LIVIA fmb%2E11%2E135 (1).pdf

8/10/2019 LIVIA fmb%2E11%2E135 (1).pdf

1/17

73Future Microbiol. (2012) 7(1), 7389

F ut ur eM

i cr o b i ol o

g y

part of

10.2217/FMB.11.135 R Edwards ISSN 1746-0913

Infectious diseases are the second-leading causeof death globally and the most signicant causeof death in children [1,2] . Antibiotics representone of the largest therapeutic categories used inthe treatment of infectious diseases caused bybacteria, but the successful use of any therapeuticagent is compromised by the potential develop-ment of tolerance or resistance to that compoundfrom the time it is rst employed. In clinicalenvironments, pathogenic and commensal bac-teria are challenged with high concentrations ofantibiotics and bacteria have become resistantto most of the antibiotics developed[3]. Withinthe hospital setting, resistance pathogens oftenemerge within a few years after a new antibioticis introduced (F IGURE 1) . The spread of resistantbacteria and resistance genes depends on differ-ent factors but the major pressure is antibioticusage [4] . Nosocomial (hospital-linked) infec-tions result in approximately 100,000 deathsand cost more than US$25 billion per year inthe USA alone. Worldwide, it is estimated that510% of patients entering hospitals develop aninfection as a result of their stay[101] . Despitethe problem of antibiotic resistance in infec-tious bacteria, little is known regarding thediversity, distribution and origins of resistancegenes, especially for the unculturable majorityof environmental bacteria.

Mechanisms of antibiotic resistanceResistance to antibiotics can be caused by four

general mechanisms(F IGURE 2) [5]:

n The inactivation or modification of theantibiotic;

n An alteration in the target site of the antibioticthat reduces its binding capacity;

n The modication of metabolic pathways tocircumvent the antibiotic effect;

n

The reduced intracellular antibiotic accumula-tion by decreasing permeability and/orincreasing active efux of the antibiotic.

Bacteria can develop resistance to antibiot-ics by mutating existing genes (vertical evolu-tion) [6,7] , or by acquiring new genes from otherstrains or species (horizontal gene transfer)[8,9] .The sharing of genes between bacteria by hori-zontal gene transfer occurs by many differentmechanisms. Mobile genetic elements, includingphages, plasmids and transposons mediate thistransfer, and in some circumstances the presenceof low levels of the antibiotic in the environmentis the key signal that promotes gene transfer[10] , perhaps ensuring that the whole microbialcommunity is protected from the antibiotic[11] .

Detection of antibiotic resistance Antibiotic resistance is a highly selectable phe-notype and can be detected using growth inhibi-tion assays performed in broth or by agar discdiffusion. In a dilution-based growth inhibitionassay, the MIC of an antibiotic can be calcu-lated for each bacterial isolate and the organ-

ism is typically interpreted as being susceptible

Insights into antibiotic resistancethrough metagenomicapproaches

Robert Schmieder & Robert Edwards*Computational Science Research Center & Department of Computer Science, San Diego State University,San Diego, CA 92182, USA*Author for correspondence:Department of Computer Science, San Diego State University, San Diego, CA92182, USAMathematics and Computer Science Division, Argonne National Laboratory, 9700 South Cass Ave, Argonne,IL 60439, USA Tel.: +1 619 594 1672 n Fax: +1 619 594 6746 n [email protected]

The consequences of bacterial infections have been curtailed by the introduction ofa wide range of antibiotics. However, infections continue to be a leading cause ofmortality, in part due to the evolution and acquisition of antibiotic-resistance genes.Antibiotic misuse and overprescription have created a driving force inuencing theselection of resistance. Despite the problem of antibiotic resistance in infectious bacteria,little is known about the diversity, distribution and origins of resistance genes, especiallyfor the unculturable majority of environmental bacteria. Functional and sequence-

based metagenomics have been used for the discovery of novel resistance determinantsand the improved understanding of antibiotic-resistance mechanisms in clinical andnatural environments. This review discusses recent ndings and future challenges in thestudy of antibiotic resistance through metagenomic approaches.

Keywords

n aminoglycosiden amphenicol n antibioticresistance n b -lactamn functional metagenomicsn metagenome n sequence-based metagenomicsn tetracycline

R evi ew

8/10/2019 LIVIA fmb%2E11%2E135 (1).pdf

2/17

Future Microbiol. (2012) 7(1)74 future science group

or resistant to the antibiotic. Of course, thereis a gradation of resistance and some classica-tion schemes include one or more intermediatelevels. For clinically important bacteria, diag-nostic laboratories perform phenotypic-basedanalyses using standardized susceptibility testingmethods, usually in accordance with those pub-lished by the Clinical and Laboratory StandardsInstitute [102] .

To produce results, culture-based approachescan take 12 days for fast-growing bacteria,such as Escherichia coli or Salmonella, but sev-eral weeks for slow-growing bacteria, such as

Mycobacterium tuberculosis . Moreover, culturingonly works for a fraction of microbes; althoughmost pathogens can be cultured because ofour years of experimental experience, the vastmajority of microbes cannot grow outside theirhost environment, including pathogens such asChlamydia or Trypanosomes . Newer moleculardetection techniques for resistance such as quan-titative PCR (qPCR) [12] or microarrays [13,14]are able to determine the presence of specicresistance genes and have improved diagnosis by

providing results within hours. However, these

culture-independent approaches only target well-studied pathogens or resistance-causing genesand cannot easily be used for broad-spectrumscreening.

Research targeted towards antibiotics fortreating infectious diseases and the risks to

human health posed by antibiotic resistancehave focused mainly on the clinical setting[15,16] .To fully understand the development and dis-semination of resistance, we need to address thestudy of antibiotics and their resistance genesnot just in clinics but in natural (nonclinical)environments as well[1719] .

Before next-generation sequencing, antibioticresistance genes were typically isolated fromenvironmental samples by cloning from culturedbacteria or by PCR amplication[20] . Thosemethods ignored potential antibiotic resistance

reservoirs because most bacteria are not cultur-able, and PCR detection depends on primersthat are based on known resistance genes anddoes not readily allow for the discovery of novelgenes. The development of culture-independenttechniques was required to identify novel resis-tance genes and access the genetic diversity ofmost bacteria.

Metagenomics is one of the more modernapproaches that overcome the limitations ofmethods based on culturing or amplication(B OX 1) [21] . This approach is a powerful tool todescribe the genetic potential of a communityand to identify the types of microbes present ina community, as well as the presence or absenceof genes or genetic variations responsible for anti-biotic resistance. Using metagenomics, severalnovel antibiotic resistance genes have been iden-tied, including resistance tob -lactams [22,23] ,tetracycline[24] , aminoglycosides[20,23] and bleo-mycin [25] . In the next sections, we describe thedifferent metagenomic approaches that are usedto identify antibiotic-resistance genes.

Culture-independent study of resistance

through metagenomicsIn 1985, Pace et al . were the rst to proposethe direct cloning of environmental DNA toclassify unculturable microorganisms[26] . Therst successful function-driven screening ofmetagenomic libraries, termed zoolibraries bythe authors, was conducted in 1995 [27] . Theterm metagenomics was coined in 1998 byHandelsman et al ., referring to the function-based analysis of mixed environmental DNAspecies [21] . Initially, metagenomics was usedmainly to recover novel biomolecules (espe-

cially DNA) from environmental microbial

Penicillin

Streptomycin

Tetracycline

Erythromycin

Vancomycin

MethicillinAmpicillin

Linezolid

Daptomycin

Chloramphenicol

Aminoglycosides

Macrolides

Glycopeptides

-lactams

Oxazolidinones

Lipopeptides

Kanamycin

Nalidixic acid

Streptogramins Synercid

Norfloxacin

Amphenicols

Quinolones

Tetracyclines

Class Antibiotic Years from introduction to clinical resistance

5 10 15 20 25 30

8 years average

0

Augmentin

Figure 1. Antibiotic resistance evolution showing the rapid development ofresistance for several classes of antibiotics. The bars mark the time from theintroduction of an antibiotic to the clinic until the rst clinical case of resistance tothat antibiotic was reported. The high number of available antibiotics during the1950s and 1960s attributed to the longer times until the rst resistance case wasreported (e.g., for nalidixic acid). If the time frames for the same antibiotic differedbetween the data sources, the shorter time frame was used.Data taken from [9497] .

Review Schmieder & Edwards

8/10/2019 LIVIA fmb%2E11%2E135 (1).pdf

3/17

www.futuremedicine.com 75future science group

assemblages. The development of next-genera-tion sequencing techniques led to an alternativeapproach where a fraction of the DNA in thesample was sequenceden masse , without regardto cloning. This approach was sometimes calledrandom community genomics, which also

became known as metagenomics. Metagenomicsequencing represented a powerful alternative torRNA sequencing for analyzing complex micro-bial communities [20,28] and has a tremendousimpact on the study of microbial diversity inenvironmental and clinical samples. Nowadays,the field of metagenomics can roughly bedivided into two different approaches: func-tional metagenomics and sequence-basedmetagenomics(F IGURE 3) .

Functional metagenomics

Functional metagenomics involves cloning andheterologous expression of environmental DNAin a surrogate host with coupled activity-basedscreening to discover functions of genes thatmight not be obvious from their sequence. Bycreating a functional metagenomic library in which cloned genomic fragments are expressedand selecting directly for resistance to antibiotics,traditional challenges associated with studyinggenes of unknown sequence were circumvented.The metagenomic analysis revealed novel anti-biotic resistance proteins that were previouslyof unknown function and unrecognizable bysequence alone. Functional metagenomics hasbeen used to identify genes encoding proteinsthat inactivate antibiotics, genes encoding mul-tidrug efux pumps and genes conferring resis-tance to the folate antagonist trimethoprim.Functional metagenomics has additionally beenapplied to cultured isolates of a community[29] .

Sequence-based metagenomicsSequence-based metagenomics involves extract-ing and random sequencing of DNA directlyfrom the environment, including the DNA ofuncultured bacteria. Typically, the eukaryoticcells, bacteria, viruses, and free DNA are sepa-rated by size (using ltration or centrifugation),and total DNA extracted from the appropriatefraction. A sample of the DNA is sequenced,and that sample is assumed to be a random frac-tion of the whole community. The metagenomicsequences are then compared to the knownsequences that have been accumulated over theyears in national and international databanks(reference sequences) to identify resistancegenes and/or mutations that are known to cause

resistance (F IGURE 3 & 4) . Using a wide range of

reference resistance genes, the resistance poten-tial for multiple antibiotics can be predictedfrom a single metagenome. The metagenomicsequences additionally represent the diversity ofthe community, including strains that cannotbe cultured, valuable information for the studyof community changes as a result of antibiotictreatment.

To date, more than one thousand differentmetagenomes have been sequenced from a largevariety of environments, such as soil, oceanand human gut (F IGURE 5 ) . In addition, extinctspecies such as the woolly mammoth[30] andthe Neanderthals [31] have been analyzed bysequence-based metagenomic approaches.

The volume of sequence data generated inthe last few years spawned a new generation of

X

Decreased influx

Increased efflux

Aminoglycosides-lactamsFluoroquinolonesGlycopeptidesMacrolidesRifamycinsTetracyclines

AminoglycosidesAmphenicolsAntifolates-lactamsGlycopeptidesRifamycins

Aminoglycosides-lactamsMacrolidesQuinolonesTetracyclines

Aminoglycosides-lactams

Antibiotic

Target site alteration

Target amplification

SulfonamidesTrimethoprim

Antibiotic inactivation

Figure 2. Mechanisms of antibiotic resistance in bacteria. The classes ofantibiotics affected by each of the mechanisms are listed in the boxes.

Box 1. n The generic term antibiotic is used to denote any class of organic molecules that

inhibits or kills microbes by specic interactions with bacterial targets, withoutany consideration of the source of the particular compound or class

n The term antibiotic resistome was proposed for the collection of all antibioticresistance genes in microorganisms, from both pathogenic and nonpathogenicbacteria

n Metagenome describes the genetic potential of a community n Sequence-based metagenomics involves extracting and random sequencing of

DNA directly from the environment without the need for culturing n Functional metagenomics involves cloning and heterologous expression of

environmental DNA in a surrogate host with coupled activity-based screening n The term microbiome has been suggested by Nobel laureate Joshua Lederberg

to describe the collective genome of our indigenous microbes n Metatranscriptomics is a culture-independent method by which the collective

RNA transcript of an entire community of microorganisms is analyzed

Insights into antibiotic resistance through metagenomic approaches Review

8/10/2019 LIVIA fmb%2E11%2E135 (1).pdf

4/17

8/10/2019 LIVIA fmb%2E11%2E135 (1).pdf

5/17


In recent years, large-scale projects have beenstarted as part of the International HumanMicrobiome Consortium to investigate thehuman microbiota. The Human MicrobiomeProject (HMP), an initiative of the NIH, waslaunched in 2008 with the mission to sequence,

characterize, and analyze as much of the humanmicrobiota as possible in 5 years[39,40] . Part ofthe HMP is to learn about microbes throughmetagenomics by sampling from ve body sites:the mouth, nasal cavity, skin, gut and (in women)vagina of healthy donors. The HMP also funded15 demonstration projects for 1 year to look at themicrobiome in correlation to human health andawarded continued funding to eight of the proj-ects. In addition to the HMP, the Metagenomicsof the Human Intestinal Tract (MetaHIT) proj-ect [41] applies metagenomics to the study of the

human intestinal tract of European and Asianpopulations. A recent MetaHIT metagenomic study of

124 individuals suggests the existence of a com-mon core human gut microbiome [41] , but thiscore may exist more at the level of shared func-tional genes rather than shared taxa[42] . An openquestion in microbial ecology is whether selec-tion operates at the level of taxa or the level ofthe functions that those taxa perform. Despite the

core microbiome, there is a subject-to-subject vari-ability in the composition of the gut microbiotaamong humans. Numerous internal and externalfactors, including diet, geography, host physiol-ogy, disease state, and features of the gut itselfcontribute to the community composition of the

gut microbiota, and the HMP and related stud-ies will signicantly extend our understanding ofthese relationships[36,43,44] .

Humans have directly exposed their bodies toantibiotics in medicines and indirectly throughagriculture and cleaning or beauty products. Onaverage, an adult in the USA makes two outpa-tient visits per year and 15.3% of these visits resultin the prescription of an antibiotic[45] . Antibioticsprescribed during clinical visits may have long-term consequences on the human microbiota;community level studies demonstrated that

the gut microbiota recovery is incomplete afterrepeated antibiotic perturbation and that treat-ment might cause a shift to a different, but stablecommunity [46] .

Antibiotic-resistant strains persist in the humanhost environment in the absence of selective pres-sure [4,47] . For example, Sommeret al . character-ized the functional resistance reservoir in twounrelated healthy individuals who had not beentreated with antibiotics for at least 1 year[48] . The

Raw metagenomic sequences Example program

Filter unwanted sequences

De novo assembly

Identify sequences similar to reference genes

Reference assembly

(adapters, tags, low quality ends, ...)

(low quality, duplicates, contamination, ...)

Resistance gene abundance Identify resistance mutations

Resistance pathways

Resistance potential

PRINSEQFASTX-ToolkitDeconSeq

Newbler, SOAPdenovo

BLAST, BWA

KEGG

ARDB, NR

Newbler, SOAP2

Gene abundance

Cleaned metagenomic sequences

Trim sequences

P r e p r o c e s s i n g

P o s t - p r o c e s s i n g

Database

Figure 4. A proposed bioinformatics methodology to identify antibiotic resistance fromsequence-based metagenomes. The methodology includes preprocessing steps, where a numberof freely available tools can be applied, an analytical step where the sample is compared to databasessuch as the NCBI NR or the ARDB, and post-processing steps to identify the resistance potential. Weencourage the reader to consult online resources, such as SEQanswers [105] and BioStar [106] , forup-to-date information of current programs and databases.ARDB: Antibiotic resistance database; NR: Nonredundant protein database.


8/10/2019 LIVIA fmb%2E11%2E135 (1).pdf

6/17


subjects microbiota was analyzed using functionalscreening of metagenomic DNA and genomicDNA from 572 bacterial isolates. The sequencingof inserted clones conferring resistance to 13 dif-ferent antibiotics revealed 95 unique inserts that were evolutionarily distant from known resistancegenes. This diverse gene pool of resistance genes inthe commensal microbiota of healthy individualscould potentially lead to the emergence of newresistant pathogenic strains. It may also suggestthat barriers exist to lateral gene transfer betweenthe bacteria with the novel resistance genes andreadily cultured human pathogens[49] . We donot know where the resistance mechanisms comefrom, but they are easily acquired. For example, astudy with a newborn showed that methicilin andouroquinolone resistance genes and multidrugefux pumps could be found in the meconium3 days after birth[50] . At day 6, teicoplanin resis-tance genes were found in the infant fecal sample.In this particular individual, a fever at day 92 wasfollowed by the presence ofb -lactamase genes anda reduction in bacterial load in the stool, eventhough antibiotics were not reported to be used

at that time.

Similar to the human gut microbiata, animportant attribute of the oral microbiota is itsability to act as a reservoir of antibiotic resistantorganisms. Oral bacteria can easily reach otherbody sites by swallowing or via the bloodstreamand spread to other individuals by, for exam-ple, coughing and kissing. Therefore, resistantoral bacteria have the opportunity for rapiddissemination.

A novel tetracycline resistance determinant,tet37 , was identied by Diaz-Torreset al . whoconstructed metagenomic libraries from themicrobiota of the human oral cavity[24] , andthen 3 years later, they applied functionalmetagenomics to identify genes encoding antibi-otic resistance in the oral microbiota of 60 adulthumans [51] . Clones resistant to tetracycline andamoxycillin were detected in each library, andthe former was the most common resistancein each library. They did not report on theirpatients antibiotic usage, and so it is not knownhow that affected their experiments.

Our oral and fecal microbiota are distinctlydifferent, both in the organisms present and the

resistance mechanisms found in those organisms.

0.01

0.10

1.00

10.00

100.00

1,000.00

10,000.00

350

700

1050

1400

1750

2100

C o s

t p e r

M b ( $ )

N um

b er

of m

e t a g en

omi cl i b r ar i e

si n

S RA

MetaHIT

HMP

First submission to SRA

MPSP

HMP launched

First Roche/454sequencer installed

Illumina/Solexa systemcommercialized

SOLiD system launched

Sanger sequencing

M o o r e s La w

Next-generation sequencing

2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

Roche/454Illumina/SolexaABI SOLiDCost per Mb

Figure 5. Overview of sequencing cost and number of metagenomic libraries submitted to the international Sequence ReadArchive. The cost of sequencing is based on data provided by the US National Human Genome Research Institute [107] , illustrating themore than logarithmic decrease and sudden change when the sequencing centers transitioned to next-generation sequencingtechnologies. The cost of sequencing is compared to Moores Law, which describes a long-term trend in the computer hardware industrythat involves the doubling of computing power every 2 years (here halving of sequencing cost every 2 years). The stacked bars show thenumber of metagenomic libraries that were submitted to the Sequence Read Archive (as of 1 July 2011) with a specied study type ofMetagenomics and a library strategy of WGS (whole genome shotgun sequencing). The sequencing technologies used to generate themetagenomic libraries are marked as shown in the legend. As of July 2011, only one metagenomic library in this category was generatedusing the SOLiD system. Major milestones in next-generation sequencing and major submissions (>100 libraries) are highlighted.HMP: Human Microbiome Project; MetaHIT: Metagenomics of the Human Intestinal Tract; MPSP: Marine Phage Sequencing Project;SRA: Sequence Read Archive.


8/10/2019 LIVIA fmb%2E11%2E135 (1).pdf

7/17


Sevilleet al . investigated the presence and preva-lence of various tetracycline and erythromycinresistance genes in the metagenomic DNA iso-lated from human oral and fecal samples from sixdifferent European countries[52] . Although thesubjects did not receive antibiotics for 3 months,

the prole of resistance genes detected in the fecalmetagenome differed considerably from thosedetected in the oral samples, again suggestingthat exposure to antibiotics has long term affectson the microbiota.

The fate of antibiotics used in human medi-cine is largely unknown. After secretion fromthe human, some antibiotics are broken down byresistance, some are light, or temperature labile, while others may exist for a long time in the envi-ronment. As we shall see later, antibiotics in theenvironment may promote the development of

novel mechanisms of antibiotic resistance.Antibiotic resistance in soil microbialcommunities

There are as many microorganisms in 1g of soilas there are humans on the entire planet[53] , andmuch less than 1% of those microbes are read-ily culturable by current methods[54,55] . Soil ismost likely the ecosystem where antibiotic syn-thesis originally evolved[56] and soils from diverselocations around the world such as forest, prai-rie, agriculture and urban have been explored todevelop new clinical and medicinal applications. As a result, over 80% of antibiotics in clinical usetoday originated from soil bacteria, either directlyas natural products or as their semi-syntheticderivatives[57] .

The high density of antibiotic-producingbacteria makes soil a likely source of diverseantibiotic resistance determinants [20,23,58,59] .Molecular resistance mechanisms between clini-cal pathogens and the common soil bacteriumStreptomyces were rst shown to be similar in1973 [60] . Since then, numerous parallels havebeen identied between soil microorganisms andclinically important strains, and the abundanceof pathogens that can survive in soil results in apotent mixture that can give rise to the emergenceof antibiotic resistance in the clinical setting.In recent years, metagenomic approaches havebeen implemented to characterize the diversityand prevalence of resistance in soil bacteria[61] .

The metagenomic studies of soil environmentsshow evidence that multiple diverse mechanismsfor resistance are associated with microbes fromdifferent soil samples(T ABL E 1) . These studiesalso suggested that soil microbes are resistant to

most antibiotics, and the broad activity of these

resistance genes might afford protection againstnewly developed antibiotics.

It is assumed that many soil bacteria are natu-rally resistant to antibiotics such asb -lactams[22,62] . Allen et al . identied genes that medi-ate resistance to b -lactam antibiotics from

remote Alaskan soil with no known exposure toanthropogenically derived antibiotics[22] . Theyperformed functional metagenomics to allowcomparison of antibiotic resistance betweenthe Alaskan soil and soils subjected to humanactivity, and to search for orfenicol resistancegenes[59] . Torres-Corts et al . applied functionalmetagenomics to soil samples and identied agene belonging to a new type of reductase con-ferring resistance to trimethoprim, a syntheticantibiotic that interferes with the production oftetrahydrofolic acid[63] .

Antibiotic resistance is common in soil bacte-ria. Nine clones expressing antibiotic resistancein E. coli to ve different aminoglycoside anti-biotics and one clone expressing tetracyclineresistance were identied in a metagenomiclibrary from soil samples[20] . These resistanceproteins were less than 60% identical to previ-ously published sequences, suggesting that soilmicroorganisms harbor more genetic diversitythan previously assumed[63] . More recently,446,000 clones comprising a soil-derivedmetagenomic library were screened for genesconferring resistance to tetracycline,b -lactams,or aminoglycoside antibiotics, which resultedin the identication of 13 different antibiotic-resistant clones[23] . The low level of sequencesimilarity from these clones suggests that theseproteins may play a different role in naturalenvironments. For example, it was proposed thatantibiotics could, in reality, be signal moleculesthat help shape the structure of microbial com-munities [6466] . Alternatively, resistance maybe associated with adaption to, and survival in,nutrient-poor environments, and some bacteriacan use antibiotics as a food source, presumablyvia resistance mechanisms they have developed[67] . Further studies on a more diverse subset ofstrains, especially slow-growing strains and thosedifcult to culture will be important to clarifythe role of resistance in bacteria.

Antibiotic resistance associated withwastewater treatment plants

Wastewater treatment plants are interfacesbetween different environments and, therefore,provide an opportunity for mobile elements(including resistance) to mix between pathogens,

opportunistic pathogens, and environmental


8/10/2019 LIVIA fmb%2E11%2E135 (1).pdf

8/17


bacteria [68] . The presence of antibiotics in sew-age selects for resistance markers that are able tospread through the microbial community and as aresult, antibiotic-resistant bacteria can potentiallydisseminate their resistance genes widely amongmembers of the endogenous microbial community(F IGURE 6) . The sludge products of urban and rural wastewater treatment plants are increasingly usedto fertilize agricultural crops, dispersing unknownamounts of resistance genes and antibiotics that

withstand standard sewage treatment.

The level of antibiotic resistance in activatedsludge is not known, and the few studies to datehave been inconclusive. The activated sludgeprocess may promote cellular interactions amongall the diverse microorganisms in the sewage,but it is also a highly competitive environment where the fraction of pathogens decreases dueto competition with other microbes or throughpredation. A function- and sequence-basedmetagenomic approach was used to identify

antibiotic resistance determinants carried on

Table 1. Summary of resistance genes identied through functional metagenomics.

Environment Number of clones screened(total library size)

Number ofresistant clones

Resistant to (predictedmechanisms)

Ref.

Soil (plano silt loam) 1,158,000(4.1 Gb)

91

Aminoglycoside (inactivation)Tetracycline (efux)

[20]

Soil (apple orchard) 446,000(13 Gb)

319

Aminoglycoside (inactivation)Tetracycline (efux)b -lactam (inactivation)

[23]

Soil (Alaskan) 714,000(12.4 Gb)

14 b -lactam (inactivation) [22]

Soil (Alaskan) -

(13.2 Gb)1 Amphenicol (efux) [59]

Soil (loamy) 550,000(3.6 Gb)

2324

Aminoglycoside (inactivation)b -lactam (inactivation)Amphenicol (efux)Antifolate (inactivation)

[63]

Activated sludge -

(1.8 Gb)

1161(1.2 Mb)

1

170

Aminoglycoside (inactivation)

b -lactam (inactivation)Amphenicol (inactivation)-

[69]

Activated sludge 96,000(3.2 Gb)

3 Glycopeptide (inactivation) [25]

Human (fecal, saliva) - #

(9.3 Gb)3562408121330191

AminoglycosideAmino acid derivativeAmphenicolb -lactamTetracycline

[48]

Human(plaque, saliva)

450 (-)

18 Tetracycline [24]

Human(plaque, saliva)

1860 (-)

143258

Aminoglycosideb -lactamTetracycline

[51]

Insect (midgut disect) 38,500(0.3 Gb)

231

Macrolide (efux)b -lactam (inactivation)Amphenicol

[29]

Organic pig (fecal) 9000(0.1 Gb)

10 Tetracycline [87]

Resistance genes could only be identied in 13 clones.Library from [22] used.Phage metagenomic library. Two of three resistant clones were assumed to be from the same region due to identical sequences.#Insert length between 1000 and 3000 bp.Insert length between 800 and 3000 bp.One clone was resistant to all three antibiotic classes in a secondary screen.


8/10/2019 LIVIA fmb%2E11%2E135 (1).pdf

9/17


bacterial chromosomes, plasmids or viruses within activated sludge(T ABLE 1) [69] . Functionalscreening of the viral metagenomic librarydid not yield any antibiotic resistant clones,and therefore, a sequence-based approach wasemployed that identied six clone inserts with

homology to known resistance genes. In con-trast to some of the studies we have discussed,the degree of similarity between the plasmid- orphage-derived antibiotic resistance determinantsand the reference sequences was greater thanthat for other genes that were tested. However,none of the sequenced clones conferred theirpredicted antibiotic resistance inE. coli . Thereare a number of reasons that the genes maynot confer resistance toE. coli , including theincorrect expression in that strain, or the incom-plete cloning of the sequence. In another study,

activated sludge used to treat industrial wastewater polluted with phenolic compounds, wasscreened for bleomycin resistance genes using ametagenomics approach and two novel genes were identied[25] .

A sequence-based metagenomic approach of wastewater using Roche/454 pyrosequencingshowed a high diversity of plasmids and resis-tance genes in that sample[70] . The metagenomicsequences representing resistance genes, espe-cially fromb -lactamase protein families, wereidentied by searches against reference databases with known antibiotic resistance genes and byscreening for particular motifs in the proteinsequences. These motifs and sequences weresimilar to enzymes that were previously identi-ed in pathogenic bacteria isolated from hospi-tal patients with diverse infections, but were nottested for activity.

High levels of antibiotic resistance genes havealso been found in bacteria that live in river sedi-ments downstream from a wastewater treatmentplant using sequence-based metagenomics[71] .The resistance genes made up almost 2% ofthe DNA samples taken. Clearly, the survival

of antibiotics and antibiotic resistance genesthrough the sewage treatment process and inthe aquatic environment merits further study,perhaps with comparisons between areas wheredifferent antibiotics are routinely used in theclinical setting.

Antibiotic resistance in marine microbialcommunities

Antibiotic resistant bacteria have been found widely in aquatic environments[72,73] . As wediscussed earlier for terrestrial environments,

resistant organisms in marine environmentscan occur following the introduction of anti-biotics from agricultural runoff or wastewatertreatment plants. Aquaculture farms are a signif-icant source of antibiotics contaminating marineenvironments, because antibiotics are added toovercome the diseases that predominate in thehigh animal density cages. There is evidence thatantibiotic resistance can also occur in marineenvironments without the addition of antibioticcontamination. For instance, the same resistancegenes found in clinical human pathogens havebeen reported among pristine ecosystems with-out a history of antibiotic contamination[74,75] .For example, a sequence-based metagenomicstudy of the microbial community associated with the coralPorites astreoides showed the pres-ence of uoroquinolone-resistance genes[76] .The authors proposed that the coral harbors

Human

Food animals WastewatertreatmentAgriculture

Aquaculture

F i s h

R un

of f

S l u d g e

c o n t a c t

Dr i nk i n

gw

a t er

& s wi mmi n

g

M a n u r

e

A

B

C

D i r e c t

& m e a t

R a i n f a l l

Soil

V e g e t a b l e s & f r u i t

River/sea/lake

h o u s e

h o l d

s

I n d u s

t r i a l a

n d

Figure 6. Potential antibiotic resistance gene dissemination. The arrows indicate possiblepoints of dissemination among different environments. Supporting metagenomic studies are markedas follows: (A) [82] , (B) [71] and (C) [48,52] .


8/10/2019 LIVIA fmb%2E11%2E135 (1).pdf

10/17


specialized microbiota that may protect thecoral from pathogens by producing antibiotics;however, there was no evidence of uoroquino-lone production in the metagenomes and there was no evidence of a human-derived source ofuoroquinolones into the environment.

Animal- & agriculture-related resistanceOnly approximately half of antibiotics producedeach year are used for human health. The otherhalf is used for industrial farm animal produc-tion. Some are used for therapeutics, but mostlythe antibiotics are added at subtherapeutic levelsto feedstuffs where they act as growth promoters[77] . The use of subtherapeutic doses of antibiot-ics is extremely controversial it helps to keepthe costs of meat and poultry in the USA low,perhaps reduces the overall number of animals

that need to be reared, reduces fat in the meat,and reduces environmental impacts, but at thecost of providing an enormous source for theevolution of antibiotic resistance elements thatmay affect human health.

Antibiotics were rst approved for the use asgrowth promoters in food animal production inthe early 1950s, and since then the extensive useof subtherapeutic levels of antibiotics has beenshown to promote antibiotic-resistance[78,79] andanimal feed has been found to contain antibi-otic resistance genes[80] . In 1986, Sweden wasthe rst country to ban the use of antibioticgrowth promoters in food animal production,and 20 years later the EU banned the feedingof antibiotics to livestock for growth promotion(although they are still used therapeutically). Inthe USA, cattle are usually treated with monen-sin, which is not directly related to antibioticsused in human therapeutics, while poultry areoften fed bacitracin and tetracycline, which areboth used in human medicine.

Because the issues surrounding the agricul-tural use of antibiotics are complex, and thespread of resistance mechanisms might poten-tially affect human health, metagenomicshas been used to try to understand the owof resistance genes associated with this sourceof antibiotics. One of the problems with agri-cultural antibiotic use is that the antibioticis not restricted solely to the animal or farm where it is applied: residues from farms cancontain antibiotic-resistance genes that maycontaminate natural environments, wastewatertreatment plants, and as we have seen, impactenvironmental microbiota [81] . The fecal com-ponent of cattle manure carries a large com-

munity of bacteria and is thus a natural vehicle

for transmission of bacteria into the environ-ment and onto other animals [82] . In particular,a variety of insects feed on cow manure, andinsect guts are environmental reservoirs of anti-biotic resistant bacteria with the potential fordissemination of resistance genes[29] . Manure

applied to agricultural elds can also con-taminate the surface of food crops, includingvegetables and fruits[83] and has led to manyof the well publicized outbreaks of food-borneillness in the USA and elsewhere. In addition,antibiotic-resistance genes can be aerosolizedand carried away from the farm; they havebeen found downwind (but not upwind) ofa large agricultural facility[84] . A sequence-based metagenomic study of the chicken cecummicrobiome showed that resistance to tetracy-clines and uoroquinolones were dominating

that environment, both classes of antibioticsthat are routinely used in poultry production[85] . A further comparison of the virulence-associated sequences showed that the chickencecum was more similar to the microbiome ofbovine rumen than to that of mouse cecumor human fecal samples[85] , perhaps not sur-prising given the physiological differences inthe intestine, rumen and cecum. Other studiessuggest that the use of antibiotics in agricul-ture may have resulted in the spread of strains,such as vancomycin-resistant enterococci inboth farm animals exposed to antibiotics andhumans in contact with the animals [78,86] , as well as the maintenance of antibiotic-resistancegenes in apparently antibiotic-free animals[87] .

Detailed discussion of the problems and chal-lenges associated with agricultural antibiotic usecan be found at the Pew Charitable Trusts web-site [103] , and at the Animal Health Institutes website[104] .

Ongoing challenges for detectingantibiotic resistance in environmentalsamples

As we have discussed, metagenomics can beused to identify antibiotic resistance genes in theenvironment, and has increased our understand-ing of the sources and roles of these genes innature. However, there are problems associated with metagenomics, and in the next sections wediscuss some of those limitations.

Functional metagenomicsFunctional screening, where fragments arecloned and expressed, may be hampered bythe clone and insert size compared to the total

metagenome size. Antibiotic resistance can be


8/10/2019 LIVIA fmb%2E11%2E135 (1).pdf

11/17


encoded by multiple genes that are required to work together (e.g., vancomycin resistance),by a single gene (e.g.,bla , the gene encodingb -lactamase), or by a point mutation in house-keeping genes (e.g., gyrA, the gene that encodesDNA gyrase). Each of these results in difculties

and challenges for metagenomic screens.The results of functional metagenomicsare dependent upon each genes ability to beexpressed in surrogate hosts, typicallyE. coli[88] .Resistance genes are regulated by genetic ele-ments that may not be recognized by the surro-gate hosts gene expression machinery, the codonusage may not be appropriate for expression,post-translational modications may be missing,or the expressed protein may not fold correctly.Heterologous gene expression can also result infalse positives since the foreign gene may interact

in novel ways with the cellular machinery.The range of media types, antibiotic concen-trations and incubation methods used to mea-sure resistance levels make it difcult to compareresults between environments[18] . Diaz-Torreset al ., for example, found that there are problems with the expression of certain tetracycline resis-tance genes found in the human oral microbiome when E. coli was used as a host[51] .

In addition to heterologous gene expressionissues, genes and their encoded proteins areadapted to different environmental conditionsthat might preclude efcient enzyme activity at37C in standard laboratory media. For exam-ple, genes extracted from Alaskan soil microbesthat live at temperatures much lower than 30C[22,59] , and insect midguts, where microbes livein an environment with a pH of 12.4 [29] . Itis possible that many of the genes in func-tional metagenome libraries are not expressedin E. coli , resulting in an underestimate of thefrequency of resistance determinants in envi-ronmental samples. Some resistance geneshave been found using a wider range of expres-sion systems and hosts, but global sequencingapproaches indicate a large discrepancy betweenpredicted and detected resistance genes[89] . Thelimitations ofE. coli -based screening call for thebroader use of a wider range of hosts includingnot only Gram-negative, but also Gram-positivehost species.

Another study was not able to identify strep-tomycin-resistant clones in functional screens,despite their presence in the metagenomiclibrary veried using PCR and culturable bac-teria [23] . In addition to unidentied resistancegenes, a large number of resistant clones may be

false positives[63] .

Sequence-based metagenomicsThe biggest challenge with sequence-basedmetagenomics is the large number of sequencesthat show no signicant similarity to previouslysequenced genes or organisms; without knownreference sequences, resistance genes cannot be

easily identied in the metagenomes. The strongselection for antibiotic resistance alleles resultsin convergent evolution the adaption of verydifferent genes to perform the same function[16,90,91] . We noted above that many resistancegenes identied in functional screens have lowsimilarity to known genes, but with sequence-based approaches we are generally limited toonly identify things we already know.

The current sequence-based metagenomicapproaches need to be evaluated based on thecomplexity of technical procedures, robustness,

accuracy and cost. The preparation of a samplelibrary requires multiple molecular biology stepsand, depending on the technology, up to 4 daysto complete. The range of data volumes leads toprocessing times from a few minutes to multiplehours, emphasizing the need for sufcient com-putation power. The data analysis requires bothexpertise in bioinformatics and a more advancedinformatics infrastructure.

High-throughput sequencing technologiesgenerate data that currently challenge data stor-age, management and processing, demandingaccess to supercomputing resources or cloud-based computing services for efcient handling. Advances are needed in data transfer and man-agement, standardization of data formats, andintegration of different types of data[92] . Theamount of data is even challenging nationaldata warehouses, such as the Sequence Read Archive, which announced that it will limit dataarchiving to a specic subset of next-generationsequence data starting from October 2011.

The sequence-based identication of a resis-tance gene requires separate functional con-rmation. The metagenomic sequences onlysuggest the presence of an enzyme that mayencode antibiotic resistance, but do not nec-essarily conrm that a gene is functionallyexpressed or that it does not encode an alter-native function in its host. With the advance-ment of mRNA extraction from environmentalsamples, metatranscriptomics may become themain method for the detection of functionalresistance genes[93] .

Whether the methodology employed in micro-biota characterizations are faithfully reproducingthe community composition, and not distort-

ing it due to experimental bias or sequencing


8/10/2019 LIVIA fmb%2E11%2E135 (1).pdf

12/17


artifacts is still an open question. However, assequencing becomes cheaper, researchers willbe able to sequence deeper and at multiple timepoints to address possible biases, and to answerthe important questions about the sources, evo-lution, and effects of antibiotic resistance genes

that we have touched on here.

Conclusion We are dependent on antibiotics for the treat-ment of infectious diseases and they are criti-cal for the success of advanced surgical proce-dures, such as organ and prosthetic transplants. Ant ibiotic-resistance mechanisms create anenormous clinical and financial burden onhealthcare systems worldwide. Despite theproblem of antibiotic resistance in infectiousbacteria, little is known about the diversity,

distribution and origins of resistance genes,especially for the unculturable majority ofenvironmental bacteria.

There exists high-level resistance both toantibiotics that have for decades served asgold standard treatments and to those onlyrecently approved for human use. The studyof the environmental resistance reservoir usingmetagenomic approaches will provide an early warning system for future clinica lly relevantantibiotic-resistance mechanisms. Knowledgeof such natural variations will complementstudies on clinical isolates to guide the rationaldevelopment of next-generation antibiotics that will be active against resistant strains.

Functional and sequence-based metagenom-ics have been used for the discovery of novelresistance determinants and the improvedunderstanding of antibiotic resistance mecha-nisms. Metagenomic sequence data can beused to generate sample-specic and temporalantibiotic resistance proles to facilitate anunderstanding of the ecology of the micro-bial communities in an environment as well asthe epidemiology of antibiotic resistance genetransport between and among environments.

Traditional approaches to antibiotic resistancetypically concentrate on human pathogens.There is a clear need to expand the focus toinclude nonpathogenic bacteria in antibioticresearch. This may allow researchers to pre-dict resistance before it emerges clinically; todevelop diagnostic techniques, and to build newtherapeutic strategies to counteract resistancebefore it emerges in human pathogens. Withrespect to functional metagenomics, we need todevelop and apply new approaches to cultivate

the previously uncultivated and rare members of

the microbial communities to assign functionsto the vast number of unknown or hypotheticalgenes, and to develop novel genetic systems thatallow screening of the vast array of microbeson earth to identify antibiotic-resistance genes.

Sequence-based metagenomics allows the

comparison of microbial communities fromdifferent hosts to investigate differences in theresponse to antibiotics and to select the mostfavorable antibiotic to reduce side effects forcertain hosts.

The transition of next-generation sequencinginto clinical diagnostics is in the early stagesof development in large reference laboratoriesand is being leveraged for applications thatrequire large amounts of sequence informa-tion. Sequence-based metagenomic data hasthe ability to combine antibiotic-resistance gene

abundance data with community compositionand metabolic pathway information to providea more complete prole of specic samples.This information is of great importance forthe implementation of rational administrationguidelines for antibiotic therapies. We may alsouse these techniques to learn how human-asso-ciated microbes can be manipulated by antibi-otics or probiotics to reduce our dependencyon antibiotics, provide alternatives to existingantibiotics, and extend the lifetime of currentand new antibiotics.

Many questions remain: the roles of the envi-ronmental reservoirs in clinical resistance devel-opment are still hypothetical; we have little orno evidence that any of the putative resistancegenes identied in the environmental studieshave been mobilized into pathogenic bacteriaand expressed as resistance phenotypes; we donot know how to rapidly identify new resis-tance genes; and functional approaches are stillrelatively low throughput.

Future perspectiveThe application of metagenomics will not onlyfacilitate future identication of novel resistancegenes, but will be used for the prediction offuture evolution of antibiotic resistance and will enable further studies on genetic elementsparticipating in resistance gene transfer.

Within the next few years, we wil l see theapplication of metagenomics or sequence-basedtechnologies in almost every part of life andbiomedical sciences, in particular in clinicaldiagnostic settings. To date, most sequence-based metagenomic studies are performed inresearch laboratories using the Roche/454 and

Illumina systems(F IGURE 5) . With the promises of


8/10/2019 LIVIA fmb%2E11%2E135 (1).pdf

13/17


third-generation sequencing technologies, suchas higher throughput (amounts of DNA thatcan be processed per unit time) with increasedaccuracy, longer read lengths, lower cost,smaller amounts of starting material and shorterturnaround times (time to result), these tech-

nologies are anticipated to transition into clin-ical-diagnostics use over the next several years.In June 2011, the US FDA held a public meet-ing to discuss the use of ultra high-throughputsequencing for clinical diagnostic applications.In addition to sequencing technologies, the bio-informatics analysis of the sequence data was amajor focus of the meeting. The large amountsof sequence data generated with next-generationsequencing technologies pose a bioinformaticschallenge for the clinical laboratory to providedata processing and interpretation to the clini-

cians. New algorithms, visualization tools, anddata abstractions will be needed to cope withthe challenges presented by this data.

High-throughput sequencing allows forglobal analysis of commensal populations andpotential identication of distinct signatures inthe human microbiome. As sequencing becomesless expensive, repeated sampling of the metage-nome of a microbial community can be used toevaluate changes in the community over time,and whole-body metagenome approaches canbe applied in large-scale research and clinicalstudies. For example, urine and feces providetwo windows that reect the health status ofthe human body, and sequencing the microbi-omes associated with these samples will providea snapshot of overall health. We will also movetowards personalized prescriptions, a subsetof the genome-inspired personalized medi-cine. Antibiotic treatment of high-risk patients

will not stop, but metagenomics will help usunderstand the effects of different classes ofantibiotics on the composition and function ofcommensal and pathogenic microbial popula-tions. These insights will allow the switch frombroad- to narrow-spectrum antibiotics, antibi-

otics that target very specic pathogens, andvery specic courses of treatment instead of thestandard 710 days. In addition, a new categoryof therapeutics will be developed that target themicrobiota and modulate microbemicrobe andmicrobehuman interactions. These therapeu-tics will allow us to alter the composition of themicrobial community, endowing it with newfunctions, and use the sequences of the micro-biota to diagnose disease. In the future, thismight be integrated into daily life; for example,DNA sequencers in toilet bowls that automati-

cally follow changes of the human microbiomeand suggest appropriate antibiotic treatment.

Executive summary

Mechanisms of antibiotic resistance n Resistance to antibiotics can be caused by four general mechanisms (inactivation, alteration of the target, circumvention of the target

pathway or efux of the antibiotic) and bacteria can develop resistance by mutating existing genes, or by acquiring new genes fromother strains or species.

Detection of antibiotic resistance n Antibiotic resistance is a highly selectable phenotype and can be detected using growth inhibition assays; however, culture-based

approaches can take days to weeks for slow-growing bacteria and only work for a fraction of microbes. n Antibiotic resistance genes were typically isolated by cloning from cultured bacteria or by PCR amplication. Those methods missed

potential antibiotic resistance reservoirs because of uncultivable bacteria or dependence on known resistance genes. n Metagenomics overcomes the limitations of culture-dependent techniques and is a powerful tool to identify novel resistance genes and

to describe the presence or absence of genes or genetic variations responsible for antibiotic resistance.

Culture-independent study of resistance through metagenomics n Functional metagenomics involves cloning and expression of DNA in a surrogate host with coupled activity-based screening.

n Sequence-based metagenomics involves extracting and random sequencing of DNA directly from the environment. The metagenomicsequences are then compared to known sequences to identify resistance genes and/or mutations.

AcknowledgementsThe authors would like to thank Linda Wegley, MollySchmid and Anca Segall for helpful comments,suggestions and thought provoking discussions.

Financial & competing interests disclosureThis work was funded by NSF grants DBI:0850356 from the Division of Biological Infrastructure andDEB:1046413 from the Division of EnvironmentalBiology (both to R Edwards). The authors have noother relevant afliations or nancial involvementwith any organization or entity with a nancial inter-est in or nancial conict with the subject matter ormaterials discussed in the manuscript apart from thosedisclosed.

No writing assistance was utilized in the productionof this manuscript.


8/10/2019 LIVIA fmb%2E11%2E135 (1).pdf

14/17


Executive summary (cont.)

Antibiotic resistance in human-associated microbes n It is estimated that there are ten-times as many microbes in and on any given human as there are cells in that persons body and the

majority of those have not yet been cultured. Large-scale sequencing projects such as the HMP and MetaHIT have been started as partof the International Human Microbiome Consortium to investigate the human microbiota.

n Antibiotics may have long-term consequences on the human microbiota, which cannot fully recover after repeated antibioticperturbation, and might cause a shift to a different, but stable community.

n Our oral and fecal microbiota are distinctly different, both in the organisms present and the resistance mechanisms in those organisms.

Antibiotic resistance in soil microbial communities n There are an estimated 10 9 microorganisms per gram of soil and less than 1% of those microbes are readily culturable by current

methods. The high density of antibiotic-producing bacteria makes soil a reservoir for antibiotic resistance. n Metagenomic studies of soil environments identied novel and known genes that mediate resistance to different classes of antibiotics,

including soil from remote locations with no known exposure to antibiotic drugs. The resistance proteins had low similarity to previouslypublished sequences suggesting that soil microorganisms harbor more genetic diversity than previously assumed.

Antibiotic resistance associated with wastewater treatment plants n Wastewater treatment plants are interfaces between different environments and provide an opportunity for resistance to mix between

pathogens, opportunistic pathogens, and environmental bacteria. The sludge products are increasingly used to fertilize agriculturalcrops, dispersing unknown amounts of resistance genes and antibiotics.

n Activated sludge showed evidence for a high diversity of resistance genes and allowed the identication of novel resistance genes. n The survival of antibiotic-resistance genes through the sewage treatment process and in the aquatic environment merits further study.

Antibiotic resistance in marine microbial communities n Antibiotic-resistant bacteria have been found widely in aquatic environments and resistance genes found in clinical human pathogens

have been identied in marine microbes.

Animal- & agriculture-related resistance n Antibiotics are extensively used for animal farming and for agricultural purposes and may have effects on the selection of resistance. n Residues from farms can contain antibiotic resistance genes that may impact environmental microbiota.

Ongoing challenges for detecting antibiotic resistance in environmental samples n The results of functional metagenomics are dependent upon each genes ability to be expressed in surrogate hosts. The limitations of

Escherichia coli -based screening most likely resulted in an underestimate of the frequency of resistance determinants in environmentalsamples. In addition, a range of media types, antibiotic concentrations and incubation methods make it difcult to compare resultsbetween environments or laboratories.

n The biggest challenge with sequence-based metagenomics is the large number of sequences that show no signicant similarity assequence-based approaches are generally limited to only identify things we already know. High-throughput sequencing technologiesgenerate data that currently challenge data storage, management and processing, demanding access to supercomputing resources andexpertise in bioinformatics.

n Sequence-based identication of a resistance gene requires separate functional conrmation; however, with the advancement ofmRNA extraction from environmental samples, metatranscriptomics may become be the main method for the detection of functionalresistance genes.

Conclusion n Functional and sequence-based metagenomics have been used for the discovery of novel resistance determinants and the improved

understanding of antibiotic-resistance mechanisms.n Sequence-based metagenomic data has the ability to combine antibiotic resistance gene abundance data with community composition

and metabolic pathway information to provide a more complex prole of specic samples. n The study of the environmental resistance reservoir using metagenomic approaches could provide an early warning system for future

clinically relevant antibiotic resistance mechanisms.

ReferencesPapers of special note have been highlighted as:n of interest

1. Fauci AS, Touchette NA, Folkers GK.Emerging infectious diseases: a 10-yearperspective from the National instituteof allergy and infectious diseases.Emerging Infect. Dis. 11, 519525(2005).

2. Spek J, Novotn J, Rezanka T, Demain AL.Do we need new antibiotics? The search fornew targets and new compounds. J. Ind. Microbiol . Biotechnol . 37, 12411248 (2010).

3. Hancock REW. The end of an era?Nat. Rev.Drug Discov. 6, 28 (2007).

4. Jernberg C , Lfmark S, Ed lund C, Jansson JK. Long-term impacts of antibiotic exposureon the human intestinal microbiota.

Microbiolog y. 156, 32163223 (2010).

5. McManus MC. Mechanisms of bacterialresistance to antimicrobial agents. Am. J.Health Syst. Pharm. 54, 14201433 (1997).

6. Crumplin GC, Odell M. Development ofresistance to ooxacin.Drugs 34(Suppl. 1),S1S8 (1987).

7. Martinez JL , Baquero F. Mutation frequenciesand antibiotic resistance. Antimicrob. AgentsChemother. 44, 17711777 (2000).


8/10/2019 LIVIA fmb%2E11%2E135 (1).pdf

15/17


8. Palmer KL, Kos VN, Gilmore MS.Horizontal gene transfer and the genomics ofenterococcal antibiotic resistance.Curr. Opin. Microbiol . 13, 632639 (2010).

9. Hegstad K, Mikalsen T, Coque TM, WernerG, Sundsfjord A. Mobile genetic elementsand their contribution to the emergence ofantimicrobial resistantEnterococcus faecalis and Enterococcus faecium. Clin. Microbiol.Infect. 16, 541554 (2010).

10. Jeters RT, Wang GR, Moon K, ShoemakerNB, Salyers AA . Tetracycline-associatedtranscriptional regulation of transfer genes ofthe Bacteroides conjugative transposonCTnDOT. J. Bacter iol. 191, 63746382(2009).

11. Ochiai K, Yamanaka T, Kimura K, SawadaO. [Inheritance of drug resistance (and itstransfer) betweenShigella strains and betweenShigella and E. coli strains]. Nippon Iji Shinpo

(Jap. M.J.) 1861, 3436 (1959).12. Shaheen BW, Wang C, Johnson CM,

Kaltenboeck B, Boothe DM. Detection ofuoroquinolone resistance level in clinicalcanine and felineEscherichia coli pathogensusing rapid real-time PCR assay.Vet. Microbiol . 139, 379385 (2009).

13. Volokhov D, Chizhikov V, Chumakov K,Rasooly A. Microarray ana lysis oferythromycin resistance determinants. J. Appl. Microbiol . 95, 787798 (2003).

14. Zhu LX, Zhang ZW, Wang C et al . Use of aDNA microarray for simultaneous detectionof antibiotic resistance genes amongstaphylococcal clinical isolates. J. Clin. Microbiol . 45, 35143521 (2007).

15. Martinez JL, Snchez MB, Martnez-SolanoL et al . Functional role of bacterial multidrugefux pumps in microbial naturalecosystems.FEMS Microbiol. Rev. 33,430449 (2009).

16. Aminov RI. The role of antibiot ics andantibiotic resistance in nature.Environ. Microbiol . 11, 29702988 (2009).

17. Cantn R. Antibiotic resistance genes fromthe environment: a perspective throughnewly identied antibiotic resistancemechanisms in the clinical setting.Clin. Microbiol. Infect. 15(Suppl. 1), S20S25(2009).

18. Allen HK, Donato J, Wang HH, Cloud-Hansen KA, Davies J, Handelsman J. Call ofthe wild: antibiotic resistance genes in naturalenvironments.Nat. Rev. Microbiol. 8,251259 (2010).

n Review of the dist ribution of antibioticresistance elements in the environment.

19. Wright GD. Ant ibiotic resistance in theenvironment: a link to the clinic?Curr. Opin. Microbiol . 13, 589594 (2010).

20. Riesenfeld CS, Goodman RM, Handelsman J. Uncultured soil bacteria a re a reservoir ofnew antibiotic resistance genes.Environ. Microbiol . 6, 981989 (2004).

n First use of functional metagenomics toidentify novel resistance genes.

21. Handelsman J, Rondon MR, Brady SF,Clardy J, Goodman RM. Molecularbiological access to the chemistry ofunknown soil microbes: a new f rontier fornatural products. Chem. Biol. 5, R245R249(1998).

22. Allen HK, Moe L A, Rodbumrer J, Gaarder A, Ha ndelsman J. Fu nctional metagenomicsreveals diverseb -lactamases in a remote Alaskan soil.ISME J. 3, 243251 (2009).

n First report of a bifunctional b -lactamaseidentied using functional metagenomics.

23. Donato JJ, Moe LA, Converse BJet al .Metagenomic analysis of apple orchard soilreveals antibiotic resistance genes encodingpredicted bifunctional proteins. Appl.Environ. Microbiol. 76, 43964401 (2010).

n Identied bifunctional proteins usingfunctional metagenomics.

24. Diaz-Torres ML, McNab R, Spratt DAet al .Novel tetracycline resistance determinantfrom the oral metagenome. Antimicrob. AgentsChemother. 47, 14301432 (2003).

25. Mori T, Mizuta S, Suenaga H, Miyazak i K.Metagenomic screening for bleomycinresistance genes. Appl. Environ. Microbiol. 74,68036805 (2008).

26. Pace NR, Stahl DA, Lane DJ, Olsen GJ. Analyzing natural microbial populations byrRNA sequences. ASM News 51, 412 (1985).

27. Healy FG, Ray RM, Aldrich HC, Wilk ie AC,Ingram LO, Shanmugam KT. Direct isolationof functional genes encoding cellulases fromthe microbial consortia in a thermophilic,anaerobic digester maintained onlignocellulose. Appl. Microbiol. Biotechnol. 43,667674 (1995).

28. Tringe SG, Rubin EM. Metagenomics: DNAsequencing of environmental samples.Nat.Rev. Genet. 6, 805814 (2005).

29. Allen HK, Cloud-Hansen KA, Wolinsk i JMet al . Resident microbiota of the gypsy mothmidgut harbors antibiotic resistancedeterminants. DNA Cell Biol. 28, 109117(2009).

30. Poinar HN, Schwarz C, Qi J et al .Metagenomics to paleogenomics: large-scalesequencing of mammoth DNA.Science 311,392394 (2006).

31. Noonan JP, Coop G, Kudarava lli Set al .Sequencing and analysis of Neanderthalgenomic DNA. Science 314, 11131118(2006).

32. Markowitz VM, Ivanova NN, Szeto Eet al .IMG/M: a data management and analysissystem for metagenomes.Nucleic Acids Res. 36, D534D538 (2008).

33. Meyer F, Paarmann D, DSouza Met al . Themetagenomics RAST server a publicresource for the automatic phylogenetic andfunctional analysis of metagenomes.BMCBioinformatics 9, 386 (2008).

34. Sun S, Chen J, Li Wet al . Communitycyberinfrastructure for advanced microbialecology research and analysis: the CAMERAresource.Nucleic Acids Res. 39, D546D551(2011).

35. Leinonen R, Sugawara H, Shumway M. Thesequence read archive.Nucleic Acids Res. 39,D19D21 (2011).

36. Ley RE , Peterson DA, Gordon JI. Ecologicaland evolutionary forces shaping microbialdiversity in the human intestine.Cell 124,

837848 (2006).37. Bckhed F, Ley RE, Sonnenburg JL, Peterson

DA, Gordon JI. Host-bacterial mutualism inthe human intestine. Science 307, 19151920(2005).

38. Eckburg PB, Bik EM, Bernstein CNet al .Diversity of the human intestinal microbialora. Science 308, 16351638 (2005).

39. Turnbaugh PJ, Ley RE, Hamady M,Fraser-Liggett CM, Knight R, Gordon JI.The human microbiome project.Nature 449,804810 (2007).

40. Peterson J, Garges S, Giovanni Met al . The

NIH human microbiome project. GenomeRes. 19, 23172323 (2009).

41. Qin J, Li R, Raes Jet al . A human gutmicrobial gene catalogue established bymetagenomic sequencing.Nature 464, 5965(2010).

42. Turnbaugh PJ, Hamady M, Yatsunenko Tet al. A core gut microbiome in obese andlean twins.Nature 457, 480 484 (2009).

43. Ley RE, Bckhed F, Turnbaugh P, LozuponeCA, Knight RD, Gordon JI. Obesity altersgut microbial ecology.Proc. Natl Acad. Sci.USA 102, 1107011075 (2005).

44. Ley RE, Lozupone CA, Hamady M, KnightR, Gordon JI. Worlds within worlds:evolution of the vertebrate gut microbiota.Nat. Rev. Microbiol. 6, 776788 (2008).

45. Roumie CL, Halasa NB, Grijalva CGet al. Trends in antibiotic prescribing for adults inthe United States 1995 to 2002. J. Gen.Intern. Med. 20, 697702 (2005).

46. Dethlefsen L, Relman DA. Incompleterecovery and individualized responsesof the human distal gut microbiota torepeated antibiotic perturbation.Proc. Natl Acad. Sci. USA 108(Suppl. 1), S4554S4561

(2011).


8/10/2019 LIVIA fmb%2E11%2E135 (1).pdf

16/17


47. Jakobsson HE, Jernberg C, A ndersson A F,Sjlund-Karlsson M, Jansson JK, EngstrandL. Short-term antibiotic treatment hasdiffering long-term impacts on the humanthroat and gut microbiome. PLoS ONE 5,E9836 (2010).

48. Sommer MOA, Dantas G, Church GM.Functional characterization of the antibioticresistance reservoir in the human microora.Science 325, 11281131 (2009).

n Functional multidrug resistance screeningof metagenomic and genomic DNAhighlights the under-sampling of resistancegenes by culture-dependent methods.

49. Sommer MOA, Church GM, Dantas G. Thehuman microbiome harbors a diverse reservoirof antibiotic resistance genes.Virulence 1,299303 (2010).

50. Koenig JE, Spor A, Scal fone Net al. Succession of microbial consortia in thedeveloping infant gut microbiome.Proc. Natl Acad. Sci. USA 108(Suppl. 1), S4578S4585(2011).

51. Diaz-Torres ML, Villedieu A, Hunt Net al. Determining the antibiotic resistancepotential of the indigenous oral microbiota ofhumans using a metagenomic approach.FEMS Microbiol. Lett. 258, 257262 (2006).

52. Seville LA, Patterson AJ, Scott KPet al. Distribution of tetracycline and erythromycinresistance genes among human oral and feca lmetagenomic DNA. Microb. Drug Resist. 15,159166 (2009).

53. Trevors JT. One gram of soil: a microbialbiochemical gene library. Antonie. Van.Leeuwenhoek 97, 99106 (2010).

54. Singh BK, Millard P, Whiteley AS, Murrell JC. Unravelling rhizosphere-microbia linteractions: opportunities and limitations.Trends Microbiol. 12, 386393 (2004).

55. Vogel TM, Simonet P, Jansson JKet al. TerraGenome: a consortium for thesequencing of a soil metagenome.Nat. Rev. Micro. 7, 252 (2009).

56. DCosta VM, Grifths E, Wright GD.Expanding the soil antibiotic resistome:

exploring environmental diversity.Curr.Opin. Microbiol. 10, 481489 (2007).

57. Martn MF, Liras P. Organization andexpression of genes involved in thebiosynthesis of antibiotics and othersecondary metabolites. Ann. Rev. Microbiol. 43, 173206 (1989).

58. DCosta VM, McGrann K M, Hughes DW, Wright GD. Sampling the ant ibioticresistome.Science 311, 374377 (2006).

59. Lang KS, Anderson JM, Schwarz S, Will iamson L , Handelsman J, Singer RS.Novel orfenicol and chloramphenicol

resistance gene discovered in Alaskan soil by

using functional metagenomics. Appl.Environ. Microbiol. 76, 53215326 (2010).

60. Benveniste R, Davies J. A minoglycosideantibiotic-inactivating enzymes inactinomycetes similar to those present inclinical isolates of antibiotic-resistant bacteria.Proc. Natl Acad. Sci. USA 70, 22762280(1973).

61. Monier JM, Demanche S, Delmont TO,Mathieu A, Vogel TM, Simonet P.Metagenomic exploration of antibioticresistance in soil.Curr. Opin. Microbiol. 14,229235 (2011).

62. Demanche S, Sanguin H, Pot Jet al. Antibiotic-resistant soil bacteria in transgenicplant elds.Proc. Natl Acad. Sci. USA 105,39573962 (2008).

63. Torres-Corts G, Milln V, Ramrez-SaadHC, Nisa-Martnez R, Toro N, Martnez- Abarca F. Characterization of novel ant ibiotic

resistance genes identied by functionalmetagenomics on soil samples.Environ. Microbiol . 13, 11011114 (2011).

64. Linares JF, Gustafsson I, Baquero F, Martinez JL. A ntibiotics as intermicrobial signalingagents instead of weapons.Proc. Natl Acad.Sci. USA 103, 1948419489 (2006).

65. Yim G, Wang HH, Davies J. Antibiotics assignalling molecules.Philos. Trans. R. Soc.Lond. B. Biol. Sci. 362, 11951200 (2007).

66. Fajardo A, Martnez JL. Antibiotics as signalsthat trigger specic bacterial responses.Curr.Opin. Microbiol. 11, 161167 (2008).

67. Dantas G, Sommer MOA, Oluwasegun RD,Church GM. Bacteria subsisting onantibiotics.Science 320, 100103 (2008).

68. Szczepanowski R, Linke B, Krahn Iet al. Detection of 140 clinically relevantantibiotic-resistance genes in the plasmidmetagenome of wastewater treatment plantbacteria showing reduced susceptibility toselected antibiotics. Microbiolog y155,23062319 (2009).

69. Parsley LC, Consuegra EJ, Kakirde KS, Land AM, Harper WF Jr, Liles MR. Identicationof diverse antimicrobial resistance

determinants carried on bacterial, plasmid, orviral metagenomes from an activated sludgemicrobial assemblage. Appl. Environ. Microbiol . 76, 37533757 (2010).

70. Schlter A, Krause L, Szczepanowski R,Goesmann A, Phler A. Genetic diversity andcomposition of a plasmid metagenome from a wastewater treatment plant . J. Biotechnol. 136,6576 (2008).

71. Kristiansson E, Fick J, Janzon Aet al. Pyrosequencing of antibiotic-contaminatedriver sediments reveals high levels of resistanceand gene transfer elements.PLoS ONE 6,E17038 (2011).

72. Baquero F, Martnez JL, Cantn R. Antibioticsand antibiotic resistance in water environments.Curr. Opin. Biotechnol. 19, 260265 (2008).

n Integrated perspective of antibiotics andresistance genes in aquatic environments.

73. Xi C, Zhang Y, Marrs CFet al . Prevalence ofantibiotic resistance in drinking watertreatment and distribution systems. Appl.Environ. Microbiol. 75, 57145718 (2009).

74. Pallecchi L, Bartoloni A, Paradisi F, RossoliniGM. Antibiotic resistance in the absence ofantimicrobial use: mechanisms andimplications.Exp. Rev. Anti. Infect. Ther. 6,725732 (2008).

75. Bartoloni A, Pallecchi L, Rodrguez Het al. Antibiotic resistance in a very remote Amazonas community.Int. J. Antimicrob. Agents 33, 125129 (2009).

76. Wegley L, Edwards R, Rodriguez-Brito B, LiuH, Rohwer F. Metagenomic analysis of themicrobial community associated with the coralPorites astreoides.Environ. Microbiol. 9,27072719 (2007).

77. Pappas G. An animal farm called extended-spectrumb -lactamase: antimicrobial resistanceas a zoonosis.Clin. Microbiol. Infect. (2011).

78. Bager F, Madsen M, Christensen J, AarestrupFM. Avoparcin used as a growth promoter isassociated with the occurrence of vancomycin-resistantEnterococcus faecium on Danishpoultry and pig farms.Prev. Vet. Med. 31,95112 (1997).

79. Mirzaagha P, Louie M, Sharma R, Yanke LJ,Topp E, McA llister TA. Distribution andcharacterization of ampicillin- and tetracycline-resistantEscherichia coli from feedlot cattle fedsubtherapeutic antimicrobials.BMC Microbiol. 11, 78 (2011).

80. Lu K, Asano R, Davies J. Antimicrobialresistance gene delivery in animal feeds.Emerging Infect. Dis 10, 679683 (2004).

81. Martinez JL. Environmental pollution byantibiotics and by antibiotic resistancedeterminants. Environ. Pollut. 157, 28932902(2009).

82. Durso LM, Harhay GP, Bono JL, Smith TPL.

Virulence-associated and antibiotic resistancegenes of microbial populations in cattle fecesanalyzed using a metagenomic approach. J. Microbiol. Methods 84, 278282 (2011).

83. Berger CN, Sodha SV, Shaw RKet al . Freshfruit and vegetables as vehicles for thetransmission of human pathogens.Environ. Microbiol. 12, 23852397 (2010).

84. Gibbs SG, Green CF, Tarwater PM, ScarpinoPV. Airborne antibiotic resistant andnonresistant bacteria and fungi recovered fromtwo swine herd conned animal feedingoperations. J. Occup. Environ. Hyg. 1, 699706

(2004).


8/10/2019 LIVIA fmb%2E11%2E135 (1).pdf

17/17

LIVIA fmb%2E11%2E135 (1).pdf

Documents

Transcript of LIVIA fmb%2E11%2E135 (1).pdf