Samir Jana

CAF exosomes directly inhibit T-cell proliferation. A, Schematic of experimental design. B and C, Proliferation of CFSE-labeled murine T cells 72 hours after concomitant stimulation (Stim.) with anti-CD3/CD28 antibodies and 10 μg/1 × 105 cells cancer cell or CAF exosomes (Exo.) from the MCaP line. Unstimulated (Unstim.) T cells used as a negative control. Dot plot of CD8 T cells shown in B. Proliferation of gated CD8 and CD4 T cells shown in C. D, Representative contour plot of granzyme B+ CD8+ T cells in the presence of anti-CD3/CD28 antibodies and cancer cell or CAF exosomes from the MCaP line for 72 hours. Quantification shown in E. F, Cancer cell or CAF exosomes (10 μg/1 × 105 cells) from 4T1 tumor explants were added (10 μg/1 × 105 cells) to single-cell suspensions from lymph nodes along with anti-CD3/CD28 antibodies. Proliferation of gated CD8 and CD4 T cells after 72 hours shown. G, Exosomes from human normal fibroblasts (NF) or human luminal CAFs were added (10 μg/1 × 105 cells) to single-cell suspensions from human PBMCs concomitantly with anti-human CD3/CD28 antibodies. Proliferation of gated CD8 and CD4 T cells shown. Experiments depict three technical replicates from at least three biological replicates. Significance was tested using one-way ANOVA with Tukey honestly significant difference post hoc test. [A, Created in BioRender. Jones, D. (2026) https://BioRender.com/nbcfuj2.]

Cancer-associated fibroblasts (CAF), a major component of the breast tumor microenvironment, drive immune evasion in various cancers by promoting T-cell exclusion and dampening T-cell activation. Previous studies have implicated CAF-derived soluble factors in mediating these immunosuppressive effects. In this study, we investigated whether exosomes secreted by CAFs could suppress T-cell activity. Inhibition of global exosome secretion in breast tumor–bearing mice significantly reduced tumor growth and increased tumor-infiltrating T cells with lower exhaustion marker expression. Conversely, administration of CAF-derived exosomes into tumors produced the opposite effects. Moreover, CAF exosomes associated with T cells in vivo and impaired T-cell activation and cytotoxic potential in ex vivo assays. Proteomic and biochemical analyses of T cells exposed to CAF exosomes revealed dampened early T-cell receptor signaling. Mass spectrometry identified an extracellular matrix (ECM) signature on CAF exosomes. Depleting type I and type V collagens from CAF exosomes restored T-cell proliferation, whereas overexpression of collagen in cancer cells led to its incorporation into exosomes, which suppressed T-cell activation. These findings suggest that a signaling bridge between CAF exosomes and T cells, mediated by collagen, promotes T-cell dysfunction, contributing to immune evasion in breast cancer.Significance:Our data provide the first evidence that ECM proteins associate with mouse and human breast CAF-derived exosomes and directly impair T-cell activation and cytotoxicity. These findings suggest that signaling between collagen-rich CAF exosomes and T cells contribute to local and systemic T-cell dysfunction.

Intratumoral blockade of exosome secretion significantly reduces tumor growth.

Purified extracellular matrix is sufficient to inhibit CD8+ T cell proliferation in vitro.

ECM from CAF exosomes is critical for attenuating T-cell proliferation. A, Western blot of COL1A1 and COL1A2 in MCaP cancer cells (top) and derived exosomes (bottom) from parental cells (transduced with vector control) or cells overexpressing (OE) COL1A1 and or COL1A1/COL1A2 vectors. B, Exosomes from MCaP cancer cells expressing vector control (VC), COL1A1, and/or COL1A1/COL1A2 were added to CFSE-labeled T cells along with anti-CD3/CD28 antibodies. CFSE dilution was measured by flow cytometry 72 hours later. C and D, Exosomes from cohorts in B were added to isolated CD8 T cells along with anti-CD3/CD28 antibodies. Flow cytometry measurements of granzyme B (C) and IFNγ (D) were made after 72 hours. Significance was tested using one-way ANOVA with Tukey honestly significant difference post hoc test. E, Quantification of serum exosomes during MCaP tumor progression at 12 weeks after implantation. F, Western blot analysis of epithelial, ECM, and exosomal marker proteins from purified serum exosomes of tumor-naïve mice (n = 4) and mice bearing MCaP tumors (n = 5) 12 weeks after implantation. G, Single-cell suspensions from lymph nodes of naïve BALB/c mice were incubated for 72 hours with serum exosomes from tumor-naïve mice and mice bearing MCaP tumors 12 weeks after implantation. Flow cytometry was performed to measure the percentage of CD8 (left) and CD4 (right) T cells proliferating in response to concomitant exosome and anti-CD3/CD28 antibody stimulation. H, Growth curve of MCa-P1362 tumors injected weekly with PBS or 20 μg weekly of CAF exosomes. I, Quantification of CD8 (left) and CD4 (right) T cells in tumor-draining lymph nodes and primary tumors of PBS-injected and CAF exosome–injected mice. P value calculated by unpaired Student t test (E, G, and I) and two-way ANOVA (H). *, P < 0.05; ****, P < 0.0001.

Exosomal ECM from CAFs attenuates T-cell proliferation. A, Mean difference plot showing the log fold change and average abundance of each protein from CAF exosomes. Proteins with fold changes >1 are highlighted [red, up after proteinase K (PK) or blue, down after PK treatment]. Top 10 labeled. B, Gene set enrichment analysis identified proteins associated with ECM organization in CAF exosomes. C, Western blot of COL1A1 and COL5A2 proteins in cells (top) and in exosomes (bottom) of normal fibroblasts (NF) from the MFP of naïve animals and CAFs from the MCaP murine cell line. D, Western blot of COL1A1 and COL5A2 proteins in cells (top) and in exosomes (bottom) of untransformed normal human fibroblasts (NF) and CAFs from luminal breast cancer. E and F, Protein content based on BCA assay (E) and extracellular vesicle composition (F) of fractions 1–12 collected by iodixanol density gradient centrifugation. G, Western blot of fractions from ultracentrifugation, probed with antibodies against COL1A1, COL5A2, exosome markers Alix and CD63, and the serum protein albumin. H, Western blot of COL1A1 (top) and COL5A2 (bottom) in MCaP CAFs with empty vector (vector) or targeted sgRNA (sg) vectors. I, Exosomes from MCaP CAFs with empty vector or targeted sgRNA constructs were added to CFSE-labeled T cells along with anti-CD3/CD28 antibodies. CFSE dilution in T cells was measured by flow cytometry 72 hours later. J, Western blot of CAF exosomes after collagenase I digestion for 1 hour at indicated concentrations. K, CAF exosomes after digestion with collagenase I at indicated concentration were added to CFSE-labeled T cells along with anti-CD3/CD28 antibodies. CFSE dilution was measured by flow cytometry 72 hours later. Significance was tested using one-way ANOVA with Tukey honestly significant difference post hoc test (I and K). L, Naïve CD8+ T cells isolated from Balb/c mice were activated with anti-CD3/CD28 antibodies and nucleofected 72 hours later with Cas9 and sgRNAs targeting SHP2, DDR1, or LAIR-1, or with mock nucleofection (mock). Knockdown efficiency was assessed by Western blot. M, A subset of T cells from each cohort described in (L) was labeled with CFSE and subsequently stimulated with CAF-derived exosomes and anti-CD3/CD28 antibodies. Proliferation was assessed by flow cytometry 72 hours later. CFSE-labeled T cells stimulated with anti-CD3/CD28 antibodies alone served as positive controls (Stim.), whereas unstimulated cells (“Unstim.”) served as negative controls. Significance was tested using one-way ANOVA with Tukey honestly significant difference post hoc test.

Genetic or pharmacologic reduction of CAF exosome production partially restores CD8+ T cell proliferative capacity.

T cells associate with breast tumor exosomes in vitro and in vivo. A, Representative NanoSight characterization showing size distribution and concentration of exosomes purified from mammary adenocarcinoma-P1362 (MCaP). Mean (black trace) ± (red trace) SEM data are shown, with the size of the key peaks annotated. B, Representative electron microscopy image of MCaP exosomes. Scale bar, 100 nm. C, Western blot of indicated proteins from lysates of cells and exosomes (Exo.) from MCaP and 4T1 cancer cells. D, Confocal microscopy image of murine T cells incubated for indicated time points with PKH26-labeled exosomes for 30 minutes. Control (0′) shown of left. Scale bar, 2.5 μm. E, Flow cytometry to assess PKH26-labeled exosome uptake by CD3/CD28-stimulated CD8 T cells 24 hours after incubation (red shade) compared with untreated T cells (black shade). F, 20 μg of PKH26-labeled exosomes (red) were injected into the second MFP of naïve BALB/c mice. After 24 hours, the draining axillary lymph node was fixed, sectioned, and stained with an anti-CD3 antibody to detect T cells (green). DAPI (blue) detects all nucleated cells. Scale bar, 15 μm. G, 30 μg of PKH26-labeled exosomes (red) were injected into the footpads of naïve BALB/c mice 6 days after control (right footpad) or clodronate liposome (left footpad) injection. Twenty-four hours after exosome injection, single-cell suspensions from the draining popliteal lymph node were stained with an anti-CD3 APC antibody to detect PKH26+ T cells by flow cytometry. H and I, On days 14 and 16 after implantation (pooled from two separate experiments), uptake of CD63-mCHERRY exosomes by T cells in primary tumor of mice bearing 4T1 cells transfected with CD63-mCHERRY constructs was measured by flow cytometry. Tumors formed from nontransduced parental 4T1 cells used as a comparator. P value calculated by unpaired Student t test (G and I). ***, P < 0.001; ****, P < 0.0001.

Density gradient fractions from conditioned medium are characterized by exosome markers and specific fractions inhibit T cell proliferation.

Pharmacologic inhibition of SHP-2, but not SHP-1, partially relieves CAF exosome mediated suppression of T cell proliferation.