Rapid film-based determination of antibiotic susceptibilities of Mycobacterium tuberculosis strains by using a luciferase reporter phage and the Bronx Box.

Detecting antibiotic resistance in Mycobacterium tuberculosis is becoming increasingly important with the global recognition of drug-resistant strains and their adverse impact on clinical outcomes. Current methods of susceptibility testing are either time-consuming or costly; rapid, reliable, simple, and inexpensive methods would be highly desirable, especially in the developing world where most tuberculosis is found. The luciferase reporter phage is a unique reagent well-suited for this purpose: upon infection with viable mycobacteria, it produces quantifiable light which is not observed in mycobacterial cells treated with active antimicrobials. In this report, we describe a modification of our original assay, which allows detection of the emitted light with a Polaroid film box designated the Bronx Box. The technique has been applied to 25 M. tuberculosis reference and clinical strains, and criteria are presented which allow rapid and simple discrimination among strains susceptible or resistant to isoniazid and rifampin, the major antituberculosis agents.

Contrasting effects of identical nutrients given parenterally or enterally after 70% hepatectomy: bacterial translocation.

High mortality occurs in rats with 70% hepatectomy fed intravenous (IV) total parenteral nutrition (TPN; 13.9% glucose, 4.17% amino acids, 1.46% fat, electrolytes, trace minerals, and vitamins providing 216 kcal.kg-1.d-1) but not when the identical nutrients are given at the same rate enterally (gastrostomy). We hypothesized that a difference in bacterial translocation (BT) was a contributing factor to this phenomenon. Forty-five male Sprague-Dawley rats (300-360 g) were divided into five groups and underwent the following: control (no operation), sham (intraperitoneal [IP] pentobarbital anesthesia, central venous and gastrostomy catheters, laparotomy, sham hepatectomy), standard oral feeding (SOF), TPN (IV nutrients), and total enteral nutrition (TEN; gastrostomy). The SOF, TPN, and TEN groups had IP pentobarbital anesthesia, central venous and gastrostomy catheters, and 70% hepatectomy. Postoperatively, control and SOF (both catheters plugged) rats ate a commercial rat chow and drank tap water ad libitum pre- and postoperatively. The sham, TPN, and TEN groups were given the identical infusate composition as above, but the nutrient concentrations were cut in half (110 kcal/kg) and three-quarters (165 kcal/kg) on postoperative days 1 and 2, respectively. At the end of postoperative day 2, all rats were euthanized. BT to mesenteric lymph nodes (MLNs), liver, spleen, and lungs was significantly higher in the TPN rats compared with all other groups, except that BT to the MLNs was similar in the TPN and TEN groups. Bacteremia was found only in the TPN rats. BT in TPN rats with 70% hepatectomy was significantly greater 48 h after operation than in those fed the identical nutrients enterally at the same rate; this correlates with the previously reported significantly greater mortality in rats with 70% hepatectomy receiving TPN.

Single local instillation of nonviable Staphylococcus aureus or its peptidoglycan ameliorates glucocorticoid-induced impaired wound healing.

An excess in glucocorticoid steroids, either from endogenous or exogenous sources, has been shown to inhibit wound repair. Key to this impairment is a diminution of the inflammatory response to wounding, fibroplasia, capillary formation, reparative tissue collagen accumulation, and wound breaking strength. Because a single local application at operation of nonviable Staphylococcus aureus or its peptidoglycan increases all of these processes in normal rats, we hypothesized that nonviable S. aureus and S. aureus peptidoglycan would each ameliorate glucocorticoid-induced impaired healing. Sprague-Dawley male rats aseptically received two 7 cm paravertebral skin incisions and underwent subcutaneous implantation of polyvinyl alcohol sponges. Two glucocorticoids were used: hydrocortisone, 8 mg intramuscularly, daily beginning 1 day before operation and continuing during the postoperative period; or a single dose of a long-acting preparation of methylprednisolone, 6 or 8 mg intramuscularly, on the day before operation. Controls received intramuscular injections of saline solution at the same respective times. At the time of the operation, one incision and the polyvinyl alcohol sponges on one side of the animal were instilled with saline solution while the incision and sponges on the opposite side were instilled with nonviable S. aureus (hydrocortisone study) or S. aureus peptidoglycan (two methylprednisolone studies). The data showed that, at postoperative day 7, the single local application at wounding of nonviable S. aureus or S. aureus peptidoglycan increased wound breaking strength in the control rats by factors of 1.6 in the hydrocortisone experiment and 1.4 and 1.6 in the methylprednisolone studies. These treatments prevented (in hydrocortisone-treated rats) or mitigated (in methylprednisolone-treated rats) the glucocorticoid-induced decrease in wound breaking strength. In addition, these treatments prevented the glucocorticoid-induced decreases in the inflammatory (largely mononuclear cells) response to wounding and in the accumulation within the polyvinyl alcohol sponge of reparative tissue fibroblasts, capillaries, and collagen.

Accelerating effects of nonviable Staphylococcus aureus, its cell wall, and cell wall peptidoglycan.

We have previously reported that local application of viable Staphylococcus aureus dramatically accelerates wound healing, but viable Staphylococcus epidermidis does not. Because the S. aureus effect occurred in the absence of infection and because the cell walls of the two bacterial species differ, we hypothesized that nonviable S. aureus, its cell wall, and its cell wall component(s) would accelerate healing. Nonviable S. aureus was prepared by chemical and physical means, and its cell wall and peptidoglycan was prepared from heat-killed cultures. In a large number of experiments, nonviable S. aureus (independent of the strain's protein A content), its cell wall, and peptidoglycan when instilled locally at the time of wounding each significantly increased the breaking strength of rat skin incisions (tested both in the fresh state and after formalin fixation). These agents also enhanced subcutaneous polyvinyl alcohol sponge reparative tissue collagen accumulation, generally by a factor of two. Histologic features of treated and control incisions were similar. In contrast, the reparative tissue of treated sponges contained more neutrophils, macrophages, capillaries, and collagen. These experimental data thus confirm our previous studies, as well as our hypothesis, and extend these observations of enhanced wound healing to specific fractions of the bacterial cell wall.

Restriction fragment length polymorphism analysis of Cryptococcus neoformans isolates from environmental (pigeon excreta) and clinical sources in New York City.

Restriction fragment length polymorphism analysis of environmental (pigeon excreta) and clinical Cryptococcus neoformans var. neoformans isolates in a limited geographic area distinguished 6 strains among 8 environmental isolates and 12 strains among 17 clinical isolates. Clusters of patients with three strains types accounted for 47% of clinical isolates. Despite this diversity, two strains were shared by environmental and clinical isolates.

False-negative cerebrospinal fluid cryptococcal latex agglutination tests for patients with culture-positive cryptococcal meningitis.

Three cases of false-negative cerebrospinal fluid latex agglutination test results for patients with culture-positive cryptococcal meningitis are reported. False-negative results occurred in settings of low cryptococcal antigen concentrations in cerebrospinal fluid and were dependent on the latex agglutination test kit used. Investigation of each case revealed that prozone phenomena or interference from bound antibody or protein could not account for the false-negative results.

Effect of prestorage white cell reduction on bacterial growth in platelet concentrates.

Platelet concentrates stored with and without autologous white cells were produced from units of whole blood that had been purposefully contaminated with bacteria immediately after phlebotomy. The blood was inoculated with one of five species of bacterium at either 10 or 50 colony-forming units per mL. The growth of the organisms was quantified throughout the conventional 5-day, 22 degrees C storage period of the platelet concentrates. One species, Klebsiella pneumoniae, failed to grow in any of the components. The remaining species, Staphylococcus epidermidis, S. aureus, Enterococcus faecalis, and Salmonella enteritidis, achieved log growth after 1 day of storage and reached a relative maximum concentration by Day 3. Although the concentration of bacteria immediately after inoculation was lower in the units reduced in white cells by filtration, no significant differences were observed thereafter. Data from this in vitro study support the concept that prestorage white cell reduction of platelet concentrates should not increase the likelihood of transfusion-induced septicemia.

Persistence of initial infection in recurrent Cryptococcus neoformans meningitis.

Patients with cryptococcal meningitis tend to have recurrences of infection. Although the original strain of Cryptococcus neoformans is assumed to persist in recurrent infections, this assumption has not been tested. Southern blot hybridisation with two genomic DNA probes and pulsed-field electrophoresis of intact chromosomes were used to investigate the genetic relation between initial and relapse isolates of C neoformans from patients with recurrent cryptococcal meningitis. Eleven isolates were obtained from four patients (three with AIDS, one with leukaemia). Isolates from each patient could be distinguished from those of the other patients; however, each patient's initial and recurrence isolates were clonally related. Our results provide strong evidence that clinical recurrences of cryptococcal meningitis result from persistence of the original infecting strain.

Comparison of maternal sera, cord blood, and neonatal sera for detecting presumptive congenital syphilis: relationship with maternal treatment.

The incidence of congenital syphilis has increased rapidly over the past few years. Most infected mothers and their newborns are asymptomatic at birth and diagnosis depends on serologic testing during pregnancy and at delivery. This study was initiated to compare maternal sera, cord blood, and neonatal sera for detecting presumptive congenital syphilis and to assess the role of maternal treatment (administration of penicillin to the mother at least 1 month before delivery) on the serologic results at the time of delivery. The serologic results from all live deliveries complicated by a positive maternal and/or neonatal test for syphilis during a 12-month period were compared using chi 2 analysis and multiple comparisons for proportions. Of 3306 livebirths, 73 (2.2%) were complicated by a positive maternal or neonatal serology. At delivery, the serologic test was positive in 68 (94%) of 72 maternal sera, 30 (50%) of 60 cord sera, and 43 (63%) of 68 neonatal sera. In the absence of maternal treatment, 95% of the maternal sera, 66% of the cord blood samples, and 86% of the neonatal sera were positive. If the mother had been treated, 94% of maternal sera, 36% of cord sera, and 39% of neonatal sera were positive. Cord blood and neonatal sera appear to be inferior to maternal sera for detecting prenatal exposure to syphilis. Cord serology is also inferior to neonatal serology at 2 to 3 days of age. The most effective way to identify newborns at risk for congenital syphilis is to obtain a maternal serologic diagnosis during pregnancy and to test maternal and neonatal sera at delivery.

Prevention of growth of Yersinia enterocolitica in blood by polyester fiber filtration.

The ability of polyester white cell-reduction blood filters to prevent the growth of Yersinia enterocolitica in units of donated blood was studied. Sixteen units of freshly drawn blood were inoculated with 10, 50, 100, or 150 colony-forming units (CFU) per mL of a clinical isolate of Y. enterocolitica (serotype O:3). The units were subsequently fractionated into red cell concentrate and resuspended in AS-1 or AS-3 solution. One-half of the red cell concentrates in each solution were filtered within 15 hours of phlebotomy and stored for 42 days. The remaining units served as unfiltered controls. Bacterial growth was monitored by weekly cultures and, on the last storage day, by the presence of endotoxin and the formation of methemoglobin. One hundred twelve primary cultures (560 plates) were performed. Units collected in AS-1 and filtered remained sterile when initially inoculated with 50 CFU or less. Filtered units spiked with 100 CFU or less and collected in AS-3 remained sterile throughout their shelf life. All unfiltered units supported bacterial growth and the formation of endotoxin and methemoglobin. The filtration of freshly donated blood proves to limit the growth of Y. enterocolitica in red cell components.

Extensive allelic variation in Cryptococcus neoformans.

The orotidine monophosphate pyrophosphorylase (OMPPase) gene locus of the DNA of 13 Cryptococcus neoformans var. neoformans strains, including 10 recent clinical isolates, was studied by using restriction fragment length polymorphisms and nucleotide sequence analysis. The OMPPase locus (URA5) is highly polymorphic, and at least six alleles were identified. The nucleotide sequences of some alleles differed by up to 5%. The majority of the nucleotide polymorphisms in the protein-coding region occurred at the third codon position and were silent. The low frequency of replacement nucleotide substitutions relative to silent nucleotide substitutions implied that there is strong selection against amino acid changes in OMPPase. The allelic variation suggested that there is extensive genomic diversity among C. neoformans clinical isolates from one geographic area. The various alleles are potentially useful markers in the study of the population structure, epidemiology, and pathogenesis of C. neoformans strains.

Comparison of methods for the immunological identification of Neisseria gonorrhoeae in clinical specimens using commercially-obtained reagents.

Neither an immunofluorescent nor a co-agglutination method was adequately sensitive for the detection of Neisseria gonorrhoeae when commercially-obtained reagents were used to test oxidase positive organisms taken directly from Transgrow medium. The sensitivity of co-agglutination, the better of the two methods, was improved from 83% to 96% when organisms were subcultured for colony isolation prior to identification. False positive results were obtained with both methods.

Piperacillin: evaluation against inpatient and gentamicin-resistant bacteria.

When species-matched collections, each consisting of 128 bacterial strains, were compared, inpatient organisms were found to be significantly more resistant to piperacillin than outpatient organisms. A collection of 143 gentamicin-resistant gram-negative bacilli was significantly more resistant to piperacillin than a species-matched collection of gentamicin-sensitive organisms. Piperacillin resistance was transferred by conjugation from isolates of 8 different Enterobacteriaceae species to a recipient Escherichia coli strain. In each of 67 mating experiments in which transfer of carbenicillin resistance by conjugation occurred, resistance to ticarcillin, azlocillin and piperacillin was also transferred. As in previous studies, piperacillin was found to be markedly more active against non-fastidious gram-negative bacilli than carbenicillin, ticarcillin or azlocillin. Nevertheless, piperacillin resistance, apparently plasmid mediated, was detected in multiple species.

Susceptibility of streptococci to newer tetracyclines and cephalosporins and to other antimicrobial agents.

Two hundred and thirty-nine strains of Streptococcus including 71 strains of Group A, 81 strains of Group B, 69 strains of enterococci, and 18 strains of S. pneumoniae were tested against 12 antimicrobial agents using an agar dilution method. Cefamandole was the most active cephalosporin tested. Doxycycline and minocycline were more active than tetracycline, although the tetracyclines were considerably less inhibitory than the cephalosporins. Regression line analysis of zones of inhibition versus minimal inhibitory concentration values for tetracycline and minocycline showed the tetracycline disc to be unacceptable for predicting the susceptibility of the Group A Streptococcus to minocycline. Minimal inhibitory concentrations for clindamycin, erythromycin, chloramphenicol, nitrofurantoin, and spectinomycin are also given.

Sensitivity of gentamicin-resistant Enterobacteriaceae to cefamandole and cefoxitin.

A collection of gentamicin-resistant Enterobacteriaceae strains was significantly more resistant to cefamandole than a species-matched collection of gentamicin-sensitive organisms. Cefamandole and gentamicin resistance could be simultaneously transferred by conjugation from four different species, Citrobacter freundii, Escherichia coli, Klebsiella pneumoniae and Serratia marcescens, to a recipient E. coli strain. Plasmids specifying linked resistance to cefamandole and gentamicin are thus commonly present in the environment of our medical center despite the fact that cefamandole has never been used here. Gentamicin-resistant organisms were not more resistant to cefoxitin than gentamicin-sensitive isolates. Conjugal transfer of cefoxitin resistance could not be demonstrated. Relative resistance to cefoxitin was nevertheless common among Enterobacter and Serratia isolates.

Laboratory evaluation of a multitest system for identification of gram-negative organisms.

Seventy-two oxidase-positive or nonfermentative organisms, or both, all of which could be identified with reasonable certainty by alternative means, were used to challenge the OXI/FERM tube. There was 91% concurrence of identification between the two methods. Three observers, working independently, agreed upon the interpretation of each of the tests in the OXI/FERM system from 91% to 100% of the time. Duplicate OXI/FERM tubes inoculated with the same strain yielded identical interpretations 95% of the time for all tests except citrate utilization (89%). The OXI/FERM tube thus appeared to be an accurate identification method and yielded test results that were easily interpreted and adequately reproducible.

In vitro antibiotic sensitivity of Moraxella species.

Minimal inhibitory concentrations of 17 antibacterial agents for 34 Moraxella strains were determined using a plate dilution method. A strain of Moraxella nonliquefaciens was found which produced beta-lactamase and was resistant to ampicillin and carbenicillin but not to cephalothin. Several strains were relatively resistant to erythromycin and sulfisoxazole. Disk sensitivity tests could be used to reliably predict penicillin and erythromycin resistance but not sulfisoxazole resistance.