Accumulation of Sulfonamide Resistance Genes in Arable Soils Due to Repeated Application of Manure Containing Sulfadiazine

Abstract

Two soils were amended three times with pig manure. The abundance of sulfonamide resistance genes was determined by quantitative PCR 2 months after each application. In both soils treated with sulfadiazine-containing manure, the numbers of copies of sul1 and sul2 significantly increased compared to numbers after treatments with antibiotic-free manure or a control and accumulated with repeated applications.

The transmission of antibiotic resistance through agriculture may have a greater impact on human health than hospital transmission of resistance (25), as gene transfer and the environmental spread of antibiotic resistance genes might substantially contribute to antibiotic resistance in the human microbiome, which exacerbates the threat of infections with antibiotic-resistant bacteria (10). Antibiotic resistance genes on transferable plasmids are introduced via manure into agroecosystems and can persist regardless of the viability of the introduced host cell due to horizontal gene transfer (9). Pig manure often contains considerable amounts of administered antibiotics that pose a selective pressure on bacterial communities during storage of manure and after application to soil (5, 27). Sulfonamides are among the most widely used veterinary antibiotics in the European Union (19), of which especially sulfadiazine (SDZ) has a high usage and potential to enter the environment (6). Resistance to sulfonamides is mediated mainly by the genes sul1 and sul2, coding for dihydropteroate synthases which are insensitive to sulfonamides (24). The genes occur in a wide range of bacterial species, because they are often located on transposable elements of self-transferable or mobilizable broad-host-range plasmids (8, 16, 23). Feeding experiments performed with ¹⁴C-labeled SDZ showed that more than 96% of the SDZ administered to pigs was excreted within 10 days and did not decrease during storage of the manure over 6 months (13). A survey of 15 field-scale piggery manures showed that piggery manure represents a hot spot for antibiotic resistance genes as well as broad-host-range plasmids (2-4). In a soil microcosm experiment, significant effects of manure containing SDZ on the abundance of the sulfonamide resistance genes sul1 and sul2 were detected even 2 months after application, compared to the abundance in untreated soil or soil treated with manure devoid of antibiotics (14, 15). Also, the frequency of capturing mobile genetic elements conferring SDZ resistance was increased by manure containing SDZ. The effects of a single manure application on bacterial resistance levels decreased over time due to diminished selective pressure, and eventually the resistance level of untreated soil was restored. Farmers typically apply manure to their fields several times a year (22), which might lead to an accumulation of resistance genes and antibiotic compounds in soil. However, this has not been investigated so far (9). Only Knapp et al. showed increases in the numbers of resistance genes over recent decades in archived soil samples from agricultural sites but could not correlate them with manure application due to confounding factors (17). Therefore, it was the objective of this study to test whether repeated applications of manure containing SDZ results in an accumulation of sul genes in soil bacterial communities.

Manure from healthy mature pigs was applied to topsoil samples (Ap horizon) of two German agricultural soils that substantially differed physically and chemically (20). Soil M was a silt loam (Orthic Luvisol) which had no history of previous manure applications. Soil K was a loamy sand (Gleyic Cambisol) fertilized in previous years with manure. Four treatments were prepared in pots with 100 g of 2-mm-sieved soil. To each pot, 4 g of manure with or without SDZ was added, so that an initial concentration of 0, 10, or 100 mg kg⁻¹ soil was achieved for treatment S0, S10, or S100, respectively. The amount of manure roughly corresponded to 30 m³ ha⁻¹, and treatment S10 corresponded to a realistic maximal input of sulfonamides according to agricultural practice (13, 21). The soil moisture content was set to 55% of the soil's water-holding capacity. All treatments were set up individually in five independent replicates per soil and per sampling time. The loosely covered pots were incubated at 15°C in the dark. The weight loss of the microcosms was compensated for by adding water to the soil surface twice a week. Manure was applied at day 0, day 63, and day 133. The soil samples analyzed were taken 60 days after each manure application. Total community DNA was extracted using the FastPrep FP120 bead-beating system for cell lysis and the FastDNA spin kit for soil (Q-Biogene, Carlsbad, CA). The copy numbers of sul1, sul2, and 16S rRNA genes were quantified by 5′-nuclease assays using quantitative real-time PCR as previously described (14, 15). Copy numbers of sul genes were related to 16S rRNA gene copy numbers and log transformed (Fig. 1). Treatment effects were analyzed by analysis of variance (ANOVA) using the MIXED procedure for repeated measures of SAS package 9.2 (SAS Institute, Cary, NC).

FIG. 1.

Effects of antibiotic-free manure and manure containing sulfadiazine (SDZ¹⁰ or SDZ¹⁰⁰, 10 or 100 mg kg⁻¹ soil) on the abundances of the sulfonamide resistance genes sul1 and sul2 relative to ribosomal rrn gene abundances in two soils, as quantified by 5′-nuclease assays using quantitative real-time PCR. Manure was applied at days 0, 63, and 133. Samples were analyzed 2 months after each application (days 60, 123, and 193). Error bars indicate the standard deviations (n = 5). Values over arrows show the percentages of alterations due to the repeated applications of the respective manures within a treatment (±estimated errors of slopes), as revealed by linear-regression analysis.

Two months after each application of antibiotic-free manure, the levels of sul1 genes in soil M were significantly increased compared to levels in untreated soil by about 1 order of magnitude (P < 0.0001). The abundance of sul2 genes was below the detection limit in untreated soil M at all time points analyzed but became detectable after the addition of manure (Fig. 1). Probably due to the long history of manure fertilization, soil K responded less to manure, which had no effect on sul1 (P = 0.2) and only a slight effect on sul2 (P = 0.03) abundances. The treatments with SDZ-containing manure (S10, S100) resulted in significantly higher abundance levels of both sul genes than were obtained with the treatments with antibiotic-free manure (P < 0.0001), and this was observed in both soils.

Linear-regression analysis revealed no change over time or a trend toward a slight decrease in sul gene copy numbers for both untreated soils (Fig. 1). Only moderate increases in sul gene abundance were detected during the course of repeated applications of antibiotic-free manure, which were within the error range for sul1 in soil K and sul2 in soil M. In contrast, significant accumulations of both sul genes were observed for both soils in the course of repeated applications of SDZ-containing manure. The increases ranged between 40% and 107% of log units per manure application. Two months after the third application of SDZ-containing manure to soil M, the abundance of sul1 increased by more than 3 orders of magnitude in comparison to the abundance in untreated soil M. A maximum of −1 log of sul1 and sul2 per ribosomal gene copy was reached, meaning that roughly 10% of the soil bacteria became resistant to sulfonamides. The accumulation was less strong in soil K but still exceeded 1 order of magnitude. For soil M, but not for soil K, the accumulation was more pronounced after treatment S100 than after S10. This indicated that the bacterial community of soil M that had previously received only mineral fertilizer was less adapted to manure-SDZ treatment and thus responded more strongly.

The desorbable and hence potentially bioavailable fraction of SDZ (26) was sequentially extracted with 0.01 M CaCl₂ and methanol and quantified by high-performance liquid chromatography (12). As previously shown (13), the easily extractable fraction of SDZ rapidly declined after the manure was spread on soil but was still detectable 2 months after application (data not shown). The average concentration of SDZ in soil K increased after each manure application of treatments S10 and S100 and in soil M after treatment S10 (Fig. 2). The dissipation of SDZ in the soil matrix did not occur quickly enough to prevent the accumulation of effective concentrations in soil. Notably, adsorption and/or absorption of SDZ by microorganisms cannot be excluded as a putative mechanism of dissipation. In contrast to what occurred with the other samples, after the S100 treatment of soil M, the concentration of SDZ declined in the period after the third manure application. This might indicate the development of microbial SDZ degraders in this treatment. The contribution of reversibly sequestered residues to the bioavailable fraction is not yet clear, but they might well serve as a long-term reservoir for SDZ in soils treated with SDZ-containing manure (11).

FIG. 2.

Sulfadiazine concentrations in soil microcosms after treatment S10 (upper graph) and S100 (lower graph) during the 60-day periods after each of three manure applications. Black boxes represent the temporal averages for each period, which were calculated from interpolated concentrations with a time step of 1 day. Box whisker plots show the medians, quartiles, and ranges of measured concentrations.

The results obtained from the soil microcosm experiment supported the hypothesis that repeated applications of manure increase the abundance of sulfonamide resistance genes in soil. Obviously, this was not so much caused by the addition of sul genes originating from manure (S0 treatment) but rather by the selective pressure exerted by bioavailable SDZ in soil. Manure typically contains substantial numbers of resistant bacteria, with 10⁻³ to 10⁻² sul1 copies and 10⁻² to 10⁻¹ sul2 copies per 16S rRNA gene copy (1, 14, 15). These numbers were high enough to result in a temporary increase in sul genes in soil after manure application in earlier studies (14, 15). The substrate from manure allows for the growth of populations carrying sul genes under selective pressure of SDZ. Increased selection rates could be observed even at low antibiotic concentrations (13, 18), but without the addition of the substrate, the bacteriostatic SDZ could not affect soil bacterial communities (7, 28). Compared to the rate of vertical transmission of sul genes by the growth of resistant populations, the rate of horizontal transmission in soil is presumably very low. However, the horizontal transfer of resistance is an important factor in the dissemination of resistance, as bacteria from manure may not be well adapted to the soil environment (9, 12, 13). The horizontal transfer of sul genes may have initially resulted in intraspecific, resistant subpopulations in soil, if they were not already present, and subsequently sul genes may have become relatively more abundant by the selective effect of SDZ-containing manure. This hypothesis is supported by the significant increase in the number of resistance genes found in the soil treated with SDZ after the repeated applications, whereas the levels of sul genes in the soil treated with only manure were rather constant and independent of the number of applications of manure.

Overall, our data indicated that the accumulation of resistance genes in environmental bacterial communities and the consequences of the potential long-term persistence of active pharmaceutical residues in soil should be considered more in risk assessments, which previously not only have been impeded by a shortage of suitable data but also have rather ignored important mechanisms, like the cross-species spread of resistance by plasmid transfer. As humans are continuously exposed to bacteria in the environment, the accumulation of resistance genes in soil due to the spreading of manure is likely to contribute to the threat of antimicrobial resistance in the therapy of infectious diseases.