Cell type-specific proximity labeling of organ secretomes reveals energy balance-dependent proteomic remodeling.
Kaja PlucińskaiD, Charlotte R. WayneiD, Henry SanfordiD, Boby MathewiD, Nathalie RopekiD, Stephanie M. AdaniyaiD, Corey ModeliD, Nicolás Gómez-BanoyiD, Ksenia Morozova, Xiongwen CaoiD, Jeffrey M. FriedmaniD, Ken H. LohiD, Paul CoheniD, Ekaterina V. VinogradovaiD
Abstract
Intercellular communication is critical for maintaining organismal metabolic homeostasis. Here, we present a new method enabling temporally controlled, cell type-specific labeling of secreted and membrane proteins in key metabolic tissues. The method employs a genetically encoded proximity-labeling strategy by targeting a Cre-dependent TurboID ligase to the endoplasmic reticulum (ER) in ES cell-derived mice. Expression of TurboID in liver, adipose tissue, and spleen enabled the characterization of organ-specific ER proteomes at baseline and in response to fasting, inflammation, and dietary obesity, revealing tissue-and perturbation-specific changes and augmenting our understanding of how the proteomes of individual tissues change to regulate systemic energy balance. This comprehensive resource represents an important advance toward understanding both how cell-to-cell communication changes in response to energy homeostasis and how it contributes to these alterations. This method is broadly applicable and provides a means for identifying biomarkers and therapeutic targets across a wide range of tissues.
02 /Specification & Classification
Click any tag to browse matching papers- Citations
- 4
- FWCI
- 13.58
- Field percentile
- top 1%
- Organism
- Assay
- Research Focus
- Approach
In-depth summary
Claude Sonnet 4.6 · xhigh effort · 8 Nov 2026 · from the full text of version 1 (12 Jan 2026) · CC BY 4.0
A Cre-dependent ER-targeted TurboID knock-in mouse enables cell type-specific secretory proteomics, mapping adipocyte and hepatocyte responses to fasting, LPS and obesity, with B cells profiled at baseline.
Conventional mass spectrometry cannot trace which cell type secretes a low-abundance circulating protein, and AAV-delivered proximity labeling brings off-target and inflammatory artifacts. This work replaces viral delivery with a germline Cre-dependent TurboID-KDEL allele and uses it to map secretory proteomes across energy states.
Key results
ShownER-targeted TurboID labeling with TMT proteomics and ROC-based filtering recovered secretory pathway proteomes from adipose tissue and liver, with many enriched proteins carrying predicted signal peptides.
iWAT: 2,460 proteins identified, 1,550 passed ROC threshold, 376 with SignalP signal peptides. Liver: 2,608 identified, 1,318 passed, 343 with signal peptides. Cre-negative controls were used to exclude background. · Cell type-specific ER-proximity labeling; In vivo models for profiling secretory pathway…
ShownFasting and LPS remodel adipocyte and hepatocyte secretory proteomes differently, and LPS drives the broader change in both tissues. In liver, fasting favors gluconeogenic and fatty acid metabolism proteins, while LPS raises acute-phase proteins.
Differentially expressed proteins (p < 0.05, FC > 1.5): iWAT fasting 181; iWAT LPS 341 down and 131 up; liver fasting 142; liver LPS 340. Liver LPS: SAA1 4.7 log2FC. Western blots confirmed CES1D and C6 decreases in liver. · Fasting and LPS trigger distinct remodeling of…; Fasting and LPS trigger distinct remodeling of…; Shared and tissue-specific responses to fasting…
ShownCell type-specific plasma labeling captured circulating proteins across a wide concentration range, including sub-nanomolar proteins. Tissue-plasma concordance was higher for hepatocytes than adipocytes.
279 detected proteins overlapped a published plasma atlas, spanning APOA1 at 47.5 nmol/mL to TSP4 at 0.7 pmol/mL. Tissue-plasma Pearson R: liver 0.74 (LPS) and 0.68 (fasting); adipose 0.51 and 0.54. · Changes in circulating hepatokines and…
ShownDiet-induced obesity changed the adipocyte secretory proteome in a depot- and duration-dependent way, with larger changes at 15 weeks. GPR50 was among the most strongly induced proteins, and qPCR matched the protein-level regulation for the genes tested.
Significant proteins (p < 0.05, FC > 1.5): 15 weeks 668 iWAT and 809 eWAT; 6 weeks 252 iWAT and 431 eWAT. At 15 weeks GPR50 was up 2.3 log2FC (iWAT) and up to 5.0 log2FC (eWAT). · The white adipocyte tissue secretory proteome in…; The white adipocyte tissue secretory proteome in…; Depot-specific differences in DIO mice
SuggestedLinking energy-balance-regulated plasma proteins to UK Biobank data found associations with cardiometabolic disease for several of them, including gamma-synuclein (SYUG), which is elevated in obesity. These are correlations between basal human plasma levels and prevalent disease only.
65 regulated proteins had disease associations in a 53,026-person cohort. T2D: SYUG p = 3.66 x 10^-31, OR 1.64; SERPINF1 p = 3.47 x 10^-40, OR 4.5; APOA4 p = 1.23 x 10^-76, OR 3.23. · Secreted factors impacted by energy balance…
Methods, data and reproducibility
Cre-dependent V5-TurboID-KDEL was knocked into Rosa26 and crossed to Adiponectin-, Albumin- and CD19-Cre lines. Biotinylated proteins were streptavidin-enriched from tissue or plasma, TMTpro 16-plex labeled, and quantified by LC-MS/MS/MS with ROC filtering of true ER/secreted proteins.
- Mouse TurboID-KDEL proteomes of iWAT, liver and plasma under basal, 48 h fasting and LPS (1 mg/kg i.p.) conditions, plus basal spleen · mouse (C57BL/6J, male) · n = n = 6-8 mice per condition (physiology)
- iWAT, eWAT and plasma proteomes of Adipo-TurboID-KDEL mice fed chow or 60% HFD for 6 or 15 weeks · mouse (C57BL/6J, male) · n = minimum n=4 per condition; n = 5 mice per condition (plasma)
- Published mouse plasma absolute concentrations from multiple reaction monitoring (Michaud et al. 2018) used for dynamic range comparison · mouse · n = 385 target analytes; 243 detected here
- UK Biobank Olink plasma proteome-phenome atlas of prevalent disease associations (Bonferroni-corrected) · human (UK Biobank) · n = 53,026 individuals
- Methods & tools
- TurboID proximity labeling, TMTpro 16-plex, Orbitrap Eclipse LC-MS/MS/MS, Integrated Proteomics Pipeline (IP2), ProLuCID, DTASelect v2.0, ROC-based enrichment filtering, SignalP and OutCyte, GOATOOLS (GO Slim, Fisher's exact test), CRAPome contaminant filter, GraphPad Prism and R, Proteome-phenome atlas (UK Biobank)
- Data & accessions
- http://proteome-phenome-atlas.com/
- Reagents & environment
- Mouse UniProt release-2017_07 database; DTASelect v2.0; GO Slim file (goslim_generic.obo) downloaded April 24, 2024; peptide-level FDR below 1%.
- Reproducibility gaps
- Raw proteomic data and processing code are stated as 'will be deposited' to PRIDE and GitHub, with no accession or URL given. Mouse lines require a material transfer agreement.
Limitations (5)
- MS coverage depends on TurboID expression levels, which vary by Cre driver, and on the cellular heterogeneity within organs. · stated by the authors
- Very low-abundance proteins and polypeptides from non-canonical open reading frames remain hard to detect with current instruments and pipelines. · stated by the authors
- The proteomic data nominate candidates only; functional studies are needed to establish causal roles, for example for SYUG, GPR50 and G6PE. · stated by the authors
- All animals were male, and differential expression used nominal p < 0.05 with FC > 1.5 (two-sample t-test). No protein-level multiple-testing correction is described, so false positives are likely in the large hit lists. · noted in this summary
- The UK Biobank links are cross-species and cross-sectional (basal human plasma levels vs prevalent disease). They do not test whether the mouse energy-balance changes occur in humans. · noted in this summary
Who should read:Read this if you want a genetically encoded, AAV-free method for cell type-specific secretome proteomics in vivo, or a resource of adipocyte and hepatocyte responses to fasting, LPS and obesity.
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
- Dynamic tracking and identification of tissue-specific secretory proteins in the circulation of live mice
- Cell type-selective secretome profiling in vivo
W. Wei, N. M. Riley, … J. Z. Long·bioRxiv·Sep 18, 2020
- Proteomics of protein trafficking by in vivo tissue-specific labeling
I. A. Droujinine, D. Wang, … N. Perrimon·bioRxiv·Apr 15, 2020
- Organism-wide secretome mapping uncovers pathways of tissue crosstalk in exercise
Wei Wei, Nicholas McIlvain Riley, … Jonathan Z. Long·bioRxiv·Nov 21, 2022
- Method to Identify Liver-to-Brain Region-Specific Crosstalk Using TurboID
Matthew C. Juber, Arvand Asghari, … Matthew J. Potthoff·bioRxiv·Sep 7, 2026
- Split-TurboID enables contact-dependent proximity labeling in cells
K. F. Cho, T. C. Branon, … A. Y. Ting·Proceedings of the National Academy of SciencesPeer-reviewed·May 18, 2020