Discussion
The target of antibiotic resistance genes in natural environments should be considered, specially if we consider the evidence that these genes have a long evolutionary history in natural environments
( ( class="ltx_cite" data-bib-text="@article{Aminov_2009,
doi = {10.1111/j.1462-2920.2009.01972.x},
url = {http://dx.doi.org/10.1111/j.1462-2920.2009.01972.x},
year = 2009,
...
author = {Rustam I. Aminov},
title = {The role of antibiotics and antibiotic resistance in nature},
journal = {Environmental Microbiology}
}" data-bib-key="Aminov_2009" contenteditable="false">
Aminov 2009,
class="ltx_cite" data-bib-text="@article{Hall_2004,
doi = {10.1016/j.drup.2004.02.003},
url = {http://dx.doi.org/10.1016/j.drup.2004.02.003},
year = 2004,
...
title = {Evolution of the serine $\upbeta$-lactamases: past,
present and future},
journal = {Drug Resistance Updates}
}" data-bib-key="Hall_2004" contenteditable="false">
Hall 2004,
class="ltx_cite" data-bib-text="@article{Garau_2005,
doi = {10.1128/aac.49.7.2778-2784.2005},
url = {http://dx.doi.org/10.1128/aac.49.7.2778-2784.2005},
year = 2005,
...
author = {G. Garau and A. M. Di Guilmi and B. G. Hall},
title = {Structure-Based Phylogeny of the Metallo-~-Lactamases},
journal = {Antimicrobial Agents and Chemotherapy}
}" data-bib-key="Garau_2005" contenteditable="false">
Garau 2005). Thus, natural environments appear such as a reservoir of potential ARGs to pathogenic bacteria
( ( class="ltx_cite" data-bib-text="@article{Berglund_2015,
doi = {10.3402/iee.v5.28564},
url = {http://dx.doi.org/10.3402/iee.v5.28564},
year = 2015,
...
author = {Björn Berglund},
title = {Environmental dissemination of antibiotic resistance genes and correlation to anthropogenic contamination with antibiotics},
journal = {Infection Ecology {\&} Epidemiology}
}" data-bib-key="Berglund_2015" contenteditable="false">
Berglund 2015,
class="ltx_cite" data-bib-text="@article{Versluis_2015,
doi = {10.1038/srep11981},
url = {http://dx.doi.org/10.1038/srep11981},
year = 2015,
...
author = {Dennis Versluis and Marco Maria D'Andrea and Javier Ramiro Garcia and Milkha M. Leimena and Floor Hugenholtz and Jing Zhang and Ba{\c{s}}ak Öztürk and Lotta Nylund and Detmer Sipkema and Willem van Schaik and Willem M. de Vos and Michiel Kleerebezem and Hauke Smidt and Mark W.J. van Passel},
title = {Mining microbial metatranscriptomes for expression of antibiotic resistance genes under natural conditions},
journal = {Sci. Rep.}
}" data-bib-key="Versluis_2015" contenteditable="false">
Versluis 2015) that cannot be ignored.
The results obtained in this study show the high content of b-lactamase genes in each analyzed environments, but a deep analysis indicate important differences in the diversity of those genes. Thus, soil and glacier metagenomes exhibit the higher diversity of b-lactamase genes (40 different b-lactamase genes in average), followed by wastewater metagenomes (32 different b-lactamase genes). The high diversity of b-lactamases in soils is in accordance with a previous metagenomic-based study of antibiotic resistance genes in the environment
( ( class="ltx_cite" data-bib-text="@article{Nesme_2014,
doi = {10.1016/j.cub.2014.03.036},
url = {http://dx.doi.org/10.1016/j.cub.2014.03.036},
year = 2014,
...
author = {Joseph Nesme and S{\'{e}}bastien C{\'{e}}cillon and Tom~O. Delmont and Jean-Michel Monier and Timothy~M. Vogel and Pascal Simonet},
title = {Large-Scale Metagenomic-Based Study of Antibiotic Resistance in the Environment},
journal = {Current Biology}
}" data-bib-key="Nesme_2014" contenteditable="false">
Nesme 2014), in which soil metagenomes showed the most diverse pool of ARGs than human feces, ocean, cow and chicken gut and artic snow metagenomes. In the opposite, human gut, ocean and cow gut metagenomes, exhibit the lowest diversity values (19, 17 and 12 different type of b-lactamase genes respectively). Richness was other important diversity index analyzed, and our results indicate that not always a high level of diversity correlated with richness. Thus, soil metagenomes, glaciers and wastewater metagenomes were the most b-lactamase diverse environments; however the most rich b-lactamase environments were agricultural soils, human gut, non-agricultural soils and wasterwater (in average 0.0106, 0.0105, 0.0093 and 0.0092% of reads identified like b-lactamases from the total number of read per environment). In contrast, glacier metagenomes that exhibited a high b-lactamase diversity, show in average an intermediate b-lactamase richness (0.0068% of total number of reads).
Differences in b-lactamase content of agricultural and non-agricultural soil metagenomes could be attributed to agricultural practices such as manure amendment, fertilization, irrigation and more.
Thus, Thus, class="ltx_cite" data-bib-text="@article{Udikovic_Kolic_2014,
doi = {10.1073/pnas.1409836111},
url = {http://dx.doi.org/10.1073/pnas.1409836111},
year = 2014,
...
author = {Nikolina Udikovic-Kolic and Fabienne Wichmann and Nichole A. Broderick and Jo Handelsman},
title = {Bloom of resident antibiotic-resistant bacteria in soil following manure fertilization},
journal = {Proceedings of the National Academy of Sciences}
}" data-bib-key="Udikovic_Kolic_2014" contenteditable="false">
Udikovic-Kolic 2014 et al. (format) recently demonstrate an increase in ARG (specially in b-lactamase content) in soils amendment with manure compared to the ARG content in soil fertilized with NPK. Interestingly, the cattle that produced the manure had not been treated with antibiotics. The authors concluded that manure application produced a bloom of certain bacterial resistant populations without connection to antibiotic exposure.
The presence of b-lactamase genes in low anthropogenic impacted environment such glaciers, have been reported previously.
The analysis of the supporting information accompanying the metagenome sequences, clearly indicate that different factor affect the b-lactamase diversity values here obtained. In this context is important indicate that despite that all the metagenomes share the same type of technology used in the sequencing process, differences in methodology can be expected. In the same context, despite that metagenomes can be grouped according different environments, some differences can affect the sequencing results; thus, in the ocean environment, metagenomes related to the Sydney harbour metagenome project (see Table S1) were obtained at 0,3 m in depth, in the estuary of Sydney, and metagenomes related to Amazon continuum metagenomes project were obtained at depths ranged from 3.63 to 4.47 meters, and the samples were obtained at different points, from seawater close to the cost line until seawater far from the cost line (open ocean). For instance, some studies have been show the differences in the AR profile across different water layers in different lake samples
( ( data-bib-text="@article{Mudryk_2000,
doi = {10.1006/ecss.2000.0719},
url = {http://dx.doi.org/10.1006/ecss.2000.0719},
year = 2000,
...
title = {Occurrence and Activity of Lipolytic Bacterioneuston and Bacterioplankton in the Estuarine Lake Gardno},
journal = {Estuarine,
Coastal and Shelf Science}
}" data-bib-key="Mudryk_2000"
contenteditable="false">Mudryk 2000, contenteditable="false" style="cursor: pointer">href="#Mudryk_2000">Mudryk 2000, )
In the same way,
href="#Di_Cesare_2015">Cesare 2015). In relation to soil metagenomes, agricultural metagenomes were exposed to different agricultural practices that can affect the microbiological content on soil and therefor, affect the genomic content on soil.
Finally, Ilustrating the last point, Agga 2015 (formato) demonstrated that manure applications to agricultural soils affect the bacterial community and the profile of ARGs in soils. Finally, in the case of gut metagenomes, clearly the health situation and/or the drug supply can affect both the microbiological and genomic content. Our analysis take into account those variability, and steps oriented to reduce it were applied.
Supporting the differences in b-lactamase content and diversity, our indicator species analysis show habitat-specificity for some b-lactamase genes. Interestingly, four b-lactamase genes (blaEBR, CfxA, mecA HGI) showed a high faithfulness of occurrence in the human gut environment; those genes have been previously identified only in humans (both health and sick)
( ( class="ltx_cite" data-bib-text="@article{Bellais_2002,
doi = {10.1128/aac.46.10.3223-3227.2002},
url = {http://dx.doi.org/10.1128/aac.46.10.3223-3227.2002},
year = 2002,
...
title = {{EBR}-1,
a Novel Ambler Subclass B1 ~-Lactamase from Empedobacter brevis},
journal = {Antimicrobial Agents and Chemotherapy}
}" data-bib-key="Bellais_2002" contenteditable="false">
Bellais 2002,
class="ltx_cite" data-bib-text="@article{Avelar_2003,
doi = {10.1016/s0923-2508(03)00093-7},
url = {http://dx.doi.org/10.1016/s0923-2508(03)00093-7},
year = 2003,
...
author = {K{\'{a}}tia Eliane Santos Avelar and Koko Otsuki and Ana Carolina Paulo Vicente and Jessica Manya Bittencourt Dias Vieira and Geraldo Renato de Paula and Regina Maria Cavalcanti Pilotto Domingues and Maria Candida de Souza Ferreira},
title = {Presence of the {cfxA} gene in Bacteroides distasonis},
journal = {Research in Microbiology}
}" data-bib-key="Avelar_2003" contenteditable="false">
Avelar 2003,
class="ltx_cite" data-bib-text="@article{Sommer_2009,
doi = {10.1126/science.1176950},
url = {http://dx.doi.org/10.1126/science.1176950},
year = 2009,
...
author = {M. O. A. Sommer and G. Dantas and G. M. Church},
title = {Functional Characterization of the Antibiotic Resistance Reservoir in the Human Microflora},
journal = {Science}
}" data-bib-key="Sommer_2009" contenteditable="false">
Sommer 2009,
class="ltx_cite" data-bib-text="@article{Ballhausen_2014,
doi = {10.1128/aac.02731-13},
url = {http://dx.doi.org/10.1128/aac.02731-13},
year = 2014,
...
author = {B. Ballhausen and A. Kriegeskorte and N. Schleimer and G. Peters and K. Becker},
title = {The {mecA} Homolog {mecC} Confers Resistance against ~-Lactams in Staphylococcus aureus Irrespective of the Genetic Strain Background},
journal = {Antimicrobial Agents and Chemotherapy}
}" data-bib-key="Ballhausen_2014" contenteditable="false">
Ballhausen 2014) or in hospital and municipal wastewater systems
( ( class="ltx_cite" data-bib-text="@article{Bockelmann_2008,
doi = {10.1128/aem.01649-08},
url = {http://dx.doi.org/10.1128/aem.01649-08},
year = 2008,
...
author = {U. Bockelmann and H.-H. Dorries and M. N. Ayuso-Gabella and M. Salgot de Marcay and V. Tandoi and C. Levantesi and C. Masciopinto and E. Van Houtte and U. Szewzyk and T. Wintgens and E. Grohmann},
title = {Quantitative {PCR} Monitoring of Antibiotic Resistance Genes and Bacterial Pathogens in Three European Artificial Groundwater Recharge Systems},
journal = {Applied and Environmental Microbiology}
}" data-bib-key="Bockelmann_2008" contenteditable="false">
Bockelmann 2008,
class="ltx_cite" data-bib-text="@article{Schwartz_2003,
doi = {10.1111/j.1574-6941.2003.tb01073.x},
url = {http://dx.doi.org/10.1111/j.1574-6941.2003.tb01073.x},
year = 2003,
...
title = {Detection of antibiotic-resistant bacteria and their resistance genes in wastewater,
surface water, and drinking water biofilms},
journal = {{FEMS} Microbiology Ecology}
}" data-bib-key="Schwartz_2003" contenteditable="false">
Schwartz 2003), but there are no reports of their occurrence in natural environments. In addition, this analysis was performed when environments were categorized according degrees of human impact (see results); thus, the four previous genes that shown a high probability of be present in human gut metagenomes, exhibit again the same trend, this mean show the same high probability to be present in the category of higher anthropogenic impact, but more interesting is that in the category of the lowest anthropogenic impact, that include glacier and non-agricultural soil metagenomes, 49 b-lactamase genes show a high faithfulness of occurrence. On the whole, our results suggest that highly anthropogenic impacted environments select for only few specific b-lactamases, while less impacted environments, probably with a lower level of selective pressure, may content lower levels of b-lactamases but with a high b-lactamase diversity.
Our gene network results show a high number of connections between nodes from the same type of environment, but only few connections between nodes of two different environments; suggesting that b-lactamase genes show a high intra-habitat mobility, but a limited inter-habitat transferability. Our findings are in concordance with metagenomic (Fondi et al. 2016 and
class="ltx_cite" data-bib-text="@article{Li_2015,
doi = {10.1038/ismej.2015.59},
url = {http://dx.doi.org/10.1038/ismej.2015.59},
year = 2015,
...
author = {Bing Li and Ying Yang and Liping Ma and Feng Ju and Feng Guo and James M Tiedje and Tong Zhang},
title = {Metagenomic and network analysis reveal wide distribution and co-occurrence of environmental antibiotic resistance genes},
journal = {The {ISME} Journal}
}" data-bib-key="Li_2015" contenteditable="false">
Li 2015,
2015, class="ltx_cite" data-bib-text="@article{Agga_2015,
doi = {10.1371/journal.pone.0132586},
url = {http://dx.doi.org/10.1371/journal.pone.0132586},
year = 2015,
...
editor = {Zhi Zhou},
title = {Antimicrobial-Resistant Bacterial Populations and Antimicrobial Resistance Genes Obtained from Environments Impacted by Livestock and Municipal Waste},
journal = {{PLOS} {ONE}}
}" data-bib-key="Agga_2015" contenteditable="false">
Agga 2015) and more classical approaches (Forsberg
2014, 2014, class="ltx_cite" data-bib-text="@article{Smillie_2011,
doi = {10.1038/nature10571},
url = {http://dx.doi.org/10.1038/nature10571},
year = 2011,
...
author = {Chris S. Smillie and Mark B. Smith and Jonathan Friedman and Otto X. Cordero and Lawrence A. David and Eric J. Alm},
title = {Ecology drives a global network of gene exchange connecting the human microbiome},
journal = {Nature}
}" data-bib-key="Smillie_2011" contenteditable="false">
Smillie 2011), which reported an specificity of ARGs with analyzed environments, with a high degree of intra-habitat mobility of ARGs and few probabilities of genetic ARG transfer between different environmnets. In this context, different factors such as MGEs (specially in b-lactamase flanking regions), founder effect, ecological connectivity, fitness cost or second-order selection may limit the gene transfer of ARGs between environments
( ( class="ltx_cite" data-bib-text="@article{Mart_nez_2012,
doi = {10.3389/fmicb.2011.00265},
url = {http://dx.doi.org/10.3389/fmicb.2011.00265},
year = 2012,
...
OXY, OKP and LEN protein variation databases (
Gatica et al., 2016). Sequences deposited into EX-B were compared using
BioEdit 7.2.5 software and checked for non-redundant sequences; producing a
database containing 1566 non-redundant β-lactamase sequences. The EX-B
database is available for download at (
http://app.agri.gov.il/eddie/EX-B_Database/EX-B_Database.fastal).
Blast searches
Homology searches among metagenomic and b-lactamase sequences were performed using blastx from the BLAST suite
( ( class="ltx_cite" data-bib-text="@article{Gish_1993,
doi = {10.1038/ng0393-266},
url = {http://dx.doi.org/10.1038/ng0393-266},
year = 1993,
...
author = {Warren Gish and David J. States},
title = {Identification of protein coding regions by database similarity search},
journal = {Nature Genetics}
}" data-bib-key="Gish_1993" contenteditable="false">
Gish 1993). Only hits with a percent of identity higher than 50%, bit score higher than 30 and a e-value lower than 1e-4 were considered real hits. The 50% of identity was used under the consideration that the databases used to construct the EX-B database are highly based in clinical data and generally it not consider environmental data.
Hits analysis
Hits obtained by BLAST were analyzed in PC-ORD 5.0 (McCune and Mefford, 2011).The hits were used to construct a matrix of present hits, according b-lactamase gene type, in each metagenome. The data was relativizated by weighting by ubiquity and transformed by square root. Outliers were identified and removed of analysis and a distance matrix based on Bray-Curtis distance was constructed to downstream analysis. In addition, the distance matrix constructed was used to obtain diversity indexes, multi response permutation process (MRPP), indicator species analysis and non metric multidimensional scaling.
The Distance matrix from the previous step was used to construct a b-lactamase gene network using EDENetwork 2.18
( ( class="ltx_cite" data-bib-text="@article{Kivel__2014,
doi = {10.1111/1755-0998.12290},
url = {http://dx.doi.org/10.1111/1755-0998.12290},
year = 2014,
...
title = {{EDENetworks}: A user-friendly software to build and analyse networks in biogeography,
ecology and population genetics},
journal = {Molecular Ecology Resources}
}" data-bib-key="Kivel__2014" contenteditable="false">
Kivelä 2014). Graphical visualization and statistical test were performed using Cytoscape 3.4.0
( ( class="ltx_cite" data-bib-text="@article{Shannon_2003,
doi = {10.1101/gr.1239303},
url = {http://dx.doi.org/10.1101/gr.1239303},
year = 2003,