Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2018 Jan 15;6(1):11.
doi: 10.1186/s40168-017-0387-y.

In-depth resistome analysis by targeted metagenomics

Val F Lanza  1   2   3   4 , Fernando Baquero  1   2   3 , José Luís Martínez  2   4 , Ricardo Ramos-Ruíz  5 , Bruno González-Zorn  6 , Antoine Andremont  7 , Antonio Sánchez-Valenzuela  1 , Stanislav Dusko Ehrlich  8   9 , Sean Kennedy  8   10 , Etienne Ruppé  7   8 , Willem van Schaik  11   12 , Rob J Willems  11 , Fernando de la Cruz  13   14 , Teresa M Coque  15   16   17
Affiliations

In-depth resistome analysis by targeted metagenomics

Val F Lanza et al. Microbiome. .

Abstract

Background: Antimicrobial resistance is a major global health challenge. Metagenomics allows analyzing the presence and dynamics of "resistomes" (the ensemble of genes encoding antimicrobial resistance in a given microbiome) in disparate microbial ecosystems. However, the low sensitivity and specificity of available metagenomic methods preclude the detection of minority populations (often present below their detection threshold) and/or the identification of allelic variants that differ in the resulting phenotype. Here, we describe a novel strategy that combines targeted metagenomics using last generation in-solution capture platforms, with novel bioinformatics tools to establish a standardized framework that allows both quantitative and qualitative analyses of resistomes.

Methods: We developed ResCap, a targeted sequence capture platform based on SeqCapEZ (NimbleGene) technology, which includes probes for 8667 canonical resistance genes (7963 antibiotic resistance genes and 704 genes conferring resistance to metals or biocides), and 2517 relaxase genes (plasmid markers) and 78,600 genes homologous to the previous identified targets (47,806 for antibiotics and 30,794 for biocides or metals). Its performance was compared with metagenomic shotgun sequencing (MSS) for 17 fecal samples (9 humans, 8 swine). ResCap significantly improves MSS to detect "gene abundance" (from 2.0 to 83.2%) and "gene diversity" (26 versus 14.9 genes unequivocally detected per sample per million of reads; the number of reads unequivocally mapped increasing up to 300-fold by using ResCap), which were calculated using novel bioinformatic tools. ResCap also facilitated the analysis of novel genes potentially involved in the resistance to antibiotics, metals, biocides, or any combination thereof.

Conclusions: ResCap, the first targeted sequence capture, specifically developed to analyze resistomes, greatly enhances the sensitivity and specificity of available metagenomic methods and offers the possibility to analyze genes related to the selection and transfer of antimicrobial resistance (biocides, heavy metals, plasmids). The model opens the possibility to study other complex microbial systems in which minority populations play a relevant role.

Keywords: Antimicrobial resistance; Differential abundance analysis; Metagenomics; Resistome; Targeted metagenomics.

PubMed Disclaimer

Conflict of interest statement

Ethics approval and consent to participate

Clinical samples were used after obtaining the approval of the study by the Institutional Review Board of the Hospital Bichat, and all written informed consents from the enrolled subjects, in compliance with national legislation and the Code of Ethical Principles for Medical Research Involving Human Subjects of the World Medical Association (Declaration of Helsinki).

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Figures

Fig. 1
Fig. 1
ResCap analysis workflow. Processed reads are mapped against the reference database. SAM files are parsed to extract the reads unequivocally mapped and those ambiguously mapped to determine the genes unequivocally detected and to form the allele network. The allele network is built using all the study’s SAM files. The MGCs determined from the allele network were used to perform the statistical analysis of abundance and diversity. Finally, a differential analysis was performed with the abundance data
Fig. 2
Fig. 2
Allele network: nodes of the network represents individual genes that are mapped by some read. Edges between nodes represent reads that mapped on both nodes that link. Individual nodes are genes that are unequivocally identified. Gene clusters are mainly composed of different variants of the same gene (alleles). The Mapping Gene Cluster (MGC) is defined using the Markov cluster algorithm (MCL)
Fig. 3
Fig. 3
ResCap Performance Summary. Panel a represents the gain function in reads per kilobase per million of reads of each detected gene between MSS protocol (abscissa axis) and ResCap (ordinate axis). Genes only detected by ResCap are represented by the dot cluster in the initial values of the abscissa axis. Data distribution of the platform efficiency evaluating b the number of mapped reads per million of sequenced reads against a canonical (well known) gene data set; and c the number of detected genes per million of sequenced reads using as reference the well-known gene data set. Fecal samples were differentiated according to the source. Data distribution of the platform efficiency evaluating d the number of mapped reads per million of sequenced reads against the three canonical gene groups and e the number of detected genes per million of sequenced reads using as reference the three canonical gene groups
Fig. 4
Fig. 4
Longitudinal coverage distribution. The figure shows the comparison of longitudinal coverage distribution between protocols in each sample. Distributions are represented by density parameter and expressed by the number of genes (ordinate axis) and the coverage percent (abscissa axis)
Fig. 5
Fig. 5
Quantification of unequivocally mapping reads. The figure shows the comparative of the quantification of reads mapping on just one gene (or allele). First, the abundance of reads that are unequivocally mapped on one gene (a). Second, the number of genes (or MGC) that have almost one read that maps unequivocally (b). Box plots are differentiated for MSS protocol and ResCap protocol
Fig. 6
Fig. 6
(1) Abundance and diversity of antibiotic resistances. Comparison of ResCap and MSS protocol in antibiotic resistance data. Antibiotic resistance genes were classified among nine antibiotic families (AGly: aminoglycosides, Bla: beta-Lactams, Flq: fluoroquinolones, Gly: glycopeptides, MLS: macrolides, Phe: phenicols, Sul: sulfonamides, Tet: tetracyclines and Tmt: trimethoprim). Abundance (a) was measured as read per kilobase per million reads that mapped against genes or allele-cluster genes of each antibiotic resistance family. Diversity (b) was measured as the number of detected genes per million reads of each antibiotic resistance family (2) Abundance and diversity of relaxases. Comparison of ResCap and MSS protocol in relaxases dataset. Relaxases were classified in nine protein families (MOBB, MOBC, MOBF, MOBH, MOBP1, MOBP2, MOBQ, MOBT, and MOBV). Abundance (a) was measured as read per kilobase per million reads that mapped against genes or allele-cluster genes of each relaxase family. Diversity (b) was measured as a number of detected genes per million reads of each relaxase family. (3) Abundance and diversity of biocide and metal resistances. Comparison of ResCap and MSS protocol in biocide and metal resistance data. Biocide and metal resistance genes were classified by the type of detoxified targets. Abundance (a) was measured as read per kilobase per million reads that mapped against genes or allele-cluster genes of each target family. Diversity (b) was measured as a number of detected genes per million reads of each target family.
Fig. 7
Fig. 7
Differential study plots. Panel a shows the distribution of abundance variation between swine and human AbR resistomes (left), metal and biocide resistome (middle), and mobilome (right) in the form of volcano plots (fold change vs p value) using the different approaches MSS (top) and ResCap (bottom). Left and right branches in the volcano plot refers to higher abundance in humans and swine, respectively. Abscissa axis reflects the relative abundance between humans and swine samples. Positive values represent MGCs more abundant in swine than in human samples. Negative values represent MGCs more abundant in human than in swine samples and the values near to zero represent the MGCs with similar abundance between samples. Panel b summarizes the number of statistically significant MGCs of humans, swines, and the genes in common between them using both approaches: MSS (top) and ResCap (bottom). Panel c shows the Venn diagrams between approaches of differentially detected MGCs (top) and commonly (in both sets) detected MGCs (bottom)
Fig. 8
Fig. 8
Assembly statistics. Assembly statistics was calculated by Quast software. Statistic summary of the main assembly variables; the number of contigs (all and longer than 1 kb), number of genes per sequenced Mb, the size of longest contig, the length of the assembled metagenome per sequenced Mb, and the N50 (the shortest contig length at 50% of the metagenome). Coverage data were calculated as the total sequenced bases divided by the total length (without normalizing). Length per Mb and genes per Mb were normalized by the total amount of megabases sequenced by each sample
Fig. 9
Fig. 9
Functional annotation distribution. Assembled genes are classified as ResCapDB, UniProtKB, or novel genes (see “Methods”). All assessed genes have a maximum e value of 10−100 with some of the genes included in the ResCap design database. The figure shows the comparison between human and swine samples and between MSS and ResCap approaches
See this image and copyright information in PMC

References

    1. Årdal C, Outterson K, Hoffman SJ, Ghafur A, Sharland M, Ranganathan N, Smith R, Zorzet A, Cohn J, Pittet D, Daulaire N, Morel C, Rizvi Z, Balasegaram M, Dar OA, Heymann DL, Holmes AH, Moore LS, Laxminarayan R, Mendelson M, Røttingen JA. International cooperation to improve access to and sustain effectiveness of antimicrobials. Lancet. 2016;387:296–307. doi: 10.1016/S0140-6736(15)00470-5. - DOI - PubMed
    1. Årdal C, Baraldi E, Ciabuschi F, Outterson K, Rex JH, Piddock LJV, Findlay D; DRIVE-AB Steering Committee. To the G20: incentivising antibacterial research and development. Lancet Infect Dis. 2017:799-801. - PubMed
    1. Dar OA, Hasan R, Schlundt J, Harbarth S, Caleo G, Dar FK, Littmann J, Rweyemamu M, Buckley EJ, Shahid M, Kock R, Li HL, Giha H, Khan M, So AD, Bindayna KM, Kessel A, Pedersen HB, Permanand G, Zumla A, Røttingen JA, Heymann DL. Exploring the evidence base for national and regional policy interventions to combat resistance. Lancet. 2016;387(10015):285-95. - PubMed
    1. Bax R, Bywater R, Cornaglia G, Goossens H, Hunter P, Isham V, et al. Surveillance of antimicrobial resistance--what, how and whither? Clin Microbiol Infect. 2001;7:316–325. doi: 10.1046/j.1198-743x.2001.00239.x. - DOI - PubMed
    1. Laxminarayan R, Duse A, Wattal C, Zaidi AKM, Wertheim HFL, Sumpradit N, et al. Antibiotic resistance-the need for global solutions. Lancet Infect Dis. 2013;13:1057–1098. doi: 10.1016/S1473-3099(13)70318-9. - DOI - PubMed

Publication types