Pseudomonas aeruginosa variants isolated from patients with cystic fibrosis are killed by a bactericidal protein from human polymorphonuclear leukocytes

Antioxidant enzyme expression in clinical isolates of Pseudomonas aeruginosa: identification of an atypical form of manganese superoxide dismutase

Expression of superoxide dismutases (FeSOD and MnSOD) and catalases by laboratory strains of Pseudomonas aeruginosa is modulated by exogenous factors. Whether clinical isolates behave similarly and whether antioxidant enzyme expression influences P. aeruginosa virulence remain unclear. Fifty-seven P. aeruginosa blood culture isolates, plus seven pairs of blood and local-site isolates, were examined for FeSOD, MnSOD, and catalase production in vitro. Under iron-replete growth conditions FeSOD and catalase activities were maximized. MnSOD was not detected. FeSOD and catalase activity decreased under iron-limited growth conditions, whereas MnSOD activity appeared. SOD and catalase activity did not change with site of isolation or by patient. MnSOD could not be expressed by one isolate due to a missense mutation in sodA that produced a premature stop codon. Eleven percent of the isolates expressed a novel, rapidly migrating MnSOD that was associated with missense mutations in the normal stop codon of sodA. We conclude that clinical P. aeruginosa isolates vary little in FeSOD and catalase expression. Some strains produce a newly described MnSOD variant, whereas one is deficient in MnSOD production. The absence of MnSOD expression in a P. aeruginosa strain causing invasive human disease indicates that MnSOD is probably not essential for P. aeruginosa virulence.

Structure-function relationships in novel peptide dodecamerswith broad-spectrum bactericidal and endotoxin-neutralizing activities

A series of designed peptide 33-mers (betapep peptides) areknown to be bactericidal [Mayo, Haseman, Ilyina and Gray (1998)Biochim. Biophys. Acta 1425, 81-92]. Here dodecapeptides (SC-1-SC-8), which 'walk through' the sequence ofbetapep-25, were investigated for their ability to kill Gram-negativeand -positive bacteria and to neutralize endotoxin. SC-4 (KLFKRHLKWKI I-NH(2); the -NH(2) at the right of each sequenceindicates amidation of the C-terminal carboxylate group) is the mosteffective, more so than betapep-25, at killing Gram-negative bacteriawith nanomolar LD(50) values. Against Gram-positive bacteria,SC-4 also shows good activity with submicromolar LD(50)values. Leakage studies indicate rapid bacterial membrane permeability,with t(1/2) valuesof 10-15 min. SC-4 in the micromolar range also effectivelyneutralizes endotoxin and is not haemolytic below 10(-4)M. For all SC peptides, CD and NMR data indicate the presence of both 3(10)- and alpha-helix. For SC-4, nuclear-Overhauser-effect-based computational modelling yields an amphipathic helix with K1, K4,R5, and K8 arrayed on the same face (K is lysine, R is arginine). Activity differences among SC peptides and single-site variants of SC-4allow some structure-function relationships to be deduced. Relative to other known bactericidal peptides in the linear peptide,helix-forming category, SC-4 is the most potent broad-spectrumantibacterial identified to date. The present study contributes to thedevelopment of agents involved in combating the ever-recurring problemof drug-resistant micro-organisms.

Designed beta-sheet-forming peptide 33mers with potent human bactericidal/permeability increasing protein-like bactericidal and endotoxin neutralizing activities

Novel peptide 33mers have been designed by incorporating beta-conformation stabilizing residues from the beta-sheet domains of alpha-chemokines and functionally important residues from the beta-sheet domain of human neutrophil bactericidal protein (B/PI). B/PI is known for its ability to kill bacteria and to neutralize the action of bacterial endotoxin (lipopolysaccharide, LPS) which can induce septic shock leading to eventual death. Here, the goal was to make short linear peptides which demonstrate good beta-sheet folding and maintain bioactivity as in native B/PI. A library of 24 peptide 33mers (betapep-1 to betapep-24) were synthesized with various amino acid substitutions. CD and NMR data acquired in aqueous solution indicate that betapep peptides form beta-sheet structure to varying degrees and self-associate as dimers and tetramers like the alpha-chemokines. Bactericidal activity toward Gram-negative Pseudomonas aeruginosa was tested, and betapep-19 was found to be only about 5-fold less potent (62% kill at 1.2 x 10(-7) M) than native B/PI (80% kill at 2.9 x 10(-8) M). At LPS neutralization, betapep-2 and -23 were found to be most active (66-78% effective at 1.2 x 10(-6) M), being only about 50-100-fold less active than B/PI (50% at 1.5 x 10(-8) M). In terms of structure-activity relations, beta-sheet structural stability correlates with the capacity to neutralize LPS, but not with bactericidal activity. Although a net positive charge is necessary for activity, it is not sufficient for optimal activity. Hydrophobic residues tend to influence activities indirectly by affecting structural stability. Furthermore, results show that sequentially and spatially related residues from the beta-sheet domain of native B/PI can be designed into short linear peptides which show good beta-sheet folding and retain much of the native activity. This research contributes to the development of solutions to the problem of multiple drug-resistant, opportunistic microorganisms like P. aeruginosa and of agents effective at neutralizing bacterial endotoxin.

Correlation of Pseudomonas aeruginosa virulence factors from clinical and environmental isolates with pathogenicity in the neutropenic mouse

The potential pathogenicity of a microorganism is a major concern for Health Canada evaluators, who will be processing new biotechnology products under the Canadian Environmental Protection Act. Potential pathogenicity is generally predicted by the results of animal pathogenicity studies. In an attempt to define surrogate data for an animal model, this study was initiated. Pseudomonas aeruginosa isolates from clinical and environmental sources were screened for their pilus type, serotype, lipopolysaccharide type, ability to evade host responses, and production of toxin A, exoenzyme S, elastase, phospholipase C, and total protease. The 50% lethal dose (LD50) of the same isolates was determined in the neutropenic mouse model of infection. An attempted correlation was drawn between each (or combinations) of the virulence determinants and the LD50. Stepwise linear regression showed that the presence of high levels of exoenzyme S in association with elastase or phospholipase C, or to a minor extent toxin A, was correlated with low numbers of bacteria required to elicit an LD50. No correlation between any of the other factors examined and virulence was detected. The data suggest that an in vitro high level of exoenzyme S production could be used as surrogate information for neutropenic mouse modelling; however, the levels of all of the extracellular enzymes should be considered when making such an assessment.

Augmentation of oxidant injury to human pulmonary epithelial cells by the Pseudomonas aeruginosa siderophore pyochelin

Pseudomonas aeruginosa causes acute and chronic infections of the human lung, with resultant tissue injury. We have previously shown that iron bound to pyochelin, a siderophore secreted by the organism to acquire iron, is an efficient catalyst for hydroxyl radical (HO.) formation and augments injury to pulmonary artery endothelial cells resulting from their exposure to superoxide (O2.) and/or H2O2. Sources for O2-. and H2O2 included phorbol myristate acetate (PMA)-stimulated neutrophils and pyocyanin. Pyocyanin, another P. aeruginosa secretory product, undergoes cell-mediated redox, thereby forming O2-. and H2O2. In P. aeruginosa lung infections, damage to airway epithelial cells is probably more extensive than that to endothelial cells. Therefore, we examined whether ferripyochelin also augments oxidant-mediated damage to airway epithelial cells. A549 cells, a human type II alveolar epithelial cell line, was exposed to H2O2, PMA-stimulated neutrophils, or pyocyanin, and injury was determined by release of 51Cr from prelabeled cells. Ferripyochelin significantly increased (> 10-fold) oxidant-mediated cell injury regardless of whether H2O2, neutrophils, or pyocyanin was employed. Apo-pyochelin was not effective, and ferripyochelin was not toxic by itself at the concentrations employed. Spin trapping with alpha-(4-pyrridyl-1-oxide)-N-t-butyl-nitrone-ethanol confirmed the generation of HO., and injury was decreased by a variety of antioxidants, including superoxide dismutase, catalase, and dimethylthiourea. These data are consistent with the hypothesis that the presence of ferripyochelin at sites of P. aeruginosa lung infection could contribute to tissue injury through its ability to promote HO.-mediated damage to airway epithelial cells.

B/PI-derived synthetic peptides: synergistic effects in tethered bactericidal and endotoxin neutralizing peptides

Human neutrophil bactericidal protein (B/PI) is known for its ability to kill bacteria and to neutralize the action of endotoxin. Short linear peptides derived from residues 80-109 have been synthesized and their bactericidal and endotoxin neutralizing activities have been assayed. A series of 'walk-through' decapeptides, overlapping 3 to 4 residues, indicates that endotoxin neutralizing and partial bactericidal activities can be localized within the N- and C-terminal portions, respectively, of the 80-109 sequence. Bactericidal activity toward Pseudomonas aeruginosa was localized in central peptides of the walk-through series and greatest in peptide 90-99. By using longer peptides, residues 86-104 and 82-108, both bactericidal and endotoxin neutralizing activities are significantly enhanced. Bactericidal activity of peptide 82-108 is now only 6-fold less than that of parent B/PI and 9-fold more potent than peptide 86-104. The 82-108 peptide was 7-fold more active at endotoxin neutralization than 86-104 but showed less enhanced activity, being approx. 470-times less active than B/PI. Cyclized 82-108 peptide retained bactericidal activity but did not improve in capacity to neutralize endotoxin.

Bactericidal activity of synthetic peptides based on the structure of the 55-kilodalton bactericidal protein from human neutrophils

Short (10- to 11-mer) hydrophilic peptides based on the structure of the 55-kDa bactericidal protein (BP55, B/PI, and CAP57) from human neutrophil granules were identified from the hydropathy plot of the 456-amino-acid sequence predicted from the nucleotide sequences of cDNA clones for BP55 and B/PI. Peptides corresponding to amino acid residues 90 to 99 (peptide #90-99), 86 to 99, or 90 to 102 of BP55 were bactericidal toward 5 x 10(6) Pseudomonas aeruginosa cells at 0.6 x 10(-5) to 1.5 x 10(-5) M and killed an Escherichia coli rough strain at 3 x 10(-5) M. The #90-99 peptide with a cysteine added at the amino terminus (C#90-99) was approximately 10 times more active than #90-99, killing P. aeruginosa at 1.5 x 10(-6) M. Peptides representing amino acid residues 27 to 37, 118 to 127, and 160 to 170 and the first 10 amino acids of the signal sequence for BP55 were not bactericidal. When coupled to either keyhole limpet hemocyanin or ovalbumin protein carriers through the thiol group, the C#90-99 peptide was not diminished on a molar basis in its capacity for killing of P. aeruginosa. Two other relatively hydrophilic peptides with an added amino-terminal cysteine, peptides C#227-236 and C#418-427, were not bactericidal at 1.2 x 10(-4) M or at 100 times the effective bactericidal concentration of C#90-99. The C#90-99 peptide killed E. coli at 1.5 x 10(-5) M, or at 10 times the concentration required to kill an equal number of P. aeruginosa cells. Although Pseudomonas cepacia and Staphylococcus aureus were resistent to killing by the parent BP55 molecule, they were susceptible to the C#90-99 and #90-99 peptides in the same concentration range as was E. coli. When all peptides were compared for the ability to neutralize E. coli O55:B5 endotoxin in a Limulus amoebocyte lysate assay, the C#227-236, C#418-427, and #160-170 peptides completely inhibited gelation at a 10(-4) M concentration. All other synthetic peptides, including bactericidal peptide #90-99 and its congeners, lacked endotoxin-neutralizing activity at the highest concentration tested (4.5 x 10(-4) M). A hybrid of the C#227-236 and #90-99 peptides (CHybrid) was identical to the C#227-236 peptide component in effectiveness for carrying out endotoxin neutralization and was fivefold better than the #90-99 peptide in its capacity for killing P. aeruginosa.

The role of Pseudomonas aeruginosa elastase in corneal ring abscess formation in pseudomonal keratitis

In order to identify the causative factors of ring abscess, which is the characteristic feature of pseudomonal keratitis, pseudomonal endotoxin, exotoxin A, and elastase were each separately injected into guinea pig cornea. There was no formation of ring abscess. Injection of living Pseudomonas aeruginosa strains IFO3455 and Takamatsu which produce all three molecules, clearly induced ring abscess. In contrast, when heat-killed bacteria strain IFO3455 or living bacteria of the non-elastase-producing strain PA103 were injected, ring abscess was not induced. Furthermore, when living bacteria strain IFO3455 were injected with anti-elastase antibody or a protease inhibitor, ovomacroglobulin, ring abscess formation was significantly inhibited. Histological examination demonstrated that the ring abscess was a dense accumulation and aggregation of polymorphonuclear leukocytes (PMN) with debris of cells and lamellae in the deep stroma at the corneal margins, suggesting prevention of PMN migration to the central lesion. The presence of anti-elastase antibody or a specific elastase inhibitor facilitated PMN migration towards living bacteria strain IFO3455 in an in vitro model. These results indicate that pseudomonal elastase is a necessary but not sufficient factor in the formation of ring abscess in pseudomonal keratitis.

Effector mechanisms of intestinally induced immunity to Pseudomonas aeruginosa in the rat lung: role of neutrophils and leukotriene B4

This paper investigates the effector mechanisms of immune clearance in the lungs of rats immunized against mucoid Pseudomonas aeruginosa. After the gut-associated lymphoid tissue was primed and after a subsequent pulmonary challenge with live bacteria, significantly accelerated bacterial clearances from the lung and raised levels of anti-P. aeruginosa antibodies in sera (immunoglobulin G [IgG], IgA, and IgM) and bronchoalveolar lavages (IgG and IgA) were observed for all immune animals. These changes were associated with enhanced recruitment, chemotaxis, chemokinesis, phagocytic indices, and chemiluminescence of pulmonary polymorphonuclear neutrophils (PMN). In the alveolar spaces of immune animals, an increase in the level of PMN recruitment was not associated with higher levels of leukotriene B4 (LTB4). In contrast, in nonimmune animals that were intratracheally infected with P. aeruginosa, the levels of recruitment and activity of alveolar PMN were lower than those in immune rats but PMN infiltration correlated with a significant increase in the synthesis of LTB4 in the alveolar space. In pulmonary tissue, LTB4 synthesis for both groups was elevated. These findings suggest that accelerated clearance of mucoid P. aeruginosa from the lungs of intestinally immunized rats is due at least in part to factors that induce the enhancement of PMN recruitment and activity in the alveolar space. The mediators that regulate this enhanced response remain unknown but do not seem to include LTB4. The high levels of LTB4 measured in the bronchoalveolar lavages and pulmonary tissues from nonimmune animals infected with live bacteria implicate LTB4 as an important amplifier of the inflammatory response during acute pulmonary infections with mucoid P. aeruginosa in unimmunized hosts.

Human bactericidal/permeability-increasing protein and a recombinant NH2-terminal fragment cause killing of serum-resistant gram-negative bacteria in whole blood and inhibit tumor necrosis factor release induced by the bacteria

The bactericidal/permeability-increasing protein (BPI) of neutrophils and BPI fragments neutralize the effects of isolated Gram-negative bacterial lipopolysaccharides both in vitro and in vivo. Since endotoxin most commonly enters the host as constituents of invading Gram-negative bacteria, we raised the question: Can BPI and its bioactive fragments also protect against whole bacteria? To determine whether the bactericidal and endotoxin-neutralizing activities of BPI/fragments are expressed when Gram-negative bacteria are introduced to the complex environment of whole blood we examined the effects of added BPI and proteolytically prepared and recombinant NH2-terminal fragments on: (a) the fate of serum-resistant encapsulated Escherichia coli, Klebsiella pneumoniae, and Pseudomonas aeruginosa that survive the antibacterial actions of whole blood and (b) the ability of these bacteria to trigger cytokine release. Added BPI in nanomolar concentrations killed each of three encapsulated strains of E. coli and in closely parallel fashion inhibited tumor necrosis factor (TNF) release. Holo-BPI and its NH2-terminal fragment were equipotent toward a rough LPS chemotype K1-encapsulated strain, but the fragment was substantially more potent than holo-BPI toward two encapsulated smooth LPS chemotype strains. TNF release induced by K. pneumoniae and P. aeruginosa was also inhibited by both holo-BPI and fragment but, at the protein concentrations tested, P. aeruginosa was killed only by the fragment and K. pneumoniae was not killed by either protein. The bactericidal action of BPI/fragment toward E. coli is inhibited by C7-depleted serum, but accelerated by normal serum, indicating that BPI, acting in synergy with late complement components, enhances extracellular killing of serum-resistant bacteria. Thus, BPI and an even more potent NH2-terminal fragment may protect against Gram-negative bacteria in the host by blocking bacterial proliferation as well as endotoxin-mediated effects, not only as components of the intracellular antibacterial arsenal of the neutrophil, but also as potentially therapeutic extracellular agents.

Comparison of granule proteins from human polymorphonuclear leukocytes which are bactericidal toward Pseudomonas aeruginosa

Killing of Pseudomonas aeruginosa by a 55-kDa bactericidal protein (BP 55), a 30-kDa protein (BP 30), cathepsin G, elastase, and proteinase 3 has been compared. P. aeruginosa was resistant to killing by elastase and proteinase 3. BP 55 at a 50% lethal dose (LD50) of 0.23 micrograms of protein per 5 x 10(6) bacteria per ml killed P. aeruginosa and was far more active than BP 30 and cathepsin G. The LD50s of BP 30 and cathepsin G were 16.9 and 28.3 micrograms of protein per 5 x 10(6) bacteria per ml, respectively. Preincubation of BP 55 or BP 30 with lipopolysaccharide (LPS) from P. aeruginosa inhibited bactericidal activity. The N-terminal amino acid sequence of BP 55 and BP 30 revealed no relationship between the two proteins. However, a monoclonal antibody (AHN-15) reacted with both proteins by Western immunoblot. The bactericidal activity of cathepsin G toward P. aeruginosa appeared to be dependent on the availability of the active site of the enzyme; bactericidal activity was inhibited by phenylmethylsulfonyl fluoride (PMSF) and by the specific cathepsin G inhibitor, Z-Gly-Leu-Phe-CH2Cl. The enzyme and bactericidal activities of cathepsin G were also inhibited by LPS from P. aeruginosa. LPS from P. aeruginosa was shown to be a competitive inhibitor of the enzyme activity of cathepsin G. Elastase enzyme activity was also inhibited noncompetitively by LPS, but the enzyme was not bactericidal. We have concluded that all three bactericidal proteins (BP 55, BP 30, and cathepsin G) may bind to the LPS of the outer membrane of P. aeruginosa. It appears that the enzyme active site must be available for cathepsin G to kill P. aeruginosa and that the active site may be involved in the binding of cathepsin G to P. aeruginosa.