br Fig Construction and verification of the functional BBB A
Fig. 2. Construction and verification of the functional BBB. (A) Diagram showing the sequenced processes of cell seeding to generate the BBB. (i) hBMVECs were introduced into the vascular channels; scale bar, 100 lm. (ii) Human astrocytes (HA) were introduced into the Nocodazole parenchyma chamber; scale bar, 100 lm. (iii) The intact BBB was composed of hBMVECs co-cultured with HA under the flow shear force (0.1 ll/min); scale bar, 200 lm. (B) Representative staining of endothelial cells on the side wall of the
BBB. hBMVECs were labeled with CD31 (green) to confirm the integrity of the endothelium on the side wall; scale bar, 50 lm. (C) Expression of TJ proteins ZO-1 (green) and VE-cadherin (red) in hBMVECs alone and in the intact BBB (hBMVEC + HA + shear force); Images were captured with a confocal microscope. Scale bar, 20 lm. (D) TEER measurement in the BBB group and hBMVECs alone group every 12 h for 4 days. The maximal value of TEER was 1120 ± 92 X cm2 in the BBB group and 483 ± 40 X cm2 in the hBMVECs alone group. Data are presented as mean ± SD, and n = 3 for each group. (E) Representative images show the ROI set (white circles) and permeable FITC-
conjugated dextran (green) at 40 kDa diffusing across the BBB into the brain compartment over 12 h. The quantitative graphs show the permeability (%) of dextran across the BBB into the brain compartment under the condition of hBMVECs alone or with the BBB group. **p < 0.01; scale bar, 100 lm. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
determined by counting the stained cells in five randomly selected fields with a light microscope. And the transendothelial ability was standardized by making a ratio of experimental groups cell num-ber to the control group cell number (100%).
The NSCLC cell line PC9 was engineered to stably express GFP-luciferase fusion protein through transfection of a triple modality
plasmid vector (Suzhou GenePharma Co., Ltd., China). Athymic female BALB-c-nu mice aged between 4 and 6 weeks were pur-chased from Beijing Vital River Laboratory Animal Technology Co., Ltd., China. After anesthetizing with ketamine (100 mg/kg body weight; Sigma, USA) and xylazine (10 mg/kg body weight; Sigma, USA), approximately 106 transfected cells in 100 ml PBS were injected into the left ventricle of each mouse. Brain coloniza-tion was analyzed in vivo by bioluminescence imaging (BLI). After retro-orbital injection of D-Luciferin (150 mg/kg body weight; Pro-
mega, USA), images were acquired with an IVIS Spectrum Xenogen machine (PerkinElmer, USA). The Living Image software (version 2.50) was used to analyze the bioluminescence images. All exper-iments involving the use of animals in this study were approved by Dalian Medical University Licensing Committee.
2.5. Surgical and serum samples
Surgical specimens were obtained from 12 patients diagnosed with NSCLC, from whom 6 primary lung tumors and 6 brain metas-tases were collected after surgical removal. In addition, 6 primary brain tumors (2 for glioma, 2 for meningioma, and 2 for other types) were used as non-metastatic controls. Serum samples were har-vested following a standard operation procedure. In brief, periph-eral blood (2 mL each subject) was collected in a serum separator tube (SST) and allowed to clot for 30 min before centrifugation for 15 min at approximately 1000 g/min. Immediately after centrifuga-tion, serum was transferred to clean polypropylene tubes and stored at 80 LC. Written informed consent was obtained from all participants. This study was approved by the Ethics Review Com-mittee of the Second Hospital of Dalian Medical University.
2.6. Statistical analysis
All quantitative data are presented as the mean ± standard devi-ation (SD), and were collected from at least 3 parallel in vitro sam-ples or 6 in vivo samples. All statistical analyses were conducted using the GraphPad Prism software (version 5.0; GraphPad Soft-ware Inc., USA). To compare differences between two groups, t tests were used, while for more than 2 groups, one-way ANOVA (analysis of variance) followed by Tukey’s multiple comparisons was used. The survival curve was calculated using the Kaplan-Meier method. A p value of 0.05 was defined as statistically significant.
3.1. Construction of a multi-organ microfluidic chip for lung cancer-derived BM
To recapitulate the key events of lung cancer-derived BM in vitro, we constructed a multi-organ bionic microfluidic chip con-sisting of two organ chip units: an upstream ‘‘lung” and a down-stream ‘‘brain” unit characterized by a functional ‘‘BBB” structure (Fig. 1A). The upstream ‘‘lung” adopted the sandwich structure for co-culturing multiple cell types, and two vacuum channels were precisely controlled by micro-pumps on both sides to simu-late the human respiratory rhythm. The lung cancer cells grew in the bionic microenvironment and intravasated the pulmonary ves-sels, as previously described . The downstream metastasis tar-get organ, ‘‘brain”, was comprised of a brain parenchyma chamber surrounded by two vascular channels, one of which was connected to the upstream unit as the experimental group while the other one served as the control group. The vascular channels and the parenchymal chamber communicated through micro-gaps to achieve cell interaction and allow penetration of the tumor cells into the BBB. In general, we built a dynamic microfluidic platform that can reproduce a series of continuous pathological processes of BM, from the development and intravasation of primary cancer in situ to extravasation and colonization of cancer at the metastasis site.