• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of aemPermissionsJournals.ASM.orgJournalAEM ArticleJournal InfoAuthorsReviewers
Appl Environ Microbiol. Aug 2009; 75(16): 5363–5372.
Published online Jul 6, 2009. doi:  10.1128/AEM.00658-09
PMCID: PMC2725463

Potentially Pathogenic Bacteria in Shower Water and Air of a Stem Cell Transplant Unit[down-pointing small open triangle]

Abstract

Potential pathogens from shower water and aerosolized shower mist (i.e., shower aerosol) have been suggested as an environmental source of infection for immunocompromised patients. To quantify the microbial load in shower water and aerosol samples, we used culture, microscopic, and quantitative PCR methods to investigate four shower stalls in a stem cell transplant unit at Barnes-Jewish Hospital in St. Louis, MO. We also tested membrane-integrated showerheads as a possible mitigation strategy. In addition to quantification, a 16S rRNA gene sequencing survey was used to characterize the abundant bacterial populations within shower water and aerosols. The average total bacterial counts were 2.2 × 107 cells/liter in shower water and 3.4 × 104 cells/m3 in shower aerosol, and these counts were reduced to 6.3 × 104 cells/liter (99.6% efficiency) and 8.9 × 103 cells/m3 (82.4% efficiency), respectively, after membrane-integrated showerheads were installed. Potentially pathogenic organisms were found in both water and aerosol samples from the conventional showers. Most notable was the presence of Mycobacterium mucogenicum (99.5% identity) in the water and Pseudomonas aeruginosa (99.3% identity) in the aerosol samples. Membrane-integrated showerheads may protect immunocompromised patients from waterborne infections in a stem cell transplant unit because of efficient capture of vast numbers of potentially pathogenic bacteria from hospital water. However, an in-depth epidemiological study is necessary to investigate whether membrane-integrated showerheads reduce hospital-acquired infections. The microbial load in shower aerosols with conventional showerheads was elevated compared to the load in HEPA-filtered background air in the stem cell unit, but it was considerably lower than typical indoor air. Thus, in shower environments without HEPA filtration, the increase in microbial load due to shower water aerosolization would not have been distinguishable from anticipated variations in background levels.

Hospital water supplies are frequently inhabited with environmental waterborne microbes, including bacteria (Legionella pneumophila, Pseudomonas aeruginosa, Mycobacterium avium, Stenotrophomonas maltophilia, and Achromobacter spp.) and fungi (Aspergillus spp. and Fusarium spp.) (4, 13, 63). Although water storage tanks may be cleaned annually and residual chlorine levels of water contents maintained, these measures alone cannot prevent the formation of biofilms along inert surfaces of the tank and piping systems. Biofilms attached to living and inert surfaces consist of complex communities of microbes that produce glycocalyx polysaccharides, which protect bacteria from desiccation, chemical treatments, and immunologic attack (25). They can form quickly and have been found in dental water lines only a few weeks after installation (14). Finally, biofilms can harbor pathogens that are periodically released through sloughing of fringe layers (25, 70). Common point-of-use sources of potential exposure to waterborne microbes contained in biofilms in the health care setting include showerheads, water tanks, faucets, aerators, water fountains, and ice machines (2, 4, 13, 31, 34, 38, 68).

While microbes found in water usually pose no risk for healthy individuals, they can be opportunistic pathogens capable of causing serious and life-threatening infections in severely immunocompromised individuals. Patients with cancers of the blood, lymph, and bone marrow (leukemia, lymphoma, and myeloma) are frequently treated with intense chemotherapy, irradiation, and/or stem cell transplant, resulting in neutropenia and profound immunosuppression. Stem cell transplant patients are encouraged to bathe or shower daily before and after transplant to help maintain skin integrity; these patients also almost universally have a central venous access device for the administration of chemotherapy and other medications. Opportunistic microbes in shower water may contaminate the central venous catheter and provide a mechanism for bacterial invasion into the bloodstream (48). Once a patient has become infected, treatment of these organisms may be more difficult because of antibiotic resistance. Thus, severely immunosuppressed patients are at risk of significant morbidity and mortality from exposure to health care-acquired environmental pathogens.

Although there is no data on bacteria in hospital showers, the shower environment, particularly the aerosolized shower mist (i.e., shower aerosol), has long been suspected as a source of opportunistic pathogens (4, 10). Inhalation of aerosolized pathogenic bacteria in the shower may result in respiratory infections and dissemination of organisms from the lungs into the bloodstream. The advent of high-throughput sequencing now allows for qualifying the bacterial composition in aerosols (23, 28). Angenent et al. (5) investigated the aerosols generated from an indoor therapy pool environment in which multiple staff members contracted Mycobacterium avium infections and hypersensitivity pneumonitis (i.e., swelling of alveoli due to an immunologic reaction to airborne particles). A fraction (>30%) of the bacteria in pool water was identified as M. avium, which preferentially partitioned into the aerosol (>80%) (5). In a previous study conducted by Bollin et al. (10), Legionella pneumophila isolates were collected from shower and sink aerosols; however, to our knowledge, there are no published studies that document the effectiveness of membrane-integrated showerheads in decreasing the microbial load in indoor air.

We investigated whether shower stalls in a stem cell transplant unit in a hospital could be a source of potential pathogens by quantifying and qualifying the bacterial load, colony count, and bacterial DNA in shower water and air. An engineering control that consisted of a showerhead with an integrated 0.2-μm-pore-size membrane was utilized to ascertain whether conventional hospital showers aerosolize bacteria and therefore whether a significant increase in the bacterial load was observed compared to HEPA-filtered background air. A total of four shower stalls in individual rooms in a stem cell transplant unit were evaluated during two seasonal sampling periods (two stalls per season). Each shower stall was sampled for 3 days with a conventional showerhead in place and then for 3 days with a membrane-integrated showerhead installed in the same shower. Our goal was to determine the overall mitigation effectiveness of utilizing membrane-integrated showerheads in reducing the presence of potentially pathogenic bacteria from shower water and aerosols. Patient infections were not evaluated during the course of this study.

MATERIALS AND METHODS

Shower environment.

Barnes-Jewish Hospital is a 1,228-bed tertiary care teaching hospital affiliated with the Washington University School of Medicine in St. Louis, MO. The stem cell transplant unit has 26 private rooms with positively pressured, HEPA-filtered air. Each room (54 m3) has a private, attached bathroom (9.0 m3) with a sink, commode, and curtain-closing shower stall (2.2 m3). The hall door, internal restroom door, and shower curtains were kept closed during sampling. The rooms were occupied by patients, but the patients did not use the shower during the sampling process. Housekeeping personnel cleaned each restroom and shower daily before sampling. The stem cell transplant unit is supplied with water from six independent hot water risers with each riser feeding between four and six rooms; sample showers were selected from the different risers. The St. Louis City Water Division supplies the hospital water tanks with domestic water. Water storage tank temperatures are maintained between 49 and 52°C. Each tank is cleaned annually by hospital maintenance; however, chlorine levels are not managed by hospital maintenance personnel (based on historic records, the levels are ~0.1 mg/liter free chlorine and ~0.3 mg/liter total chlorine).

Experimental design.

On a daily basis, three background air samples, three shower aerosol samples, and one shower water sample were collected from each shower stall, during two seasons (two stalls in the winter and two stalls in the summer of 2007). Data were collected for 6 days from each shower stall: 3 days with a conventional showerhead in place, followed by 3 days with a Pall-Aquasafe water filter (AQF7S) (Ann Arbor, MI) showerhead installed, which has an integrated 0.2-μm-pore-size membrane (i.e., membrane-integrated showerhead). For each day of sampling, the three similar aerosol samples were pooled for daily quantitative culturing and microscopy methods, but all 3 days of collected samples were pooled prior to DNA extraction for the molecular biology techniques due to the low levels of DNA present in the air of the HEPA-filtered rooms. Likewise, bacteria in the water samples were quantified through culturing and microscopy on a daily basis, and samples were pooled prior to DNA extraction. Thus, quantitative PCR and 16S rRNA gene sequencing were combined samples from 3 days of sampling. For each method of quantification, a statistical analysis was performed using the R Project for statistical computing software (www.r-project.org). Three-way analysis of variance tests were used to determine the significance of season, filtration, and sample type for each method of quantification. A correlation analysis was also conducted to determine the correlation between the data from the total bacterial counts and quantitative PCR. For this analysis, it was assumed that each data set had a bivariate normal distribution.

Sampling conditions.

Aerosol samples were collected with swirling aerosol collectors (SACs) (Biosampler, AGI-30; SKC, Eighty Four, PA) for 90 min, allowing a total sample volume of 1.125 m3 to pass through each of three samplers that were run simultaneously. The SACs were filled with 20 ml of sterile phosphate-buffered saline (PBS) (Cellgro, Manassas, VA) and maintained on a sampling stand about 1.5 m from the floor. In addition to the SACs, three flow meters (RMA-22; Dwyer, Michigan City, IN) were used to control the flow rate pulled through the SACs. Two high-volume vacuum pumps (2688VE44 and 2669CE44; Thomas, Lake Zurich, IL) were connected to the flow meters and SACs with neoprene tubing (Masterflex, Vernon Hills, IL). This tubing ran under the hospital room and bathroom doors from the hallway to the shower stall. A daily site blank sample was collected by preparing one SAC on location with 20 ml of PBS and immediately pouring it out into the sampling vial. Daily background aerosol samples were collected from the shower stall prior to turning on the shower. Shower aerosol samples were taken at the same location while the shower was running at a water temperature of 33 to 43°C, which is the human comfort zone in terms of shower water temperature. One liter of shower water was collected for analysis while the shower was running during the shower aerosol sampling period.

DNA extraction and PCR amplification.

One 100-ml shower water sample and a pooled 40-ml SAC fluid sample (from three samplers) were concentrated each day with 0.2-μm-pore-size, gamma-sterilized microfunnel filters (4803 Pall; Pall, Ann Arbor, MI). Samples were then immediately stored at −80°C until elution from the three daily filters (in one tube) with 2 ml of elution buffer (5). DNA was extracted from the elution wash using a bead-beating and phenol-chloroform extraction protocol (40). The 16S rRNA genes were amplified by a 30-cycle touchdown PCR. The 30-cycle touchdown PCR consisted of an initial 2-min denaturing step at 94°C, followed by 20 cycles, with 1 cycle consisting of 30 s at 92°C, 90 s at a temperature from 65°C to 45°C (temperature decreased 1°C per cycle), and 90 s at 72°C. These touchdown cycles were followed by 10 cycles, with 1 cycle consisting of 30 s at 92°C, 30 s at 45°C, and 30 s at 72°C, and a final extension step of 15 min at 72°C. The 50-μl solutions contained 1.25 units of GoTaq (Promega Corp., Madison, WI), 0.4 pmol/μl (each) forward and reverse primers (8F [5′-AGAGTTTGATCCTGGCTCAG-3′] and 1391R [5′-GACGGGCGGTGWGTRCA-3′]), 0.5 mM MgCl2, 0.2 mM (each) deoxynucleoside triphosphates, 0.8 mg/ml bovine serum albumin, and 2 μl of template. Positive- and negative-control reactions were included with each reaction set. Shower aerosol and negative-control PCR products were ethanol precipitated and reamplified with an additional 20 cycles of PCR, with 1 cycle consisting of 30 s at 92°C, 30 s at 45°C, and 30 s at 72°C, with the same reaction solution. Negative controls were negative for the repeated data sets.

Bacterial quantification: colony counts, total bacterial counts, and qPCR.

Microfil S filtration devices (0.22-μm pore size; Millipore, Billerica, MA) were used to filter 100 ml of each water sample. The filters were placed on heterotrophic tryptic soy agar plates and incubated at 35°C before quantifying the CFU. Direct cell counts were conducted by an epifluorescence microscopy procedure (5). In short, 45 ml of water or 20 ml of SAC fluid was filtered through 0.22-μm black polycarbonate filter (GE Water, Trevose, PA), stained with 1 μM 4′,6′-diamidino-2-phenylindole (DAPI) (Sigma, St. Louis, MO), and rinsed. Cells were then counted with an epifluorescence microscope (Olympus BX41, Olympus, Center Valley, PA). Quantitative PCR (qPCR) was used to determine the bacterial DNA load within each sample. Wells contained 2 μl of extracted template and 23 μl of SYBR green mix (ABgene, Rockford, IL) supplemented with 0.25 U UDP-N-glycosidase, and 10 μM (each) universal bacterial primers (forward primer [5′-TCCTACGGGAGGCAGCAGT-3′] and reverse primer [5′-GGACTACCAGGGTATCTAATCCTGTT-3′]) (47). Samples were analyzed with a Stratagene Mx3000P qPCR system (Cedar Creek, TX), using the program outlined by Nadkarni et al. (47) (i.e., 40 cycles, with 1 cycle consisting of 15 s at 95°C, 1 min at 60°C, and 30 s at 72°C with data collection temperatures of 85°C to 88°C and analysis of a final melting curve). To generate a standard curve for qPCR, DNA was extracted from Escherichia coli and quantified using a PicoGreen double-stranded DNA quantitation kit (Molecular Probes, Eugene, OR). The standard curve of E. coli DNA ranged from 4.7 pg to 4.7 ng with an R2 of 0.987.

Cloning and sequence analysis.

16S rRNA gene surveys were conducted for the pooled shower water samples that combined the first 3 days of sampling prior to installation of the membrane-integrated showerhead. Gene surveys were also performed for two pooled aerosol samples taken during the summer without the membrane-integrated showerhead in place. For these samples, PCR amplicons were gel purified (Montage DNA gel extraction kit, catalog no. LSKGEL050; Millipore, Billerica, MA) and sent to the Genome Sequencing Center, Washington University School of Medicine, for cloning, purification, and Sanger sequencing on an ABI 3730 capillary sequencer (Applied Biosystems, Foster City, CA). 16S rRNA gene sequences were edited and assembled into consensus sequences using PHRED and PHRAP aided by XplorSeq (33). Bases with a PHRAP quality score of <20 were trimmed from the data set before they were aligned using the NAST online tool (19). Chimeras were detected using Bellerophon (37) and removed. Nonchimeric sequences were compared to the Greengenes public database and aligned with the computer application software ARB (43) to determine the rRNA secondary structure information with phylogenetic identification. A sequence identity of 95% or greater was considered to mean organisms of a similar phylum, whereas an identity of 97% or greater was considered to mean the same species. The generated distance matrices were used in the DOTUR program to assess the number of operational taxonomic units (OTUs) and cluster them by pair-wise identity (percent identity [ID]) with a furthest-neighbor algorithm and a precision of 0.01 (58). Assignment of the majority of sequences to phyla was based on their position after parsimony insertion to the ARB dendrogram in the Greengenes database (20). Phylum classifications were double checked, and the percent ID was calculated for nonchimeric sequences with the Ribosomal Database Project II (RDP) (16). Phylogenetic trees using a heuristic algorithm were created with PAUP (72). Finally, UniFrac (42) was used to compare the sampled environments.

DNA extracted from representative colonies grown on tryptic soy agar plates was PCR amplified with universal bacterial primers (8F and 1391R) and gel purified (Montage DNA gel extraction kit, catalog no. LSKGEL050; Millipore, Billerica, MA). The single culture 16S rRNA gene amplicons were sent to Retrogen Inc. (San Diego, CA) for DNA sequencing with 8F as the primer. We obtained a single sequence with a length of ~700 nucleotides for the environmental isolates and did, therefore, not build a consensus sequence or use them for our phylogenetic analyses. These sequences were identified through RDP (16).

RESULTS

Bacterial quantification.

The number of culturable bacteria in shower water from a conventional showerhead and a membrane-integrated showerhead ranged from 639 to 27,800 CFU/liter and from 3 to 6 CFU/liter, respectively (Table (Table1).1). The levels in the shower aerosol samples, however, were below the limits of detection of our methods. We did obtain total bacterial counts for the shower aerosol samples because the total bacterial counts (culturable and nonculturable) found with an epifluorescence microscope are much higher than the number of culturable bacteria (for shower water, this was on average ~17,000 times higher [Table [Table1]).1]). We observed between 4,600 and 41,000 cells/m3 in aerosol samples from the shower stall, while there were between 5.1 × 104 and 5.1 × 107 cells/liter in shower water (Table (Table1).1). Our statistical analysis with data from Table Table11 indicated a significant difference between sample type (water and aerosol samples; P = 0.008) and an effect of installing the membrane-integrated showerhead (P = 0.008), but there was not a significant seasonal effect (P = 0.27) or riser effect (P = 0.51). To further compare our quantitative data and to gauge the effect of a conventional showerhead versus a membrane-integrated showerhead, we averaged the data from the two seasons and the four different risers (Fig. (Fig.1).1). This resulted in an averaged 99.0% (±1.3% [standard deviation]) (P = 0.0008) decrease in the number of culturable bacteria in shower water after membrane-integrated showerheads were installed. Similarly, the averaged total bacterial counts in shower water decreased from 2.2 × 107 cells/liter (±1.6 × 107) to 6.3 × 104 cells/liter (±2.1 × 104) before and after installation of the membrane-integrated showerheads, respectively, which constituted a decrease of almost 3 orders of magnitude (99.6% [±0.36%]; P = 0.02) of bacterial cells in shower water (Fig. (Fig.1A).1A). Using the PBS and the site blank as estimated backgrounds for the shower water, we found that the bacterial concentration in water collected with the membrane-integrated showerhead was lower than these background solutions (Fig. (Fig.1A).1A). For the shower aerosol, the averaged total bacterial counts decreased from 3.4 × 104 cells/m3 (±1.2 × 104) to 8.9 × 103 cells/m3 (±3.9 × 103) before and after installation of the membrane-integrated showerheads, respectively (Fig. (Fig.1B),1B), which is an 82.4% (±2.3%; P = 0.07) decrease. Furthermore, the averaged total bacterial counts for aerosols with the membrane-integrated showerheads were at the same level as the counts for the background aerosols (8.9 × 103 [±3.9 × 103] and 1.1 × 104 [±4.1 × 103], respectively; Fig. Fig.1B).1B). These reductions in total bacterial cells were comparable with qPCR for which the number of targeted amplicons decreased from an average of 7.6 × 102 pg DNA/liter to 4.5 pg DNA/liter in shower water and an average of 4.0 pg DNA/m3 to 1.75 pg DNA/m3 in aerosol samples before and after installation of the membrane-integrated showerhead, respectively (nonaveraged data are shown in Table Table1).1). This corresponds to a 99.0% (±1.4%; P = 0.01) decrease in the shower water and a 70.5% (±16%; P = 0.02) decrease in the shower aerosol. Indeed, our statistical analysis showed a strong correlation between the data from epifluorescence microscopy and qPCR methods (R2 = 0.979; P < 0.0001).

FIG. 1.
Averaged total bacterial counts for shower water (A) and shower air (B) with conventional showerheads and with membrane-integrated showerheads. The total bacterial counts (number of cells per liter or per m3) were determined by direct epifluorescence ...
TABLE 1.
Summary of bacterial quantification results for bacteria in shower water and air samples from a stem cell transplant unita

Sequence analyses.

We characterized four shower water samples and two aerosol samples in shower stalls in the stem cell transplant unit by sequencing 954 nearly full-length 16S rRNA genes, which were acquired directly from the environment without a culturing step. These water and aerosol samples were pooled samples from three sampling days with a conventional showerhead. The DNA concentration for the winter aerosol samples and for all samples with the membrane-integrated showerhead was below the level required for a successful cloning step in the sequencing pipeline at the genome-sequencing center (even after two PCR amplification steps), and thus, we were not able to acquire gene surveys for these samples. All shower water samples were diverse in terms of their bacterial phylum composition (Fig. 2A, B, E, and F). Some similarities between the water samples were found—the phyla Actinobacteria, Bacteroidetes, Cyanobacteria, Clostridia, and Proteobacteria were identified in all samples even though the percentage of each phylum was different for samples (Fig. (Fig.2).2). Several bacterial phyla were found only in a single sample: the phylum Erysipelotrichales (10%) in a summer water sample from riser A and TM7 (2%) in a winter water sample from riser D. A small number of nonclassifiable bacteria (2% to 9%) were found in each of the water samples (Fig. (Fig.2).2). Only bacteria from the phylum Proteobacteria were identified in the aerosol samples with Alpha-, Beta-, and Gammaproteobacteria in the summer aerosol sample from riser A, and Alpha- and Betaproteobacteria in the summer aerosol sample from riser B (Fig. 2C and D).

FIG. 2.
Phylogenetic distributions of bacterial 16S rRNA gene sequences. The percentages of phylum or class distribution for samples from conventional showerheads are shown for the following samples: summer water samples from hot water riser A (247 sequences) ...

Further phylogenetic analyses were performed to classify individual 16S rRNA gene sequences. We found 444 different operational taxonomic units for the 954 nonchimeric sequences. The RDP algorithm was used to identify the sequences from shower water and aerosol samples from conventional showerheads with a specific interest to identify potentially pathogenic bacteria (Table (Table2).2). The most notable potential pathogens identified were Mycobacterium mucogenicum in shower water samples and Pseudomonas aeruginosa in the aerosol samples. To further verify the relatedness of these important potential pathogens to sequences in public databases, we performed maximum-likelihood, maximum-parsimony, and neighbor-joining analyses on the sequences that were identified within two groups of microbes: (i) the suborder Corynebacterineae of the phylum Actinobacteria to include closely related mycobacteria (Fig. (Fig.3B);3B); and (ii) the class Gammaproteobacteria to include the Pseudomonas and Legionella genera (Fig. (Fig.3C).3C). These analyses produced similar trees, with similar bootstrap support at resolved branches (data not shown). For the Corynebacterineae, 88 16S rRNA gene sequences from the summer water sample from riser A were most closely aligned to published M. mucogenicum and Mycobacterium phocaicum sequences (Fig. (Fig.3B).3B). We also retrieved 18 sequences from colonies with different morphologies, which were grown from bacteria in shower water with conventional showerheads: from the summer water samples, we identified three sequences from riser A and three sequences from riser B, which were closely related to M. mucogenicum (≥99.5% ID). Therefore, this pathogenic strain was cultured under laboratory conditions and consisted of ~35% (88/247) of the directly retrieved sequences and 90% (9/10) of the cultured sequences from the summer water sample from riser A. Thus, this potential pathogen was present in the summer water samples at very high numbers, and at least some of the cells were viable. The other important potential pathogen identified was P. aeruginosa in the summer aerosol sample from riser A. This species was much less abundant compared to the Mycobacterium species when directly retrieved from the air sample (3/250 sequences) and was not identified in the water sample because of the presence of more abundant bacteria.

FIG. 3.
Phylogenetic relationships of bacterial sequences from shower water and shower aerosol samples for four hot water risers in a stem cell transplant unit during two seasons. (A) Unrooted phylogenetic tree showing all sequences and their origin; (B) maximum-likelihood ...
TABLE 2.
Summary of potentially pathogenic bacterial sequences for bacteria in shower water and air samples from a stem cell transplant unita

The individual sequences, their evolutionary relationship, and sample distribution are visualized in a comprehensive phylogenetic tree to identify commonalities between samples (Fig. (Fig.3A).3A). Related sequences were found in the shower water and shower aerosol samples within the class Betaproteobacteria, the class Gammaproteobacteria, the order Rhizobiales (Alphaproteobacteria), and the genus Sphingopyxis (Alphaproteobacteria) (Fig. (Fig.3A).3A). However, none of the related sequences in either water or aerosol samples were >97% identical (i.e., from a single species) to each other. This is illustrated in Fig. Fig.3C3C for the class Gammaproteobacteria for which the sequences in the water samples were closely related to the genera Acinetobacter, Legionella, Rickettsiela, and Aquicella, while the aerosol samples were closely related to P. aeruginosa. We also found that the species composition in the water samples varied considerably between seasons (Fig. (Fig.3A)3A) and that the summer water samples were more similar to each other than to the winter water samples (Fig. (Fig.4).4). Clustering of the samples based on the OTUs in the community also verified that the shower water and aerosol communities were more different from each other than the bacteria in different water samples regardless of the sampling season (Fig. (Fig.44).

FIG. 4.
Clustered sample sets with weighted and normalized UniFrac analysis. The scale bar represents branch length units.

DISCUSSION

The membrane-integrated showerhead reduced the microbial load in shower water and aerosols to background water and air microbial concentrations.

The observed numbers of culturable bacteria in the shower water samples from the conventional showerhead were lower than the CDC's recommended maximum containment level of 500 CFU/ml (67), while total bacterial counts were within published findings for stored tap water cell counts (107 to 108 cells/liter) (57). Therefore, the bacterial levels in the shower water in this stem cell transplant unit with conventional showerheads were within a typical range for a hospital. The installation of the 0.2-μm-pore-size membrane-integrated showerhead lowered the microbial load considerably in the shower water to background water and air microbial concentrations that were similar to or lower than our lab-filtered buffer solutions (Fig. (Fig.1A1A).

In contrast, the microbial loads in the shower aerosols were much lower than typical indoor and outdoor environments due to the use of pressurized HEPA-filtered air in the rooms in the sampled stem cell transplant unit. The shower aerosol from the conventional showerhead contained an averaged total bacterial count of 3.4 × 104 cells/m3 (~17 microbes per breath—assuming a tidal volume of 500 ml), which is approximately 1 order of magnitude lower than an aerosol sample taken in a single-family home (2.1 × 105 cells/m3; 105 microbes per breath) (27); 2 orders of magnitude lower than in an indoor hospital therapy pool (106 cells/m3; 500 microbes per breath) (5); 1 to 3 orders of magnitude lower than an outdoor aerosol sample collected in Salt Lake City, UT (105 to 107 cells/m3; 50 to 5,000 microbes per breath) (52); and 1 order of magnitude lower than the outdoor air in East St. Louis, IL (1.5 × 105 cells/m3; 75 microbes per breath). The latter total bacterial count is of interest because it allows us to compare our findings with the outdoor air of St. Louis, MO. During a year-long, daily sampling campaign of the outdoor air in East St. Louis, IL, which is seven miles east of the hospital, Rauer (53) measured an arithmetic average total bacterial count of 1.5 × 105 (±1.1 × 104) cells/m3. He also found a seasonal effect on the total bacterial counts with higher counts (1.8 × 105 [±1.8 × 104]) cells/m3 during the summer months. This would have resulted in an inhalation of ~90 microbes per breath on an average summer day. Thus, the microbial load in the indoor environment of our stem cell transplant unit (17 microbes per breath when the conventional shower was running) was considerably lower than that of the outdoor environment seven miles east of Barnes-Jewish hospital, in East St. Louis, IL.

In addition, the culturable bacterial levels in the ambient air of the stem cell transplant unit were also lower than in typical indoor environments, but the levels were similar for other stem cell transplant units. Angenent et al. (5) found the ratio of total bacterial cells to the CFU count from air to be between 500 and 5,000, while Radosevich et al. (52) reported a ratio of 1,250 for air. Our ratios in shower water were even higher than that (Table (Table1).1). Others have discussed a ratio higher than 100 for soil and water (3, 49). Therefore, due to the low sample volume and a relatively high ratio of total bacterial cells to CFU in our study, it is not a surprise that the CFU counts in our aerosol samples were below detection. On the basis of a ratio of 1,000 for air, which is in agreement with previously published studies (5, 52), we estimate the CFU counts of our aerosol samples to be 34 and 10 CFU/m3 before and after installation of the membrane-integrated showerhead, respectively. This estimate is similar compared to the value for rooms in a stem cell transplant unit in Taiwan (≤32 CFU/m3) (41), but it is much lower than in a Polish pneumonia ward (296 to 530 bacterial and fungal CFU/m3) (6). Thus, even with a conventional showerhead, the culturable bacterial levels in the shower aerosols from shower stalls in the stem cell transplant unit was very low.

The installation of the membrane-integrated showerhead decreased the microbial load to background levels (104 cells/m3) of the indoor air of the stem cell transplant unit during a showering event (four to six microbes per breath). The relatively low background levels of bacteria in the indoor air were helpful in this study because we were able to show a relatively small increase of ~2.5 × 104 cells/m3 in microbial load for the shower aerosol when the conventional showerhead was on. Such an increase would not have been statistically significant for a study of a typical indoor air environment with 105 to 106 cells/m3. This may explain why, to our knowledge, no other studies with a significant increase in shower aerosol bacterial levels were found in the literature.

Potentially pathogenic bacteria were present in shower water and aerosol samples from the stem cell transplant unit.

The most notable potential pathogens in shower water were the relatively high numbers of M. mucogenicum sequences in the summer. However, no mycobacteria were detected in the winter water samples. Species within the Mycobacteriaceae family are known for their robustness due to a waxy outer membrane layer, which protects them from disinfectants, such as chlorine in domestic water (24, 29). M. mucogenicum was isolated from the water system in a French hospital, where it was linked to two terminal infections in immunocompromised patients (2). In addition, Mycobacterium spp. in air have caused aerosol-related health problems (5, 28, 71), however, no mycobacterial sequences were identified in our summer aerosol samples. Other sequences in the shower water samples were identified as Mycobacterium spp., including Mycobacterium gordonae (98.5% ID). Identification of Mycobacterium spp. through 16S rRNA gene surveys is difficult because they are very closely related to one another with only 0- to 7-bp differences within the 16S rRNA gene between species, while in most other microbes this difference is between 5 and 15 bp (65). Therefore, Hussein et al. (37a) required a ≥99% homology to database sequences before species identification in their study of nontuberculosis mycobacteria in hospital waters. Even with such close to perfect homology, our maximum-likelihood phylogenetic tree showed that the sequences were indistinguishable between M. mucogenicum and M. phocaicum due to a 100% identical 16S rRNA gene sequence (Fig. (Fig.3B).3B). M. phocaicum is also pathogenic and is associated with chronic pneumonia (1). Here, we identified our sequences as M. mucogenicum because RDP matched our directly extracted and cultured gene sequence closest to this species. However, we realize that for a true identification between M. mucogenicum and M. phocaicum, other genes or enzymes must be targeted.

Three sequences out of a total of 250 sequences (~1%) in summer aerosols from riser A were identified as Pseudomonas aeruginosa (99.3% ID). P. aeruginosa is commonly linked with ventilator-associated pneumonia, as biofilms with P. aeruginosa grow on endotracheal tubes (7, 30). The other pathogen from a summer aerosol sample (from riser B) was identified as Bosea thiooxidans (98.6% ID), which is a member of the Alphaproteobacteria initially isolated from agricultural soil (62). Recently, this species has been identified as a close relative to multiple pathogenic aquatic organisms isolated from hospital water supplies in French and Swiss hospital water systems (39, 64). Finally, a strain of Bacillus cereus (99.9% ID) was cultured from winter water samples (riser C). B. cereus strains are associated with food-borne illnesses and Bacillus anthracis-related virulence genes that can be detrimental to immunocompromised individuals (61). Dohmae et al. (22) identified hospital towels and laundering facilities, including the rinse water, to be sources of B. cereus. Even though we sequenced up to ~250 clones from our samples, the diversity in environmental samples is high enough that we identify only the most abundant microbes. Less prevalent microbes have the ability to infect patients but will be identified only when the number of sequences per sample increases. Here, the less prevalent B. cereus was found in the water sample by culturing even though it was not identified with the 16S rRNA gene survey, because we enriched for this viable organism by the growth conditions used in our lab. The phenomenon of enrichment by culturing has been described in detail previously (56).

Legionella sequences were found in the shower water.

The threat of Legionella species as environmental contaminants has come to the forefront since it was first isolated in July 1976 (21). To date, there are over 70 members of the genus Legionella that inhabit natural aquatic environments as intracellular parasites to protozoa (12). Thus far, ~40% of the identified Legionella species are human pathogens, and L. pneumophila has been isolated in over 90% of culture-confirmed cases of legionellosis. Another 9% of legionellosis cases have been caused by Legionella spp., such as Legionella longbeachae, L. bozemanii, L. feeleii, L. dumoffii, L. wadsworthii, and L. anisa, and by Tatlockia micdadei (46). A total of six water sequences (two from summer water samples from riser A, three from winter water samples from riser C, and one from a winter water sample from riser D) were identified as Legionella (Fig. (Fig.3C).3C). None of these were positively identified (>97% ID) to a known Legionella species in RDP, and they did not align closely to any Legionella species in the Greengenes database (Fig. (Fig.3C).3C). This genus is ubiquitous in natural and man-made aquatic environments (9, 11, 12, 15, 17, 21, 35, 46), and finding sequences of this genus does not automatically indicate pathogenesis. To conclude whether the identified Legionella species in the shower water poses a threat to hospital patients, isolation (by culturing) and characterization of the strain are necessary.

Cyanobacterial sequences in the shower water are a marker for St. Louis drinking water.

The concept of an endosymbiotic origin of chloroplasts within plants has been accepted (54). In this study, 51% (110/214) of the cyanobacterial sequences found in the water samples were further identified as 16S rRNA genes from chloroplasts of eukaryotic algae. Algae are ubiquitous in surface waters that supply drinking water treatment facilities (36, 69), and therefore, the presence of these cells in the hospital water is not surprising and represents an environmental marker for drinking water in St. Louis, MO. In the summer shower water sample from riser A, 95% (101/106) of cyanobacterial sequences were from chloroplast of the genus Bacillariophyta (diatoms) and matched closest to an environmental uncultured sequence collected from an Indian wetland (98.9% ID). Previous studies have reported similar findings in freshwater rivers and estuaries (18).

Proteobacterial sequences were common in water and aerosol samples.

The phylum Proteobacteria, especially the classes Alpha-, Beta-, and Gammaproteobacteria, is the dominant microbial group that was identified in bulk water distribution systems and drinking water biofilms (9, 26, 45, 59). In the present study, 28% of all the water sequences were Proteobacteria (160/553), and 78% of these were Alphaproteobacteria (124/160). The nitrifying organisms present within Proteobacteria are of particular interest because nitrification along with extended residence times, such as in storage tanks, are known to deplete chloramine residuals and ultimately lead to increased microbial growth (9, 60), resulting in an increase in growth of potentially pathogenic organisms in biofilms. Nitrifying organisms comprised 5% (29/553) of the identified water sequences in this study. We were, therefore, not surprised to find Proteobacteria sequences in the shower water. However, we had not anticipated the Proteobacteria to overwhelm the community in the summer aerosol samples, especially since the source water samples from the summer showed an abundance of Mycobacteriaceae sequences. Species in the family Mycobacteriaceae are known to selectively partition out from standing source water (e.g., pool water) into the aerosol by the “bubble burst” mechanism from the water film due to their hydrophobic cell membranes (5, 50). Here, we did not find this phenomenon, possibly because shower aerosol formation can be explained by a different mechanism, a mechanism similar to “jet mist” generation. Our work verified the work of other gene surveys of air, which found Proteobacteria to be the most common phylum. Specifically, three out of the seven published outdoor air surveys showed Proteobacteria to be exceeding 60% of the bacterial composition (8, 32, 44). In addition, the indoor air of a shopping mall showed similar percentages of Proteobacteria (>60%) (66), while the indoor air in modern Finnish houses consisted of 44 to 50% Proteobacteria (55). Research is necessary to understand the mechanisms of proteobacterial enrichment in air.

Should membrane-integrated showerheads be used to protect extremely immunocompromised patients?

We have shown that membrane-integrated showerheads in shower stalls in a stem cell transplant unit reduced the microbial load in water and aerosol. However, does this warrant the investment of ~$30 for each membrane-integrated showerhead per week (they must be replaced every 7 days) and will they prevent infections in the stem cell transplant patient population? Most of the potentially pathogenic sequences (88/247) in our 16S rRNA gene survey (out of 92/954) were found in one shower water sample (summer water sample from riser A). This potential pathogen, which was identified as M. mucogenicum, was also cultured from these summer shower water samples (with a conventional showerhead), and therefore, at least some of the M. mucogenicum bacteria were viable. Considering the average total bacterial count of 2.2 × 107 cells/liter in the shower water from the conventional showerhead, membrane-integrated showerheads in stem cell transplant units may prevent the transmission of waterborne pathogens from showerheads to extremely vulnerable patients. A further study is needed to determine the incidence of waterborne infections with and without the membrane-integrated showerhead and therefore to firmly conclude whether recommendation for this engineering control is warranted. Conversely, it is unlikely that these showerheads will be able to reduce the occurrence of infections from shower aerosols in the immunocompromised population due to a combination of the low microbial load of shower air (with a conventional showerhead) and the relatively low occurrence of potentially pathogenic bacterial species in the shower aerosol samples.

Acknowledgments

This work was supported by a Barnes-Jewish Hospital Foundation grant (grant 00541-1105-01). Financial support for S.D.P. was through a 2005 graduate research fellowship from the National Science Foundation.

We thank the personnel and patients of the stem cell transplant unit of Barnes-Jewish Hospital in St. Louis, MO, for their patience and support. Finally, we thank Miriam Rosenbaum, Matt Agler, and Jeff Fornero for helpful suggestions.

Footnotes

[down-pointing small open triangle]Published ahead of print on 6 July 2009.

REFERENCES

1. Adékambi, T., P. Berger, D. Raoult, and M. Drancourt. 2006. rpoB gene sequence-based characterization of emerging non-tuberculous mycobacteria with descriptions of Mycobacterium bolletii sp. nov., Mycobacterium phocaicum sp. nov. and Mycobacterium aubagnense sp. nov. Int. J. Syst. Evol. Microbiol. 56:133-143. [PubMed]
2. Adékambi, T., C. Foucault, B. La Scola, and M. Drancourt. 2006. Report of two fatal cases of Mycobacterium mucogenicum central nervous system infection in immunocompetent patients. J. Clin. Microbiol. 44:837-840. [PMC free article] [PubMed]
3. Amann, R. I., W. Ludwig, and K. H. Schleifer. 1995. Phylogenetic identification and in situ detection of individual microbial cells without cultivation. Microbiol. Rev. 59:143-169. [PMC free article] [PubMed]
4. Anaissie, E. J., S. R. Penzak, and M. C. Dignani. 2002. The hospital water supply as a source of nosocomial infections. Arch. Intern. Med. 162:1483-1492. [PubMed]
5. Angenent, L. T., S. T. Kelley, A. St. Amand, N. R. Pace, and M. T. Hernandez. 2005. Molecular identification of potential pathogens in water and air of a hospital therapy pool. Proc. Natl. Acad. Sci. USA 102:4860-4865. [PMC free article] [PubMed]
6. Augustowska, M., and J. Dutkiewicz. 2006. Variability of airborne microflora in a hospital ward within a period of one year. Ann. Agric. Environ. Med. 13:99-106. [PubMed]
7. Augustyn, B. 2007. Ventilator-associated pneumonia. Crit. Care Nurse 27:32-40. [PubMed]
8. Baertsch, C., T. Paez-Rubio, E. Viau, and J. Peccia. 2007. Source tracking aerosols released from land-applied class B biosolids during high-wind events. Appl. Environ. Microbiol. 73:4522-4531. [PMC free article] [PubMed]
9. Berry, D., C. Xi, and L. Raskin. 2006. Microbial ecology of drinking water distribution systems. Curr. Opin. Biotechnol. 17:297-302. [PubMed]
10. Bollin, G. E., J. F. Plouffe, M. F. Para, and B. Hackman. 1985. Aerosols containing Legionella pneumophila generated by shower heads and hot-water faucets. Appl. Environ. Microbiol. 50:1128-1131. [PMC free article] [PubMed]
11. Borchardt, J., J. H. Helbig, and P. C. Luck. 2008. Occurrence and distribution of sequence types among Legionella pneumophila strains isolated from patients in Germany: common features and differences to other regions of the world. Eur. J. Clin. Microbiol. Infect. Dis. 27:29-36. [PubMed]
12. Brettar, I., and M. G. Hofle. 2008. Molecular assessment of bacterial pathogens—a contribution to drinking water safety. Curr. Opin. Biotechnol. 19:274-280. [PubMed]
13. Casini, B., P. Valentini, A. Baggiani, F. Torracca, S. Frateschi, L. C. Nelli, and G. Privitera. 2008. Molecular epidemiology of Legionella pneumophila serogroup 1 isolates following long-term chlorine dioxide treatment in a university hospital water system. J. Hosp. Infect. 69:141-147. [PubMed]
14. Chate, R. A. 2006. An audit improves the quality of water within the dental unit water lines of three separate facilities of a United Kingdom NHS trust. Br. Dent. J. 201:565-569. [PubMed]
15. Christen, R. 2008. Identifications of pathogens—a bioinformatic point of view. Curr. Opin. Biotechnol. 19:266-273. [PubMed]
16. Cole, J. R., B. Chai, R. J. Farris, Q. Wang, A. S. Kulam-Syed-Mohideen, D. M. McGarrell, A. M. Bandela, E. Cardenas, G. M. Garrity, and J. M. Tiedje. 2007. The ribosomal database project (RDP-II): introducing myRDP space and quality controlled public data. Nucleic Acids Res. 35:D169-D172. [PMC free article] [PubMed]
17. Coscolla, M., and F. Gonzalez-Candelas. 2007. Population structure and recombination in environmental isolates of Legionella pneumophila. Environ. Microbiol. 9:643-656. [PubMed]
18. Crump, B. C., E. V. Armbrust, and J. A. Baross. 1999. Phylogenetic analysis of particle-attached and free-living bacterial communities in the Columbia river, its estuary, and the adjacent coastal ocean. Appl. Environ. Microbiol. 65:3192-3204. [PMC free article] [PubMed]
19. DeSantis, T. Z., P. Hugenholtz, K. Keller, E. L. Brodie, N. Larsen, Y. M. Piceno, R. Phan, and G. L. Andersen. 2006. NAST: a multiple sequence alignment server for comparative analysis of 16S rRNA genes. Nucleic Acids Res. 34:W394-W399. [PMC free article] [PubMed]
20. DeSantis, T. Z., P. Hugenholtz, N. Larsen, M. Rojas, E. L. Brodie, K. Keller, T. Huber, D. Dalevi, P. Hu, and G. L. Andersen. 2006. Greengenes, a chimera-checked 16S rRNA gene database and workbench compatible with ARB. Appl. Environ. Microbiol. 72:5069-5072. [PMC free article] [PubMed]
21. Diederen, B. M. 2008. Legionella spp. and Legionnaires' disease. J. Infect. 56:1-12. [PubMed]
22. Dohmae, S., T. Okubo, W. Higuchi, T. Takano, H. Isobe, T. Baranovich, S. Kobayashi, M. Uchiyama, Y. Tanabe, M. Itoh, and T. Yamamoto. 2008. Bacillus cereus nosocomial infection from reused towels in Japan. J. Hosp. Infect. 69:361-367. [PubMed]
23. Doring, G., M. Ulrich, W. Muller, J. Bitzer, L. Schmidtkoenig, L. Munst, H. Grupp, C. Wolz, M. Stern, and K. Botzenhart. 1991. Generation of Pseudomonas aeruginosa aerosols during hand-washing from contaminated sink drains, transmission to hands of hospital personnel, and its prevention by use of a new heating device. Zentralbl. Hyg. Umweltmed. 191:494-505. [PubMed]
24. du Moulin, G. C., K. D. Stottmeier, P. A. Pelletier, A. Y. Tsang, and J. Hedley-Whyte. 1988. Concentration of Mycobacterium avium by hospital hot water systems. JAMA 260:1599-1601. [PubMed]
25. Edwards, C. 2000. Problems posed by natural environments for monitoring microorganisms. Mol. Biotechnol. 15:211-223. [PubMed]
26. Emtiazi, F., T. Schwartz, S. M. Marten, P. Krolla-Sidenstein, and U. Obst. 2004. Investigation of natural biofilms formed during the production of drinking water from surface water embankment filtration. Water Res. 38:1197-1206. [PubMed]
27. Fabian, M. P., S. L. Miller, T. Reponen, and M. T. Hernandez. 2005. Ambient bioaerosol indices for indoor air quality assessments of flood reclamation. J. Aerosol Sci. 36:763-783.
28. Falkinham, J. O., III. 2003. Mycobacterial aerosols and respiratory disease. Emerg. Infect. Dis. 9:763-767. [PMC free article] [PubMed]
29. Falkinham, J. O., III, C. D. Norton, and M. W. LeChevallier. 2001. Factors influencing numbers of Mycobacterium avium, Mycobacterium intracellulare, and other mycobacteria in drinking water distribution systems. Appl. Environ. Microbiol. 67:1225-1231. [PMC free article] [PubMed]
30. Favre-Bonte, S., E. Chamot, T. Kohler, J. A. Romand, and C. van Delden. 2007. Autoinducer production and quorum-sensing dependent phenotypes of Pseudomonas aeruginosa vary according to isolation site during colonization of intubated patients. BMC Microbiol. 7:33. [PMC free article] [PubMed]
31. Ferroni, A., L. Nguyen, B. Pron, G. Quesne, M. C. Brusset, and P. Berche. 1998. Outbreak of nosocomial urinary tract infections due to Pseudomonas aeruginosa in a paediatric surgical unit associated with tap-water contamination. J. Hosp. Infect. 39:301-307. [PubMed]
32. Fierer, N., Z. Liu, M. Rodriguez-Hernandez, R. Knight, M. Henn, and M. T. Hernandez. 2008. Short-term temporal variability in airborne bacterial and fungal populations. Appl. Environ. Microbiol. 74:200-207. [PMC free article] [PubMed]
33. Frank, D. N. 2008. Xplorseq: a software environment for integrated management and phylogenetic analysis of metagenomic sequence data. BMC Bioinformatics 9:420. [PMC free article] [PubMed]
34. Garcia-Nunez, M., N. Sopena, S. Ragull, M. L. Pedro-Botet, J. Morera, and M. Sabria. 2008. Persistence of Legionella in hospital water supplies and nosocomial Legionnaires' disease. FEMS Immunol. Med. Microbiol. 52:202-206. [PubMed]
35. Golovlev, E. L. 2000. General and molecular ecology of Legionella. Mikrobiologiia 69:5-12. (In Russian.) [PubMed]
36. Henderson, R. K., A. Baker, S. A. Parsons, and B. Jefferson. 2008. Characterisation of algogenic organic matter extracted from cyanobacteria, green algae and diatoms. Water Res. 42:3435-3445. [PubMed]
37. Huber, T., G. Faulkner, and P. Hugenholtz. 2004. Bellerophon: a program to detect chimeric sequences in multiple sequence alignments. Bioinformatics 20:2317-2319. [PubMed]
37a. Hussein, Z., O. Landt, B. Wirths, and N. Wellinghausen. 2009. Detection of non-tuberculous mycobacteria in hospital water by culture and molecular methods. Int. J. Med. Microbiol. 299:281-290. [PubMed]
38. Kassis, I., I. Oren, S. Davidson, R. Finkelstein, G. Rabino, T. Katz, and H. Sprecher. 2007. Contamination of peripheral hematopoeitic stem cell products with a Mycobacterium mucogenicum-related pathogen. Infect. Control Hosp. Epidemiol. 28:755-757. [PubMed]
39. La Scola, B., M. N. Mallet, P. A. Grimont, and D. Raoult. 2003. Bosea eneae sp. nov., Bosea massiliensis sp. nov. and Bosea vestrisii sp. nov., isolated from hospital water supplies, and emendation of the genus Bosea (Das et al. 1996). Int. J. Syst. Evol. Microbiol. 53:15-20. [PubMed]
40. Ley, R. E., P. J. Turnbaugh, S. Klein, and J. I. Gordon. 2006. Microbial ecology: human gut microbes associated with obesity. Nature 444:1022-1023. [PubMed]
41. Li, C. S., and P. A. Hou. 2003. Bioaerosol characteristics in hospital clean rooms. Sci. Total Environ. 305:169-176. [PubMed]
42. Lozupone, C., M. Hamady, and R. Knight. 2006. Unifrac—an online tool for comparing microbial community diversity in a phylogenetic context. BMC Bioinformatics 7:371. [PMC free article] [PubMed]
43. Ludwig, W., O. Strunk, R. Westram, L. Richter, H. Meier, Yadhukumar, A. Buchner, T. Lai, S. Steppi, G. Jobb, W. Forster, I. Brettske, S. Gerber, A. W. Ginhart, O. Gross, S. Grumann, S. Hermann, R. Jost, A. Konig, T. Liss, R. Lussmann, M. May, B. Nonhoff, B. Reichel, R. Strehlow, A. Stamatakis, N. Stuckmann, A. Vilbig, M. Lenke, T. Ludwig, A. Bode, and K.-H. Schleifer. 2004. ARB: a software environment for sequence data. Nucleic Acids Res. 32:1363-1371. [PMC free article] [PubMed]
44. Maron, P. A., D. Lejon, E. Carvalho, K. Bizet, P. Lemanceau, L. Ranjard, and C. Mougel. 2005. Assessing genetic structures and diversity of airborne bacterial communities by DNA fingerprinting and 16S rDNA clone library. Atmos. Environ. 39:3687-3695.
45. Martiny, A. C., H. J. Albrechtsen, E. Arvin, and S. Molin. 2005. Identification of bacteria in biofilm and bulk water samples from a nonchlorinated model drinking water distribution system: detection of a large nitrite-oxidizing population associated with Nitrospira spp. Appl. Environ. Microbiol. 71:8611-8617. [PMC free article] [PubMed]
46. Muder, R. R., and V. L. Yu. 2002. Infection due to Legionella species other than L. pneumophila. Clin. Infect. Dis. 35:990-998. [PubMed]
47. Nadkarni, M. A., F. E. Martin, N. A. Jacques, and N. Hunter. 2002. Determination of bacterial load by real-time PCR using a broad-range (universal) probe and primers set. Microbiology 148:257-266. [PubMed]
48. O'Grady, N. P., M. Alexander, E. P. Dellinger, J. L. Gerberding, S. O. Heard, D. G. Maki, H. Masur, R. D. McCormick, L. A. Mermel, M. L. Pearson, I. I. Raad, A. Randolph, and R. A. Weinstein. 2002. Guidelines for the prevention of intravascular catheter-related infections. Infect. Control Hosp. Epidemiol. 23:759-769. [PubMed]
49. Pace, N. R. 1997. A molecular view of microbial diversity and the biosphere. Science 276:734-740. [PubMed]
50. Parker, B. C., M. A. Ford, H. Gruft, and J. O. Falkinham III. 1983. Epidemiology of infection by nontuberculous mycobacteria. IV. Preferential aerosolization of Mycobacterium intracellulare from natural waters. Am. Rev. Respir. Dis. 128:652-656. [PubMed]
51. Reference deleted.
52. Radosevich, J. L., W. J. Wilson, J. H. Shinn, T. Z. DeSantis, and G. L. Andersen. 2002. Development of a high-volume aerosol collection system for the identification of air-borne micro-organisms. Lett. Appl. Microbiol. 34:162-167. [PubMed]
53. Rauer, D. 2005. Characterization and monitoring of ambient biological PM-2.5 at the St. Louis-Midwest Supersite. M.S. thesis. Washington University, St. Louis, MO.
54. Reith, M., and R. A. Cattolico. 1986. Inverted repeat of Olisthodiscus luteus chloroplast DNA contains genes for both subunits of ribulose-1,5-bisphosphate carboxylase and the 32,000-dalton Q(B) protein: phylogenetic implications. Proc. Natl. Acad. Sci. USA 83:8599-8603. [PMC free article] [PubMed]
55. Rintala, H., M. Pitkaranta, M. Toivola, L. Paulin, and A. Nevalainen. 2008. Diversity and seasonal dynamics of bacterial community in indoor environment. BMC Microbiol. 8:56. [PMC free article] [PubMed]
56. Rittmann, B. E. 2002. The role of molecular methods in evaluating biological treatment processes. Water Environ. Res. 74:421-427. [PubMed]
57. Sakamoto, C., N. Yamaguchi, M. Yamada, H. Nagase, M. Seki, and M. Nasu. 2007. Rapid quantification of bacterial cells in potable water using a simplified microfluidic device. J. Microbiol. Methods 68:643-647. [PubMed]
58. Schloss, P. D., and J. Handelsman. 2005. Introducing DOTUR, a computer program for defining operational taxonomic units and estimating species richness. Appl. Environ. Microbiol. 71:1501-1506. [PMC free article] [PubMed]
59. Schmeisser, C., C. Stockigt, C. Raasch, J. Wingender, K. N. Timmis, D. F. Wenderoth, H. C. Flemming, H. Liesegang, R. A. Schmitz, K. E. Jaeger, and W. R. Streit. 2003. Metagenome survey of biofilms in drinking-water networks. Appl. Environ. Microbiol. 69:7298-7309. [PMC free article] [PubMed]
60. Srinivasan, S., G. W. Harrington, I. Xagoraraki, and R. Goel. 2008. Factors affecting bulk to total bacteria ratio in drinking water distribution systems. Water Res. 42:3393-3404. [PubMed]
61. Stenfors Arnesen, L. P., A. Fagerlund, and P. E. Granum. 2008. From soil to gut: Bacillus cereus and its food poisoning toxins. FEMS Microbiol. Rev. 32:579-606. [PubMed]
62. Subrata, K. D., A. K. Mischra, B. J. Tindall, F. A. Rainey, and E. Stackebrandt. 1996. Oxidation of thiosulfate by a new bacterium, Bosea thiooxidans (strain BI-42) gen. nov., sp. nov.: analysis of phylogeny based on chemotaxonomy and 16S ribosomal DNA sequencing. Int. J. Syst. Bacteriol. 46:981-987. [PubMed]
63. Szewzyk, U., R. Szewzyk, W. Manz, and K. H. Schleifer. 2000. Microbiological safety of drinking water. Annu. Rev. Microbiol. 54:81-127. [PubMed]
64. Thomas, V., N. Casson, and G. Greub. 2007. New Afipia and Bosea strains isolated from various water sources by amoebal co-culture. Syst. Appl. Microbiol. 30:572-579. [PubMed]
65. Tortoli, E. 2003. Impact of genotypic studies on mycobacterial taxonomy: the new mycobacteria of the 1990s. Clin. Microbiol. Rev. 16:319-354. [PMC free article] [PubMed]
66. Tringe, S. G., T. Zhang, X. Liu, Y. Yu, W. H. Lee, J. Yap, F. Yao, S. T. Suan, S. K. Ing, M. Haynes, F. Rohwer, C. L. Wei, P. Tan, J. Bristow, E. M. Rubin, and Y. Ruan. 2008. The airborne metagenome in an indoor urban environment. PLoS One 3:e1862. [PMC free article] [PubMed]
67. U.S. Environmental Protection Agency. 2003. National primary drinking water standards. EPA publication no. 816F03016. Office of Water, U.S. Environmental Protection Agency, Washington, DC.
68. Vaerewijck, M. J., G. Huys, J. C. Palomino, J. Swings, and F. Portaels. 2005. Mycobacteria in drinking water distribution systems: ecology and significance for human health. FEMS Microbiol. Rev. 29:911-934. [PubMed]
69. von Hertzen, L., T. Laatikainen, T. Pitkanen, T. Vlasoff, M. J. Makela, E. Vartiainen, and T. Haahtela. 2007. Microbial content of drinking water in Finnish and Russian karelia—implications for atopy prevalence. Allergy 62:288-292. [PubMed]
70. Walker, J. T., and P. D. Marsh. 2007. Microbial biofilm formation in DUWS and their control using disinfectants. J. Dent. 35:721-730. [PubMed]
71. Wan, G. H., S. C. Lu, and Y. H. Tsai. 2004. Polymerase chain reaction used for the detection of airborne mycobacterium tuberculosis in health care settings. Am. J. Infect. Control 32:17-22. [PubMed]
72. Wilgenbusch, J. C., and D. Swofford. 2003. Inferring evolutionary trees with PAUP*. Curr. Protoc. Bioinformatics, unit 6.4. doi:.10.1002/0471250953.bi0604s00 [PubMed] [Cross Ref]

Articles from Applied and Environmental Microbiology are provided here courtesy of American Society for Microbiology (ASM)
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...