Gram-negative sepsis driven by lipopolysaccharide (LPS) has detrimental outcomes, especially in neonates. The neutrophil-derived bactericidal/permeability-increasing protein (BPI) potently neutralizes LPS. Interestingly, polymorphism of the BPI gene at position 645 (rs4358188) corresponds to a favorable survival rate of these patients in the presence of at least one allele 645 A as opposed to 645 G. When we exploited the existing X-ray crystal structure, the corresponding amino acid at position 216 was revealed as surface exposed and proximal to the lipid-binding pocket in the N-terminal domain of BPI. Our further analysis predicted a shift in surface electrostatics by a positively charged lysine (BPI_216K) exchanging a negatively charged glutamic acid (BPI_216E). To investigate differences in interaction with LPS, we expressed both BPI variants recombinantly. The amino acid exchange neither affected affinity towards LPS nor altered bactericidal activity. However, when stimulating human peripheral blood mononuclear cells, BPI_216K exhibited a superior LPS-neutralizing capacity (IC₅₀ 12.0 ± 2.5 pM) as compared to BPI_216E (IC₅₀ 152.9 ± 113.4 pM, p = 0.0081) in respect to IL-6 secretion. In conclusion, we provide a functional correlate to a favorable outcome of sepsis in the presence of BPI_216K.

Early diagnosis of invasive pulmonary aspergillosis (IPA) is crucial to prevent lethal disease in immunocompromized hosts. So far, lipopolysaccharide binding protein (LBP) and bactericidal/permeability-increasing protein (BPI) levels have not been evaluated as biomarkers for IPA. IL-8, previously introduced as a biomarker for IPA, was also included in this study. Bronchoalveolar lavage fluid (BALF) of IPA patients and control patients with non-infectious lung disease was collected according to clinical indications. Measurements in BALF displayed significantly higher levels of LBP (p < 0.0001), BPI (p = 0.0002) and IL-8 (p < 0.0001) in IPA compared to control patients. Receiver operating characteristic curve analysis revealed higher AUC for LBP (0.98, 95% CI 0.95–1.00) than BPI (0.84, 95% CI 0.70–0.97; p = 0.0301). Although not significantly different, AUC of IL-8 (0.93, 95% CI 0.85–1.00) also tended to be higher than AUC for BPI (p = 0.0624). When the subgroup of non-hematological patients was analyzed, test performance of LBP (AUC 0.99, 95% CI 0.97–1.00), BPI (AUC 0.97, 95% CI 0.91–1.00) and IL-8 (AUC 0.96, 95% CI: 0.90–1.00) converged. In conclusion, LBP and—to a lesser extend—BPI displayed high AUCs that were comparable to those of IL-8 for diagnosis of IPA in BALF. Further investigations are worthwhile, especially in non-hematological patients in whom sensitive biomarkers for IPA are lacking.