Fucosyltransferase 4 and 7 mediates adhesion of non-small cell lung cancer cells to brain-derived endothelial cells and results in modification of the blood-brain-barrier: in vitro investigation of CD15 and CD15s in lung-to-brain metastasis

Abstract

Purpose: Metastatic non-small cell lung (NSCLC) cancer represents one of the most common types of brain metastasis. The mechanisms involved in how circulating cancer cells transmigrate into brain parenchyma are not fully understood. The aim of this work was to investigate the role of fucosylated carbohydrate epitopes CD15 and sialyated CD15s in cancer adhesion to brain-derived endothelial cells and determine their influence in blood-brain barrier (BBB) disruption METHODS: Three distinct, independent methods were used to measure brain endothelial integrity and include voltohmmeter (EVOM™), impedance spectroscopy (CellZscope®) and electric cell-substrate impedance sensing system (ECIS™). Two fucosyltransferases (FUT4 and 7) responsible for CD15 and CD15s synthesis were modulated in four human cancer cell lines (three lung cancer and one glioma).

Results: Overexpression of CD15 or CD15s epitopes led to increase in adhesion of cancer cells to cerebral endothelial cells compared with wild-type and cells with silenced CD15 or CD15s (p < 0.01). This overexpression led to the disruption of cerebral endothelial cell monolayers (p < 0.01). Knockdown of FUT4 and FUT7 in metastatic cancer cells prevented disruption of an in vitro BBB model. Surprisingly, although the cells characterised as 'non-metastatic', they became 'metastatic' -like when cells were forced to over-express either FUT4 or FUT7.

Conclusions: Results from these studies suggest that overexpression of CD15 and CD15s could potentiate the transmigration of circulating NSCLC cells into the brain. The clinical significance of these studies includes the possible use of these epitopes as biomarkers for metastasis.

Keywords: Brain metastasis; CD15; CD15s; FUT4; FUT7; NSCLC.

Fig. 1

CD15 and CD15s expression in FUT4/FUT7-(cDNA and shRNA) transfected cells. Target cells (SEBTA-001, NCI-H1299, COR-L105 and UP-007) were transfected, either with plasmids containing FUT4 or FUT7-cDNA to induce CD15 or CD15s overexpression, or with FUT4 or FUT7-shRNA constructs for CD15 or CD15s knockdown. a, cFUT4/CD15. b, dFUT7/CD15s. a, b Representative confocal images showing localisation of a CD15 or b CD15s (red) cells grown on coverslips. c, d Representative FC analysis of CD15 or CD15s expression. The histograms show percentages of CD15/CD15s positive cells in wild type controls (blue) where no treatment applied, vehicle controls were transfected with non-coding plasmid (black), cells transfected with shRNA for FUT4 or FUT7 (red) and in cells transfection with cDNA for FUT4 or FUT7 (green). (**P < 0.01) difference compared to the wild type. N = 3. No treatment was applied to wild type controls and vehicle controls were transfected with non-coding plasmids. Scale bar = 20 μm

Fig. 2

Adhesion of cancer cells following genetic manipulation of CD15 and CD15s. Adhesion of cancer cells following FUT4/CD15 and FUT7/CD15s manipulation on metastatic NSCLC SEBTA-001 (a, e), NCI-H1299 (b, f) and non-metastatic NSCLC: COR-L105 (c, g), UP-007 (d, h). Cancer cells were incubated for 90 min on a monolayer of activated brain endothelial cells (hCMEC/D3). Non-adherent cancer cells were washed off followed by lysis of cells and quantified via a microplate reader at 480–520 nm. (**P < 0.01), N = 3

Fig. 3

Effects of FUT4/CD15 and FUT7/CD15s modulation on intact brain endothelial monolayer assessed by EVOM™. Resistance of hCMEC/D3 monolayer in response to cancer cells following FUT4/CD15 (a–d) and FUT7/CD15s (e, f) genetic manipulation on SEBTA-001 (a, e), NCI-H1299 (b, f), COR-L105 (c, g) and UP-007 (d, h). TEER values decreased when cell lines (except for UP-007) were transfected with FUT4/CD15 were placed on an established endothelial cell layer. Similar data were observed for cells transfected with FUT7/CD15s.The black bars represents the resistance of endothelial monolayer before adding cancer cells, light-grey bars to TEER values 2 h post-addition of cancer cells, dark-grey bars to TEER values 24-h post-addition of cancer cells and white bars to TEER values 48-h post addition of applying cancer cells. a–c; *p < 0.01, **p < 0.05. N = 3

Fig. 4

Effects of FUT4/CD15 and FUT7/CD15s modulation in cells on brain endothelial monolayer measured in real-time. Representative histograms depicting the changes in TEER values, pre and post addition of cancer cells using the automated system (CellZscope®). The green line represents endothelial cells (control) with no addition of cancer cells, red line represents FUT4/CD15 overexpressor cells (a–d) or FUT7/CD15s (e–h), blue line represents FUT4/CD15 shRNA transfected cells (a–d) or FUT7/CD15s shRNA cells (e–h), yellow line representing the wild type and black line representing the blank. Bar chart represents quantified TER values (*P < 0.05), N = 3. WT wild type, KO knockdown, OE overexpression

Fig. 5

FUT4 and FUT7 knockdown in metastatic cancer cells prevents disruption of intact brain endothelial monolayer. FUT4 in SEBTA-001 (a) and NCI-H1299 (b). FUT7 in SEBTA-001 (c) and NCI-H1299 (d). a–d Representative histograms showing resistance of endothelial barrier in real time, pre and post addition of cancer cells measured by (ECIS™). Green line represents control (no cancer cells), Red line represents FUT4/CD15 (a, b) or FUT7/CD15s (c, d) overexpressing cancer cells, blue line represents knockdown cell lines FUT4/CD15 (a, b) or FUT7/CD15s (c, d), yellow line represents wildtype cancer cells and black line represents blank no cells. (a–d) Bar charts represent quantified resistance values at 24-h pre-post-addition of cancer cells (*P < 0.05). Lower panel a–d) Representative microscopic images demonstrating the changes of integrity in brain endothelial monolayer, at time point of 24 h post-addition of cancer cells. Red arrows point to areas of endothelial cell monolayer disruption. Images were obtained using a phase contrast microscope (× 4). All studies N = 3