Review papers with raw data transparency

Visual paper analysis — "Targeting FSP1 triggers ferroptosis in lung cancer" (Nature 2025)

Figure source: TCGA analysis reported in Papagiannakopoulos et al.; median survival values taken from main figure legend (Fig.2b) and main text TCGA survival (Fig.2b)

Interpretation: the study combines orthogonal in vivo genetics, orthotopic/xenograft/PDX pharmacology, quantitative lipidomics and biochemical rescues to argue FSP1-dependent ferroptosis suppression is essential for LUAD tumorigenesis Multi-modal experimental design

Mechanistic chain supported by data: genetic Fsp1 or Gpx4 loss → increased oxidized PUFA-PLs and 4‑HNE staining → increased TUNEL/loss of tumour burden → rescue by lipid RTAs (LIP1, vitamin E) or reducing PUFA-PL supply (Acsl4 KO) → pharmacologic inhibition of FSP1 phenocopies genetic loss in vivo Evidence for ferroptosis mechanism

Critical strengths (evidence-based)

• In vivo-first design using autochthonous KP GEMMs with tumour-specific CRISPR: stronger physiological relevance than cell-line–only studies GEMM CRISPR approach
• Orthogonal rescue experiments (LIP1, vitamin E, Acsl4 KO) strongly support lipid-peroxidation–driven ferroptosis as causal, not correlative Rescue experiments
• Quantitative lipidomics (oxPE/oxPC) and CoQ redox measurements provide mechanistic biochemical readouts consistent with known FSP1 biochemistry (CoQ recycling) Epilipidomics and CoQ data

Main weaknesses, blind-spots, and caveats

1. Translatability & toxicity: systemic GPX4 inhibition is known to be highly toxic in mammals; the therapeutic window for FSP1 inhibitors requires comprehensive toxicology (authors acknowledge patenting and early-stage compounds) Safety caveat (main text)
2. Immune microenvironment: reported flow-cytometry profiling showed no gross changes in major immune subsets after icFSP1, but deeper single-cell immune profiling and functional immune assays are missing; ferroptosis can release pro-inflammatory signals (see ferroptosis secretome literature) and could have complex pro‑ or anti‑tumor immune effects that remain unresolved Limited immune profiling noted
3. Model selection bias: strong effect across many models (GEMM, xenografts, PDX) is compelling, but majority of in vivo work is in KRAS-driven contexts; robustness across broad patient-derived, genetically diverse LUAD or non-lung cancers needs broader validation (authors present companion melanoma work by Palma et al.) Generality limited to tested models
4. On-target validation of icFSP1: authors used an icFSP1-resistant FSP1(Q319K) allele to show on-target activity in vivo — good practice — but detailed ADME/PK, off-target panels, and toxicology data are not publicly available in the paper's main figures (likely in supplements) icFSP1 on-target control (Q319K)
5. CRISPR KO limitations: tumour-specific CRISPR delivery is powerful, but potential sgRNA off-targets and selection of clones in cell lines can introduce artifacts; authors validated KO by IHC/western blot and used multiple sgRNAs to mitigate this risk (good practice) CRISPR validation details

How robust is the mechanistic inference?

The mechanistic claim that FSP1 protects tumours by regenerating reduced CoQ and preventing propagation of PUFA-PL peroxidation is directly supported by decreased CoQ9H2/CoQ9 ratios and accumulation of oxidized PE/PC species in Fsp1 KO tumours, plus functional rescues (Acsl4 KO, vitamin E, LIP1) that are specific to lipid radical handling; re-expression of membrane-targeted FSP1 (myristoylation) is necessary for protection, consistent with published FSP1 biochemistry CoQ & FSP1 localization data

Comparison to contemporaneous literature

Two points of context from recent studies: (1) FSP1 was first identified as a GPX4-independent ferroptosis suppressor and shows biochemical CoQ oxidoreductase activity (Bersuker/Doll 2019); (2) recent work shows context-specific FSP1 reliance during metastasis/lymphatic colonization (Palma et al., 2025), consistent with the present in vivo dependency, and other groups report development of diverse FSP1 inhibitors with variable in vivo stability — placing this paper as a major in vivo demonstration of FSP1 as a therapeutic target Position in field (this study)

Concrete next experiments to strengthen/translate the claim

1. GLP toxicology and chronic dosing of icFSP1 (IV/PO) in two species (rodent, non-rodent), with full histopathology, hematology, and CoQ redox readouts in major organs.
2. Single-cell RNA + lipidomics of treated tumours to map which cell types accumulate lipid peroxides and whether ferroptosis induction triggers immunogenic signals (DAMPs, prostaglandins) that alter anti-tumour immunity (connects to secretome studies) Need for single-cell immune-lipidomics
3. Combination therapy studies: icFSP1 + immune checkpoint blockade and icFSP1 + inhibitors of CoQ biosynthesis (statins/ mevalonate pathway) to test synergy and potential for reduced dosing.
4. Evaluate FSP1 dependence across large PDX panels and organoids representing the genomic diversity of LUAD (EGFR, KRAS, STK11, KEAP1, TP53 cases) with paired ex vivo lipidomics to identify biomarkers predicting response (e.g., baseline CoQ redox, PUFA-PL content, ACSL4 expression).

Summary judgements (concise, evidence-weighted)

• Conclusion validity: High — multi-model genetic and pharmacologic in vivo evidence plus biochemical lipidomics support the causal claim that FSP1 suppresses ferroptosis in LUAD and that FSP1 inhibition triggers ferroptosis in vivo Causal evidence
• Translational potential: Promising but conditional — the PDX icFSP1 efficacy is encouraging, but clinical translation depends on safety, PK/PD, and wider-tested generality across human tumour genotypes.

All claims above are directly tied to experimental evidence in the paper; where broader interpretation or translational inferences were made I cited mechanistic data (lipidomics, CoQ), rescue experiments (LIP1/VitE/Acsl4 KO), and icFSP1 pharmacology in PDX — see inline citations for direct excerpts and methods.