Hypersaline sediments used as a source of halophilic microorganisms were collected from two extreme salt lakes, Chott El Djerid (CJ) and Sebkhet El Melah (SM), located in the south of Tunisia (N:33°57.272’ E: 008°24.393′ and N: 33°23.705′ E: 010°54.878, respectively) during the wet season (February). Samples were secured in sterile bags and stored at +4°C until utilization.

All experiments were carried out in minimal salt medium supplemented with TBZ as the sole carbon source. The detailed composition of the minimal salt medium was (per liter): 150 g NaCl, 0.5 g K₂HPO₄, 1 g NH₄Cl₂, 2 g MgCl₂, 0.1 g CaCl₂.

50 ppm of TBZ was added of either native or irradiated. An analytical standard of TBZ (≥99% purity, Sigma-Aldrich) was used for analytical purposes analyses and culture experiments. pH and temperature were adjusted at 7.2 and 38°C, respectively.

Aqueous solutions of TBZ were irradiated by a Cobalt 60 γ-source. Irradiation processes were performed in the National Centre of Nuclear Sciences and Technologies (CNSTN) at Sidi-Thabet, Tunisia. For an accurate control of the irradiation time, a specific determination of the dose rate is necessary. Dose rates of 26.31 and 6.81 Gy/min were determined with Red Perspex dosimeters (M’Garrech et al., 2012). The doses applied were 1, 2, and 4 kGy. The method of dose rate measurement is described in Ounalli et al. (2017). It is important to mention that when gamma rays interact with water, unstable and stable oxidizing and reducing species are formed and will act as modifiers of the TBZ structure.

Single-chamber 750 ml glass reactors equipped with a three-electrode system were used for microbial anodes formation. A carbon felt (6 × 1 × 1 cm³, Mersen) was used as a working electrode (WE). A platinum grid (Heraeus) was used as the auxiliary electrode (AE), and a saturated calomel reference electrode (+0.24 V/SHE, Origalys) was located between the WE and the AE. The three electrodes were connected to a multichannel Potentiostat/Galvanostat (OGF 500, Origalys) managed by the OrigaMaster five software (Figure 1). Constant polarization at −0.10 V vs. SCE was applied on the WE. Chronoamperometric maximum current densities (CA) of microbial bioanodes were determined as indicated in our previous studies (Askri et al., 2019–2020).

The global 3D structure of biofilms and the distribution of microorganisms on the carbon fibers of anodes were investigated using epifluorescence microscopy. At the end of the 7 days of microbial anodes formation, a first part of the WE covered by THS enriched biofilms was stained immediately with acridine orange 0.01% (A6014 Sigma) for 10 min then washed and dried at room temperature for 24 h. Biofilms architecture was observed with a Carl Zeiss Axio ImagerM2 microscope motorized in Z and equipped with a UV light source (HBO 200) and a Zeiss 09 filter (HP450-490 exciter, FT-10 reflector, LP520 barrier filter). Image acquisition is performed with a Zeiss digital monochrome camera (AxioCammm RM). With the help of the Zen software, the images are processed to obtain projections from the acquisition of image stacks scanning several tens of µm in biofilm thickness.

After 7 days of operation, both the anodic biofilm and the bacterial pool present in the electrolyte were collected from the reactor in which the bioelectrochemical degradation of nTBZ occurred by using CJ sediments as inoculum (CJ nTBZ experiment). Culture depended approach was carried on to phylogenetically characterize the existing community in the sediments leading to TBZ degradation and current production. For molecular identification, portion of the bioanode (about 2 cm) and 5 ml of electrolyte were separately added to sterile test tube containing physiological solution (0.9% NaCl). After a serial dilution (10^–1 to 10^–7), in the culture medium described above supplemented with 15% agar, successive subcultures were performed at 38°C until isolation of pure strains. Genomic DNA from pure strains was extracted by Organic (Phenol–Chloroform) method (Li et al., 2007). DNA quality and quantity were measured by a Thermo Scientific NanoDrop™ ND-2000 Spectrophotometer (Thermo Fisher Scientific, Waltham, MA, United States). PCR reactions and sequencing were performed based on 16S rRNA gene using universal bacteria primers S-D-Bact-0008-a-S-20/S-D-Bact1495-a-A-20 (Ettoumi et al., 2013). 16S rDNA sequencing was carried out in the research and technology transfer support unit at the BorjCedria Biotechnology Center. DNA sequences were edited using BioEdit software. The resulting 16S rDNA sequences were used via EzBioCloud (Yoon et al., 2017) to determine the taxonomy of the strains. The phylogenetic dendrograms was constructed using the phylogeny. fr website.

Two distinct types of extreme Tunisian hypersaline sediments (THS) were compared as a source of halophilic microorganisms for 1) forming anodic biofilms on carbon-felt electrodes, and 2) biodegrading native TBZ (nTBZ) and TBZ pre-irradiated (iTBZ) with increasing gamma irradiation doses. Both forms of TBZ were provided as the exclusive substrate, it is the only possible source of energy and organic carbon for the microorganisms. Overall, eight electrochemical experiments were performed with a synthetic FPWW. The performance of the THS for current production and TBZ degradation was evaluated using nTBZ and iTBZ separately. Current densities, cumulated charges, Coulombic efficiencies and COD removals determined from different experiments for both sources of microorganisms are summarized in Table 1. The detail of the eight chronoamperometric curves is shown in Figure 2. A control experiment without the presence of electrodes was also conducted with sediments from CJ. The SM sediment shows lower performance compared to the CJ sediment particularly in terms of current densities and Coulombic efficiencies. A decrease in current densities exceeding 70 and 90% is noted for nTBZ and iTBZ-4kGy, respectively (Table 1). As the experiments were performed under the same conditions, microorganisms hosting the SM sediments could be unable to degrade TBZ even in its irradiated form. As demonstrated in Table 1, CJ sediment supported higher COD removal (>80%) than SM sediments. The removal efficiency of nTBZ (>80%) was comparable to that of azo-dye (85%) (Karuppiah et al., 2018) and better than those of petroleum hydrocarbons (49.38%) (Mohanakrishna et al., 2019) and Ibuprofen (64.28%) (Wang et al., 2022).

Table 1 showed that initial COD of iTBZ may increase with elevated irradiation doses (1 and 2 kGy). In contrast, a declining trend in COD removal rate efficiency, accompanied by a decrease in current generation, was observed. Contrary to what was expected irradiation pretreatment of TBZ had prominent negative effect on both the COD removal and the current generation. In fact, with CJ as inoculum, the 7 days COD removal rate of 50 ppm nTBZ was 1.78 and 1.77 times of those for iTBZ-1.0 kGy and iTBZ-2.0 kGy, respectively. Similarly, maximum current density of nTBZ was 4.21 and 4.77 times of those monitored for iTBZ-1.0 kGy and iTBZ-2.0 kGy, respectively.

In general, the evolution of the current density as a function of time (Figure 2) is well consistent with what is known about the formation of an anodic biofilm on carbon-felt electrodes. Thus, taking the experiments with nTBZ as an illustration evolution of the electrochemical activity is analogous to the successive formation and growth phases of electroactive bacterial biofilms on the working electrode (Baranitharan et al., 2015). Latent phase, at zero current, during this period, the bacteria adapted to the sessile lifestyle by progressive colonization at the surface of the anode using nTBZ as electron donors. Followed by a low anodic current that slowly increased to attend their maximum explained by the nTBZ or iTBZ oxidation. The production of current continued to increase to reach maximum current densities at the end of the experiment of 0.95 and 3.66 mA/m², for SM and CJ inoculums, respectively. Previous studies using CJ as inoculum, with lactate (5 g/L) or real textile wastewater as energy sources, have demonstrated the ability to design electroactive bioanodes with current production of 6.98 ± 0.06 and 12.50 ± 0.20 A/m², respectively. Additionally, the COD removal rate was 91 ± 3% when real textile wastewater containing azo-dyes was used as energy substrate (Askri et al., 2019; Askri et al., 2020). In this study, a comparable COD removal was obtained (>80%) in particular with nTBZ. However, the current production remains very low (0.003 A/m²).

The increase of the irradiation doses on the iTBZ resulted in a very important decrease in current generation in the case of both sources of inoculum (Figure 2 and Table 1). And even an almost zero current production is recorded for the pre-treated iTBZ-4.0 kGy. However, the COD removal rate is 75.85%. It is both very similar to that obtained with nTBZ in the bioelectrochemical experiments and to that resulting from the biological degradation experiments of iTBZ-4.0 kGy without electrode. In all the cases studied, the gamma irradiation did not improve the anodic current resulting from the oxidation of iTBZ. LC-MS analysis of the irradiation degradation products and the evaluation of their toxicity towards microorganisms could clarify the observed effect of the gamma irradiation of TBZ.

In fact, the monitoring of nTBZ and iTBZ biodegradation by measuring the optical density at specific wavelengths demonstrated a degradation rate less than 50% in different synthetic FVPWW containing iTBZ as substrate. However, the degradation rate was >80% when nTBZ was used with CJ as inoculum. Despite this high degradation rate of nTBZ, the current density produced remains really weak. This could be explained by the fact that several bacteria in the synthetic FVPWW non-anode adherent microorganisms and called “planktonic” consume almost all of the nTBZ in the synthetic FVPWW without participating in the current production. Consequently, the competition for oxidizing nTBZ between electroactive and non-electroactive bacterial species results in a low current density in chronoamperometric experiments. It is particularly important to note that microbial electrochemical processes involving bioremediation of recalcitrant pollutants do not have energy as the main output (Liang et al., 2020). However, weak electrostimulation (ES) could be necessary for enhancing the microbial transformation of nTBZ. Ailijiang et al. (2016) reported that ES (2 mA) is potentially effective for biofilm reactors treating phenol-containing wastewater. More recently Wang et al. (2022), demonstrated that weak ES enhanced the microbial transformation of anti-inflammatory recalcitrant drugs such as ibuprofen and naproxen. The weak ES could influence the growth of cells and metabolic behaviors of microorganisms. In this study, the electrostimulated biodegradation removal rate of nTBZ after 7 days was greater than 80%, which corresponds to an abatement of the concentration of 40 ± 0.5 ppm of nTBZ with CJ as inoculum. In terms of bioremediation these results could be very conclusive compared to those already obtained by Perruchon et al. (2017) using bacterial consortium isolated from wastewater disposal site adjacent to a citrus fruit packaging plant and 20 ppm of TBZ as initial concentration.

Halophilic sediments such as CJ have shown in previous works their efficiency in designing halotolerant bioanode able to degrade azo dyes in textile wastewater with more than 90% of COD removal rate (Askri et al., 2020). Moreover Pugazhendi et al. (2020) reached a COD removal efficiency of up to 89% when used halophilic microorganisms to treat seafood industrial wastewater. Similarly, Feng et al. (2015) reported enhanced removal of recalcitrant p-Fluoronitrobenzene when a MEC was operated at highly saline conditions. In view of the findings CJ sediment is a promising potential source of tenacious exoelectrogenic halophilic microorganisms for enriching the microbial communities on the anode to degrade nTBZ or iTBZ. Consequently, CJ sediments are the subject of more in-depth investigations in this study.

It is clear at this stage of understanding that the conversion of TBZ, whether irradiated or not, is not linked to the direct oxidative action of an electroactive biofilm established on the surface of the carbon felt. The recorded current densities, and especially the low Coulombic efficiency calculations (Table 1), rather suggest an electrostimulation activity that would boost the biodegradation activity of TBZ. This electrostimulation, effective via an electrode polarized at −0.10 V/SCE, clearly allows an increase in the TBZ removal rate of 1% in the least convincing case of iTBZ-4.0 kGy and up to 38% in the most conclusive case of nTBZ removal.

In any case, even if an anodic electroactive biofilm does not seem to play a major role, the external surface of the carbon-felt electrodes was still imaged in search of microorganisms that could be electrostimulated by the presence of the polarized electrode. At the end of the chronoamperometries at -0.10 V/SCE, the carbon felt working electrodes were removed from the reactors containing nTBZ, iTBZ with SM and CJ as inoculums. As a reminder, these electrodes produced maximum current densities of 0.95–3.66 mA/m², and TBZ removal rates of 81 to 48.99%. The electrodes have been stained with a fluorescent dye that is selective for nucleic acids, and therefore able to detect any microorganisms on the surface of the carbon fibers of the felt. The external surfaces of the electrodes were imaged by epifluorescence microscopy (Figure 3).

Despite the very low current densities detected, all electrodes appear to host a biofilm of microorganisms. In general, the colonization is mostly centred on the surface of the carbon felt fibers without any real continuous proliferation or interconnection between the carbon fibers. This is already characteristic of saline environments where microorganisms are more likely to form electroactive biofilms strongly sheathing the fibers of conductive materials (Rousseau et al., 2014; Askri et al., 2019). Although in our work the biofilms are not very electroactive, the findings seem to conclude that the structural organization of the biofilms is the same.

In terms of biofilm density on the electrode surface, the two inocula induce different colonization rates in the presence of either more or less irradiated iTBZ. In the case of SM THS, colonization appears to be maximal in the presence of nTBZ or low-irradiated iTBZ (iTBZ-1.0 kGy), then the colonization rate appears to collapse as the TBZ irradiation dose increases. In contrast, CJ THS positively respond to iTBZ. The colonization of the carbon felt even increased as the irradiation dose of the iTBZ was intensified. These findings may suggest that the gamma degradation products of TBZ inhibit the growth of microorganisms from the TSH SM, while on the contrary they promote the development of those from TSH CJ.

At the end of the operation period of the bioelectrochemical degradation of nTBZ using CJ as inoculum, biofilm samples on the anode surface and electrolyte were collected. Culture-dependent approach was applied to isolate bacterial strains community involved in TBZ degradation. As reported in Figure 6 at phylum level, the bacterial community was mainly presented by two phyla Proteobacteria (75%) and Firmicutes (25%). At species level Bacillus velezensis, Bacillus spizizenii, and Bacillus inaquosorum were isolated as Firmicutes strains. Then, Halomonas smyrnensis, Halomonas halophila, Halomonas salina, Halomonas organivorans and Halomona skoreensis as Proteobacteria strains.

Previous studies have shown the ability of these isolated strains to degrade several refractory toxic molecules (azo dyes, pesticides and hydrocarbons) and able to tolerate heavy metals. B. velezensis was reported as azo dye degrading bacterium (Bafana et al., 2008). Vörös et al. (2019) examined the ability of B. velezensis to tolerate heavy metals. It is also reported as pesticides degrading bacterium. Similarly, several study (Ishag et al., 2016; Ishag et al., 2017; Abdelbagi et al., 2018) reported the ability of B. Inaquosorum in the degradation of different pesticides such as chlorpyrifos, malathion, dimethoate, pendimethalin, endosulfan alpha and endosulfan beta. Moreover, B. Inaquosorumis shown to be involved in the treatment of recalcitrant petrochemicals (Ahmadi et al., 2019). Halomonas genus as well, has a considerable potential ability for degrading aromatic compounds of saline contaminated waters (CortiMonzón et al., 2018). It has been demonstrated that those isolates could degrade polycyclic aromatic hydrocarbons such as pyrene, fluoranthene, fluorine (Nanca et al., 2018; Govarthanan et al., 2020). In addition, Halomonas sp. are widely explored as dyes removal bacteria, they were used for decolorization of aqueous solution containing azo dyes such as Reactive Black 5 (RB5), Remazol Brilliant Violet 5 R (RV5), and Reactive Orange 16 (RO16) (Montañez-Barragán, et al., 2020). As these, strains have proven to be effective and live naturally in an environment highly contaminated by pesticides and have the potential to eliminate pesticides from contaminated waters, especially wastewaters from pesticide industries and landfill sites. Overall bacterial community structure at the genus level isolated from reactors containing TBZ was predominantly composed of Halomonas species (75%). These species were previously cited as electroactive bacteria and able to grow in high levels of salinity and possess particular electrochemical properties (Askri et al., 2019; Uma Maheswari et al., 2019; Liu and Wu, 2021).