The role of superoxide anions in the development of distant tumour recurrence

Abstract

We hypothesise that reactive oxygen species (ROS) released from activated polymorphonuclear leucocytes during surgery play a crucial role in enhanced tumour recurrence seen after surgery. Therefore, the effect of ROS on adhesion of tumour cells to microvascular endothelium in a reproducible human in vitro model was studied. Preincubation of microvascular endothelial cells with the superoxide anion producing xanthine-xanthine oxidase complex significantly increased adhesion of the human colon carcinoma cells HT29 (167% vs control, P < 0.01), Caco2 (164% vs control, P < 0.01) and of the pancreas carcinoma cells PanC1 (180% vs control, P < 0.01). Addition of the antioxidant enzymes superoxide dismutase or catalase significantly decreased tumour cell adhesion (P < 0.01). Exposure of endothelial cells to superoxide anions increased the apoptotic rate to 7.9 times the normal rate. Additionally, exposure increased expression of the endothelial adhesion molecules E-Selectin, ICAM-1, and VCAM-1 of maximally 170% vs control (P < 0.01). In conclusion, this study shows that superoxide anions promote the adherence of tumour cells to the microvasculature by inducing endothelial apoptosis that subsequently induces the expression of various adhesion molecules for tumour cells. This indicates that by tackling the production of ROS preventing tumour recurrence at distant sites might be feasible.

Figure 1

Calibration curves showing the fluorescence measurements (OD 485–530 nm) of calcein-AM-labelled tumour cells in order to quantify the amount of tumour cells. Data represent mean ±s.e.m. of duplicate wells.

Figure 2

Validation of superoxide anion production by the ferricytochrome c reduction assay. Xanthine–xanthine oxidase was added to the wells with or without 400 U ml⁻¹ SOD immediately followed by the addition of 75 μM cytochrome c. Continuous measurements at 550–540 nm were carried out for 125 min. Data represent mean absorbance values (OD 540-550 nm) ±s.e.m. of quadriplate wells.

Figure 3

Tumour cell adhesion to MEC after preincubation of MEC with X–XO at varying time intervals. Means (n=6; % vs control) ±s.e.m. are shown. ^*P<0.01 vs control.

Figure 4

Adhesion of HT29 (A), Caco2 (B), and PanC1 (C) after 12 h preincubation of MEC with X, XO, and X–XO. The antioxidant enzymes SOD, catalase (cat) and the combination of both were added during the preincubation. Measurement of adhesion to an empty well (plastic) was determined as a negative control. Means (n=6; % vs control) ±s.e.m. are shown. ^*P<0.01 vs control; ^**P<0.01 vs X–XO; §P<0.05 vs X–XO.

Figure 5

Adhesion of untreated HT29 vs HT29 preincubated with X–XO for 12 h (HT29+X–XO) to untreated endothelium (control) and endothelium preincubated with X–XO (X–XO). Means (n=6; % vs control) ±s.e.m. are shown. ^*P<0.01 vs control.

Figure 6

(A) MEC proliferation assay. DNA in ng ml⁻¹ of MEC after 0, 1, 2, and 3 day(s) incubation with X, XO, and X–XO. Bars represent means (n=4). (B) Photographs of MEC monolayers: in medium (normal) and in medium preincubated with X–XO for 12 h. Comparable number of cells are seen.

Figure 7

Measurement of apoptosis by ELISA in MEC after preincubation with X, XO, and X–XO for 12 h. Data represent mean absorbance values (OD 550 nm) ±s.d. of triplate wells. ^*P<0.01 vs control.

Figure 8

Adhesion molecule expression on MEC. After 2–24 h preincubation with X–XO, EIA with anti-E-Selectin, anti-ICAM,- and anti-VCAM-antibodies was performed. Bars represent the mean absorbance values (OD 405 nm) ±s.d. of quadriplate wells. ^*P<0.01 vs control.

Figure 9

Adhesion molecule expression on HT29 colon carcinoma cells. After 12 h preincubation with X–XO, EIA was performed. Bars represent the mean absorbance values (OD 405 nm) ±s.d. of quadriplate wells.