Imaging and tracking of tumor extracellular vesicles to unravel the progression of ovarian carcinoma using fluorescent membrane probes

Ting Wang【王婷】 a b 1, Qingyuan Liu【刘清源】 a b 1, Xingya Chen【陈星雅】 a b, Yueyue Zhao【赵月月】 a b, Yan Chen a b, Rui Wang a b, Fabiao Yu【于法标】 a b, Yanlong Xing【邢艳珑】 a b

a

Key Laboratory of Hainan Trauma and Disaster Rescue, The First Affiliated Hospital of Hainan Medical University, Hainan Medical University, Haikou 571199, China

b

Engineering Research Center for Hainan Bio-Smart Materials and Bio-Medical Devices, Key Laboratory of Emergency and Trauma, Ministry of Education, Key Laboratory of Hainan Functional Materials and Molecular Imaging, College of Emergency and Trauma, Hainan Medical University, Haikou 571199, China

Sensors and Actuators B: Chemical

Highlights

Introduction

Extracellular vesicles (EVs) are lipid bilayer vesicles that are secreted into the extracellular environment by all known organisms [1], [2]. EVs have been extensively investigated owing to the vital role in transporting bioactive cargos and mediating the intercellular communication to regulate various biological processes [3], [4], [5], [6]. Therefore, knowledge of EV biology and application has gained wide interest and grown rapidly [7], [8], [9], [10]. Typically, EVs have been employed as potential targets to study the mechanism of cargo transport of cancer cells and their influence on tumor development [11], [12]. However, investigation on EVs' function faces challenges, for instance, lack of powerful live imaging tools, efficient labelling strategies or suitable animal models [13], [14], [15]. In addition, EVs preserve heterogeneity due to complex biogenesis, which put hurdles in EV biology and pathology research including cancer [16], [17], [18].

Recent years, it has been reported that a multitude of noninvasive imaging methods has been investigated to explore the spatio-temporal dynamics of nanoparticles in vitro and in vivo [19], [20], [21], by combining with advanced labeling strategies [22], [23]. For instance, fluorescently labelled EVs have been visualized at cellular level or in vivo [24], [25], using lipid membrane dyes such as PKH67, DiR [26], Membright dyes [14], or using fluorescent EVs secreted from genetically encoded cells [15], [27], [28]. However, the visualized tracking EVs biodistribution and the assessment on EVs’ function in tumor development has only been sparsely investigated due to the above-mentioned challenges, which limits the insights into EVs pathophysiology, especially in single vesicle level [29].

Among the various cancer types, ovarian carcinoma is the gynecologic malignancy with the highest case-to-fatality ratio, putting threaten on female health worldwide [30]. It has been reported that EVs exert crucial function in shuttle molecules to receipt cells and influence cancer development [11], [31]. In particular, the homing targeting ability of EVs contributes to cancer growth and metastases [32]. However, there is still a lack of direct evidence on the function of ovarian cancer EVs in affecting tumor progression.

Scheme 1. Schematic illustration of imaging and tracking of tumor extracellular vesicles to unravel the progression of ovarian carcinoma using fluorescent membrane probes.

Fig. 1. Characterization of the OC-EVs before and after fluorescent dye labelling. (a) Molecular structure of the membrane binding probe MemBright. (b) TEM of OC-EVs, OC-EVs-Mem560 and OC-EVs-PKH67. Enlarged images of single EVs displayed in the bottom showed the respective boxed area of images in the top. Scale bars, 100 nm. (c) Western blot analysis of EVs with anti-CD63, anti-TSG101, anti-EpCAM, using GAPDH as loading control. (d) NTA results of i) OC-EVs ii) OC-EVs-Mem560 (average diameter of 155.7 nm with peak diameter of 129.5 nm) and iii) OC-EVs-PKH67 (average diameter of 153.7 nm and peak diameter of 144.6 nm) showing the concentration (y axis) and diameter (x axis).

Fig. 2. (a) Schematic diagram of cell membrane labelling with GFP (HO23 cells, top) or mCherry (EpCAM overexpression, OVCAR3 cells, bottom) and the secretion of transfected EVs. (b) Live cell confocal fluorescence imaging of i) HO23-GFP cells and ii) OVCAR3-mCherry cells, the enlarged images of boxes area and the 3D imaging of cells. Scale bars, 5 μm. (c) NTA results of EVs derived from i) HO23 cells ii) HO23-GFP cells and iii) OVCAR3-mCherry cells. (d) TEM of HO23-EVs, HO23-EVs-GFP and OC-EVs-mCherry. Scale bars, 100 nm. (e) Western blot analysis of three types of EVs using CD63, TSG101 and EpCAM antibodies, using GAPDH as loading control.

Fig. 3. (a) Fluorescence imaging of macrophages (RAW264.7) cocultured with OC-EVs-PKH67. Figures show the blue, green, merge channels and figure inset of individual cells. (b) Fluorescence imaging of OC-EVs-PKH67 internalized in macrophages at different time and the quantitative result. (c) Dynamic tracking of the interaction between macrophage-Mem560 and OC-EVs-PKH67. (d) The XZ and YZ display of 3D imaging of the cell in Fig. c. Time-dependent fluorescence imaging of the interaction between (e) Macrophage-GFP and OC-EVs-Mem560 (f)-(g) showed interaction of OC-EVs-PKH67 with neutrophil-Mem560, HO23-Mem560 and OVCAR3-Mem560. (i) Comparison of different internalization time of OC-EVs by various cells. Scale bars, 5 μm.

The following is the Supplementary material related to this article Movie S1 -S5.

Fig. 4. (a) Schematic illustration of co-incubation of OVCAR3-mCherry cells and HO23-GFP cells. Time-lapsed confocal imaging of (b) OVCAR3-mCherry cells and (c) HO23-GFP cells. Scale bars, 5 μm. (d) Time-lapsed imaging of the internalization of OC-EVs-mCherry by a HO23-GFP cell within 84 s with time interval of 12 s. Scale bars, 5 μm. (e) Live-cell imaging of EVs exchange and uptake between OVCAR3-mCherry and HO23-GFP cells. Magnification of boxes in merged panels are shown in enlarged images. i) EVs released from two cell lines were observed in surrounding regions (while and yellow arrowhead). ii) EVs secreted from HO23-GFP cell were observed inside the OVCAR3-mCherry cell (yellow arrowhead). iii) EVs secreted from OVCAR3-mCherry cell were found inside the HO23-GFP cell (white arrowhead). Scale bars, 20 μm. (f) and (g) The 3D imaging of OVCAR3-mCherry cell and internalized HO23-EVs-GFP, and HO23-GFP cell with captured OC-EVs-mCherry.

Fig. 5. In vivo and ex vivo imaging of OC-EVs. (a) Bioimaging plan after using i) OC-EVs-Mem560 and ii) OC-EVs-mCherry via I. P. or I. V. injection. (b) Ex vivo fluorescence (FL) imaging of organs (brain, lung, heart, liver, kidney, spleen and ovary) harvested from the mice treated with i) Mem560 dye and ii) OC-EVs-Mem560, I. P. injection. iii) Mem560 dye and iv) OC-EVs-Mem560, I. V. injection. (c) and (d) are quantification results of images displayed in Fig. b ii) and iv), analyzed by recording the photons/second/steradian (ph/s/sr) of each organ and normalized to that of the injected dose based on the fluorescence intensity. Data shown as mean (n = 3). (e) Ex vivo imaging of main organs harvested from the mice treated with OC-EVs-mCherry via i) I. P. injection or ii) I. V. injection. (f) and (g) quantification of imaging results displayed in Fig. e. Data shown as mean (n = 3).

Fig. 6. Investigation on the function of OC-EVs on tumor progression in mice models. (a) Schematic illustration of the experimental plan. (b) Photographs of the representative mice of the control (day 0), model (day 0), PBS (day 14) and OC-EVs-mCherry (day 14) groups. (c) Ultrasonic imaging of the mice displayed in Fig. b, images showed left (L) and right (R) ovaries from coronal, transverse views, and the blood signal. Scale bars, 1 mm. (d) Photographs of the reproductive system of the mice in Fig. b. The size of left ovary was measured by a vernier caliper. (e) H&E-stained slices of the brain, liver and ovary of the mice in Fig. b. Scale bars, 50 μm. (f) Ovarian volumes (mm3) of the left (L) and right (R) ovaries of the mice treated with PBS and OC-EVs-mCherry, measured by ultrasonic imaging (day1-day 11) and vernier caliper (day 14). (g) Ovarian tumor growth speed calculated from Fig. f and displayed in mm3/d. (h) Photographs (left) and fluorescence imaging (right) of organs harvested from the mice treated with PBS and OC-EVs-mCherry.

Conclusions

In this study, we employed two types of membrane probes to obtain fluorescently labelled EVs and successfully visualized OC-EVs in vitro and in vivo. Based on the direct labeling of EVs by lipophilic membrane probes (PKH67 and Mem560), we recorded the single vesicle internalization by various cells (including macrophage, neutrophil, ovarian cell and ovarian cancer cell). Using indirect labeling of vesicles by harvesting EVs from lentivirus (GFP and EpCAM:mCherry) transfected cells, we observed the one-way delivery and bidirectional exchange of EVs between normal ovarian cells and cancer cells. These data indicated the liability of cancer EVs to interact with cells of the same origin, mainly attributing to the homing characteristics of EVs. By administrating fluorescently labelled OC-EVs into mice models, the hepatic and renal clearance process, as well as the homing effect of EVs were unveiled. Furthermore, by injecting OC-EVs into xenotransplanted ovarian tumor bearing nude mice, the function of OC-EVs in promoting tumor growth was verified. In general, our findings supported the EV-mediated intercellular communication, which were complex, multi-directional, far-reaching, and homing. We also confirmed the metabolism routes of EVs in vivo and the essential function of EVs in accelerating tumor development. Future studies would focus on the underling molecular mechanism dominated in EVs influence on tumor progression, which would provide potential targets for tumor treatment.