Bacterial community structures and biodegradation kinetic of Tiamulin antibiotic degrading enriched consortia from swine wastewater

Abstract

The antibiotic tiamulin (TIA) is common and widely used medication for dysentery eradication in swine productions. Tiamulin persists in livestock manure, and its residues have been found in various environment. This work obtained four tiamulin-degrading enriched bacterial consortia from a covered anaerobic lagoon system and a stabilized pond system of swine farms. Tiamulin was efficiently removed by the enriched cultures at the concentrations between 2.5 and 200 mg/L, with a removal of 60.1–99.9% during 16 h and a degradation half-life of 4.5–15.7 h. The stabilized pond system cultured with taimulin solely could eliminate tiamulin at the highest rates. The logistic substrate degradation model fit most of the experimental data. Next-generation amplicon sequencing was conducted, and it was found that the bacterial community was significantly impacted by the inoculum source, nutrient addition, and high tiamulin concentrations. Principal coordinate analysis (PCoA) indicated the similarity of bacterial communities in the original enriched samples and the 2.5 mg/L tiamulin-removed cultures. The 200 mg/L consortia were rather different and became similar to the other 200 mg/L consortia from different sources and cultures without nutrient supplementation. Shannon and Simpson indices suggested a reduction in bacterial diversity at high concentrations. The microbes that had high growth in the most efficient enriched culture, or which were abundant in all samples, or which increased with higher tiamulin concentrations were likely to be the major tiamulin-degrading bacteria. This is the first report suggested the possible roles of Achromobacter, Delftia, Flavobacterium, Pseudomonas, and Stenotrophomonas in tiamulin degradation.

Electronic supplementary material

The online version of this article (10.1007/s40201-019-00426-2) contains supplementary material, which is available to authorized users.

Introduction

Veterinary antibiotics are essentially used in swine production to prevent and treat disease, to improve the efficiency of feeding, and to promote growth. Based on data from different countries, the antibiotics were consumed approximately 27–730 mg/Kg meat/year [1]. Among the veterinary antibiotics, tiamulin, a derivative of the diterpene antibiotic of the pleuromutilin class (previously the macrolide class), is often used in swine diets to solve problems of dysentery and other bacterial infections. In the US and UK, 1435 kg tiamulin were sold in 2000 [2]. Investigation of 55 pig farms in Vietnam showed that the use of tiamulin accounted for 5.74% of antibiotics used [3]. Tiamulin is also one of the most frequently used antibiotics throughout the life of swine in Thailand.

Tiamulin has been discovered in various environments. Previous studies reported the detection of tiamulin in swine manure in Germany at a concentration of 43 μg/kg [4], in swine wastewater in China at 2–5 μg/L [5, 6], and in the US at 133–182 ng/L [7]. Furthermore, tiamulin was observed in Chinese river samples at concentrations ranging from 0.2–5.8 ng/L [8–10]. Tiamulin was also discovered in shallow groundwater below a cattle wastewater lagoon in the US at a concentration of 29 ng/L [7]. The presence of tiamulin in activated sludge could constrain microorganisms, especially nitrifiers and anaerobes [11].

Some studies observed the elimination of tiamulin, but the removal efficiency was not sufficient. A previous study [12] indicated that the half-life of tiamulin in aerobic conditions with an initial concentration of 2000 μg/kg in agricultural soil was 16 days. The degradation continued until day 62 and became stable approximately 50 μg/kg without further degradation after 180 days. However, tiamulin remained unchanged in anaerobic manure storage under anaerobic conditions for 180 days [13]. Another study reported the ability of particular wood-rot fungal strains to eliminate 10 mg/L tiamulin from 65 to 95% over a period of 12 days by manganese peroxidase (MnP) [14]. Using the aerobic biodegradability test with the activated sludge organisms, ISO 9408:1999, no degradation started until the 20th day and reached only 40 ± 2% in the 43 days of the test, after which it stopped [11]. Abiotic degradation has not been detected (≤ 2%). Due to environmental persistence, the biodegradation of tiamulin should be improved. Key microorganisms contributing to tiamulin degradation should be identified.

This study aimed to investigate the potential role of swine farm bacteria in degrading tiamulin aerobically and to identify the consortium by using an enrichment technique and next-generation sequencing. The bacterial communities were compared between the culture conditions at different tiamulin concentrations. A better understanding of the microbial community can contribute to further improve on the removal of residual antibiotics in swine waste and the environment.

Materials and methods

Enrichment of tiamulin-degrading bacteria

Swine wastewater samples were collected from various locations at two swine farms, including a covered anaerobic lagoon and a stabilization pond systems. The samples from the same system were combined in equal proportion and added as an inoculum for enrichment to a designated flask at 10% (v/v). The mineral salt medium (MSM) composed of 3.5 g/L K₂HPO₄, 0.5 g/L NaCl, 1.5 g/L KH₂PO₄, 0.15 g/L MgSO₄.7H₂O, and 0.5 g/L (NH₄)₂SO₄ was used. This study was conducted initially in two conditions for the two water samples; i.e., (1) 4 mg/L tiamulin fumarate (Vetranal Sigma, US) as the sole carbon source and (2) 4 mg/L tiamulin fumarate and 2 mg/L nutrient broth (Sigma-Aldrich, US) as carbon sources. All enriching flasks were incubated on a shaker at 200 rpm, at room temperature, and in dark conditions.

Because tiamulin could inactivate bacteria, introduction of tiamulin in the enrichment systems needed to perform by gradually increasing the tiamulin concentration. In every subculturing event, tiamulin concentration was increased by 1–2 mg/L until it reached 20 mg/L. For the condition (2), NB concentration was decreased at every event by 0.5 mg/L to become NB-free medium in the flasks of fifth subculture and later. Tiamulin concentration was maintained at 20 mg/L without nutrient broth until 7 months period, resulting in total 14 subcultures.

The subculturing event was conducted after a 14-d interval for four times. Ten percent (v/v) of each enrichment culture was transferred to a new flasks containing fresh MSM supplemented with the carbon sources for each condition. After the fifth subculture, transfer amount of the cell suspension was decreased to 1% (v/v) when the cell numbers reached 10⁵ CFU/mL. The bacteria were enumerated on plate counting agar (HiMedia Laboratories, India) in triplicate after incubating at room temperature for two days.

Tiamulin degradation at different concentrations

The tiamulin degradation kinetics of the four enriched cultures were carried out in a batch process at different initial tiamulin concentrations of 2.5, 5, 10, 20, 30, 50, 70, 100, 150, and 200 mg/L in MSM. The experiments were performed in triplicate. The incubation was conducted on a shaker at a rate of 200 rpm at room temperature (25–30 °C) and a neutral pH (7.0–7.5) in the dark. Abiotic control (sterile MSM medium with tiamulin) was performed in parallel with the tiamulin degradation experiment at the same concentration. Liquid samples in culturing flasks were harvested every 2 h for 16 h. Tiamulin residue in the collected samples was determined using HPLC analysis. The initial bacterial cell concentration was approximately 1 × 10⁸ cell/mL and the average cell growth yield was 3.14 × 10⁶ cell/ng tiamulin. The first-order kinetic model, the logistic model, and the logarithmic model, together with the corresponding kinetic rate constant (k) and degradation half-life (D50, hour), were defined by eqs. 1–9 following the procedures of Simkins and Alexander (1984) [15]. Model parameter k was estimated by matching the model with the experimental data using the least squares and GRG (Generalized Reduced Gradient) nonlinear methods implemented in Solver software in MS Excel. The best-fitted model was the one which obtained the highest root square value (R²) was chosen for each culturing condition.

$$\text{First order}:S=S_{0}exp\left(-k_{1}t\right)$$ 1

$$k_1=\frac{{\mu }_{\mathit{max}}{X}_0}{K_s}$$ 2

$$D50=\frac{\mathit{ln}2}{k1}$$ 3

$$\text{Logistic}:S=\frac{S_0+X_0}{1+\left(\frac{X_0}{S_0}\right)\mathit{exp}\left[k_2\left(S_0+X_0\right)t\right]}$$ 4

$$k_2=\frac{{\mu }_{\mathit{max}}}{K_s}$$ 5

$$D50=\frac{ln\left(\frac{S_0}{X_0}+2\right)}{k_2\left(S_0+X_0\right)}$$ 6

$$\text{Logarithmic}:S=S_0+X_0\left[1-exp\left(k_{3}t\right)\right]$$ 7

$$k_3={\mu }_{\mathit{max}}$$ 8

$$D50=\frac{1}{k_3}ln\left(1+\frac{\frac{S_0}{2}}{X_0}\right)$$ 9

where S is the tiamulin concentration at any time (mg/L), S₀ is the initial tiamulin concentration (mg/L), t is incubating period (h), X₀ is the amount of substrate required to produce a population density equal to the initial cell concentration (mg/L), μ_max is the maximum specific growth rate (h⁻¹), and K_s is the half-saturation constant equal to the substrate concentration at which the specific growth rate (μ) reaches half of the μ_max value. The units of k₁, k₂, and k₃ were h⁻¹, L/mg·h, and h⁻¹, respectively.

Tiamulin analysis was performed using an Agilent 1100 series HPLC system with diode array detectors (Agilent Technologies, USA) by the method following Cancho-Grande et al. (2001) [16] with some modifications as shown in Supplementary material.

16S rRNA gene amplicon sequencing and data analysis.

Bacterial 16S rRNA gene libraries of original enriched cultures and the degrading tests at 2.5 and 200 mg/L were constructed using primers 515F and 806R [17]. The PCR reaction mixture was performed using KAPA HiFi HotStart ReadyMix (KAPA, US) with the following program: 95 °C for 5 min; 28 cycles of 95 °C for 30 s, 55 °C for 45 s, and 68 °C for 30 s; and 68 °C for 7 min. The amplicons were purified by using AMPure XP beads (Agencourt Bioscience, US). The libraries were constructed, indexes were ligated (Nextera XT Index Kit, Illumina Inc., US), and libraries were quantified using a QFX Fluorometer (DeNovix Inc., US) and then normalized and pooled. Amplicon sequencing was performed on an Illumina MiSeq sequencer at Omics Sciences and Bioinformatics Center (BKK, Thailand) using paired-end (2 × 150 bp) sequencing.

The raw data were sorted based on the barcode sequence, after which the barcodes and primers were removed. Sequencing quality filters were applied by using FASTX-Toolkit. Reads with nucleotide-quality scores <30 were discarded, and short sequences (< 200 bp) were excluded. High-quality reads were then clustered into operational taxonomic units (OTUs) at 97% similarity using QIIME 1.9.0 [17] and chimeras were excluded using UCHIME [18]. The metagenomic sequences have been deposited in the Sequence Read Archive (SRA) under the accession number SRP132663, and the sample information was deposited with the accession number PRJNA431340. The alpha diversity was assessed by computing the following indices: sequence number, OTU richness, Chao1 index [19], Shannon index [20], and Simpson index [21]. The beta diversity was further calculated with the weighted UniFrac algorithm via Principal Coordinates Analysis (PCoA) [22].

Results

Biodegradation of tiamulin by bacterial enrichment cultures

After seven months with fourteen subcultures, four enrichment cultures, including A, AN, S, and SN, were obtained. Cultures A and AN originated from a covered anaerobic lagoon system supplied with tiamulin only and with nutrient broth at the beginning, respectively. S and SN cultures were brought from a stabilization pond system that was further cultivated in tiamulin only or supplemented with nutrient broth. The abilities of the four bacterial enriched cultures to degrade tiamulin at various initial concentrations of 2.5–200 mg/L were examined.

The tiamulin concentrations in most treatments were significantly reduced (Fig. 1) with efficiencies of 60.1–99.9% within the 16-h experimental period. For concentrations above 50 mg/L, a longer incubation time was needed to observe complete degradation for the less effective cultures (A, AN, and SN) (Fig. 1–2). The culture S eliminated tiamulin at the highest rate at all concentrations. The degradation half-life (D50 or t_1/2) values were from 4.5–10 h, while the values for other cultures were above 8 h (Fig. 2 and Supplementary Table A1). Among the three enriched cultures, the degradation by culture AN was slightly better at high concentrations than the degradations by cultures A and SN, which were very similar to each other.

Fig. 1.

Tiamulin degradation by bacterial enriched cultures in different initial concentrations. Line is the model fit curve to the experimental data (Table A1). The error bar is the standard error

Fig. 2.

Tiamulin degradation half-life (D50) at different initial tiamulin concentrations

Almost all degrading data were best described using the logistic model, except for the three lowest concentrations of culture S that were more similar to the first-order kinetics and the three highest concentrations of culture SN that matched well with logarithmic kinetics (Table A1 and Fig. 1). All the selected models shown were well matched to the experimental data with root square values above 0.94. Logistic and logarithmic kinetics have been used to describe the degradation rates accompanied by cell growth, which often contain a lag period before starting to degrade [15]. This study also observed cell growth up to log 2–3 orders (the data are not shown). Based on the results, the D50 values and the lag period increased with higher tiamulin concentrations, suggesting more stringent conditions for the bacterial cultures to degrade the compound. For all abiotic sterile controls, the remaining concentrations were quite constant at the initial concentrations (Fig. 1). Biodegradation is the only major mechanism for tiamulin elimination.

The enriched samples supplemented with nutrient broth (AN and SN) degraded low concentrations of tiamulin at slower rates (high D50) than did the samples enriched with only tiamulin (A and S) (Fig. 2); nonetheless, the AN culture was better than the A culture at the concentrations above 20 mg/L. The enrichment without other carbon sources at low tiamulin concentrations promoted the growth of effective tiamulin-degrading bacteria better than the enrichment with other carbon sources, especially in the case of the culture S. However, other carbon sources used to enrich the culture AN promoted efficient growth of tiamulin-degraders at high concentrations, as the rate was slightly above that of culture A. Different degrading behaviors suggest different key players in complex bacterial communities. It was interesting to investigate the major microorganisms contributing to the degradation of tiamulin under different conditions.

Bacterial community analysis by 16S rRNA amplicon sequencing

16S rRNA gene amplicon sequencing was conducted to investigate the bacterial community in the original enriched cultures and after the degradation tests at 2.5 and 200 mg/L. The reason of choosing these concentrations because the polymerase chain reaction – denatured gradient gel electrophoresis (PCR-DGGE) has been conducted before and found the high similarity between the tiamulin degrading consortia at concentrations 2.5 and 50 mg/L (Supplementary material Fig. A1). Sequencing results revealed that Proteobacteria dominated all of the samples (Supplementary Fig. A2) as they contributed up to 87.3–99.8% in all samples. Previous studies also indicate that Proteobacteria was dominant in activated sludge and increased in abundance under other antibiotic treatments [10, 23, 24]. Among these, Betaproteobacteria occupied a large proportion of cases (49.7–93.2%), except for A samples where Gammaproteobacteria were quite comparable (40.6–54%) and SN and SN 2.5 samples in which Alphaproteobacteria levels were high (45.3–62.1%). The samples from the stabilization ponds mostly had higher Alphaproteobacteria (5.3–62.1%) than the samples from the anaerobic digestion system (1.4–7.9%). Following Proteobacteria, the Bacteroidetes phylum was highly abundant, as it was 0.1–12.1% in all samples. Cytophagia were only observed in SN samples, while Flavobacteria were only observed in AN, S, and SN200 samples. For Flavobacteria, the density was rather increased at high concentrations, except in the A samples (Supplementary Table A2). Because S and AN cultures mostly degraded tiamulin at faster rates than the others (Fig. 2), some bacteria in the genus Flavobacteria may have a high ability to degrade tiamulin. In addition, phylum Actinobacteria was observed in only 0.1% of the S2.5 sample. These three phyla were also observed to have similar proportions in a mixed bacterial consortium derived from rhizosediments of estuarine plants capable of degrading enrofloxacin [25]. Our study indicated the major role of Proteobacteria in degrading tiamulin and possibly some Flavobacteria.

Figure 3 and Table A2 demonstrate the relative abundance of taxa at the rank genus level for 16S rRNA gene sequencing reads. Six major genera consisting of Achromobacter, Delftia, Pseudomonas, Pandoraea, Stenotrophomonas, and Hyphomicrobium were found to have large numbers more than 5% of the total read sequences. Only Achromobacter, Pseudomonas, and Stenotrophomonas were present in all samples. This indicated that they were likely able to degrade tiamulin. Achromobacter density increased with tiamulin concentration increased, whereas the density of Stenotrophomonas gradually decreased. Some microbes were specific to sources such as Kaistia, Sphingomonas, Delftia, and Hylemonella,that were found only in stabilization pond samples. Different swine waste sources resulted in variation in microbes in enrichment cultures.

Fig. 3.

Relative abundance of taxa at rank genus for 16S rRNA gene sequencing reads more than 1.0% of total reads from bacterial enriched cultures (A, AN, S, SN) and after degradation tests at tiamulin concentrations of 2.5 and 200 mg/L

In addition, the bacterial community in enriched cultures that were first supplied with the nutrient broth and tiamulin were different from the one cultured in tiamulin only. However, the community at 200 mg/L became similar to the cultures without nutrient addition. Pandoraea, Chryseobacterium, and an unidentified genus in Methylophilaceae occupied a high proportion in AN and AN 2.5 samples, while the proportions decreased in the AN 200 sample and became similar to those in sample A with high Achromobacter. In the case of the stabilization pond, SN and SN 2.5 had high proportions of Hyphomicrobium, an unknown genus in Rhizobiaceae, and Leadbetterella. Again, those proportions decreased and had high Delftia and Flavobacterium instead, similar to all S samples. These results suggested that there was an effect of supplementing nutrient broth at the beginning of enrichment on bacterial communities; however, when the cultures were exposed to high concentrations, some efficient tiamulin-degrading microorganisms might outgrow other organisms and become the dominant species.

The comparison of the bacterial communities by principal coordinate analysis (PCoA) using the weighted UniFrac distance (Fig. 4) explained 88.55% of the variation in the data. PC 1 extracted almost ¾ (63.34%) of the variation in the entire data set, and PC 2 and PC 3 extracted similar variations in the data, which were 14.87% and 10.34%, respectively. The bacterial consortia of the samples originated from the same enrichment culture, regardless of the different tiamulin concentrations, were located in close proximity. Except for the SN 200 sample, which was rather distant; however, all of the SN samples were similar in terms of a high PC2 factor. This indicates the similarity of microbial diversity. The bacterial communities of 200 mg/L in the AN, S, and SN enriched cultures were also distant to the other enriched samples from the same source. They were closer to the other 200 mg/L cultures. The microorganisms in the 200 mg/L culture were rather different from the original and 2.5 mg/L samples. Supplementary Table A3 shows the alpha diversity index, including Shannon and Simpson analyses, and supports the decrease in diversity for the samples with 200 mg/L culturing conditions and originating from the stabilized pond system. The effect was not pronounced in the samples from the anaerobic lagoon system. There is an effect of high tiamulin concentration (200 mg/L) on the bacterial community. Therefore, both sample sources and substrate concentrations affected on tiamulin degrading consortia.

Fig. 4.

Weighted UniFrac principal coordinates analysis (PCoA) of bacterial community structures in original enriched cultures (A, AN, S, and SN) and after degradation tests at tiamulin concentrations of 2.5 and 200 mg/L

Because we enriched these cultures for more than seven months, many microbes remaining in our systems should be able to utilize tiamulin as a main carbon source. The microbes that notably had high proportions in only the S sample, which had the highest biodegradation rates, were likely to be effective tiamulin degrader. This suggested important roles for Flavobacterium and Delftia. In addition, the microbes that were always present in all samples and increased significantly at high concentrations (Acromobacter and Delftia) were likely to be effective tiamulin degraders. Moreover, the microbes that were present in all samples at high proportions (Achromobacter, Pseudomonas, and Stenotrophomonas) also had a strong possibility of being the main tiamulin degraders. Our analysis suggests that Flavobacterium, Delftia, Achromobacter, Pseudomonas, and Stenotrophomonas were most likely to be effective tiamulin-degrading bacteria in swine wastewater. However, the tiamulin-degrading bacteria are not only limited to these strains. Other microbes identified in this study may also be able to degrade tiamulin. Further degradation studies or enzyme and genome analyses on bacterial isolates should be conducted to investigate the abilities of our suggested genera to degrade tiamulin.

Discussion

The current study obtained effective tiamulin-degrading enriched. The results show that tiamulin was almost completely degraded by 16 h under aerobic conditions. Tiamulin was able to be degraded by wastewater microorganisms, and this was in agreement with other studies which revealed the aerobic biodegradability of activated sludge by 40% with an initial tiamulin concentration of 100 mg/L over 43 days [11]. A long lag phase was observed as degradation started after the 20th day [11]. However, in our study a much faster degradation period of less than a day with a lag period of a few hours was experienced. This could be because enriched cultures that acclimated tiamulin approximately 7 months were applied in this study, leading to a reduction in the lag period. Another study applied a 4-month-enriched activated sludge culture with only macrolide azithromycin (AZI) antibiotics to degrade AZI, clarithromycin, and erythromycin at 100 mg/L [26]. The results showed high removal efficiencies of >99% within 150 h. Therefore, cultures that are acclimated or enriched should be used to enhance antibiotic degradation instead of freshly introduced inoculum cultures.

The tiamulin-degrading characteristics at various concentrations of the enriched cultures originated from swine wastewater aligned well with the logistic kinetics. Only the three lowest concentrations of culture S and the three highest concentrations of culture SN fit with the first-order kinetics and logarithmic kinetics, respectively. Based on Simkins and Alexander (1984) [15], the necessary conditions for the logistic kinetics and logarithmic kinetics are (1) the substrate used for cell growth and (2) S₀ < < K_s or S₀ > > K_s. The necessary conditions for first-order kinetic growth are (1) substrate not available for cell growth, (2) X₀ > > S₀, and (3) S₀ < < K_s. In fact, cell growth was observed during the course of degradation. The substrate utilization kinetic models applied for cell growth should provide more realistic data. However, the reason for first-order kinetics in the three lowest concentrations of culture S might be because the kinetic conditions also limit very high cell densities (X₀) and very low substrate concentrations (S₀). The initial inoculum cell number of 10⁸ cells/mL might be too high for the enriched culture S growing with 2.5–10 mg/L of tiamulin. Most of the substrate would quickly disappear to saturate the uptake system of cells; hence, the nongrowth degradation kinetics were fitted as first-order for these cultures. On the other hand, for the case of the three highest concentrations of culture SN, the logarithmic kinetics matched well; this was also possible, as the kinetic conditions indicate the applicability with S₀ > > K_s. Because the tiamulin-degrading efficiency of enriched culture SN was the lowest compared to the others judging from the high D50 values in Fig. 2, the half-saturated constant (K_s) should be relatively low. The tiamulin substrate concentrations of 100–200 mg/L might be too high for this culture SN. For the rest of the degradation tests that fit the logistics model, the condition of S₀ < < K_s would have applied, suggesting that the tiamulin concentration was sufficient compared to the relatively high efficiency of the enriched cultures.

Although the logistics model was less common than the first-order rate kinetics model, it has been used in the literature on antibiotic biodegradation. For example, Terzic et al., (2018) [26] observed the degradation of azithromycin (AZI) in the macrolide group, which agreed with the logistics model, while erythromycin (ERY) and clarithromycin (CLA) were fitted with the first-order kinetics model. The authors explained that AZI adsorbed into biomass well, as it had high Kd value which was different from other tested compounds. This may lead to an under-estimation of the available substrate concentration for the cells; therefore, it fit with the condition S₀ < < K_s. Another study indicated the mineralization of 10 mg/L sulfadiazine by the Microbacterium lacus strain SDZm4 with the logistics model [27]. The authors described that this model accounts for the initial stage of exponential growth of the population and then a second stage of limited substrate availability. This again correlates with the condition of low substrate concentration compared to the cells’ degradation abilities (S₀ < < K_s). The first-order model was suggested by many previous reports on the reduction of antibiotics in activated sludge, such as cefalexin, sulfamethoxazole, sulfadiazine, norfloxacin, ofloxacin, and ciprofloxacin (100 μg/L each) [28] and the removal of sulfadiazine and sulfamethoxazole (4–20 mg/kg) in agricultural soil [29]. It is notable that the antibiotic concentrations used in the previous studies were quite low compared to this work with no lag period; this might lead to a different substrate degradation kinetics model than our results.

Based on the amplicon sequencing results, microorganisms in the genera Achromobacter, Flavobacterium, Delftia, Stenotrophomonas, and Pseudomonas were likely to be the major tiamulin-degrading bacteria in swine wastewater. Many previous studies reported the antibiotic-degrading function or the detection of these genera in an antibiotic-degrading environment, which could suggest the possible functions of these bacteria on tiamulin degradation. Achromobacter isolates were able to degrade tetracycline [30], sulfamethazine [31], and sulfamethoxazole [32]. For Flavobacteria, F. saccharophilum has been observed to degrade validamycin A [33]. It was also detected in estuarine sediment high in oxytetracycline [23], enrofloxacin [34], and sulfonamide [35], and in upflow biological aerated filter treating ciprofloxacin [36]. Some Delftia was isolated from norfloxacin and spiramycin-containing wastewater [37]. Delftia acidovorans was also found to be enriched in the internal loop photobiodegradation reactor treating sulfamethoxazole [38]. In addition, the strain KV29 (capable of eliminating crystal violet) was found to contain the plasmid pKV29 that has typical IncP-1β backbone gene modules similar to those conferring antibiotic resistance [39]. Stenotrophomonas maltophilia is an important multidrug resistant that has emerged as an important nosocomial pathogen with high mortality rate. It resisted to various antibiotics including trimethoprim-suphamethoxazole, fluoroquinolones, β-lactams, tetracycline derivatives, aminoglycosides, and polymyxins [40]. Stenotrophomonas maltophilia DT1 was also found to transform tetracycline [41]. Stenotrophomonas acidaminiphila and Stenotrophomonas nitritireducens degraded amoxicillin [42]. Many strains of Pseudomonas putida was able to use triclosan [43], sulfamethoxazole [44], and ciprofloxacin [45]. Pseudomonas aeruginosa was able to use erythromycin [46] and sulfamethoxazole [44]. Pseudomonas psychrophila HA-4 used sulfamethoxazole [47]. Pseudomonas fluorescens MC46 degraded triclocarban [48]. In addition, Pseudomonas aeruginosa displays various antibiotic resistance such as carbapenem and fluoroquinolone and it also known to acquire actively genetic mutations for further resistance [49].

Nevertheless, biodegradation ability is not broadly distributed among all members in the genus, demonstrated in the case of Achromobacter denitrificans PR1 which was the only strain among the six tested Achromobacter strains capable of degrading sulfamethoxazole [32]. Thus, it cannot clearly be concluded based on the literature studies. Further bacterial isolation, enzyme detection, or genomic studies are needed to clarify the roles of these microbes in tiamulin degradation. In addition, it is important to note that because this study identified these strains from the enrichment cultures, their proportion and real function in natural setting require further investigation.

Conclusion

Four enrichment cultures, including A (covered anaerobic lagoon source), AN (covered anaerobic lagoon source, nutrient broth), S (stabilization pond source), and SN (stabilization pond source, nutrient broth), were obtained after a long acclimation period in aerobic conditions. They were able to remove tiamulin at concentrations of 2.5–200 mg/L with efficiencies of 60.1–99.9% over 16 h. The degradation half-life values were between 4.5–15.7 h. The culture S eliminated tiamulin at the highest rates in all concentrations. The logistic substrate degradation model best described most of the experimental results. Bacterial community structures were significantly different based on the source of the inoculum, nutrient addition at the beginning of the enrichment process, and high tiamulin concentrations. Proteobacteria dominated in all samples and accounted for 87.3–99.8%. Principal coordinate analysis (PCoA) indicated that the bacterial communities in 2.5 mg/L degradation tests were not much different from the original enriched cultures; however, the consortia at 200 mg/L were obviously different and were located closer to the other 200 mg/L consortia and the cultures without nutrient supplementation. The analysis by Shannon and Simpson suggested a reduction in bacterial diversity at high concentrations. The microbes that had high proportions in enriched culture S, or were present in all samples which increased in number under high tiamulin concentrations, or had high proportions in all samples, were likely to be the major tiamulin-degrading bacteria. The results suggested important roles of Achromobacter, Delftia, Flavobacterium, Pseudomonas, and Stenotrophomonas in tiamulin degradation.

Funding

This study was partially funded by Office of Higher Education Commission (OHEC) and the S&T Postgraduate Education, Research Development Office (PERDO) for the support of their research program in Hazardous Substance Management in the Agricultural Industry through the Center of Excellence on Hazardous Substance Management. One researcher was partially financial support from Chulalongkorn Graduate School Thesis Grant. Some instrument used in this study was acquired through the King Mongkut’s University of Technology Thonburi 55th Anniversary Commemorative Fund.

Compliance with ethical standards

Conflict of interest

The authors declare that they have no competing interests.