A filamentous fungal strain, KC0924g, exhibiting a high ENR degradation ability, was isolated from the rhizosphere sediment sample. Colonies of strain KC0924g on PDA were 47 to 58 mm in diameter after 7 days at 25°C, with entire edges. The aerial mycelium was thick, white to yellow and olivaceous black in reverse (Fig. 1a). The colony morphologies of strain KC0924g grown on four other different media (OA, CMA, MEA, and PCA) are shown in Fig. S1, and their characteristics are described in the figure legend. Strain KC0924g possessed thick-walled conidia and an acremonium-like synanamorph (Fig. 1bc). The above morphological characteristics of strain KC0924g were similar to those of the model strain Humicola pulvericola CBS 723.97T (Wang et al. 2019). A BLASTN search against the sequences on the National Center for Biotechnology Information (NCBI) website revealed that the ITS gene sequence of strain KC0924g shared 98.99% similarity to Humicola sp. JS-0112 and Chaetomium sp. TR160. Hence, the rDNA-ITS sequence alone cannot distinguish the taxonomic status of strain KC0924g. It was found that the 5.8S nrRNA gene region and the D1/D2 domains of the 28S nrDNA (ITS&LSU), the second largest subunit of the DNA-directed RNA polymerase II gene region (rpb2) and the partial beta-tubulin gene region (tub2) were used as DNA barcodes to differentiate Humicola species and related genera in Chaetomiaceae (Wang et al. 2016). Subsequently, the phylogenetic tree (Fig. 1d) built based on the ITS&LSU-rpb2-tub2 gene sequences of strain KC0924g and related strains revealed that strain KC0924g clustered with Humicola pulvericola CBS 723.97T (Wang et al. 2019). Combining the morphological characteristics with phylogenetic analysis, the ENR degrader was identified as a member of the genus Humicola and deposited at the China Center for Type Culture Collection (deposition number: CCTCC M 2020752).
Fungi of the genus Humicola are common and widespread and are known to have biotechnological and industrial potential (Ibrahim et al. 2021). In recent years, the genus Humicola has been found to be valuable for environmental remediation. For example, Humicola sp. 2 WS1 showed marked arsenic biomethylation potential. Through bioaugmentation, Humicola sp. 2 WS1 not only attenuated the stressful effects of arsenic on Bacopa monnieri L. in the rhizosphere environment but also promoted the growth of B. monnieri (Tripathi et al. 2020). In this study, Humicola sp. KC0924g was isolated from the rhizosphere sediment sample of the submerged macrophyte V. spiralis. Submerged macrophytes play an essential role in maintaining aquatic ecosystem stability and are widely used in the remediation of contaminated water and sediment (Dai et al. 2014). Further work is required to explore the effect of Humicola sp. KC0924g and V. spiralis interaction on sediment remediation. Admittedly, this is the first report that a Humicola species can degrade the antibiotic ENR.
Degradation of ENR by strain KC0924g
Temperature and pH are two critical environmental factors that impact xenobiotic degradation. This is due to the influence of temperature on the growth of microorganisms and the activity of certain catabolic enzymes, as well as the impact of the initial pH of the medium on the movement of compounds to microorganisms and the alteration in the solubility of contaminants. Therefore, the effects of different temperatures and the initial pH value of the medium on ENR degradation were studied separately. Among five temperatures ranging from 15°C to 40°C, the optimum degradation temperature was observed at 28°C with maximum degradation of 89.6%. When the temperature lowered to 15°C, the rate of degradation decreased to 69.4%. Subsequently, with an increase in the temperature to 40℃, the degradation efficiency surpassed 76.9%. Overall, Humicola sp. KC0924g exhibited relatively broad temperature adaptability for ENR degradation. The degradation rate of ENR affected by temperature was listed in the following order: 28℃>20℃≥35℃>40℃>15℃ (Fig. 2a). The optimum temperature for photosynthesis in temperate submerged plants is usually between 25 ~ 32°C (Barko et al. 1982, Santamaría and van Vierssen 1997, Pedersen et al. 2013). As can be seen from the results, the optimal temperature interval for ENR degradation by Humicola sp. KC0924g was consistent with that for the growth of temperate submerged plants. Subsequently, the effect of different pH values (ranging from 5.0 to 11.0) on the degradation of ENR by Humicola sp. KC0924g was investigated at the optimum temperature of 28°C. The degradation rate was not significantly different when the initial pH was 7.0, 8.0 and 9.0 (p>0.05), and the maximum degradation rate was as high as 98.2% at pH of 9.0. The degradation rate decreased significantly when the initial pH was below 6.0 or above 10.0. The degradation extents of ENR affected by pH were ranked in the following order: pH9.0 ≥ pH8.0 ≥ pH 7.0>pH 10.0>pH 6.0 (Fig. 2b). The above experiments showed that this strain could degrade ENR efficiently under neutral to slightly alkaline conditions. The optimal pH range required for fish and shrimp growth in aquaculture is between 7.5 and 8.0 (Liu and Wang 2012). During this range, ENR can be efficiently degraded by strain KC0924g. The dynamic degradation curve of ENR (initial concentration of 1 mg L− 1) under the optimal condition (28°C and pH 7.0) by strain KC0924g is shown in Fig. 2c. The degradation rate increased immediately during the first 12 h, then gradually slowed down between 12 h and 72 h and remained stable after 72 hours.
Identification of the metabolites and proposed ENR degradation pathways
Nine metabolites were preliminarily detected in the catabolic supernatant by means of ultra-performance liquid chromatography coupled with high-resolution electrospray ionization mass spectrometry (UPLC-HRMS) (shown in Fig. 3). All identified peaks were found to correspond to protonated derivatives of the theoretical values of the corresponding compounds with errors between − 1.39 and 0.04 ppm (shown in Table 1). In contrast, no corresponding mass spectrum peaks of these metabolites were found in the control. Four metabolic pathways were proposed based on the identified metabolites (Fig. 4). In Pathway I, ENR underwent transformation through the splitting of the piperazine moiety, resulting in the production of Compounds C1 and C2. According to Adjei et al. (2006), it was believed that the antibacterial activities of Compounds C1 and C2 were reduced due to the cracking of the piperazine ring. In Pathways II, III, and IV, ENR underwent biotransformation through hydroxylation of the aromatic core, oxidative decarboxylation, and oxidative defluorination to generate C3, C4, and C5, respectively. For C5, further pathways existed. These included pathway iv1 and pathway iv2. In pathway iv1, C6 was created by further hydroxylation of C5 at the aromatic core. In pathway iv2, C7 was produced by oxidative removal of the cyclopropyl group from the quinoline ring of C6. Afterward, the piperazine ring of C7 slowly degraded through C-N bond cleavage, yielding C8 and C9.
Table 1
Spectral data for ENR and postulated biodegradation metabolites determined from UPLC-HRMS
Compound | Peak mass [M + H]+(m/z) | Display Formula | Theo. mass [M + H]+(m/z) | Delta (ppm)* | RT (min) | References |
---|
ENR | 360.1713 | C19H23O3N3F | 360.17180 | -1.39 | 11.25 | |
C1 | 334.1559 | C17H21O3N3F | 334.15615 | -0.88 | 3.99 | (Wetzstein et al. 1997, Karl et al. 2006) |
C2 | 263.0826 | C13H12O3N2F | 263.08265 | -0.14 | 1.81 |
C3 | 376.1663 | C19H23O4N3F | 376.16671 | -1.20 | 1.21 |
C4 | 332.1768 | C18H23O2N3F | 332.17688 | -0.19 | 1.11 |
C5 | 358.1761 | C19H24O4N3 | 358.17613 | 0.04 | 3.69 |
C6 | 374.1708 | C19H24O5N3 | 374.17105 | -0.64 | 0.83 |
C7 | 318.1446 | C16H20O4N3 | 318.14483 | -0.66 | 3.10 | This study |
C8 | 249.0870 | C12H13O4N2 | 249.08698 | -0.09 | 12.41 | This study |
C9 | 235.0713 | C11H11O4N2 | 235.07133 | -0.25 | 10.91 | This study |
*Generally, a mass error between − 5 ppm and 5 ppm is acceptable (Blake et al. 2011).
Previous studies have shown that basidiomycetes fungi decay wood and degrade certain xenobiotics due to their ability to produce hydroxyl radicals (H2O2), which trigger an extracellular Fenton-type reaction(Jensen et al. 2001). A total of 137 ENR metabolites produced by Gloeophyllum striatum and other basidiomycetous fungi were identified, in which the principal routes for ENR metabolism include hydroxylation of the aromatic core, oxidative decarboxylation or defluorination, or decomposition of the piperazine moiety and N-oxidation (Wetzstein et al. 1997, Karl et al. 2006, Wetzstein et al. 2006, Parshikov and Sutherland 2012). In this study, ENR major metabolites catalyzed by the ascomycete fungi Humicola sp. KC0924g were similar but differed slightly from those of Gloeophyllum striatum. The presence of the fluorine atom at the R6 position, the piperazine ring at the R7 position, and the cyclopropyl group at the R1 position in FQs improves the antibacterial effect of the drug and reduces the compatibility of the drug with microbial degradation (Van Caekenberghe and Pattyn 1984, Bhatt and Chatterjee 2022). In Pathway IV to iv2, ENR first underwent hydroxyl substitution to remove the fluorine group, then the cyclopropyl group was removed, and finally, the piperazine ring gradually cracks. As the three essential functional groups in the ENR molecule were consecutively degraded, we speculated that this pathway might diminish the antibacterial efficacy of ENR while enhancing its degradability. However, due to the low concentration of the metabolites, the information obtained from UPLC-HRMS alone is not sufficient to fully analyze their chemical structures. The exact stereochemistry of these trace metabolites will be determined in future work by enriching and purifying the metabolites followed by NMR spectroscopy.
Degradation of other fluoroquinolones, malachite green, and leucomalachite green by strain KC0924g
Strain KC0924g showed degradative capabilities against additional types of FQs, such as including CIP, NOR, and OFL. The removal rates of CIP, NOR, and OFL by strain KC0924g after 72 h of incubation were 86.4%, 70.2%, and 46.4%, respectively (Fig. 2d). It was found that the different types of FQs have similar molecular configurations and follow similar biodegradation pathways, but slight differences in their molecular structures or different toxicities to microorganisms could lead to altered degradation rates (Parshikov and Sutherland 2012, Bhatt and Chatterjee 2022). FQs have similar chemical structures but different toxicities to microorganisms, which may explain why KC0924g is able to degrade different classes of FQs at different rates.
Furthermore, strain KC0924g could also degrade MG and LMG (Fig. 2d and Fig. S2). The degradation rate of MG reached 99.9% without accumulating LMG and that of LMG reached 87.5%. MG is a synthetic triphenylmethane dye with strong bactericidal, fungicidal and parasiticidal properties (Srivastava et al. 2004). As a result, MG was once extensively applied in fisheries to prevent and control diseases, such as saprolegniasis, white spot disease, and ciliate disease (Bergwerff and Scherpenisse 2003, Srivastava et al. 2004). LMG is the major reductive metabolite of MG, which persists in fish tissues and the environment for long periods of time (Li et al. 2012). The use of MG in aquaculture is now strictly prohibited as MG and LMG have been shown to be toxic and carcinogenic (Lin et al. 2016). However, residues of MG and LMG are often detected in the aquaculture environment. Therefore, MG must be removed from the environment to prevent potential harm to human health. The ability of strain KC0924g to degrade many common contaminants in the aquaculture environment makes it a good candidate for environmental remediation.
Residual antibacterial activity assay
The antibacterial activity of the supernatants against the gram-negative bacterium E. coli DH5α (non-resistant) was measured directly by means of the agar diffusion method. The residual antibacterial activity was expressed as a percentage of the initial drug (1 mg L− 1). After 72 hours of degradation by strain KC0924g, the antibacterial activities of the four FQs (ENR, CIP, NOR and OFL) in the supernatants decreased by approximately half, with residual antibacterial activities of 51.57%, 56.63%, 37.73% and 44.72%, respectively (shown in Fig. 5). In addition, a control with the supernatant of strain KC0924g after 72 h incubation in MM medium (without FQs) showed no antibacterial activity against E. coli DH5α, ruling out the effect of endogenous fungal antimicrobials in this experiment. Parent drugs that have not been completely degraded and degradation products with antimicrobial groups (e.g. piperazine groups) may account for the residual antibacterial activity in the supernatants. Čvančarová et al. (2015) showed that the supernatant of FQs degraded by ligninolytic fungi still exhibited high residual antibacterial activity against gram-negative and gram-positive strains that were not resistant to FQs in the environment. Wetzstein et al. found that the residual antibacterial activity of CIP supernatant against gram-negative Escherichia coli ATCC 8739 was attenuated after 13 weeks of degradation by species of basidiomycetes (Wetzstein et al. 1999). After degradation by a Thermous sp. isolate, the residual antibacterial activity of the supernatant of four FQs against Escherichia coli K12 decreased by 20–40% (Pan et al. 2018).
Cu2+ treatment increased the efficiency of ENR removal by strain KC0924g
It was observed that resting cells of strain KC0924g treated with Cu2+ displayed a higher efficiency of ENR removal in comparison to untreated resting cells. The removal of ENR by cells of strain KC0924g with or without Cu2+ treatment could be fitted by the first-order kinetic model. Table 2 presents the values for k (0.0308, 0.0763) and R2 (0.931, 0.969), demonstrating that the first-order kinetic model corresponds to the data well. The half-life of ENR (1 mg L− 1) measured in Cu2+-treated cells reached 3.9 h, compared to 9.8 h obtained in untreated cells. Moreover, the removal rate of ENR in Cu2+-treated cells within 24 h reached 98.9%, a higher value than the 77.39% measured in untreated cells. The negligible change in ENR concentration upon treatment with Cu2+ alone (in the absence of microorganisms) confirmed that microorganisms dominate ENR degradation. The results showed that the addition of Cu2+ could greatly improve the degradation efficiency of ENR by strain KC0924g. This study is consistent with some previous studies indicating that low levels of Cu2+ may promote xenobiotic degradation. For example, Xu and Wang (2014) found that low concentrations of Cu2+ stimulated the degradation efficiency of decabromodiphenyl ether by the white rot fungus Phlebia lindtneri JN45(Xu and Wang 2014). Similarly, Sangare et al. (2014) demonstrated that Cu2+ is an activator of aflatoxin B1 degradation by Pseudomonas aeruginosa N17-1(Sangare et al. 2014).
Table 2
Kinetic parameters and removal of ENR (1 mg L− 1) by strain KC0924g (induced by Cu2+) after a 24 h incubation
Cu2+ (µM) | k (h− 1) | T1/2 (h) | R2 | Removal (%) |
---|
0 | 0.0308 | 9.8 | 0.931 | 77.39 |
500 | 0.0763 | 3.9 | 0.969 | 98.94 |
k-Kinetic removal rate constant (h− 1); |
T1/2-Degradation half-life (h); |
R2-correlation coefficient.
Comparative transcriptome analysis of strain KC0924g with and without Cu2+ treatment
Six libraries were constructed from mRNA of control (CK) and Cu2+-treated (CU) samples to investigate the transcriptomic responses to Cu2+ in strain KC0924g. These libraries each generated approximately 53 million, 74 million, 53 million, 53 million, 57 million and 51 million raw reads, and after quality control, over 88% and 83% of clean reads were obtained from CK samples (CK-1, CK-2 and CK-3) and Cu2+-treated samples (CU-1, CU-2 and CU-3), respectively (Table S3). After de novo assembly using Trinity software, 24531 unigenes were obtained, with an average length of 1513 bp and an N50 of 2519 bp. The number and percentage of unigenes with significant similarity to sequences in the NR, GO, COG, KEGG and SWISS-prot databases are listed in Table S4. Generally, 21,527 unigenes were annotated in at least one database, accounting for 87.75%. Comparative transcriptome analysis was performed to explore the differentially expressed genes (DEGs) between the CK and CU groups. A total of 11570 genes were identified as DEGs in one group relative to the other group. In the CU group, 275 genes were upregulated and 11,295 genes were downregulated compared to the CK group (Fig. 6a, b), indicating that Cu2+ has a repressive impact on the expression of the majority of DEGs. Gene Ontology (GO) enrichment analysis was conducted to identify the primary functional categories of DEGs. As shown in Fig. 6c and Table S5, the DEGs were mainly enriched in the GO categories of biological processes (BP) and cellular component (CC). The top three enriched items were plastid (genes number = 1509, cellular component category, GO:0009536), chloroplast (genes number = 1481, cellular component category, GO:0009507) and developmental process (genes number = 1361, biological processes category, GO:0032502). The Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis of the DEGs identified 37 significantly enriched pathways (P<0.05), of which ribosome (ko03010), carbon metabolism (ko01200) and protein processing in the endoplasmic reticulum (ko04141) were the top three enriched pathways (Fig. 6d and Table S6).
Most of the detected DEGs involved in xenobiotic biodegradation and metabolism were downregulated. However, only two genes encoding FAD-dependent monooxygenase (TRINITY_DN2180_c0_g1:K00480) and a hypothetical protein (TRINITY_DN13314_c0_g1) were found to be upregulated (Table S7). FAD-dependent monooxygenase (FMO), an oxidoreductase with flavin adenine dinucleotide or flavin mononucleotide as a coenzyme, is involved in diverse biochemical redox reactions and the synthesis of intricate natural products (van Berkel et al. 2006, Huijbers et al. 2014). In nature processes, FMO catalytic activity initiates the breakdown of xenobiotics, and bacteria have also used this catalytic activity to degrade or inactivate clinically used antibiotics(Reis et al. 2021). A class A FMO, bacterial tetracycline destructase (TetX), has been discovered to hydroxylate the C11a of tigecycline, causing it to become inactivated by producing 11a-hydroxytigecycline (Volkers et al. 2011). Rifampicin monooxygenase (Rox) is another class A FMO that was propsed to catalyze the hydroxylation of C2 in rifampicin, resulting in the generation of 2’-N-hydroxy-4-oxo-rifampicin based on an NMR analysis of the Rox crystal structure(Hoshino et al. 2010). On the other hand, BVMO (Baeyer-Villiger monooxygenase) is a class B FMO that has been discovered to oxidize the carbonyl moiety of the β-lactam ring (Minerdi et al. 2016). Molecular docking studies have identified active sites within the crystal structure that interact with vancomycin and/or methicillin (Hwang et al. 2018). Furthermore, the sulfonamide monooxygenase (SadA/SadC) constitutes a two-component FMO system that facilitates the conversion of sulfamethoxazole to products that lack antimicrobial activity (Ricken and Kolvenbach 2017). The upregulated FMO (TRINITY_DN2180_c0_g1) in this research probably accommodates the ENR for oxidative reactions. Nonetheless, further investigation is necessary to unravel the precise role of this FMO.
A previous report indicated that laccase and cytochrome P450 enzymes play roles in FQs degradation by white-rot fungi (Prieto et al. 2011, Gao et al. 2018). Through transcriptome analysis, we found that only one laccase gene (TRINITY_DN6638_c0_g1) was upregulated in the CU group, and no significantly upregulated genes encoding cytochrome P450 enzymes were found in the same group. The BLAST results indicated that the amino acid sequence similarity of this particular sequence with the laccase sequences from Trametes versicolor and Phanerochaete chrysosporium were 24.26% and 25.99%, respectively (Fig. 7). For comparison, the amino acid sequence similarity of the laccases from Trametes versicolor and Phanerochaete chrysosporium was 26.91%, and both laccases have been previously observed to biotransform FQs. The laccase activity of strain KC0924g was not detected, possibly due to low quantity and activity of extracellular laccase. Futher attempts to increase the laccase concentration will be made to study its catalytic function. The transcriptome analysis of strain KC0924g highlighted potential biodegradation-related enzymes, which could assist in future investigation of the microbial degradation molecular mechanism of FQs.