Research ArticleCELL BIOLOGY

Mechanisms of nuclear content loading to exosomes

See allHide authors and affiliations

Science Advances  20 Nov 2019:
Vol. 5, no. 11, eaax8849
DOI: 10.1126/sciadv.aax8849
  • Fig. 1 Characterization of nuclear-derived content from ovarian cancer exosomes.

    (A) TCGA pan-cancer ploidy analysis of 20 cancer types. n = 62 [kidney chromophobe (KICH)], n = 418 [brain low-grade glioma (LGG)], n = 7 [pancreatic cancer (PAAD)], n = 138 [pheochromocytoma (PCPG)], n = 353 [prostate adenocarcinoma (PRAD)], n = 184 [thyroid carcinoma (THCA)], n = 543 [glioblastoma (GBM)], n = 415 [kidney clear cell carcinoma (KIRC)], n = 61 [uveal melanoma (UVM)], n = 415 [uterine endometrial carcinoma (UCEC)], n = 257 [skin cutaneous melanoma (SKCM)], n = 501 [head and neck squamous carcinoma (HNSC)], n = 155 [kidney papillary carcinoma (KIRP)], n = 330 [stomach adenocarcinoma (STAD)], n = 940 [breast cancer (BRCA)], n = 187 [liver hepatocellular carcinoma (LIHC)], n = 396 [colon adenocarcinoma (COAD)], n = 34 [cervical cancer (CESC)], n = 85 [adrenocortical carcinoma (ACC)], n = 158 [renal adenocarcinoma (READ)], n = 435 [lung squamous carcinoma (LUSC)], n = 544 [ovarian cancer (OV)], n = 429 [lung adenocarcinoma (LUAD)], n = 144 [bladder cancer (BLCA)], and n = 55 [uterine carcinosarcoma (UCS)]. (B) Cryo-EM image of the exosomes isolated from OVCAR-5 cells. Scale bars, 100 nm. (C) NTA for the exosomes isolated from OVCAR-5 cells. (D) Western blot analysis of exosome markers in OVCAR-5. TSG101, Alix, and CD63 are used as exosome markers, and GRP94 is used as a marker of cellular contamination. TCL, total cell lysate. (E) Pie chart of cellular compartment proteins resulting from MS analysis in OVCAR-5 cell–derived exosomes. Nuclear components are highlighted in red: 1, endoplasmic reticulum; 2, endosome; 3, Golgi; 4, cell surface; 5, mitochondrion; 6, proteasome; 7, vacuole; 8, spliceosomal complex. (F) Counts of the cellular compartment origin of proteins resulting from MS analysis in OVCAR-5 cell–derived exosomes. The x axis represents the categories of cellular compartments. Nuclear proteins identified in chromosome and nucleus are highlighted in red. (G) CNVs of both the exosomal DNA (inner red circle) and cellular DNA (outer blue circle), both derived from OVCAR-5 cells, are displayed on a chromosome map generated using Circos (v0.69.3). The outermost circle represents human chromosomes with coordinates (megabases). The green and red histograms inside the blue and red inner circles represent copy number alterations identified by cnvkit. The larger the bar on the track, the larger the copy number alteration (log scale). Green bars represent amplification events, and red bars represent deletions. (H) A Venn diagram of all the CNVs overlapping between the exosomal and cellular DNA derived from OVCAR-5 cells. (I) Representative plots of OVCAR-5 exosomes from flow cytometry analysis. Top left: Particles are shown as black dots, and exosomes are in the green area. Right: Each dot indicates single exosomes stained with CellMask Green (Ch02), and the red gate indicates DNA-positive particles stained with DRAQ5 (Ch11). Bottom left: Snapshots of individually stained exosomes. (A) and (B) are the exosomes present in the areas indicated in the right panel. (A) represents the DNA-positive exosomes, and (B) represents the negative exosomes. (J) Representative gate images of OVCAR-5 exosomes from imaging flow cytometry analysis. Left: Each green dot indicates a single exosome, and the blue gate indicates a Lamin A/C–positive population. Right: All dots are from DNA-positive exosomes, and the green gate indicates a Lamin A/C–positive population.

  • Fig. 2 Promoting MN formation increases DNA-carrying exosomes.

    (A) Immunofluorescence (IF) images of FTE and OVCAR-5 cells. The inset shows a magnified image of a micronucleated OVCAR-5 cell in the right panel. Scale bars, 50 μm. (B) Quantification of MN cells in FTE, OVCAR-5, and OVCAR-8 cells. MN counting is described in Materials and Methods. The experiment was performed in three independent biological replicates, and the average of the fold changes was calculated. Error bars are represented as SD. Statistical significance was determined by conducting an unpaired Student’s t test. (C) Representative images of FTEexo and OVCAR-5exo from imaging flow cytometry analysis. The gates in both graphs indicate the DNA-positive population. (D) Population of DNA-positive exosomes in FTEexo, OVCAR-5exo, and OVCAR-8exo. The experiment was performed in three independent biological replicates, and the average of the fold changes was calculated. Error bars represent SD. Statistical significance was determined by conducting an unpaired Student’s t test. (E) Representative images of nuclei from OVCAR-5 cells. Nuclei were IF-stained with Lamin A/C antibody. Scale bars, 50 μm. (F) Quantification of MN cells in OVCAR-5 and OVCAR-8 cells treated with DMSO, topotecan, and olaparib. The experiment was performed in three independent biological replicates, and the average of fold changes was calculated. Error bars represent SD. Statistical significance was determined by conducting an unpaired Student’s t test. *P < 0.05, **P < 0.01. (G) Representative images of OVCAR-5exo in imaging flow cytometry analysis. (H) Population of DNA-positive exosomes in OVCAR-5exo and OVCAR-8exo. Parental cells were treated with DMSO, olaparib, or topotecan for 48 hours. FTEexo, OVCAR-5exo, and OVCAR-8exo indicate exosomes derived from FTE, OVCAR-5, and OVCAR-8 cells, respectively. The experiment was performed in three independent biological replicates, and the average of fold changes was calculated. Error bars represent SD. Statistical significance was determined by conducting an unpaired Student’s t test. *P < 0.05.

  • Fig. 3 In vivo promotion of nExos with genotoxic drugs in ovarian cancer.

    (A) Schematic protocol for topotecan treatment. (B) Topotecan was administered intraperitoneally. All mice were euthanized on day 30. n of vehicle control–treated mice = 4 and n of topotecan-treated mice = 4. (C) Representative image of tumor tissue sections stained with hematoxylin and eosin (H&E). Scale bars, 700 μm (left panel) and 100 μm (middle panel). Black arrowhead indicates MN. (D) Left upper panel: Representative image of tumor tissue stained with DAPI and phalloidin. Scale bar, 50 μm. Right upper panel: Representative image of cell segmentation with Vectra imaging software. Lower panel: Representative image of the detected MN, indicated by white arrowhead. (E) Quantification of MN cells. MN counting is described in Materials and Methods. n of vehicle control–treated mice = 3 and n of topotecan-treated mice = 3. Error bars represent SD. Statistical significance was determined by conducting an unpaired Student’s t test. (F) Nanoparticle tracking analyses for exosomes derived from plasma and ascites. (G) Representative cryo-EM images of ascites exosomes from OVCAR-5 intraperitoneal model. Scale bar, 100 nm. (H) Western blot of mouse plasma exosomes (n = 3). (I) Population of DNA-positive exosomes in serum from non–tumor-bearing mouse and OVCAR-5 intraperitoneal model. n = each 6. Error bars represent SD. Statistical significance was determined by conducting unpaired Student’s t test. (J) Population of DNA-positive exosomes in serum and ascites from the mice treated with topotecan or vehicle. n = ascites, each 3 and serum, each 4. Error bars represent SD. Statistical significance was determined by conducting unpaired Student’s t test. (K) Representative images of serum exosome in imaging flow cytometry analysis.

  • Fig. 4 MN and nExo contain similar protein content.

    (A) Representative images of nuclear fraction and MN-enriched fraction. All samples were obtained from OVCAR-5 cells and stained with DAPI. The white box in the middle image is a magnified view. Scale bar, 50 μm. (B) Counts of protein cellular compartment of origin resulting from MS analysis in exosome and MN-enriched fraction. Samples were obtained from OVCAR-5 cells. The x axis represents the categories of cell components. Proteins from nuclear and chromosome compartments highlighted in yellow. (C) Venn diagram showing overlapping proteins between exosome and MN-enriched fraction. Samples were obtained from OVCAR-5 cells. Lamin A/C, Histone H2A/B, importin, and heat shock protein (HSP) 70/90 were included in the overlapped 127 proteins. (D and E) Serial confocal images of mCherry-LaminA/C–expressing OVCAR-5 cells. Nuclei were stained with DAPI. CD63 and CD9 were IF-stained with antibodies as described in Materials and Methods. Scale bars, 5 μm. (F) Representative images from time-lapse imaging in mCherry-LaminA/C–expressing OVCAR-5 cells. Cell membrane was stained with CellMask Deep Red. Cells stably expressed green fluorescent protein (GFP)–CD63 by lentiviral infections as described in Materials and Methods. Scale bar, 3 μm. (G) Representative images of time-lapse imaging in mCherry-Histone H2B–expressing OVCAR-5 cells. Cells were transiently transfected with mEmerald-CD9 plasmid as described in Materials and Methods. Scale bar, 4 μm.

  • Fig. 5 Cargo of disrupted MN is loaded into nExos.

    (A) Confocal image of OVCAR-5 cells. Nuclei were stained with DAPI, and Lamin A/C and CD63 were IF-stained as described in Materials and Methods. Scale bars, 5 μm. (B) Serial confocal stack images of OVCAR-5 cells. Nuclei were stained with DAPI, and Lamin A/C and CD9 were IF-stained with antibodies as described in Materials and Methods. Scale bars, 10 μm (upper panels) and 3 μm (lower panels). (C to E) Representative images of TEM of OVCAR-5 cells. Black arrowheads indicate the disrupted nuclear envelop of MN. Scale bars, 2 μm (C) and 500 nm (D and E). (F) Serial confocal images of OVCAR-5 cells. Nuclei were stained with DAPI. IF staining for Lamin A/C and EEA1 was performed with antibodies as described in Materials and Methods. Scale bars, 5 μm (upper panel) and 2.5 μm (lower panel). (G) Western blot of OVCAR-5 cells with CD63 knockdown (CD63 KD). CTRL indicates the OVCAR-5 cells transfected with a scramble shRNA sequence. Densitometry analysis is quantified in the bar chart. (H) Representative images of exosome from CD63-knockdown OVCAR-5 cells in imaging flow cytometry analysis. The fold change of nExo population. The experiment was performed in three independent biological replicates, and the average of fold changes was calculated. Error bars represent SD. Statistical significance was determined by conducting an unpaired Student’s t test. (I) The samples of OVCAR-5 cells with IP experiments for CD63 were analyzed with 2% agarose gel electrophoresis, and DNAs were visualized by ethidium bromide staining. (J) Western blot of OVCAR-5 cells with IP experiments for CD63. CTRL-IP indicates the OVCAR-5 cells treated with negative control immunoglobulin G (IgG) in the Universal Magnetic Co-IP Kit (54002, Active Motif).

  • Fig. 6 Detection of nExo in clinical samples.

    (A and B) NTA for exosomes was derived from plasma and ascites from patients with high-grade serous ovarian carcinoma, along with representative cryo-EM images. Scale bars, 100 nm. (C and D) Pie charts of the cellular compartment of origin of proteins based on MS analysis from plasma- and ascites-derived exosomes. Nuclear components are highlighted in red. (C) “Others” includes cytoskeleton, mitochondrion, ribosome, and vacuole. (D) “Others” includes mitochondrion, nucleus, organelle lumen, ribosome, and vacuole. (E) Representative tissue image of high-grade serous ovarian carcinoma stained with H&E. Black arrowhead indicates MN. Scale bars, 100 μm. (F) Population of micronucleated cells in high-grade serous ovarian carcinoma tissue. Pretreated tissue slides were obtained and analyzed as described in Materials and Methods. Each dot indicates one view for analysis. n = 4. (G) Representative imaging flow cytometry images of exosomes obtained from ascites of patients with high-grade serous ovarian carcinoma. (H) Venn diagram showing overlapping exonic mutated genes between tumor and ascites exosomes. Samples were obtained from patients with HGSC. DROSHA, LIG4, MACROD2, SATB1, RASSF6, and BIRC2 were included in the 43 overlapping genes. (I) Read counts of all chromosomes in plasma- and ascites-derived exosomal DNA and tumor DNA. All three cases were from patients with advanced-stage HGSC. (J) CNV status in gDNA in ascites-derived exosomal DNA and corresponding primary tumor from patients with HGSC. Profiles demonstrate somatic chromosomal gains (upper) and losses (lower), as well as normal polymorphisms.

Supplementary Materials

  • Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/5/11/eaax8849/DC1

    Fig. S1. Optimization of imaging flow cytometry analyses for exosomes.

    Fig. S2. Characterization of nExo.

    Fig. S3. Analyses of cell MN and exosomes treated with genotoxic drugs.

    Fig. S4. Imaging of MN in ovarian cancer cells.

    Fig. S5. Characterization of nExo and MN in human samples.

    Fig. S6. WGS for clinical samples.

    Fig. S7. Proposed model for the mechanism underlying nExo synthesis.

    Movie S1. Time-lapse imaging in mCherry-LaminA/C–expressing OVCAR-5 cells.

    Movie S2. Time-lapse imaging in mCherry-Histone H2B–expressing OVCAR-5 cells.

  • Supplementary Materials

    The PDFset includes:

    • Fig. S1. Optimization of imaging flow cytometry analyses for exosomes.
    • Fig. S2. Characterization of nExo.
    • Fig. S3. Analyses of cell MN and exosomes treated with genotoxic drugs.
    • Fig. S4. Imaging of MN in ovarian cancer cells.
    • Fig. S5. Characterization of nExo and MN in human samples.
    • Fig. S6. WGS for clinical samples.
    • Fig. S7. Proposed model for the mechanism underlying nExo synthesis.
    • Legends for movies S1 and S2

    Download PDF

    Other Supplementary Material for this manuscript includes the following:

    • Movie S1 (.mp4 format). Time-lapse imaging in mCherry-LaminA/C–expressing OVCAR-5 cells.
    • Movie S2 (.mp4 format). Time-lapse imaging in mCherry-Histone H2B–expressing OVCAR-5 cells.

    Files in this Data Supplement:

Stay Connected to Science Advances

Navigate This Article