Research ArticleBIOPHYSICS

Exosome-templated nanoplasmonics for multiparametric molecular profiling

See allHide authors and affiliations

Science Advances  06 May 2020:
Vol. 6, no. 19, eaba2556
DOI: 10.1126/sciadv.aba2556
  • Fig. 1 Templated nanoplasmonics for multiparametric profiling of exosomes.

    (A) Schematic of the TPEX platform. The technology is designed to measure exosomal markers and comprises three functional steps. Exosomes are first labeled with fluorescent molecular probes and AuNP. While AuNP remain well dispersed when associated with nonvesicle, free proteins, they assemble onto exosome periphery, through electrostatic interactions. Excess unbound probes and AuNP are not removed. In the presence of gold salt, the AuNP serve as seeds for in situ gold growth. The dispersed AuNP experience a small growth and a slight shift in their absorbance spectra, leading to minimal changes in the fluorescence signals of probes. The exosome-bound AuNP, on the other hand, develop into a nanoshell; this nanostructure is templated by the vesicle dimension and demonstrates a large red shift in its plasmonic resonance to effectively quench the fluorescence signal of probes bound onto the same vesicle. The TPEX fluorescence signal is thus multiparametric, for both exosomal biophysical characteristics and biomarker compositions. (B) Transmission electron micrographs of TPEX products. In the presence of free proteins, AuNP remained well dispersed (before) and demonstrated a small particle growth after treatment with gold salt (after). When incubated with exosomes, AuNP bound to vesicle periphery (before) and developed into large spherical particles after gold growth (after). Scale bars, 20 nm. (C and D) Photographs of the microfluidic device and the smartphone-based optical detector. Absorbance and fluorescence measurements could be performed on the integrated platform through different light-emitting diode (LED) sources and filter configurations. Scale bar, 1 cm.

  • Fig. 2 TPEX absorbance analysis.

    (A) Optical simulations with different-sized templates. On the basis of microscopy characterization of the formed TPEX nanostructures, we simulated the plasmonic resonance peaks of gold nanoshells developed on different-sized templates (left). For exosome-sized templates (30 to 150 nm; shaded red), the resultant plasmonic peaks locate predominantly at >600 nm. Red dotted line indicates the mean peak wavelength, formed from this range of template diameters, and locates to 750 nm. Electric field distributions at 750 nm were mapped for single AuNP (bare or particles associated with free proteins) and gold nanoshell (exosome-templated), formed after gold growth (right). Ø indicates particle diameter after gold growth. The simulations confirmed that nanoshells templated to exosome dimension could generate a strong plasmonic resonance at 750 nm |E|, electric field norm. (B) Tuning of the TPEX responsive range to template diameter. We incubated different-sized templates with AuNP of different diameters to form gold nanoshells. The TPEX absorbance measurement (A) is defined as the ratio of absorbance at 750 and 540 nm and its difference (ΔA) before and after gold growth. Using the 9-nm AuNP, the TPEX response range could be optimized to match exosome dimension, so as to maximize exosome-induced signals. (C) Experimental evaluation with biological samples. Exosomes derived from human colorectal adenocarcinoma (DLD-1) were spiked into vesicle-depleted fetal bovine serum (dFBS) and subjected to TPEX analysis with 9-nm AuNP. In all reactions, we measured the resultant absorbance (left) and diameter changes (right). Diameter changes were performed through dynamic light scattering analysis. Only samples containing exosomes demonstrated a large signal increment, while reactions in phosphate-buffered saline (PBS) (i.e., bare AuNP) and that in dFBS (i.e., free proteins) showed negligible changes. (D) Correlation of TPEX absorbance analysis with exosome concentration. Exosomes derived from four cell lines (DLD-1, HTC116, MKN45, and SNU484) were counted through nanoparticle tracking analysis (NTA) and evaluated by the TPEX absorbance analysis. All measurements were performed in triplicate, and the data are displayed as means ± SD in (B) to (D). *P < 0.05 and ***P < 0.0005, Student’s t test. NS, not significant; a.u., arbitrary units.

  • Fig. 3 Multiplexed fluorescence analysis of exosome molecular markers.

    (A) TPEX fluorescence analysis. To evaluate whether TPEX nanoshell can be used to quench colocalized fluorescent probes, we prepared PDA nanoparticles as well-defined size templates and conjugated the particles with fluorescent dyes [Alexa Fluor 647 (A647)]. We treated the templates with TPEX reaction and measured the resultant changes in fluorescence (ΔF; top) and absorbance (ΔA; bottom). Both analyses showed a similar trend and demonstrated a template size-responsive range optimized for exosome diameters. (B) Assay specificity to exosome markers. We incubated whole exosomes (derived from DLD-1) that contain CD63 (top) and free CD63 (bottom) with fluorescent aptamers (anti-CD63 and scrambled control) for TPEX measurements. Only whole exosomes showed significant signals, while free CD63 samples demonstrated negligible signals. Of the different fluorescent dyes tested [fluorescein isothiocyanate (FITC), rhodamine B (RB), and A647], aptamers modified with A647 (emission of 665 nm, most closely matched to a TPEX absorbance of 750 nm) demonstrated the largest signal difference. (C) Multiplexed profiling of exosome markers. Exosomes were incubated with different fluorescent aptamers, either individually (singleplex) or as a mixture (multiplex), for TPEX analysis. The multiplex fluorescence spectrum agreed with the singleplex spectra (top) and showed accurate marker expression profiles across cell lines (bottom). (D) Molecular detection sensitivity. The limit of detection was determined by titrating a known quantity of exosomes and measuring their associating TPEX signal for CD63. The detection limit of enzyme-linked immunosorbent assay (ELISA) was independently assessed on the basis of chemiluminescence. All measurements were performed in triplicate, and fluorescence analysis was normalized against respective sample-matched scrambled controls. The data are displayed as means ± SD in (A), (B), and (D). *P < 0.05, **P < 0.005, and ***P < 0.0005, Student’s t test.

  • Fig. 4 Exosome analysis in complex background.

    (A) TPEX analysis of mock clinical samples. Samples were prepared by spiking exosomes, derived from six human lines into vesicle-depleted human serum. In these spiked samples, we measured exosome marker CD63 and putative cancer markers including CD24, EpCAM, and MUC1. All protein measurements of the spiked samples were performed by multiplex TPEX analysis on a microfluidic platform, as well as conventional singleplex sandwich ELISA. The analyses were compared against marker signatures of pure exosomes (obtained from exosomes before spiking). For each marker analyzed, the TPEX analysis showed a better concordance to reflect the expression trends across cell lines. (B) Correlation of TPEX measurements with pure exosome signatures. The TPEX detection showed a good correlation to the pure exosome analysis (left), while the conventional ELISA measurements performed on the same spiked samples showed a significantly poorer correlation (right). All measurements were performed in triplicate, against respective sample-matched scrambled controls. The data are assay-normalized and displayed as means in (A) and as means ± SD in (B).

  • Fig. 5 TPEX analysis of patient prognosis.

    (A) Analysis of protein markers in clinical cancer ascites (n = 20; 12 colorectal cancer and 8 gastric cancer) using multiplex TPEX for measurement of vesicle-associated target markers (top) and conventional singleplex ELISA for measurement of total target markers (bottom). TPEX analysis showed different protein expression profiles as compared to the ELISA analysis. (B and C) Receiver operating characteristic (ROC) curves of the TPEX (B) and ELISA (C) regression models on ascites samples of colorectal cancer (left), gastric cancer (middle), and both cancer types (right). ROC curves were constructed using individual markers or a combination of the target markers (mix). The TPEX analysis showed a higher accuracy in prognosis classification across both cancers as compared to the ELISA assay. All measurements were performed in triplicate, against respective sample-matched scrambled controls. The data are assay-normalized and displayed as means in (A).

Supplementary Materials

  • Supplementary Materials

    Exosome-templated nanoplasmonics for multiparametric molecular profiling

    Xingjie Wu, Haitao Zhao, Auginia Natalia, Carine Z. J. Lim, Nicholas R. Y. Ho, Chin-Ann J. Ong, Melissa C. C. Teo, Jimmy B. Y. So, Huilin Shao

    Download Supplement

    This PDF file includes:

    • Figs. S1 to S10

    Files in this Data Supplement:

Stay Connected to Science Advances

Navigate This Article