henhao
Sign in
Cancer BiologybioRxiv
Preprint

De novo lipogenesis controls antigen presentation and metastatic immune escape in triple-negative breast cancer.

Myung Soo KoiD, Henry Withers, Divya Ramchandani, Jason McCormickiD, Sebastian E. CarrascoiD, Pius Amartey, Arshdeep SinghiD, Liron YoffeiD, Glenn E. Simmons Jr.iD, Guoan Zhang, Wen Hong SheniD, Vivek MittaliD

PDFDOICodeData
Post to X
Threads on XPost to BlueskyThreads on Bluesky
In-depth full-text digest available · Jump to digest ↓

Abstract

Metastatic dissemination is a major cause of mortality in triple-negative breast cancer (TNBC), yet how tumor metabolic state regulates immune surveillance during metastasis remains poorly understood. Here, we identify de novo lipogenesis as a tumor-intrinsic determinant of immune visibility and metastatic progression. Genetic or pharmacologic disruption of fatty acid synthase (FASN) suppresses metastasis with limited effects on primary tumor growth and enhances tumor-cell major histocompatibility complex class I (MHC-I) antigen presentation and CD8+ T cell activation. Mechanistically, FASN-dependent lipogenesis sustains epidermal growth factor receptor (EGFR) palmitoylation, selectively maintaining PI3K-AKT-mTOR signaling without altering EGFR abundance, cell-surface localization, or MAPK activation. Disruption of FASN-dependent lipogenesis, EGFR palmitoylation, or mTOR signaling restores peptide-MHC-I presentation and enhances antigen-specific CD8+ T cell recognition and cytotoxicity. FASN loss increases the abundance of multiple antigen-processing machinery proteins without corresponding changes in MHC-I transcript abundance, supporting transcription-independent regulation of antigen presentation downstream of PI3K-AKT-mTOR signaling. Tumor-cell B2M depletion substantially restores metastatic outgrowth and circulating tumor cells following FASN loss, identifying MHC-I antigen presentation as a major mediator of lipogenesis-dependent metastatic immune escape. In patient-derived TNBC organoids and independent human breast cancer cohorts, elevated lipogenic gene expression correlated with PI3K-AKT-mTOR-related transcriptional programs and reduced immune-surveillance programs, without significant differences in MHC-I transcriptional signatures. Together, these findings define a tumor-intrinsic immunometabolic circuit in which de novo lipogenesis sustains EGFR palmitoylation and PI3K-AKT-mTOR signaling to regulate antigen presentation, thereby modulating tumor-cell immune visibility during metastatic progression.

02 /Specification & Classification

Click any tag to browse matching papers
Citations
0
OpenAlex Topics

In-depth summary

✦ Multi-agent verified full-text synthesis

Claude Sonnet 4.6 · xhigh effort · 8 Nov 2026 · from the full text of version 4 (30 Oct 2026) · CC BY 4.0

FASN-driven lipogenesis sustains EGFR palmitoylation and PI3K-AKT-mTOR signaling, suppressing tumor MHC-I antigen presentation; blocking it curbs TNBC metastasis in mice via CD8 T cell recognition.

How tumor metabolic state couples to antigen presentation during metastasis was unclear, and it was unknown whether FASN-derived palmitate acts through protein palmitoylation. This work links lipogenesis to EGFR palmitoylation, selective PI3K-AKT-mTOR signaling and MHC-I protein abundance without MHC-I transcript changes.

Key results

  1. ShownGenetic loss of FASN strongly reduced spontaneous lung metastasis and improved survival in immunocompetent mice, and TVB-3166 also lowered lung metastatic burden, while primary tumor growth was barely affected.

    Kaplan-Meier survival n = 15 per group, log-rank P = 0.0005; histology of lung burden n = 9 mice; BLI n = 20 mice; TVB-3166 given at 5 mg/kg daily (BLI n = 8 mice) also lowered lung metastasis with little effect on primary tumors. · De novo lipogenesis is required for metastatic…; In vivo FASN inhibition; Figure legends

  2. ShownFASN loss raised MHC-I and antigen-processing proteins and surface MHC-I without changing MHC-I transcripts. It also increased SIINFEKL-H-2Kb presentation and OT-I CD8 T cell killing of OVA-expressing cells.

    Proteomics (3,516 proteins identified) showed higher MHC-I and antigen-processing proteins in FASN KO tumors; H2-K1, H2-D1 and B2M qPCR unchanged (n = 5); sorted GFP+ cell surface MHC-I n = 17 WT vs 18 KO. · De novo lipogenesis suppresses tumor antigen…; Global proteomic analysis; Figure legends

  3. ShownDoxycycline-inducible B2M knockdown largely restored lung metastatic outgrowth and circulating tumor cells in FASN-deficient tumors. It also dampened the CD8 T cell activation, implicating MHC-I presentation as a major mediator.

    Lung metastatic burden n = 7 FASN WT, 9 FASN KO, 7 FASN KO/B2M KD mice; CTC counts n = 5, 5 and 7; IFNg+ CD8 T cells n = 5, 5 and 6. The text gives no numeric effect size. · De novo lipogenesis suppresses tumor antigen…; Orthotopic model with inducible B2M knockdown; Figure legends

  4. ShownPalmitoylation-deficient EGFR (C1025A, C1122A) mimicked FASN loss, lowering PI3K-AKT-mTOR signaling with MAPK intact and raising surface MHC-I; OVA assays with C1025A showed stronger OT-I responses. Rapamycin and Torin1 raised MHC-I, with unchanged transcripts shown for rapamycin.

    Rapamycin (20 nM) and Torin1 (100 nM) raised surface MHC-I; EGFR mutants had normal total and surface EGFR; OT-I killing and cytokine assays n = 4 per group; B2M knockdown abolished the T cell effects; HLA transcripts unchanged. · EGFR palmitoylation links de novo lipogenesis to…; Signaling assays; Figure legends

  5. SuggestedIn basal-like breast tumors, higher lipogenesis scores tracked with PI3K-AKT-mTOR programs and lower immune-response signatures, and in TNBC organoids FASN correlated with RPTOR. TCGA MHC-I signature scores did not differ. These are correlative.

    TCGA basal-like n = 166 (77 Lipo-high, 89 Lipo-low); METABRIC n = 329 (175 high, 154 low); FASN-RPTOR in tumor organoids r = 0.758, P = 0.0294, n = 8 (normal r = 0.724, P = 0.1663, n = 5); MHC-I signature P = 0.157. · A lipogenic−PI3K−AKT−mTOR state is associated…; TCGA and METABRIC analysis; Figure legends

Methods, data and reproducibility

FASN or EGFR knockouts, EGFR palmitoylation-site mutants, pharmacologic inhibitors and inducible B2M knockdown were tested in orthotopic murine EO771-ML1 tumors and in human and mouse cell lines, with RNA-seq, proteomics, LC-MS lipid profiling and OT-I T cell assays. Human TCGA, METABRIC and organoid transcriptomes were analyzed by GSVA and GSEA.

  • Orthotopic EO771-ML1 tumor and metastasis experiments in immunocompetent mice (FASN WT/KO, TVB-3166, B2M knockdown) · mouse (C57BL/6J) · n = n = 10 per group at implantation; analysis group sizes vary
  • Human MDA-MB-231-LM2 and MDA-MB-231 cell models (FASN and EGFR genotypes, mTOR inhibitors) · human · n = n = 3 to 8 biological replicates per assay
  • RNA-seq of FASN WT/KO tumors and sorted GFP+/GFP- tumor cells · mouse · n = n = 5 biological replicates per group · GSE347931
  • Label-free proteomics of FASN WT and KO primary tumors · mouse · n = 3,516 proteins identified · PXD084210
  • RNA-seq of patient-derived TNBC tumor organoids and normal breast organoids · human · n = n = 8 tumor, n = 5 normal · GSE349342
  • Public scRNA-seq atlas of primary breast tumors (Single Cell Portal) · human · n = 26 tumors, 130,246 cells · GSE176078
  • TCGA primary basal-like breast tumors, lipogenesis-high vs low · human · n = n = 166 (77 high, 89 low)
  • METABRIC PAM50 basal-like tumors, lipogenesis-high vs low · human · n = n = 329 (175 high, 154 low)
Methods & tools
CRISPR/Cas9 knockout, TVB-3166 and TVB-2640 (FASN inhibitors), Acyl-biotin exchange assay, OT-I CD8 T cell cytotoxicity assay, LC-MS lipid profiling (MS-DIAL v4.9), MaxQuant v2.1, STAR v2.7.9a, DESeq2, clusterProfiler / fgsea, GSVA v2.6.6, TIMER2.0 and xCell (immunedeconv v2.1.4), GraphPad Prism v11.0.1
Code
https://github.com/omnideconv/immunedeconv
Data & accessions
https://www.cancer.gov/tcgahttps://compbio.cn/timer2/
Reagents & environment
Mouse RNA-seq: STAR v2.7.9a, DESeq2 v1.38.3, clusterProfiler v4.6.2. TCGA: TCGAbiolinks 2.40.0, DESeq2 1.52.0, fgsea 1.38.0, GSVA 2.6.6. METABRIC: MetaGxBreast v1.32.0, limma 3.68.5. Also MaxQuant v2.1 and FlowJo v10/v11.
Reproducibility gaps
No custom code was developed. Other data are available from the corresponding author on reasonable request. The reagent, primer and guide tables are in supplementary tables.
Limitations (6)
  • Newly synthesized palmitate was not traced into palmitoylated EGFR, so it is unresolved whether it directly supplies the substrate or maintains a separate lipid pool. · stated by the authors
  • How PI3K-AKT-mTOR signaling controls antigen-processing machinery abundance is unresolved; post-transcriptional synthesis or degradation, and trafficking, are suggested but untested. · stated by the authors
  • After direct vascular delivery, FASN-deficient cells gave comparable lung tumor burden at the examined endpoint, although CD8 T cell and dendritic cell infiltration rose. · noted in this summary
  • Whether FASN inhibition improves checkpoint blockade in vivo, and how systemic metabolic and immune effects shape responses, remains to be determined. · stated by the authors
  • Human datasets do not establish EGFR palmitoylation or post-transcriptional MHC-I regulation; they only show transcriptional associations. · stated by the authors
  • Mechanistic and immune work rests mainly on EO771-ML1 and MDA-MB-231-LM2 lines, and the organoid correlation uses 8 tumor and 5 normal samples, so power and generality are limited. · noted in this summary

Who should read:Read this if you study tumor metabolism, antigen presentation or metastatic immune escape and want evidence linking FASN, EGFR palmitoylation and mTOR to MHC-I in TNBC.

Reproducibility ledger

Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.

Sign in to file a reproducibility report.

Related papers

Similarity Map
SelectedSame fieldCross-field
THISSteigerwald '23Stirling '23Lorito '23Okada '22Sinha '23Sim '24
69% semantic matchCancer Biology
H. Steigerwald, T. Bozzetti, M. Tams+14 more·bioRxiv·Oct 1, 2023