Chlorinated aliphatic compounds (CAHs) are widespread contaminants that pose serious environmental and health concerns. Due to extensive applications, careless handling and storage, spills, and leakages, CAHs are among the most abundant classes of groundwater contaminants (Hatt and Löffler, 2012). Chlorinated solvents, including perchloroethylene (PCE), migrate downward to form dense non-aqueous phase liquids (DNAPLs), which constitute a source of contamination that may last for decades in groundwater (McCarty et al., 2007). Remediation technologies conventionally consist of separation techniques (i.e., ion exchange, reverse osmosis, or nanofiltration) (Altalyan et al., 2016) or permeable reactive barriers (Obiri-Nyarko et al., 2014). Since the discovery of the biological reductive dechlorination (RD) process, many bioengineering solutions have been successfully applied for in situ CAH removal. RD is an anaerobic process that allows bacterial growth with CAHs’ biodegradation up to the harmless ethylene in the presence of electron donors (i.e., hydrogen or fermentable organic substrates) (Hug and Edwards, 2013; Paul et al., 2016). RD occurs through a step-by-step process where highly chlorinated compounds, including PCE and trichloroethylene (TCE), are sequentially reduced to cis-dichloroethylene (cis-DCE), vinyl chloride (VC), and then to ethylene. VC biodegradation is the RD’s limiting step. Indeed, uncompleted anaerobic dechlorination might occur, resulting in VC accumulation, which is more toxic than the parent compounds (Dolinová et al., 2017). However, VC can be removed further through oxidative processes when specialized microorganisms are present in the aquifer.

Several organohalide-respiring bacteria (OHRB) are known for their dechlorinating ability. Among them, Dehalococcoides mccartyi is responsible for dechlorination of highly chlorinated solvents (i.e., PCE) to ethylene (Adrian et al., 2000; Maphosa et al., 2010; Hug et al., 2013; Jugder et al., 2016; Matturro et al., 2016). RD occurs through the metabolic activity of reductive dehalogenase enzymes (i.e., TceA, BvcA, and VcrA) typical of D. mccartyi strains involved in various RD steps (Saiyari et al., 2018). D. mccartyi (by 16S rRNA gene quantification) and reductive dehalogenase genes are commonly used as biomarkers for monitoring of RD processes, both on laboratory and field scales (Jugder et al., 2015; Fincker and Spormann, 2017). Moreover, several aerobic VC-assimilating bacteria have been isolated from CAH-contaminated groundwater (Fullerton et al., 2014; Liang et al., 2017). Some microorganisms can adapt to low oxygen concentrations and form aerobic niches in polluted aquifers (Coleman et al., 2002). Among them, ethylene-oxidizing bacteria (ethenotrophs) may utilize VC directly as a carbon and energy source; furthermore, they can co-metabolize VC while utilizing ethylene as a growth substrate (Verce et al., 2001; Jin and Mattes, 2008; Mattes et al., 2010). Moreover, methane-oxidizing bacteria (methanotrophs) fortuitously oxidize VC during aerobic growth in methane (Chang and Alvarez-Cohen, 1995; Semrau, 2011). The enzymes involved in VC oxidation are alkene monooxygenase (AkMO), whose alpha subunit is encoded by the etnC gene (Hartmans and De Bont, 1992; Verce et al., 2000), and epoxyalkane coenzyme M transferase (EaCoMT) encoded by the etnE gene (Coleman and Spain, 2003; Mattes et al., 2005). The etnC and etnE genes are currently used as biomarkers for monitoring ethenotroph populations on laboratory and field scales (Mattes et al., 2015). Thus, coupling reductive and oxidative dechlorination with engineered systems is a promising strategy for CAH removal, mainly because it can ensure the complete removal of highly chlorinated ethylenes (i.e., PCE), avoiding secondary contamination by more toxic and mobile low-chlorinated by-products (i.e., VC).

Among the bioremediation strategies proposed for restoration of groundwater contaminated by chlorinated solvents, bioelectrochemical systems (BESs) emerged in the past decade as an effective solution. BESs allow prompting reductive dechlorination, overcoming the use of fermentable organic substrates that can introduce uncontrollable side reactions (Wang et al., 2020). In BESs, electrons are directly provided (in the cathode) or accepted (in the anode) to perform the dechlorination process. Despite bioelectroremediation technologies having been deeply investigated and optimized under various conditions, little is still known about the composition and dynamics of microorganisms involved in reactions occurring in electrodes (Aulenta et al., 2008, 2010; Strycharz et al., 2010; Wang et al., 2015; Chen et al., 2018).

Notably, our research group has recently proposed an innovative BES aiming at sequential reductive/oxidative PCE dechlorination occurring in two separate reactors where PCE is first reduced (cathodic chamber) to TCE, cis-DCE, and VC, and then chlorinated by-products (i.e., VC) are oxidized (anodic chamber) to achieve complete PCE removal (Zeppilli et al., 2019, 2021; Dell’Armi et al., 2021a). The system has been tested with various PCE-contaminated feeding solutions, including an anaerobic mineral medium (MM) as the optimum condition on laboratory scale, synthetic groundwater (SG) to mimic the anion species content of contaminated aquifers, and, ultimately, real contaminated groundwater (RG) collected from an industrial site located in Northern Italy.

Here, we report the microbiome compositions and dynamics occurring in the sequential reductive/oxidative BES operated under various feeding conditions. Moreover, the main biomarkers of reductive (D. mccartyi 16S rRNA and tceA, bvcA, vcrA genes) and oxidative (etnE and etnC genes) dechlorination have been monitored and quantified. Possible metabolic mechanisms occurring in the BES have been discussed.

In this study, the microbiome established in a BES for sequential anaerobic/aerobic PCE dechlorination was characterized under various feeding conditions and at increasing level of complexity in terms of anion content and possible effects on redox reactions. Together with microbial characterization of the effluent of the reductive and oxidative compartments, the abundances of the biomarker genes involved in reductive (D. mccartyi and reductive dehalogenase genes) and oxidative (etnE and etnC) dechlorination were also estimated under all the tested conditions.

The various feeds affected the reactor performances and microbiome compositions.

The use of the MM feed showed the best kinetic performances as expected having the optimum composition for cultivation of anaerobic dechlorinating cultures. Indeed, the MM feed was also used for start-up of the reactor and acclimation of the bacterial community in the BES (Zeppilli et al., 2019, 2021), ensuring high PCE removal efficiencies and avoiding the occurrence of competitive reactions such as sulfate or nitrate reduction. Diversely, the use of SG, where sulfate and nitrate were added, caused a slight decrement in PCE removal efficiency introducing sulfate and nitrate reduction processes. Nevertheless, as highlighted in a previous study, the presence of sulfate and nitrate reduction reactions in the cathodic compartment of the BES did not involve a competition for the reducing power with the RD reaction (Dell’Armi et al., 2021a). In line with kinetic performances, in the reductive compartment of the BES fed with MM or SG, D. mccartyi was abundant (2.5E + 07 and 1.8E + 07 16S rRNA gene copies/L respectively under MM and SG feeding) and was mainly represented by tceA-carrying strains. D. mccartyi was responsible for the biological RD occurring in the reductive compartment, likely via an indirect electron transfer mechanism, but this microorganism has not been considered an electroactive bacterium until now (Matturro et al., 2021). Regarding microbiome composition, Epsilonbacteraeota members were dominant (50.8 and 89.5% under MM and SG, respectively) and mostly affiliated to one ASV that showed 99.5% of identity to Sulfuricurvum kujiensestrain YK-3 (GenBank: AB080644) (Kodama and Watanabe, 2003). The Sulfuricurvum species is frequently found in hydrocarbon-contaminated environments, and it is a facultative anaerobic, chemolithoautotrophic, sulfur-oxidizing, and nitrate-reducing bacterium (Kodama and Watanabe, 2004; Han et al., 2012). It can be grown anaerobically using sulfide, elemental sulfur, thiosulfate, and hydrogen as electron donors and nitrate as an electron acceptor under anaerobic conditions (Kodama and Watanabe, 2004). The system configuration might explain the role of the Sulfuricurvum species found in the reactor. Indeed, as mentioned in previous studies, the unrecovered current of the system is likely associated with parasitic reductive/oxidative reactions occurring between the working electrode (cathode) and the inner counter electrode (anode) that promotes electron loops (Zeppilli et al., 2019). Therefore, some ionic species can be exchanged between different parts of the reactor, establishing simultaneous reductive and oxidative reactions (Zeppilli et al., 2021). Thus, hydrogen sulfide, already present in the MM feeding solution or obtained by sulfate reduction occurring during SG feeding, could be bioelectrochemically oxidized into elemental sulfur and then oxidized by the Sulfuricurvum species by nitrate reduction (Supplementary Figure 2). However, in order to confirm the role of the Sulfuricurvum species found in the BES and its metabolic functions and interactions, additional whole metagenomic analyses are necessary. Moreover, the occurrence of methylotrophs and methanotrophs microorganisms under MM or SG feeding is in line with the methane production observed in the reactor. However, defining separately the anaerobic or aerobic metabolic pathways occurring in each chamber of the BES is challenging because of the presence of a counter electrode in each reactor of the BES that allow the establishment of various redox niches and, consequently, co-existence of oxidizing and reducing microorganisms in the chamber that are not separable. Indeed, sampling for biomolecular analysis was possible on the outlet of each reactor and obviously included all microorganisms present in the bulk.

The RG feeding caused a change in the microbiome composition of the reductive compartment involving higher biodiversity and decrement in D. mccartyi abundance (8.78E + 06 16S rRNA gene copies/L, decreased by one order of magnitude) compared to the previous feeding conditions. Organisms affiliated with Clostridiaceae, rich in the original community of the aquifer, decreased considerably their relative abundance in the reductive reactor at the end of the RG operation.

On the other hand, in the Alphaproteobacteria group, Beijerinckiaceae was the most abundant family with RG feeding, mainly followed by Gemmataceae, Pirellulaceae, and Mycobacteriaceae. The presence of low-chlorinated intermediates in RG may have stimulated the selection of this community in the cathodic chamber. The methanotrophic bacterial group of Beijerinckiaceae included the genera Methylocystis, alphaI cluster, and Bosea. Interestingly, the Methylocystis genus comprises aerobic, obligate, and facultative methylotrophic and methanotrophic bacteria capable of utilizing methane and methanol as sole carbon and energy sources (Jung et al., 2020). Members of this genus inhabit different soils, peatlands, and freshwater sediments, and they have been previously shown to be capable of degrading TCE and cis-DCE via aerobic metabolic or co-metabolic pathways (Uchiyama et al., 1989; Nakajima et al., 1992; Dolinová et al., 2017). Pirellula-like and Gemmataceae are chemoorganotrophic and ubiquitous aquatic bacteria are often detected in anoxic or micro-oxic habitats despite taxonomy describing representatives of these bacteria as strictly aerobic (Fuerst, 1995; Kulichevskaya et al., 2020). Conversely, diverse strains of Mycobacterium sp. have been previously associated with aerobic ethyleneotrophs considering their ability to grow on VC as the sole carbon and energy source (Mattes et al., 2010). These bacteria are indigenous and widely distributed in chlorinated-ethylene-contaminated sites adapted to low oxygen level (Coleman et al., 2002). Moreover, the presence of the Rhodobacter genus, known for its excellent iron reduction ability, may explain the iron reduction reaction, a characteristic of this condition (Weber et al., 2006).

Regarding the oxidative reactor, the MM community showed abundance of the Leptospiraceae, Rhodobacteraceae, and Rhodocyclaceae families. The presence of the Rhodocyclaceae family is of great interest, because its members are commonly found in polluted aquifers, which are involved in denitrification processes and in degradation of a wide range of aromatic compounds (Oren, 2014). In addition, the genera Mycobacterium and Methylocystis were present in this condition, in accordance with the high cis-DCE and VC removal efficiency achieved.

Conversely, the aerobic community with SG feeding included members of the Propionibacteriaceae family, uncultured Saccharimonadales and Sulfurospirillum, and the more abundant genus Sulfuricurvum. Members of the Propionibacteriaceae family have been isolated from a broad range of different habitats including chlorosolvent-contaminated soils and groundwater, but they deserve more investigations on their biodegradation potential (Bae et al., 2006; Stackebrandt, 2014). Likewise, very little is known about the ecology of the Saccharimonadales group despite Saccharibacteria having been shown to possess very small genomes and cell sizes, and it has therefore been proposed to live in symbiosis with other microorganisms depending on co-metabolism (Lemos et al., 2019). The Sulfurospirillum genus instead, consists of versatile, often microaerophilic bacteria growing with many different growth substrates. Few strains are able to use halogenated compounds as electron acceptors. The wide range of possible growth substrates enables many Sulfurospirillum species to thrive in polluted habitats, which is reflected by the presence of these bacteria in many contaminated sites (Goris and Diekert, 2016). This versatile metabolic capability suggests that it might be involved in the global dechlorination capability of the microbiome.

When the oxidative reactor was fed with RG, further enrichment of the Sulfurospirillum genus (38%), Rhodocyclaceae family members, Mycobacterium, and Methylocystis was obtained. All these microorganisms having good degradative capacities may have contributed to the removal of chlorinated compounds, in accordance with the excellent VC and cis-DCE removal efficiency detected under this feeding condition (Table 5). Besides, the presence and abundance of VC-oxidizing bacteria were further confirmed by the absolute quantification of functional genes involved in the oxidative dechlorination pathway, which increased by two orders of magnitude (up to 2.38E + 08 gene copies/L) in this feeding condition compared to the previous ones.

Overall, our findings are in line with various studies conducted on BESs applied to contaminated groundwater that highlighted the presence of complex interactions in the sulfur cycle during the electrobioremediation process (Rakoczy et al., 2013; Daghio et al., 2016, 2017). In bioelectrochemical treatments, in fact, many substances may act as electron acceptors besides the anode, such as sulfate, nitrate, and iron(III) (Zhang et al., 2010). They capture electrons during the substrate degradation process and participate in the cycles of nitrogen, sulfur, and iron. For instance, sulfate may be reduced to sulfide, and sulfide can be oxidized to elemental sulfur through biological or chemical processes. Sulfur may then be reduced again to sulfide, which can be re-oxidized, thus constituting a sulfur cycle that promotes current generation as reported in the literature (Dutta et al., 2009). Thus, sulfate/sulfide can serve as an electron shuttle to deliver electrons from microorganisms to the electrode and vice versa (Zhao et al., 2006; Dutta et al., 2009; Rakoczy et al., 2013).

This study highlighted that the feeding solutions affected the microbial composition of the BES and the rate of its dechlorination activities. The peculiarity of the BES configuration allowed for the simultaneous presence of diverse redox niches in each compartment allowing for complete PCE and by-product dechlorination. The complex interactions between these degrading microorganisms ensured excellent system performance, even with the RG. Despite sulfate, nitrate, and iron reduction reactions having not directly competed for reducing power with the RD process, the abundance of these anions promoted in the reductive reactor the predominance of some microbial groups linked with the sulfur cycle, especially members of the Sulfuricurvum genus. The microbiome in the oxidative compartment included both ethyleneotroph and methanotroph populations such as Mycobacterium and Methylocystis. They are able to achieve metabolic and co-metabolic degradation of low-chlorinated products, effectively preventing VC accumulation in the system. Overall, the sequential anaerobic/aerobic bioelectrochemical system fed with RG showed a great dechlorination potential, leading to the complete mineralization of chlorinated pollutants, with relevant implications for system optimizations on a field scale for sustainable and cost-effective bioremediation.

The data presented in this study are deposited in the GenBank repository, accession number: PRJNA839891 (BioSample accession numbers: SAMN28557297–SAMN28557302).

MD, BM, MZ, and ED: investigation, validation, and writing of the original draft. SC: investigation and validation. MM, MP, and SR: supervision and funding acquisition. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.