Assessing and removing the effect of unwanted technical variations in microbiome data.

Varying technologies and experimental approaches used in microbiome studies often lead to irreproducible results due to unwanted technical variations. Such variations, often unaccounted for and of unknown source, may interfere with true biological signals, resulting in misleading biological conclusions. In this work, we aim to characterize the major sources of technical variations in microbiome data and demonstrate how in-silico approaches can minimize their impact. We analyzed 184 pig faecal metagenomes encompassing 21 specific combinations of deliberately introduced factors of technical and biological variations. Using the novel Removing Unwanted Variations-III-Negative Binomial (RUV-III-NB), we identified several known experimental factors, specifically storage conditions and freeze-thaw cycles, as likely major sources of unwanted variation in metagenomes. We also observed that these unwanted technical variations do not affect taxa uniformly, with freezing samples affecting taxa of class Bacteroidia the most, for example. Additionally, we benchmarked the performances of different correction methods, including ComBat, ComBat-seq, RUVg, RUVs, and RUV-III-NB. While RUV-III-NB performed consistently robust across our sensitivity and specificity metrics, most other methods did not remove unwanted variations optimally. Our analyses suggest that a careful consideration of possible technical confounders is critical during experimental design of microbiome studies, and that the inclusion of technical replicates is necessary to efficiently remove unwanted variations computationally.

Mucosal microbiotas and their role in stem cell transplantation.

Mucosal microbiotas and their role in stem cell transplantation. Patients with hematological disorders such as leukemia often undergo allogeneic hematopoietic stem cell transplantation, and thereby receive stem cells from a donor for curation of disease. This procedure also involves immunosuppressive and antimicrobial treatments that disturb the important interactions between the microbiota and the immune system, especially at mucosal sites. After transplantation, bacterial diversity decreases together with a depletion of Clostridia, and shifts toward predominance of Proteobacteria. Infectious and inflammatory complications, such as graft-versus-host disease, also interfere with patient recovery. This review collects and contextualizes current knowledge of the role of mucosal microbiotas at different body sites in stem cell transplantation, proposes underlying mechanisms, and discusses potential clinical value of bacterial markers for improved treatment strategies.

Standard Sample Storage Conditions Have an Impact on Inferred Microbiome Composition and Antimicrobial Resistance Patterns.

Storage of biological specimens is crucial in the life and medical sciences. Storage conditions for samples can be different for a number of reasons, and it is unclear what effect this can have on the inferred microbiome composition in metagenomics analyses. Here, we assess the effect of common storage temperatures (deep freezer, -80°C; freezer, -20°C; refrigerator, 5°C; room temperature, 22°C) and storage times (immediate sample processing, 0 h; next day, 16 h; over weekend, 64 h; longer term, 4, 8, and 12 months) as well as repeated sample freezing and thawing (2 to 4 freeze-thaw cycles). We examined two different pig feces and sewage samples, unspiked and spiked with a mock community, in triplicate, respectively, amounting to a total of 438 samples (777 Gbp; 5.1 billion reads). Storage conditions had a significant and systematic effect on the taxonomic and functional composition of microbiomes. Distinct microbial taxa and antimicrobial resistance classes were, in some situations, similarly affected across samples, while others were not, suggesting an impact of individual inherent sample characteristics. With an increasing number of freeze-thaw cycles, an increasing abundance of Firmicutes, Actinobacteria, and eukaryotic microorganisms was observed. We provide recommendations for sample storage and strongly suggest including more detailed information in the metadata together with the DNA sequencing data in public repositories to better facilitate meta-analyses and reproducibility of findings. IMPORTANCE Previous research has reported effects of DNA isolation, library preparation, and sequencing technology on metagenomics-based microbiome composition; however, the effect of biospecimen storage conditions has not been thoroughly assessed. We examined the effect of common sample storage conditions on metagenomics-based microbiome composition and found significant and, in part, systematic effects. Repeated freeze-thaw cycles could be used to improve the detection of microorganisms with more rigid cell walls, including parasites. We provide a data set that could also be used for benchmarking algorithms to identify and correct for unwanted batch effects. Overall, the findings suggest that all samples of a microbiome study should be stored in the same way. Furthermore, there is a need to mandate more detailed information about sample storage and processing be published together with DNA sequencing data at the International Nucleotide Sequence Database Collaboration (ENA/EBI, NCBI, DDBJ) or other repositories.

A Peek into the Plasmidome of Global Sewage.

Plasmids can provide a selective advantage for microorganisms to survive and adapt to new environmental conditions. Plasmid-encoded traits, such as antimicrobial resistance (AMR) or virulence, impact the ecology and evolution of bacteria and can significantly influence the burden of infectious diseases. Insight about the identity and functions encoded on plasmids on the global scale are largely lacking. Here, we investigate the plasmidome of 24 samples (22 countries, 5 continents) from the global sewage surveillance project. We obtained 105-Gbp Oxford Nanopore and 167-Gbp Illumina NextSeq DNA sequences from plasmid DNA preparations and assembled 165,302 contigs (159,322 circular). Of these, 58,429 carried genes encoding for plasmid-related and 11,222 for virus/phage-related proteins. About 90% of the circular DNA elements did not have any similarity to known plasmids. Those that exhibited similarity had similarity to plasmids whose hosts were previously detected in these sewage samples (e.g., Acinetobacter, Escherichia, Moraxella, Enterobacter, Bacteroides, and Klebsiella). Some AMR classes were detected at a higher abundance in plasmidomes (e.g., macrolide-lincosamide-streptogramin B, macrolide, and quinolone) compared to the respective complex sewage samples. In addition to AMR genes, a range of functions were encoded on the candidate plasmids, including plasmid replication and maintenance, mobilization, and conjugation. In summary, we describe a laboratory and bioinformatics workflow for the recovery of plasmids and other potential extrachromosomal DNA elements from complex microbiomes. Moreover, the obtained data could provide further valuable insight into the ecology and evolution of microbiomes, knowledge about AMR transmission, and the discovery of novel functions. IMPORTANCE This is, to the best of our knowledge, the first study to investigate plasmidomes at a global scale using long read sequencing from complex untreated domestic sewage. Previous metagenomic surveys have detected AMR genes in a variety of environments, including sewage. However, it is unknown whether the AMR genes were present on the microbial chromosome or located on extrachromosomal elements, such as plasmids. Using our approach, we recovered a large number of plasmids, of which most appear novel. We identified distinct AMR genes that were preferentially located on plasmids, potentially contributing to their transmissibility. Overall, plasmids are of great importance for the biology of microorganisms in their natural environments (free-living and host-associated), as well as for molecular biology and biotechnology. Plasmidome collections may therefore be valuable resources for the discovery of fundamental biological mechanisms and novel functions useful in a variety of contexts.

Microbiota long-term dynamics and prediction of acute graft-versus-host disease in pediatric allogeneic stem cell transplantation.

BACKGROUND: Patients undergoing allogeneic hematopoietic stem cell transplantation (HSCT) exhibit changes in their gut microbiota and are experiencing a range of complications, including acute graft-versus-host disease (aGvHD). It is unknown if, when, and under which conditions a re-establishment of microbial and immunological homeostasis occurs. It is also unclear whether microbiota long-term dynamics occur at other body sites than the gut such as the mouth or nose. Moreover, it is not known whether the patients' microbiota prior to HSCT holds clues to whether the patient would suffer from severe complications subsequent to HSCT. Here, we take a holobiont perspective and performed an integrated host-microbiota analysis of the gut, oral, and nasal microbiota in 29 children undergoing allo-HSCT. RESULTS: The bacterial diversity decreased in the gut, nose, and mouth during the first month and reconstituted again 1-3 months after allo-HSCT. The microbial community composition traversed three phases over 1 year. Distinct taxa discriminated the microbiota temporally at all three body sides, including Enterococcus spp., Lactobacillus spp., and Blautia spp. in the gut. Of note, certain microbial taxa appeared already changed in the patients prior to allo-HSCT as compared with healthy children. Acute GvHD occurring after allo-HSCT could be predicted from the microbiota composition at all three body sites prior to HSCT. The reconstitution of CD4+ T cells, TH17, and B cells was associated with distinct taxa of the gut, oral, and nasal microbiota. CONCLUSIONS: This study reveals for the first time bacteria in the mouth and nose that may predict aGvHD. Monitoring of the microbiota at different body sites in HSCT patients and particularly through involvement of samples prior to transplantation may be of prognostic value and could assist in guiding personalized treatment strategies. The identification of distinct bacteria that have a potential to predict post-transplant aGvHD might provide opportunities for an improved preventive clinical management, including a modulation of microbiomes. The host-microbiota associations shared between several body sites might also support an implementation of more feasible oral and nasal swab sampling-based analyses. Altogether, the findings suggest that the microbiota and host factors together could provide actionable information to guiding precision medicine. Video Abstract.

Reconstruction of ancient microbial genomes from the human gut.

Loss of gut microbial diversity1-6 in industrial populations is associated with chronic diseases7, underscoring the importance of studying our ancestral gut microbiome. However, relatively little is known about the composition of pre-industrial gut microbiomes. Here we performed a large-scale de novo assembly of microbial genomes from palaeofaeces. From eight authenticated human palaeofaeces samples (1,000-2,000 years old) with well-preserved DNA from southwestern USA and Mexico, we reconstructed 498 medium- and high-quality microbial genomes. Among the 181 genomes with the strongest evidence of being ancient and of human gut origin, 39% represent previously undescribed species-level genome bins. Tip dating suggests an approximate diversification timeline for the key human symbiont Methanobrevibacter smithii. In comparison to 789 present-day human gut microbiome samples from eight countries, the palaeofaeces samples are more similar to non-industrialized than industrialized human gut microbiomes. Functional profiling of the palaeofaeces samples reveals a markedly lower abundance of antibiotic-resistance and mucin-degrading genes, as well as enrichment of mobile genetic elements relative to industrial gut microbiomes. This study facilitates the discovery and characterization of previously undescribed gut microorganisms from ancient microbiomes and the investigation of the evolutionary history of the human gut microbiota through genome reconstruction from palaeofaeces.

Metagenomics-Based Proficiency Test of Smoked Salmon Spiked with a Mock Community.

An inter-laboratory proficiency test was organized to assess the ability of participants to perform shotgun metagenomic sequencing of cold smoked salmon, experimentally spiked with a mock community composed of six bacteria, one parasite, one yeast, one DNA, and two RNA viruses. Each participant applied its in-house wet-lab workflow(s) to obtain the metagenomic dataset(s), which were then collected and analyzed using MG-RAST. A total of 27 datasets were analyzed. Sample pre-processing, DNA extraction protocol, library preparation kit, and sequencing platform, influenced the abundance of specific microorganisms of the mock community. Our results highlight that despite differences in wet-lab protocols, the reads corresponding to the mock community members spiked in the cold smoked salmon, were both detected and quantified in terms of relative abundance, in the metagenomic datasets, proving the suitability of shotgun metagenomic sequencing as a genomic tool to detect microorganisms belonging to different domains in the same food matrix. The implementation of standardized wet-lab protocols would highly facilitate the comparability of shotgun metagenomic sequencing dataset across laboratories and sectors. Moreover, there is a need for clearly defining a sequencing reads threshold, to consider pathogens as detected or undetected in a food sample.

Addressing Learning Needs on the Use of Metagenomics in Antimicrobial Resistance Surveillance.

One Health surveillance of antimicrobial resistance (AMR) depends on a harmonized method for detection of AMR. Metagenomics-based surveillance offers the possibility to compare resistomes within and between different target populations. Its potential to be embedded into policy in the future calls for a timely and integrated knowledge dissemination strategy. We developed a blended training (e-learning and a workshop) on the use of metagenomics in surveillance of pathogens and AMR. The objectives were to highlight the potential of metagenomics in the context of integrated surveillance, to demonstrate its applicability through hands-on training and to raise awareness to bias factors. The target participants included staff of competent authorities responsible for AMR monitoring and academic staff. The training was organized in modules covering the workflow, requirements, benefits and challenges of surveillance by metagenomics. The training had 41 participants. The face-to-face workshop was essential to understand the expectations of the participants about the transition to metagenomics-based surveillance. After revision of the e-learning, we released it as a Massive Open Online Course (MOOC), now available at https://www.coursera.org/learn/metagenomics. This course has run in more than 20 sessions, with more than 3,000 learners enrolled, from more than 120 countries. Blended learning and MOOCs are useful tools to deliver knowledge globally and across disciplines. The released MOOC can be a reference knowledge source for international players in the application of metagenomics in surveillance.

Specific gut microbiome members are associated with distinct immune markers in pediatric allogeneic hematopoietic stem cell transplantation.

BACKGROUND: Increasing evidence reveals the importance of the microbiome in health and disease and inseparable host-microbial dependencies. Host-microbe interactions are highly relevant in patients receiving allogeneic hematopoietic stem cell transplantation (HSCT), i.e., a replacement of the cellular components of the patients' immune system with that of a foreign donor. HSCT is employed as curative immunotherapy for a number of non-malignant and malignant hematologic conditions, including cancers such as acute lymphoblastic leukemia. The procedure can be accompanied by severe side effects such as infections, acute graft-versus-host disease (aGvHD), and death. Here, we performed a longitudinal analysis of immunological markers, immune reconstitution and gut microbiota composition in relation to clinical outcomes in children undergoing HSCT. Such an analysis could reveal biomarkers, e.g., at the time point prior to HSCT, that in the future could be used to predict which patients are of high risk in relation to side effects and clinical outcomes and guide treatment strategies accordingly. RESULTS: In two multivariate analyses (sparse partial least squares regression and canonical correspondence analysis), we identified three consistent clusters: (1) high concentrations of the antimicrobial peptide human beta-defensin 2 (hBD2) prior to the transplantation in patients with high abundances of Lactobacillaceae, who later developed moderate or severe aGvHD and exhibited high mortality. (2) Rapid reconstitution of NK and B cells in patients with high abundances of obligate anaerobes such as Ruminococcaceae, who developed no or mild aGvHD and exhibited low mortality. (3) High inflammation, indicated by high levels of C-reactive protein, in patients with high abundances of facultative anaerobic bacteria such as Enterobacteriaceae. Furthermore, we observed that antibiotic treatment influenced the bacterial community state. CONCLUSIONS: We identify multivariate associations between specific microbial taxa, host immune markers, immune cell reconstitution, and clinical outcomes in relation to HSCT. Our findings encourage further investigations into establishing longitudinal surveillance of the intestinal microbiome and relevant immune markers, such as hBD2, in HSCT patients. Profiling of the microbiome may prove useful as a prognostic tool that could help identify patients at risk of poor immune reconstitution and adverse outcomes, such as aGvHD and death, upon HSCT, providing actionable information in guiding precision medicine.

Author Correction: Abundance and diversity of the faecal resistome in slaughter pigs and broilers in nine European countries.

In the version of this Article originally published, the surname of author Oksana Lukjancenko was spelt incorrectly as 'Lukjacenko'. This has now been corrected.

Abundance and diversity of the faecal resistome in slaughter pigs and broilers in nine European countries.

Antimicrobial resistance (AMR) in bacteria and associated human morbidity and mortality is increasing. The use of antimicrobials in livestock selects for AMR that can subsequently be transferred to humans. This flow of AMR between reservoirs demands surveillance in livestock and in humans. We quantified and characterized the acquired resistance gene pools (resistomes) of 181 pig and 178 poultry farms from nine European countries, sequencing more than 5,000 Gb of DNA using shotgun metagenomics. We quantified acquired AMR using the ResFinder database and a second database constructed for this study, consisting of AMR genes identified through screening environmental DNA. The pig and poultry resistomes were very different in abundance and composition. There was a significant country effect on the resistomes, more so in pigs than in poultry. We found higher AMR loads in pigs, whereas poultry resistomes were more diverse. We detected several recently described, critical AMR genes, including mcr-1 and optrA, the abundance of which differed both between host species and between countries. We found that the total acquired AMR level was associated with the overall country-specific antimicrobial usage in livestock and that countries with comparable usage patterns had similar resistomes. However, functionally determined AMR genes were not associated with total drug use.

Genomics-Based Identification of Microorganisms in Human Ocular Body Fluid.

Advances in genomics have the potential to revolutionize clinical diagnostics. Here, we examine the microbiome of vitreous (intraocular body fluid) from patients who developed endophthalmitis following cataract surgery or intravitreal injection. Endophthalmitis is an inflammation of the intraocular cavity and can lead to a permanent loss of vision. As controls, we included vitreous from endophthalmitis-negative patients, balanced salt solution used during vitrectomy and DNA extraction blanks. We compared two DNA isolation procedures and found that an ultraclean production of reagents appeared to reduce background DNA in these low microbial biomass samples. We created a curated microbial genome database (>5700 genomes) and designed a metagenomics workflow with filtering steps to reduce DNA sequences originating from: (i) human hosts, (ii) ambiguousness/contaminants in public microbial reference genomes and (iii) the environment. Our metagenomic read classification revealed in nearly all cases the same microorganism that was determined in cultivation- and mass spectrometry-based analyses. For some patients, we identified the sequence type of the microorganism and antibiotic resistance genes through analyses of whole genome sequence (WGS) assemblies of isolates and metagenomic assemblies. Together, we conclude that genomics-based analyses of human ocular body fluid specimens can provide actionable information relevant to infectious disease management.

Draft Genome Sequence of Acinetobacter johnsonii C6, an Environmental Isolate Engaging in Interspecific Metabolic Interactions.

Acinetobacter johnsonii C6 originates from creosote-polluted groundwater and performs ecological and evolutionary interactions with Pseudomonas putida in biofilms. The draft genome of A. johnsonii C6 is 3.7 Mbp and was shaped by mobile genetic elements. It reveals genes facilitating the biodegradation of aromatic hydrocarbons and resistance to antimicrobials and metals.

A sampling and metagenomic sequencing-based methodology for monitoring antimicrobial resistance in swine herds.

OBJECTIVES: Reliable methods for monitoring antimicrobial resistance (AMR) in livestock and other reservoirs are essential to understand the trends, transmission and importance of agricultural resistance. Quantification of AMR is mostly done using culture-based techniques, but metagenomic read mapping shows promise for quantitative resistance monitoring. METHODS: We evaluated the ability of: (i) MIC determination for Escherichia coli; (ii) cfu counting of E. coli; (iii) cfu counting of aerobic bacteria; and (iv) metagenomic shotgun sequencing to predict expected tetracycline resistance based on known antimicrobial consumption in 10 Danish integrated slaughter pig herds. In addition, we evaluated whether fresh or manure floor samples constitute suitable proxies for intestinal sampling, using cfu counting, qPCR and metagenomic shotgun sequencing. RESULTS: Metagenomic read-mapping outperformed cultivation-based techniques in terms of predicting expected tetracycline resistance based on antimicrobial consumption. Our metagenomic approach had sufficient resolution to detect antimicrobial-induced changes to individual resistance gene abundances. Pen floor manure samples were found to represent rectal samples well when analysed using metagenomics, as they contain the same DNA with the exception of a few contaminating taxa that proliferate in the extraintestinal environment. CONCLUSIONS: We present a workflow, from sampling to interpretation, showing how resistance monitoring can be carried out in swine herds using a metagenomic approach. We propose metagenomic sequencing should be part of routine livestock resistance monitoring programmes and potentially of integrated One Health monitoring in all reservoirs.

Comparative genomics of toxigenic and non-toxigenic Staphylococcus hyicus.

The most common causative agent of exudative epidermitis (EE) in pigs is Staphylococcus hyicus. S. hyicus can be grouped into toxigenic and non-toxigenic strains based on their ability to cause EE in pigs and specific virulence genes have been identified. A genome wide comparison between non-toxigenic and toxigenic strains has never been performed. In this study, we sequenced eleven toxigenic and six non-toxigenic S. hyicus strains and performed comparative genomic and phylogenetic analysis. Our analyses revealed two genomic regions encoding genes that were predominantly found in toxigenic strains and are predicted to encode for virulence determinants for EE. All toxigenic strains encoded for one of the exfoliative toxins ExhA, ExhB, ExhC, or ExhD. In addition, one of these regions encoded for an ADP-ribosyltransferase (EDIN, epidermal cell differentiation inhibitor) and a novel putative RNase toxin (polymorphic toxin) and was associated with the gene encoding ExhA. A clear differentiation between toxigenic and non-toxigenic strains based on genomic and phylogenetic analyses was not apparent. The results of this study support the observation that exfoliative toxins of S. hyicus and S. aureus are located on genetic elements such as pathogenicity islands, phages, prophages and plasmids.

The metabolically active subpopulation in Pseudomonas aeruginosa biofilms survives exposure to membrane-targeting antimicrobials via distinct molecular mechanisms.

Biofilms are reported to be inherently refractory toward antimicrobial attack and, therefore, cause problems in industrial and medical settings. Pseudomonas aeruginosa biofilms contain subpopulations that exhibit high metabolic activity and subpopulations that exhibit low metabolic activity. We have found that membrane-targeting antimicrobials such as colistin, EDTA, SDS, and chlorhexidine specifically kill the inactive subpopulation in P. aeruginosa biofilms, whereas the active subpopulation survives exposure to these compounds. Because treatment of P. aeruginosa biofilms with the membrane-targeting compounds colistin, EDTA, SDS, and chlorhexidine resulted in the same spatial distribution of live and dead bacteria, we investigated whether tolerance to these compounds originated from the same molecular mechanisms. Development of colistin-tolerant subpopulations was found to depend on the pmr genes encoding lipopolysaccharide modification enzymes, as well as on the mexAB-oprM, mexCD-oprJ, and muxABC-opmB genes encoding antimicrobial efflux pumps, but does not depend on the mexPQ-opmE efflux pump genes. Development of chlorhexidine-tolerant subpopulations was found to depend on the mexCD-oprJ genes, but does not depend on the pmr, mexAB-oprM, mexPQ-opmE, or muxABC-opmB genes. Tolerance to SDS and EDTA in P. aeruginosa biofilms is linked to metabolically active cells, but does not depend on the pmr, mexAB, mexCD, mexPQ, or muxABC genes. Our data suggest that the active subpopulation in P. aeruginosa biofilms is able to adapt to exposure to membrane-targeting agents through the use of different genetic determinants, dependent on the specific membrane-targeting compound.

Inactivation of the rhlA gene in Pseudomonas aeruginosa prevents rhamnolipid production, disabling the protection against polymorphonuclear leukocytes.

Many of the virulence factors produced by the opportunistic human pathogen Pseudomonas aeruginosa are quorum-sensing (QS) regulated. Among these are rhamnolipids, which have been shown to cause lysis of several cellular components of the human immune system, e.g. monocyte-derived macrophages and polymorphonuclear leukocytes (PMNs). We have previously shown that rhamnolipids produced by P. aeruginosa cause necrotic death of PMNs in vitro. This raises the possibility that rhamnolipids may function as a 'biofilm shield'in vivo, which contributes significantly to the increased tolerance of P. aeruginosa biofilms to PMNs. In the present study, we demonstrate the importance of the production of rhamnolipids in the establishment and persistence of P. aeruginosa infections, using an in vitro biofilm system, an intraperitoneal foreign-body model and a pulmonary model of P. aeruginosa infections in mice. Our experimental data showed that a P. aeruginosa strain, unable to produce any detectable rhamnolipids due to an inactivating mutation in the single QS-controlled rhlA gene, did not induce necrosis of PMNs in vitro and exhibited increased clearance compared with its wild-type counterpart in vivo. Conclusively, the results support our model that rhamnolipids are key protective agents of P. aeruginosa against PMNs.

Insight into the microbial multicellular lifestyle via flow-cell technology and confocal microscopy.

Biofilms are agglomerates of microorganisms surrounded by a self-produced extracellular matrix. During the last 10 years, there has been an increasing recognition of biofilms as a highly significant topic in microbiology with relevance for a variety of areas in our society including the environment, industry, and human health. Accordingly a number of biofilm model systems, molecular tools, microscopic techniques, and image analysis programs have been employed for the study of biofilms under controlled and reproducible conditions. Studies using confocal laser scanning microscopy (CLSM) of biofilms formed in flow-chamber experimental systems by genetically color-coded bacteria have provided detailed knowledge about biofilm developmental processes, cell differentiations, spatial organization, and function of laboratory-grown biofilms, in some cases down to the single cell level. In addition, the molecular mechanisms underlying the increased tolerance that biofilm cells often display towards antibiotic treatment are beginning to be unravelled.

Tolerance to the antimicrobial peptide colistin in Pseudomonas aeruginosa biofilms is linked to metabolically active cells, and depends on the pmr and mexAB-oprM genes.

Bacteria living as biofilm are frequently reported to exhibit inherent tolerance to antimicrobial compounds, and might therefore contribute to the persistence of infections. Antimicrobial peptides are attracting increasing interest as new potential antimicrobial therapeutics; however, little is known about potential mechanisms, which might contribute to resistance or tolerance development towards these compounds in biofilms. Here we provide evidence that a spatially distinct subpopulation of metabolically active cells in Pseudomonas aeruginosa biofilms is able to develop tolerance to the antimicrobial peptide colistin. On the contrary, biofilm cells exhibiting low metabolic activity were killed by colistin. We demonstrate that the subpopulation of metabolically active cells is able to adapt to colistin by inducing a specific adaptation mechanism mediated by the pmr operon, as well as an unspecific adaptation mechanism mediated by the mexAB-oprM genes. Mutants defective in either pmr-mediated lipopolysaccharide modification or in mexAB-oprM-mediated antimicrobial efflux were not able to develop a tolerant subpopulation in biofilms. In contrast to the observed pattern of colistin-mediated killing in biofilms, conventional antimicrobial compounds such as ciprofloxacin and tetracycline were found to specifically kill the subpopulation of metabolically active biofilm cells, whereas the subpopulation exhibiting low metabolic activity survived the treatment. Consequently, targeting the two physiologically distinct subpopulations by combined antimicrobial treatment with either ciprofloxacin and colistin or tetracycline and colistin almost completely eradicated all biofilm cells.