Article pubs.acs.org/est
Long-Term Impact of Field Applications of Sewage Sludge on Soil Antibiotic Resistome Wan-Ying Xie,† Steve P. McGrath,‡ Jian-Qiang Su,§ Penny R. Hirsch,‡ Ian M. Clark,‡ Qirong Shen,† Yong-Guan Zhu,§ and Fang-Jie Zhao*,†,‡ †
Jiangsu Key Laboratory for Organic Waste Utilization, Jiangsu Collaborative Innovation Center for Solid Organic Waste Resource Utilization, College of Resources and Environmental Sciences, Nanjing Agricultural University, Nanjing 210095, China ‡ Rothamsted Research, Harpenden, Hertfordshire AL5 2JQ, United Kingdom § Key Lab of Urban Environment and Health, Institute of Urban Environment, Chinese Academy of Sciences, Xiamen 361021, China S Supporting Information *
ABSTRACT: Land applications of municipal sewage sludge may pose a risk of introducing antibiotic resistance genes (ARGs) from urban environments into agricultural systems. However, how the sewage sludge recycling and application method influence soil resistome and mobile genetic elements (MGEs) remains unclear. In the present study, high throughput quantitative PCR was conducted on the resistome of soils from a field experiment with past (between 1994 and 1997) and annual (since 1994) applications of five different sewage sludges. Total inputs of organic carbon were similar between the two modes of sludge applications. Intrinsic soil resistome, defined as the ARGs shared by the soils in the control and sludge-amended plots, consisted of genes conferring resistance to multidrug, β-lactam, Macrolide-Lincosamide-Streptogramin B (MLSB), tetracycline, vancomycin, and aminoglycoside, with multidrug resistance genes as the most abundant members. There was a strong correlation between the abundance of ARGs and MGE marker genes in soils. The composition and diversity of ARGs in the five sludges were substantially different from those in soils. Considerable proportions of ARGs and MGE marker genes in the sludges attenuated following the application, especially aminoglycoside and tetracycline resistance genes. Annual applications posed a more significant impact on the soil resistome, through both continued introduction and stimulation of the soil intrinsic ARGs. In addition, direct introduction of sludge-specific ARGs into soil was observed especially from ARG-rich sludge. These results provide a better insight into the characteristics of ARG dissemination from urban environment to the agricultural system through sewage sludge applications.
■
on the overall effect of sewage sludge applications on the indigenous soil resistome.14 Ecological succession in the soil resistome is related to multidimensional factors such as exogenous loading of ARGs, varying the microbial community by organic matter introduction, changes in selective pressure and horizontal gene transfer (HGT).15−17 Persistence or attenuation of ARGs following sludge or manure applications have been observed in various environments.18,19 For instance, significant accumulation of sul1 and sul2 was observed in soils two months after receiving a sequential amendment of sulfadiazine-contained manure in a pot experiment.20 In soil microcosms amended with municipal wastewater solids, macrolide resistance gene erm(B) was found to attenuate significantly in around 2 weeks.21 These studies,
INTRODUCTION The issue of antibiotic resistance genes (ARGs) has drawn intense attention in recent years due to their increasing prevalence and the potential risk to human health.1−3 As a major source of antibiotics, the soil environment sustains a large reservoir of intrinsic ARGs,2,4 the dynamic of which could be disturbed by anthropogenic activities.2,5 Wastewater treatment plants (WWTPs) gathering the partially metabolized and unused antibiotics from urban environment are hotspots of antibiotic resistance proliferation.6−10 Due to the ban on sea disposal in the EU, sewage sludges from WWTPs are increasingly being disposed on land.11 Appropriate uses of sewage sludge on agricultural land can increase soil organic matter and recycle nutrients to the soil, but may also introduce highly abundant ARGs into agricultural systems.12 On the basis of the direct interrogation of ARGs in sewage sludge, a number of reports suggested that applications of sewage sludge on fields may lead to the spread and enrichment of ARGs in soils.12−14 However, there are few field studies to allow a comprehensive assessment © 2016 American Chemical Society
Received: Revised: Accepted: Published: 12602
April 29, 2016 November 6, 2016 November 16, 2016 November 16, 2016 DOI: 10.1021/acs.est.6b02138 Environ. Sci. Technol. 2016, 50, 12602−12611
Article
Environmental Science & Technology
Carlsbad, CA) according to the manufacturer’s protocol. DNA extraction was conducted five times for each sample (2.5 g in total), and the extracts were combined and purified with a PowerClean DNA Clear-up Kit (MoBio Laboratories, Carlsbad, CA). The total DNA was eluted with 100 μL of solution 7 (10 mM Tris without EDTA, pH 8) provided in the clear-up kit and stored at −80 °C until further analysis. DNA quality was checked in a NanoDrop 2000C spectrophotometer (Thermo Scientific, Wilmington, U.S.A.), and the concentration was measured with a Quant-iT PicoGreen double-stranded DNA (dsDNA) assay kit (Invitrogen, U.S.A.). High-Throughput Quantitative PCR and Data Analysis. HT-qPCR was performed according to previous studies employing the platform of SmartChip Real-time PCR system (Wafergen Inc., USA), including a MSND multisample nanodispenser for sample dispersion on Smart Chips and a SmartChip Cycler for qPCR.8,12 The design and validation of the original batch of primer sets were conducted by Looft et al.22 Subsequent studies have added more primer sets using similar protocols through BLAST homology search, amplification efficiency checks, type strain tests, and case studies.3,9,12,24 Primer sets targeting 213 ARGs (in some cases, multiple primers were used to target one gene in different bacterial groups,22 resulting in 285 pairs of individual primers), nine marker genes for mobile genetic elements (MGEs) including eight transposase genes and one universal class I integronintegrase gene (intI-1),25 and one 16S rRNA gene were included in the present study (Table S3). The gene cintI-1 for clinical class 1 integron-integrase was also quantified. The cintI1 data were used for correlation analysis with ARGs, but not included in the calculations of the total abundance of MGE marker genes due to the possible overlapped amplifying range with intI-1.26 The process of PCR was conducted on Wafergen Smart Chip with 5,184 nanowells containing 100-nL PCR mixtures in each well, employing PCR conditions described previously.8 DNA concentrations from all samples were adjusted to around 50 ng μL−1, and the final DNA concentration in the reaction well was around 5 ng μL−1. The theoretical detection limit is one copy per reaction well, which equates to two copies ng−1 DNA based on the amount of DNA added to each well. Technical triplicates for each sample and one reagent control were included on one chip for quality control. To minimize the possible bias caused by different amplifying lengths of the primers, amplifying efficiency of 80%−120% was set for data validation; the majority (66%) of all amplifications were within the efficiency range of 95%−105%. Because no standard curve was available for the calculation of the actual detection limit, Ct cut value of 31 was used as another validation criterion as in the previous reports employing the same methods.8,12,24 Biological replicates (3 for each treatment) were randomly tested on separate chips and amplifications in less than 3 replicates were taken as negative and discarded. The absolute abundances of ARGs were obtained in three steps: first, copy number of each gene on the chip was calculated by the follow equation: gene copy number = 10 (31−Ct)/(10/3); second, divide the ARG copy number on the chip by that of the 16S rRNA gene; third, multiply the value from the second step by the absolute copy numbers of 16S rRNA gene.12 The absolute copy numbers of 16S rRNA genes were quantified as described previously.9 For soil samples, the Ct values from the 3 biological replicates were used for calculating fold changes of genes (FC values) compared with the control soil.12 ARG profiles were grouped into eight defined classes with three major resistance mechanisms according
among others, demonstrated significant disturbance of the soil resistome due to human activities. However, these pot or microcosm-based studies are short-term in nature and deal with only a small subset of ARG members. To gain a better understanding of the dynamics of soil resistome, it is necessary to investigate the fates of a broad range of ARGs under field conditions. The advent of high throughput quantitative PCR (HT-qPCR) targeting an array of hundreds of ARGs has greatly improved our understanding of the resistomes in multiple environmental matrices, such as manure or compost, manure-amended soils, and soils irrigated with reclaimed wastewater.3,8,22 Therefore, this technology should enable more comprehensive insights into the dissemination of ARGs from urban to agricultural environments. In the present study, HT-qPCR was employed to assess the ARG profile in soil amended with five types of sewage sludges with divergent ARG compositions and abundances in a field experiment. Sludges were applied either in 4 large doses during 1994−1997, or annually in small doses for 21 years up to 2014. By profiling the ARGs in the sludges and soils, we aim to (1) track the fates of different ARGs in the sludges following the applications; and (2) reveal the long-term impact of sludge amendment on the soil resistome under field conditions.
■
MATERIALS AND METHODS Field Experiment and Soil Sampling. The field experiment was conducted at Woburn, Bedfordshire, U.K. The soil is a sandy loam, classified as Arenosol according to the FAO soil taxonomy system and Typic Udipsamment according to the U.S. soil taxonomy system. Basic information on the soil has been described in a previous study23 and selected soil properties are presented in Supporting Information Table S1. The experiment began in 1994 with the aim of investigating the long-term effect of heavy metals contained in sewage sludges on soil fertility and microbial activity.23 Five types of sewage sludge cakes (coded as S1−S5) produced in different wastewater treatment plants with different characteristics (Table S1) were applied in two different ways, as past applications (P_S1− P_S5) in four large doses between 1994 and 1997 or as annual applications (A_S1−A_S5) between 1994 and 2014 of approximately 1/25th of the amounts of the same sludges to realize the same total dose as in the past applications in 25 years (Table S1). The sludges were evenly incorporated into the topsoil (0−25 cm) by using a Celli spading machine.23 After 1997, these sludges were air-dried and stored for use during the annual application period. The input of total organic carbon into soils from the five sludges was comparable in each application mode (Table S1). A control receiving no sludge was included in the experiment. There were 11 treatments in total used in the present study. Each treatment was replicated in three plots (6 × 8 m2) in a randomized block design. The plots were cultivated to 23 cm depth and wheat was grown annually during the whole period and received N, P, K fertilizers according to requirements assessed by soil analyses over time. Soils were collected from the plowing depth by auger in February 2015. Six cores were collected from each plot and bulked. Samples were stored at −80 °C until further use. Soil properties were determined as shown in Table S2. DNA Extraction and Purification. Portions of frozen soil samples were freeze-dried, homogenized by vigorous shaking in 15 mL plastic vials and stored at −80 °C. Approximately 0.5 g (dry weight, DW) air-dried sludge or freeze-dried soil was extracted with a PowerSoil DNA Isolation Kit (MoBio Laboratories, 12603
DOI: 10.1021/acs.est.6b02138 Environ. Sci. Technol. 2016, 50, 12602−12611
Article
Environmental Science & Technology to a newly published study (Table S3), in which some genes originally in the classes of “FCA” (fluoroquinolone, quinolone, florfenicol, chloramphenicol, and amphenicol resistance genes) and “Other/Efflux” in previous studies have been newly grouped in the class of “multidrug”.3,8,9,12 All qPCR methods currently used have some limitations.27 In the case of the HT-qPCR array method, the main limitations are the uniform PCR conditions for all primers and relatively high detection limits.27 These limitations become less critical for the purpose of relative comparisons between different sludge treatments. Statistical Analysis. The means and standard deviations (SD) were calculated by Microsoft Office Excel 2010. Multiple comparisons were based on one-way ANOVA in SPSS 16.0. Least significant difference (LSD, P < 0.05) test was used for comparisons between treatment means. In the case of comparisons of gene abundances between treatments, log transformed data were employed. Structure (principle component analysis, PCA) and diversity (index of Inverse Simpson) of the ARG profiles were analyzed with log transformed data in R (version 3.2.0)28 with the “Vegan” (2.3−0) package. Heatmaps of ARG profiles with gene abundances normalized to the corresponding absolute abundance of 16S rRNA genes and log transformed were constructed in R using the package “pheatmap”. All the correlation analyses between gene levels were based on log transformed data of absolute abundances. Venny (2.0.1)29 was employed to group gene members belonging to different sections with different fates from sludges to soils.
Figure 1. Absolute abundances (copies g−1 dry soil) of ARGs and MGE marker genes in the sludges, sludge-amended soils, and the control soil. Asterisks indicate significant differences (P < 0.05) between the two modes of sludge applications. Different lower- and upper-case letters above the bars indicate significant differences (P < 0.05) among the past application treatments and among the annual application treatments, respectively. Statistical analyses were based on log10-transformed data.
■
RESULTS Abundances of ARGs and MGE Marker Genes. The Abundances of ARG and MGE marker genes in the sludges (copies g−1 dry sludge) were generally 1−2 orders of magnitude higher than those in the soils (copies g−1 dry soil), except for S4 and S5, in which MGE marker genes and ARGs, respectively, were less abundant than those in the soils amended with these sludges (Figure 1). Compared with the control soil receiving no sludge, both past and annual applications of sludge generally resulted in increased abundances of ARGs and MGE marker genes (Figure 1). In general, annual applications produced a greater effect than past applications, although not all comparisons were statistically significant. Among the five sludges, annual applications of S1 (A_S1) exhibited the highest effect and boosted ARGs and MGE marker genes by 3.6 and 7.6 fold in abundance, respectively, compared with the control. The abundances of ARG and MGE marker genes in soils amended with S5 were the lowest among all sludge treatments in the two modes of application (Figure 1). Abundances of ARGs in sludgeamended soils showed significant correlation to total sludge inputs (copies ha−1) of ARGs (Person, r = 0.80, P = 0.005). However, level of MGE marker genes in sludge-amended soils showed no significant (P = 0.055) correlation with their total inputs by sludge application. No significant correlation between ARG abundance and the content of soil organic carbon or soil microbial biomass was observed across all soil samples (data not shown). Diversity of ARGs. ARGs in S1, S2, and S3 were more diverse in terms of α-diversity with higher numbers detected (58−91 genes) than those in the corresponding soils (41−69 genes), but the opposite was observed in treatments with S4 and S5 (30 and 24 ARGs for S4 and S5, respectively, compared with 34−48 genes in soils amended with these two sludges) (Figures S1a and S2). Annual sludge applications with the
5 different sludges all significantly (P < 0.01) enhanced the soil ARG α-diversity, showing more significant (P < 0.01) boosting effect than past applications, in which only soils amended with ARG-diverse sludges (P_S1, P_S2, and P_S3) showed a higher diversity than the control (Figure S2). Principal component analysis (PCA) showed that sludges and soils exhibited significantly different structures of ARG compositions (Adonis test, P < 0.01; Figure S3) along PC1, which explains 39.8% of the variance. Dissimilarity among the 5 sludges was more obvious than that among soils on PC1. The close distribution of soil samples along PC1 was probably due to the robust influence of background soil resistome. Differences along PC2 (explaining 18.3% of the variance) were observed among sludges as well as between control soil and sludge-amended soils (Figure S3), indicating responses of soil ARG structure to sludge amendments. The ARGs in sludges and soils confer resistance to all the eight defined classes of antibiotics linked to 3 major resistance mechanisms, and there are clear differences between sludges and soils (Figure 2). Multidrug, tetracycline, aminoglycoside, and MLSB resistance genes were most abundant in sludges S1, S2, S3, and S5, while aminoglycoside accounted for 91.1% in S4 due to the high detection of aphA1(aka kanR). The proportions of aminoglycoside resistance and tetracycline resistance genes decreased substantially from the sludges to the corresponding sludge-amended soils (Figure 2a). Soils amended with different sludges from the two modes of application all showed relatively similar patterns of ARG classes and mechanisms, with multidrug-resistance (MDR) genes that operate via antibiotic efflux dominating (82.7%) in all soils (Figure 2b). Significant differences between the two modes of sludge application were mainly observed for soils amended with S1, where A_S1 harbored more genes related to the resistance of 12604
DOI: 10.1021/acs.est.6b02138 Environ. Sci. Technol. 2016, 50, 12602−12611
Article
Environmental Science & Technology
applications, especially in A_S1. In total, there are 31 ARGs in subgroups A and B shared by all soils including the control; these can be considered to represent the intrinsic resistance genes in the soil. These genes confer resistance to multidrug, aminoglycoside, β-lactam, MLSB, tetracycline, and vancomycin, within which MDR genes are dominant (Figure 3d). Genes in subgroups C are more abundant in the soils receiving annual applications of sludges (especially A_S1) than those receiving past applications, in terms of number or abundance. Genes in subgroup D appear to be randomly distributed or treatmentspecific. Fate of ARGs and MGE Marker Genes from Sludge to Soil. Among the 105 ARGs and eight MGE marker genes detected in the sludges, only 10 ARGs and 3 MGE marker genes were common to all five different sludges, most of which became undetectable following the applications to the soils (Figure S4). To aid the interpretation of the fate of ARGs from sludges to soil, we used Venn diagram to group the genes into seven sections (Figure 4) and calculated the number of genes, their abundance percentages and the enrichment (Fold change) compared with the control soil. Soils treated with either past or annual applications of sludges were both considered as sludgeamended soils (Table 1). Genes in Section 1 include the unique ARGs in the sludges which became undetectable following the application. These easily attenuated genes encompassed all the eight defined resistance classes, with aminoglycoside/tetracycline resistance genes being predominant (Table S4). A considerable portion (29.3−97.4%) of the ARGs in the sludges are in this section, especially for S4 (97.4% in abundance). A significant negative linear relationship (R2 = 0.94, P < 0.01) was observed between the abundance of MGE marker genes and the proportion of unique ARGs in the sludges (Figure 5a), suggesting that the attenuation of sludge ARGs following applications, especially with respect to diversity loss, decreased with the increasing MGE abundance in sludges. Genes in Section 2 (Table 1; Table S5) likely represent those introduced directly from sludges to soils, because they are common only in sludges and soils amended with sludges. These are the sludge-specific genes that survived in soils. The number of genes in this category is small (
