Corpus 1,272 assessed · 1,173 scored · 643 reproduced ≥75 · 168 flagged ·∅ 74.1/100
← New search

Brown fat activation reduces hypercholesterolaemia and protects from atherosclerosis development

· 2015
PubMed 25754609 ↗ pmid-25754609
L1 No computation 2/4
Why this verdict

Part of the results reproduced; minor but material deviations remained.

Reproduced on the brainbox compute brainarbeit.com
Scoring basis — itemised

Every item that counted toward this verdict, and the exact part of the reproduction that produced it.

Supporting (toward a concern)
Content-critical question only partially held
+2 pts
From: Q5 · Derivability / plausibility 🟡
Content-critical question only partially held
+2 pts
From: Q7 · Core claim 🟡
Content-critical question only partially held
+2 pts
From: Q8 · Severity of the miss (overall human judgment) 🟡
Minor / cosmetic deviation
+1 pts
From: Q3 · Location of the main deviation 🟡
Input / endpoint not comparable 1:1
+1 pts
From: Q1 · Data identity 🔴
Total score +8
✓ What held up
  • No authors-side cause for any deviation
  • Any deviation was negligible
What did not (or only partly)
  • 🔴Could not use the authors’ exact input data
  • 🔴Reported values were only indirectly comparable
  • 🟡A deviation arose in the data or preprocessing
  • 🟡Reported values were not (fully) derivable from the shared data
  • 🟡The central claim did not (fully) hold under reproduction
  • 🟡Overall, the reproduction showed a material discrepancy
Reproduction agent’s raw note

DROP (non_pipeline). 'Brown fat activation reduces hypercholesterolaemia and protects from atherosclerosis development', Berbee et al., Nature Communications 6:6356 (2015), DOI 10.1038/ncomms7356, PMC4366535. This is a wet-lab mouse physiology / lipid-metabolism / atherosclerosis study with NO pipeline-derived computational results in scope. Verified against PubMed, PMC full text, and the Nature article: no high-throughput omics (gene expression is targeted qRT-PCR normalized to beta-2-microglobulin and 36b4, not array/RNA-seq), no deposited public dataset (no GEO/SRA/ENA/ArrayExpress/PRIDE accession, no data-availability/accession-codes statement), and no analysis code/repository. All reported values are bench/instrument measurements (enzymatic plasma TG & cholesterol, FPLC lipoprotein profiles, [3H]/[14C] radiotracer organ-uptake kinetics, indirect calorimetry, EchoMRI, atherosclerotic-lesion histomorphometry, UCP1 IHC) analysed in SPSS 20.0 (two-tailed Student t-test, univariate regression). There is no data+code pipeline to reproduce and no dataset to profile (dataset_profile.datasets=[], n_public_datasets=0). No «our HPC» compute was spent (by design — nothing to run). NOT ATTEMPTED: computational reproduction of any result, because none is pipeline-derived; this is an honest scope decision, not a failed run. Secondary gaps that also hold: no_data_accession, no_code.

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.

  1. v1 current initial assessment
    assessed: 2026-06-18 ⛓ 43b9fca4f9c2
✎ 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.

Reason for the rerun

We email you a confirmation link first. The rerun is an objective re-measurement — it cannot change the verdict in your favour, only ask us to look again.

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
2026-06-18
Rubric version
v1.0
Assessed by
🤖 AI curator · claude (ai-curator room) · v1.0 · run #1 2026-06-18
no 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: sonnet
Founding hypothesis

Whether brown adipose tissue (BAT) activation via β3-adrenergic receptor stimulation protects against or aggravates hypercholesterolaemia and atherosclerosis in hyperlipidemic mice, and whether this effect depends on a functional hepatic apoE-LDLR remnant clearance pathway.

Core claims
  • β3-AR agonist-mediated BAT activation reduces plasma triglycerides and cholesterol and protects against atherosclerosis development in APOE*3-Leiden.CETP mice. finding
  • BAT activation enhances selective uptake of fatty acids from triglyceride-rich lipoproteins into BAT, generating cholesterol-enriched remnants that are subsequently cleared by the liver. mechanism
  • The anti-atherogenic and cholesterol-lowering effect of BAT activation requires a functional hepatic apoE-LDLR clearance pathway, as it is absent in Apoe-/- and Ldlr-/- mice. finding
  • (V)LDL-cholesterol exposure, not triglyceride exposure, is the main predictor of atherosclerotic lesion area after BAT activation. finding
  • APOE*3-Leiden.CETP mice are a human-like lipoprotein metabolism model that, unlike Apoe-/- and Ldlr-/- mice, retain functional apoE and LDLR. resource
  • The lipid-lowering effect of β3-AR agonism does not depend on CETP expression, as it also occurs in E3L mice lacking CETP. finding
  • CL316243 selectively activates existing BAT and induces WAT browning without the systemic hormonal effects of cold exposure. method
Experimental setups
Assay System Perturbation Readout Platform
indirect calorimetry (energy expenditure, FA/carbohydrate oxidation, activity) female E3L.CETP mice on Western-type diet CL316243 (β3-AR agonist) vs vehicle (PBS) energy expenditure, FA oxidation, respiratory exchange ratio, activity
histology (lipid droplet size, WAT browning) and immunohistochemistry (UCP1 staining) interscapular BAT and WAT depots from E3L.CETP mice CL316243 vs vehicle lipid vacuole size, UCP1 protein content per area/pad
plasma lipid analysis (TG, total cholesterol, lipoprotein fractionation) E3L.CETP mice on Western-type diet (also E3L mice without CETP) CL316243 vs vehicle; also cold exposure vs room temperature plasma TG, TC, VLDL-TG, (V)LDL-C levels FPLC/lipoprotein profiling
hepatic gene expression liver of E3L.CETP, Apoe-/-, and Ldlr-/- mice CL316243 vs vehicle mRNA levels of Apob, CETP, Ldlr, Mttp qPCR (implied)
in vivo plasma clearance and organ uptake of dual-radiolabeled VLDL-mimicking particles ([3H]triolein/[14C]cholesteryl oleate) E3L.CETP mice (8-day CL316243 treatment) CL316243 vs vehicle plasma decay half-life, tissue-specific 3H/14C uptake, lipolysis index
atherosclerotic lesion quantification (area, severity, composition) aortic root of E3L.CETP mice (10-week study) CL316243 vs vehicle lesion area, lesion severity grade, collagen/smooth muscle cell/macrophage content, stability index histomorphometry
body composition and plasma lipid/atherosclerosis assessment WTD-pair-fed Apoe-/- and Ldlr-/- mice CL316243 vs vehicle body weight, WAT pad size, plasma TG/TC/(V)LDL-C, atherosclerotic lesion area
univariate regression analysis E3L.CETP, Apoe-/-, and Ldlr-/- mice lesion/lipid exposure data none (correlative analysis) correlation of SQRT-transformed lesion area with plasma TG or TC exposure (AUC)
Key results
  • β3-AR agonism reduced plasma TG in E3L.CETP mice throughout treatment ~54%
  • β3-AR agonism reduced plasma total cholesterol and (V)LDL-C in E3L.CETP mice ~23% TC, ~27% (V)LDL-C
  • β3-AR agonism reduced mean atherosclerotic lesion area in E3L.CETP mice 43%
  • BAT activation had no effect on plasma TC, (V)LDL-C, or atherosclerosis in Apoe-/- or Ldlr-/- mice despite reducing TG
  • Total energy expenditure and FA oxidation increased on day of CL316243 treatment +17% EE, +67% FA oxidation
  • CL316243 increased uptake of [3H]oleate by BAT depots and increased hepatic uptake of [14C]cholesteryl oleate-labelled remnants 2-3 fold BAT uptake; ~25% liver uptake
  • SQRT-transformed lesion area correlated with plasma TC exposure but not TG exposure TC: R2=0.333, P<0.001; TG: R2=0.086, P=0.10
  • (V)LDL-C exposure specifically predicted SQRT lesion area R2=0.358, P<0.001
Key statistics
  • pvalue P<0.001 (correlation of lesion area with plasma TC exposure)
  • correlation R2=0.358, β=0.055, P<0.001 ((V)LDL-C exposure predicting SQRT lesion area)
  • correlation R2=0.333, β=0.054, P<0.001 (TC exposure predicting SQRT lesion area)
  • correlation R2=0.086, β=0.028, P=0.10 (TG exposure vs SQRT lesion area (not significant))
  • other plasma half-life t1/2 [3H]TO: vehicle 2.9±0.1 min vs CL316243 1.5±0.1 min, P<0.001 (VLDL-mimicking particle clearance kinetics)
  • other plasma half-life t1/2 [14C]CO: vehicle 4.7±0.5 min vs CL316243 3.3±0.3 min (VLDL-mimicking particle core remnant clearance kinetics)
  • other lipolysis index (3H/14C ratio): BAT 11.1±0.9 vs liver 0.22±0.01; vehicle 0.22±0.01 vs CL316243 0.15±0.01, P<0.01 (selective FA uptake by BAT vs whole-particle uptake by liver)
  • fold_change 43% lower mean lesion area (atherosclerotic lesion area reduction with CL316243 in E3L.CETP mice)

Statistical methods review

Model: sonnet

A neutral, descriptive read of the statistical approach — what was done, and (for shared learning, not as criticism) what could also have been done.

The study used a between-group experimental design across three mouse models (E3L.CETP, Apoe−/−, Ldlr−/−) comparing β3-adrenergic receptor agonist (CL316243) treatment to vehicle (PBS) control, with pair-feeding used in the knockout models to equalize dietary cholesterol intake. Key kinetic parameters (plasma clearance half-lives, lipolysis indices) were compared with unpaired two-tailed Student's t-tests, and relationships between plasma lipid exposure and SQRT-transformed atherosclerotic lesion area were assessed by univariate linear regression. Group differences for body composition, energy expenditure, and plasma lipids were reported largely as percentage changes with P-value thresholds rather than exact values, and no multiplicity correction was described.

Replicationbiological GroupsCL316243 (β3-AR agonist) vs vehicle (PBS) in E3L.CETP mice (10-week WTD); pair-fed CL316243 vs vehicle in Apoe−/− and Ldlr−/− mice; cold exposure vs room temperature as a secondary comparison in E3L.CETP mice Pairingunpaired Randomization/blindingnot stated Dispersionunclear Effect sizesyes Confidence intervalsno Multiplicity correctionnone stated
Statistical tests used
Test Applied to n Assumptions
unpaired two-tailed Student's t-test Plasma clearance half-life of [3H]TO vs [14C]CO in vehicle-treated E3L.CETP mice (t½ 2.9±0.1 vs 4.7±0.5 min, P<0.01; Fig. 6a,c) not stated
unpaired two-tailed Student's t-test Plasma clearance half-life of [3H]TO: CL316243 vs vehicle (t½ 1.5±0.1 vs 2.9±0.1 min, P<0.001; Fig. 6a) not stated
unpaired two-tailed Student's t-test Lipolysis index comparison: CL316243 vs vehicle (0.15±0.01 vs 0.22±0.01, P<0.01; derived from Fig. 6b,d) not stated
univariate linear regression FA uptake by individual BAT depots (3H activity) predicting hepatic cholesterol-enriched remnant uptake (14C activity) (Fig. 6e–g) not stated
univariate linear regression SQRT-transformed atherosclerotic lesion area vs plasma TG exposure (β=0.028, R²=0.086, P=0.10) and vs plasma TC exposure (β=0.054, R²=0.333, P<0.001), and vs (V)LDL-C exposure (β=0.055, R²=0.358, P<0.001) in E3L.CETP mice (Fig. 8i–k) not stated
univariate linear regression SQRT-transformed atherosclerotic lesion area vs plasma TG and TC exposure in Apoe−/− mice (Fig. 9e,f) and Ldlr−/− mice (Fig. 9k,l) not stated
Approaches that could also have been used
  • Numerous endpoints (body composition, energy expenditure, plasma lipids, tissue uptake, lesion metrics) were compared across groups, each apparently with separate tests and no stated multiplicity control
    Could also: A false discovery rate correction (e.g., Benjamini-Hochberg) applied across related endpoint families could also be used — FDR control is a standard tool for quantifying the expected proportion of false positives among many simultaneously tested hypotheses; it complements the biological coherence of the findings by providing a formal statistical bound
  • Univariate linear regressions were run separately for TG exposure and TC (or (V)LDL-C) exposure as predictors of SQRT lesion area
    Could also: A multivariable regression including both TG and TC simultaneously could also be used — Multivariable regression quantifies the independent contribution of each lipid fraction after adjusting for the other; it is especially informative when TG and TC are correlated, as they often are under the treatment studied
  • Dispersion was reported as mean ± an unlabeled value throughout
    Could also: Explicitly labeling the dispersion measure as SD or SEM, or reporting 95% confidence intervals, would also convey spread — SD describes the distribution of individual values, while 95% CIs directly reflect estimation uncertainty; explicit labeling allows readers to interpret the range of variation and judge the precision of group estimates, particularly for small animal study sample sizes
  • Square-root transformation was applied to lesion area before linear regression
    Could also: A log transformation or a nonparametric rank correlation (e.g., Spearman's rho) could also be applied to address non-normality and heteroscedasticity in lesion area data — Different variance-stabilizing transformations suit different distributional shapes; reporting sensitivity of the association to the choice of transformation, or using a distribution-free approach, adds robustness information
  • Per-group sample sizes (n) are not reported in the text
    Could also: Explicit reporting of n per group, along with a priori power calculations or post-hoc power estimates, could also accompany the statistical results — Reported n allows readers to gauge estimation uncertainty and reproducibility, and pre-specified power calculations contextualize whether the study was designed to detect the primary effect at a given confidence level
  • Time-course plasma lipid data (Fig. 4a,c) spanning 10 weeks were compared between groups, with results summarized as overall percentage reductions
    Could also: A linear mixed-effects model or repeated-measures ANOVA with a post-hoc correction could also formally test treatment-by-time interactions across the longitudinal measurements — Mixed-effects models account for within-animal correlation across time points, yielding valid standard errors under the repeated-sampling design; they also allow testing whether the treatment effect emerged gradually or was stable from the first measurement

What was reproduced

The exact results taken into scope, with each reported value next to the value our attempt produced.

Scope assessment — pmid-25754609

Title: Brown fat activation reduces hypercholesterolaemia and protects from atherosclerosis development Venue: Nature Communications 6:6356 (2015) · DOI 10.1038/ncomms7356 · PMCID PMC4366535 Authors: Berbée JFP, Boon MR, Khedoe PPSJ, Bartelt A, Schlein C, Worthmann A, Kooijman S, Hoeke G, Mol IM, John C, Jung C, Vazirpanah N, Brouwers LPJ, Gordts PLSM, Esko JD, Hiemstra PS, Havekes LM, Scheja L, Heeren J, Rensen PCN.

Verdict: DROP — non_pipeline

This publication contains no pipeline-derived computational results and ships no public high-throughput dataset and no analysis code. It is an experimental (wet-lab) mouse physiology / lipid-metabolism / atherosclerosis study. There is nothing to reproduce computationally from deposited data + code, which is the defined scope of this study.

Evidence (read from PMC full text + PubMed)

  • No omics / high-throughput data. Methods describe only targeted molecular biology and physiology. Gene expression is by qRT-PCR (normalized to β2-microglobulin and 36b4), not RNA-seq or microarray. No sequencing, proteomics, or array profiling anywhere.
  • No data accession. No GEO / ArrayExpress / SRA / ENA / PRIDE / figshare / Zenodo accession appears in the paper. No "Data availability" / "Accession codes" statement exists (2015 paper, predates Nat Commun mandatory data statements).
  • No code. No repository, no script, no computational workflow is referenced.
  • Statistics: unpaired two-tailed Student's t-test and univariate regression, computed in SPSS 20.0 for Windows on directly measured bench values — not a reproducible data+code pipeline.

Reported results, and why each is OUT OF SCOPE

Result class Method (paper) In scope? Why
Plasma TG / total / HDL / LDL cholesterol enzymatic colorimetric assays no instrument/bench measurement; no deposited data
Lipoprotein distribution FPLC fractionation no instrument output, manual integration; not deposited
Atherosclerotic lesion area / severity histology + histomorphometry no manual image quantification; no images/data deposited
BAT/organ uptake of TG-derived FA / cholesterol [3H]/[14C] radiotracer kinetics no radioisotope bench assay; no deposited data
Energy expenditure / fat & carb oxidation indirect calorimetry no instrument output; no deposited data
Body composition EchoMRI no instrument output
Hepatic gene expression qRT-PCR no targeted bench assay; no deposited data/code
UCP1 staining, lipid-droplet size immunohistochemistry / morphometry no manual histology

Every reported number originates from a bench instrument or manual quantification and was analysed in SPSS. None is regenerable from a public dataset + pipeline.

Pipelines named per in-scope result

None. There are no in-scope (pipeline-derived) results.

Decision

Record drop_reason = non_pipeline (root cause); secondary gaps no_data_accession + no_code also hold. No «our HPC» compute spent — by design, since there is no data or code to run. Honest drop per BRIEF rule 6.

Figures / tables: Fig.2Fig.1figure

No individual results have been recorded for this entry yet.

Assessments & scoring basis

Each contributor’s verdict, the per-question basis, and the auditable, itemised worksheet behind it.

🤖 AI curator · claude (ai-curator room) · v1.0 L1 50/100

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.

🔴1. Data identity
🔴2. Endpoint comparability
🟡3. Location of the main deviation
🟢4. Cause of the deviation
🟡5. Derivability / plausibility
🟢6. Severity of the deviation
🟡7. Core claim
🟡8. Severity of the miss (overall human judgment)
Scoring basis — itemised

Every item that counted toward this verdict, and the exact part of the reproduction that produced it.

Supporting (toward a concern)
Content-critical question only partially held
+2 pts
From: Q5 · Derivability / plausibility 🟡
Content-critical question only partially held
+2 pts
From: Q7 · Core claim 🟡
Content-critical question only partially held
+2 pts
From: Q8 · Severity of the miss (overall human judgment) 🟡
Minor / cosmetic deviation
+1 pts
From: Q3 · Location of the main deviation 🟡
Input / endpoint not comparable 1:1
+1 pts
From: Q1 · Data identity 🔴
Total score +8

This is an honest non-pipeline DROP: Berbée et al. (Nat Commun 6:6356, 2015) is a wet-lab mouse atherosclerosis/lipid study whose claims C1–C6 are enzymatic, radiotracer, calorimetry and histomorphometry measurements analysed in SPSS 20.0, with no deposited dataset and no code. The barrier sits entirely on data/code availability (q1/q2 red), not on the authors' methodology or any suspicious value — so q5 is yellow (untestable, not fabrication-suspect) and q7 yellow (central claim neither confirmed nor refuted). Nothing was computed, hence no factual deviation (q6 green); overall a non-critical out-of-scope drop that a human should simply log rather than treat as a discrepancy.

🤝
Reproduced automatically — and fairly

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-4-8

Measured resources invested to assess this paper — sanitised (machine class only, no job ids/paths). Compute = HPC accounting (SLURM); tokens = the AI agent's session.

44.3 k
tokens (I/O) · 2.7 M incl. cache
5 min
runtime
Per-job HPC accounting not captured for this run — the runtime shown is the reproduction’s measured wall-clock time.