A Comparison of Human Neutrophils Acquired from Four Experimental Models of Inflammation.

Abstract

Defects in neutrophil function have been implicated in a wide spectrum of clinical conditions. Several models are employed to study activated human neutrophils akin to those found at a site of inflammation. These include whole blood (WB) ex vivo stimulation with lipopolysaccharide (LPS) and in vivo techniques: cantharidin blister, skin windows and intra-dermal injection of UV-killed E.coli (UVKEc). Neutrophils obtained from these have never been compared. We compared the activation status of neutrophils from each technique in order to inform the optimal model for use in human studies. Healthy male volunteers were randomised to undergo one of the four techniques (n = 5/group). LPS: WB stimulated with 1ng/ml of LPS for 4 hours. Cantharidin: 12.5μl of 0.1% cantharidin elicited a single blister, aspirated at 24 hours. Skin windows: four 6mm mechanical-suction blisters created, de-roofed and an exudate-collection chamber placed over the windows for 4 hours before aspiration. UVKEc: 1.5 x 107 UVKEc injected intra-dermally. A single 10mm mechanical-suction blister formed and aspirated at 4 hours. Unstimulated WB used as the control. Flow cytometry was used to determine activation status using CD16, CD11b, CD54, CD62L and CD88. Functional status was assessed with a phagocytosis assay. The pattern of neutrophil activation was similar in all models. Neutrophil CD11b was elevated in all models, most markedly in UVKEc (p<0.0001), and CD54 was also elevated but only significant in the LPS model (p = 0.001). CD62L was significantly reduced in all 4 models (p<0.0001) and CD88 was also suppressed in all. There were no changes in CD16 in any model, neither was there any significant difference in the phagocytic capacity of the neutrophils. In summary, there are no significant differences in activation marker expression or phagocytic capacity in the neutrophils obtained from each technique. Therefore we believe whole blood stimulation is the best model in experimentally challenging inpatient populations.