Clinical isolates of Streptococcus pneumoniae that exhibit tolerance of vancomycin.

The ability of Streptococcus pneumoniae to escape lysis and killing by vancomycin, a property termed "tolerance," has recently been noted in a laboratory strain of the species. Vancomycin tolerance in clinical isolates represents a potential new health risk. We determined the prevalence of vancomycin and penicillin tolerance among 116 clinical isolates of pneumococci by monitoring lysis and viability after exposure to the respective antibiotic for 4 hours. Eight percent of the strains were tolerant to penicillin and 3% were tolerant to vancomycin. The 3 vancomycin-tolerant isolates also had a high ratio of minimum bactericidal concentration to minimum inhibitory concentration, in contrast to nontolerant strains. They were of serotype 9V and had reduced susceptibility to penicillin. Only 1 was also tolerant to penicillin. Growth rate and ability to divide were not affected in the 3 vancomycin-tolerant strains, and they all lysed with deoxycholate, which indicates autolysin production. Vancomycin tolerance among clinical isolates of pneumococci will necessitate tracking to determine the magnitude of the evolving health risk, since tolerance may contribute to treatment failure (in particular, cases of meningitis, in which bactericidal activity is critical for eradication) and since it may also be a favored background for acquisition of resistance of vancomycin.     

Penicillin tolerance in group B streptococci isolated from infected neonates

Four (4%) of 100 isolates of group B streptococci from the blood or cerebrospinal fluid of infected neonates were shown to be tolerant to penicillin by significantly slower rates of killing at penicillin levels that were 16 times greater than the minimal inhibitory concentrations. Different killing rates between nontolerant and tolerant streptococci were observed with penicillin only in the logarithmic phase of growth. Nontolerant strains in the stationary phase and tolerant strains in both the logarithmic and stationary phases were killed at similar rates, which were slower than the rates for nontolerant streptococci in the logarithmic phase. When incubated with [14C]lysine-labeled preparations of cell wall, freeze-thaw extracts of tolerant strains released less radioactivity than did those of nontolerant strains. This activity of nontolerant streptococci was maximal during logarithmic growth and was heat-labile, a result which presumably reflected the activity of autolytic enzyme(s).     

Mechanism of action of penicillin: triggering of the pneumococcal autolytic enzyme by inhibitors of cell wall synthesis.

During penicillin treatment of an autolysin defective mutant pneumococcus we have observed three novel phenomena: (i) Growth of the mutant cultures is inhibited by the same concentrations of penicillin that induce lysis in the wild type. (ii) Mutant bacteria treated with the minimum growth inhibitory concentration of penicillin will lyse upon the addition of wild-type autolysin to the growth medium. Chloramphenicol and other inhibitors of protein synthesis protect the cells against lysis by exogenous enzyme. Sensitivity of the cells to exogenous autolysin requires treatment with penicillin or other inhibitors of cell wall synthesis (e.g., D-cycloserine or fosfonomycin) since exogenous autolysin alone has no effect on bacterial growth. (iii) Treatment with penicillin (or other inhibitors of cell wall synthesis) causes the escape into the medium of a choline-containing macromolecule that has properties suggesting that it contains pneumococcal lipoteichoic acid (Forssman antigen). Each one of these three phenomena (growth inhibition, sensitization to exogenous autolysin, and leakage of lipoteichoic acid) shows the same dose response as that of the penicillin-induced lysis of wild-type pneumococci. On the basis of these findings we propose a new hypothesis for the mechanism of penicillin-induced lysis of bacteria. It is suggested that inhibition of cell wall synthesis by any means triggers bacterial autolytic enzymes by destabilizing the endogenous complex of an autolysin inhibitor (lipoteichoic acid) and autolytic enzyme. Escape of lipoteichoic acid-like material to the growth medium is a consequence of this labilization. Chloramphenicol protects bacteria against penicillin-induced lysis by interfering with the activity of the autolytic enzyme, rather than by depleting the concentration of the enzyme at the cell surface.     

Lipoteichoic acid: a specific inhibitor of autolysin activity in Pneumococcus.

The choline-containing pneumococcal lipoteichoic acid (Forssman antigen) is a powerful inhibitor of the homologous autolytic enzyme, an N-acetylmuramyl-L-alanine amidase (EC 3.5.1.28, MUCOPEPTIDE AMIDOHYDROLASE). Low concentrations of deoxycholate can reverse the inhibition. Wall teichoic acid preparations are inactive at several hundred-fold higher concentrations. Activation of an inactive form of autolysin by in vitro incubation with choline-containing cell walls is also inhibited by lipoteichoic acid. Addition of lipoteichoic acid to the growth medium of pneumococcal cultures causes chain formation, resistance to stationary phase lysis, and penicillin tolerance. It is suggested that a physiological role of lipoteichoic acids may be in the in vivo control of autolysin activity.     

Penicillin resistance and defective lysis in clinical isolates of pneumococci: evidence for two kinds of antibiotic pressure operating in the clinical environment

Seventy percent of clinical isolates of penicillin-resistant pneumococci also exhibit defective lysis when treated with penicillin exceeding the minimal inhibitory concentration (MIC). To provide a possible explanation for the frequent association of these two traits, we exposed penicillin-susceptible pneumococci to two kinds of antibiotic pressures in the laboratory. Treatment of cultures with cycles of high concentrations of penicillin (20 X MIC) followed by growth of the survivors in drug-free medium selected for lysis-defective mutants that died only slowly during antibiotic treatment but had unchanged MICs. Exposure to sustained, low levels of penicillin produced resistant mutants, with elevated MICs, that lysed normally with penicillin. We suggest that the cyclic antibiotic exposure generally used in the clinical setting may select primarily for enhanced survival. From these survivors a second type of antibiotic exposure--sustained antibiotic concentrations just above the MIC (concentrations that may be restricted to the tail-end trough of a dosing interval)--selects for penicillin-resistant mutants.     

Capsular expression in Streptococcus pneumoniae negatively affects spontaneous and antibiotic-induced lysis and contributes to antibiotic tolerance

Penicillin and vancomycin induce a lytic response in Streptococcus pneumoniae that requires the N-acetylmuramyl-l-alanine amidase LytA. We show that clinical isolates of pneumococci of capsular serotypes 1, 4, 6B, and 23F were generally less lytic to penicillin than pneumococci of serotypes 14 and 3. In addition, most 9V isolates were less lytic to vancomycin, compared with isolates of other serotypes. Parent-mutant pairs expressing and not expressing capsular serotypes 2, 4, and 9V were compared for antibiotic-induced lysis. The nonencapsulated variants were considerably more lytic after beta-lactam and/or vancomycin treatment, and antibiotic tolerance was seen only in the context of capsule expression. Conversion from a nonlytic to a lytic phenotype, after loss of capsule expression, required an intact lytA autolysin gene. Exogenous addition of purified LytA gave a lower lytic response in capsulated strains, compared with that in nonencapsulated mutants. Spontaneous autolysis in stationary phase also was negatively affected by capsule expression in an autolysin-dependent manner. Long-term starvation in the stationary phase of the vancomycin- and penicillin-tolerant isolate I95 yielded nonencapsulated mutants that had lost antibiotic tolerance and were lytic to penicillin and vancomycin. The 9V capsular locus of I95 and one of these stationary phase-selected mutants were completely sequenced. The only difference found was a 1-bp frameshift deletion in the cps9vE gene of the lytic mutant, encoding a uridine diphosphate-glucosyl-1-phosphate transferase. Two additional independently isolated lytic mutants of I95 from the stationary phase also contained mutations in the same region of cps9vE, which identified it as a mutational hot spot. This report demonstrates that capsular polysaccharides negatively influence the lytic process and contribute to antibiotic tolerance in clinical isolates of pneumococci.     

Antimicrobial resistance in Streptococcus pneumoniae: an overview.

Clinical resistance to penicillin in Streptococcus pneumoniae was first reported by researchers in Boston in 1965; subsequently, this phenomenon was reported from Australia (1967) and South Africa (1977). Since these early reports, penicillin resistance has been encountered with increasing frequency in strains of S. pneumoniae from around the world. In South Africa strains resistant to penicillin and chloramphenicol as well as multiresistant strains have been isolated. Similar patterns of resistance have been reported from Spain. Preliminary evidence points to a high prevalence of resistant pneumococci in Hungary, other countries of Eastern Europe, and some countries in other areas of Europe, notably France. In the United States most reports of resistant pneumococci come from Alaska and the South, but resistance is increasing in other states and in Canada. Pneumococcal resistance has also been described in Zambia, Japan, Malaysia, Pakistan, Bangladesh, Chile, and Brazil; information from other African, Asian, and South American countries is not available. The rising prevalence of penicillin-resistant pneumococci worldwide mandates selective susceptibility testing and epidemiological investigations during outbreaks.     

Role of tolerance in cloxacillin treatment of experimental Staphylococcus aureus endocarditis

The role of Staphylococcus aureus tolerance was investigated in endocarditis in rats. The efficacies of cloxacillin, gentamicin, and a combination of the two were compared for animals infected with a tolerant strain, its kill-sensitive variant, or a nonisogenic nontolerant strain of S. aureus. Cloxacillin was significantly less effective for treating the tolerant than for the nontolerant strains. The addition of gentamicin to cloxacillin reduced bacterial numbers in endocardial vegetations for the tolerant strain comparable to the reduction by cloxacillin alone for the nontolerant strains, but had no additional effect for the nontolerant strains. Isolates from animals infected with the tolerant or nontolerant strains during antibiotic treatment remained tolerant or nontolerant. These results show that the in vitro phenomenon of tolerance is relevant in vivo.     

Tolerance in Staphylococcus aureaus: evidence for bacteriophage role

The conversion of nontolerant Staphylococcus aureus to a tolerant organism is reported. It had been shown previously that a nontolerant strain produces tolerant progeny when it is incubated in media that contain a bacteria-free filtrate derived from cultures of each of the three tolerant stains of S. aureus. The tolerant progeny retain the characteristic tolerance upon serial subculture, and cell-free filtrates prepared from such subcultures are capable of converting nontolerant organisms to tolerant bacteria. A study of filtrates made from one of the tolerant strains revealed that all of the converting activity was sedimented by ultracentrifugation. DNase I and RNase A did not reduce the activity of the sediment, but Proteinase K completely diminished it. Equilibrium centrifugation of filtrates in cesium chloride showed that the conversion activity was maximal in the fraction with a density of approximately 1.445 +/- 0.015 g/ml. When a portion of this fraction was examined with the electron microscope, hexagonal bacteriophages (20 nm in diameter) were seen at this density. This observation seems to be evidence that a bacteriophage is involved in conversion of S. aureus from a nontolerant to a tolerant organism.     

Tolerance to vancomycin in pneumococci: detection with a molecular marker and assessment of clinical impact

BACKGROUND: Vancomycin is often added to therapy for meningitis caused by Streptococcus pneumoniae. Tolerant bacteria without classic resistance that escape killing by multiple antibiotics have been reported sporadically. We determined the prevalence of tolerance to vancomycin in pneumococci and its effect on the outcome of meningitis. METHODS: Archival samples of 215 nasopharyngeal (NP) and 113 meningitis isolates were tested for the killing efficacy of vancomycin. Specific DNA sequence changes in a transporter locus were identified for tolerant isolates. Similar tests were conducted prospectively on 517 NP isolates from healthy children. RESULTS: In archival isolates, tolerance to vancomycin was detected in 3.7% of NP and 10.6% of invasive isolates. Patients with meningitis caused by tolerant isolates had a worse estimated 30-day survival than did patients with meningitis caused by nontolerant isolates (49% vs. 86%; P = .048); 62.5% of tolerant archival NP isolates harbored a specific sequence change for pep27 and vex2 (P = .021). Prospective analysis of 517 carriage isolates indicated that 8.1% were tolerant to vancomycin and that 82.1% of tolerant isolates harbored the specified marker gene sequences (P = .001). CONCLUSIONS: Tolerance to vancomycin exists in the population of pneumococci. Tolerant isolates are associated with meningitis of increased mortality, and these isolates can be tracked by specific marker sequences in 2 genes.     

Abnormal Autolytic Enzyme in a Pneumococcus with Altered Teichoic Acid Composition

Pneumococci in which the choline component of the cell wall teichoic acid was replaced by ethanolamine contain an abnormal autolytic enzyme that has a low molecular weight and low activity in contrast to the enzyme typical of choline-containing bacteria that has a high molecular weight and high activity. The abnormal autolysin can be converted to the normal (cholinetype) enzyme by incubation in vitro with choline-containing cell walls.     

Penicillin tolerance in beta-streptococci isolated from patients with tonsillitis

Beta-streptococci isolated from patients with acute tonsillitis were tested for penicillin tolerance defined as an MBC/MIC ratio greater than or equal to 16. 11/18 strains recovered from patients with clinical treatment failure were tolerant to penicillin in comparison with 0/15 strains from successfully treated patients. The MBC/MIC ratio was less than 16 for all strains versus cefadroxil but above that ratio for many strains versus clindamycin, doxycycline and erythromycin. We suggest that penicillin tolerance may be one reason to treatment failures in individuals with streptococcal tonsillitis and that other antibiotics could be used to treat these patients since penicillin tolerance is not correlated to a general increase in antibiotic resistance.     

Vancomycin-tolerance among clinical isolates of Streptococcus pneumoniae in Mississippi during 1999-2001

BACKGROUND: In recent years, infection with Streptococcus pneumoniae has been a serious worldwide health concern. Antimicrobial-resistant S pneumoniae is increasing in incidence worldwide, posing a potentially serious threat. Resistance to beta-lactams, macrolides, and trimethoprim-sulfamethoxazole represents a major problem in the treatment of pneumococcal infections METHODS: Our laboratory conducted a survey of local resistance patterns in S pneumoniae. Clinical isolates from two separate respiratory seasons were collected from representative geographic areas in Mississippi (totaling 28 hospitals) and were tested for antimicrobial resistance to penicillin, amoxicillin, ceftriaxone, cefuroxime, azithromycin, clindamycin, tetracycline, trimethoprim-sulfamethoxazole, levofloxacin, gatifloxacin, moxifloxacin, and vancomycin using reference methods. Vancomycin-tolerant strains of S pneumoniae were initially identified as those in which the vancomycin MIC was 0.5 microg/mL. Strain tolerance was confirmed by time kill studies RESULTS: For the 1999-2000 respiratory season, 318 isolates were available for testing; for 2001-2002, 166 isolates were available. Of the 484 total isolates tested, two isolates were identified as having increased tolerance to vancomycin. A greater than 2 log10 difference in viability between the tolerant isolates and the nontolerant isolates of S pneumoniae was observed in time kill studies CONCLUSIONS: Two vancomycin-tolerant isolates of S pneumoniae were identified and characterized. Antibiotic tolerance is defined as the ability of bacteria to survive but not proliferate in the presence of an antibacterial agent. Tolerance to vancomycin is particularly significant when the incidence of penicillin tolerance or resistance is high. In addition, tolerance to vancomycin is not detected by routine in vitro susceptibility testing.     

Microbiological and clinical significance of a new property of defective lysis in clinical strains of pneumococci

A pneumococcal isolate that caused relapsing meningitis in a patient infected with human immunodeficiency virus (HIV) was found to display an unusual response to penicillin--rapid death but a striking lack of cellular lysis. This lytic defect was also detected in all four pneumococcal isolates from three additional HIV-infected patients and in more than half of the clinical isolates from patients with bacteremia. In a rabbit model of meningitis, the lysis-defective strain remained cryptic, with a delay of 5 h in the onset of leukocytosis in cerebrospinal fluid. A marked burst of leukocytosis was associated with ampicillin-induced lysis of a lysis-sensitive strain but not of a lysis-defective strain. Pneumococcal clinical isolates have different lytic responses to penicillin; defective lysis may adversely affect the course of meningitis, an observation suggesting that autolysins play a role in modulating infectious diseases.     

The epidemiology of penicillin resistance in Streptococcus pneumoniae.

Pneumococci are not intrinsically resistant penicillin or other commonly used antibiotics. Penicillin-resistant strains were not encountered until 1965 when two strains were identified in Boston. At that time, the resistance was of a minor degree and its significance was not recognized by the authors. In 1971, resistant strains were encountered in New Guinea and, by the late 1970s, penicillin-resistant pneumococci had spread worldwide. By the early 1980s, areas where more than 10% of isolates were found to be penicillin resistant included Israel, Poland, Spain, South Africa, New Guinea, and the United States from New Mexico to Alaska. In this decade a number of countries including South Korea, Hungary, and Spain have reported dramatic increases in penicillin resistance. Penicillin resistance, once acquired by a virulent pneumococcal clone, has the ability to spread from country to country and to other continents in a relatively short time. Coincident with the increasing penicillin resistance has been the development of resistance to a wide variety of other antibiotics, including other cephalosporins, macrolides, trimethoprim-sulfamethoxazole, tetracycline, and chloramphenicol. Some strains are so highly resistant as to significantly impair our ability to treat patients with meningitis and to choose an appropriate oral agent for the treatment of pneumococcal otitis media.     

Penicillin therapy for treatment of experimental endocarditis caused by viridans streptococci in animals

We studied the efficacy of penicillin and penicillin combined with streptomycin in the treatment of experimental endocarditis caused by viridans streptococci that are susceptible, tolerant, or relatively resistant to penicillin. Rabbits with experimental endocarditis were treated with procaine penicillin (1.5 X 10(5) U/kg) administered twice daily or with procaine penicillin (1.5 X 10(5) U/kg) plus streptomycin (20 mg/kg) administered twice daily for five days. Compared with control animals, animals treated with penicillin alone experienced a significant reduction (P less than .001) of colony forming units per gram of cardiac valve vegetations when infected with streptococci that are susceptible, tolerant, or resistant to penicillin. This antibiotic alone was less effective against streptococci that were tolerant or resistant to penicillin than against streptococci susceptible to the drug (P less than .01). The combination of penicillin and streptomycin was more effective therapy than was penicillin alone in animals with penicillin-tolerant or penicillin-resistant streptococci causing endocarditis (P less than .01). Penicillin-streptomycin therapy was less active against penicillin-resistant strains than against either penicillin-tolerant (P less than .04) or penicillin-susceptible (P less than .01) strains. The results of our study suggest that tolerance or relative resistance to penicillin in strains of viridans streptococci influences the response to therapy with penicillin alone or penicillin combined with streptomycin in the treatment of experimental endocarditis caused by viridans streptococci.     

Antimicrobial resistance in Streptococcus pneumoniae: a South African perspective.

Resistance to penicillin among South African strains of Streptococcus pneumoniae increased from 4.9% in 1979 to 14.4% in 1990. Except for resistance to co-trimoxazole (44%), resistance to other antimicrobial agents remained relatively low. Multiply resistant strains belonged mainly to serovars 6B, 19A, 14, and, more recently, 23F. Use of chloramphenicol to treat meningitis caused by strains relatively resistant to penicillin proved to be unsatisfactory, probably because of the inadequate bactericidal activity of chloramphenicol against these strains. Spread of penicillin-resistant nasopharyngeal strains in pediatric wards was most common among children who received antimicrobial therapy. Penicillin-binding protein (PBP) patterns were shown to vary in resistant clinical strains. Interspecies transfer of penicillin resistance between Streptococcus mitis and S. pneumoniae was demonstrated and antigenic homology was found in PBPs 1A and 2B of strains belonging to these species. Restriction enzyme mapping following DNA amplification of the PBP 2B gene revealed six arrangements among South African strains within serogroup 19. Despite extensive studies in South Africa and several other countries, many questions with regard to the global problem of antimicrobial resistance among S. pneumoniae strains remain unanswered, especially those that relate to prevalence in developing regions of the world.     

Trends in antimicrobial resistance and serotype distribution of blood and cerebrospinal fluid isolates of Streptococcus pneumoniae in South Africa, 1991-1998.

OBJECTIVE: Since 1979, the South African Institute for Medical Research (SAIMR) has served as the national reference center for pneumococcal serotyping and monitoring of antibiotic resistance trends. This study documents trends in antimicrobial resistance in pneumococci isolated from blood or cerebrospinal fluid (CSF) between 1991 and 1998 in South Africa. METHODS: Pneumococcal isolates (n = 7406) from either blood or CSF were sent to the SAIMR reference laboratory for serotyping. The isolates were evaluated for resistance to penicillin, chloramphenicol, tetracycline, erythromycin, clindamycin, and rifampicin. RESULTS: Resistance to one or more antibiotics increased significantly from 19% in 1991 to 1994 to 25% in 1995 to 1998 in all ages, and in children from 32% to 38% (P < 10[-6]). Although penicillin resistance did not increase in children (28.1% vs. 28.9%), penicillin resistance in all ages increased from 9.6% to 18.0%. Significant increases in resistance to chloramphenicol, tetracycline, erythromycin, and rifampicin also were seen in both groups. Multiple resistance increased significantly, from 2.2% to 3.8%. The proportion of isolates with intermediate or high-level penicillin resistance remained constant during the surveillance period. Erythromycin resistance, predominantly expressed as simultaneous resistance to erythromycin and clindamycin, increased from 1.6% to 2.6%. The percentage of erythromycin-resistant isolates that were resistance to erythromycin alone increased from 10.6% to 28.7%, suggesting the emergence of mefE-mediated resistance. In children 2 years of age and younger, although serogroup 6 remained the most common, there were significant increases in serogroups 19, 18, and 13. The percentage of the total invasive pneumococcal disease in this population that is caused by serogroups found in the nonavalent pneumococcal conjugate vaccine (serogroups 1, 4, 5, 6B, 9V, 14, 18C, 19F, 23F) increased from 72% to 91%. CONCLUSIONS: Antibiotic resistance in the pneumococcus is increasing in South Africa, although the proportion of strains with high-level penicillin resistance has not increased. New conjugate vaccines may not only decrease the burden of all pneumococcal disease but, in addition, lower the incidence of antibiotic-resistant disease in South Africa.     

Treatment and diagnosis of infections caused by drug-resistant Streptococcus pneumoniae.

Drug-resistant strains of Streptococcus pneumoniae have now been reported from all continents and have become the predominant pathogens in some areas; many strains are resistant to multiple agents. Because of the importance of pneumococci in the etiology of meningitis, the criteria used to assess susceptibility are conservative: strains with minimal inhibitory concentrations (MICs) of penicillin G of less than or equal to 0.06 microgram/mL are regarded as susceptible, those with MICs of 0.1-1 microgram/mL are considered intermediately resistant, and those with MICs of greater than 1 microgram/mL are designated highly resistant. The diagnosis, treatment, and prevention of infections due to resistant pneumococci are the subjects of this review. Methods of susceptibility testing have now been well defined for pneumococci. Screening for penicillin resistance with 1-microgram oxacillin disks is recommended for all clinically significant isolates. The activity of other beta-lactam agents against penicillin-resistant strains has been documented, and the MICs of a number of non-beta-lactam agents have been determined as well. Treatment of resistant pneumococcal infections depends on the site of infection, the degree of resistance to penicillin G, the resistance of the infecting strain to other agents, the severity of disease, the presence of underlying conditions, and the dose and route of administration of antimicrobial agents. Current recommendations for treatment are based on retrospective case studies, and adequate prospective studies providing more definitive data are needed. Prevention of pneumococcal infections in children less than 2 years of age and in the elderly remains a problem. Improved vaccines must be developed for this purpose.     

Streptococcus pneumoniae deficient in pneumolysin or autolysin has reduced virulence in meningitis

BACKGROUND: The role played by pneumolysin and autolysin in pneumococcal meningitis is poorly understood. METHODS: A rat model was used to investigate the disease, in which surgical implantation of a cisternal catheter allowed bacterial instillation and cerebrospinal fluid (CSF) sampling. RESULTS: CSF infection of rats with wild-type pneumococci caused meningitis within 26 h, whereas isogenic mutants that do not express pneumolysin (DeltaPly) or autolysin (LytA(-)) caused very mild or no disease. Wild-type infections resulted in pneumococci in the CSF and cortical homogenates, but a minority of the rats infected with DeltaPly or LytA(-) had bacteria in these locations at 26 h. Leukocyte numbers in the CSF were similar after infection with all pneumococci; however, neutrophils and monocytes predominated after wild-type infection, whereas lymphocytes and atypical lymphocytes predominated after infection with the mutants. Wild-type pneumococci caused disruption to the ependyma, but this was not observed in rats infected with DeltaPly or LytA(-). Cells surrounding the ventricles in wild type-infected animals expressed caspase 3, and astrocytes had hypertrophy; both findings were absent in rats infected with the mutants. CONCLUSIONS: This study provides strong in vivo evidence that pneumolysin and autolysin play crucial roles in the pathogenesis of pneumococcal meningitis.