Preincubation of Haemophilus influenzae with subinhibitory concentrations of macrolides: influence on human neutrophil chemiluminescence

Abstract

Preincubation of Haemophilus influenzae with antibiotics may influence opsonophagocytosis as studied by chemiluminescence. Two strains of H. influenzae (strain 1 [type b] and strain 2 [uncapsulated]) were pretreated with erythromycin, roxithromycin, clarithromycin, and azithromycin for 1 h in Haemophilus test medium (the last 25 min was either without serum or with 10% fresh serum or 10% decomplemented serum). Human neutrophils were stimulated with a pretreated or control inoculum at four different bacterium/neutrophil ratios and tested for luminol chemiluminescence with an LKB luminometer. The results were normalized for bacterium/neutrophil ratio and compared by the two-sided Wilcoxon test. Pretreatment of bacteria with one-half of the MICs of erythromycin, clarithromycin, and roxithromycin produced nonsignificant (P > 0.05) increases in the chemiluminescence response (means of 23% for strain 1 and 4% for strain 2). Pretreatment with azithromycin at one-half of the MIC produced an increase in the chemiluminescence response induced by serum-opsonized strain 1 (320% +/- 36% [mean +/- standard error of the mean]) and strain 2 (107% +/- 20%) (P < 0.05). This increase was concentration dependent: for strain 1, 60% +/- 18% at one-fourth of the MIC to 440% +/- 41% at the MIC; for strain 2, 10% +/- 5% at one-fourth of the MIC to 300% +/- 20% at the MIC. For strain 1, the maximal increase with azithromycin pretreatment (at the MIC) required opsonization with fresh serum. Opsonization with decomplemented serum was associated with a 53% +/- 21% increase; this increase was 28% +/- 3% in the absence of serum. For strain 2, azithromycin reduced the lag phase of the chemiluminescence response induced by the absence of serum but did not alter the chemiluminescence response in the presence of decomplemented serum. A significant contribution of soluble factors in the enhanced response observed with bacteria preincubated with azithromycin was excluded. The increase of the chemiluminescence response with azithromycin pretreatment was probably due to improvement in complement-dependent opsonization for strain 1 and to improvement in both serum-independent and serum-dependent opsonization for strain 2.