The aim of this study was to select technetium-99m labelled peptides that can discriminate between bacterial infections and sterile inflammations. For this purpose, we first assessed the binding of various 99mTc-labelled natural or synthetic peptides, which are based on the sequence of the human antimicrobial peptide ubiquicidin (UBI) or human lactoferrin (hLF), to bacteria and to leucocytes in vitro. In order to select peptides that preferentially bind to bacteria over host cells, radiolabelled peptides were injected into mice intraperitoneally infected with Klebsiella pneumoniae (K. pneumoniae) and the amount of radioactivity associated with the bacteria and with the leucocytes was quantitated. The next phase focussed on discrimination between bacterial infections and sterile inflammatory processes using 99mTc-labelled peptides in mice intramuscularly infected with various bacteria (e.g. multi-drug-resistant Staphylococcus aureus) and in animals that had been injected with lipopolysaccharides (LPS) of bacterial origin to create a sterile inflammatory process. Also, we studied the distribution of 99mTc-labelled UBI 29-41 and UBI 18-35 in rabbits having an experimental thigh muscle infection with K. pneumoniae and in rabbits injected with LPS. Based on the results of our in vitro and in vivo binding assays, two peptides, i.e. UBI 29-41 and UBI 18-35, were selected as possible candidates for infection imaging. The radiolabelled peptides can detect infections with both gram-positive and gram-negative bacteria in mice as early as 5-30 min after injection, with a target-to-non-target (T/NT) ratio between 2 and 3; maximum T/NT ratios were seen within 1 h after injection. In rabbits, high T/NT ratios (>5) for 99mTc-labelled UBI 29-41 were observed from 1 h after injection. No accumulation of the selected 99mTc-labelled UBI-derived peptides was observed in thighs of mice and rabbits previously injected with LPS. Scintigraphic investigation into the biodistribution of 99mTc-labelled UBI peptides revealed that these peptides were rapidly removed from the circulation by renal excretion. Similar data were observed for 99mTc-labelled defensin 1-3. Our data for 99mTc-labelled hLF and related peptides indicate that these compounds are less favourable for infection detection. Taken together, 99mTc-labelled UBI 18-35 and UBI 29-41 enable discrimination between bacterial infections and sterile inflammatory processes in both mice and rabbits. Based on their characteristics, we consider these peptides the candidates of preference for detection of bacterial infections in man.

In light of the need for new antifungal agents, the candidacidal activities of human lactoferrin (hLF) and synthetic peptides representing the first, hLF(1-11), and second, hLF(21-31), cationic domains of its N terminus were compared. The results revealed that hLF(1-11) was more effective in killing fluconazole-resistant Candida albicans than hLF(21-31) and much more effective than lactoferrin, as determined microbiologically and by propidium iodide (PI) staining. By using hLF(1-11) and various derivatives, it was found that the second and third residues of the N terminus of hLF(1-11) were critical for its candidacidal activity. Detailed investigation to elucidate the mechanism of action of hLF(1-11) revealed a dose-dependent release of ATP by Candida upon exposure to hLF(1-11). Our observations that sodium azide reduced the PI uptake and candidacidal activity of hLF(1-11) and that, upon exposure to hLF(1-11), the fluorescent dye rhodamine 123 first accumulated inside the mitochondria and later was released into the cytoplasm indicate that the peptide triggers the energized mitochondrion. Furthermore, oxidized ATP, which interferes with the interaction of ATP with its extracellular receptors, blocked the candidacidal action of hLF(1-11), as measured microbiologically and by PI staining. Addition of ATP (or analogues) was not a sufficient stimulus to kill C. albicans or to act synergistically with suboptimal concentrations of the peptide. The main conclusions are that the first two arginines at the N terminus of hLF are critical in the candidacidal activity of hLF(1-11) and that extracellular ATP is essential but not sufficient for the peptide to exert its candidacidal activity.

Neutrophil defensins (or human neutrophil peptides-HNP) are major constituents of the azurophilic granules of human neutrophils and have been shown to display broad-spectrum antimicrobial activity. Other activities of these defensins, which are released from stimulated neutrophils, include cytotoxic, stimulatory, and chemotactic activities toward a variety of target cells. We studied the potential use of HNP-1 for antibacterial therapy of experimental bacterial infections in mice. In experimental peritoneal Klebsiella pneumoniae infections in mice, HNP-1 injection was shown to markedly reduce bacterial numbers in the infected peritoneal cavity 24 h after infection. This antibacterial effect was found to be associated with an increased influx of macrophages, granulocytes, and lymphocytes into the peritoneal cavity. These leukocytes appeared to be a requirement for the antibacterial effect, since in leukocytopenic mice administration of HNP-1 did not display antibacterial activity. HNP-1 treatment also reduced bacterial numbers in experimental K. pneumoniae or Staphylococcus aureus thigh muscle infections. In this model, radiolabeled HNP-1 was found to accumulate at the site of infection, whereas most of the injected HNP-1 was rapidly removed from the circulation via renal excretion. These results demonstrate that neutrophil defensins display marked in vivo antibacterial activity in experimental infections in mice and that this activity appears to be mediated, at least in part, by local leukocyte accumulation. ( J. Clin. Invest. 1998. 102:1583-1590.)

In light of the need for new antifungal regimens, we report that at noncandidacidal concentrations, the lactoferrin-derived peptide hLF(1-11), which is highly active against fluconazole-resistant Candida albicans, acts synergistically with fluconazole against this yeast and a fluconazole-sensitive C. albicans strain as well as C. glabrata, C. krusei, C. parapsilosis, and C. tropicalis. When these yeasts were exposed to hLF(1-11) for 5 min and then incubated with fluconazole, they were killed effectively, while no candidacidal activity was observed when they were incubated first with fluconazole and then exposed to the peptide, indicating that the candidacidal activity is initiated by the peptide while fluconazole is only required during the effector phase. Investigations of the effect of azide, which inhibits mitochondrial respiration, on the activity of combinations of hLF(1-11) and fluconazole against fluconazole-resistant C. albicans revealed that it inhibits this activity, even when added during the effector phase only. As expected, azide inhibited the accumulation of rhodamine 123 in mitochondria and the production and release of ATP by C. albicans that occurred upon exposure to the combination of hLF(1-11) and fluconazole. Accordingly, oxidized ATP (oATP), an antagonist of ATP receptors, completely blocked the candidacidal activity of the hLF(1-11)-fluconazole combination, whereas oATP did not block the activity when its presence was restricted to the effector phase. The candidacidal activity of combinations of hLF(1-11) and fluconazole, which is initiated by the peptide through the involvement of energized mitochondria, renders fluconazole-resistant C. albicans sensitive to this azole.

We previously showed that the energized mitochondrion and extracellular ATP are essential for the candidacidal activity of the N-terminal peptide of human lactoferrin, subsequently referred to as hLF(1-11). The present study focuses on the involvement of internal thiols and reactive oxygen species (ROS) in the candidacidal activity exerted by hLF(1-11). Our results reveal that hLF(1-11) reduced the internal thiol level of Candida albicans by 20%. In agreement, N-acetyl-L-cysteine (NAC), which is a precursor of glutathione and an ROS scavenger, inhibited the candidacidal activity of hLF(1-11). In addition, azodicarboxylic acid bis(N,Ndimethylamide) (diamide), which oxidizes internal thiols, was candidacidal. Furthermore, hLF(1-11) increased the level of ROS production by C. albicans in a dose-dependent manner, and a correlation between ROS production and candidacidal activity was found. 6-Hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid (trolox), which is an ROS scavenger, partially inhibited the hLF(1-11)-induced, but not the diamide-triggered, candidacidal activity. It is of interest that hLF(1-11) and diamide acted synergistically in killing C. albicans and in ROS production. In agreement, oxidized ATP, an irreversible inhibitor of extracellular ATP receptors, partially blocked the hLF(1-11)-induced, but not the diamide-triggered, candidacidal activity. Finally, the hLF(1-11)-induced activation of mitochondria was inhibited by NAC, indicating that internal thiols and ROS affect mitochondrial activity. Therefore, the candidacidal activity of hLF(1-11) involves both generation of ROS and reduction of internal thiols.