(1) Our experiments consider intracellular microcystins and extracellular microcystins dissolved in water using non-sonicated and sonicated cells, respectively, and (2) two acute microcystin concentrations in water.

Radiation-Induced Microcystin Attenuation and Monitoring (1) We use an ample interval of doses of 0, 3, 5, 10 and 15 kGy, and (2) bioassays with Cladocerans to evaluate the microcystin toxicity attenuation.

We express the experimental results as an ad hoc attenuation coefficient, as a function of doses imparted to microcystin water solutions at different concentrations and doses.

From this body of new and fuller information on microcystins exposed to electrons and gamma radiation, we contemplate the following environmental aspects: (1) remediation of water sources contaminated by microcystins using ionizing radiation (gammas and electrons), illustrated by a working example; (2) a better conceptual and experimental understanding of the behavior of microcystins as contaminants, and (3) contribution to ameliorate strategies to combat chronic contamination scenarios of the environment.

Damaging processes to DNA and large molecules, as e.g. proteins and cyanotoxins, are initiated by direct impact of radiation on water molecules, leading to free radical formation by breaking down chemical bonds. Additionally, incidental radiation gives rise to low-energy secondary electrons (smaller than ≈ 20 eV) at significant fluxes. Approximately 5 × 104 secondary electrons are produced per 1 MeV of transferred energy 1920, playing an important role in inducing DNA and protein damage.

Electrons, in turn, are charged particles transferring most of their energies to the aqueous medium molecules, that is, water itself plus larger molecules therein dissolved, thus producing copious fluxes of secondary electrons. In fact, the stopping power of gammas in water is much weaker than that of electrons, explaining why their production of secondary electrons is considerably smaller.

We used the cyanobacteria strain of toxin-producing Microcystis aeruginosa Kutzing 1846, NPLJ-4, and a non-toxic strain of Microcystis sp. supplied by Dr. Armando Vieira at the Federal University of São Carlos, Brazil. Both strains were cultivated in 6 L Erlenmeyer flasks containing 4 L ASM-1 medium 21, with the pH adjusted between 7.0 and 7.5. The strains were maintained under a light:dark photoperiod of 12:12 h and at a constant temperature of 25.0 ± 0.2 °C. The biological material was harvested at the late exponential phase of growth, centrifuged (Quimis-Q 222T) and lyophilized at −80 °C (Lyophilizer LB 5000 TT-TERRONI-FAUVEL) until fully dehydrated.

The crude cyanobacterial extract samples were prepared with 100 mg of freeze-dried cells and 1 L of distilled water (stock solution). Two separate procedures, referred to in this study as “sonicated” and “non-sonicated”, were performed in order to evaluate the degradation of microcystin through ionizing radiation. First the solution was homogenized and centrifuged, and then the toxins were sonicated in aqueous medium, and finally the microcystin in the supernatant was exposed to radiation. After homogenization, the crude extract of the non-sonicated samples (toxin inside the cell) was first directly exposed to radiation and then sonicated and centrifuged.

Both sonicated and non-sonicated samples were irradiated with two types of ionizing radiations: gamma and electron beams at doses of 0 (control), 3, 5, 10 and 15 kGy. Gamma radiation was carried out with a ⁶⁰Co gamma source facility (Gammabeam, model 650 from MSD Nordion, Ottawa, Canada) at a rate of 1.98 kGy h^-1. Irradiation with electron beams (LINAC, model Dynamitron DC 1500/25/4, job 188) was performed with an energy of 1.174 MeV and at the following doses: 3 kGy for 03 min. at 6.70 kGy s^-1 dose rate; 5 kGy for 03 min. at 11.16 kGy s^-1 dose rate; 10 kGy for 03 min. at 22.32 kGy s^-1 dose rate; and 15 kGy for 06 min. at 16.74 kGy s^-1 dose rate.

Toxicity tests of M. aeruginosa subjected to gamma radiation were performed with Ceriodaphnia silvestrii (Cladocera, Crustacea). For irradiation with electron beams, bioassays included C. silvestrii and two additional species: Argyrodiaptomus furcatus and Notodiaptomus iheringi (Copepoda, Calanoida).

Stock cultures of C. silvestrii Daday 1902 have been maintained for several years at the Ecotoxicology Laboratory of the Federal University of São Carlos, Brazil, reproducing parthenogenetically. The individuals used in this experimental setup originate from a single female to eliminate variation among clones. Cultures were maintained in 2 L glass beakers, and the medium used (modified from 22) was reconstituted water (pH 7.0–7.6; conductivity 160 mS cm^-1 and hardness 40 to 48 mg CaCO₃ L^-1) and renewed three times weekly. Cladocerans were fed daily with Pseudokirchneriellasubcapitata at a concentration of 3 × 105 cell mL^-1 and a yeast suspension (1 mL.L^-1) prepared with active commercial dried yeast (Fleischmann’s^®) and distilled water (0.25 g in 50 mL^-1). Fifty individuals per beaker were kept under a light: dark photoperiod of 12:12 at 25 ± 1 °C in an incubator. All experiments were initiated with third brood neonates (≤ 24 h old) derived from healthy parent stock.

Two week before beginning the experiments, we evaluated the health and sensitivity of the cultures by acute toxicity tests with the reference compound sodium chloride, NaCl 22.

Acute toxicity tests were carried out with C. silvestrii, A. furcatus and N. iheringi to analyze their survival during exposure to aqueous crude extracts of irradiated and non-irradiated Microcystis aeruginosa (NPLJ-4 strain), detecting immobilization (i.e., paralysis) of the organisms exposed. The concentrations are given as milligrams of freeze-dried material per liter of water. The results of the acute toxicity tests with C. silvestrii, for instance, were analyzed using the Trimmed Spearman-Karber statistical program 23, obtaining 48-h EC₅₀, that is, the average effective concentration causing immobility in 50% of the organisms exposed to the toxic agent over a period of 48 hours.

Acute toxicity tests were carried out with sonicated and non-sonicated non-toxic Microcystis sp. irradiated with gamma radiation and electron beams at doses 3, 5, 10 and 15 kGy. None of the samples showed toxicity to Ceriodaphnia silvestrii.

All tests performed with the control dose 0 kGy of the toxic culture represented by strain NPLJ-4 showed toxic effects on C. silvestrii, with a mean value of 48-h EC₅₀ equal to 31.4 mg.L^-1 dry weight of freeze-dried material and a 95% confidence interval of 26.0–38.0 mg.L^-1 dry weight of freeze-dried material (Table 1).

The crude extracts irradiated with doses of 3, 5, 10 and 15 kGy gamma radiation systematically exhibited a higher 48-h EC₅₀ than the non-irradiated crude extracts. Both sonicated and non-sonicated samples reflected a loss of toxicity that may be attributed to efficient degradation of cyanotoxins. At the doses 10 kGy and 15 kGy, sonicated and non-sonicated crude extracts exhibited 48-h EC₅₀ with values higher than the control. Statistical analysis of 48-h EC₅₀ obtained in this study are shown in Table 1 and reveal that, with the exception of irradiation with 10 kGy, there was no difference between non-sonicated and sonicated crude extracts (p = 0.05, Student’s t-test) 24.

For bioassays carried out with sonicated crude extracts and irradiated with electron beams at doses of 3, 5, 10 and 15 kGy, the dose of 3 kGy obtained only 48-h EC₅₀ (70.6 mg.L^-1 dry weight of freeze-dried material), and we observed immobility of neonate of C. silvestrii. For non-sonicated crude extract irradiated with electron beams with doses of 3 and 5 kGy, we observed immobility of Ceriodaphniasilvestrii, presenting a 48-h EC₅₀ of 73.9 and 62.2 mg.L^-1 dry weight of freeze-dried material, respectively. Results for immobility tests of acute toxicity to Ceriodaphniasilvestrii (Crustacea, Cladocera) following irradiation with gammas and electrons are shown in Table 2.

We believe that attenuation of the toxin occurred for crude extract sonicated and irradiated with 5 kGy due to immobilities occurring in non-significant numbers (i.e., an amount insufficient to obtain 48-h EC₅₀). For non-sonicated crude extract at the same dose, the percentage of immobile neonates was elevated. The percentage of immobile individuals (data not shown in Table 2) in sonicated crude extract exposed to electron beams (toxins in aqueous medium) was 1.5%, and non-sonicated crude extract (toxin inside the cell) was 21.5% of the individuals. This behavior could be attributed to the presence of the membrane during irradiation. In fact, the membrane is an additional mass interposed between the electron beams and the toxin within the cells, and thus the dose effectively imparted to the toxin could be lower. This is not the case in irradiation with gammas due to the high penetrability of this radiation.

Other results from bioassays were obtained from toxins irradiated with electron beams. For instance, Borrely and collaborators 25, using Ceriodaphniadubia, Daphnia similis and Vibrio fischeri as test organisms, observed high effectiveness of electron beams in the degradation of the surfactant dodecyl pbenzene sulphonate acid (LAS). They showed that exposing waste from two wastewater treatment plants to electron beams proved to be highly effective for mitigating toxicity 25. Also, Romanelli and collaborators demonstrated reduced toxicity of the surfactant sodium dodecyl sulfate (SDS) when it was diluted with water and subjected to irradiation with electron beams at doses of 3, 6, 9 and 12 kGy 26. The evaluation was performed via acute toxicity tests on two test organisms, the bacterium Vibrio fischeri and the cladoceran Daphnia similis, finding the dose of 6 kGy to be sufficient in mitigating this highly toxic substance.

In the treatment of sonicated and non-sonicated crude extracts, we used four replicates, and the concentrations in the ecotoxicity tests with C. silvestrii were 5.7, 11.3, 22.5, 45.0 and 90.0 mg.L^-1 dry weight of freeze-dried material. For crude extracts irradiated with both gammas and electron beams, the immobility of neonate C. silvestrii was better observed at the higher concentrations of 45 and 90 mg.L^-1 dry weight of freeze-dried material. For a concerted visualization of the results, Table 2 shows total values of C. silvestrii neonate immobilities after acute toxicity tests (48 h). For sonicated crude extracts exposed to 3 kGy of gamma radiation, the number of immobilities was almost three times greater than the value for the crude extracts irradiated with electron beams, at a concentration of 45 mg.L^-1 dry weight of freeze-dried material. With this same concentration and at doses of 5, 10 and 15 kGy, we observed immobility only for the crude extracts irradiated with gamma, and no immobility for those crude extracts irradiated with electron beams. At a concentration of 90 mg.L^-1 dry weight of freeze-dried material and in crude extracts sonicated and irradiated with gamma radiation, the percentage of immobility of organisms occurred after the acute test (48 h). Immobility of organisms was 100% at doses of 3 and 5 kGy, while for 10 and 15 kGy the value was 98.8% and 90%, respectively. The non-sonicated crude extracts showed a number of neonate immobilities greater than sonicated crude extracts after the acute test (48 h). Immobility of organisms occurred in crude extracts exposed to gamma radiation at doses of 3, 5 and 10 kGy and toxin concentration of 45 mg.L^-1. One single immobile individual was observed after 48 hours for acute toxicity tests with C. silvestrii in four replicates, for crude extract irradiated with 15 kGy.

A cursory visual inspection of Table 1 and Figure 1 shows that the attenuation of microcystin is substantially more efficient following exposure to electron beams, relative to gamma radiation, for both sonicated and non-sonicated cells; that is, A_e >> A_γfor the entire dose interval.

If electrons as charged particles have quite small penetration lengths in matter, they also display the ability to deposit a large amount of energy in the aqueous medium and molecules, leading to the formation ofsecondary electrons192027. As pointed out in section 2.1.2, low-energy secondary electrons have the capacity to inflict damage on DNA and proteins. Secondary electrons are produced by gammas at much smaller amounts.

Our laboratory previously verified this damaging characteristic of electrons 28 and found that plasmids irradiated at intermediate and high doses of electrons are severely shattered into quite small fragments. It is thus clear that toxins exposed to electron beams would be subjected to the same fate of plasmids; they shatter in a rather efficient process to debilitate many of their functions.

Experiments with non-sonicated and sonicated samples correspond to the irradiation of cells with and without membranes. The implications of irradiation conditions are as follows: (1) innon-sonicated cells (NS), the toxin molecules are confined within the cell interior, where membranes constitute a dense attenuation material for the incoming radiation; (2) in sonicated cells (S), no membrane blocks the incoming radiation, and the toxin molecules are dispersed into the medium.

Table summarizes the results of the Toxin Attenuation Parameters, corresponding to all experimental configurations; note that A(S) > A(NS), for exposure to both gammas and electrons, and with microcystin concentrations of 45 mg.L^-1 and 90 mg.L^-1. The only exception occurred at 3 kGy for 45 mg.L^-1 of microcystin, where A(NS) > A(S) by an amount of only 10%, approximately, but within the range of experimental uncertainties.

The results in Table 3show that toxinattenuation for microcystin samples at concentrations of 45 mg.L^-1 is much higher than at concentrations of 90 mg.L^-1. This is a counterintuitive finding. However, as demonstrated in APPENDIX-A, the average energy absorbed by each target molecule, ε₀, is inversely proportional to ρ_t (mL^-1), the target density of toxin molecules; that is, ε₀~ 1/ρ_t. Thus, ε₀(45) > ε₀(90), implying A(45) > A(90).

The characteristics of the outcomes (i.e., the number of mobile and immobile individuals) are a consequence of both (1) the Microcystis biophysical conditions and (2) the amount ingested by the Cladocerans.

The Microcystis biophysical conditions refer to the degree of damage inflicted by radiation on its microcystin content, which is a deterministic process, particularly at the doses used in this study. However, the amount of ingested microcystin is a statistically fluctuating quantity, since it is driven by the erratic movement of the Cladocerans in the solution, as described by a simple random walk (similar to a diffusive Brownian motion). Therefore, thevaluesof ingested microcystin at each time is a random variable, thus characterizing the whole process as stochastic.

The Toxin Attenuation Parameters, A_i(%), were defined in Equation 1, Section 3.4. The stochastic nature of the attenuation process is discussed in Section 4.4. The deduction of a statistically driven expression for A_i presented in APPENDIX-B, Equation B-4, is

A_i(%)=100.(1–exp(–λD))

where the parameter λ is a sole property of the interaction radiation/microcystin. In this case, λ⁻¹ = D₀ is the dose necessary to attenuate 63% of the initial microcystin cell population N₀.

The data points in Figure 2 and Figure 3 correspond, respectively, to the following experimental conditions: (a) irradiation with gammas, sonicated samples and a concentration of 45 mg/L, and (b) irradiation with electron beams, sonicated samples and concentrations of 45 and 90 mg/L. The curves in Figure 2 and Figure 3 were obtained from a single-parameter best-fitting standard procedure, with λ as the parameter, and imposing A_i = 0 for D = 0. Fitting results are,

(i) λ_γ = 0.24 ± 0.05 (kGy)^-1 (for 45 mg.L^-1); therefore (see Figure 2),

A_γ(%)=100.(1–exp(–0.24D)),and

(ii) λ_e= 0.29 ± 0.04 (kGy)^-1 (for 90 mg.L^-1); therefore (see Figure 3),

A_e(%)=100.(1–exp(–0.29D)]

Therefore,

D₀(e) = 3.45 ± 0.04 (kGy) and D₀(γ) = 4.17 ± 0.05 (kGy)

The fact that D₀(e) < D₀(γ) indicates the higher lethality of electron beams, which is consequence of both (1) their substantial energy transfer to matter and (2) their prolific production of secondary electrons. This is consistent with the present findings of A_e > A_γ.