ABSTRACT
We analyzed 69 bacterial isolates, comprising seven species of gram-negative bacterial rods and three species of coagulase-negative staphylococci, recovered from both the hands of caretakers and their environment in households sampled in upper Manhattan. Repetitive sequence-based PCR and dendrogram analysis were used to determine strain similarity. Greater than 25% of individual species of Acinetobacter, Enterobacter, and coagulase-negative staphylococci recovered from the hands and immediate environment within each household shared the same genotype. This study is the first to demonstrate the frequency of bacteria shared within community households. These strains may serve as potential reservoirs for either community- or hospital-acquired infections.
Published studies that examine the microbial flora in community settings are sparse in comparison to those within hospital environments (1, 14). Nosocomial infections can emanate from endemic hospital strains or from microorganisms endogenous in the community. Recently, nosocomial transmission of a community-associated strain of methicillin-resistant Staphylococcus aureus (CA-MRSA) among postpartum women has been reported (28), underscoring the clinical importance of determining the frequency of shared or similar bacterial genotypes within households. Studies examining the microbial ecology of microorganisms within households are scarce but are clinically relevant, because they enhance our understanding of pathogen transmission within both community and hospital settings. The emergence of nosocomial infections due to gram-negative rods (GNR) is of paramount importance. Infections due to Acinetobacter species, particularly Acinetobacter baumannii, causing nosocomial bacteremia and pneumonia have a high mortality rate (9, 11). However, there is a lack of published reports concerning the epidemiology of clonal transmission of A.baumannii and other GNR within community settings.
The advent of molecular technology has facilitated the examination of strain similarities. Molecular methods available for genotyping isolates are separated into those using DNA amplification by PCR and non-nucleic acid amplification techniques (26, 35). Both pulsed-field gel electrophoresis and repetitive sequence-based PCR (rep-PCR) analyze large parts of the bacterial genome, and they offer greater discriminatory power than those that employ the use of small regions of single genes or operons (e.g., rRNA) (26, 33). PCR typing methods using specific primers designed on the basis of the repetitive and conserved sequences in bacteria and stringent annealing conditions generate complex fingerprint patterns for strain typing.
Some reports have shown that results obtained using rep-PCR are superior to those obtained using manual or automated ribotyping (6, 8, 19). The two types of repetitive elements most commonly used to differentiate a vast variety of strains are the enterobacterial repetitive intergenic consensus and repetitive extragenic palindromic sequences (7, 16, 31). The rep-PCR technology was chosen for strain typing in this study because of its rapidity, relatively low cost, ease of use, and applicability to typing a wide variety of strains, e.g., S. aureus (32), Streptococcus species (8, 23), Clostridium difficile (30), Salmonella enterica serovar Enteritidis (5), Shigella species (18), Enterococcus faecalis (20), Enterobacter aerogenes (21), Burkholderia cepacia (19), Stenotrophomonas maltophilia (22), A. baumannii (3), Pasteurella multocida (10), Vibrio parahaemolyticus (34), Lactobacillus johnsonii (33), Helicobacter pylori (4), and Bartonella henselae (29).
The goal of the present study was to determine if bacterial isolates recovered from the hands of caretakers were similar or different from those obtained from a designated environmental site within their households. To the best of our knowledge, this is the first study of its kind to determine the prevalence and frequency of bacterial strains, particularly GNR, shared within community settings.
MATERIALS AND METHODS
Specimen collection.A total of 238 primary caretakers living in inner-city multigenerational households in northern Manhattan, New York, with an immigrant population, predominantly from the Dominican Republic, were enrolled in this study. The participants were part of a larger clinical trial to determine the effect of hygiene practices on infectious disease symptoms (15).
Hand cultures were obtained during a home visit using a modification of the “glove juice” method (14). Briefly, a randomly selected hand was cultured immediately after washing by placing in a sterile polyethylene bag containing 50 ml of sampling solution (0.075 M phosphate buffer, pH 7.9, containing 0.1% polysorbate 80 and 0.1% sodium thiosulfate). The hand was massaged through the wall of the bag for 60 seconds by a trained data collector, and samples were hand delivered to the clinical microbiology laboratory for processing.
To detect environmental reservoirs, an area about the size of a quarter was sampled from the table used for eating meals, using a premoistened cotton-tipped swab (Mini-Tip culturette collection and transport system; Becton Dickinson Microbiology Systems, Sparks, MD). This environment site was chosen since the subjects reported eating most of their meals within the home at the table and, therefore, it reflected a high-contact/high-traffic area.
After collection, swabs were placed into 2 ml of sampling solution, capped, and delivered to the clinical microbiology laboratory.
Specimen processing.All specimens were processed within 2 to 3 h of collection or refrigerated for up to 12 h before processing. Samples were diluted 10-fold in 0.85% normal saline, up to 10−3, and plated on five different medium types to optimize the recovery of bacterial and fungal organisms. The enriched and selective agar media included Columbia agar with 5% sheep blood, MacConkey, colistin-nalidixic acid, Sabouraud's with chloramphenicol and gentamicin, and bile esculin (Becton Dickinson Microbiology Systems, Sparks, MD). The plates were incubated at 35°C and examined daily for microbial growth over 48 h.
Species identification.Gram-negative and gram-positive bacteria were identified to the species level using the MicroScan system (Dade Behring Inc., Deerfield, IL). A high percentage (≥95%) was utilized as the acceptance criterion for identification by MicroScan. Staphylococci were additionally identified by coagulase and Staphaurex (Remel Europe Ltd., Kent, United Kingdom.). All isolates were frozen at −70°C for further analyses.
Epidemiologic strain typing.From each study household, paired specimens containing the same bacterial species isolated from both the hand of the primary caretaker and the environmental area were evaluated. For each pair, isolates were genotyped using rep-PCR (DiversiLab DNA fingerprinting kits; Spectral Genomics, Houston, TX) to determine strain similarities.
Rep-PCR.Cultures were incubated overnight (37°C), and the DNA was extracted from 1 μl of culture with the UltraClean microbial DNA isolation kit (Mo Bio Laboratories, Inc., Solana Beach, CA). The DNA samples were standardized to 25 ng/μl, and genomic integrity was verified using conventional agarose gel electrophoresis. DNA (50 ng) from staphylococci was amplified using the DiversiLab Staphylococcus kit (Spectral Genomics) and from GNR using a DiversiLab DNA fingerprinting kit and UprimeE1 plus UprimeB2 primers.
The following thermal cycling parameters were used: initial denaturation at 94°C (2 min), 35 cycles of denaturation at 94°C (30 s), annealing at 55°C (or 45°C for staphylococci) (30 s), extension at 70°C (90 s), and a final extension at 70°C (3 min). The amplicons were separated by 1.5% agarose gel electrophoresis, and digital images of the gels were captured with the AlphaImager 2200 (Alpha Innotech Corp., San Leandro, CA). Samples were analyzed with BioNumerics software package using the Pearson's correlation coefficient and unweighted pair group method with arithmetic means to create dendrograms (Applied Maths, Inc., Austin, TX). The relatedness was determined by cluster analysis and guidelines provided by the manufacturer. Isolates were categorized as different, similar, or indistinguishable. Because homogeneous organisms (such as MRSA) have less variation between genomes while heterogeneous organisms (such as Pseudomonas and Acinetobacter) have a larger natural presence in the environment and have greater genomic variations, the following criteria were applied to determine relatedness. In general, “different” was defined as <95% similarity and two band differences for homogeneous organisms or three band differences for heterogeneous organisms. “Similar” was defined as <97% similarity and one band difference for homogeneous organisms or up to two band differences for heterogeneous organisms. “Indistinguishable” was defined as >95% similarity and no banding differences, including no variation in intensities of individual bands, although overall intensities may differ. DNA from a Pantoea agglomerans strain was included as a positive kit control, while water was included as a negative control for the performance of rep-PCR.
RESULTS
Bacterial strains.There were 69 instances in which the same bacterial species were recovered from both the hand and the environmental surface sampled (Table 1). Of these, 81% were GNR and 19% were gram-positive cocci (GPC). Among the GNR matched pairs, 29 were Acinetobacter species (52%), of which 15 A. baumannii and 14 Acinetobacter lwoffii isolates were found. The Acinetobacter strains were susceptible to the majority of antimicrobial agents tested, and results have been published elsewhere (37). The remaining GNR included 12 Enterobacter agglomerans, 2 Enterobacter cloacae, 9 Klebsiella pneumoniae, 3 Pseudomonas fluorescens/P. putida, and 1 S. maltophilia isolate. All GPC matched pairs were coagulase-negative staphylococci and included nine Staphylococcus warneri, three Staphylococcus epidermidis, and one Staphylococcus haemolyticus isolate.
Rep-PCR analysis of paired bacterial strains from hands and household environments
Strain differentiation by rep-PCR.DNA fragment patterns by rep-PCR were obtained for each matched bacterial species. The dendrogram analyses revealed significant correlations between hand and the environmental isolates. Similar strains were identified between the paired isolates from the hands and environment by using a Pearson's correlation coefficient of >95% similarity.
(i) Relationships between paired hand and environmental strains.Results indicated that nearly 29% of all paired hand and environmental isolates tested were similar (26.8% GNR and 36.5% GPC) (Table 1). For GNR, A. baumannii, 6/15 (40%) matched hand and environmental pairs were similar or indistinguishable (Ab1, Ab2, Ab6, Ab7, Ab12, and Ab15) and 9/15 (60%) pairs were different (Table 1; Fig. 1). For A. lwoffii, 2/14 (14%) matched pairs were similar and 12/14 (86%) pairs were different. For E. agglomerans, 1/12 (8%) paired isolates were similar (Ea11) and 11/12 (92%) pairs were different (Table 1; Fig. 2). Of two paired E. cloacae isolates, one was similar and one was different. For K. pneumoniae, 4/9 (44%) paired isolates were similar and 5/9 (56%) pairs were different (Table 1). For P. fluorescens/P. putida, 1/3 (33%) were similar and 2/3 (67%) pairs were different (Table 1).
Rep-PCR gel images and dendrogram of A. baumannii strains isolated from hands (h) and the environment (e), showing the degree of strain similarity.
Rep-PCR gel images and dendrogram of E. agglomerans strains isolated from hands (h) and the environment (e), showing the degree of strain similarity.
One pair of S. maltophilia isolates was different. Among GPC, 2/9 (22%) S. warneri matched pairs were similar (Sw4 and Sw9) (Fig. 3) and 7/9 (78%) were different. For S. epidermidis, 2/3 (67%) matched pairs were similar and 1/3 (33%) pairs were different. The single matched pair of S. haemolyticus was similar.
Rep-PCR gel images and dendrogram of S. warneri strains isolated from hands (h) and the environment (e), showing the degree of strain similarity.
(ii) Relationships among strains from different households.The majority of gram-negative and gram-positive bacteria recovered from different households were unique strains that were readily distinguishable by rep-PCR. Despite their isolation from different households, some strains of A. baumannii (Ab10e, Ab14e and Ab2h, Ab2e, Ab3h, Ab4h, and Ab5h), E. agglomerans (Ea3e, Ea7h, and Ea2h), and S. warneri (Sw2e, Sw3e, Sw8e, Sw1e and Sw7h, Sw4h, Sw6h, Sw4e, Sw9h, and Sw9e) displayed indistinguishable patterns (Fig. 1, 2, and 3, respectively). This may reflect the presence of clonal strains within the community.
DISCUSSION
While the majority of paired isolates occurred as unique, source-specific strains, greater than 25% of Acinetobacter species (Fig. 1), E. agglomerans (8%) (Fig. 2), and S. warneri (22%) (Fig. 3) organisms isolated from the hands and environments of individual households shared common genotypes. Although Acinetobacter colonization on human skin and fingertips and prolonged survival in the environment (27 days) has been reported (2, 12, 17), this is the first study to document same-strain sharing between hands and environmental surfaces within a community setting. Over the past two decades, Acinetobacter species have become virulent pathogens, responsible for nosocomial infections and outbreaks, particularly in intensive care units (36). Therefore, the majority of published studies have concentrated on the hospital epidemiology of these organisms (27) and not on those within the community. Interestingly, MRSA, previously limited to causing nosocomial infections, has emerged as a virulent pathogen in the community among individuals without established risk factors for MRSA infection (24, 25). Saiman et al. recently reported that a virulent CA-MRSA strain had caused a nosocomial outbreak among postpartum women in our medical center (28), underscoring the importance of transmission of community strains in hospital settings.
Nosocomial A. baumannii isolates, compared to those isolated from the community, have been shown to be multidrug resistant and clonally related (12, 13, 37), suggesting that the hospital environment serves as the reservoir for epidemic A. baumannii strains. Although A. baumannii isolates in the community are more drug susceptible, they also are clonal and could be encountered as potential pathogens within the hospital environment (37).
To the best of our knowledge, no data have been reported demonstrating shared genotypes of A. baumannii, E. agglomerans, and S. warneri between the hands of caretakers and their household environments. The sampled environmental surface represented only one high-contact/high-traffic source area within the home; we cannot comment on whether there are other reservoirs harboring genetically similar shared species among other areas within the home.
Because transmission of community strains (e.g., recent example of CA-MRSA) into the hospital can occur, we speculate that these strains may comprise a natural reservoir of short- or long-term duration within the community that could serve as a source for nosocomial infections.
ACKNOWLEDGMENTS
This study was funded in part by the National Institutes of Health, National Institute of Nursing Research, RO1 NR05251 (Home Hygiene Intervention to E.L.L.).
FOOTNOTES
- Received 5 April 2005.
- Returned for modification 16 June 2005.
- Accepted 7 July 2005.
- Copyright © 2005 American Society for Microbiology
