilution.
For the CRISPR–Cas9 genetic screens, FSP1GFP-P2A-BFP and FSP1GFP-P2A-BFP/RFKKO knock-in lines stably expressing Cas9 were generated by transduction with lentiCRISPRv2 hygro virus (a gift from B. Stringer; Addgene, plasmid 98291), encoding either SAFE-control or RFK-targeting sgRNA guides. Cells were selected and maintained in medium containing 375 µg ml−1 hygromycin. Active Cas9 expression was validated by flow cytometry analysis following infection with a self-cutting mCherry plasmid, which expresses mCherry and an sgRNA targeting the mCherry gene.
RFKKO, FLAD1KO, NFE2L1KO, NFE2L2KO, UBR4KO, KCMF1KO, CUL1KO, RNF4KO, RNF113AKO, ARIH1KO and RNF8KO lines were generated by transduction with pMCB320 virus (a gift from M. Bassik; Addgene, plasmid 89359) encoding the appropriate sgRNA guides. FSP1GFP-P2A-BFP or FSP1GFP-P2A-BFP/RFKKO knock-in lines expressing Cas9 were then selected in medium containing 1 µg ml−1 puromycin and analyzed by flow cytometry.
RFKKO and FLAD1KO lines were generated by transduction with lentiCRISPRv2 hygro virus encoding the appropriate sgRNA guides. NCI-H460, NCI-H520, NCI-H441 and NCI-H1703 lines expressing NLS–mCherry (described below) were then selected in medium containing 375 µg ml−1 hygromycin.
Lentiviral particles for transduction were generated by cotransfection of lentiCRISPRv2 plasmids with second-generation lentiviral packaging plasmids (pMD2.g and psPAX2) or pMCB320 plasmids with third-generation lentiviral packaging plasmids (pMDLg/pRRE, pRSV-Rev and pMD2.G) into HEK293T cells using TransIT-LT1 transfection reagent (Mirus) according to the manufacturer’s instructions. Lentiviral medium was collected 72 h after transfection, passed through a 0.45-µm syringe filter and used to infect cells in the presence of 8 µg ml−1 polybrene.
Generation of Dox-inducible cell lines
U-2 OS FSP1KO lines were generated as previously described16. Briefly, U-2 OS cells were transfected with px459 (Addgene, plasmid 48139) encoding FSP1 sgRNA guide 2, followed by selection in medium containing 1 µg ml−1 puromycin and isolation of individual clones using cloning rings.
Expression lines were generated by transduction of U-2 OS FSP1KO cells with pLenti CMV rtTA3 Blast (w756-1; a gift from E. Campeau; Addgene, plasmid 26429), followed by selection in medium containing 10 µg ml−1 blasticidin. FSP1KO rtTA3-expressing cells were subsequently infected with pMCB497-pTRE-FSP1-GFP-poly(A)-Blast virus containing either WT, D41A, N49A or D285N FSP1 construct (described below). FSP1–GFP-expressing cells were enriched by two rounds of FACS of the GFP-positive populations upon induction with 1 µg ml−1 Dox for 24 h.
RFKKO and FLAD1KO lines were generated by transduction with lentiCRISPRv2 hygro virus (Addgene, plasmid 98291) and selection in medium containing 375 µg ml−1 hygromycin. The RNF8KO cell line was generated by transduction with lentiCRISPRv2 puro (Addgene, plasmid 98290) virus and selection in medium containing 1 µg ml−1 puromycin.
RNF8 S-tag rescue lines (described below) were generated by transduction with pLenti CMV/TO Zeo DEST (644-1; Addgene, plasmid 17294) virus containing either WT or C406S RNF8 and selected in medium containing 250 µg ml−1 zeocin.
All lentiviral particles for transduction were generated by cotransfection with third-generation lentiviral packaging plasmids (pMDLg/pRRE, pRSV-Rev and pMD2.G) into HEK293T cells, unless otherwise described.
Plasmids and siRNA
Cloning of all expression plasmids and the GFP-P2A-BFP donor plasmid was performed using restriction-enzyme-independent fragment insertion by megaprimer cloning.
To generate the FSP1GFP-P2A-BFP knock-in donor plasmid, 800-bp homology arms flanking the FSP1 stop codon were amplified from U2OS genomic DNA and inserted into pUC57 plasmid (a gift from R. Tjian, UC Berkeley). To introduce the GFP-P2A-BFP insert, primers containing 15-bp overlaps with each of the FSP1 homology arms were used to amplify a codon-optimized GFP-P2A-BFP gBlock (Integrated DNA Technologies). This insert was then cloned in frame of the FSP1 stop codon within the pUC57 plasmid containing the FSP1 homology arms. The protospacer-adjacent motif site that corresponds to FSP1 sgRNA guide 1 was subsequently mutated in the donor sequence using site-directed mutagenesis primers without changing the encoded amino acids to prevent cutting of the integrated donor sequence by Cas9.
WT FSP1 was cloned and inserted in frame 5′ to 3′ to replace the Insig1 gene sequence in pMCB497-pTRE-Insig1-GFP-poly(A)-Blast plasmid (a gift from R. Kopito, Stanford University). D41A, N49A and D285N mutant FSP1–GFP was subsequently generated using site-directed mutagenesis. pDONR221 RNF8 (DNASU, plasmid HsCD00042754) was modified by insertion of a C-terminal S-tag and stop codon at the C terminus of RNF8. RNF8 C406S S-tag was subsequently generated using site-directed mutagenesis. All entry plasmids were cloned into pLenti CMV/TO Zeo DEST by Gateway recombination cloning (Thermo Fisher Scientific).
mCherry-NLS-P2A-bleoR was modified from CSII-prEF1a-mCherry-3×NLS (a gift from J. Skotheim; Addgene, plasmid 125262). Briefly, a P2A sequence and the bleomycin resistance gene were inserted after mCherry–3×NLS in the original plasmid using the NEBuilder Hifi DNA assembly master mix (New England Biolabs, E2621S).
For in vitro protein translation, WT and D41A, N49A and D285N mutant FSP1 were cloned and inserted in frame 5′ to 3′ to replace the PURExpress DHFR control gene sequence (New England Biolabs, E6800). To introduce the C-terminal Strep tag insert, primers containing 15-bp overlaps between the stop codon of WT FSP1 were used to amplify a codon-optimized STREP gBlock (Integrated DNA Technologies).
For recombinant protein expression, WT and D285N mutant FSP1 lacking the ATG start codon were inserted into the pET-His6-Tev vector (a gift from S. Gradia; Addgene, plasmid 29653) C-terminal to the His6–TEV tag.
CRISPR sgRNA sequences targeting FSP1 were designed using the CRISPR guide design tool by Benchling (https://www.benchling.com). sgRNAs targeting RFK, FLAD1, NFE2L1, NFE2L2, UBR4, KCMF1, RNF4, CUL1, RNF113A, ARIH1, RNF8 and SAFE-control were selected on the basis of the effective score from the human CRISPR–Cas9 KO libraries used for genetic screening. The oligonucleotide sequences preceding the protospacer motif were as follows: sgFSP1-1, 5′-TGAGGCAGTCTCCACCTTGA-3′; sgFSP1-2, 5′-GAATCGGGAGCTCTGCACG-3′; sgRFK, 5′-GGTCTCAGGTAGCCAACAA-3′; sgFLAD1, 5′-GAGTGGAACTGAACAGCT-3′; sgNFE2L1, 5′-GTCCAGGTCTATGCTCTT-3′; sgNFE2L2, 5′-GTAGCCCCTGTTGATTTAGA-3′; sgUBR4, 5′-GTGCACAGGCCCTTGATT-3′; sgKCMF1, 5′-GTGCATCTTGTTATGAAAG-3′; sgRNF4, 5′-GGGCTCTGGCAGTGACTAGG-3′; sgCUL1, 5′-GGAATTATATAAACGACTTA-3′; sgRNF113A, 5′-GCTTTCTCCAGGAAAGGCGG-3′; sgARIH1, 5′-GGAGGAGGATTACCGCTACG-3′; sgRNF8, 5′-GACTGTAGGACGAGGATT-3′; sgSAFE control, 5′-AAATTTCATGGGAAAATAG-3′. Guide sequences were cloned into the BbsI restriction sites of vector px459 or px330, into the BsmBI restriction sites of lentiCRISPRv2 or between the BstXI and BlpI restriction sites of pMCB320.
The siRNA oligo smart pools targeting FSP1, RFK and FLAD1, as well as a nontargeting control (siCNTL), were purchased from Horizon Discovery and cells were transfected with the RNAiMAX reagent (Invitrogen) according to the manufacturer’s instructions.
Genomic DNA sequencing and indel analysis
Primers were designed to flank either the stop codon or predicted Cas9 cut site with the forward primer ~250 bp upstream of the cut site.
For genomic DNA extraction, cells were washed twice and scraped off plates in cold DPBS and pelleted at 500g for 5 min. Genomic DNA was extracted from cell pellets and purified using the Qiagen blood mini kit (Qiagen, 51104) according to the manufacturer’s protocol.
PCR was performed on an Applied Biosystems thermal cycler using Q5 high-fidelity 2× master mix (New England Biolabs, M0492S). The TIDE oligonucleotide sequences used were as follows: Fwd-FSP1-Cterm, 5′-CCGAGGCGGGCAGATCACCT-3′; Rev-FSP1-Cterm, 5′-GGCTCCCCG-ACCGTGTGATGG-3′; Fwd-RNF8-Exon2, 5′-CTACTTGGTGGTTCTCAAGACAGC-3′; Rev-RNF8-Exon2, 5′-GTCTCTGGGAGCTTCACCTCA-3′.
Amplicons were separated on 2% agarose–TAE gels and purified using the QIAquick gel extraction kit (Qiagen, 28706). Amplicons were sequenced at QuintaraBio and sequences were assessed for indels using the TIDE analysis tool (http://shinyapps.datacurators.nl/tide/).
Chemicals and reagents
We purchased 1S,3R-RSL3 (19288), FSEN1 (38025), Fer1 (17729), idebenone (15475), DFO (14595), ZVAD(OMe)-FMK (27421) and necrostatin 1 (11658) from Cayman Chemical. 3-Methyladenine (189490), Baf-A1, Streptomyces griseus (196000), BSA (A8806), oleic acid (O1383), trichloroacetic acid (T6399), potassium hydroxide (221473), (−)-riboflavin (R9504), riboflavin 5′-monophosphate sodium salt hydrate (F8399), FAD disodium salt hydrate (F6625), emetine dihydrochloride hydrate (E2375), cycloheximide (239763-M) and CoQ1 (C7956) were all purchased from Sigma-Aldrich. Blasticidin S HCl (A1113903), puromycin (A1113803), hygromycin B (10687010), zeocin (R25001), BODIPY 581/591 C11 (D3861), CellMask deep red plasma membrane stain (C10046) and SYTOX green dead cell stain (34860) were all purchased from Thermo Fisher Scientific. LipiBlue (LD01) was purchased from Dojindo Laboratories. SCR7 (S7742), CB5083 (S8101), MLN7243 (S8341) and MG132 (S2619) were purchased from Selleck Chemicals. NADH (481913) and Dox hyclate (D9891) were purchased from Millipore Sigma. Transfection reagents used in this study include: Lipofectamine RNAiMAX transfection reagent (13778150, Thermo Fisher Scientific), Polybrene (107689, Millipore Sigma), X-tremeGENE HP DNA transfection reagent (6366244001, Roche) and TransIT-LT1 transfection reagent (MIR2300, Mirus).
Western blotting
Cells were washed twice with PBS, lysed in 1% SDS, sonicated for 20 s and incubated for 5 min at 100 °C. The cell lysate was then centrifuged for 10 min at 15,000g to remove any cell debris. Protein concentrations were determined using the BCA protein assay kit (Thermo Fisher Scientific) and equal amounts of protein by weight were combined with Laemmli buffer, boiled for 5 min at 100 °C, separated on 4–20% polyacrylamide gradient gels (Bio-Rad Laboratories) and transferred onto nitrocellulose membranes (Bio-Rad Laboratories). Membranes were blocked in PBS with 0.1% Tween-20 (PBSt) containing 5% (w/v) dried milk for 60 min. Membranes were washed twice with water and incubated overnight at 4 °C in Tris-buffered saline with 0.1% Tween-20 (TBSt) containing 3% BSA (Sigma-Aldrich) and primary antibodies. After washing with PBSt, membranes were incubated at room temperature for 60 min in PBSt containing fluorescent secondary antibodies.
For ubiquitination blots, proteins were separated on 4–20% polyacrylamide gradient gels (Bio-Rad Laboratories) and transferred onto PVDF membranes (Bio-Rad Laboratories). Membranes were blocked in TBS containing 3% (w/v) BSA for 60 min. Membranes were then incubated overnight at 4 °C in TBSt containing 3% BSA and primary antibodies. After washing with TBSt, membranes were incubated at room temperature for 60 min in TBSt containing fluorescent secondary antibodies. All immunoblots were imaged on a LI-COR imager (LI-COR Biosciences).
The following blotting reagents and antibodies were used: anti-FSP1 (Santa Cruz, sc-377120), anti-GFP (Sigma-Aldrich, SAB1305545), GAPDH (Cell Signaling, D4C6R), anti-PLIN2 (Abcepta, AP5118c), anti-Derlin2 (a kind gift from Y. Ye, National Institutes of Health), anti-NFE2L2 (Santa Cruz, sc-365949) anti-RFK (Proteintech, 15813-1-AP), anti-FLAD1 (Santa Cruz, sc-376819), anti-AIFM1 (Santa Cruz, sc-13116), anti-NQO1 (Proteintech, 11451-1-AP), anti-DHCR24 (Proteintech, 10471-1-AP), anti-CYB5R3 (Santa Cruz, sc-398043), anti-GSR (Santa Cruz, sc-133245), anti-TXNRD1 (Santa Cruz, sc-28321), anti-SQLE (Santa Cruz, sc-271651), anti-DHODH (Santa Cruz, sc-166348), anti-S-Tag (Novus Biologicals, NB600-511) and anti-Ub (Enzo Life Sciences, BML-PW0930).
Flow cytometry
For flow cytometry analysis, U-2 OS cells were seeded in six-well plates. When subjected to treatments, FSP1GFP-P2A-BFP cells were treated the next day with 5 mM 3-MA, 250 nM Baf-A1, 5 µM CB5083, 10 µM MLN7243 and 10 µM MG132 for 6 h and/or varying concentrations of RSL3 for 24 h, while FSP1KO cells expressing an inducible, GFP-tagged FSP1 construct were treated with 1 µg ml−1 Dox and/or 1 µM MG132 for 24 h. For cells supplemented with exogenous flavin metabolites (vitamin B2, FMN or FAD), FSP1KO cells expressing an inducible, GFP-tagged FSP1 construct were grown in vitamin B2-depleted medium supplemented with 50 µM of each indicated flavin metabolite in the presence of 1% DMSO58 for a total of 72 h, where FSP1 expression was induced with 1 µg ml−1 Dox 24 h before harvesting. Cells were dissociated from plates using TrypLE Express (Gibco) and resuspended in DMEM containing 10% FBS. Cells were then pelleted by centrifugation at 500g for 5 min, resuspended in DPBS and placed on ice.
For all flow cytometry assays, fluorescence was analyzed using an LSR Fortessa (BD Biosciences) collecting at least 10,000 events per sample. The following filter sets were used: FITC (GFP, BODIPY C11ox), Pacific blue (BFP) and Texas red (mCherry, BODIPY C11non-ox). Data were analyzed using FlowJo version 10.10.0 (FlowJo).
Fluorescence microscopy
For super-resolution microscopy of live cells, U-2 OS FSP1GFP-P2A-BFP knock-in cells were seeded in 35-mm imaging dishes (CellVis, D35-14-1.5-N) 48 h before the start of imaging. The next day, cells were incubated with 200 µM oleate for 24 h. Lipid droplets were stained with 500 nM LipiBlue for 30 min and the plasma membrane was stained with 5 µg ml−1 CellMask deep red for 10 min before imaging. Lattice-SIM images were acquired on a Zeiss Elyra7 super-resolution fluorescence microscope, equipped with dual scientific complementary metal–oxide–semiconductor PCO Edge 4.2 cameras for simultaneous two-channel acquisition, with a ×63 (numerical aperture: 1.4) oil objective. For each focal plane, 13 phase images were acquired. Lattice structured illumination microscopy (SIM) reconstruction was performed with the SIM processing tool of ZEN 3.0 SR (Black, version 16) software.
Lipid droplet fractionation
Two 500-cm2 plates of U-2 OS FSP1GFP-P2A-BFP knock-in cells were incubated in 200 µM oleate–BSA complex for 24 h. Cells were collected by scraping into PBS and centrifuged for 10 min at 500g. Cell pellets were resuspended in cold hypotonic lysis medium (HLM; 20 mM Tris-HCl pH 7.4 and 1 mM EDTA) supplemented with 1× cOmplete mini EDTA-free protease inhibitor cocktail (Sigma-Aldrich), incubated on ice for 10 min, lysed using 20 passes through a syringe affixed to a 23G needle and centrifuged at 1,000g for 10 min. The supernatant was subsequently transferred to Ultra-Clear ultracentrifuge tubes (Beckman-Coulter), diluted with 60% sucrose in HLM to a final concentration of 20% sucrose in HLM and overlaid with 4 ml of 5% sucrose in HLM buffer followed by 4 ml of HLM buffer. Overlaid samples were centrifuged for 30 min at 15,000g in an ultracentrifuge using a SW41 swinging bucket rotor. Buoyant fractions were collected using a tube slicer (Beckman-Coulter), additional fractions were pipetted from the top of the sucrose gradient in 1-ml increments and pellets were resuspended in 1 ml of HLM. Next, 100 µl of 10% SDS was added to each fraction, yielding a final concentration of 1% SDS. Samples were then sonicated for 15 s and incubated for 10 min at 65 °C. Buoyant fractions were incubated at 37 °C for 1 h and sonicated every 20 min, followed by a final incubation at 65 °C for 10 min.
Cell viability and death analysis
U-2 OS cells were plated in duplicate or triplicate at a density of 2,000 cells per well in a black 96-well plates (Corning, 3904) 48 h before imaging. To induce expression of FSP1, cells were treated with 1 µg ml−1 Dox the next day. For exogenous supplementation of flavin metabolites, cells were first grown in vitamin-B2-depleted medium supplemented with 50 µM of each indicated flavin metabolite in the presence of 1% DMSO for a total of 72 h and then plated as stated above. After 48 h, the medium was replaced with fresh medium containing 30 nM SYTOX green dead cell stain, Dox (if needed) and the indicated drugs at different doses. The plates were immediately transferred to an IncuCyte S3 live-cell imaging system (Sartorius) enclosed in an incubator set to 37 °C and 5% CO2. Three images per well were captured in the green, red and phase channels every 3 h over a 24 h period.
NCI-H460, NCI-H520, NCI-H441 and NCI-H1703 cells were plated in triplicate at a density of 1,500, 5,000, 2,500 and 2,000 cells per well, respectively, 24 h before imaging and assessed as described above.
The ratio of SYTOX green-positive objects (dead cells) to phase objects (total cells) or mCherry-positive objects (live cells) was quantified using the Sartorius IncuCyte 2020A image analysis software. For each treatment condition, the SYTOX-to-mCherry-object ratio (WT, FSP1GFP-P2A-BFP or lung cancer lines) or SYTOX-to-phase-object ratio (FSP1–GFP expression lines) was plotted against the 24-h imaging interval and the average area under the curve was plotted as a function of drug concentration (for example, RSL3) using GraphPad Prism version 10. To compare cell death between FSP1–GFP expression lines (WT, RFKKO, FLAD1KO and no-Dox control), the SYTOX-object counts at the 24-h time point from 1 µM RSL3 treatment were normalized by the maximum value for each cell line (that is, the phase-object counts at the 0-h time point).
Genome-wide U-2 OS FSP1GFP-P2A-BFP CRISPR–Cas9 screen
Genome-wide CRISPR–Cas9 screens were performed using the Bassik human CRISPR KO library (Addgene, pooled libraries 101926–101934). The library consists of nine sublibraries, comprising a total of 225,171 elements, including 212,821 sgRNAs targeting 20,549 genes (~10 sgRNAs per gene) and 12,350 negative control sgRNAs. To generate lentiviral particles, each sublibrary was cotransfected with third-generation lentiviral packaging plasmids into HEK293T cells. Medium containing lentivirus was collected 48 and 72 h after transfection, combined and filtered. U-2 OS FSP1GFP-P2A-BFP cells stably expressing Cas9 were transduced with lentiviral packaged sublibraries (one sublibrary at a time) in 8 µg ml−1 polybrene. After 72 h of growth, infected cells were selected in medium containing 1 µg ml−1 puromycin until over 90% of cells were mCherry positive (according to flow cytometry). Cells were then recovered for 3–5 days in medium lacking puromycin and frozen in liquid nitrogen.
For the screen, library infected cells were thawed (one sublibrary at a time) and expanded at 1,000× coverage (1,000 cells per library element). On the day of the sort, cells were dissociated using 0.25% trypsin–EDTA (Gibco), collected by centrifugation at 400g for 5 min and washed once with DPBS. Cells were resuspended in phenol-red-free medium (HyClone, 16777-406) supplemented with 3% FBS and 1% BSA (fatty acid free) and kept on ice until FACS.
Cells were sorted on a BD Aria Fusion equipped with four lasers (488 nm, 405 nm, 561 nm and 640 nm). sgRNA-expressing mCherry+ cells were then sorted into the brightest and dimmest 30% GFP+ populations in 15-ml conical flasks containing DMEM with 4.5 g L−1 glucose and L-glutamine supplemented with 10% FBS. For each sublibrary sort, at least 1,000× as many cells as guides were collected. Sorted cells were collected by centrifugation at 1,000g for 10 min and washed once with DPBS; pellets were frozen at −80 °C until genomic DNA extractions. PCR amplicon preparation and deep sequencing were performed according to Mathiowetz et al.66.
Secondary pooled (batch retest) screens using a custom sgRNA library
The custom sgRNA batch retest library contains 9,350 elements with 7,350 sgRNAs targeting 735 genes (~10 sgRNAs per gene) and 2,000 negative control sgRNAs. Guide sequences were from the Bassik human CRISPR KO library and the library construction protocol was previously described32,33.
U-2 OS FSP1GFP-P2A-BFP cells stably expressing Cas9 were transduced with the packaged library as described above. For each screen, cells were thawed and expanded at >1,000× coverage. For all screens, cells were seeded into 500-cm2 plates at 1,000-fold library coverage. For inducing lipid peroxidation, cells were treated the following day with 100 nM RSL3 for 24 h. Cells were screened by FACS as described above with minor modifications. Briefly, sgRNA-expressing mCherry+ cells were sorted as follows: GFP screen (brightest 30% GFP+ and dimmest 30% GFP+), BFP screens (brightest 30% BFP+ and dimmest 30% BFP+) and GFP:BFP screens (brightest 30% GFP:BFP+ ratio and dimmest 30% GFP:BFP+ ratio).
Sensitized degradation screen with U-2 OS FSP1GFP-P2A-BFP/RFKKO reporter cell line
The custom UBAL degradation library contains 20,710 elements with 18,710 sgRNAs targeting 1,871 genes (~10 sgRNAs per gene) and 2,000 negative control sgRNAs as previously described59.
U-2 OS FSP1GFP-P2A-BFP/RFKKO cells stably expressing Cas9 were transduced with the packaged library as described above. Cells were thawed and maintained at >1,000× coverage. For all screens, cells were seeded into 500-cm2 plates at 1,000-fold library coverage. sgRNA-expressing, mCherry+ cells were then sorted to the brightest and dimmest 30% GFP:BFP+ ratio by FACS.
CRISPR–Cas9 screen data analysis
Sequence reads were aligned to the sgRNA reference library using Bowtie 2 software. For each gene, a gene effect and score (likely maximum effect size and score) and P values were calculated using the casTLE statistical framework as previously described32,33.
Functional interactions and protein–protein interactions for high-confidence candidate regulators were identified using the GO enrichment analysis and STRING interaction network database.
FENIX assay
Egg phosphatidylcholine liposomes (extruded to 100 nm, diluted to 1.03 mM) in cPBS (pH 7.4, 12 mM, Chelex-100 treated) were vortexed with STY-BODIPY (1.03 µM in DMSO). This solution was aliquoted into a black 96-well plate containing each test compound followed by manual mixing. DTUN was added to the wells for initiation as the last step, followed by mixing. The plate was then incubated at 37 °C for 3 min followed by mixing for 3 min before data were acquired by fluorescence (λex = 488 nm, λem = 518 nm). From PMC-inhibited reactions, Ri was determined to be (6.4 ± 0.1) × 10−10 M s−1 under these exact conditions67.
To reduce riboflavin, DMSO was added to a vial containing 3.32 mM riboflavin and sonicated at 37 °C until full dissolution. The solution was sparged with N2 and kept under an inert atmosphere protected from light. The solution was subsequently diluted to 1.91 mM with cPBS that contained three equivalents of sodium dithionite (sparged with N2), then serially diluted immediately and added to the plate before the addition of DTUN.
FAD measurements
A total of 1 × 106 FSP1GFP-P2A-BFP or FSP1KO cells expressing the indicated FSP1–GFP construct and/or sgRNA were seeded into 10-cm2 plates. The following day, FSP1KO cells were treated with medium containing 1 µg ml−1 Dox for 24 h and/or 50 µM of each indicated exogenous flavin metabolite in the presence of 1% DMSO when necessary. Cells were washed twice with cold PBS and collected by scraping. For recombinant FSP1 proteins, 10 µg of WT or D285N variants were resuspended in 100 µl of cold FAD assay buffer. All samples were prepared for FAD fluorescence measurements using the FAD assay kit (Abcam, ab204710) according to the manufacturer’s protocol. FAD concentrations were calculated using a standard curve and normalized to the total cell lysate (500 µg) or total protein level (10 µg) in each sample. Four biological replicates were performed for each condition.
Emetine chase
WT, FSP1GFP-P2A-BFP or Dox-induced FSP1KO cells expressing FSP1–GFP in the indicated sgRNA background were washed once with PBS, treated with 75 µM emetine for the indicated times, harvested by centrifuging at 1,500g for 5 min and analyzed by immunoblotting.
Transcriptional shutoff ‘chase’ or protein turnover assays
U-2 OS FSP1KO cells expressing the indicated FSP1–GFP construct and/or sgRNA were seeded into six-well plates. The following day, cells were treated with medium containing 1 µg ml−1 Dox for 12 h to induce expression of FSP1–GFP. Cells were then washed 1× with DPBS and replenished with complete medium to initiate Dox washout and measure reporter protein turnover at the indicated times. MG132 controls were treated with 1 µM MG132 during both Dox induction and washout.
For flow cytometry analysis, the median fluorescence intensity (MFI) was calculated using FlowJo version 10. Cells without Dox were run as controls for background fluorescence and the average MFI for the no-Dox controls was subtracted from the MFI of the Dox-positive samples at all time points. The GFP percentage remaining was calculated as the MFI at t = X divided by the MFI at t = 0 h. One-phase decay curves were drawn using GraphPad Prism version 10 using the default settings.
In vitro protein translation
All samples were prepared using the PURExpress in vitro protein synthesis kit (New England Biolabs, E6800) according to the manufacturer’s protocol with minor modifications. Then, 250 ng of FSP1 variants (WT, D41A, N49A or D285N) or WT FSP1–Strep template DNA, 20 U of RNase inhibitor murine (New England Biolabs, M0314) and 10 µM of exogenous flavin metabolites (vitamin B2, FMN or FAD) were used in each sample as indicated and incubated at 37 °C for 4 h. To test the incorporation of FAD into FSP1 upon translation, 5 µM of cycloheximide was used to stop the translation reaction. For Strep-tag enrichment, samples were diluted to 500 µl in 50 mM Tris-HCl pH 8.0, incubated with 10 µl of Strep-Tactin XT 4Flow resin (IBA Lifesciences, 2-5010-002) for 30 min at 4 °C, washed with 1 ml of 50 mM Tris-HCl pH 8.0, 100 mM NaCl and 0.5 mM EDTA, eluted with 50 µl of wash buffer containing 50 mM biotin and buffer-exchanged using Pierce protein concentrators (3-kDa molecular weight cutoff (MWCO)) into 50 mM NaH2PO4 pH 7.4. Two biological replicates were performed for each reaction.
Protein purification
WT and D285N FSP1 constructs were expressed in LOBSTR-BL21(DE3)-RIL Escherichia coli cells (Kerafast, EC1002) in Luria–Bertani medium supplemented with appropriate antibiotics at 37 °C to an optical density at 600 nm of ~0.5–0.6 and transferred to 18 °C for 16 h after induction with 0.5 mM IPTG. Bacterial pellets were harvested, flash-frozen, resuspended in lysis buffer A (50 mM NaH2PO4 pH 8.0, 300 mM NaCl, 10 mM MgCl2, 10 mM imidazole, 1 mg ml−1 lysozyme and 1 mM PMSF), lysed by sonication and clarified by centrifugation at 60,000g for 30 min at 4 °C.
All subsequent purification steps were carried out at 4 °C. The clarified lysate was passed through an Econo-Column chromatography column (Bio-Rad, 7371512) packed and equilibrated with HisPur Ni-NTA agarose resin (Thermo Fisher Scientific, 88221). Bound proteins were washed four times with EQ buffer (50 mM NaH2PO4 pH 8.0, 200 mM NaCl, 10 mM MgCl2 and 10 mM imidazole) and eluted using EQ buffer containing 250 mM imidazole. The eluate was concentrated and further purified on a HiLoad 16/600 Superdex 75 Pg gel-filtration column in size-exclusion chromatography buffer (50 mM NaH2PO4 pH 7.4, 150 mM NaCl and 2 mM MgCl2). After the final chromatography step, all proteins were concentrated in Amicon Ultra spin concentrators (30-kDa MWCO), buffer-exchanged into storage buffer (50 mM NaH2PO4 pH 7.4, 150 mM NaCl and 0.5 mM TCEP), aliquoted, flash-frozen in liquid N2 and stored at −80 °C. Concentrations of all proteins were determined by Bradford assay.
NADH oxidation (FSP1 activity) assay
To measure NADH consumption, 2.5 pmol of immunopurified FSP1–GFP isolated from indicated cell lines or recombinant WT or D285N mutant FSP1 at the indicated concentrations were combined with 500 µM NADH and 400 µM CoQ1 in a total volume of 200 µl in 50 mM sodium phosphate pH 7.4. A reduction in absorbance at 340 nm, corresponding to NADH oxidation, was determined over the course of 3 h at 37 °C. All measurements were taken using a Tecan Spark multimode microplate reader (Tecan).
FAD absorption spectroscopy
To assess the presence of protein-bound FAD, 100 µg of recombinant WT or D285N mutant FSP1 protein or 10 µg of free FAD control was resuspended in a total volume of 200 µl in 50 mM sodium phosphate pH 7.4. Absorption spectra were acquired using a Tecan Spark multimode microplate reader (Tecan). All spectra were acquired over a range of 350 to 500 nm with a wavelength step size of 2 nm, manual gain of 100 and 30 flashes at 25 °C. The relative absorption measured was normalized by setting the maximum value to 1 of each given sample.
CD spectroscopy
CD spectra were collected on a Jasco J-1500 spectrophotometer. For spectra, samples were approximately 7.2 µM protein in 10 mM sodium phosphate pH 7.4, 150 mM NaCl and 0.1 mM TCEP in a 1-mm quartz cuvette at 20 °C. Far-ultraviolet (UV) CD spectra were acquired from 300 to 200 nm with a scan rate of 10 nm min−1, digital integration time of 16 s and 1-nm bandwidth. Three spectra were acquired with each sample.
Protein thermal shift (Sypro orange) assay
To measure thermal stability, 5 pmol of immunopurified FSP1–GFP isolated from indicated cell lines or 100 µg of recombinant WT or D285N mutant FSP1 were combined with 10× SYPRO orange protein gel stain in a total volume of 25 µl in 50 mM sodium phosphate pH 7.4. An increase in relative fluorescence, corresponding to protein thermal unfolding, was determined over a temperature range of 20 °C to 80 °C in increments of 0.5 °C for 10 s. All measurements were taken using the CFX96 Touch real-time PCR detection system (Bio-Rad laboratories). The relative fluorescence units measured were normalized by setting the maximum value to 1 of each given sample. Three biological replicates were performed for each condition.
Immunoprecipitations
U-2 OS FSP1KO cells expressing the indicated FSP1–GFP construct and/or sgRNA were seeded into 10-cm2 plates. The following day, cells were treated with medium containing 1 µg ml−1 Dox and/or 1 µM MG132 for 24 h. Cells were washed twice with cold PBS, collected by scraping and spun down at 1,000g for 10 min. Cell pellets were resuspended in 100 µl of lysis buffer C (50 mM Tris-HCl pH 7.5, 150 mM NaCl, 0.5 mM EDTA, 0.5% IGEPAL 630 and 2 mM N-ethylmaleimide supplemented with 1× cOmplete mini EDTA-free protease inhibitor cocktail). Lysates were sonicated for 30 s and cleared by centrifuging at 20,000g for 15 min. Protein concentrations were determined using BCA protein assay.
For immunopurification of FSP1–GFP proteins, 5 mg of precleared lysates were incubated with 50 µl of ChromoTek GFP-Trap magnetic agarose (Proteintech, gtma) at 4 °C for 1 h with end-over-end rotation. Beads were washed three times with PBS and 5 µl of bound proteins were eluted by adding 2× Laemmli buffer and incubating at 100 °C for 5 min. Immunopurified FSP1–GFP concentrations were calculated using a standard curve of recombinant FSP1 and normalized by densitometry using the anti-FSP1 antibody.
For immunoprecipitation of FSP1 and RNF8 interactions, 1 mg of precleared lysates were incubated with 10 µl of ChromoTek GFP-Trap magnetic agarose or 10 µl of S-protein agarose (Millipore Sigma, 69704) at 4 °C for 1 h with end-over-end rotation. Beads were washed three times with lysis buffer C and bound proteins were eluted by adding 2× Laemmli buffer and incubating at 100 °C for 5 min.
Statistics and reproducibility
All data shown are the mean ± s.e.m. or median ± s.e.m. and the numbers (n) in each figure legend represent two independent biological replicates unless stated otherwise. For comparison across multiple experimental groups, P values were calculated using a two-way analysis of variance and adjusted using the Bonferroni correction for multiple comparisons (****P < 0.0001, ***P < 0.001, **P < 0.002 and *P < 0.033) using GraphPad Prism version 10. No statistical method was used to predetermine sample size. No data were excluded from the analyses. The experiments were not randomized. The investigators were not blinded to allocation during experiments or outcome assessment.
Materials availability
All unique/stable reagents generated in this study are available from the lead contact with a completed materials transfer agreement.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
