Incidence of neutrophil extracellular traps (NETs) in different membrane oxygenators: pilot in vitro experiments in commercially available coated membranes

Abstract

Neutrophil extracellular traps (NETs) were detected in blood samples and in cellular deposits of oxygenator membranes during extracorporeal membrane oxygenation (ECMO) therapy and may be responsible for thrombogenesis. The aim was to evaluate the effect of the base material of gas fiber (GF, polymethylpentene) and heat exchange (HE) membranes and different antithrombogenic coatings on isolated granulocytes from healthy volunteers under static culture conditions. Contact of granulocytes with membranes from different ECMO oxygenators (with different surface coatings) and uncoated-GFs allowed detection of adherent cells and NETotic nuclear structures (normal, swollen, ruptured) using nuclear staining. Flow cytometry was used to identify cell activation (CD11b/CD62L, oxidative burst) of non-adherent cells. Uncoated-GFs were used as a reference. Within 3 h, granulocytes adhered to the same extent on all surfaces. In contrast, the ratio of normal to NETotic cells was significantly higher for uncoated-GFs (56-83%) compared to all coated GFs (34-72%) (p < 0.001) with no difference between the coatings. After material contact, non-adherent cells remained vital with unchanged oxidative burst function and the proportion of activated cells remained low. The expression of activation markers was independent of the origin of the GF material. In conclusion, the polymethylpentene surfaces of the GFs already induce NET formation. Antithrombogenic coatings can already reduce the proportion of NETotic nuclei. However, it cannot be ruled out that NET formation can induce thrombotic events. Therefore, new surfaces or coatings are required for future ECMO systems and long-term implantable artificial lungs.

Keywords: ECMO; Membrane oxygenators; NETosis; NETs; Neutrophil granulocytes.

Fig. 1

Contact of isolated granulocytes with surfaces of different GF-types (test material) resulted in granulocyte adhesion and NET forming. A Representative sample of adherent and DAPI-stained granulocytes on a PLS-GF with normal (1) and NETotic nuclei (swollen nuclei (2), ejection of DNA (3)), (80x magnification). B Quantification of cell density of adhering granulocytes on GFs from different oxygenator types (300 randomly selected microscopic images from 10 gas fibres of each sample). Cells were evenly distributed and cell density was comparable (p = 0.230). C Quantification of the ratio of NETotic to all analysed nuclei (215 non-overlapping microscopic images). The percentage of NETotically altered cells was significantly higher in the uncoated-GF compared to all other GFs. (**p < 0.01). Uncoated and coated GFs exhibit a significantly higher number of NETotically altered granulocytes compared to the negative control on glass surfaces (p < 0.01). Data are median (IQR). Statistics, one-way ANOVA

Fig. 2

NET formation on PLS-GFs. Adherent granulocytes on a PLS-GF (80x magnification) after 3 h incubation with A buffer (negative control) and B PMA (positive control). PMA induced NET formation presenting different nuclear morphologies (predominantly completely ejected DNA). C PMA-treated granulocytes on a PLS-GF formed NETs that were removed after DNAse I treatment (D)

Fig. 3

Contact of isolated granulocytes with test materials did not affect viability (A), did not induce ROS production (B), only marginally activate the cells (C, D) without alteration of expression of CD11b and CD62L (E–H). Granulocytes in the supernatant of the adhesion experiments (n = 6) were vital (> 98%) (A), and showed no ROS production (B). The proportion of CD62L-/CD11b + cells (= activated cells) (C) and CD62L + /CD11b− cells (= non-activated cells) (D) remained below 10%. The median fluorescence intensity (MFI) of CD11b (E, F) and CD62L (G, H) remained at the level of reference (not-stimulated cells without contact to test materials). TNF/fMLP and PMA, stimulation of reference cells with TNF/fMLP or PMA as described in the method section. Statistics, two-way ANOVA on variance. (Gas fibre membranes (GF), heat exchanger membranes (HE))