Antibacterial mechanisms of the lower respiratory tract. I. Immunoglobulin synthesis and secretion.

Immunoglobulin synthesis and secretion have been studied in the rabbit lower respiratory tract, both in the normal state and after infection with Diplococcus pneumoniae or Listeria monocytogenes. In vitro synthesis of immunoglobulin and specific antibody was assessed by incorporation of 14C-labeled amino acids into protein. Lower respiratory tract secretions and serum were analyzed for immunoglobulin and antibody against the infecting organism. Normal respiratory tract produced small quantities of immunoglobulin, most of which was IgG. After bacterial infection of the lower respiratory tract, there was a marked increase in local synthesis of immunoglobulin, especially IgG. Specific antibody of IgG class was produced in all lungs infected with listeria by the 11th day, and in lungs infected with pneumococcus by the 8th day. Secretions from all normal and infected lower respiratory tracts contained IgA and IgG. The IgA to IgG ratios in secretions of normal animals, and animals infected with listeria or pneumococcus, were 2.3, 2.5, and 2.6, respectively. Sera of animals infected with L. monocytogenes contained specific antibody of IgG class but lacked IgA antibody, whereas secretions had both IgA and IgG class antibody against listeria. Similarly, sera of animals infected with D. pneumoniae had IgG class antibody but no IgA antibody, whereas only IgA antibody was found in secretions. The evidence that locally synthesized immunoglobulin (especially IgA), including specific antibody, is secreted into the lower respiratory tract lumen is discussed. Further definition of the role of "local" antibacterial antibody in the respiratory tract is of considerable importance.

Cell-mediated immunity after bacterial infection of the lower respiratory tract.

Lower respiratory tract and systemic cell-mediated immunity have been studied in rabbits after infection with Listeria monocytogenes or Diplococcus pneumoniae. Respiratory tract cell-mediated immunity was evaluated by direct and indirect assays of migration inhibitory factor (MIF) production. Systemic delayed hypersensitivity was determined by means of intradermal testing with appropriate antigens. Aerosol exposure to listeria was followed by markedly increased numbers of free lower respiratory tract cells. These cells manifested antigen-stimulated inhibition of migration (mean inhibition of migration = 30.4%). Pneumococcal pneumonia was associated with similar but less dramatic changes. Intravenous administration of organisms was uncommonly followed by inhibition of lower respiratory tract cells in direct migration assays. Fractionated MIF, as well as crude supernates of antigen-stimulated lower respiratory tract and lymph node lymphocytes from animals exposed to listeria aerosols, caused inhibition of normal alveolar macrophage migration. MIF, produced by lymph node lymphocytes, has a molecular weight of approximately 65,000 and is inactivated by chymotrypsin or neuraminidase. Delayed dermal hypersensitivity to listeria antigen was observed in 54 of 55 animals exposed to listeria aerosols and in all 9 animals infected by the intravenous route. Delayed dermal reactions to pneumococcal sonicate antigen (but not capsular polysaccharide) followed D. pneumoniae respiratory tract infection in 19 of 28 animals, and was elicited in 5 of 6 animals after intravenous infection. Both local (macrophage migration inhibition) and systemic delayed hypersensitivity followed bacterial infection of the lower respiratory tract. MIF activity was shown to be one mechanism for inhibition of alveolar macrophage migration.

Characteristics and mechanisms of azithromycin accumulation and efflux in human polymorphonuclear leukocytes.

Azithromycin achieves prolonged, high tissue concentrations in spite of low serum levels and obviously must be active at tissue sites of infection to be effective. These unique features prompted us to evaluate the interactions of azithromycin and human polymorphonuclear leukocytes (PMN). Uptake of radiolabeled antibiotic by PMN was determined by a velocity-gradient centrifugation technique and expressed as the ratio of cellular to extracellular drug concentration (C/E). Azithromycin was massively accumulated by human PMN (C/E=387.2 at 2 h). Uptake was not influenced by inhibitors of cellular metabolism, but phagocytosis slightly inhibited the entry of azithromycin into PMN. After removal of extracellular drug, the release (efflux) of azithromycin from PMN was extremely slow. Agents which neutralize lysosomal pH, preventing protonation and trapping of azithromycin, markedly increased antibiotic efflux. Active concentration and prolonged retention of azithromycin by phagocytic cells should allow delivery and subsequent release of accumulated drug at sites of infection.

Spontaneous peritonitis: a reappraisal.

Spontaneous bacterial peritonitis occurs frequently in cirrhotic patients. In order to define more accurately the spectrum of this disease, 55 cases of spontaneous peritonitis were compared with 56 controls (patients with negative ascitic fluid cultures). Of several presenting symptoms, only vomiting (p less than 0.01), fever (p less than 0.05), and severe gastrointestinal bleeding (p less than 0.05) were more prevalent in cases than controls. There were no physical signs and no laboratory studies that separated the two groups except for elevated serum amylases in controls. Studies of peritoneal fluid were rarely helpful and cell counts overlapped in the cases and controls. Spontaneous peritonitis is usually seen in patients with severe liver disease, but there are few distinctive symptoms, signs, or laboratory findings. The mortality rate is high, and it is uncertain from our data that antibiotic therapy alters this prognosis.

Current challenges in antibiotic resistance.

The striking, widespread increase in bacterial resistance to antibiotics is an issue of great concern. Worldwide emergence of antibiotic resistances in our common gram-positive coccal pathogens is probably the most serious problem we have in the realm of bacterial infections. The most important of these organisms are penicillin-resistant Streptococcus pneumoniae, vancomycin-resistant Enterococcus, and methicillin- (and now vancomycin-) resistant Staphylococcus aureus. Although known by the above names, all of these organisms are multidrug-resistant. Beta-lactam and vancomycin resistances in gram-positive cocci are caused by altered cell wallñbinding sites with decreased affinity for the drug. Another serious problem is that of resistance in certain gram-negative bacilli due to extended-spectrum beta-lactamase production. These antibiotic resistances in common pathogens have made antimicrobial therapy of many infections extremely difficult or virtually impossible in some instances. The extensive, and often inappropriate, use of antibiotics in the U.S. and worldwide is the major factor in the emergence and spread of antimicrobial resistance. Microbial mechanisms, epidemiology, clinical importance, treatment, and prevention of these antibiotic resistance problems are discussed.

Inhibition of cell-free oxidative bactericidal activity by erythrocytes and hemoglobin.

Sickle cell anemia and other chronic hemolytic anemias are associated with an increased frequency of bacterial infections. There is evidence to suggest that in hemolytic states massive erythrocyte (RBC) ingestion by macrophages interferes with their antibacterial function, thereby predisposing infection. Stimulated by this possibility, we recently demonstrated that erythrophagocytosis by macrophages markedly inhibited intracellular killing of bacteria, and that zymosan-stimulated superoxide generation and chemiluminescence were also suppressed by RBC ingestion. We examined the effects of RBC components on generation of chemiluminescence, superoxide, and bactericidal activity by cell-free oxidative systems. Generation of chemiluminescence by hypoxanthine-xanthine oxidase was depressed in the presence of human RBC lysate or column-fractionated hemoglobin but not crystallized human hemoglobin (methemoglobin) (peak cpms of 15,522 [P = 0.00024], 28,360 [P = 0.0088], and 50,041 [P = 0.37], respectively, compared with 59,898 for positive controls). Similarly, hypoxanthine-xanthine oxidase production of superoxide was inhibited in the presence of column-fractionated human hemoglobin (43.8 versus 17.4 nmol per tube, P = 0.000001). A cell-free bactericidal system, acetaldehyde and xanthine oxidase with or without myeloperoxidase and Cl-, was markedly inhibited by column-purified hemoglobin. For example, after 2 h of incubation, surviving numbers of Staphylococcus aureus were: control (buffer only), 2.5 X 10(6)/ml; bactericidal system, none; bactericidal system plus hemoglobin, 2.2 X 10(6)/ml (P less than or equal to 0.03, bactericidal system versus other systems). Our studies have documented that interactions between RBC (hemoglobin) and reactive products of oxygen metabolism inhibit oxidative bactericidal mechanisms in cell-free systems as well as in macrophages.(ABSTRACT TRUNCATED AT 250 WORDS)

Interactions of dirithromycin with human polymorphonuclear leukocytes.

Dirithromycin, a new macrolide antibiotic, achieves prolonged, high levels in tissue. We previously demonstrated that certain macrolides are highly concentrated within phagocytic cells. This background information prompted us to evaluate the interactions of dirithromycin and human polymorphonuclear leukocytes (PMNs). After incubation with radiolabeled dirithromycin, antibiotic uptake by PMNs was determined by a velocity-gradient centrifugation technique and was expressed as the ratio of the cellular to the extracellular drug concentration (C/E). Dirithromycin was avidly accumulated by PMNs (C/E, 5 at 15 min, 10 at 30 min, 19 at 1 h, and 35 at 2 h). Uptake was dependent on cell viability, physiologic environmental temperature, and pH (optimum 8.6), but was not influenced by potential competitive inhibitors of membrane transport. Incubation with sodium cyanide caused an increase in dirithromycin accumulation by PMNs. Ingestion of microbial particles (mimicking in vivo infection) modestly inhibited the entry of dirithromycin into PMNs. After removal of extracellular drug, the efflux (release) of dirithromycin from PMNs was slow; only 10% was released within the first 30 min. This prolonged retention of dirithromycin within phagocytic cells might allow delivery and release of accumulated drug at sites of infection. The impact of intraphagocytic dirithromycin on cellular function was also evaluated. In a manner similar to that of other highly concentrated, weakly basic antibiotics, dirithromycin inhibited the respiratory burst response (superoxide production) in stimulated PMNs. The presence of dirithromycin slightly increased the intraphagocytic killing of Staphylococcus aureus in human PMNs. These interactions of dirithromycin with phagocytic cells may promote the extraphagocytic, and possibly the intraphagocytic, killing of infecting organisms.

Influence of pentoxifylline and its derivatives on antibiotic uptake and superoxide generation by human phagocytic cells.

Pentoxifylline modulates multiple activities of stimulated polymorphonuclear neutrophils (PMNs), including the respiratory burst response and membrane transport of certain substances (e.g., nucleosides). We found that several weakly basic antibiotics are highly concentrated by human PMNs and that these drugs also inhibit the respiratory burst response (by a mechanism different from that of pentoxifylline). Since both pentoxifylline and antibiotics will be administered to some patients with serious infections, we have evaluated several types of interactions between these drugs and human PMNs and have attempted to identify the mechanisms that produce alterations in cellular function. Roxithromycin, dirithromycin, and clindamycin were avidly concentrated by PMNs. Pentoxifylline and two derivatives (HWA-448 [torbafylline] and HWA-138 [albifylline]) increased the uptake of these antibiotics by PMNs, both in the resting state and during phagocytosis. Pentoxifylline, HWA-448, HWA-138, and the highly concentrated antibiotics each exerted an inhibitory effect on the stimulated respiratory burst response in PMNs. The combination of both pentoxifylline and a modulatory antibiotic (roxithromycin or clindamycin) inhibited superoxide production to a greater extent than either agent alone. This additive effect might be expected, since pentoxifylline and the modulatory antibiotics influence the respiratory burst activation pathway at different sites. The ability of pentoxifylline to augment the entry of antibiotics into neutrophils has important therapeutic implications. The consequences of this phenomenon might include improved intracellular bactericidal activity as well as efficient antibiotic delivery and release at sites of infection.

The effect of ethanol on adherence and phagocytosis by rabbit alveolar macrophages.

Several defects in host defense mechanisms of chronic alcoholics have been described, but the effect of ethanol on the alveolar macrophage has not been defined. Monolayers of rabbit alveolar macrophages established on plastic culture dishes were exposed to various agents to evaluate the maintenance of adherence. Ethanol produces a rapid onset loss of adherence in this system. The loss of adherence is (1) dose-dependent, (2) transient despite the continued presence of ethanol, and (3) due to a change in the macrophage. Similar rapid-onset, transient loss of adherence is produced by osmolar agents, inducers of cAMP, and colchicine. Phagocytosis of latex particles by rabbit alveolar macrophages in suspension is also transiently decreased by ethanol. These results suggest that inhibition of alveolar macrophage adherence and phagocytosis may be mediated by one or a combination of three mechanisms: osmolar forces altering the cell membrane, changes in cyctic nucleotide concentrations, or disruption of microtubules. Ethanol may act through some combination of these mechanisms to produce functional abnormalities of alveolar macrophages.

Deficiency of serum bactericidal activity against Salmonella typhimurium in sickle cell anaemia.

Systemic salmonellosis is a recognized complication of sickle cell anemia (SCA). In our initial study of SCA host defences against salmonella, we evaluated the bactericidal activity of serum against Salmonella typhimurium. When compared to controls, sera from eight out of nineteen SCA patients were deficient in bactericidal function. Levels of factor B, haemolytic complement and agglutinating antibody were similar in SCA and control sera. However, abnormalities that might theoretically account for the decreased antibacterial activity were observed in many SCA sera. These abnormal findings included: (a) defective function of the alternative complement pathway (decreased bacterial killing in the presence of Mg EGTA); (b) low serum C3 concentration; and (c) decreased total iron-binding capacity (TIBC), with a resultant increase in per cent saturation of iron-binding capacity. Of these deficiencies only the abnormal alternative pathway function was significantly associated with decreased serum bactericidal activity. A suggested function of serum bactericidal activity is prevention of bacteraemia by susceptible organisms. Thus diminished serum bactericidal capacity may increase the risk of Salmonella bacteraemia in some individuals with sickle cell disease.

Antibiotic entry into human polymorphonuclear leukocytes.

Since bacteria which survive within phagocytes may produce serious infection, antibiotics which inactivate these intracellular organisms are needed. To establish those factors which mediate entry of antimicrobial agents into human phagocytes, we studied the uptake of 13 radiolabeled antibiotics by peripheral blood polymorphonuclear leukocytes (PMN). At intervals during a 2-h incubation period, antibiotic uptake by PMN was determined by means of velocity gradient centrifugation, which separates the cell-associated antibiotic from the extracellular antibiotic. Penicillin G and three cephalosporin antibiotics penetrated PMN poorly. The ratio of cellular concentration to extracellular concentration (C/E) of these drugs was less than 0.01 to 0.5. For gentamicin and isoniazid, the C/E values were approximately 0.8 to 1.0. Chloramphenicol, rifampin, and lincomycin, antibiotics with good lipid solubility, were concentrated twofold (C/E = 2) in PMN. Ethambutol (C/E = 5), clindamycin (C/E = 11), and two erythromycin preparations (C/E = 10 to 13) were markedly concentrated within PMN. Clindamycin uptake was rapid: greater than 70% of the total drug entry occurred within the first minute. Accumulation of clindamycin and erythromycin was an active, energy-requiring process, dependent at least in part upon glycolysis. Clindamycin entered PMN by means of an active membrane transport system which was saturable and had a high binding affinity (Km = 2 mM) and maximum velocity of uptake (Vmax = 5 nmol/45 s per 10(6) cells). These observations, together with studies of the biological consequences of intracellular antibiotics, should lead to more effective therapy for infection due to intracellular pathogens..

Effect of phagocyte membrane stimulation on antibiotic uptake and intracellular bactericidal activity.

Phagocytosis stimulated a substantial increase in clindamycin uptake and a smaller increase in rifampin accumulation by polymorphonuclear leukocytes. The effect of this increased antibiotic uptake on intraphagocytic bactericidal activity was evaluated. Although zymosan stimulated antibiotic uptake by polymorphonuclear leukocytes, neither zymosan nor formyl-methionyl-leucyl-phenylalanine enhanced the ability of clindamycin or rifampin to kill ingested staphylococci. Properties other than antibiotic uptake are important in determining intraphagocytic bactericidal activity.

Effects of phagocytosis on antibiotic and nucleoside uptake by human polymorphonuclear leukocytes.

The ability of antibiotics to enter phagocytes during infection with facultative intracellular organisms was investigated using an in vitro model. Human polymorphonuclear leukocytes (PMNLs) were incubated with ingestible particles or phorbol myristate acetate (PMA), after which radiolabeled antibiotics were added to the cell suspension. Antibiotic uptake, determined by a velocity-gradient centrifugation technique, was expressed as the cellular:extracellular (C/E) antibiotic concentration ratio. Phagocytosis or PMA exposure enhanced PMNL clindamycin uptake (for example, C/E ratio of 12 for control vs 30 after zymosan). Entry of penicillin was unaffected and erythromycin uptake was slightly decreased after phagocytosis. Because clindamycin uptake by phagocytes is mediated by the nucleoside transport system, adenosine uptake after phagocytosis was studied. Adenosine uptake was stimulated by phagocytosis, and this increase was inhibited by clindamycin. Thus, clindamycin uptake, mediated by the nucleoside transport system, was augmented by phagocytosis. This marked stimulation of a membrane transport system by phagocytosis has not been previously described.

Membrane transport of clindamycin in alveolar macrophages.

The use of antibiotics which can penetrate phagocytic cells and kill intracellular organisms is desirable in the treatment of chronic facultative bacterial infections. Recently, we reported that several antibiotics were selectively concentrated by rabbit alveolar macrophages. Clindamycin accumulation was especially marked. In the present study we evaluated the plasma membrane transport (initial uptake) of clindamycin in alveolar macrophages. The transport of clindamycin is an active process, as documented by requirements for cellular viability, elevated environmental temperature, metabolic energy, and establishment of the 40- to 50-fold cellular/extracellular gradient. Energy for membrane transport of the drug depended at least in part upon mitochondrial oxidative respiration and cell membrane Na-K pump activity. Kinetic analysis of active clindamycin transport revealed it to be saturable, with a high binding affinity (Km = 1 mM) and a high velocity of uptake (Vmax = 15.8 nmol/45 s per 10(6) cells). Clindamycin uptake was not influenced by the presence of hexose or amino acids, but was inhibited by nucleosides (adenosine, puromycin). Decreased clindamycin transport in the presence of puromycin was typical of competitive inhibition (increased Km, unchanged Vmax). Conversely, competitive inhibition of adenosine transport by clindamycin was documented. Thus, clindamycin is transported into alveolar macrophages via the nucleoside system. The potential biological consequences of this unique antibiotic transport mechanism are of interest.

Contrasts between phagocyte antibiotic uptake and subsequent intracellular bactericidal activity.

An ideal antibiotic for therapy of infections due to facultative intracellular organisms would enter phagocytes readily and kill intracellular bacteria. We have examined the consequences of antibiotic uptake by human polymorphonuclear lymphocytes (PMN) on intraphagocytic bactericidal activity, using antibiotics which differ markedly in their ability to enter PMN. After ingestion of Staphylococcus aureus, PMN were evaluated in regard to uptake of antibiotics and survival of intraphagocytic bacteria in the presence or absence of these drugs. Except for erythromycin, the uptake of which was slightly decreased, the entry of tested antibiotics into PMN was increased or unchanged after ingestion of S. aureus. Clindamycin and erythromycin, which achieved high cellular levels in PMN, failed to produce a significant reduction in viable intraphagocytic S. aureus during 3 h of antibiotic exposure. In contrast, rifampin, which was concentrated severalfold by phagocytes, was able to kill intracellular staphylococci. Gentamicin and penicillin G penetrated PMN rather poorly. However, while gentamicin demonstrated efficient intraphagocytic killing of bacteria, penicillin had no intracellular effect during the 3-h incubation period. These observations document that the ability to enter phagocytes is only one of the factors which determine the intracellular antibacterial activity of an antibiotic.

Uptake of antibiotics by human polymorphonuclear leukocyte cytoplasts.

Enucleated human polymorphonuclear leukocytes (PMN cytoplasts), which have no nuclei and only a few granules, retain many of the functions of intact neutrophils. To better define the mechanisms and intracellular sites of antimicrobial agent accumulation in human neutrophils, we studied the antibiotic uptake process in PMN cytoplasts. Entry of eight radiolabeled antibiotics into PMN cytoplasts was determined by means of a velocity gradient centrifugation technique. Uptakes of these antibiotics by cytoplasts were compared with our findings in intact PMN. Penicillin entered both intact PMN and cytoplasts poorly. Metronidazole achieved a concentration in cytoplasts (and PMN) equal to or somewhat less than the extracellular concentration. Chloramphenicol, a lipid-soluble drug, and trimethoprim were concentrated three- to fourfold by cytoplasts. An unusual finding was that trimethroprim, unlike other tested antibiotics, was accumulated by cytoplasts more readily at 25 degrees C than at 37 degrees C. After an initial rapid association with cytoplasts, cell-associated imipenem declined progressively with time. Clindamycin and two macrolide antibiotics (roxithromycin, erythromycin) were concentrated 7- to 14-fold by cytoplasts. This indicates that cytoplasmic granules are not essential for accumulation of these drugs. Adenosine inhibited cytoplast uptake of clindamycin, which enters intact phagocytic cells by the membrane nucleoside transport system. Roxithromycin uptake by cytoplasts was inhibited by phagocytosis, which may reduce the number of cell membrane sites available for the transport of macrolides. These studies have added to our understanding of uptake mechanisms for antibiotics which are highly concentrated in phagocytes.