This study clearly demonstrated the ability of the Bacillus strain (AMRI-03) isolated from eutrophic lake, to degrade ATX-a toxin. To date, only one study identified a Pseudomonas strain responsible for ATX-a degradation (Kiviranta et al. 1991). Our study is, therefore, the second one to recognize and identify a bacterial strain as ATX-a degrader. In our study, AMRI-03 strain showed no lag phase when was grown with ATX-a. This contradicts the results of a previous study reporting a lag phase during ATX-a degradation by bacterial populations (Kiviranta et al. 1991). The absence of lag phase by AMRI-03 strain was also observed during CYN biodegradation (Mohamed and Alamri 2012). This is in agreement with the suggestion that cyanotoxin biodegradation occurs rapidly by strains with a previous history of cyanotoxins in water bodies (Mohamed and Alamri 2012; Smith et al. 2008).This could be the situation of our strain which could have been previously exposed to this cyanotoxin contained in cyanobacterial booms.
Our results also showed that strain AMRI-03 entered a decline phase earlier (3 days) in control (i.e., without ATX-a) compared to 6 days in ATX-treated cultures. This indcates that this bacterium could employ ATX-a as a carbon and nitrogen source for its growth, after the depletion of the natural organic matter found in lake water used in our batch experiments. Accordingly, previous studies demonstrated that bacteria can grow in preconditioned sediments (Klitzke et al. 2010) and lake water with CYN without any addition of natural organic matter (Mohamed and Alamri 2012), possibly using of CYN as a carbon source.
Interestingly, strain AMRI-03 showed faster complete ATX-a degradation at higher initial ATX-a concentrations (50 & 100 µg L− 1) than at lower concentrations (1–20 µg L− 1). Similar results of bacterial biodegradation were also reported for another cyanotoxin (CYN), where complete degradation occurred at higher initial concentrations but not observed at lower concentrations (Mohamed and Alamri 2012; Smith et al. 2008). It seems that this is the most common trait of the biodegradation of many organic pollutants. It was found that genes involved in the biodegradation of aniline by Pseudomonas sp. were induced by the presence of aniline (Thomas and Peretti 1998), the biphenyl degradation genes were induced by the presence of biphenyl (Ohtsubo et al. 2000).
This reflects that ATX-a can enter the bacterial cell and acts as an inducer activating the genes expressing the enzymes involved in the toxin degradation. In this respect, it has been reported that the cyanotoxin, microcystin after uptake into Sphingopyxis cells, is firstly hydrolyzed from cyclic form to linear intermediate by the microcystin-degrading enzyme, MlrA, in the periplasmic space, and this linear microcystin is then degraded in the periplasmic space by the enzyme MlrB into tetrapeptide, which in turn is degraded by the enzyme, MlrC in the cytoplasm (Maseda et al. 2012). Unlikely, ATX-a degradation by our Bacillus strain AMRI-03 (Gram positive bacteria) could not occur in the periplasm, as the periplasm is peculiar to the gram-negative bacteria. Instead, Gram-positive bacteria secrete degradative enzymes into the surrounding environment, and these enzymes act beyond the cell to digest or alter toxic substances to forms that are harmless to cells (Zuber et al. 2006).
In our study, the results of MS/MS analysis revealed the presence of degradation product with m/z182 in the supernatant of Bacillus culture (AMRI-03) exposed to ATX-a for 2 days and after complete degradation of ATX-a (i.e., 4 days). Based on other publications describing ATX-a and its analogues, this degradation product could be identified as epoxy-ATX-a (Kaminski et al. 2021; Liu et al. 2022; James et al. 2005). Previous studies have demonstrated that epoxy-ATX-a is formed by oxidation of ATX-a with UV-A/TiO2 photocatalysis (Kaminski et al. 2021) or FeIII –B*/H2O2 catalytic oxidation system (Liu et al. 2022). In our study, Bacillus strain (AMRI-03) could oxidize ATX-a through producing extracellular enzymes such as monooxygenases that introduce one oxygen atom derived from molecular oxygen (Arora et al. 2010) into ATX-a converting it to epoxy-ATX-a (Fig. 4). Notably, some strains of B. subtilis were found to produce P450 monooxygenase which catalyzed the epoxidation of linoleic acid without further conversion of the epoxidation product (Hou 2006). However, further research on Bacillus strain (AMRI-03) is needed to identify the genes and enzymes involving in ATX-a degradation.
What is also noteworthy here in our study is that the degradation product, epoxy-ATX-a resulting from the biodegradation of ATX-a by strain AMRI-03, was not toxic to T. platyurus. Our results are thus in agreement with previous studies that epioxy-ATX-a is nontoxic (Kaminski et al. 2021; Liu et al. 2022; James et al. 2005). Our results also showed that strain AMRI-03 which had highest homology with Bacillus subtilis (Alamri 2010), did not exhibited any toxicity to T. platyurus. This is in accordance with other studies reporting that the majority of B. subtilis are non-pathogenic and non-toxigenic to humans (Lefevre et al. 2017). Therefore, B. subtilis has been widely used in the production of enzymes and chemicals for biotechnological and industrial applications (Su et al. 2020).
As our bacillus strain (AMRI-03) is nontoxic, it could be applied in slow sand filter to remove ATX-a in drinking water treatment plants. However, microcosm or mesocosm experiments should be set up to study the potential effects of this bacterial strain on water quality. Previously, Eleuterio and Batista demonstrated the feasibility of using drinking water biofilters containing microcystin degrading bacteria to remove microcystins from waters (Eleuterio and Batista 2010).
In our study, the statistical analysis revealed a positive correlation (R = 0.89) between the initial ATX –a concentrations and its degradation rate. The maximum rate (12.5 µg L− 1 day− 1) was obtained at higher initial ATX-a concentrations (50 and 100 µg L− 1). Our results thus support the evidence that cyanotoxin biodegradation by bacterial populations is initial-concentration-dependent (Mohamed and Alamri, 2012; Alamri, 2010; Smith et al., 2008). However, no significant difference in ATX-a degradation rate was observed between the highest two initial concentrations of ATX-a (50 and 100 µg L− 1). This finding can be due to the possible existence of a threshold ATX-a concentration that is required to induce ATX-a biodegradation by bacteria. The highest ATX-a degradation rate by strain AMRI-03 (12.5 µg L− 1 day− 1) is lower than those obtained by Pseudomonas sp. (2–10 µg mL− 1 day− 1) (Kiviranta et al., 1991) and by sediment bacteria (0.6 µg mL− 1 day− 1). This discrepancy could be due to the difference in bacterial strains involved in ATX-a biodegradation. On the other hand, the ATX-a degradation rate by strain AMRI-03 can be compared with that of CYN toxin (1.25-50 µg L− 1 day− 1) obtained by the same strain at initial CYN concentrations (10–300 µg L− 1) (Mohamed and Alamri 2012).
Our results also showed that ATX-a degradation rate by strain AMRI-03 was temperature-dependent. The highest degradation rate (10µg L− 1 day− 1) was obtained at higher temperatures (25 & 30ºC), and the decrease in temperature below 20ºC slowed down ATX-a degradation rate by a factor of 2–10. Such high optimum temperature for ATX-a degradation by strain AMRI-03 is reasonable, as this strain was isolated from a warm lake. Other studies have also demonstrated similar temperature effects for biodegradation of other cyanotoxins including microcystin (Ho et al. 2007; 2010) and CYN (Mohamed and Alamri, 2012; Smith et al., 2008; Klitzke and Fastner, 2012). These studies suggested three possible mechanisms for the effect of low temperatures on cyanotoxin degradation rates including: delaying the expression of cyanotoxin degrading genes, decreasing the activity of the enzymes involved in cyanotoxins degradation and/or decreasing the rate of cyanotoxin diffusion into the cell.
Besides temperature, pH exerted noticeable effect on ATX-a biodegradation rate by strain AMRI-03. The highest degradation rates were obtained at pH8 and 7(10 & 8.33 µg L− 1 day− 1, respectively), and this rate decreased slightly at pH 9 (7.14 µg L− 1 day− 1). Conversely, both lower and higher pH levels (pH6 & 10) reduced sharply this biodegradation rate to levels of 0.44 and 1.56 µg L− 1 day− 1. This indicates that ATX-a degradation may rely on the optimum pH for the activity of its degradation enzyme. Concomitantly, Smith and Sutton found in laboratory experiments using reservoir water with sediment microbial populations that the half-life of ATX-a was 5 days at neutral pH, while it was 14 days at pH8 and 10 and lasted 21 days at pH4 (Smith and Sutton, 1993). Similar effects of pH on cyanotoxin biodegradation were also documented for CYN by the same strain (Mohamed and Alamri, 2012). The authors recorded highest CYN degradation rates at pH7 and 8, (16.7 and 15.6 µg L− 1 d− 1), and this rate decreased sharply at higher pH levels (pH10 & 11). Despite ATX-a was not completely degraded by strain AMRI-03 at higher pH levels, it may contribute to ATX-a degradation under alkaline conditions, particularly at pH9.