• 2022-05
  • 2022-04
  • 2021-03
  • 2020-08
  • 2020-07
  • 2020-03
  • 2019-11
  • 2019-10
  • 2019-09
  • 2019-08
  • 2019-07
  • br in the contralateral channel


    in the contralateral channel bearing tumor-free cells. Immunofluo-rescence staining showed that the expression of TJ proteins in the vascular that interacted with tumor cells was vastly diminished, indicating that the TJs were degraded during the process of tumor cells crossing the BBB (Fig. 3B). We also performed epithelial phe-notypic staining (E-cadherin (E-Cad), N-cadherin (N-Cad), and Vimentin) on the tumor cells in each step (Fig. S2A). In addition, the fibroblasts Amiloride HCL in the in situ microenvironment were stained for a-smooth muscle Amiloride HCL (a-SMA) to determine if the cells turned into cancer-associated fibroblasts (CAFs) (Fig. S2B), while the astro-cytes in the metastatic site microenvironment were stained for glial fibrillary acidic protein (GFAP) to detect activation (Fig. S2C). Our results demonstrate that the behavior and phenotype of tumor cells and microenvironmental mesenchymal cells during metastasis on the chip can recapitulate the changes of these cells during the course of in vivo metastasis, suggesting that our chip platform is an effective tool to study lung cancer brain metastasis.
    3.4. Highly brain metastatic lung cancer cells are more prone to BBB extravasation compared to parental cells
    Next, we used multiple human cancer cell lines to validate our chip platform by comparing extravasation of lung cancer cells with differed BM abilities. Mice were inoculated with the human lung cancer cell line PC9 and the brain metastatic subpopulation was extracted to isolate PC9 clones with more metastatic activity. After 
    two rounds of in vivo selection and in vitro expansion (Fig. 4A(i)), the PC9-BrM3 cells were selected as the highly brain metastatic cell populations, and exhibited consistently enriched brain meta-static activity. The PC9-BrM3 cells formed BM in 94.1% of animals with a 3–4 week tumor formation cycle, while the parental PC9 line exhibited a BM efficiency of 21.4% with a 6–7 week tumor for-mation cycle (Fig. 4A(ii)). Further, mice inoculated with PC9-BrM3 cells had significantly shorter survival times compared to mice inoculated with the parental line (Fig. 4B).
    Since extravasation through the BBB is an important and unique indicator of BM, we compared the ability of PC9-BrM3 cells and parental cells to penetrate the endothelium using both the classic Transwell two-dimensional culture model and our dynamic bionic chip model. Consistently, both models showed that the highly brain metastatic PC9-BrM3 cells were more capable of penetrating the endothelial cells compared to the parental cells (Fig. 4C), and the wound healing assays showed that PC9-BrM3 cells were also more migratory compared to PC9 cells with a more obvious mes-enchymal phenotype (Fig. S3A, B). Many studies have shown that matrix metalloproteinases (MMPs), especially MMP-2 and MMP-9, play a critical role in promoting BBB leakage by degrading TJs and basement membrane proteins [40,41]. Therefore, we measured the expression of MMP-2 and MMP-9 in brain metastatic subpop-ulations and parental cells, respectively. We found that MMP-2 and MMP-9 protein expression was significantly up-regulated after increasing rounds of BM (Fig. 4D). Taken together, the highly brain
    Fig. 4. Highly-brain metastatic cells more aggressively extravasate the BBB compared to parental cells. (A)(i) In vivo selection scheme for the isolation of metastatic populations (BrM1, BrM2, and BrM3) from the PC9 lung adenocarcinoma line. (ii) Frequency and time circle of BM after inoculation of cells in nude mice. (B) Survive curve of nude mice after inoculation of indicated cells. (C) Representative images showing the GFP-expressing PC9/PC9-BrM3 cells and red-labeled hBMVEC cells in the trans-endothelia on Transwell (upper; scale bar, 20 lm) and after penetrating the BBB on the chip (lower; scale bar, 50 lm). (D) Representative western blot images showing MMP-9 and MMP-2 expression in brain metastatic cells and parental cells. The bar graphs are summarized results from 3 independent experiments. *p < 0.05; **p < 0.01; ***p < 0.001. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
    metastatic lung cancer cells recapitulated the salient features of BM and were more capable of extravasating the BBB compared to the parental cells.
    3.5. AKR1B10 is up-regulated in lung cancer brain metastasis
    AKR1B10 is a multi-functional protein expressed in many tumors. To investigate changes in AKR1B10 expression in brain metastatic NSCLC cell lines, we quantified mRNA and protein levels. We found that there was no statistical difference in AKRB10 mRNA levels (Fig. S4A), but the protein levels determined by west-ern blot assay (cellular AKR1B10) and ELISA (secreted AKR1B10 in supernatant) were significantly up-regulated in brain metastatic subpopulations compared to the parent cells (Fig. 5A, B). Since the chip we constructed can mimic the course of in vivo BM, we compared the expression level of AKR1B10 in lung cancers with and without BM on the chip using immunofluorescent staining. Compared with the upstream primary tumor cells, the down-stream brain metastasized lung cancer cells had up-regulated AKR1B10 expression (Fig. 5C), suggesting a good correlation