Practical implications of erythromycin resistance gene diversity on surveillance and monitoring of resistance

Practical implications of erythromycin resistance gene diversity on surveillance and monitoring... Abstract Use of antibiotics in human and animal medicine has applied selective pressure for the global dissemination of antibiotic-resistant bacteria. Therefore, it is of interest to develop strategies to mitigate the continued amplification and transmission of resistance genes in environmental reservoirs such as farms, hospitals and watersheds. However, the efficacy of mitigation strategies is difficult to evaluate because it is unclear which resistance genes are important to monitor, and which primers to use to detect those genes. Here, we evaluated the diversity of one type of macrolide antibiotic resistance gene (erm) in one type of environment (manure) to determine which primers would be most informative to use in a mitigation study of that environment. We analyzed all known erm genes and assessed the ability of previously published erm primers to detect the diversity. The results showed that all known erm resistance genes group into 66 clusters, and 25 of these clusters (40%) can be targeted with primers found in the literature. These primers can target 74%–85% of the erm gene diversity in the manures analyzed. antibiotic, resistance, target, environment, surveillance, genes INTRODUCTION Antibiotic resistance is a global challenge, with increasing resistance to antibiotics threatening our ability to treat both human and animal diseases (WHO). Antibiotic use in human medicine and animal agriculture has increased environmental reservoirs of antibiotic resistance genes, which in turn has increased the risk of transmission of antibiotic-resistant bacteria to both humans and animals (McEwen and Fedorka-Cray 2002; Vaz-Moreira et al.2014). This linkage has resulted in the prioritization of understanding how resistance moves from environmental sources to clinical pathogens and the associated influence of human activity. To understand the movement of antibiotic resistance in the environment, we need accessible tools that can provide large-scale surveillance of resistance in diverse environmental samples. Molecular microbiology advances have allowed us to leverage amplification and subsequent sequencing of DNA that encodes for antibiotic resistance genes, resulting in our awareness of an incredibly diverse global reservoir of environmental “resistomes”. Generally, metagenomic shotgun sequencing is a costly tool for antibiotic gene surveillance as it provides information on ‘all’ genes in an environmental sample. Among these genes, only a fraction (0.01%–1%) are related to antibiotic resistance, resulting in a significant majority of sequences from metagenomes not readily usable for resistance detection (Shi et al.2013; Li et al.2015). A promising alternative to metagenomic sequencing is high-throughput amplicon qPCR assays, such as the Wafergen Smartchip that has been previously used for several resistance surveillance studies (Shi et al.2013; Wang et al.2014; Karkman et al.2016; Muziasari et al.2016; Stedtfeld et al.2017). Unlike the broad scope of metagenomic sequencing, high-throughput qPCR assays target a suite of genes using primers and can quantify hundreds of targeted resistance genes and multiple samples simultaneously (e.g. one Wafergen Smartchip contains 5184 assays). Consequently, the price per gene or sample of these assays for resistance gene detection is orders of magnitude less than metagenomic sequencing, making it more conducive to large-scale surveillance. A significant limitation of this technology is the need to develop primer-based assays for each targeted gene of interest that are effective for high-throughput amplification conditions. We are increasingly aware that certain genes may be more related to the risks of the emergence or persistence of resistance than others. For example, integrons and sulfonamide resistance genes have been used to detect anthropogenic contaminants (Wang et al.2014; Gillings et al.2015). Further, specific environments (mammalian gut, manure, wastewater, etc.) have been observed to be enriched in antibiotic resistance genes relative to soil or water environments (Chee-Sanford et al.2009; Koike et al.2010; Garder, Moorman and Soupir 2014; Joy et al.2014; Luby, Moorman and Soupir 2016), suggesting that these environments are potential reservoirs of resistance genes. Among the hundreds of genes associated with antibiotic resistance that are observed in environmental metagenomes, selecting the key targets relevant to the spread of resistance is a significant and important opportunity. In this study, we demonstrate how we have chosen specific genes that are the most effective among previously targeted genes to serve as indicators for antibiotic resistance and to understand resistance hotspots and transmission. This framework, while developed for agriculturally impacted environments, can be broadly applied to the selection of genes from varying resistance gene classes and environments. Specifically, this effort focuses on understanding the diversity of erythromycin ribosomal methylase (erm) gene and the most relevant gene targets for understanding the spread of erm-associated resistance from manure sources to the environment. Erm genes encode resistance to macrolide antibiotics, which have long been used to treat Gram-positive and certain Gram-negative pathogens infecting humans, swine and cattle (Roberts et al.2008; Pyörälä et al.2014). Broadly, macrolide antibiotics act by binding to the 23S subunit of the bacterial ribosome, causing premature release of peptides during translation. The erm genes cause resistance by methylating rRNA at the active site, reducing the ability of macrolide antibiotics to bind to the ribosome (Weisblum 1998; Vester and Douthwaite 2001). Erm-mediated resistance to macrolides has also been observed to confer resistance against other antibiotics, including lincosamide and streptogramin B (MLSB resistance) (Leclercq and Courvalin 1991). The widespread use of macrolides and their relevance for both animal and human health has resulted in a research emphasis on erm genes and their bacterial hosts as key targets for understanding the development of resistance and its spread in agricultural environments. Previously, erm genes have been detected in various agricultural settings, including swine manure, lagoon water, soils, surface and subsurface drainage from fields, and groundwater surrounding and underlying animal production facilities (Chen et al.2007; Knapp et al.2010; Koike et al.2010; Joy et al.2013, 2014; Whitehead and Cotta 2013; Fahrenfeld et al.2014; Garder, Moorman and Soupir 2014; Soni et al.2015; Luby, Moorman and Soupir 2016). Most of our previous knowledge of erm genes and their associated amplicon targets have stemmed from the characterization and sequencing of bacterial isolates and their phenotypic resistance to MLSB antibiotics (Pyörälä et al.2014). A total of 21 unique classes of erm genes have been identified based on sequence homology to protein-coding erm sequences from cultured bacteria (Roberts et al.1999). More recently, metagenomic analyses of DNA from the total microbial community in environmental samples has expanded what is known about erm diversity beyond these 21 classes, showing that the erm class of genes is comprised of numerous sequence variants from diverse bacterial hosts (Fang et al.2015; Li et al.2015). These sequence variants are present in a range of abundances depending on their environment of origin. The focus of this study was to better understand the diversity of erm genes and to target the gene variants that could be indicative of resistance originating from manure and spreading to agricultural soil and water environments. MATERIALS AND METHODS Phylogenetic analysis of erm genes Gene sequencing sharing high similarity to ermA, ermB, ermC and ermF were obtained from publicly available databases. The Ribosomal Database Project Fungene Repository (Fish et al.2013) was used to obtain ermB- and ermC-associated sequences. It was required that sequences share 97% amino acid sequence coverage to established HMM protein models for Fungene gene families “Resfam_ermA”, “Resfam_ermB” and “Resfam_ermC” (Version 8.8). Additionally, ermF gene nucleotide sequences were obtained from proteins listed in the ARDB-Antibiotic Resistance Genes Database (version 1.1, July 3, 2009) (Liu and Pop 2009) and associated with the annotation “ermF”. All erm-associated sequences were combined and clustered at 99% nucleotide similarity using CD-HIT (v4.6.1c) (Li and Godzik 2006; Fu et al.2012), resulting in 66 unique clusters. One representative sequence for each cluster was identified by CD-HIT and was aligned using Muscle (v3.8.31) (Edgar 2004) with the following parameters: gap open –400, gap extend 0, clustering method UPGMB. A maximum-likelihood phylogenetic tree was constructed from this alignment using FastTree (v2.1.8) (Price, Dehal and Arkin 2010) with default parameters. Taxonomy was identified based on annotations in the NCBI non-redundant nucleotide database (NCBI Resource Coordinators 2017). To consider an erm gene sequence to be associated with a previously targeted PCR primer sequence, both forward and reverse primers were required to share 100% nucleotide similarity over a minimum of 17 bp of the primer length. Manure metagenomic datasets The presence of erm genes was characterized in swine and cattle manures. For swine manure, DNA was extracted from two biological replicates (three technical replicates each) of swine manure originating from Iowa State University's Northeast Research and Demonstration Farm, near Nashua, IA (43.0° N, 92.5° W). Metagenomic libraries were prepared and sequenced at Iowa State University DNA Sequencing Facility on a HiSeq 2500 instrument (Illumina, San Diego, CA) according to manufacturer's instructions. These datasets are deposited in the NCBI SRA as project SRP109083 (Table S1, Supporting Information). Sequences were compared to representatives of erm genes described above (BLAST, v2.4.0+) (Camacho et al.2009). Sequences were annotated as erm genes if they matched the representative sequence within a cluster with a minimum e-value of 1e-5 and if both paired-end reads matched the same representative target. The abundance of erm sequences in each sample was calculated as the total number of reads meeting these criteria. Cattle manure metagenomes were obtained from a previously published study of antibiotic resistant genes in commercial cattle as they moved through the process of beef production from feedlot entry to slaughter (Noyes et al.2016). The presence of erm sequences in these samples was determined by the total number of reads that shared sequence homology (BLAST, v2.4.0+, e-value 1e-5) to the best matched erm representative sequence. Similarly, metagenomes from human-impacted (Fitzpatrick and Walsh 2016) and pristine environment (Staley et al.2013) were aligned against erm sequences and considered a match if alignment scores resulted in e-value scores of at least 1e-5. RESULTS A total of 5648 erm DNA sequences were identified from annotated genes based on sequence similarity to well-characterized erm genes and were clustered at 99% nucleotide similarity to identify 66 unique erm variant clusters. A representative sequence of each cluster was defined as the longest consensus sequence in each cluster as determined by a greedy incremental clustering algorithm (see Methods, Table 1). These representative sequences were aligned and used to construct a phylogenetic tree describing the diversity of erm genes (Fig. 1). Based on sequence homology, the resulting erm gene clusters encompass the majority of erm genes studied in previous literature: ermA, ermB, ermC, ermF, ermG and ermT (reviewed in Roberts et al.2008). Among the gene clusters, a cluster associated with ermA was the most represented in our erm gene database (Cluster 15, 3542 genes), followed by an ermB cluster (Cluster 18, 1387 genes), and then an ermC cluster (Cluster 30, 399 genes). These three gene clusters comprise 94% of erm genes and are evidence to biases in the previous characterization of erm genes towards specific gene variants. Beyond the three most abundant gene clusters, the next most represented cluster (Cluster 11, 50 genes) is not well-characterized (e.g. most similar to unannotated erm gene clusters in our database) and is most closely related to genes belonging to Streptococcus agalactiae strain TR7 (100% nucleotide identity). Most clusters (53 of 66) are associated with five or less gene sequences, demonstrating that much of what we know of specific erm gene families is based on very few characterized representatives. Figure 1. View largeDownload slide Maximum likelihood phylogenetic tree of 66 erm sequence clusters based on 99% nucleotide similarity of 5648 DNA sequences extracted from known erm genes described in existing databases. Clusters that contain gene targets from existing PCR primers (see Table 2) are highlighted in color. The relative number of sequences comprising each cluster among the 5648 DNA sequences is also shown. Figure 1. View largeDownload slide Maximum likelihood phylogenetic tree of 66 erm sequence clusters based on 99% nucleotide similarity of 5648 DNA sequences extracted from known erm genes described in existing databases. Clusters that contain gene targets from existing PCR primers (see Table 2) are highlighted in color. The relative number of sequences comprising each cluster among the 5648 DNA sequences is also shown. Table 1. Erm gene clusters identified from 5648 erm sequences. For each cluster, the most representative gene is referenced by its NCBI accession number in NCBI nucleotide and protein databases. Cluster (this study)  NCBI protein accession no.  NCBI nucleotide accession no.  Description in NCBI GenBank  Organism  Cluster 0  BAJ34818  AB601890  Erythromycin resistance protein  Photobacterium damselae subsp. piscicida  Cluster 1  KNF08983  LGSS01000004  rRNA (adenine-N(6)-)-methyltransferase  Clostridium purinilyticum  Cluster 2  AFS78141  CP003326  rRNA (adenine-N(6)-)-methyltransferase  Clostridium acidurici 9a  Cluster 3  ABW20380  CP000853  rRNA (adenine-N(6)-)-methyltransferase  Alkaliphilus oremlandii OhILAs  Cluster 4  KKS60599  LCDU01000003  rRNA (Adenine-N(6)-)-methyltransferase  Candidate division WWE3 bacterium GW2011_GWF2_42_42  Cluster 5  KKS35651  LCCU01000032  rRNA (Adenine-N(6)-)-methyltransferase  Candidate division WWE3 bacterium GW2011_GWF1_42_14  Cluster 6  AJB79756  CP010391  Hypothetical protein  Klebsiella pneumoniae  Cluster 7  EKD94896  AMFJ01010665  Hypothetical protein  Uncultured bacterium  Cluster 8  KKU26033  LCLY01000007  rRNA (Adenine-N(6)-)-methyltransferase  Microgenomates group bacterium GW2011_GWA2_46_16  Cluster 9  CCQ93859  CARA01000062  rRNA adenine N-6-methyltransferase  Clostridium ultunense Esp  Cluster 10  BAP00917  AP013353  Dimethyladenosine transferase  Mycoplasma californicum HAZ160_1  Cluster 11  CNJ04734  CQCN01000003  Dimethyladenosine transferase  Streptococcus agalactiae  Cluster 12  AAA27431  M17808.1  ermF  Bacteroides fragilis  Cluster 13  AAA63165  M62487.1  ermF  Bacteroides fragilis  Cluster 14  EEO52603  ACAB02000055.1  ermF  Bacteroides sp. D1  Cluster 15  CCJ25599  HE579073  rRNA adenine N-6-methyltransferase  Staphylococcus aureus subsp. aureus ST228  Cluster 16  CCX90994  CAXH010000024  Ribosomal RNA small subunit methyltransferase A  Succinatimonas sp. CAG:777  Cluster 17  EGV00599  AFXA01000001  Dimethyladenosine transferase rRNA modification enzyme  Mycoplasma columbinum SF7  Cluster 18  EFY03905  AEVN01000118  rRNA adenine N-6-methyltransferase  Phascolarctobacterium succinatutens YIT 12067  Cluster 19  ACB90575  CP001033  Erythromycin ribosome methylase  Streptococcus pneumoniae CGSP14  Cluster 20  EJY36237  AMBI01000188  rRNA adenine N-6-methyltransferase  Enterococcus faecium 510  Cluster 21  AKB11102  CP011096  16S rRNA methyltransferase  Mycoplasma synoviae ATCC 25204  Cluster 22  ACD66486  CP001080  Dimethyladenosine transferase  Sulfurihydrogenibium sp. YO3AOP1  Cluster 23  EIB96299  AICL01000010  rRNA methylase  Lactobacillus salivarius SMXD51  Cluster 24  EEP60650  ABZS01000069  Dimethyladenosine transferase  Sulfurihydrogenibium yellowstonense SS-5  Cluster 25  AFV15157  JQ655732  Erythromycin  Clostridium perfringens  Cluster 26  KDE45359  JFKK01000007  16S rRNA methyltransferase  Mycoplasma hyosynoviae  Cluster 27  ADM89794  CP002161  Putative dimethyladenosine transferase  Candidatus Zinderia insecticola CARI  Cluster 28  KER55751  JPHP01000035  SAM-dependent methlyltransferase  Bacteroides fragilis  Cluster 29  AAR27225  AY357120  N-methyltransferase  Streptococcus pyogenes  Cluster 30  AIU96746  KF831357  ErmC  Staphylococcus aureus  Cluster 31  ACG57739  CP001130  Ribosomal RNA adenine methylase transferase  Hydrogenobaculum sp. Y04AAS1  Cluster 32  AAO20906  AF205068  erm44  Lactobacillus reuteri  Cluster 33  AFH70049  CP003045  rRNA adenine N-6-methyltransferase  Staphylococcus aureus subsp. aureus 71193  Cluster 34  ACC94310  EU595407  ErmB  Uncultured Enterococcus sp.  Cluster 35  AAF86219  AF242872  ErmB  Enterococcus faecium  Cluster 36  CDZ75671  LM997412  rRNA adenine N-6-methyltransferase  Peptoniphilus sp. ING2-D1G  Cluster 37  EOK35943  ASEN01000042  rRNA adenine N-6-methyltransferase  Enterococcus faecalis EnGen0332  Cluster 38  EZX88180  JIYN01000027  rRNA adenine N-6-methyltransferase  Staphylococcus aureus GD2010-052  Cluster 39  CEI83544  CDGG01000001  rRNA adenine N-6-methyltransferase  Oceanobacillus oncorhynchi  Cluster 40  CEJ95855  LN680996  23S RNA methylase for macrolide-lincosamide-streptogramin B resistance  Staphylococcus fleurettii  Cluster 41  CAD32685  AJ488494  Erythromycin resistance protein  Lactobacillus fermentum  Cluster 42  EIY35985  AGXG01000023  rRNA adenine N-6-methyltransferase  Bacteroides cellulosilyticus CL02T12C19  Cluster 43  EDV04163  ABJL02000008  Hypothetical protein  Bacteroides intestinalis DSM 17393  Cluster 44  AHH55321  KC790462  rRNA adenine N-6-methyltransferase  Streptococcus suis  Cluster 45  BAB20748  AB014481  ErmGM  Staphylococcus aureus  Cluster 46  KAC49299  JIQI01000041  rRNA adenine N-6-methyltransferase  Staphylococcus aureus VET0243R  Cluster 47  CDQ41560  CCDP010000003  rRNA adenine N-6-methyltransferase  Virgibacillus massiliensis  Cluster 48  AGK85210  KC405064  Erythromycin ribosome methylase  Haemophilus parasuis  Cluster 49  BAC12877  BA000028  Erythromycin resistance protein  Oceanobacillus iheyensis HTE831  Cluster 50  AAC37034  L42817  rRNA methyltransferase  Bacteroides thetaiotaomicron  Cluster 51  EJD65709  AFSU01000133  Hypothetical protein  Bacillus sp. 916  Cluster 52  CAJ43792  AM159501  rRNA methylase  Staphylococcus saprophyticus  Cluster 53  EZS04927  JILJ01000152  rRNA adenine N-6-methyltransferase  Staphylococcus aureus VET0436R  Cluster 54  CCG55258  HE775264  Ribosomal RNA adenine methylase Erm(43)  Staphylococcus lentus  Cluster 55  EJY20540  AMBD01000117  rRNA adenine N-6-methyltransferase  Enterococcus faecium C1904  Cluster 56  CAE18145  AJ579365  rRNA methylase  Staphylococcus sciuri  Cluster 57  KIJ86993  JXBG01000010  SAM-dependent methlyltransferase  Staphylococcus saprophyticus  Cluster 58  EKB53568  AGZE01000039  Hypothetical protein  Facklamia ignava CCUG 37419  Cluster 59  CDS14986  LK392593  23S rRNA methylase  Staphylococcus xylosus  Cluster 60  AJK31391  KJ728534  Ribosomal RNA adenine methylase variant  Staphylococcus xylosus  Cluster 61  AJK31388  KJ728533  Ribosomal RNA adenine methylase  Staphylococcus saprophyticus  Cluster 62  KIO72601  JXLU01000090  rRNA adenine N-6-methyltransferase  Bacillus thermoamylovorans  Cluster 63  KKD22675  LATV01000011  SAM-dependent methlyltransferase  Staphylococcus cohnii subsp. cohnii  Cluster 64  EVJ59956  JBER01000028  rRNA adenine N-6-methyltransferase  Staphylococcus aureus GGMC6053  Cluster 65  EDU98728  ABIY02000132.1  ermF  Bacteroides coprocola DSM 17136  Cluster (this study)  NCBI protein accession no.  NCBI nucleotide accession no.  Description in NCBI GenBank  Organism  Cluster 0  BAJ34818  AB601890  Erythromycin resistance protein  Photobacterium damselae subsp. piscicida  Cluster 1  KNF08983  LGSS01000004  rRNA (adenine-N(6)-)-methyltransferase  Clostridium purinilyticum  Cluster 2  AFS78141  CP003326  rRNA (adenine-N(6)-)-methyltransferase  Clostridium acidurici 9a  Cluster 3  ABW20380  CP000853  rRNA (adenine-N(6)-)-methyltransferase  Alkaliphilus oremlandii OhILAs  Cluster 4  KKS60599  LCDU01000003  rRNA (Adenine-N(6)-)-methyltransferase  Candidate division WWE3 bacterium GW2011_GWF2_42_42  Cluster 5  KKS35651  LCCU01000032  rRNA (Adenine-N(6)-)-methyltransferase  Candidate division WWE3 bacterium GW2011_GWF1_42_14  Cluster 6  AJB79756  CP010391  Hypothetical protein  Klebsiella pneumoniae  Cluster 7  EKD94896  AMFJ01010665  Hypothetical protein  Uncultured bacterium  Cluster 8  KKU26033  LCLY01000007  rRNA (Adenine-N(6)-)-methyltransferase  Microgenomates group bacterium GW2011_GWA2_46_16  Cluster 9  CCQ93859  CARA01000062  rRNA adenine N-6-methyltransferase  Clostridium ultunense Esp  Cluster 10  BAP00917  AP013353  Dimethyladenosine transferase  Mycoplasma californicum HAZ160_1  Cluster 11  CNJ04734  CQCN01000003  Dimethyladenosine transferase  Streptococcus agalactiae  Cluster 12  AAA27431  M17808.1  ermF  Bacteroides fragilis  Cluster 13  AAA63165  M62487.1  ermF  Bacteroides fragilis  Cluster 14  EEO52603  ACAB02000055.1  ermF  Bacteroides sp. D1  Cluster 15  CCJ25599  HE579073  rRNA adenine N-6-methyltransferase  Staphylococcus aureus subsp. aureus ST228  Cluster 16  CCX90994  CAXH010000024  Ribosomal RNA small subunit methyltransferase A  Succinatimonas sp. CAG:777  Cluster 17  EGV00599  AFXA01000001  Dimethyladenosine transferase rRNA modification enzyme  Mycoplasma columbinum SF7  Cluster 18  EFY03905  AEVN01000118  rRNA adenine N-6-methyltransferase  Phascolarctobacterium succinatutens YIT 12067  Cluster 19  ACB90575  CP001033  Erythromycin ribosome methylase  Streptococcus pneumoniae CGSP14  Cluster 20  EJY36237  AMBI01000188  rRNA adenine N-6-methyltransferase  Enterococcus faecium 510  Cluster 21  AKB11102  CP011096  16S rRNA methyltransferase  Mycoplasma synoviae ATCC 25204  Cluster 22  ACD66486  CP001080  Dimethyladenosine transferase  Sulfurihydrogenibium sp. YO3AOP1  Cluster 23  EIB96299  AICL01000010  rRNA methylase  Lactobacillus salivarius SMXD51  Cluster 24  EEP60650  ABZS01000069  Dimethyladenosine transferase  Sulfurihydrogenibium yellowstonense SS-5  Cluster 25  AFV15157  JQ655732  Erythromycin  Clostridium perfringens  Cluster 26  KDE45359  JFKK01000007  16S rRNA methyltransferase  Mycoplasma hyosynoviae  Cluster 27  ADM89794  CP002161  Putative dimethyladenosine transferase  Candidatus Zinderia insecticola CARI  Cluster 28  KER55751  JPHP01000035  SAM-dependent methlyltransferase  Bacteroides fragilis  Cluster 29  AAR27225  AY357120  N-methyltransferase  Streptococcus pyogenes  Cluster 30  AIU96746  KF831357  ErmC  Staphylococcus aureus  Cluster 31  ACG57739  CP001130  Ribosomal RNA adenine methylase transferase  Hydrogenobaculum sp. Y04AAS1  Cluster 32  AAO20906  AF205068  erm44  Lactobacillus reuteri  Cluster 33  AFH70049  CP003045  rRNA adenine N-6-methyltransferase  Staphylococcus aureus subsp. aureus 71193  Cluster 34  ACC94310  EU595407  ErmB  Uncultured Enterococcus sp.  Cluster 35  AAF86219  AF242872  ErmB  Enterococcus faecium  Cluster 36  CDZ75671  LM997412  rRNA adenine N-6-methyltransferase  Peptoniphilus sp. ING2-D1G  Cluster 37  EOK35943  ASEN01000042  rRNA adenine N-6-methyltransferase  Enterococcus faecalis EnGen0332  Cluster 38  EZX88180  JIYN01000027  rRNA adenine N-6-methyltransferase  Staphylococcus aureus GD2010-052  Cluster 39  CEI83544  CDGG01000001  rRNA adenine N-6-methyltransferase  Oceanobacillus oncorhynchi  Cluster 40  CEJ95855  LN680996  23S RNA methylase for macrolide-lincosamide-streptogramin B resistance  Staphylococcus fleurettii  Cluster 41  CAD32685  AJ488494  Erythromycin resistance protein  Lactobacillus fermentum  Cluster 42  EIY35985  AGXG01000023  rRNA adenine N-6-methyltransferase  Bacteroides cellulosilyticus CL02T12C19  Cluster 43  EDV04163  ABJL02000008  Hypothetical protein  Bacteroides intestinalis DSM 17393  Cluster 44  AHH55321  KC790462  rRNA adenine N-6-methyltransferase  Streptococcus suis  Cluster 45  BAB20748  AB014481  ErmGM  Staphylococcus aureus  Cluster 46  KAC49299  JIQI01000041  rRNA adenine N-6-methyltransferase  Staphylococcus aureus VET0243R  Cluster 47  CDQ41560  CCDP010000003  rRNA adenine N-6-methyltransferase  Virgibacillus massiliensis  Cluster 48  AGK85210  KC405064  Erythromycin ribosome methylase  Haemophilus parasuis  Cluster 49  BAC12877  BA000028  Erythromycin resistance protein  Oceanobacillus iheyensis HTE831  Cluster 50  AAC37034  L42817  rRNA methyltransferase  Bacteroides thetaiotaomicron  Cluster 51  EJD65709  AFSU01000133  Hypothetical protein  Bacillus sp. 916  Cluster 52  CAJ43792  AM159501  rRNA methylase  Staphylococcus saprophyticus  Cluster 53  EZS04927  JILJ01000152  rRNA adenine N-6-methyltransferase  Staphylococcus aureus VET0436R  Cluster 54  CCG55258  HE775264  Ribosomal RNA adenine methylase Erm(43)  Staphylococcus lentus  Cluster 55  EJY20540  AMBD01000117  rRNA adenine N-6-methyltransferase  Enterococcus faecium C1904  Cluster 56  CAE18145  AJ579365  rRNA methylase  Staphylococcus sciuri  Cluster 57  KIJ86993  JXBG01000010  SAM-dependent methlyltransferase  Staphylococcus saprophyticus  Cluster 58  EKB53568  AGZE01000039  Hypothetical protein  Facklamia ignava CCUG 37419  Cluster 59  CDS14986  LK392593  23S rRNA methylase  Staphylococcus xylosus  Cluster 60  AJK31391  KJ728534  Ribosomal RNA adenine methylase variant  Staphylococcus xylosus  Cluster 61  AJK31388  KJ728533  Ribosomal RNA adenine methylase  Staphylococcus saprophyticus  Cluster 62  KIO72601  JXLU01000090  rRNA adenine N-6-methyltransferase  Bacillus thermoamylovorans  Cluster 63  KKD22675  LATV01000011  SAM-dependent methlyltransferase  Staphylococcus cohnii subsp. cohnii  Cluster 64  EVJ59956  JBER01000028  rRNA adenine N-6-methyltransferase  Staphylococcus aureus GGMC6053  Cluster 65  EDU98728  ABIY02000132.1  ermF  Bacteroides coprocola DSM 17136  View Large Next, we evaluated the diversity of bacteria carrying these erm genes by identifying the taxonomic origin of potential bacterial hosts associated with each erm gene sequence (Table 1; Fig S1, Supporting Information). In general, the majority of known erm gene sequences were associated with Firmicutes (98%), followed by Proteobacteria (0.6%) and Bacterioidetes (0.6%). While ermF and ermG genes were observed to be carried by only Bacteriodetes, ermA, ermB, ermC and ermT genes were associated primarily with Firmicutes (Fig S1, Supporting Information). Within the Firmicutes, ermB genes were associated mainly with the order Lactobacillales, while ermA and ermT genes were associated with members of the Bacillales order (Fig S2, Supporting Information). These results demonstrate a wide range of potential host diversity for erm genes and highlight the impact of the choice of primer gene targets selecting for or against specific host bacteria. Historically, erm genes have been extensively targeted for qPCR quantification of gene abundances in the environment (Table 2), and we evaluated the ability of previously published PCR primers to detect the erm gene diversity described above by computationally hybridizing the primer sequences from the literature with the representative erm gene sequences in our database. Overall, published primer pairs were 100% similar to 25 of the representative sequences of erm clusters (Fig. 1). Generally, well-characterized gene clusters (e.g. containing the most known gene sequences) were observed to be associated with previous primer development. Several clusters were not associated with previously published primer targets, very likely due to the few well-characterized erm sequences within these clusters. Previously, observed diversity in natural samples have weak correlations with well-characterized genes (Choi et al.2017), suggesting that primer targets selected based on the most well-studied genes may not be effective in environmental samples. Table 2. Previously published PCR primer and gene targets for erm genes. Gene  Cluster  Primers design  Papers citing primers  ermA  15  Patterson et al.2007  b    n/aa  Sutcliffe et al.1996  Martel et al.2003a, Jackson et al.2004, Luthje and Schwarz 2006, Garofalo et al.2007, Chenier and Juteau 2009, Zou et al.2011, Di Cesare et al.2012, Hoang et al.2013, Lerma et al.2014    15, 56  Jensen et al.1999  Aarestrup et al.2000a,b, Jensen et al.2002, Petersen and Dalsgaard 2003, Whitehead and Cotta 2013    n/a  Chen et al.2007  Sharma et al.2009, Just et al.2011, Alexander et al.2011, Wang et al.2012, Holman and Chenier 2013, Wang et al.2015, Xu et al.2016    15, 56  Koike et al.2010  Ekizoglu et al.2013  ermB  n/a  Sutcliffe et al.1996  Martel et al.2003a,b, Cauwerts et al.2007, Ahmad et al.2011, Hoang et al.2013    18, 19, 20, 25, 28, 29, 32, 34, 35, 42  Jensen et al.1999  De Leener et al.2005, Whitehead and Cotta 2013    18, 19, 20, 25, 28, 29, 32, 34, 35, 42  Chen et al.2007  Sharma et al.2009, Chen et al.2010, Alexander et al.2011, Just et al.2011, Kalmokoff et al.2011, Negreanu et al.2012, Holman and Chenier 2013, Beukers et al.2015, Wang et al.2015, Sandberg and LaPara 2016, Xu et al.2016    18, 19, 20, 25, 29, 32, 35, 42  Patterson et al.2007  Knapp et al.2010    18, 19, 20, 25, 28, 29, 32, 34, 35, 42  Koike et al.2010  Ekizoglu et al.2013, Garder et al.2014, Joy et al.2013, Joy et al.2014, Soni et al.2015, Luby et al.2016  ermC  n/a  Sutcliffe et al.1996  Martel et al.2003b, Hoang et al.2013    30, 46  Jensen et al.1999  Ekizoglu et al.2013, Whitehead and Cotta 2013    23, 30, 46, 51, 52, 63, 64  Patterson et al.2007  Knapp et al.2010, Popowska et al.2012    30, 46, 51  Koike et al.2010  Luby et al.2016  ermF  12, 13  Chen et al.2007  Sharma et al.2009, Chen et al.2010, Alexander et al.2011, Kalmokoff et al.2011, Negreanu et al.2012, Wang et al.2012, Hoang et al.2013, Holman and Chenier 2013, Farenfeld et al.2014, Garder et al.2014, Luby et al.2016, Xu et al.2016    12, 13  Patterson et al.2007  Knapp et al.2010    12, 13  Koike et al.2010  Ekizoglu et al.2013, Joy et al.2013, Joy et al.2014    43, 50  Wang et al.2005  Wang et al.2005, Kalmokoff et al.2011  ermG  43, 50  Patterson et al.2007  N/A    43, 50  Koike et al.2010  Ekizoglu et al.2013  ermT  33, 41  Chen et al.2007  Sharma et al.2009, Alexander et al.2011, Kalmokoff et al.2011, Wang et al.2012, Hoang et al.2013, Garder et al.2014, Wang et al.2015  Gene  Cluster  Primers design  Papers citing primers  ermA  15  Patterson et al.2007  b    n/aa  Sutcliffe et al.1996  Martel et al.2003a, Jackson et al.2004, Luthje and Schwarz 2006, Garofalo et al.2007, Chenier and Juteau 2009, Zou et al.2011, Di Cesare et al.2012, Hoang et al.2013, Lerma et al.2014    15, 56  Jensen et al.1999  Aarestrup et al.2000a,b, Jensen et al.2002, Petersen and Dalsgaard 2003, Whitehead and Cotta 2013    n/a  Chen et al.2007  Sharma et al.2009, Just et al.2011, Alexander et al.2011, Wang et al.2012, Holman and Chenier 2013, Wang et al.2015, Xu et al.2016    15, 56  Koike et al.2010  Ekizoglu et al.2013  ermB  n/a  Sutcliffe et al.1996  Martel et al.2003a,b, Cauwerts et al.2007, Ahmad et al.2011, Hoang et al.2013    18, 19, 20, 25, 28, 29, 32, 34, 35, 42  Jensen et al.1999  De Leener et al.2005, Whitehead and Cotta 2013    18, 19, 20, 25, 28, 29, 32, 34, 35, 42  Chen et al.2007  Sharma et al.2009, Chen et al.2010, Alexander et al.2011, Just et al.2011, Kalmokoff et al.2011, Negreanu et al.2012, Holman and Chenier 2013, Beukers et al.2015, Wang et al.2015, Sandberg and LaPara 2016, Xu et al.2016    18, 19, 20, 25, 29, 32, 35, 42  Patterson et al.2007  Knapp et al.2010    18, 19, 20, 25, 28, 29, 32, 34, 35, 42  Koike et al.2010  Ekizoglu et al.2013, Garder et al.2014, Joy et al.2013, Joy et al.2014, Soni et al.2015, Luby et al.2016  ermC  n/a  Sutcliffe et al.1996  Martel et al.2003b, Hoang et al.2013    30, 46  Jensen et al.1999  Ekizoglu et al.2013, Whitehead and Cotta 2013    23, 30, 46, 51, 52, 63, 64  Patterson et al.2007  Knapp et al.2010, Popowska et al.2012    30, 46, 51  Koike et al.2010  Luby et al.2016  ermF  12, 13  Chen et al.2007  Sharma et al.2009, Chen et al.2010, Alexander et al.2011, Kalmokoff et al.2011, Negreanu et al.2012, Wang et al.2012, Hoang et al.2013, Holman and Chenier 2013, Farenfeld et al.2014, Garder et al.2014, Luby et al.2016, Xu et al.2016    12, 13  Patterson et al.2007  Knapp et al.2010    12, 13  Koike et al.2010  Ekizoglu et al.2013, Joy et al.2013, Joy et al.2014    43, 50  Wang et al.2005  Wang et al.2005, Kalmokoff et al.2011  ermG  43, 50  Patterson et al.2007  N/A    43, 50  Koike et al.2010  Ekizoglu et al.2013  ermT  33, 41  Chen et al.2007  Sharma et al.2009, Alexander et al.2011, Kalmokoff et al.2011, Wang et al.2012, Hoang et al.2013, Garder et al.2014, Wang et al.2015  a Primers did not hit any clusters. b No relevant citing papers. View Large We next evaluated the diversity of erm genes in 12 947 environmental metagenomes (Table S1, Supporting Information), resulting in the observation that significantly more erm genes are present in human-impacted environments (feces- and animal-associated soil and water) than in natural environments (Fig. 2). We also searched an additional 39 metagenomes originating from relatively pristine freshwaters along the Upper Mississippi River (Staley et al.2013, Table S1, Supporting Information), resulting in only 3 reads out of 716 million, sharing similarity (e-value < 1e-5) to erm genes. Combined, these results demonstrate that erm genes are rare in environments with minimal human impact and suggest that erm genes associated with feces or manure are ideal for tracking the spread and persistence of resistance through the environment. These results are also consistent with previous observations that manure contains abundant genes related to erm resistance and is a source of these genes into the environment (e.g. soil and water) (Chee-Sanford et al.2009; Koike et al.2010; Heuer, Schmitt and Smalla 2011; Joy et al.2013; Luby, Moorman and Soupir 2016). Figure 2. View largeDownload slide Average number of erm genes in metagenomes from various environments (see Table S1, Supporting Information). *For ermC gene, the average number of reads in animal-associated soil metagenomes was 1665 ± 659 reads. Figure 2. View largeDownload slide Average number of erm genes in metagenomes from various environments (see Table S1, Supporting Information). *For ermC gene, the average number of reads in animal-associated soil metagenomes was 1665 ± 659 reads. Consequently, we next identified erm genes in manure metagenomes. We aligned erm gene sequences against metagenomes derived from two large manure metagenomic studies (requiring nmanure > 3): swine manure collected near Nashua, IA (Luby, Moorman and Soupir 2016) and cattle manure from a previously published study (Noyes et al.2016). These manure metagenomes were strategically selected based on the number of biological replicates and sequencing depth. Three erm clusters comprised 46% and 45% of the total abundance of erm genes in swine and cattle manure, respectively (Table S1, Supporting Information). The genes associated with these most abundant clusters differed between swine and cattle manures. In swine metagenomes, sequences associated with the ermB gene cluster (Fig. 1, sharing 93%–99% similarity) captured 26% of all erm sequences, followed by ermG-associated sequences capturing 11% and ermA-associated sequences capturing 9%. In cattle metagenomes, sequences associated with ermF represented 23.5% of all erm abundances, followed by sequences associated with ermG capturing 12.4% and sequences associated with ermB capturing 9%. Only a subset of erm genes detected in manure are targeted by existing primer sets. Overall, a total of 25 out of the 66 erm clusters (40%) could be computationally detected with known primers (Table 1, Supporting Information), and these genes also encompass much of the total erm abundances observed in manure metagenomes. Collectively, if all primers were used, 74% and 85% of the total erm gene sequence diversity observed in swine and cattle metagenomes, respectively, could be detected, suggesting good coverage of these genes for PCR or qPCR assays. Specifically, in swine manure metagenomes, ermB primers could detect 29% of erm sequences, followed by ermF primers capturing 14% and ermG primers capturing 12% (Fig. 3). In cattle, ermF primers are the most effective, capturing 30% of erm sequences, followed by 21% with ermG primers, and 15% with ermB primers. Consequently, depending on the environmental sample in a study, in this case swine versus cattle manure, the choice of erm gene targets can significantly alter erm abundance estimations. For example, in swine manure, two times more erm gene abundance would be estimated if ermB primers were used instead of ermF primers. Even within the same gene clade, different primers could result in significant differences in abundance estimations, and this result is observed especially for ermC primers where a near two-fold difference in abundance estimations would result based on selection of primers from Patterson et al. (2007) versus Jensen, Frimodt-Moller and Aarestrup (1999). The selection of Patterson primers would result in the detection of genes from up to seven erm gene clusters over the two to three gene cluster detected with Jensen, Frimodt-Moller and Aarestrup (1999) or Koike et al. (2010) primers. Similar results are noted in the cattle manure, where ermC primers designed by Patterson capture 13% of the total abundance of erm sequences in the metagenomes, while Koike and Jensen primers only capture 4.4% and 2.1%, respectively. These results emphasize that the targeting of a specific erm gene, even within closely related gene variants, can significantly alter estimations of associated resistance in manures. Figure 3. View largeDownload slide Abundance of DNA sequences homologous to erm gene PCR primers as a percent of total erm abundance in swine and cattle metagenomes. Figure 3. View largeDownload slide Abundance of DNA sequences homologous to erm gene PCR primers as a percent of total erm abundance in swine and cattle metagenomes. Thus, overall, for swine manure, the most effective gene target based on abundance in swine metagenomes (26% of erm genes) originates from an ermB cluster (Cluster 25) and is associated with Clostridium perfringens. The next most abundant ermB cluster in swine (Cluster 19, most similar to a gene in Streptococcus pneumoniae CGSP14) represented only 2% of erm abundances. These results indicate that while ermB primers can target multiple strains (Fig. S1 and S2, Supporting Information), in these swine metagenomes, it is one gene cluster that specifically dominates. This gene cluster is also abundant in cattle manure metagenomes, though comprising less of total erm gene abundance (9%). Within our erm gene database, this particular sequence cluster is represented by a single gene representative and shares 100% similarity to experimental Clostridium acetobutylicum strains in the NCBI non-redundant gene database (mutant HQ683763.1 and clone HQ25744.1). The overall lack of similar homologous genes in NCBI nr suggests that this specific ermB gene is abundant in manures but is a gene for which we have few sequenced representatives. We identified this gene during our exploration of the effectiveness of current primers on manure metagenomes, and our observations suggest that this gene would benefit from further study given its prevalence. DISCUSSION Over the past 20 years, an abundance of literature has been published quantifying macrolide resistance in agricultural landscapes using qPCR approaches. However, these previous studies often use primers for erm genes designed in only a handful of publications (Table 2). Our study found that current published primer sets, used on their own, are effective at capturing only a subset of the erm diversity in manure samples. For example, if only one primer set were used, less than one-third of erm genes would be detected. To increase our ability to detect erm genes in agricultural systems, we identified the most abundant erm clusters in both swine and cattle manures, identifying the best gene targets for future studies. These genes and their associated primers are recommended for high-throughput qPCR assays that can scale the detection and quantification of these genes for antibiotic gene surveillance. In all amplicon assays, quantifying environmental abundances of gene targets is limited by the effectiveness of primer design. The results presented here emphasize that estimates of abundances of a gene of interest cannot simply be based on primers to genes that have previously been successfully detected. Rather, genes appropriate for antibiotic gene surveillance should be indicative of the spread of resistance (e.g. originate from manure but lacking from pristine environments), representative of diverse hosts (especially those with clinical risks) and accurately represent gene abundances in environmental samples. Our specific effort targeted the erm gene and evaluated the effectiveness of previously published primers sets. The increasing availability of metagenomes makes these evaluations possible, as demonstrated in this study. Although metagenomic sequencing advances will continue to provide powerful tools to understand the broad diversity of resistance in environments, metagenomes are limited by both detection rate and resolution. Short read lengths, the difficulty of assembling many resistance genes (because of their common association with mobile elements containing repeated sequences) and their presence in multiple bacterial hosts challenges the detection of resistance genes using metagenomics. Going forward, high-throughput amplicon assays with strategic gene targets and primer designs are a complementary alternative to help fill these gaps and help us understand the movement of resistance genes among complex environments. SUPPLEMENTARY DATA Supplementary data are available at FEMSEC online. FUNDING This project was supported by the AFRI food safety from the USDA National Institute of Food and Agriculture [grant no. 2016-68003-24604] and the National Pork Board [contract 13-051]. Conflict of interest. None declared. REFERENCES Aarestrup FM, Agersø Y, Ahrens P et al.  . Antimicrobial susceptibility and presence of resistance genes in staphylococci from poultry. Vet Microbiol  2000a; 74: 353– 364. Google Scholar CrossRef Search ADS   Aarestrup FM, Kruse H, Tast E et al.  . Associations between the use of antimicrobial agents for growth promotion and the occurrence of resistance among Enterococcus faecium from broilers and pigs in Denmark, Finland, and Norway. Microb Drug Resist Epidemiol Dis  2000b; 6: 63– 70. Google Scholar CrossRef Search ADS   Ahmad A, Ghosh A, Schal C et al.  . Insects in confined swine operations carry a large antibiotic resistant and potentially virulent enterococcal community. BMC Microbiol  2011; 11: 23. Google Scholar CrossRef Search ADS PubMed  Alexander TW, Yanke JL, Reuter T et al.  . Longitudinal characterization of antimicrobial resistance genes in feces shed from cattle fed different subtherapeutic antibiotics. BMC Microbiol  2011; 11: 19. Google Scholar CrossRef Search ADS PubMed  Beukers AG, Zaheer R, Cook SR et al.  . Effect of in-feed administration and withdrawal of tylosin phosphate on antibiotic resistance in enterococci isolated from feedlot steers. Front Microbiol  2015; 6: 483. Google Scholar CrossRef Search ADS PubMed  Camacho C, Coulouris G, Avagyan V et al.  . BLAST+: architecture and applications. BMC Bioinformatics  2009; 10: 421. Google Scholar CrossRef Search ADS PubMed  Cauwerts K, Decostere A, De Graef EM et al.  . High prevalence of tetracycline resistance in Enterococcus isolates from broilers carrying the erm (B) gene. Avian Pathol  2007; 36: 395– 399. Google Scholar CrossRef Search ADS PubMed  Di Cesare A, Vignaroli C, Luna GM et al.  . Antibiotic-resistant enterococci in seawater and sediments from a coastal fish farm. Microb Drug Resist  2012; 18: 502– 509. Google Scholar CrossRef Search ADS PubMed  Chee-Sanford JC, Mackie RI, Koike S et al.  . Fate and transport of antibiotic residues and antibiotic resistance genes following land application of manure waste. J Environ Qual  2009; 38: 1086. Google Scholar CrossRef Search ADS PubMed  Chen J, Michel FC, Sreevatsan S et al.  . Occurrence and persistence of erythromycin resistance genes (erm) and tetracycline resistance genes (tet) in waste treatment systems on swine farms. Microb Ecol  2010; 60: 479– 486. Google Scholar CrossRef Search ADS PubMed  Chen J, Yu Z, Michel FC et al.  . Development and application of real-time PCR assays for quantification of erm genes conferring resistance to Macrolides-Lincosamides-Streptogramin B in livestock manure and manure management systems. Appl Environ Microb  2007; 73: 4407– 16. Google Scholar CrossRef Search ADS   Chénier MR, Juteau P. Fate of chlortetracycline- and tylosin-resistant bacteria in an aerobic thermophilic sequencing batch reactor treating swine waste. Microb Ecol  2009; 58: 86– 97. Google Scholar CrossRef Search ADS PubMed  Choi J, Yang F, Stepanauskas R et al.  . Strategies to improve reference databases for soil microbiomes. ISME J  2017; 11: 829– 34. Google Scholar CrossRef Search ADS PubMed  Edgar RC. MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res  2004; 32: 1792– 7. Google Scholar CrossRef Search ADS PubMed  Ekizoglu M, Koike S, Krapac I et al.  . Phenotypic and genotypic characterization of antibiotic-resistant soil and manure bacteria adjacent to swine production facilities. Turkish J Vet Anim Sci  2013; 37: 504– 511. Google Scholar CrossRef Search ADS   Fahrenfeld N, Knowlton K, Krometis LA et al.  . Effect of manure application on abundance of antibiotic resistance genes and their attenuation rates in soil: Field-scale mass balance approach. Environ Sci Technol  2014; 48: 2643– 50. Google Scholar CrossRef Search ADS PubMed  Fang H, Wang H, Cai L et al.  . Prevalence of antibiotic resistance genes and bacterial pathogens in long-term manured greenhouse soils as revealed by metagenomic survey. Environ Sci Technol  2015; 49: 1095– 104. Google Scholar CrossRef Search ADS PubMed  Fish JA, Chai B, Wang Q et al.  . FunGene: The Functional Gene Pipeline and Repository . 2013, DOI: https://doi.org/10.3389/fmicb.2013.00291. Fitzpatrick D, Walsh F. Antibiotic resistance genes across a wide variety of metagenomes. FEMS Microbiol Ecol  2016; 92: fiv168. Google Scholar CrossRef Search ADS PubMed  Fu L, Niu B, Zhu Z et al.  . CD-HIT: accelerated for clustering the next-generation sequencing data. Bioinformatics  2012; 28: 3150– 2. Google Scholar CrossRef Search ADS PubMed  Garder JL, Moorman TB, Soupir ML. Transport and persistence of tylosin-resistant enterococci, genes, and tylosin in soil and drainage water from fields receiving swine manure. J Environ Qual  2014; 43: 1484. Google Scholar CrossRef Search ADS PubMed  Garofalo C, Vignaroli C, Zandri G et al.  . Direct detection of antibiotic resistance genes in specimens of chicken and pork meat. Int J Food Microbiol  2007; 113: 75– 83. Google Scholar CrossRef Search ADS PubMed  Gillings MR, Gaze WH, Pruden A et al.  . Using the class 1 integron-integrase gene as a proxy for anthropogenic pollution. ISME J  2015; 9: 1269– 79. Google Scholar CrossRef Search ADS PubMed  Heuer H, Schmitt H, Smalla K. Antibiotic resistance gene spread due to manure application on agricultural fields. Curr Opin Microbiol  2011; 14: 236– 43. Google Scholar CrossRef Search ADS PubMed  Hoang TTT, Soupir ML, Liu P et al.  . Occurrence of tylosin-resistant enterococci in swine manure and tile drainage systems under no-till management. water, air. Soil Pollut  2013; 224: 1754. Google Scholar CrossRef Search ADS   Holman DB, Chénier MR, PJ F-C et al.  . Impact of subtherapeutic administration of tylosin and chlortetracycline on antimicrobial resistance in farrow-to-finish swine. FEMS Microbiol Ecol  2013; 85: 1– 13. Google Scholar CrossRef Search ADS PubMed  Jackson CR, Fedorka-Cray PJ, Barrett JB et al.  . Effects of tylosin use on erythromycin resistance in enterococci isolated from swine. Appl Environ Microbiol  2004; 70: 4205– 10. Google Scholar CrossRef Search ADS PubMed  Jensen LB, Agersø Y, Sengeløv G. Presence of erm genes among macrolide-resistant Gram-positive bacteria isolated from Danish farm soil. Environ Int  2002; 28: 487– 491. Google Scholar CrossRef Search ADS PubMed  Jensen LB, Frimodt-Moller N, Aarestrup FM. Presence of erm gene classes in Gram-positive bacteria of animal and human origin in Denmark. FEMS Microbiol Lett  1999; 170: 151– 8. Google Scholar CrossRef Search ADS PubMed  Just NA, Létourneau V, Kirychuk SP et al.  . Potentially pathogenic bacteria and antimicrobial resistance in bioaerosols from cage-housed and floor-housed poultry operations. Ann Occup Hyg  2011; 71: 6926– 33. Joy SR, Bartelt-Hunt SL, Snow DD et al.  . Fate and transport of antimicrobials and antimicrobial resistance genes in soil and runoff following land application of swine manure slurry. Environ Sci Technol  2013; 47: 12081– 8. Google Scholar CrossRef Search ADS PubMed  Joy SR, Li X, Snow DD et al.  . Fate of antimicrobials and antimicrobial resistance genes in simulated swine manure storage. Sci Total Environ  2014; 481: 69– 74. Google Scholar CrossRef Search ADS PubMed  Kalmokoff M, Waddington LM, Thomas M et al.  . Continuous feeding of antimicrobial growth promoters to commercial swine during the growing/finishing phase does not modify faecal community erythromycin resistance or community structure. J Appl Microbiol  2011; 110: 1414– 1425. Google Scholar CrossRef Search ADS PubMed  Karkman A, Johnson TA, Lyra C et al.  . High-throughput quantification of antibiotic resistance genes from an urban wastewater treatment plant. FEMS Microbiol Ecol  2016; 92: 1– 7. Google Scholar CrossRef Search ADS   Knapp CW, Dolfing J, Ehlert PAI et al.  . Evidence of increasing antibiotic resistance gene abundances in archived soils since 1940. Environ Sci Technol  2010; 44: 580– 7. Google Scholar CrossRef Search ADS PubMed  Koike S, Aminov RI, Yannarell AC et al.  . Molecular ecology Of macrolide–lincosamide–streptogramin B methylases in waste lagoons and subsurface waters associated with swine production. Microb Ecol  2010; 59: 487– 98. Google Scholar CrossRef Search ADS PubMed  Leclercq R, Courvalin P. Bacterial resistance to macrolide, lincosamide, and streptogramin antibiotics by target modification. Antimicrob Agents Chemother  1991; 35: 1267– 72. Google Scholar CrossRef Search ADS PubMed  Leener EDe, Martel A, De Graef E. Molecular analysis of human, porcine, and poultry Enterococcus faecium isolates and their erm (B) genes. Appl  2005. Available at http://aem.asm.org/content/71/5/2766.short (verified 18 July 2017). Lerma LL, Benomar N, del M et al.  . Antibiotic multiresistance analysis of mesophilic and psychrotrophic pseudomonas spp. isolated from goat and lamb slaughterhouse surfaces throughout the meat production process. Appl Environ Microbiol  2014; 80: 6792– 6806. Google Scholar CrossRef Search ADS PubMed  Li B, Yang Y, Ma L et al.  . Metagenomic and network analysis reveal wide distribution and co-occurrence of environmental antibiotic resistance genes. ISME J  2015; 9: 2490– 502. Google Scholar CrossRef Search ADS PubMed  Li W, Godzik A. Cd-hit: a fast program for clustering and comparing large sets of protein or nucleotide sequences. Bioinformatics  2006; 22: 1658– 9. Google Scholar CrossRef Search ADS PubMed  Liu B, Pop M. ARDB–Antibiotic resistance genes database. Nucleic Acids Res  2009; 37: D443– 7. Google Scholar CrossRef Search ADS PubMed  Luby EM, Moorman TB, Soupir ML. Fate and transport of tylosin-resistant bacteria and macrolide resistance genes in artificially drained agricultural fields receiving swine manure. Sci Total Environ  2016; 550: 1126– 33. Google Scholar CrossRef Search ADS PubMed  Luthje P, Schwarz S. Antimicrobial resistance of coagulase-negative staphylococci from bovine subclinical mastitis with particular reference to macrolide-lincosamide resistance phenotypes and genotypes. J Antimicrob Chemother  2006; 57: 966– 969. Google Scholar CrossRef Search ADS PubMed  Martel A, Devriese L, Decostere A et al.  . Presence of macrolide resistance genes in streptococci and enterococci isolated from pigs and pork carcasses. Int J Food Microbiol  2003a; 84: 27– 32. Google Scholar CrossRef Search ADS   Martel A, Meulenaere V, Devriese LA et al.  . Macrolide and lincosamide resistance in the gram-positive nasal and tonsillar flora of pigs. Microb Drug Resist  2003b; 9: 293– 7. Google Scholar CrossRef Search ADS   McEwen SA, Fedorka-Cray PJ. Antimicrobial use and resistance in animals. Clin Infect Dis  2002; 34: S93– 106. Google Scholar CrossRef Search ADS PubMed  Muziasari WI, Pärnänen K, Johnson TA et al.  . Aquaculture changes the profile of antibiotic resistance and mobile genetic element associated genes in Baltic Sea sediments. FEMS Microbiol Ecol  2016; 92: fiw052. Google Scholar CrossRef Search ADS PubMed  NCBI Resource Coordinators. Database resources of the National Center for Biotechnology Information. Nucleic Acids Res  2017; 45: D12– 7. CrossRef Search ADS PubMed  Negreanu Y, Pasternak Z, Jurkevitch E et al.  . Impact of treated wastewater irrigation on antibiotic resistance in agricultural soils. Environ Sci Technol  2012; 46: 4800– 4808. Google Scholar CrossRef Search ADS PubMed  Noyes N, Yang X, Linke L et al.  . Resistome diversity in cattle and the environment decreases during beef production. Elife  2016; 5: e13195. Google Scholar CrossRef Search ADS PubMed  Patterson AJ, Colangeli R, Spigaglia P et al.  . Distribution of specific tetracycline and erythromycin resistance genes in environmental samples assessed by macroarray detection. Environ Microbiol  2007; 9: 703– 15. Google Scholar CrossRef Search ADS PubMed  Petersen A, Dalsgaard A. Species composition and antimicrobial resistance genes of Enterococcus spp., isolated from integrated and traditional fish farms in Thailand. Environ Microbiol  2003; 5: 395– 402. Google Scholar CrossRef Search ADS PubMed  Popowska M, Rzeczycka M, Miernik A et al.  . Influence of soil use on prevalence of tetracycline, streptomycin, and erythromycin resistance and associated resistance genes. Antimicrob Agents Chemother  2012; 56: 1434– 43. Google Scholar CrossRef Search ADS PubMed  Price MN, Dehal PS, Arkin AP. FastTree 2 – Approximately maximum-likelihood trees for large alignments. PLoS One . 2010; 5: e9490 Pyörälä S, Baptiste KE, Catry B et al.  . Macrolides and lincosamides in cattle and pigs: Use and development of antimicrobial resistance. Vet J  2014; 200: 230– 9. Google Scholar CrossRef Search ADS PubMed  Roberts MC, Dueck M, Hoban D et al.  . Update on macrolide-lincosamide-streptogramin, ketolide, and oxazolidinone resistance genes. FEMS Microbiol Lett  2008; 282: 147– 59. Google Scholar CrossRef Search ADS PubMed  Roberts MC, Sutcliffe J, Courvalin P et al.  . Nomenclature for macrolide and macrolide-lincosamide-streptogramin B resistance determinants. Antimicrob Agents Ch  1999; 43: 2823– 30. Sandberg KD, LaPara TM. The fate of antibiotic resistance genes and class 1 integrons following the application of swine and dairy manure to soils (K Smalla, Ed.). FEMS Microbiol Ecol  2016; 92: fiw001. Google Scholar CrossRef Search ADS PubMed  Sharma R, Larney FJ, Chen J et al.  . Selected Antimicrobial resistance during composting of manure from cattle administered sub-therapeutic antimicrobials. J Environ Qual  2009; 38: 567. Google Scholar CrossRef Search ADS PubMed  Shi P, Jia S, Zhang X-X et al.  . Metagenomic insights into chlorination effects on microbial antibiotic resistance in drinking water. Water Res  2013; 47: 111– 20. Google Scholar CrossRef Search ADS PubMed  Soni B, Bartelt-Hunt SL, Snow DD et al.  . Narrow grass hedges reduce tylosin and associated antimicrobial resistance genes in agricultural runoff. J Environ Qual  2015; 44: 895. Google Scholar CrossRef Search ADS PubMed  Staley C, Unno T, Gould TJ et al.  . Application of Illumina next-generation sequencing to characterize the bacterial community of the Upper Mississippi River. J Appl Microbiol  2013; 115: 1147– 58. Google Scholar CrossRef Search ADS PubMed  Stedtfeld RD, Stedtfeld TM, Waseem H et al.  . Isothermal assay targeting class 1 integrase gene for environmental surveillance of antibiotic resistance markers. 2017, DOI: https://doi.org/10.1016/j.jenvman.2017.04.079. Sutcliffe J, Grebe T, Tait-Kamradt A et al.  . Detection of erythromycin-resistant determinants by PCR. Antimicrob. Agents Chemother . 1996; 40(11): 2562– 6. Vaz-Moreira I, Nunes OC, Manaia CM et al.  . Bacterial diversity and antibiotic resistance in water habitats: searching the links with the human microbiome. FEMS Microbiol Rev  2014; 38: 761– 78. Google Scholar CrossRef Search ADS PubMed  Vester B, Douthwaite S. Macrolide resistance conferred by base substitutions in 23S rRNA. Antimicrob Agents Ch  2001; 45: 1– 12. Google Scholar CrossRef Search ADS   Wang F-H, Qiao M, Su J-Q et al.  . High throughput profiling of antibiotic resistance genes in urban park soils with reclaimed water irrigation. Environ Sci Technol  2014; 48: 9079– 85. Google Scholar CrossRef Search ADS PubMed  Wang L, Gutek A, Grewal S et al.  . Changes in diversity of cultured bacteria resistant to erythromycin and tetracycline in swine manure during simulated composting and lagoon storage. Lett Appl Microbiol  2015; 61: 245– 251. Google Scholar CrossRef Search ADS PubMed  Wang L, Oda Y, Grewal S et al.  . Persistence of Resistance to Erythromycin and Tetracycline in Swine Manure During Simulated Composting and Lagoon Treatments. Microb Ecol  2012; 63: 32– 40. Google Scholar CrossRef Search ADS PubMed  Wang Y, Wang G-R, Shoemaker NB et al.  . Distribution of the ermG gene among bacterial isolates from porcine intestinal contents. Appl Environ Microbiol  2005; 71: 4930– 4. Google Scholar CrossRef Search ADS PubMed  Weisblum B. Macrolide resistance. Drug Resist Updat  1998; 1: 29– 41. Google Scholar CrossRef Search ADS PubMed  Whitehead TR, Cotta MA. Stored swine manure and swine faeces as reservoirs of antibiotic resistance genes. Lett Appl Microbiol  2013; 56: 264– 7. Google Scholar CrossRef Search ADS PubMed  WHO. Global Action Plan on Antimicrobial Resistance . 2017. http://www.wpro.who.int/entity/drug_resistance/resources/global_action_plan_eng.pdf. Xu S, Sura S, Zaheer R et al.  . Dissipation of antimicrobial resistance determinants in composted and stockpiled beef cattle manure. J Environ Qual  2016; 45: 528– 6. Google Scholar CrossRef Search ADS PubMed  Zou L-K, Wang H-N, Zeng B et al.  . Erythromycin resistance and virulence genes in Enterococcus faecalis from swine in China. New Microbiol  2011; 34: 73– 80. Google Scholar PubMed  © FEMS 2018. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png FEMS Microbiology Ecology Oxford University Press

Practical implications of erythromycin resistance gene diversity on surveillance and monitoring of resistance

Loading next page...
 
/lp/ou_press/practical-implications-of-erythromycin-resistance-gene-diversity-on-ogecQWCS4y
Publisher
Blackwell
Copyright
© FEMS 2018.
ISSN
0168-6496
eISSN
1574-6941
D.O.I.
10.1093/femsec/fiy006
Publisher site
See Article on Publisher Site

Abstract

Abstract Use of antibiotics in human and animal medicine has applied selective pressure for the global dissemination of antibiotic-resistant bacteria. Therefore, it is of interest to develop strategies to mitigate the continued amplification and transmission of resistance genes in environmental reservoirs such as farms, hospitals and watersheds. However, the efficacy of mitigation strategies is difficult to evaluate because it is unclear which resistance genes are important to monitor, and which primers to use to detect those genes. Here, we evaluated the diversity of one type of macrolide antibiotic resistance gene (erm) in one type of environment (manure) to determine which primers would be most informative to use in a mitigation study of that environment. We analyzed all known erm genes and assessed the ability of previously published erm primers to detect the diversity. The results showed that all known erm resistance genes group into 66 clusters, and 25 of these clusters (40%) can be targeted with primers found in the literature. These primers can target 74%–85% of the erm gene diversity in the manures analyzed. antibiotic, resistance, target, environment, surveillance, genes INTRODUCTION Antibiotic resistance is a global challenge, with increasing resistance to antibiotics threatening our ability to treat both human and animal diseases (WHO). Antibiotic use in human medicine and animal agriculture has increased environmental reservoirs of antibiotic resistance genes, which in turn has increased the risk of transmission of antibiotic-resistant bacteria to both humans and animals (McEwen and Fedorka-Cray 2002; Vaz-Moreira et al.2014). This linkage has resulted in the prioritization of understanding how resistance moves from environmental sources to clinical pathogens and the associated influence of human activity. To understand the movement of antibiotic resistance in the environment, we need accessible tools that can provide large-scale surveillance of resistance in diverse environmental samples. Molecular microbiology advances have allowed us to leverage amplification and subsequent sequencing of DNA that encodes for antibiotic resistance genes, resulting in our awareness of an incredibly diverse global reservoir of environmental “resistomes”. Generally, metagenomic shotgun sequencing is a costly tool for antibiotic gene surveillance as it provides information on ‘all’ genes in an environmental sample. Among these genes, only a fraction (0.01%–1%) are related to antibiotic resistance, resulting in a significant majority of sequences from metagenomes not readily usable for resistance detection (Shi et al.2013; Li et al.2015). A promising alternative to metagenomic sequencing is high-throughput amplicon qPCR assays, such as the Wafergen Smartchip that has been previously used for several resistance surveillance studies (Shi et al.2013; Wang et al.2014; Karkman et al.2016; Muziasari et al.2016; Stedtfeld et al.2017). Unlike the broad scope of metagenomic sequencing, high-throughput qPCR assays target a suite of genes using primers and can quantify hundreds of targeted resistance genes and multiple samples simultaneously (e.g. one Wafergen Smartchip contains 5184 assays). Consequently, the price per gene or sample of these assays for resistance gene detection is orders of magnitude less than metagenomic sequencing, making it more conducive to large-scale surveillance. A significant limitation of this technology is the need to develop primer-based assays for each targeted gene of interest that are effective for high-throughput amplification conditions. We are increasingly aware that certain genes may be more related to the risks of the emergence or persistence of resistance than others. For example, integrons and sulfonamide resistance genes have been used to detect anthropogenic contaminants (Wang et al.2014; Gillings et al.2015). Further, specific environments (mammalian gut, manure, wastewater, etc.) have been observed to be enriched in antibiotic resistance genes relative to soil or water environments (Chee-Sanford et al.2009; Koike et al.2010; Garder, Moorman and Soupir 2014; Joy et al.2014; Luby, Moorman and Soupir 2016), suggesting that these environments are potential reservoirs of resistance genes. Among the hundreds of genes associated with antibiotic resistance that are observed in environmental metagenomes, selecting the key targets relevant to the spread of resistance is a significant and important opportunity. In this study, we demonstrate how we have chosen specific genes that are the most effective among previously targeted genes to serve as indicators for antibiotic resistance and to understand resistance hotspots and transmission. This framework, while developed for agriculturally impacted environments, can be broadly applied to the selection of genes from varying resistance gene classes and environments. Specifically, this effort focuses on understanding the diversity of erythromycin ribosomal methylase (erm) gene and the most relevant gene targets for understanding the spread of erm-associated resistance from manure sources to the environment. Erm genes encode resistance to macrolide antibiotics, which have long been used to treat Gram-positive and certain Gram-negative pathogens infecting humans, swine and cattle (Roberts et al.2008; Pyörälä et al.2014). Broadly, macrolide antibiotics act by binding to the 23S subunit of the bacterial ribosome, causing premature release of peptides during translation. The erm genes cause resistance by methylating rRNA at the active site, reducing the ability of macrolide antibiotics to bind to the ribosome (Weisblum 1998; Vester and Douthwaite 2001). Erm-mediated resistance to macrolides has also been observed to confer resistance against other antibiotics, including lincosamide and streptogramin B (MLSB resistance) (Leclercq and Courvalin 1991). The widespread use of macrolides and their relevance for both animal and human health has resulted in a research emphasis on erm genes and their bacterial hosts as key targets for understanding the development of resistance and its spread in agricultural environments. Previously, erm genes have been detected in various agricultural settings, including swine manure, lagoon water, soils, surface and subsurface drainage from fields, and groundwater surrounding and underlying animal production facilities (Chen et al.2007; Knapp et al.2010; Koike et al.2010; Joy et al.2013, 2014; Whitehead and Cotta 2013; Fahrenfeld et al.2014; Garder, Moorman and Soupir 2014; Soni et al.2015; Luby, Moorman and Soupir 2016). Most of our previous knowledge of erm genes and their associated amplicon targets have stemmed from the characterization and sequencing of bacterial isolates and their phenotypic resistance to MLSB antibiotics (Pyörälä et al.2014). A total of 21 unique classes of erm genes have been identified based on sequence homology to protein-coding erm sequences from cultured bacteria (Roberts et al.1999). More recently, metagenomic analyses of DNA from the total microbial community in environmental samples has expanded what is known about erm diversity beyond these 21 classes, showing that the erm class of genes is comprised of numerous sequence variants from diverse bacterial hosts (Fang et al.2015; Li et al.2015). These sequence variants are present in a range of abundances depending on their environment of origin. The focus of this study was to better understand the diversity of erm genes and to target the gene variants that could be indicative of resistance originating from manure and spreading to agricultural soil and water environments. MATERIALS AND METHODS Phylogenetic analysis of erm genes Gene sequencing sharing high similarity to ermA, ermB, ermC and ermF were obtained from publicly available databases. The Ribosomal Database Project Fungene Repository (Fish et al.2013) was used to obtain ermB- and ermC-associated sequences. It was required that sequences share 97% amino acid sequence coverage to established HMM protein models for Fungene gene families “Resfam_ermA”, “Resfam_ermB” and “Resfam_ermC” (Version 8.8). Additionally, ermF gene nucleotide sequences were obtained from proteins listed in the ARDB-Antibiotic Resistance Genes Database (version 1.1, July 3, 2009) (Liu and Pop 2009) and associated with the annotation “ermF”. All erm-associated sequences were combined and clustered at 99% nucleotide similarity using CD-HIT (v4.6.1c) (Li and Godzik 2006; Fu et al.2012), resulting in 66 unique clusters. One representative sequence for each cluster was identified by CD-HIT and was aligned using Muscle (v3.8.31) (Edgar 2004) with the following parameters: gap open –400, gap extend 0, clustering method UPGMB. A maximum-likelihood phylogenetic tree was constructed from this alignment using FastTree (v2.1.8) (Price, Dehal and Arkin 2010) with default parameters. Taxonomy was identified based on annotations in the NCBI non-redundant nucleotide database (NCBI Resource Coordinators 2017). To consider an erm gene sequence to be associated with a previously targeted PCR primer sequence, both forward and reverse primers were required to share 100% nucleotide similarity over a minimum of 17 bp of the primer length. Manure metagenomic datasets The presence of erm genes was characterized in swine and cattle manures. For swine manure, DNA was extracted from two biological replicates (three technical replicates each) of swine manure originating from Iowa State University's Northeast Research and Demonstration Farm, near Nashua, IA (43.0° N, 92.5° W). Metagenomic libraries were prepared and sequenced at Iowa State University DNA Sequencing Facility on a HiSeq 2500 instrument (Illumina, San Diego, CA) according to manufacturer's instructions. These datasets are deposited in the NCBI SRA as project SRP109083 (Table S1, Supporting Information). Sequences were compared to representatives of erm genes described above (BLAST, v2.4.0+) (Camacho et al.2009). Sequences were annotated as erm genes if they matched the representative sequence within a cluster with a minimum e-value of 1e-5 and if both paired-end reads matched the same representative target. The abundance of erm sequences in each sample was calculated as the total number of reads meeting these criteria. Cattle manure metagenomes were obtained from a previously published study of antibiotic resistant genes in commercial cattle as they moved through the process of beef production from feedlot entry to slaughter (Noyes et al.2016). The presence of erm sequences in these samples was determined by the total number of reads that shared sequence homology (BLAST, v2.4.0+, e-value 1e-5) to the best matched erm representative sequence. Similarly, metagenomes from human-impacted (Fitzpatrick and Walsh 2016) and pristine environment (Staley et al.2013) were aligned against erm sequences and considered a match if alignment scores resulted in e-value scores of at least 1e-5. RESULTS A total of 5648 erm DNA sequences were identified from annotated genes based on sequence similarity to well-characterized erm genes and were clustered at 99% nucleotide similarity to identify 66 unique erm variant clusters. A representative sequence of each cluster was defined as the longest consensus sequence in each cluster as determined by a greedy incremental clustering algorithm (see Methods, Table 1). These representative sequences were aligned and used to construct a phylogenetic tree describing the diversity of erm genes (Fig. 1). Based on sequence homology, the resulting erm gene clusters encompass the majority of erm genes studied in previous literature: ermA, ermB, ermC, ermF, ermG and ermT (reviewed in Roberts et al.2008). Among the gene clusters, a cluster associated with ermA was the most represented in our erm gene database (Cluster 15, 3542 genes), followed by an ermB cluster (Cluster 18, 1387 genes), and then an ermC cluster (Cluster 30, 399 genes). These three gene clusters comprise 94% of erm genes and are evidence to biases in the previous characterization of erm genes towards specific gene variants. Beyond the three most abundant gene clusters, the next most represented cluster (Cluster 11, 50 genes) is not well-characterized (e.g. most similar to unannotated erm gene clusters in our database) and is most closely related to genes belonging to Streptococcus agalactiae strain TR7 (100% nucleotide identity). Most clusters (53 of 66) are associated with five or less gene sequences, demonstrating that much of what we know of specific erm gene families is based on very few characterized representatives. Figure 1. View largeDownload slide Maximum likelihood phylogenetic tree of 66 erm sequence clusters based on 99% nucleotide similarity of 5648 DNA sequences extracted from known erm genes described in existing databases. Clusters that contain gene targets from existing PCR primers (see Table 2) are highlighted in color. The relative number of sequences comprising each cluster among the 5648 DNA sequences is also shown. Figure 1. View largeDownload slide Maximum likelihood phylogenetic tree of 66 erm sequence clusters based on 99% nucleotide similarity of 5648 DNA sequences extracted from known erm genes described in existing databases. Clusters that contain gene targets from existing PCR primers (see Table 2) are highlighted in color. The relative number of sequences comprising each cluster among the 5648 DNA sequences is also shown. Table 1. Erm gene clusters identified from 5648 erm sequences. For each cluster, the most representative gene is referenced by its NCBI accession number in NCBI nucleotide and protein databases. Cluster (this study)  NCBI protein accession no.  NCBI nucleotide accession no.  Description in NCBI GenBank  Organism  Cluster 0  BAJ34818  AB601890  Erythromycin resistance protein  Photobacterium damselae subsp. piscicida  Cluster 1  KNF08983  LGSS01000004  rRNA (adenine-N(6)-)-methyltransferase  Clostridium purinilyticum  Cluster 2  AFS78141  CP003326  rRNA (adenine-N(6)-)-methyltransferase  Clostridium acidurici 9a  Cluster 3  ABW20380  CP000853  rRNA (adenine-N(6)-)-methyltransferase  Alkaliphilus oremlandii OhILAs  Cluster 4  KKS60599  LCDU01000003  rRNA (Adenine-N(6)-)-methyltransferase  Candidate division WWE3 bacterium GW2011_GWF2_42_42  Cluster 5  KKS35651  LCCU01000032  rRNA (Adenine-N(6)-)-methyltransferase  Candidate division WWE3 bacterium GW2011_GWF1_42_14  Cluster 6  AJB79756  CP010391  Hypothetical protein  Klebsiella pneumoniae  Cluster 7  EKD94896  AMFJ01010665  Hypothetical protein  Uncultured bacterium  Cluster 8  KKU26033  LCLY01000007  rRNA (Adenine-N(6)-)-methyltransferase  Microgenomates group bacterium GW2011_GWA2_46_16  Cluster 9  CCQ93859  CARA01000062  rRNA adenine N-6-methyltransferase  Clostridium ultunense Esp  Cluster 10  BAP00917  AP013353  Dimethyladenosine transferase  Mycoplasma californicum HAZ160_1  Cluster 11  CNJ04734  CQCN01000003  Dimethyladenosine transferase  Streptococcus agalactiae  Cluster 12  AAA27431  M17808.1  ermF  Bacteroides fragilis  Cluster 13  AAA63165  M62487.1  ermF  Bacteroides fragilis  Cluster 14  EEO52603  ACAB02000055.1  ermF  Bacteroides sp. D1  Cluster 15  CCJ25599  HE579073  rRNA adenine N-6-methyltransferase  Staphylococcus aureus subsp. aureus ST228  Cluster 16  CCX90994  CAXH010000024  Ribosomal RNA small subunit methyltransferase A  Succinatimonas sp. CAG:777  Cluster 17  EGV00599  AFXA01000001  Dimethyladenosine transferase rRNA modification enzyme  Mycoplasma columbinum SF7  Cluster 18  EFY03905  AEVN01000118  rRNA adenine N-6-methyltransferase  Phascolarctobacterium succinatutens YIT 12067  Cluster 19  ACB90575  CP001033  Erythromycin ribosome methylase  Streptococcus pneumoniae CGSP14  Cluster 20  EJY36237  AMBI01000188  rRNA adenine N-6-methyltransferase  Enterococcus faecium 510  Cluster 21  AKB11102  CP011096  16S rRNA methyltransferase  Mycoplasma synoviae ATCC 25204  Cluster 22  ACD66486  CP001080  Dimethyladenosine transferase  Sulfurihydrogenibium sp. YO3AOP1  Cluster 23  EIB96299  AICL01000010  rRNA methylase  Lactobacillus salivarius SMXD51  Cluster 24  EEP60650  ABZS01000069  Dimethyladenosine transferase  Sulfurihydrogenibium yellowstonense SS-5  Cluster 25  AFV15157  JQ655732  Erythromycin  Clostridium perfringens  Cluster 26  KDE45359  JFKK01000007  16S rRNA methyltransferase  Mycoplasma hyosynoviae  Cluster 27  ADM89794  CP002161  Putative dimethyladenosine transferase  Candidatus Zinderia insecticola CARI  Cluster 28  KER55751  JPHP01000035  SAM-dependent methlyltransferase  Bacteroides fragilis  Cluster 29  AAR27225  AY357120  N-methyltransferase  Streptococcus pyogenes  Cluster 30  AIU96746  KF831357  ErmC  Staphylococcus aureus  Cluster 31  ACG57739  CP001130  Ribosomal RNA adenine methylase transferase  Hydrogenobaculum sp. Y04AAS1  Cluster 32  AAO20906  AF205068  erm44  Lactobacillus reuteri  Cluster 33  AFH70049  CP003045  rRNA adenine N-6-methyltransferase  Staphylococcus aureus subsp. aureus 71193  Cluster 34  ACC94310  EU595407  ErmB  Uncultured Enterococcus sp.  Cluster 35  AAF86219  AF242872  ErmB  Enterococcus faecium  Cluster 36  CDZ75671  LM997412  rRNA adenine N-6-methyltransferase  Peptoniphilus sp. ING2-D1G  Cluster 37  EOK35943  ASEN01000042  rRNA adenine N-6-methyltransferase  Enterococcus faecalis EnGen0332  Cluster 38  EZX88180  JIYN01000027  rRNA adenine N-6-methyltransferase  Staphylococcus aureus GD2010-052  Cluster 39  CEI83544  CDGG01000001  rRNA adenine N-6-methyltransferase  Oceanobacillus oncorhynchi  Cluster 40  CEJ95855  LN680996  23S RNA methylase for macrolide-lincosamide-streptogramin B resistance  Staphylococcus fleurettii  Cluster 41  CAD32685  AJ488494  Erythromycin resistance protein  Lactobacillus fermentum  Cluster 42  EIY35985  AGXG01000023  rRNA adenine N-6-methyltransferase  Bacteroides cellulosilyticus CL02T12C19  Cluster 43  EDV04163  ABJL02000008  Hypothetical protein  Bacteroides intestinalis DSM 17393  Cluster 44  AHH55321  KC790462  rRNA adenine N-6-methyltransferase  Streptococcus suis  Cluster 45  BAB20748  AB014481  ErmGM  Staphylococcus aureus  Cluster 46  KAC49299  JIQI01000041  rRNA adenine N-6-methyltransferase  Staphylococcus aureus VET0243R  Cluster 47  CDQ41560  CCDP010000003  rRNA adenine N-6-methyltransferase  Virgibacillus massiliensis  Cluster 48  AGK85210  KC405064  Erythromycin ribosome methylase  Haemophilus parasuis  Cluster 49  BAC12877  BA000028  Erythromycin resistance protein  Oceanobacillus iheyensis HTE831  Cluster 50  AAC37034  L42817  rRNA methyltransferase  Bacteroides thetaiotaomicron  Cluster 51  EJD65709  AFSU01000133  Hypothetical protein  Bacillus sp. 916  Cluster 52  CAJ43792  AM159501  rRNA methylase  Staphylococcus saprophyticus  Cluster 53  EZS04927  JILJ01000152  rRNA adenine N-6-methyltransferase  Staphylococcus aureus VET0436R  Cluster 54  CCG55258  HE775264  Ribosomal RNA adenine methylase Erm(43)  Staphylococcus lentus  Cluster 55  EJY20540  AMBD01000117  rRNA adenine N-6-methyltransferase  Enterococcus faecium C1904  Cluster 56  CAE18145  AJ579365  rRNA methylase  Staphylococcus sciuri  Cluster 57  KIJ86993  JXBG01000010  SAM-dependent methlyltransferase  Staphylococcus saprophyticus  Cluster 58  EKB53568  AGZE01000039  Hypothetical protein  Facklamia ignava CCUG 37419  Cluster 59  CDS14986  LK392593  23S rRNA methylase  Staphylococcus xylosus  Cluster 60  AJK31391  KJ728534  Ribosomal RNA adenine methylase variant  Staphylococcus xylosus  Cluster 61  AJK31388  KJ728533  Ribosomal RNA adenine methylase  Staphylococcus saprophyticus  Cluster 62  KIO72601  JXLU01000090  rRNA adenine N-6-methyltransferase  Bacillus thermoamylovorans  Cluster 63  KKD22675  LATV01000011  SAM-dependent methlyltransferase  Staphylococcus cohnii subsp. cohnii  Cluster 64  EVJ59956  JBER01000028  rRNA adenine N-6-methyltransferase  Staphylococcus aureus GGMC6053  Cluster 65  EDU98728  ABIY02000132.1  ermF  Bacteroides coprocola DSM 17136  Cluster (this study)  NCBI protein accession no.  NCBI nucleotide accession no.  Description in NCBI GenBank  Organism  Cluster 0  BAJ34818  AB601890  Erythromycin resistance protein  Photobacterium damselae subsp. piscicida  Cluster 1  KNF08983  LGSS01000004  rRNA (adenine-N(6)-)-methyltransferase  Clostridium purinilyticum  Cluster 2  AFS78141  CP003326  rRNA (adenine-N(6)-)-methyltransferase  Clostridium acidurici 9a  Cluster 3  ABW20380  CP000853  rRNA (adenine-N(6)-)-methyltransferase  Alkaliphilus oremlandii OhILAs  Cluster 4  KKS60599  LCDU01000003  rRNA (Adenine-N(6)-)-methyltransferase  Candidate division WWE3 bacterium GW2011_GWF2_42_42  Cluster 5  KKS35651  LCCU01000032  rRNA (Adenine-N(6)-)-methyltransferase  Candidate division WWE3 bacterium GW2011_GWF1_42_14  Cluster 6  AJB79756  CP010391  Hypothetical protein  Klebsiella pneumoniae  Cluster 7  EKD94896  AMFJ01010665  Hypothetical protein  Uncultured bacterium  Cluster 8  KKU26033  LCLY01000007  rRNA (Adenine-N(6)-)-methyltransferase  Microgenomates group bacterium GW2011_GWA2_46_16  Cluster 9  CCQ93859  CARA01000062  rRNA adenine N-6-methyltransferase  Clostridium ultunense Esp  Cluster 10  BAP00917  AP013353  Dimethyladenosine transferase  Mycoplasma californicum HAZ160_1  Cluster 11  CNJ04734  CQCN01000003  Dimethyladenosine transferase  Streptococcus agalactiae  Cluster 12  AAA27431  M17808.1  ermF  Bacteroides fragilis  Cluster 13  AAA63165  M62487.1  ermF  Bacteroides fragilis  Cluster 14  EEO52603  ACAB02000055.1  ermF  Bacteroides sp. D1  Cluster 15  CCJ25599  HE579073  rRNA adenine N-6-methyltransferase  Staphylococcus aureus subsp. aureus ST228  Cluster 16  CCX90994  CAXH010000024  Ribosomal RNA small subunit methyltransferase A  Succinatimonas sp. CAG:777  Cluster 17  EGV00599  AFXA01000001  Dimethyladenosine transferase rRNA modification enzyme  Mycoplasma columbinum SF7  Cluster 18  EFY03905  AEVN01000118  rRNA adenine N-6-methyltransferase  Phascolarctobacterium succinatutens YIT 12067  Cluster 19  ACB90575  CP001033  Erythromycin ribosome methylase  Streptococcus pneumoniae CGSP14  Cluster 20  EJY36237  AMBI01000188  rRNA adenine N-6-methyltransferase  Enterococcus faecium 510  Cluster 21  AKB11102  CP011096  16S rRNA methyltransferase  Mycoplasma synoviae ATCC 25204  Cluster 22  ACD66486  CP001080  Dimethyladenosine transferase  Sulfurihydrogenibium sp. YO3AOP1  Cluster 23  EIB96299  AICL01000010  rRNA methylase  Lactobacillus salivarius SMXD51  Cluster 24  EEP60650  ABZS01000069  Dimethyladenosine transferase  Sulfurihydrogenibium yellowstonense SS-5  Cluster 25  AFV15157  JQ655732  Erythromycin  Clostridium perfringens  Cluster 26  KDE45359  JFKK01000007  16S rRNA methyltransferase  Mycoplasma hyosynoviae  Cluster 27  ADM89794  CP002161  Putative dimethyladenosine transferase  Candidatus Zinderia insecticola CARI  Cluster 28  KER55751  JPHP01000035  SAM-dependent methlyltransferase  Bacteroides fragilis  Cluster 29  AAR27225  AY357120  N-methyltransferase  Streptococcus pyogenes  Cluster 30  AIU96746  KF831357  ErmC  Staphylococcus aureus  Cluster 31  ACG57739  CP001130  Ribosomal RNA adenine methylase transferase  Hydrogenobaculum sp. Y04AAS1  Cluster 32  AAO20906  AF205068  erm44  Lactobacillus reuteri  Cluster 33  AFH70049  CP003045  rRNA adenine N-6-methyltransferase  Staphylococcus aureus subsp. aureus 71193  Cluster 34  ACC94310  EU595407  ErmB  Uncultured Enterococcus sp.  Cluster 35  AAF86219  AF242872  ErmB  Enterococcus faecium  Cluster 36  CDZ75671  LM997412  rRNA adenine N-6-methyltransferase  Peptoniphilus sp. ING2-D1G  Cluster 37  EOK35943  ASEN01000042  rRNA adenine N-6-methyltransferase  Enterococcus faecalis EnGen0332  Cluster 38  EZX88180  JIYN01000027  rRNA adenine N-6-methyltransferase  Staphylococcus aureus GD2010-052  Cluster 39  CEI83544  CDGG01000001  rRNA adenine N-6-methyltransferase  Oceanobacillus oncorhynchi  Cluster 40  CEJ95855  LN680996  23S RNA methylase for macrolide-lincosamide-streptogramin B resistance  Staphylococcus fleurettii  Cluster 41  CAD32685  AJ488494  Erythromycin resistance protein  Lactobacillus fermentum  Cluster 42  EIY35985  AGXG01000023  rRNA adenine N-6-methyltransferase  Bacteroides cellulosilyticus CL02T12C19  Cluster 43  EDV04163  ABJL02000008  Hypothetical protein  Bacteroides intestinalis DSM 17393  Cluster 44  AHH55321  KC790462  rRNA adenine N-6-methyltransferase  Streptococcus suis  Cluster 45  BAB20748  AB014481  ErmGM  Staphylococcus aureus  Cluster 46  KAC49299  JIQI01000041  rRNA adenine N-6-methyltransferase  Staphylococcus aureus VET0243R  Cluster 47  CDQ41560  CCDP010000003  rRNA adenine N-6-methyltransferase  Virgibacillus massiliensis  Cluster 48  AGK85210  KC405064  Erythromycin ribosome methylase  Haemophilus parasuis  Cluster 49  BAC12877  BA000028  Erythromycin resistance protein  Oceanobacillus iheyensis HTE831  Cluster 50  AAC37034  L42817  rRNA methyltransferase  Bacteroides thetaiotaomicron  Cluster 51  EJD65709  AFSU01000133  Hypothetical protein  Bacillus sp. 916  Cluster 52  CAJ43792  AM159501  rRNA methylase  Staphylococcus saprophyticus  Cluster 53  EZS04927  JILJ01000152  rRNA adenine N-6-methyltransferase  Staphylococcus aureus VET0436R  Cluster 54  CCG55258  HE775264  Ribosomal RNA adenine methylase Erm(43)  Staphylococcus lentus  Cluster 55  EJY20540  AMBD01000117  rRNA adenine N-6-methyltransferase  Enterococcus faecium C1904  Cluster 56  CAE18145  AJ579365  rRNA methylase  Staphylococcus sciuri  Cluster 57  KIJ86993  JXBG01000010  SAM-dependent methlyltransferase  Staphylococcus saprophyticus  Cluster 58  EKB53568  AGZE01000039  Hypothetical protein  Facklamia ignava CCUG 37419  Cluster 59  CDS14986  LK392593  23S rRNA methylase  Staphylococcus xylosus  Cluster 60  AJK31391  KJ728534  Ribosomal RNA adenine methylase variant  Staphylococcus xylosus  Cluster 61  AJK31388  KJ728533  Ribosomal RNA adenine methylase  Staphylococcus saprophyticus  Cluster 62  KIO72601  JXLU01000090  rRNA adenine N-6-methyltransferase  Bacillus thermoamylovorans  Cluster 63  KKD22675  LATV01000011  SAM-dependent methlyltransferase  Staphylococcus cohnii subsp. cohnii  Cluster 64  EVJ59956  JBER01000028  rRNA adenine N-6-methyltransferase  Staphylococcus aureus GGMC6053  Cluster 65  EDU98728  ABIY02000132.1  ermF  Bacteroides coprocola DSM 17136  View Large Next, we evaluated the diversity of bacteria carrying these erm genes by identifying the taxonomic origin of potential bacterial hosts associated with each erm gene sequence (Table 1; Fig S1, Supporting Information). In general, the majority of known erm gene sequences were associated with Firmicutes (98%), followed by Proteobacteria (0.6%) and Bacterioidetes (0.6%). While ermF and ermG genes were observed to be carried by only Bacteriodetes, ermA, ermB, ermC and ermT genes were associated primarily with Firmicutes (Fig S1, Supporting Information). Within the Firmicutes, ermB genes were associated mainly with the order Lactobacillales, while ermA and ermT genes were associated with members of the Bacillales order (Fig S2, Supporting Information). These results demonstrate a wide range of potential host diversity for erm genes and highlight the impact of the choice of primer gene targets selecting for or against specific host bacteria. Historically, erm genes have been extensively targeted for qPCR quantification of gene abundances in the environment (Table 2), and we evaluated the ability of previously published PCR primers to detect the erm gene diversity described above by computationally hybridizing the primer sequences from the literature with the representative erm gene sequences in our database. Overall, published primer pairs were 100% similar to 25 of the representative sequences of erm clusters (Fig. 1). Generally, well-characterized gene clusters (e.g. containing the most known gene sequences) were observed to be associated with previous primer development. Several clusters were not associated with previously published primer targets, very likely due to the few well-characterized erm sequences within these clusters. Previously, observed diversity in natural samples have weak correlations with well-characterized genes (Choi et al.2017), suggesting that primer targets selected based on the most well-studied genes may not be effective in environmental samples. Table 2. Previously published PCR primer and gene targets for erm genes. Gene  Cluster  Primers design  Papers citing primers  ermA  15  Patterson et al.2007  b    n/aa  Sutcliffe et al.1996  Martel et al.2003a, Jackson et al.2004, Luthje and Schwarz 2006, Garofalo et al.2007, Chenier and Juteau 2009, Zou et al.2011, Di Cesare et al.2012, Hoang et al.2013, Lerma et al.2014    15, 56  Jensen et al.1999  Aarestrup et al.2000a,b, Jensen et al.2002, Petersen and Dalsgaard 2003, Whitehead and Cotta 2013    n/a  Chen et al.2007  Sharma et al.2009, Just et al.2011, Alexander et al.2011, Wang et al.2012, Holman and Chenier 2013, Wang et al.2015, Xu et al.2016    15, 56  Koike et al.2010  Ekizoglu et al.2013  ermB  n/a  Sutcliffe et al.1996  Martel et al.2003a,b, Cauwerts et al.2007, Ahmad et al.2011, Hoang et al.2013    18, 19, 20, 25, 28, 29, 32, 34, 35, 42  Jensen et al.1999  De Leener et al.2005, Whitehead and Cotta 2013    18, 19, 20, 25, 28, 29, 32, 34, 35, 42  Chen et al.2007  Sharma et al.2009, Chen et al.2010, Alexander et al.2011, Just et al.2011, Kalmokoff et al.2011, Negreanu et al.2012, Holman and Chenier 2013, Beukers et al.2015, Wang et al.2015, Sandberg and LaPara 2016, Xu et al.2016    18, 19, 20, 25, 29, 32, 35, 42  Patterson et al.2007  Knapp et al.2010    18, 19, 20, 25, 28, 29, 32, 34, 35, 42  Koike et al.2010  Ekizoglu et al.2013, Garder et al.2014, Joy et al.2013, Joy et al.2014, Soni et al.2015, Luby et al.2016  ermC  n/a  Sutcliffe et al.1996  Martel et al.2003b, Hoang et al.2013    30, 46  Jensen et al.1999  Ekizoglu et al.2013, Whitehead and Cotta 2013    23, 30, 46, 51, 52, 63, 64  Patterson et al.2007  Knapp et al.2010, Popowska et al.2012    30, 46, 51  Koike et al.2010  Luby et al.2016  ermF  12, 13  Chen et al.2007  Sharma et al.2009, Chen et al.2010, Alexander et al.2011, Kalmokoff et al.2011, Negreanu et al.2012, Wang et al.2012, Hoang et al.2013, Holman and Chenier 2013, Farenfeld et al.2014, Garder et al.2014, Luby et al.2016, Xu et al.2016    12, 13  Patterson et al.2007  Knapp et al.2010    12, 13  Koike et al.2010  Ekizoglu et al.2013, Joy et al.2013, Joy et al.2014    43, 50  Wang et al.2005  Wang et al.2005, Kalmokoff et al.2011  ermG  43, 50  Patterson et al.2007  N/A    43, 50  Koike et al.2010  Ekizoglu et al.2013  ermT  33, 41  Chen et al.2007  Sharma et al.2009, Alexander et al.2011, Kalmokoff et al.2011, Wang et al.2012, Hoang et al.2013, Garder et al.2014, Wang et al.2015  Gene  Cluster  Primers design  Papers citing primers  ermA  15  Patterson et al.2007  b    n/aa  Sutcliffe et al.1996  Martel et al.2003a, Jackson et al.2004, Luthje and Schwarz 2006, Garofalo et al.2007, Chenier and Juteau 2009, Zou et al.2011, Di Cesare et al.2012, Hoang et al.2013, Lerma et al.2014    15, 56  Jensen et al.1999  Aarestrup et al.2000a,b, Jensen et al.2002, Petersen and Dalsgaard 2003, Whitehead and Cotta 2013    n/a  Chen et al.2007  Sharma et al.2009, Just et al.2011, Alexander et al.2011, Wang et al.2012, Holman and Chenier 2013, Wang et al.2015, Xu et al.2016    15, 56  Koike et al.2010  Ekizoglu et al.2013  ermB  n/a  Sutcliffe et al.1996  Martel et al.2003a,b, Cauwerts et al.2007, Ahmad et al.2011, Hoang et al.2013    18, 19, 20, 25, 28, 29, 32, 34, 35, 42  Jensen et al.1999  De Leener et al.2005, Whitehead and Cotta 2013    18, 19, 20, 25, 28, 29, 32, 34, 35, 42  Chen et al.2007  Sharma et al.2009, Chen et al.2010, Alexander et al.2011, Just et al.2011, Kalmokoff et al.2011, Negreanu et al.2012, Holman and Chenier 2013, Beukers et al.2015, Wang et al.2015, Sandberg and LaPara 2016, Xu et al.2016    18, 19, 20, 25, 29, 32, 35, 42  Patterson et al.2007  Knapp et al.2010    18, 19, 20, 25, 28, 29, 32, 34, 35, 42  Koike et al.2010  Ekizoglu et al.2013, Garder et al.2014, Joy et al.2013, Joy et al.2014, Soni et al.2015, Luby et al.2016  ermC  n/a  Sutcliffe et al.1996  Martel et al.2003b, Hoang et al.2013    30, 46  Jensen et al.1999  Ekizoglu et al.2013, Whitehead and Cotta 2013    23, 30, 46, 51, 52, 63, 64  Patterson et al.2007  Knapp et al.2010, Popowska et al.2012    30, 46, 51  Koike et al.2010  Luby et al.2016  ermF  12, 13  Chen et al.2007  Sharma et al.2009, Chen et al.2010, Alexander et al.2011, Kalmokoff et al.2011, Negreanu et al.2012, Wang et al.2012, Hoang et al.2013, Holman and Chenier 2013, Farenfeld et al.2014, Garder et al.2014, Luby et al.2016, Xu et al.2016    12, 13  Patterson et al.2007  Knapp et al.2010    12, 13  Koike et al.2010  Ekizoglu et al.2013, Joy et al.2013, Joy et al.2014    43, 50  Wang et al.2005  Wang et al.2005, Kalmokoff et al.2011  ermG  43, 50  Patterson et al.2007  N/A    43, 50  Koike et al.2010  Ekizoglu et al.2013  ermT  33, 41  Chen et al.2007  Sharma et al.2009, Alexander et al.2011, Kalmokoff et al.2011, Wang et al.2012, Hoang et al.2013, Garder et al.2014, Wang et al.2015  a Primers did not hit any clusters. b No relevant citing papers. View Large We next evaluated the diversity of erm genes in 12 947 environmental metagenomes (Table S1, Supporting Information), resulting in the observation that significantly more erm genes are present in human-impacted environments (feces- and animal-associated soil and water) than in natural environments (Fig. 2). We also searched an additional 39 metagenomes originating from relatively pristine freshwaters along the Upper Mississippi River (Staley et al.2013, Table S1, Supporting Information), resulting in only 3 reads out of 716 million, sharing similarity (e-value < 1e-5) to erm genes. Combined, these results demonstrate that erm genes are rare in environments with minimal human impact and suggest that erm genes associated with feces or manure are ideal for tracking the spread and persistence of resistance through the environment. These results are also consistent with previous observations that manure contains abundant genes related to erm resistance and is a source of these genes into the environment (e.g. soil and water) (Chee-Sanford et al.2009; Koike et al.2010; Heuer, Schmitt and Smalla 2011; Joy et al.2013; Luby, Moorman and Soupir 2016). Figure 2. View largeDownload slide Average number of erm genes in metagenomes from various environments (see Table S1, Supporting Information). *For ermC gene, the average number of reads in animal-associated soil metagenomes was 1665 ± 659 reads. Figure 2. View largeDownload slide Average number of erm genes in metagenomes from various environments (see Table S1, Supporting Information). *For ermC gene, the average number of reads in animal-associated soil metagenomes was 1665 ± 659 reads. Consequently, we next identified erm genes in manure metagenomes. We aligned erm gene sequences against metagenomes derived from two large manure metagenomic studies (requiring nmanure > 3): swine manure collected near Nashua, IA (Luby, Moorman and Soupir 2016) and cattle manure from a previously published study (Noyes et al.2016). These manure metagenomes were strategically selected based on the number of biological replicates and sequencing depth. Three erm clusters comprised 46% and 45% of the total abundance of erm genes in swine and cattle manure, respectively (Table S1, Supporting Information). The genes associated with these most abundant clusters differed between swine and cattle manures. In swine metagenomes, sequences associated with the ermB gene cluster (Fig. 1, sharing 93%–99% similarity) captured 26% of all erm sequences, followed by ermG-associated sequences capturing 11% and ermA-associated sequences capturing 9%. In cattle metagenomes, sequences associated with ermF represented 23.5% of all erm abundances, followed by sequences associated with ermG capturing 12.4% and sequences associated with ermB capturing 9%. Only a subset of erm genes detected in manure are targeted by existing primer sets. Overall, a total of 25 out of the 66 erm clusters (40%) could be computationally detected with known primers (Table 1, Supporting Information), and these genes also encompass much of the total erm abundances observed in manure metagenomes. Collectively, if all primers were used, 74% and 85% of the total erm gene sequence diversity observed in swine and cattle metagenomes, respectively, could be detected, suggesting good coverage of these genes for PCR or qPCR assays. Specifically, in swine manure metagenomes, ermB primers could detect 29% of erm sequences, followed by ermF primers capturing 14% and ermG primers capturing 12% (Fig. 3). In cattle, ermF primers are the most effective, capturing 30% of erm sequences, followed by 21% with ermG primers, and 15% with ermB primers. Consequently, depending on the environmental sample in a study, in this case swine versus cattle manure, the choice of erm gene targets can significantly alter erm abundance estimations. For example, in swine manure, two times more erm gene abundance would be estimated if ermB primers were used instead of ermF primers. Even within the same gene clade, different primers could result in significant differences in abundance estimations, and this result is observed especially for ermC primers where a near two-fold difference in abundance estimations would result based on selection of primers from Patterson et al. (2007) versus Jensen, Frimodt-Moller and Aarestrup (1999). The selection of Patterson primers would result in the detection of genes from up to seven erm gene clusters over the two to three gene cluster detected with Jensen, Frimodt-Moller and Aarestrup (1999) or Koike et al. (2010) primers. Similar results are noted in the cattle manure, where ermC primers designed by Patterson capture 13% of the total abundance of erm sequences in the metagenomes, while Koike and Jensen primers only capture 4.4% and 2.1%, respectively. These results emphasize that the targeting of a specific erm gene, even within closely related gene variants, can significantly alter estimations of associated resistance in manures. Figure 3. View largeDownload slide Abundance of DNA sequences homologous to erm gene PCR primers as a percent of total erm abundance in swine and cattle metagenomes. Figure 3. View largeDownload slide Abundance of DNA sequences homologous to erm gene PCR primers as a percent of total erm abundance in swine and cattle metagenomes. Thus, overall, for swine manure, the most effective gene target based on abundance in swine metagenomes (26% of erm genes) originates from an ermB cluster (Cluster 25) and is associated with Clostridium perfringens. The next most abundant ermB cluster in swine (Cluster 19, most similar to a gene in Streptococcus pneumoniae CGSP14) represented only 2% of erm abundances. These results indicate that while ermB primers can target multiple strains (Fig. S1 and S2, Supporting Information), in these swine metagenomes, it is one gene cluster that specifically dominates. This gene cluster is also abundant in cattle manure metagenomes, though comprising less of total erm gene abundance (9%). Within our erm gene database, this particular sequence cluster is represented by a single gene representative and shares 100% similarity to experimental Clostridium acetobutylicum strains in the NCBI non-redundant gene database (mutant HQ683763.1 and clone HQ25744.1). The overall lack of similar homologous genes in NCBI nr suggests that this specific ermB gene is abundant in manures but is a gene for which we have few sequenced representatives. We identified this gene during our exploration of the effectiveness of current primers on manure metagenomes, and our observations suggest that this gene would benefit from further study given its prevalence. DISCUSSION Over the past 20 years, an abundance of literature has been published quantifying macrolide resistance in agricultural landscapes using qPCR approaches. However, these previous studies often use primers for erm genes designed in only a handful of publications (Table 2). Our study found that current published primer sets, used on their own, are effective at capturing only a subset of the erm diversity in manure samples. For example, if only one primer set were used, less than one-third of erm genes would be detected. To increase our ability to detect erm genes in agricultural systems, we identified the most abundant erm clusters in both swine and cattle manures, identifying the best gene targets for future studies. These genes and their associated primers are recommended for high-throughput qPCR assays that can scale the detection and quantification of these genes for antibiotic gene surveillance. In all amplicon assays, quantifying environmental abundances of gene targets is limited by the effectiveness of primer design. The results presented here emphasize that estimates of abundances of a gene of interest cannot simply be based on primers to genes that have previously been successfully detected. Rather, genes appropriate for antibiotic gene surveillance should be indicative of the spread of resistance (e.g. originate from manure but lacking from pristine environments), representative of diverse hosts (especially those with clinical risks) and accurately represent gene abundances in environmental samples. Our specific effort targeted the erm gene and evaluated the effectiveness of previously published primers sets. The increasing availability of metagenomes makes these evaluations possible, as demonstrated in this study. Although metagenomic sequencing advances will continue to provide powerful tools to understand the broad diversity of resistance in environments, metagenomes are limited by both detection rate and resolution. Short read lengths, the difficulty of assembling many resistance genes (because of their common association with mobile elements containing repeated sequences) and their presence in multiple bacterial hosts challenges the detection of resistance genes using metagenomics. Going forward, high-throughput amplicon assays with strategic gene targets and primer designs are a complementary alternative to help fill these gaps and help us understand the movement of resistance genes among complex environments. SUPPLEMENTARY DATA Supplementary data are available at FEMSEC online. FUNDING This project was supported by the AFRI food safety from the USDA National Institute of Food and Agriculture [grant no. 2016-68003-24604] and the National Pork Board [contract 13-051]. Conflict of interest. None declared. REFERENCES Aarestrup FM, Agersø Y, Ahrens P et al.  . Antimicrobial susceptibility and presence of resistance genes in staphylococci from poultry. Vet Microbiol  2000a; 74: 353– 364. Google Scholar CrossRef Search ADS   Aarestrup FM, Kruse H, Tast E et al.  . Associations between the use of antimicrobial agents for growth promotion and the occurrence of resistance among Enterococcus faecium from broilers and pigs in Denmark, Finland, and Norway. Microb Drug Resist Epidemiol Dis  2000b; 6: 63– 70. Google Scholar CrossRef Search ADS   Ahmad A, Ghosh A, Schal C et al.  . Insects in confined swine operations carry a large antibiotic resistant and potentially virulent enterococcal community. BMC Microbiol  2011; 11: 23. Google Scholar CrossRef Search ADS PubMed  Alexander TW, Yanke JL, Reuter T et al.  . Longitudinal characterization of antimicrobial resistance genes in feces shed from cattle fed different subtherapeutic antibiotics. BMC Microbiol  2011; 11: 19. Google Scholar CrossRef Search ADS PubMed  Beukers AG, Zaheer R, Cook SR et al.  . Effect of in-feed administration and withdrawal of tylosin phosphate on antibiotic resistance in enterococci isolated from feedlot steers. Front Microbiol  2015; 6: 483. Google Scholar CrossRef Search ADS PubMed  Camacho C, Coulouris G, Avagyan V et al.  . BLAST+: architecture and applications. BMC Bioinformatics  2009; 10: 421. Google Scholar CrossRef Search ADS PubMed  Cauwerts K, Decostere A, De Graef EM et al.  . High prevalence of tetracycline resistance in Enterococcus isolates from broilers carrying the erm (B) gene. Avian Pathol  2007; 36: 395– 399. Google Scholar CrossRef Search ADS PubMed  Di Cesare A, Vignaroli C, Luna GM et al.  . Antibiotic-resistant enterococci in seawater and sediments from a coastal fish farm. Microb Drug Resist  2012; 18: 502– 509. Google Scholar CrossRef Search ADS PubMed  Chee-Sanford JC, Mackie RI, Koike S et al.  . Fate and transport of antibiotic residues and antibiotic resistance genes following land application of manure waste. J Environ Qual  2009; 38: 1086. Google Scholar CrossRef Search ADS PubMed  Chen J, Michel FC, Sreevatsan S et al.  . Occurrence and persistence of erythromycin resistance genes (erm) and tetracycline resistance genes (tet) in waste treatment systems on swine farms. Microb Ecol  2010; 60: 479– 486. Google Scholar CrossRef Search ADS PubMed  Chen J, Yu Z, Michel FC et al.  . Development and application of real-time PCR assays for quantification of erm genes conferring resistance to Macrolides-Lincosamides-Streptogramin B in livestock manure and manure management systems. Appl Environ Microb  2007; 73: 4407– 16. Google Scholar CrossRef Search ADS   Chénier MR, Juteau P. Fate of chlortetracycline- and tylosin-resistant bacteria in an aerobic thermophilic sequencing batch reactor treating swine waste. Microb Ecol  2009; 58: 86– 97. Google Scholar CrossRef Search ADS PubMed  Choi J, Yang F, Stepanauskas R et al.  . Strategies to improve reference databases for soil microbiomes. ISME J  2017; 11: 829– 34. Google Scholar CrossRef Search ADS PubMed  Edgar RC. MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res  2004; 32: 1792– 7. Google Scholar CrossRef Search ADS PubMed  Ekizoglu M, Koike S, Krapac I et al.  . Phenotypic and genotypic characterization of antibiotic-resistant soil and manure bacteria adjacent to swine production facilities. Turkish J Vet Anim Sci  2013; 37: 504– 511. Google Scholar CrossRef Search ADS   Fahrenfeld N, Knowlton K, Krometis LA et al.  . Effect of manure application on abundance of antibiotic resistance genes and their attenuation rates in soil: Field-scale mass balance approach. Environ Sci Technol  2014; 48: 2643– 50. Google Scholar CrossRef Search ADS PubMed  Fang H, Wang H, Cai L et al.  . Prevalence of antibiotic resistance genes and bacterial pathogens in long-term manured greenhouse soils as revealed by metagenomic survey. Environ Sci Technol  2015; 49: 1095– 104. Google Scholar CrossRef Search ADS PubMed  Fish JA, Chai B, Wang Q et al.  . FunGene: The Functional Gene Pipeline and Repository . 2013, DOI: https://doi.org/10.3389/fmicb.2013.00291. Fitzpatrick D, Walsh F. Antibiotic resistance genes across a wide variety of metagenomes. FEMS Microbiol Ecol  2016; 92: fiv168. Google Scholar CrossRef Search ADS PubMed  Fu L, Niu B, Zhu Z et al.  . CD-HIT: accelerated for clustering the next-generation sequencing data. Bioinformatics  2012; 28: 3150– 2. Google Scholar CrossRef Search ADS PubMed  Garder JL, Moorman TB, Soupir ML. Transport and persistence of tylosin-resistant enterococci, genes, and tylosin in soil and drainage water from fields receiving swine manure. J Environ Qual  2014; 43: 1484. Google Scholar CrossRef Search ADS PubMed  Garofalo C, Vignaroli C, Zandri G et al.  . Direct detection of antibiotic resistance genes in specimens of chicken and pork meat. Int J Food Microbiol  2007; 113: 75– 83. Google Scholar CrossRef Search ADS PubMed  Gillings MR, Gaze WH, Pruden A et al.  . Using the class 1 integron-integrase gene as a proxy for anthropogenic pollution. ISME J  2015; 9: 1269– 79. Google Scholar CrossRef Search ADS PubMed  Heuer H, Schmitt H, Smalla K. Antibiotic resistance gene spread due to manure application on agricultural fields. Curr Opin Microbiol  2011; 14: 236– 43. Google Scholar CrossRef Search ADS PubMed  Hoang TTT, Soupir ML, Liu P et al.  . Occurrence of tylosin-resistant enterococci in swine manure and tile drainage systems under no-till management. water, air. Soil Pollut  2013; 224: 1754. Google Scholar CrossRef Search ADS   Holman DB, Chénier MR, PJ F-C et al.  . Impact of subtherapeutic administration of tylosin and chlortetracycline on antimicrobial resistance in farrow-to-finish swine. FEMS Microbiol Ecol  2013; 85: 1– 13. Google Scholar CrossRef Search ADS PubMed  Jackson CR, Fedorka-Cray PJ, Barrett JB et al.  . Effects of tylosin use on erythromycin resistance in enterococci isolated from swine. Appl Environ Microbiol  2004; 70: 4205– 10. Google Scholar CrossRef Search ADS PubMed  Jensen LB, Agersø Y, Sengeløv G. Presence of erm genes among macrolide-resistant Gram-positive bacteria isolated from Danish farm soil. Environ Int  2002; 28: 487– 491. Google Scholar CrossRef Search ADS PubMed  Jensen LB, Frimodt-Moller N, Aarestrup FM. Presence of erm gene classes in Gram-positive bacteria of animal and human origin in Denmark. FEMS Microbiol Lett  1999; 170: 151– 8. Google Scholar CrossRef Search ADS PubMed  Just NA, Létourneau V, Kirychuk SP et al.  . Potentially pathogenic bacteria and antimicrobial resistance in bioaerosols from cage-housed and floor-housed poultry operations. Ann Occup Hyg  2011; 71: 6926– 33. Joy SR, Bartelt-Hunt SL, Snow DD et al.  . Fate and transport of antimicrobials and antimicrobial resistance genes in soil and runoff following land application of swine manure slurry. Environ Sci Technol  2013; 47: 12081– 8. Google Scholar CrossRef Search ADS PubMed  Joy SR, Li X, Snow DD et al.  . Fate of antimicrobials and antimicrobial resistance genes in simulated swine manure storage. Sci Total Environ  2014; 481: 69– 74. Google Scholar CrossRef Search ADS PubMed  Kalmokoff M, Waddington LM, Thomas M et al.  . Continuous feeding of antimicrobial growth promoters to commercial swine during the growing/finishing phase does not modify faecal community erythromycin resistance or community structure. J Appl Microbiol  2011; 110: 1414– 1425. Google Scholar CrossRef Search ADS PubMed  Karkman A, Johnson TA, Lyra C et al.  . High-throughput quantification of antibiotic resistance genes from an urban wastewater treatment plant. FEMS Microbiol Ecol  2016; 92: 1– 7. Google Scholar CrossRef Search ADS   Knapp CW, Dolfing J, Ehlert PAI et al.  . Evidence of increasing antibiotic resistance gene abundances in archived soils since 1940. Environ Sci Technol  2010; 44: 580– 7. Google Scholar CrossRef Search ADS PubMed  Koike S, Aminov RI, Yannarell AC et al.  . Molecular ecology Of macrolide–lincosamide–streptogramin B methylases in waste lagoons and subsurface waters associated with swine production. Microb Ecol  2010; 59: 487– 98. Google Scholar CrossRef Search ADS PubMed  Leclercq R, Courvalin P. Bacterial resistance to macrolide, lincosamide, and streptogramin antibiotics by target modification. Antimicrob Agents Chemother  1991; 35: 1267– 72. Google Scholar CrossRef Search ADS PubMed  Leener EDe, Martel A, De Graef E. Molecular analysis of human, porcine, and poultry Enterococcus faecium isolates and their erm (B) genes. Appl  2005. Available at http://aem.asm.org/content/71/5/2766.short (verified 18 July 2017). Lerma LL, Benomar N, del M et al.  . Antibiotic multiresistance analysis of mesophilic and psychrotrophic pseudomonas spp. isolated from goat and lamb slaughterhouse surfaces throughout the meat production process. Appl Environ Microbiol  2014; 80: 6792– 6806. Google Scholar CrossRef Search ADS PubMed  Li B, Yang Y, Ma L et al.  . Metagenomic and network analysis reveal wide distribution and co-occurrence of environmental antibiotic resistance genes. ISME J  2015; 9: 2490– 502. Google Scholar CrossRef Search ADS PubMed  Li W, Godzik A. Cd-hit: a fast program for clustering and comparing large sets of protein or nucleotide sequences. Bioinformatics  2006; 22: 1658– 9. Google Scholar CrossRef Search ADS PubMed  Liu B, Pop M. ARDB–Antibiotic resistance genes database. Nucleic Acids Res  2009; 37: D443– 7. Google Scholar CrossRef Search ADS PubMed  Luby EM, Moorman TB, Soupir ML. Fate and transport of tylosin-resistant bacteria and macrolide resistance genes in artificially drained agricultural fields receiving swine manure. Sci Total Environ  2016; 550: 1126– 33. Google Scholar CrossRef Search ADS PubMed  Luthje P, Schwarz S. Antimicrobial resistance of coagulase-negative staphylococci from bovine subclinical mastitis with particular reference to macrolide-lincosamide resistance phenotypes and genotypes. J Antimicrob Chemother  2006; 57: 966– 969. Google Scholar CrossRef Search ADS PubMed  Martel A, Devriese L, Decostere A et al.  . Presence of macrolide resistance genes in streptococci and enterococci isolated from pigs and pork carcasses. Int J Food Microbiol  2003a; 84: 27– 32. Google Scholar CrossRef Search ADS   Martel A, Meulenaere V, Devriese LA et al.  . Macrolide and lincosamide resistance in the gram-positive nasal and tonsillar flora of pigs. Microb Drug Resist  2003b; 9: 293– 7. Google Scholar CrossRef Search ADS   McEwen SA, Fedorka-Cray PJ. Antimicrobial use and resistance in animals. Clin Infect Dis  2002; 34: S93– 106. Google Scholar CrossRef Search ADS PubMed  Muziasari WI, Pärnänen K, Johnson TA et al.  . Aquaculture changes the profile of antibiotic resistance and mobile genetic element associated genes in Baltic Sea sediments. FEMS Microbiol Ecol  2016; 92: fiw052. Google Scholar CrossRef Search ADS PubMed  NCBI Resource Coordinators. Database resources of the National Center for Biotechnology Information. Nucleic Acids Res  2017; 45: D12– 7. CrossRef Search ADS PubMed  Negreanu Y, Pasternak Z, Jurkevitch E et al.  . Impact of treated wastewater irrigation on antibiotic resistance in agricultural soils. Environ Sci Technol  2012; 46: 4800– 4808. Google Scholar CrossRef Search ADS PubMed  Noyes N, Yang X, Linke L et al.  . Resistome diversity in cattle and the environment decreases during beef production. Elife  2016; 5: e13195. Google Scholar CrossRef Search ADS PubMed  Patterson AJ, Colangeli R, Spigaglia P et al.  . Distribution of specific tetracycline and erythromycin resistance genes in environmental samples assessed by macroarray detection. Environ Microbiol  2007; 9: 703– 15. Google Scholar CrossRef Search ADS PubMed  Petersen A, Dalsgaard A. Species composition and antimicrobial resistance genes of Enterococcus spp., isolated from integrated and traditional fish farms in Thailand. Environ Microbiol  2003; 5: 395– 402. Google Scholar CrossRef Search ADS PubMed  Popowska M, Rzeczycka M, Miernik A et al.  . Influence of soil use on prevalence of tetracycline, streptomycin, and erythromycin resistance and associated resistance genes. Antimicrob Agents Chemother  2012; 56: 1434– 43. Google Scholar CrossRef Search ADS PubMed  Price MN, Dehal PS, Arkin AP. FastTree 2 – Approximately maximum-likelihood trees for large alignments. PLoS One . 2010; 5: e9490 Pyörälä S, Baptiste KE, Catry B et al.  . Macrolides and lincosamides in cattle and pigs: Use and development of antimicrobial resistance. Vet J  2014; 200: 230– 9. Google Scholar CrossRef Search ADS PubMed  Roberts MC, Dueck M, Hoban D et al.  . Update on macrolide-lincosamide-streptogramin, ketolide, and oxazolidinone resistance genes. FEMS Microbiol Lett  2008; 282: 147– 59. Google Scholar CrossRef Search ADS PubMed  Roberts MC, Sutcliffe J, Courvalin P et al.  . Nomenclature for macrolide and macrolide-lincosamide-streptogramin B resistance determinants. Antimicrob Agents Ch  1999; 43: 2823– 30. Sandberg KD, LaPara TM. The fate of antibiotic resistance genes and class 1 integrons following the application of swine and dairy manure to soils (K Smalla, Ed.). FEMS Microbiol Ecol  2016; 92: fiw001. Google Scholar CrossRef Search ADS PubMed  Sharma R, Larney FJ, Chen J et al.  . Selected Antimicrobial resistance during composting of manure from cattle administered sub-therapeutic antimicrobials. J Environ Qual  2009; 38: 567. Google Scholar CrossRef Search ADS PubMed  Shi P, Jia S, Zhang X-X et al.  . Metagenomic insights into chlorination effects on microbial antibiotic resistance in drinking water. Water Res  2013; 47: 111– 20. Google Scholar CrossRef Search ADS PubMed  Soni B, Bartelt-Hunt SL, Snow DD et al.  . Narrow grass hedges reduce tylosin and associated antimicrobial resistance genes in agricultural runoff. J Environ Qual  2015; 44: 895. Google Scholar CrossRef Search ADS PubMed  Staley C, Unno T, Gould TJ et al.  . Application of Illumina next-generation sequencing to characterize the bacterial community of the Upper Mississippi River. J Appl Microbiol  2013; 115: 1147– 58. Google Scholar CrossRef Search ADS PubMed  Stedtfeld RD, Stedtfeld TM, Waseem H et al.  . Isothermal assay targeting class 1 integrase gene for environmental surveillance of antibiotic resistance markers. 2017, DOI: https://doi.org/10.1016/j.jenvman.2017.04.079. Sutcliffe J, Grebe T, Tait-Kamradt A et al.  . Detection of erythromycin-resistant determinants by PCR. Antimicrob. Agents Chemother . 1996; 40(11): 2562– 6. Vaz-Moreira I, Nunes OC, Manaia CM et al.  . Bacterial diversity and antibiotic resistance in water habitats: searching the links with the human microbiome. FEMS Microbiol Rev  2014; 38: 761– 78. Google Scholar CrossRef Search ADS PubMed  Vester B, Douthwaite S. Macrolide resistance conferred by base substitutions in 23S rRNA. Antimicrob Agents Ch  2001; 45: 1– 12. Google Scholar CrossRef Search ADS   Wang F-H, Qiao M, Su J-Q et al.  . High throughput profiling of antibiotic resistance genes in urban park soils with reclaimed water irrigation. Environ Sci Technol  2014; 48: 9079– 85. Google Scholar CrossRef Search ADS PubMed  Wang L, Gutek A, Grewal S et al.  . Changes in diversity of cultured bacteria resistant to erythromycin and tetracycline in swine manure during simulated composting and lagoon storage. Lett Appl Microbiol  2015; 61: 245– 251. Google Scholar CrossRef Search ADS PubMed  Wang L, Oda Y, Grewal S et al.  . Persistence of Resistance to Erythromycin and Tetracycline in Swine Manure During Simulated Composting and Lagoon Treatments. Microb Ecol  2012; 63: 32– 40. Google Scholar CrossRef Search ADS PubMed  Wang Y, Wang G-R, Shoemaker NB et al.  . Distribution of the ermG gene among bacterial isolates from porcine intestinal contents. Appl Environ Microbiol  2005; 71: 4930– 4. Google Scholar CrossRef Search ADS PubMed  Weisblum B. Macrolide resistance. Drug Resist Updat  1998; 1: 29– 41. Google Scholar CrossRef Search ADS PubMed  Whitehead TR, Cotta MA. Stored swine manure and swine faeces as reservoirs of antibiotic resistance genes. Lett Appl Microbiol  2013; 56: 264– 7. Google Scholar CrossRef Search ADS PubMed  WHO. Global Action Plan on Antimicrobial Resistance . 2017. http://www.wpro.who.int/entity/drug_resistance/resources/global_action_plan_eng.pdf. Xu S, Sura S, Zaheer R et al.  . Dissipation of antimicrobial resistance determinants in composted and stockpiled beef cattle manure. J Environ Qual  2016; 45: 528– 6. Google Scholar CrossRef Search ADS PubMed  Zou L-K, Wang H-N, Zeng B et al.  . Erythromycin resistance and virulence genes in Enterococcus faecalis from swine in China. New Microbiol  2011; 34: 73– 80. Google Scholar PubMed  © FEMS 2018. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.

Journal

FEMS Microbiology EcologyOxford University Press

Published: Apr 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off