Journal of Microbiological Methods 30 (1997) 153–160 Journal of Microbiological Methods Rapid detection of Enterobacteriaceae in urine by fluorescent 16S rRNA in situ hybridization on membrane filters M.W. Mittelman a a,b, *, M. Habash 1 , a , J.-M. Lacroix 2 , a , A.E. Khoury c , M. Krajden b , d Centre for Infection and Biomaterials Research, Bell G-631, The Toronto Hospital, University of Toronto, 200 Elizabeth St., Toronto, Ontario M5 G 2 C4, Canada b Department of Clinical Biochemistry, University of Toronto, Toronto, Ontario, Canada c Hospital for Sick Children, Toronto, Ontario, Canada d Department of Microbiology, The Toronto Hospital, Toronto, Ontario, Canada Received 6 May 1997; received in revised form 30 June 1997; accepted 30 June 1997 Abstract The detection and enumeration of putatively pathogenic bacteria in urine is the single-most important diagnostic criterion for the diagnosis of urinary tract infections (UTI). We have developed a fluorescent in situ hybridization (FISH) assay which utilizes a fluorescently-tagged 16S rDNA oligonucleotide probe specific for 16S rRNA of the Enterobacteriaceae family. The technique involves fixation of a urine specimen, filtration through a 0.2 m m polycarbonate membrane, staining with the nucleic acid dye, 49,6-diamidino-2-phenylindole (DAPI), hybridization with a fluorescently tagged nucleic acid probe, and examination under epifluorescence microscopy. The technique was found to be sensitive and specific, recovering #10 3 Escherichia coli / ml within 4 h, both in spiked PBS and in filter-sterilized urine. Two non-Enterobacteriaceae, Pseudomonas aeruginosa and Staphylococcus aureus, did not react with the probe but were visualized via DAPI staining. Eighty-three urine specimens were randomly selected from the clinical laboratory and assayed using this new method. A total of 10 specimens were identified by the hospital laboratory as containing members of the Enterobacteriaceae family, including E. coli and Proteus mirabilis. All 10 of these specimens were also positive by the membrane-based FISH assay. Of those specimens characterized as either ‘no growth’ or ‘mixed organisms’ by the hospital laboratory, 24 were positive using the membrane-based FISH assay. This FISH assay for bacteriuria shows promise as a rapid technique for use in clinical diagnosis of urinary tract infections.  1997 Elsevier Science B.V. Keywords: Bacteriuria; FISH assay; Fluorescence microscopy; Membrane filter; UTI 1. Introduction Urinary tract infections (UTIs) are a significant *Corresponding author. Tel.: 11 416 3404564; fax: 11 416 3404607; e-mail: marc.mittelman@utoronto.ca 1 Present address: Department of Microbiology and Immunology, University of Western Ontario, London, Ontario. 2 Present address: Visible Genetics, Inc., Toronto, Ontario. cause of morbidity and mortality world-wide. Although over 90% of UTIs are relatively simple infections in anatomically normal patients, they nevertheless result in considerable morbidity, lost work time, and a financial burden to the health-care system [1]. In complicated UTIs, renal damage leading to kidney failure and death can occur. UTIs account for a significant proportion of all antibiotic prescriptions written in North America; indeed, in 0167-7012 / 97 / $17.00  1997 Elsevier Science B.V. All rights reserved. PII S0167-7012( 97 )00061-4 154 M.W. Mittelman et al. / Journal of Microbiological Methods 30 (1997) 153 – 160 1988, two million prescriptions were written for UTIs in Canada alone [2]. The majority of patients with UTIs complain of lower urinary tract symptoms such as painful urination, frequency, and suprapubic pain. However, these symptoms are also seen with other conditions such as vaginitis and sexually transmitted diseases. Confirmation of a UTI is made upon microbiological culture of urine from patients presenting with urinary tract symptoms. Approximately 90% of all cases are caused by Gram-negative bacteria of fecal origin [1]. The current standard for the microbiological diagnosis of UTI involves quantitative culture of the urine using a calibrated loop on selective and / or non-selective culture media. Following an 18–24 h incubation, colonies are counted and ordinarily reported in gradations of 0, or 10 000 to .10 5 cfu / ml [3]. Among symptomatic patients, particularly women, a count of $10 5 cfu / ml has been adopted as a criterion to distinguish infection from adventitious ‘contamination’. However, as many as 45% of women presenting with clinical symptoms and subsequently treated with antibiotics are culture negative with routine isolation procedures [2]. Although the culture technique remains an important diagnostic tool, there are two major problems associated with the viable plate count method. Firstly, the long incubation time means that therapy is often initiated prior to diagnosis. Secondly, the viable-count techniques may not detect fastidious pathogens or those bacteria which are viable, but not culturable [4,5]. Patients on antibiotic therapy may have antibiotic resistant microorganisms that are capable of causing disease, but incapable of growth on standard laboratory media. While the risks associated with acute infections may be sufficient to warrant initiation of antibiotic therapy prior to the availability of culture results, the clinician would benefit from a definitive diagnosis — or exclusion — of bacteriuria. Definitive diagnosis of bacteriuria can also be critical for patients suffering from end-stage renal failure or drug allergies, where prescribing of antibiotics must be carefully weighed against the risk of side effects. A number of rapid screening techniques for bacteriuria have been evaluated. These include endotoxin [6], leukocyte esterase / nitrate reduction [7], catal- ase [8], DNA probe [9] and substrate mineralization [10] assays. Each of these methods suffers from one or more disadvantages, however, including poor sensitivity and / or specificity. Direct, viable-count techniques have been employed for detecting viable, but non-culturable microorganisms in the environment. These types of techniques have been used for a number of years to detect both bacterial and eucaryotic cellular activity. The need to recover slow-growing, chlorine-damaged, and fastidious bacteria has been the driving force behind development of direct, viable-count assays [11]. Stains employed have included 2-( p-iodophenyl)-3-( p-nitrophenyl)5-phenyltetrazolium chloride [12], fluorescein diacetate [13], and 5-cyano-2,3-ditolyl tetrazolium chloride (CTC). None of these techniques specifically identify putative pathogen groups; such a technique would be important for making an accurate diagnosis of UTI. Fluorescent in situ hybridization (FISH) of bacterial 16S rRNA with oligonucleotide probes has been used for the detection of bacteria in environmental and clinical specimens. Identification of single bacterial cells within natural microbial communities has been accomplished using a series of fluorescently tagged oligonucleotide probes [14,15]. Gersdorf et al. [16] used a FISH assay to detect putative periodontal pathogens in subgingival plaque in humans. The uropathogenic protozoan, Trichomonas vaginalis, was detected in clinical samples using an in situ DNA hybridization assay. Probes specific to the species, genus, and family levels have been developed for several human pathogens [17]. It is the combination of highly conserved and variable regions within the 16S ribosome which enable the synthesis of specific probes targeted to various phylogenetic levels [18]. We have developed a membrane-based FISH assay which utilizes fluorescently-tagged 16S rDNA oligonucleotide probes specific for 16S rRNA of the Enterobacteriaceae family. This rapid, membranebased assay demonstrates good specificity and selectivity for a significant group of uro-pathogens. The assay can be designed to target various putative uropathogens and specifically quantify the total number of microorganisms, allowing for laboratoryclinical correlations. M.W. Mittelman et al. / Journal of Microbiological Methods 30 (1997) 153 – 160 2. Materials and methods 2.1. Reagents and materials 2.1.1. Bacteria Overnight cultures of clinical strains of Escherichia coli, Pseudomonas aeruginosa, and Staphylococcus aureus were prepared from nutrient broth tubes incubated at 378C. Spiking experiments were carried out over a range of bacterial concentrations, from 10 2 –10 7 cfu / ml. 2.1.2. Hybridization buffer The buffer consisted of 63 SSC (13 SSC: 0.15 M NaCl, 0.015 M sodium citrate), 20 mM Tris–HCl (pH 7.0, 258C), 0.1% SDS, 0.01% poly (A), and 0.2% diethylpyrocarbonate in 100 ml of deionized water. Double distilled water was treated with diethylpyrocarbonate (0.2%, w / v), and autoclaved for RNAse inactivation. 2.1.3. Wash buffer The buffer consisted of 0.9 M NaCl, 20 mM Tris–HCl (pH 7.2, 258C), 0.1% SDS, and 0.2% diethylpyrocarbonate in 500 ml of deionized water. 2.1.4. Fluorescent-probe The probe had the following sequence: 59 2 CAT GAA TCA CAA AGT GGT AAG CGC C 2 39 (synthesized by BioMolecular Assays, Inc., Woburn, MA) and labeled on its 59 end with tetramethylrhodamine isothiocyanate. This sequence is homologous to the bases 1458 to 1482 of the positive strand 16S rDNA sequence of Escherichia coli [19]. A search of 329 017 sequences in the BLASTn database [20] revealed that this sequence is highly conserved among members of the Enterobacteriaceae family (including the common UTI pathogen genera of Escherichia, Proteus, Yersinia, Klebsiella). A few non-Enterobacteriaceae (Plesiomonas shigelloides and an insect endosymbiont) also showed 100% homology to this oligonucleotide sequence. However, these organisms have not been previously associated with UTIs. 155 2.1.5. Gelatin-coated slides Plastic coated slides (Cell Tek Inc., Glenview, IL) were precleaned in a solution of 10% KOH (w / v in 95% ethanol) for 1 h at room temperature. The slides were rinsed with distilled water and air dried. The slides were then dipped twice in a gelatin solution (0.1% gelatin 150 bloom, BDH Chemicals Toronto, and 0.01% chromium potassium sulfate, w / v in distilled water) maintained at 708C. Following gelatin coating, the slides were dried in a vertical position overnight at 258C. 2.2. Specimen collection A total of 83 patient urine specimens were randomly selected over a 2-week period from the microbiology laboratory at The Toronto Hospital, General Division. Specimens were assigned a study number; both identity of the patients and the preliminary diagnoses were blinded to the investigators. The majority of specimens were obtained from inpatients, both with and without indwelling urinary catheters. Ten-ml aliquots were removed from specimens within 4 h of patient collection. Specimens were maintained at 48C during the interval between patient collection and aliquoting. 2.3. Methods 2.3.1. Specimen fixation and filtration One ml of the urine specimen was added to 3 ml of a 4% (w / v in PBS) paraformaldehyde solution. Paraformaldehyde was best suited for fixing the bacteria as it produced no background fluorescence, unlike glutaraldehyde which fluoresces in the same region of the spectrum as the rhodamine label found on the probe. Cloudy specimens or those containing particulate material were diluted 1:10 in PBS to facilitate filtration. After a 90 min fixation, the suspensions were transferred to a Nuclepore filtration unit containing a 0.2 m m black membrane (Corning Separations Division, Acton, MA), a 10 m m filter paper and a metal grid (Fig. 1). The fixed specimen was filtered in vacuo, then rinsed with 2 ml of RNAse-free water to remove the paraformaldehyde. 156 M.W. Mittelman et al. / Journal of Microbiological Methods 30 (1997) 153 – 160 Fig. 1. Nuclepore disposable filtration unit. The filter funnel contains up to 15 ml. 2.3.2. Staining with DAPI One ml of a 5 m g / ml 49,6-diamidino-2-phenylindole (DAPI, Polysciences, Warrington, PA) solution was placed in the filtration unit, covered with aluminium foil, and allowed to stand, for 5 min at room temperature. The DAPI solution was filtered in vacuo and the membrane was rinsed with 2 ml of RNAse-free water. The filtration unit was disassembled and the membrane removed for hybridization. Unstained samples were also prepared to determine if the fixed cells autofluoresced. 2.3.3. Transfer of the bacteria to a gelatin-coated slide Three m l of RNAse-free water were added to the well of a gelatin-coated slide. The membrane was deposited on top of the slide, bacteria-side down. The slide and membrane were placed in a vacuum desiccator for |15 min until the membrane had dried. Once dry, the membrane was carefully peeled from the surface and discarded. In conjunction with the slides from the hybridization procedure, parallel samples were stained with DAPI alone to determine the efficiency of transfer from the membrane to the gelatin-coated slides. 2.3.4. Hybridization of the fluorescent probe Eight m l of the hybridization buffer and 2 m l of the fluorescent probe (25 ng /m l) were added to the well of the gelatin-coated slides containing the transferred bacteria. The slide was placed in a 50-ml disposable centrifuge tube which was equilibrated with 5 ml of the wash buffer in an oven maintained at 458C. The hybridization reaction was allowed to proceed for 2 h at 458C. After hybridization, the slide was gently rinsed with 5 ml of the wash buffer, immersed in wash buffer, and returned to the hybridization oven for 10 min at 55.58C. (Following a series of validation experiments conducted at temperatures between 37 and 658C, the optimal temperature for post-hybridization rinsing was found to be 55.560.58C.) The rinsing step was repeated a second time. After the final rinsing, the slide was washed with 5 ml of RNAse-free water and allowed to air dry at 258C. 2.3.5. Microscopic observation A drop of low-fluorescing immersion oil was placed in the well of the slide and a coverslip was added. The slides were stored in the dark until observed under the microscope. Samples were observed under epifluorescence microscopy (LeitzLeico DMRB, Frankfurt). This microscope contained a 100 W mercury lamp, Leico PL Fluorstar 1003 / 1.3 oil immersion objective, and a fluor cluster with filters specific for DAPI and the rhodamine probe. 2.3.6. Bacterial enumeration Culturable cells were enumerated on nutrient agar by a calibrated inoculating loop technique [3]. The standard hospital laboratory technique involved direct plating of urine specimens to blood and MacConkey agars via a calibrated platinum loop (loop count). All plates were incubated 24 h at 358C. 3. Results The optimal specimen processing conditions for maximal sensitivity and specificity are summarized in Table 1. Cell fixation prior to the hybridization procedure was essential, as non-fixed cells did not hybridize with the probe. The use of gelatin-coated slides was found to be necessary for successful hybridization of the cultures. When the technique was first applied, the hybridization and washing steps were performed inside the filtration unit and the bacteria were not fixed using paraformaldehyde; no labeled or DAPIstained bacteria were found on the membrane. In the absence of gelatin, no bacteria were observed. The transfer efficiency of bacteria from membrane to M.W. Mittelman et al. / Journal of Microbiological Methods 30 (1997) 153 – 160 157 Table 1 Optimal FISH assay processing conditions for urine specimens Assay step Reagent Processing time Processing temperature, 8C Specimen fixation DAPI staining Membrane transfer 16S rRNA hybridization Washing Formaldehyde, 3% final concentration DAPI, 5 m g / ml Gelatin-coated slide 16S rRNA probe, 8 m l of 25 ng /m l solution Wash buffer solution 90 5 15 2 10 25 25 25 45 55.560.5 gelatin-coated slide was 5565% (n55). It may be possible to further optimize transfer efficiency through direct application of gelatin to the membrane immediately following specimen filtration. Hybridization stringency was significantly influenced by the washing temperature. At lower temperatures (#508C), all bacteria tested were labeled, including S. aureus and P. aeruginosa. If higher temperatures ($608C) were employed, no labeling (greater than background) was observed on S. aureus and P. aeruginosa, and only a faint labeling was observed for E. coli. At higher temperatures (658C), no labeling was observed for E. coli. The optimal temperature was found to be 55–568C; at this temperature, E. coli was strongly labeled and the other bacteria were not labeled above the level of background fluorescence. Similar results were observed when bacteria were added to filter-sterilized urine and the procedure was performed using patient urine samples spiked with |10 5 E. coli / ml. With patient urines, a slightly higher level of background fluorescence was observed. The assay required |4 h min min min h min from sample collection to imaging under the epifluorescent microscope and recording of data. Assay specificity findings are shown in Table 2. The probe hybridized with E. coli in both PBS, filter-sterilized urine, and in unfiltered patient urine specimens. Although P. aeruginosa and S. aureus demonstrated weak staining with the probe, the cells were clearly non-reactive relative to E. coli. All three test organisms were visualized with DAPI staining. Validation assay results for the three test organisms were similar over the range of bacterial titers evaluated, from 10 2 –10 7 cfu / ml. Autofluorescence of unstained control specimens (patient urine containing either Enterobacteriaceae or non-Enterobacteriaceae) was not observed. The clinical findings for both the hospital-based (culture) method and the membrane-based FISH assay are summarized in Table 3. A total of 83 specimens were tested. Enterobacteriaceae identified by the hospital were E. coli, Enterobacter aerogenes, Citrobacter koseri, Serratia marcescens, Proteus mirabilis, and Hafnia alvei. All 10 of those speci- Table 2 Results of assay validation Organism Experimental conditions 16S rRNA probe reaction a DAPI reaction b E. coli E. coli E. coli P. aeruginosa P. aeruginosa P. aeruginosa S. aureus E. coli (negative assay control) Spiked PBS Spiked filter-sterilized urine Spiked patient specimens Spiked PBS Spiked filter-sterilized urine Spiked patient specimen Spiked PBS Spiked filter-sterilized urine (unstained preparation) Positive Positive Positive Negative Negative Negative Negative Negative (no probe added) Positive Positive Positive Positive Positive Positive Positive Negative (DAPI staining not performed) a Observation of orange-red fluorescing cells under epifluorescence microscopy. Observation of blue fluorescing cells under epifluorescence microscopy. b M.W. Mittelman et al. / Journal of Microbiological Methods 30 (1997) 153 – 160 158 Table 3 Summary of FISH assay findings for clinical specimens Technique Sample size Number positive for Enterobacteriaceae (%) Number positive for non-Enterobacteriaceae putative pathogens Hospital-based, selective culturing 83 10 12.0 9 10.8 Membrane-based probe assay 83 34 41.0 0 ,1.2 mens containing Enterobacteriaceae identified by the hospital laboratory also hybridized with the probe; an additional 24 specimens hybridized with the probe, but were culture-negative (Table 3). Coagulase-negative staphylococci or Enterococcus spp., which are putative urinary tract pathogens, were reported by the hospital laboratory in 9 / 83 specimens. None of these specimens hybridized with the probe. Three of the 24 culture negative patients (%) Comments Non-Enterobacteriaceae putative pathogens were either Enterococcus spp. or coagulase-negative staphylococci 100% specificity for culture-positive Enterobacteriaceae (all with urine specimens that hybridized with the probe) had been previously reported as positive for E. coli by the hospital laboratory. Cell counts obtained by the membrane-based probe assay were in the same range as or greater than those obtained by the hospital-based culture method (Fig. 2). Total counts (DAPI) were significantly higher than the culture values for specimens that hybridized with the probe (Fig. 2) and those speci- Fig. 2. Bacteria counts associated with patient urine specimens containing one or more Enterobacteriaceae spp. The hospital culture counts are in cfu / ml; the FISH and total counts are direct microscopic counts calculated as cells / ml. The reported hospital culture values are presented as the midranges of reported values; e.g. a reported range of .10 000,50 000 cfu / ml is presented as 30 000 cfu / ml. CNS, coagulase-negative staphylococci. M.W. Mittelman et al. / Journal of Microbiological Methods 30 (1997) 153 – 160 mens with reported non-Enterobacteriaceae culture counts. White blood cells, when present in the urine, were readily stained by the DAPI; however, hybridization with the oligonucleotide probe was not observed. Stained bacterial cells were easily differentiated from white blood cells. 4. Discussion Rapid detection of urinary tract pathogens is important for the rationale and effective use of antibiotics in treating UTIs. Since the majority of urine specimens taken in general practitioners’ offices are reported as ‘culture negative’ [1,21], it is clear that significant numbers of patients may be improperly treated with antibiotics. In the absence of more rapid and quantitative assays, the risk of failing to treat symptomatic patients has outweighed the risk of unnecessary antibiotic dosing. The purpose of this study was to develop a quantitative, direct, microscopic assay for a major group of urinary tract pathogens, the Enterobacteriaceae. We have demonstrated that this new FISH assay is both specific and sensitive for members of this pathogen family; significantly, results are available in less than 4 h from sample acquisition. A significant proportion of hospital and community urine specimens submitted for bacterial culture are negative following 1–2 days incubation [21]. In our study of 83 patient urine specimens, routine culture assays were positive for putative UTI pathogens in only 19 cases (22.9%). The results of the FISH assays, while recovering significantly greater numbers of Enterobacteriaceae, also demonstrated that a significant number of clinical specimens were negative. The high specificity of this FISH assay, based on DAPI enumeration and FISH identification, suggests that this assay is reliable for prescreening the large number of specimens that would otherwise be subject to culturing. This could reduce the frequency of inappropriate antibiotic usage for uninfected patients or suggest another etiology for the clinical symptomatology. The assay should be amenable to automation via application of automated aliquoting / filtration combined with image analysis. Preliminary observations (data not shown) also suggest that white 159 blood cells can be differentiated from bacteria and enumerated by an image analysis system. Probe specificity is the single-most important criterion in assessing the utility of hybridizationbased assays. In this study, we used three clinical isolates, E. coli, P. aeruginosa, and S. aureus, to optimize the experimental conditions for specimen processing, hybridization, and washing stringency associated with this new, membrane-based assay. Although the majority of urinary tract infections are caused by members of the Enterobacteriaceae family, Enterococcus spp. (e.g. E. faecalis) and coagulase-negative staphylococci (e.g. S. saprophyticus) are responsible for a significant number of infections [3]. Therefore, consideration should be given to the development of other probes, perhaps tagged with different fluorescent markers, for use in improved membrane-based FISH assays for bacteriuria. Additional screening studies with much larger numbers of clinical UTI isolates are clearly needed to develop this new method for routine clinical diagnostic use. The use of the nucleic acid-specific dye, DAPI, enabled detection of total bacterial numbers in patient urine specimens. Total bacterial counts may be helpful in correlating the relationship between bacterial counts, the types of microorganisms, and the clinical manifestations, particularly in compromised patient populations. The significantly higher total counts seen relative to viable counts are not surprising, given the inadequacy of routine culturing techniques for recovering fastidious and / or damaged bacteria from a variety of environments [22]. In the absence of significant viable bacteria, the finding of large numbers of bacteria in urine using a directcounting technique may be of significance to the clinician. For example, this situation could indicate the presence of atypical, fastidious organisms requiring specialized culturing conditions. Fluorescence-based microbial direct-count techniques offer a number of advantages over cultural techniques in the enumeration of bacteria from clinical specimens. These include: (1) rapid availability of quantitative results; (2) broad range of detection; (3) the ability to detect non-culturable cells (e.g. antibiotic damaged cells); (4) relatively low costs per assay. This membrane-based FISH assay for bacteriuria shows promise as a rapid 160 M.W. Mittelman et al. / Journal of Microbiological Methods 30 (1997) 153 – 160 technique (,4 h) for use in clinical diagnosis of urinary tract infections. The assay is specific, and has a broad dynamic range which will allow it to assess the relationship between bacterial counts, the types of microorganisms present, and the clinical manifestations associated with UTIs. Acknowledgements This research was supported by grants from Corning Inc., Separations Division and the Natural Sciences and Engineering Research Council of Canada. The technical assistance of Mr. S. Sinnadurai and Dr. F. Soboh is acknowledged. References [1] J.C. Nickel, A practical approach to urinary tract infections, Can. J. Diag. 10 (1993) 64–80. [2] W.E. Stamm, Protocol for the diagnosis of urinary tract infection: reconsidering the criterion for significant bacteriuria, Urology 32 (1988) 6–10. [3] B.A. Forbes, P.A. Granato, Processing specimens for bacteria, in: P. Murray, E.S. Baron, M.A. Pfaller, F.C. Tenover, R.H. Yolken (Eds.), Manual of Clinical Microbiology, 6th ed. ASM, Washington, DC, 1995, pp. 265–279. [4] R.R. Colwell, Nonculturable but still viable and potentially pathogenic, Zentralbl. Bacteriol. 279 (1993) 154–156. [5] M. Pommepuy, M. Butin, A. Derrien, M. Gourmelon, R.R. Colwell, M. Cormier, Retention of enteropathogenicity by viable but nonculturable Escherichia coli exposed to seawater and sunlight, Appl. Environ. Microbiol. 62 (1996) 4621–4626. [6] M. Pezzlo, Detection of urinary tract infections by rapid methods, Clin. Microbiol. Rev. 1 (1988) 268–280. [7] M.T. Pezzlo, D. Amsterdam, J.P. Anhalt, T. Lawrence, N.J. Stratton, E.A. Vetter, E.M. Peterson, L.M.D.L. Maza, Detection of bacteriuria and pyuria by uriscreen, a rapid enzymatic screening test, J. Clin. Microbiol. 30 (1992) 680– 684. [8] M.T. Dalton, S. Comeau, B. Rainnie, K. Lambert, K.R. Forward, A comparison of the API Uriscreen with the Vitek urine identification-3 and the leukocyte esterase or nitrite strip as a screening test for bacteriuria, Diag. Microbiol. Infect. Dis. 16 (1993) 93–97. [9] C. Koenig, L.J. Tick, B.A. Hanna, Analyses of the Flashtrack DNA probe and UTIscreen bioluminescence tests for bacteria, J. Clin. Microbiol. 30 (1992) 342–345. [10] D. Colombrita, G. Ravizzola, F. Pirali, M. Manni, N. Manca, E. Savoldi, A. Turano, Evaluation of a Bactec system for urine culture screening, J. Clin. Microbiol. 27 (1989) 118– 119. [11] F.P. Yu, G.A. McFeters, Rapid in situ assessment of physiological activities in bacterial biofilms using fluorescent probes, J. Microbiol. Methods 20 (1994) 1–10. [12] P.S. Tabor, R.A. Neihof, Improved method for determination of respiring individual microorganisms in natural waters, Appl. Environ. Microbiol. 43 (1982) 1249–1255. [13] P.E. Jorgensen, T. Ericksen, B.K. Jensen, Estimation of viable biomass in wastewater and activated sludge by determination of ATP, oxygen utilization rate, and FDA hydrolysis, Water Res. 11 (1992) 1495–1501. [14] R.I. Amann, L. Krumholz, D.A. Stahl, Fluorescent oligonucleotide probing of whole cells for determinative, phylogenic, and environmental studies in microbiology, J. Bacteriol. 172 (1990) 762–770. [15] R.E. Hicks, R.I. Amann, D.A. Stahl, Dual staining of natural bacterial plankton with 49,6-diamidino-2-phenylindole and fluorescent oligonucleotide probes targeting kingdom-level 16S rRNA sequences, Appl. Environ. Microbiol. 58 (1992) 2158–2163. [16] H. Gersdorf, K. Pelz, U.B. Gobel, Fluorescence in situ hybridization for direct visualization of gram-negative anaerobes in subgingival plaque samples, FEMS Immunol. Med. Microbiol. 6 (1993) 109–114. [17] H.J. Monstein, E. Kihlstrom, A. Tiveljung, Detection and identification of bacteria using in-house broad range 16S rDNA PCR amplification and genus-specific DNA hybridization probes, located within variable regions of 16S rRNA genes, APMIS 104 (1996) 451–458. [18] D.J. Lane, B. Pace, G.J. Olsen, D.A. Stahl, M.L. Sogin, N.R. Pace, Rapid determination of 16S ribosomal RNA sequences for phylogenetic analysis, Proc. Natl. Acad. Sci. USA 82 (1985) 6955–6959. [19] K. Greisen, M. Loeffelholz, A. Purohit, D. Leong, PCR primers and probes for the 16s rRNA gene of most species of pathogenic bacteria, including bacteria found in cerebrospinal fluid, J. Clin. Microbiol. 32 (1994) 335–351. [20] S.F. Altschul, W. Gish, W. Miller, E.W. Myers, D.J. Lipman, Basic local alignment search tool, J. Mol. Biol. 215 (1990) 403–410. [21] P. Murray, P. Traynor, D. Hopson, Evaluation of microbiological processing of urine specimens: comparison of overnight versus two-day incubation, J. Clin. Microbiol. 30 (1992) 1600–1601. [22] G.A. McFeters, F.P. Yu, B.H. Pyle, P. Stewart, Physiological assessment of bacteria using fluorochromes, J. Microbiol. Methods 21 (1995) 1–13.