Skip to main page content
U.S. flag
See More

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2021 Apr 6;12(1):2048.
doi: 10.1038/s41467-021-22336-4.

Ferroptosis response segregates small cell lung cancer (SCLC) neuroendocrine subtypes

Christina M Bebber  1   2   3 , Emily S Thomas  1   2   4 , Jenny Stroh  1   2 , Zhiyi Chen  1   2 , Ariadne Androulidaki  1   2 , Anna Schmitt  2   3 , Michaela N Höhne  2   5 , Lukas Stüker  1   2 , Cleidson de Pádua Alves  1 , Armin Khonsari  1   6   7 , Marcel A Dammert  1   6   7 , Fatma Parmaksiz  1   6   7 , Hannah L Tumbrink  1   6   7 , Filippo Beleggia  3 , Martin L Sos  1   6   7 , Jan Riemer  2   5 , Julie George  1 , Susanne Brodesser  2 , Roman K Thomas  1   6   8 , H Christian Reinhardt  9 , Silvia von Karstedt  10   11   12
Affiliations

Ferroptosis response segregates small cell lung cancer (SCLC) neuroendocrine subtypes

Christina M Bebber et al. Nat Commun. .

Abstract

Loss of TP53 and RB1 in treatment-naïve small cell lung cancer (SCLC) suggests selective pressure to inactivate cell death pathways prior to therapy. Yet, which of these pathways remain available in treatment-naïve SCLC is unknown. Here, through systemic analysis of cell death pathway availability in treatment-naïve SCLC, we identify non-neuroendocrine (NE) SCLC to be vulnerable to ferroptosis through subtype-specific lipidome remodeling. While NE SCLC is ferroptosis resistant, it acquires selective addiction to the TRX anti-oxidant pathway. In experimental settings of non-NE/NE intratumoral heterogeneity, non-NE or NE populations are selectively depleted by ferroptosis or TRX pathway inhibition, respectively. Preventing subtype plasticity observed under single pathway targeting, combined treatment kills established non-NE and NE tumors in xenografts, genetically engineered mouse models of SCLC and patient-derived cells, and identifies a patient subset with drastically improved overall survival. These findings reveal cell death pathway mining as a means to identify rational combination therapies for SCLC.

PubMed Disclaimer

Conflict of interest statement

H.C.R. received consulting fees from Abbvie, AstraZeneca, Vertex and Merck and research funding from Gilead. R.K.T. is a co-founder of and was a consultant for NEO New Oncology, now part of Siemens Healthcare. R.K.T. and M.L.S. are co-founders of and consultants for PearlRiver Bio GmbH and M.L.S. receives funding from PearlRiver Bio GmbH. H.L.T. is a consultant for PearlRiver Bio GmbH. R.K.T. is a co-founder and consultant of Epiphanes Inc. R.K.T. is a stockholder of Roche, AstraZeneca, GSK, Merck, Qiagen, Novartis, Bayer and Johnson & Johnson. All other authors declare no competing interests.

Figures

Fig. 1
Fig. 1. Regulated cell death pathways are counter-selected in treatment-naïve SCLC.
a Schematic view of genes involved in extrinsic apoptosis, b necroptosis and c ferroptosis. d, e, g, h, j, k RNA-seq expression data in FPKM (fragments per kilobase of exon model per million reads mapped) from normal lung (n = 22) and mostly chemo-naïve SCLC patient samples (n = 67) were log2 transformed (+0.01) and plotted for relative expression of genes involved in extrinsic apoptosis (d, e) necroptosis (g, h) and ferroptosis (j, k) boxplot center line, mean; box limits, upper and lower quartile; whiskers min. to max. Heatmap color code indicates expression levels between each sample and the average of each gene, dendrogram shows the distance between sample populations. f The indicated seven human SCLC cell lines (n = 7) and NSCLC cell line (H460) were treated with human izTRAIL for 24 h, cell viability was determined by Cell Titer Blue, n (H82) = 6, n (L88) = 4, n (H524) = 4, n (H889) = 4, n (L303) = 3, n (H526) = 3, n (H69) = 5, n (H460) = 3. i Cells were treated with TNF (T) [10 ng/ml]/zVAD (Z) [20 µM]/Smac mimetic (S) Birinapant [1 µM] +/− Nec-1 [10 µM] for 24 h, cell death was quantified by propidium iodide (PI) uptake and flow cytometry. l Cells were treated with erastin at indicated concentrations for 24 h, cell viability was determined by Cell Titer Blue. Data are means ± SEM of three independent experiments wherever not indicated otherwise. Two-way ANOVA + Tukey’s multiple comparison test (i) and two-tailed unpaired t tests for all others, ****p < 0.0001, *p < 0.05, ns p > 0.05. Source data are provided as Source data file.
Fig. 2
Fig. 2. GPX4 expression is sufficient to protect SCLC from ferroptosis.
a The indicated murine SCLC lines (n = 3) were treated either with DMSO, RSL3 [1 µM], erastin [10 µM] alone or in combination with Ferrostatin-1 (Fer-1) [5 µM] for 24 h. Cells were washed, cultured for 5 days for recovery and stained with crystal violet. b–e Human H82 cells were treated as indicated: erastin [10 µM] ± Ferrostatin-1 (Fer-1, 5 µM)  ± deferoxamine (DFO) [100 µM], RSL3 [1 µM] ± Fer-1 ± DFO for 24 h. Cell death was determined by propidium iodide (PI) uptake and flow cytometry. f RP285.5 murine SCLC cells stably expressing Cas9 and the indicated control or gRNAs targeting GPX4 were cultured with or without Ferrostatin-1 (Fer-1) for 24 h. Cell death was quantified by propidium iodide (PI) uptake and flow cytometry. g Protein extracts were obtained from cells as in (f) cultured in the presence of Fer-1 [5 µM]. A representative western blot is shown (h, i) RP285.5 murine SCLC cells as in (f) were cultured in the presence or withdrawal of Fer-1 [5 µM] for 5 h and stained for lipid ROS accumulation using BODIPY C11. Cells were analyzed by flow cytometry and mean fluorescent intensity (MFI) was quantified. Data are means ± SEM of three independent experiments in each individual cell line or representative images or histograms were applicable. One-way ANOVA + Tukey’s multiple comparison test (b–e), Two-way ANOVA + Tukey’s multiple comparison test (f), ****p < 0.0001, ***p < 0.001, **p < 0.01, *p < 0.05. Source data are provided as Source data file.
Fig. 3
Fig. 3. SCLC neuroendocrine subtypes segregate by ferroptosis response.
a, b Eight human SCLC cell lines were treated as indicated for 24 h, cell viability was determined by Cell Titer Blue. c The indicated human SCLC cells were treated with DMSO, RSL3 [100 nM] or RSL3/Fer-1 [5 µM] for 5 h and stained for lipid ROS accumulation using BODIPY C11. Cells were analyzed by flow cytometry. Gray = DMSO treated, violet = RSL3 treated, turquoise = RSL3/Fer1 treated. d RNA-seq data of human SCLC lines (sensitive n = 3 cell lines, H524, NCIH82, H526; resistant n = 5 cell lines, NCIH889, NCIH1092, CORL88, H1836, H2171) were analyzed for differential expression between responders and non-responders, heatmap represents hierarchical clustering of FPKM (log2 + 0.01) of the 50 most differentially expressed genes. Heatmap color code indicates expression levels between each sample and the average of each gene, dendrogram shows the distance between sample populations. e ASCL1 and REST1 expression (FPKM (log2 + 0.01) comparing three sensitive and five resistant cell lines is plotted, boxplot center line, mean; box limits, upper and lower quartile; whiskers min. to max. f Gene set enrichment analysis (GSEA) of a ranked list from ferroptosis sensitive and resistant cells was performed. g Western blot of SCLC NE subtype marker expression in the indicated manually separated stickers and floaters lines (n = 3). h Manually separated stickers and floaters lines (n = 3) were treated with RSL3 [1 µM] for 24 h. DRAQ7 [0.1 µM] was added to all wells to visualize dead cells. Images were acquired every 2 h using the IncuCyte S3 bioimaging platform. i The indicated mouse RP and RPM cell lines were treated with rising concentrations of RSL3 for 24 h, cell viability was determined by Cell Titer Blue. j Cells as in (i) were analyzed for protein expression by western blotting. Representative western blots are shown k 181.5 stickers expressing either vector control or YAP1 5SA-YFP were treated with RSL3 [1 µM] for 24 h. DRAQ7 [0.1 µM] was added to all wells to visualize dead cells. Images were acquired every 2 h using the IncuCyte S3 bioimaging platform, n = 4 biological replicates. Representative western blots are shown. Data are means ± SEM of three or more independent experiments or representative images if not indicated otherwise. Two-tailed unpaired t tests, ****p < 0.0001, **p < 0.01. Source Data are provided as Source data file.
Fig. 4
Fig. 4. non-NE/NE SCLC subtypes undergo lipid metabolism remodeling.
a Heatmap showing the representation of mono-oxidized phospholipid species (PE phosphatidylethanolamine; PC phosphatidylcholine) in 181.5 stickers as compared to 181.5 floaters treated with either DMSO or RSL3 [1 µM] for 5 h and then subjected to lipidomics. Samples for each condition (n = 5) were averaged and normalized to the cell number (2.5 × 106). Each lipid species was normalized to levels detected in floaters +DMSO. One representative out of two independent experiments is shown. Heatmap color code indicates normalized lipid species levels of each sample. b–e 181.5 stickers floaters (n = 5 samples) as compared to 181.5 floaters (n = 5 samples) were analyzed for basal diacylglycerol (DAG) and ether-linked lipids by mass spectrometry. Lipid content was normalized to infused protein for each condition and replicate. Individual PUFAs (4 double bonds or more) are plotted. f RNA was isolated from three stickers/floaters lines (RP181.5; RP246.7; BYC5.1), respective cDNA obtained and qPCR performed for the indicated transcripts. g RP181.5 were subjected to the indicated siRNA-mediated knockdowns for 72 h and then treated with RSL3 [1 µM] for an additional 24 h. DRAQ7 [0.1 µM] was added to all wells to visualize dead cells. Images were acquired every 2 h using the IncuCyte S3 bioimaging platform. Dead cells/image are normalized to cell confluence at the beginning of RSL3 treatment. h Representative western blots are shown. Data are means ± SEM of three independent experiments or representative images if not indicated otherwise. Two-tailed unpaired t tests (e) Two-way ANOVA + Tukey’s multiple comparison test, ****p < 0.0001, ***p < 0.001, **p < 0.01, *p < 0.05. Source data are provided as Source data file.
Fig. 5
Fig. 5. Neuroendocrine SCLC presents with TRX pathway addiction.
a Schematic view of genes involved in the TRX anti-oxidant pathway. b H82 and H2171 cells were lyzed and TRX pathway component expression was detected by western blot. Representative western blots are shown. c Indicated cells were treated with Auranofin [1 µM] for the indicated times and subjected to redox shift assays. Densitometrical quantification of TRX redox forms is shown. d Indicated cells were treated with Auranofin for 24 h, cell viability was determined by Cell Titer Blue. e, f RP181.5 manually separated stickers and floaters were treated with Ferrostatin-1 [5 µM] for 2 h prior to adding DMSO, BSO [10 mM] or Auranofin [1 µM] for an additional 24 h. DRAQ7 [0.1 µM] (red color in image) was added to all wells to visualize dead cells. Images were acquired every 5 h using the IncuCyte S3 bioimaging platform. Scale bar = 100 µm. g RP181.5 stickers were treated with either DMSO, Auranofin [500 nM] or BSO [500 µM] for 96 h and then fixed and stained for ASCL1 (red) and counterstained with DAPI (blue). Scale bar = 100 µm. h Manually separated stickers and floaters were treated with either DMSO, BSO [10 mM] or Auranofin [1 µM] for 24 h. Cells were gated on live cells and analyzed for ASCL1 expression by flow cytometry. FSC-H, forward scatter-heights. Data are means ± SEM of three independent experiments or two (c) or representative images out of at least two independent experiments are shown. Two-way ANOVA + Tukey’s multiple comparison test, ***p < 0.001. Source data are provided as Source data file.
Fig. 6
Fig. 6. Combined ferroptosis induction and TRX pathway inhibition demonstrates broad anti-tumor activity across SCLC NE subtypes in vivo and serves as prognostic marker set in human SCLC.
a Eight-weeks-old male nude mice were injected with 50:50 mixed 181.5 stickers and floaters at a total of 1.5 × 106 cells into both flanks. Once palpable, tumors were treated either with vehicle (n = 14), Auranofin [2.5 mg/kg] (n = 10) 3× per week i.p., BSO [5 mM] (n = 10) in the drinking water or the combination (combo) (n = 14) for 2 consecutive weeks. Fold change of initial tumor size is shown. Boxplot center line, mean; box limits, upper and lower quartile; whiskers min. to max. b Sections from paraffin-embedded tumors stained for ASCL1. Representative images are shown, scale bar = 100 µm. c ASCL1 H-score was quantified, n (vehicle) = 14, n (BSO) = 9, n (Aura) = 10, n (combo) = 14. d Eight to twelve-week-old RP mice were inhaled intratracheally with 2.5 × 107 plaque-forming units (PFU) Adeno-Cre virus to initiate SCLC development. e Tumor-bearing RP-mice were treated either with vehicle (n = 8) or combined BSO [5 mM] in the drinking water and Auranofin 3× per week i.p. [2.5 mg/kg] (n = 7) for 2 consecutive weeks. Fold change in tumor volumes was determined by quantifying initial tumor volume from MRI scans as compared to tumor volume at the end of the treatment cycle using Horos software. f Representative MRI images pre and post treatment of mice as in (e). g Isolation scheme of human CDXs. h Cellular morphology of human CDXs, scale bar = 400 µm. i Two CDXs or two healthy donor PBMCs were treated with DMSO, BSO [PBMCs, 500 μM; CDXs, 50 μM], Auranofin [250 nM] or BSO [PBMCs, 500 μM; CDXs, 50 μM]/Auranofin [250 nM] for 24 h, cell viability was quantified by Cell Titer Blue (CDXs) or flow cytometric quantification of propidium iodide (PI)-negative cells (PBMCs). j Kaplan–Meier survival curves for SCLC patients (n = 77) containing low (low 1/3 n = 25, median survival 33 months) or high (high 2/3 n = 52, median survival 22.5 months) expression of GPX4 mRNA. k As in (j) expression of TXNRD1 mRNA was correlated using the same cut-off (low = 1/3, median survival 38 months; high 2/3, median survival 22.5 months). l Kaplan–Meier survival curves for SCLC patients with combined low or high GPX4 and TXNRD1 mRNA expression (low/low n = 10, median survival 43.5 months; high/high n = 16, median survival 16 months). Data are means ± SEM were applicable (i). Schemes were drawn with fully licensed Biorender.com. Two-tailed unpaired t tests (a, c), or two-sided log-rank (Mantel–Cox) test (j–i), ***p < 0.001, *p < 0.05, ns p > 0.05. Source Data are provided as Source data file.
Fig. 7
Fig. 7. SCLC non-NE/NE subtypes segregate by lipid ROS/ROS vulnerability.
Heterogeneous SCLC tumors contain NE cells expressing high levels of ASCL1 and low levels of REST1 and non-NE cells with an inverse expression pattern. Upon induction of ferroptosis by BSO, non-NE cells die due to excessive lipid ROS accumulation. NE cells, in turn, undergo ROS-dependent cell death upon TRX pathway inhibition. Surviving cellular fractions in both single treatment arms can transdifferentiate under treatment, thereby escaping therapy. Dual targeting prevents selection and plasticity in SCLC. NE neuroendocrine, ROS reactive oxygen species, GSH glutathione, TRX thioredoxin. The scheme was drawn using fully licensed Biorender.com.
See this image and copyright information in PMC

References

    1. Rudin CM, et al. Molecular subtypes of small cell lung cancer: a synthesis of human and mouse model data. Nat. Rev. Cancer. 2019;17:725. - PMC - PubMed
    1. Yang D, et al. Intertumoral heterogeneity in SCLC is influenced by the cell type of origin. Cancer Discov. 2018;8:1316–1331. doi: 10.1158/2159-8290.CD-17-0987. - DOI - PMC - PubMed
    1. Lim JS, et al. Intratumoural heterogeneity generated by Notch signalling promotes small-cell lung cancer. Nature. 2017;545:360–364. doi: 10.1038/nature22323. - DOI - PMC - PubMed
    1. Wagner AH, et al. Recurrent WNT pathway alterations are frequent in relapsed small cell lung cancer. Nat. Comms. 2018;9:3787–11. doi: 10.1038/s41467-018-06162-9. - DOI - PMC - PubMed
    1. Ireland, A. S. et al. MYC drives temporal evolution of small cell lung cancer subtypes by reprogramming neuroendocrine fate. Cancer Cell10.1016/j.ccell.2020.05.001 (2020). - PMC - PubMed

Publication types