CD74 deficiency protects against doxorubicin cardiotoxicity through RRM2-mediated regulation of ferroptosis.
Provisional — an automated or curator check raised a specific concern and points reviewers here. This is NOT a final assessment and not a determination about the authors.
The main result did not reproduce in this reproduction attempt. Where our recomputation produced values that differ from the published ones, those discrepancies are listed below. This is a single automated attempt — not peer review and not a finding of error or misconduct — and differences can also arise from data access, undocumented parameters or the computing environment. The verdict can be contested via “report an error”.
Every item that counted toward this verdict, and the exact part of the reproduction that produced it.
- ✓Same input data as the authors
- ✓Reported values were directly comparable
- 🔴A deviation arose in the data or preprocessing
- 🔴A deviation was attributed to the published material
- 🔴Reported values were not (fully) derivable from the shared data
- 🔴The deviation was non-trivial in magnitude
- 🔴The central claim did not (fully) hold under reproduction
- 🔴Overall, the reproduction showed a material discrepancy
A 0–100 reproducibility-quality score from the per-question grades, shown as a z-score: standard deviations above (+) or below (−) the mean of comparable assessments.
▸Reproduction agent’s raw note
Predominantly WET-LAB paper; only pipeline-derived content is a reanalysis of THREE third-party public GEO sets (GSE224157 mouse-heart RNA-seq, GSE244574 MCF7 microarray, GSE205551 4T1 scRNA). No authors' analysis code exists (cited 'code' link github.com/tidyverse/ggplot2 is the generic plotting library; the BRIEF data accession GSE224157 actually belongs to a different paper, PMID 37232184). Described well enough at a high level (DESeq2/limma DEGs FC>=2 & P<0.01, clusterProfiler KEGG) to attempt via standard tools (valid per P16). RESULT = DIFFERENT, not 1:1: of 6 in-scope claims, only NCOA4-up reproduces cleanly (within-tol) and TP53-up partially (direction right, sub-threshold); FOUR mismatch including two OPPOSITE-direction contradictions of the very datasets cited -- CD74 is DOWN (not up) in DOX hearts (both contrasts + raw per-sample counts agree), and GPX4 is UP (not down, P=5e-14) in DOX MCF7. SLC7A11 shows no change; ferroptosis KEGG enrichment is non-significant. FLAGS: fake/placeholder code link; data accession attributed to a different study; Methods claim 'DESeq2' applied to an Agilent MICROARRAY (invalid; we used limma). NOT attempted: GSE205551 scRNA GSVA (C4) -- 'DOX-sensitive' subset undefined; all wet-lab results (out of scope). Grades provisional for human audit, not ground truth; reanalysis of third-party data is sensitive to contrast/normalization/annotation choices, but the per-sample CD74 counts and the strongly-significant opposite GPX4 effect are hard to reconcile with the reported directions.
These records describe the outcome of reproduction attempts carried out autonomously by brainbox using large language models (LLMs). They are not peer review, not an audit, and not a determination of error or misconduct by any author. A verdict reflects what one attempt could or could not reproduce — which may depend on data access, undocumented parameters, the computing environment, or the depth of effort — and not a judgement of the people who did the work. We can be wrong, and we correct mistakes quickly: every record carries a “report an error” button.
Assessment versions
Every reproduction run is kept as an immutable version — anchored to the data as it stood, with a tamper-evident chain hash. A rerun (e.g. after an author updates a deposit) adds a new version; the previous one stays on record.
-
v1 current initial assessment Score 29assessed: 2026-06-14 ⛓ 1afe40f6cacb
✎ I am an author of this paper
Updated or fixed a deposit, or is there an erratum? Ask us to re-run the metrics. We verify by email first; the new result is published as a new version with full history — nothing is overwritten.
Provenance — full disclosure
When this reproduction was carried out, which methodology version was used, and by whom — so the record can be audited and checked independently.
- Reproduced
- 2026-06-14
- Rubric version
- v1.0
- Assessed by
-
🤖 AI curator · claude (ai-curator room) · v1.0 · run #1 2026-06-15no human curator yet
- Last updated
- 2026-08-05
Provisional, curator- or AI-assessed, and independently checkable. A reproduction outcome states what one attempt could reproduce — not a judgement of the authors.
Deep full-text extraction
Model: opusThis study tests whether the immunoregulatory protein CD74 contributes to doxorubicin-induced cardiotoxicity, hypothesizing that CD74 deletion or inhibition protects the heart by regulating RRM2-mediated ferroptosis and the RRM2/p53 cascade.
- ★ CD74 levels are elevated in plasma of DOX-exposed breast cancer patients and in DOX-challenged mouse hearts and cardiomyocytes finding
- ★ CD74 deletion mitigates DOX-induced cardiac remodeling, contractile anomaly, mitochondrial abnormalities, apoptosis, and ferroptosis finding
- ★ CD74 binds RRM2 and redistributes it from cytoplasm to plasma membrane, impairing DOX-induced repair and exacerbating injury via RRM2/p53 cascade activation mechanism
- ★ A CD74 mutant (aa 220–250) fails to aggravate DOX-induced cardiac dysfunction unlike WT CD74 finding
- ★ Protective effects of CD74 inhibition in cardiomyocytes are negated by p53 activation mechanism
- ★ The CD74 inhibitor Amifostine alleviates cardiac remodeling and functional impairment in DIC mice by reducing oxidative stress and ferroptosis finding
- ★ CD74 knockdown in RAW 264.7 macrophages attenuates DOX-instigated cardiomyocyte dysfunction in a Transwell co-culture finding
- CD74 interacts with RRM2 through its 180–215 amino acid domain to recruit RRM2 onto the cytomembrane mechanism
| Assay | System | Perturbation | Readout | Platform |
|---|---|---|---|---|
| ELISA (serum CD74 and cTnT) | plasma from breast cancer patients with/without DIC (DIC vs non-DIC) | DOX therapy | serum CD74 and cardiac troponin-T levels | BioVision and Beyotime ELISA kits |
| Echocardiography (2-D guided M-mode) | WT and CD74−/− mice | DOX 5 mg/kg i.p. weekly x4 weeks | LVEDD, LVESD, wall thickness, ejection fraction, fractional shortening, heart rate, LV mass | Vevo 2100, Visualsonics |
| Cardiomyocyte contractility and intracellular Ca2+ (Fura-2) | adult mouse cardiomyocytes (AMCMs) from WT and CD74−/− mice | DOX exposure | peak shortening, ±dL/dt, TP90, TR90, FFI, ΔFFI, Ca2+ decay | — |
| Western blot / gel blot | mouse hearts, AMCMs, patient plasma | DOX, CD74 KO | CD74, ANP, BNP, cleaved-caspase 3, Bcl2, Bax, UCP2, LC3B, p62, RRM2, p53, GPX4, SLC7A11, NCOA4 | — |
| Histology (WGA, H&E, Masson trichrome, 4-HNE, TUNEL, DCF) | WT and CD74−/− mouse myocardium | DOX 5 mg/kg weekly x4 weeks | cross-sectional area, fibrosis, oxidative stress, apoptosis, ROS | digital microscope 400x, ImageJ |
| Transmission electron microscopy, JC-1, Seahorse OCR, mitochondrial respiration | AMCMs/myocardium from WT and CD74−/− mice | DOX challenge | mitochondrial area/circularity, membrane potential, basal/maximal respiration, ATP, complex I/III, II/III, IV activity | — |
| Co-immunoprecipitation and immunofluorescence/confocal co-localization | AMCMs (CD74-Flag, RRM2-His), CD74 deletion mutants | DOX exposure; CD74 domain mutants (Δ180–215, Δ220–250, Δ180–250) | CD74-RRM2 interaction and RRM2 membrane recruitment | Leica SP8 confocal; STRING/PRISM in silico |
| Single-cell RNA-seq and bulk transcriptomics (RNA-seq) analysis | 4T1 tumor BALB/c mice (GSE205551), MCF7 cells (GSE244574), DOX-treated mouse hearts (GSE224157) | DOX treatment | differentially expressed genes, KEGG/GO enrichment, ferroptosis GSVA score, CD74 expression | Seurat, Harmony, DESeq2, clusterProfiler, GSVA |
- ▲ CD74 elevated in DOX-exposed patient plasma, mouse hearts and AMCMs (volcano/IF/blot)
- – DIC patients show reduced LVEF and elevated NT-proBNP versus non-DIC LVEF 51.3% vs 64.7%; NT-proBNP 189.7 vs 29.9 pg/mL
- ▲ CD74−/− preserves ejection fraction and fractional shortening after DOX
- ▼ CD74 deletion reduces cardiomyocyte cross-sectional area, fibrosis, TUNEL apoptosis and DCF ROS after DOX
- ▲ CD74 deletion improves mitochondrial ultrastructure, membrane potential, OCR/ATP and respiratory complex activity after DOX
- – DOX increases membrane RRM2, p53 and intracellular Fe2+, and decreases GPX4/SLC7A11 with increased NCOA4; CD74 KO attenuates these
- ▲ CD74 knockdown in macrophages attenuates DOX-induced cardiomyocyte contractile dysfunction in co-culture
- – CD74 Δ220–250 mutant fails to aggravate DOX-induced cardiac dysfunction unlike WT CD74
- mean LVEF 51.3 ± 4.9% (DIC) vs 64.7 ± 4.9% (non-DIC), P < 0.05 (breast cancer patients DIC vs non-DIC)
- mean NT-proBNP 189.7 ± 79.6 pg/mL (DIC) vs 29.9 ± 3.4 pg/mL (non-DIC), P < 0.01 (patient cardiac biomarker)
- mean age 48.6 ± 4.6 vs 46.5 ± 5.3 year, P > 0.05 (DIC vs non-DIC patient age)
- count DOX 5 mg/kg i.p. once weekly for 4 weeks (chronic DIC mouse model dosing)
- count n = 13 mice per group (echocardiography groups)
- count n = 48 cells per group (cardiomyocyte contractile measurements)
- fold_change DEGs fold-change ≥2 and P < 0.01 (transcriptomic DEG threshold)
- count saline tumors expanded 400% by Day 8 (GSE205551 4T1 tumor growth)
Statistical methods review
Model: sonnetA neutral, descriptive read of the statistical approach — what was done, and (for shared learning, not as criticism) what could also have been done.
This study used a 2×2 factorial mouse model (CD74 KO vs. WT × DOX vs. saline, chronic 4-week protocol) combined with primary adult mouse cardiomyocyte assays, a Transwell macrophage–cardiomyocyte co-culture system, and human plasma samples from breast cancer patients with or without doxorubicin-induced cardiotoxicity. Bioinformatic analyses of publicly available scRNA-seq (GSE205551) and bulk RNA-seq (GSE244574, GSE224157) datasets were performed using DESeq2, Seurat, GSVA, and clusterProfiler. All experimental outcome data are reported as mean ± SEM with significance denoted as *P < 0.05; the specific inferential tests applied to in vivo and in vitro comparisons are not named in the available text.
| Test | Applied to | n | Assumptions |
|---|---|---|---|
| DESeq2 Wald test | Differential expression analysis of bulk RNA-seq datasets GSE244574 (MCF7 cells + DOX) and GSE224157 (mouse hearts ± DOX), with DEG thresholds of fold-change ≥2 and P < 0.01 | — | not stated |
| GSVA (gene set variation analysis) | Calculation of ferroptosis pathway activity scores from GSE224157 transcriptomic data | — | not stated |
| Not explicitly named (reported as '*P < 0.05 between indicated groups') | All in vivo echocardiographic, histological, mitochondrial, and protein-level comparisons across the 4-group 2×2 factorial design (WT/CD74−/− × Saline/DOX), and Amifostine treatment groups | n = 6–13 mice per group; n = 25–48 cells per group from 3–5 mice | not stated |
| Not explicitly named (P values reported as P > 0.05, P < 0.05, P < 0.01) | Comparison of patient characteristics (age, LVEF, NT-proBNP) between DIC and non-DIC breast cancer patients | n = 12 (DIC group); non-DIC group n not separately stated | not stated |
| Not explicitly named (reported as '*P < 0.05') | In vitro AMCM contractile and Ca2+ handling properties, Transwell co-culture cardiomyocyte functional endpoints, DCF fluorescence, FerroOrange Fe2+ assay | n = 25–48 cells per group from 3–5 mice | not stated |
-
The 2×2 factorial design (genotype × DOX treatment, 4 groups) is analyzed with pairwise significance markers (*P < 0.05), but the inferential test is not named and no interaction term is reported↳ Could also: A two-way ANOVA followed by a post-hoc correction (e.g., Tukey HSD or Sidak) could also be applied to the same design — Two-way ANOVA explicitly estimates the genotype × treatment interaction term, which directly quantifies whether CD74 deletion modifies DOX's effect beyond additive contributions — a quantity of central biological interest in this study — and controls the family-wise error rate across all pairwise comparisons simultaneously
-
Cells isolated from multiple mice were analyzed as independent observations (e.g., n = 48 cells from 3–5 mice), using the cell as the statistical unit↳ Could also: A linear mixed-effects model treating mouse as a random effect and cell as a nested observation could also be used — Cells from the same animal share unmeasured biological variation, so treating them as fully independent inflates the effective sample size; a nested or mixed-effects approach accounts for within-mouse correlation and yields more conservative, biologically interpretable estimates of treatment effects
-
Dispersion is reported exclusively as SEM across all experimental figures↳ Could also: SD or a 95% CI could also be used to convey the same data — SD describes the variability of individual observations rather than precision of the mean, and 95% CIs convey both precision and effect magnitude; both are increasingly recommended by reporting guidelines (ICMJE, Nature guidelines) for small-to-moderate sample sizes to enable readers to judge biological variability and effect size
-
Statistical significance for experimental comparisons is reported as threshold categories (P < 0.05, P < 0.01, P > 0.05) rather than exact P values↳ Could also: Exact P values (e.g., P = 0.018) could also be reported alongside significance stars — Exact P values allow readers to judge the strength of evidence more finely, support reproducibility assessments, and are required or strongly encouraged by many journals and reporting standards (APA, CONSORT-style guidance) as they preserve more information than threshold-based categorization
-
Multiple independent outcome families (cardiac function, histology, apoptosis markers, mitochondrial function, ferroptosis markers, calcium handling) are all tested at P < 0.05 across the same animal groups without any stated multiplicity adjustment↳ Could also: A Benjamini-Hochberg FDR correction or a Bonferroni adjustment grouped by outcome domain could also be applied — Testing many endpoints on the same animals raises the probability of at least one false positive; a stated correction strategy — even a lenient FDR approach applied within each outcome domain — helps readers calibrate confidence in individual findings and is standard practice in multi-endpoint preclinical studies
-
DESeq2 was used for differential expression on two independent RNA-seq datasets (GSE244574, GSE224157) with no cross-validation or orthogonal method reported↳ Could also: edgeR or limma-voom could also be applied to the same count data, with concordant hits across methods highlighted — All three tools are standard for bulk RNA-seq DEG analysis and differ in their dispersion estimation strategies; reporting the overlap of significant hits across two methods is a common sensitivity check used to increase confidence in key candidates such as RRM2
Citation network
Where this publication sits in the reproducibility-weighted citation graph — what it is built on, and what is built on it. Citation data from OpenAlex.
No assessed neighbours yet — the network grows as more papers are assessed.
Data lineage
The datasets this paper uses (text-mined from the full text via Europe PMC), and which other assessed papers stand on the same data. A shared dataset is a factual link — not a judgement.
Downstream reach in the literature
2 downstream papers · 2 datasetsHow widely the datasets deposited by this paper are reused across the whole literature (Europe PMC), beyond our assessed set. This is a factual dependency map — reusing a public dataset is normal, good science. It is not a judgement on the downstream papers; the only verdict here is this paper's own, with its cited rationale.
- PDE10A Inactivation Prevents Doxorubicin-Induced Car... 2023 · 54 cites
- miR-218-5p and doxorubicin combination enhances anti... 2024 · 21 cites
What was reproduced
The exact results taken into scope, with each reported value next to the value our attempt produced.
Scope — PMID 42180553
Paper: CD74 deficiency protects against doxorubicin cardiotoxicity through RRM2-mediated regulation of ferroptosis. Acta Pharmaceutica Sinica B (2026). DOI 10.1016/j.apsb.2026.01.028 · PMCID PMC13198236.
Nature of the paper
This is predominantly a wet-lab mechanistic study (CD74→RRM2→p53/ferroptosis in mouse hearts, cardiomyocytes, IF/confocal, echocardiography, co-IP, CD74 mutants, xenografts). The computational / pipeline-derived content is a small reanalysis of three pre-existing public datasets — none generated by the authors.
Code/data link audit (control-plane screening, «host»)
- BRIEF "Code" link
github.com/tidyverse/ggplot2= the generic Grammar-of-Graphics R plotting library — NOT an analysis pipeline. The paper Methods cite only plotting/vis tools:ggplot2(volcano + bubble plots) andgithub.com/kevinblighe/EnhancedVolcano(volcano plots). No authors' analysis code repository exists. →no_codefor an own pipeline. - BRIEF "Data"
GSE224157is not the authors' own dataset — it is a third-party series (Chen Yan lab, U. Rochester, PMID 37232184, "PDE10A inactivation ... doxorubicin cardiotoxicity"). The paper reuses it. - Paper actually reanalyzes three public GEO series (all third-party):
- GSE224157 — mouse heart RNA-seq, PDE10A-WT/KO ± DOX, 12 samples,
per-sample raw gene counts shipped (
*_countsU.txt.gz). [DIC model] - GSE244574 — MCF7 breast-cancer cells ± DOX (30 µM, 24 h), Agilent one-color microarray (GPL21185), 16 samples (8 untreated / 8 DOX, 2 batches).
- GSE205551 — 4T1 TNBC scRNA-seq ± DOX (PMID 35924165). [Supporting Fig S3]
- GSE224157 — mouse heart RNA-seq, PDE10A-WT/KO ± DOX, 12 samples,
per-sample raw gene counts shipped (
Per project rule P16, applying standard third-party tools to the paper's named public data is a fully valid reproduction. The pipeline is specified at a high level: DESeq2 DEGs (FC≥2, P<0.01) + GO/KEGG (clusterProfiler) + GSVA.
In-scope (pipeline-derived) results → attempted
| id | dataset | result (reported) | pipeline | priority |
|---|---|---|---|---|
| C1 | GSE224157 | CD74 expression significantly increased in DOX/DIC mouse hearts vs control (Fig 1A) | DESeq2, WT-DOX vs WT-veh | HIGH (low-hanging) |
| C2a | GSE244574 | GPX4 downregulated in DOX-treated MCF7 (Fig 5 volcano) | limma, DOX vs untreated | HIGH |
| C2b | GSE244574 | SLC7A11 downregulated in DOX-treated MCF7 | limma | HIGH |
| C2c | GSE244574 | TP53 upregulated (significant) in DOX-treated MCF7 | limma | HIGH |
| C2d | GSE244574 | NCOA4 upregulated (significant) in DOX-treated MCF7 | limma | HIGH |
| C3 | GSE224157 | Ferroptosis pathway enriched (KEGG + GSVA) in DOX mouse hearts (Fig 5F/G) | clusterProfiler enrichKEGG mmu04216 / GSVA | MEDIUM |
Out of scope (NOT attempted)
- All wet-lab results: CD74/RRM2 co-IP, membrane redistribution, p53 western, IF/confocal, echocardiography, mitochondrial assays, CD74 aa220–250 mutant, xenografts, patient blood CD74. (Manual/experimental — not pipeline-derived.)
- C4 (hard 20%, skipped): GSE205551 scRNA-seq GSVA ferroptosis in "DOX-sensitive" cells (Fig S3) — requires reconstructing the authors' cell-subset / sensitivity classification, which is unspecified. Documented, not attempted.
Key methods flags (possible-fabrication / inconsistency notes)
- DESeq2 on a microarray: Methods state "DESeq2 was used for differential expression analysis" for both GSE244574 and GSE224157, but GSE244574 is an Agilent microarray (intensities, not counts). DESeq2 is invalid for microarray data; the correct tool is limma. We reproduce GSE244574 with limma and flag the mis-described method.
- Fake/placeholder code link (ggplot2) and a data accession attributed to a different study in the harvested metadata — both recorded above.
Compute plan
All compute on «our HPC» SLURM («infra» workdir), conda prefix-env R/Bioconductor. Data downloaded inside the job (compu
Assessments & scoring basis
Each contributor’s verdict, the per-question basis, and the auditable, itemised worksheet behind it.
An automated assessment. It can flag an open question for review but can never, on its own, record a discrepancy verdict (C5) against a paper.
Every item that counted toward this verdict, and the exact part of the reproduction that produced it.
This is a predominantly wet-lab paper whose only pipeline-derived content is a reanalysis of three public third-party GEO datasets, so data identity (q1) and endpoint comparability (q2) are fine. The problem is squarely on the authors' side: of 6 in-scope claims only NCOA4↑ reproduces cleanly and TP53↑ partially, while four mismatch — including two opposite-direction reversals (CD74 is DOWN not up in Fig 1A's data; GPX4 is UP P=4.7e-14 not down) that are robust across contrasts and raw per-sample counts, plus a non-significant ferroptosis KEGG (padj=0.605). Compounding flags — a placeholder ggplot2 code link, a misattributed GSE224157 accession, and DESeq2 named for an Agilent microarray — make the bioinformatic narrative non-derivable from the shared data and fabrication-suspect. Severity is high because the motivating premise (CD74↑ → ferroptosis) does not hold on reproduction.
Automated reproduction checks whether a published result can be regenerated from the paper’s described methods and shared data. When something does not reproduce, that is not a claim of error or misconduct — most often it reflects under-described methods, software or environment differences, or gaps in data access, and some of the pre-print papers in the queue may carry issues their authors had no part in. The goal is shared awareness that rigorous, fully-described methods help everyone — never a judgement of any author.
Are you an author? We would genuinely like to hear from you — to clarify the record, add data or code, re-run the pipeline after an accession update, and publish your response right next to the assessment. Everything here is open and auditable.
🚩 Report an error in this record
Spotted something wrong — a verdict you’d contest, a data or value error, or a private detail that slipped through? Tell us, with a short justification. Authors and readers are equally welcome to write in; we review every report.
Prefer email, or the form below not working? Contact us at support@doesitreproduce.com.
Reproduction footprint
claude-opus-4-8Measured resources invested to assess this paper — sanitised (machine class only, no job ids/paths). Compute = HPC accounting (SLURM); tokens = the AI agent's session.