Diversity and distribution of iron-oxidising bacteria belonging to Gallionellaceae in different sites of a hydroelectric power plant

Abstract

Iron (Fe) is the fourth most abundant element on the planet, and iron-oxidising bacteria (FeOB) play an important role in the biogeochemical cycle of this metal in nature. FeOB stands out as Fe oxidisers in microaerophilic environments, and new members of this group have been increasingly discussed in the literature, even though their isolation can still be challenging. Among these bacteria is the Gallionellaceae family, mainly composed of neutrophilic FeOB, highlighting Gallionella ferruginea, and nitrite-oxidiser genera. In the previous metagenomic study of the biofilm and sediments of the cooling system from the Irapé hydroelectric power plant (HPP-Irapé), 5% of the total bacteria sequences were related to Gallionellaceae, being 99% unclassified at genus level. Thus, in the present study, a phylogenetic tree based on this family was constructed, in order to search for shared and unique Gallionellaceae signatures in a deep phylogenetic level affiliation and correlated them with geomorphologic characteristics. The results revealed that Gallionella and Ferrigenium were ubiquitous reflecting their ability to adapt to various locations in the power plant. The cave was considered a hotspot for neutrophilic FeOB since it harboured most of the Gallionellaceae diversity. Microscopic biosignatures were detected only in the CS1 sample, which presented abundance of the stalk-forming Ferriphaselus and of the sheath-forming Crenothrix. Further studies are required to provide more detailed insights on Gallionellaceae distribution and diversity patterns in hydroelectric power plants, particularly its biotechnological potential in this industry.

Supplementary Information

The online version contains supplementary material available at 10.1007/s42770-024-01245-w.

Introduction

Lithotrophic iron-oxidising bacteria (FeOB) are major players in the iron (Fe) biogeochemical cycle, with the ability to catalyse the dissimilatory oxidation of ferrous to ferric iron [1]. Currently, FeOB are subdivided into four physiological groups: (1) acidophilic, aerobic FeOB; (2) neutrophilic, aerobic FeOB; (3) neutrophilic, nitrate-dependent anaerobic FeOB; and (4) anaerobic, photosynthetic FeOB [2]. Acidophilic FeOB, such as Acidithiobacillus ferrooxidans, have been exhaustively studied due to their potential in the remediation of acidic mines [3, 4], whereas the role of neutral FeOB has only been documented in the last 20 years [5, 6]. The main reasons for these differences were (i) the rapid abiotic oxidation of Fe in neutral and oxygenated habitats that might not be competitive with biotic condition, resulting in fewer studies of this group [7], and (ii) the challenge of isolating pure cultures, as they live in an opposing gradient of Fe(II) and oxygen [6]. However, it has since been proven that the oxidation process performed by microaerophilic bacteria can compete with the abiotic Fe(II), reaching up to 60% of the total Fe oxidation in neutral habitats [8]. Furthermore, molecular analysis has proven to be useful in the identification of new species in this group, as well as in the comparative studies, emphasising their ecology and genomic repertoire [1, 9–11]. Stalk, dread, and sheath morphologies can also confidently be associated with FeOB metabolism and have been described as extracellular biosignatures [12, 13].

A part of this group, Gallionellaceae-related FeOB are abundant in freshwater environments and are often retrieved from redox transition boundaries, where they benefit from both high Fe concentrations and microaerophilic conditions [6, 7, 9]. This clade contains Gallionella ferruginea, which is well studied and broadly distributed [10, 14], and Ferriphaselus [15], both stalk-forming bacteria that prevents them from encrusting with iron oxides [15, 16]. Moreover, other members of Gallionellaceae family are Sideroxydans, able to oxidise iron in humic acid-enriched cultures [17, 18], and Ferrigenium kumadai, isolated from paddy field soil by Khalifa et al. [2]. This family also harbours members that are not FeOB such as Candidatus Nitrotoga arctica and Candidatus Nitrotoga fabula, which are nitrite oxidisers [19–21].

Previously, a metagenomic study [22] is aimed to characterise microbial communities of the cooling internal system of Irapé hydroelectric power plant (Minas Gerais, Brazil), especially particular taxonomic signatures of biofilm formation in heat exchangers, associated with biofouling formation. As the region in the previous study is iron- and sulphur-rich, it can be considered a hotspot for FeOB survey. Moreover, among the obtained results, the authors unveiled Gallionellaceae OTU abundance of 5% from the total of bacteria being 99% unclassified at genus level, which reflected the high number of uncultured Gallionellaceae-related members [22]. Thus, in the current study, a phylogenetic tree was proposed in order to complement the metagenomic data on FeOB members. Finally, morphological FeOB biosignatures were explored in the samples, using scanning electron microscopy (SEM), to find other FeOB traces in this power plant.

Material and methods

Study area

The Irapé hydroelectric power plant is located in the north of Minas Gerais state, Brazil (HPP-Irapé; 16° 44′ 15″ S, 42° 34′ 30″ W) a region rich in iron and sulphur (Figure S1). HPP-Irapé belongs to Companhia Energética de Minas Gerais (CEMIG) and is considered the tallest dam in the country, 208 m high with a flooded area of 137.16 km² located in the course of the Jequitinhonha River, generating 399 MW. Four sites were sampled: two of them were located outside of the power plant (rock sediment (RS) and artificial cave sediment (CS)) (reads from both locations were deposited in the Sequence Read Archive (SRA)-NCBI under submission number PRJNA643245) and other two were taken from part of the cooling water system, previously characterised by metagenomics (submission number on SRA-NCBI: PRJNA628802) [22] (filter sediment (FS) and heat exchanger microfouling (HEM)). Two sampling campaigns were performed first in September 2015 (campaign 1) and again in February 2016 (campaign 2). The HEM samples were taken at two different times after the last clearance: HEM1_30 (30 days) and HEM2_15 (15 days). Samples were transported to the laboratory on ice and stored at − 20 °C until further processing.

Total DNA isolation extraction followed by the sequencing of the V4 variable region of 16S rRNA gene from RS1, RS2, CS1, and CS2 was performed together with the other power plant samples described in Reis et al. [22]. E.Z.N.A.® Soil DNA Kit (OMEGA Bio-Tek, Norcross, GA, USA) and the MiSeq platform (Illumina, Inc., San Diego, CA, USA) were used for the total DNA isolation and sequencing, respectively. Finally, all eight samples were geomorphologically evaluated by scanning electron microscopy (SEM) and energy-dispersive X-ray spectroscopy (EDS).

Bioinformatic analysis

16S amplicon reconstruction from next-generation sequencing (NGS) data

Quality evaluation was performed on the NGS sequence data from the metagenomic samples using MultiQC (v1.10.1) [23]. BBDuk [24] was used to trim low-quality edges and adapter sequences. An amplicon sequence variant table was generated for the 16S variants on the metagenomic samples using the DADA2 package (v1.20) [25] from the Bioconductor repository for the R platform (v4.1.0). Amplicons were taxonomically classified by the DADA2 package using the Silva reference database (v138.1) [26]. Amplicons classified as pertaining to the Gallionellaceae family were exported for phylogenetic tree construction.

Phylogenetic tree construction

Reference rRNA 16S gene sequences of 17 members of the Gallionellaceae family were obtained from the Silva Species Assignment training dataset maintained by the DADA2 package, alongside 1 reference sequence of Methylobacillus pratensis, which served as an outside group. A 16S sequence of Ferrigenium kumadai An22 was obtained from the EBI/ENA database (Entry LC065124). The evolutionary history of the 13 metagenomic sequences alongside the 19 reference sequences was inferred using the UPGMA method [27]. The evolutionary distances were computed using the Kimura 2-parameter method [28], and the rate variation among the sites was modelled with a gamma distribution (shape parameter = 1). All ambiguous positions were removed for each sequence pair (pairwise deletion option). There was a total of 1544 positions in the final dataset. Evolutionary analyses were conducted in MEGA X [29].

Geomorphological analysis of sediments and microfouling samples from HPP-Irapé

Scanning electron microscopy (SEM)

Sediments and microfouling were fixed in glutaraldehyde 2.5% and dehydrated in ethanol series ranging from 70 to 100%. Some of the CS samples were treated with bleach (NaClO 2–5%) and others in a muffle furnace at 800 °C. Aluminium stubs were mounted with 10–20 µL of all samples (RS, CS, FS, and HEM) in ethanol 100% on carbon adhesive tape. Further, samples were coated with a blend of gold–palladium (Quorum coater sc7620) in an argon atmosphere for 30–60 s. SEM sessions were performed in a high vacuum atmosphere with LM-VEGA3-TESCAN SEM. Images and biomineral measurements were acquired and made after settling adequate working distance, beam intensities, magnification, and astigmatism, as well as column and centralization of the electronic cannon, when it was necessary. The software used for those was the Vega3 Control version (version 4.2.4.0).

Energy-dispersive X-ray spectroscopy (EDS)

The biomineral chemical analysis was characterised by EDS; detected with PentaFET® Precision, X-act; and processed on Aztec 2.0 software (Oxford Instruments). The analyses were performed using the Map and Point & ID modes. Most of the analyses were performed with 20 kV, BI = 16, as they generated a greater signal in the EDS and a dead time of around 15–20%. The elements were pre-defined in the periodic table available in the software which also completed the Auto ID. Elements were first identified by Auto ID and then fixed for further analyses.

Results and discussion

FeOB oxidises iron as a result of their metabolism, which means that the members of this diverse group are able to use the electrons captured from this oxidation as their sole source of energy for growth [30]. In attempt to characterise the distribution and diversity of FeOB, mainly Gallionellaceae-related members recovered from sediments and microfouling samples in HPP-Irapé, the present analysis searched for shared and unique Gallionellaceae signatures in a deep phylogenetic level affiliation and correlated them with geomorphologic characteristics in those samples.

Figure 1 and table S1 represent the FeOB described in the literature until now, highlighting those previously found in the metagenomes of different samples from HPP-Irapé, as well as their abundance in each sample. In RS2 prevailed Acidithiobacillus (aerobic), Ferritrophicum (microaerophilic), and Leptospirillum (aerobic), which are acidophilic bacteria able to oxidise iron containing sulphide minerals, such as pyrite [31–33]. The abundance of these species is expected since the principal compounds that characterise soils with low pH are those containing iron and sulphur, commonly found in the sampling region. The Gallionellaceae genus Ferriphaselus predominated in CS1 and Acidiferrobacter in CS2. Hyphomicrobium was abundant in CS1, CS2, and FS1 and passively oxidises Fe²⁺ and precipitates FeO(OH) [34], being also associated with Mn oxidation [35]. Meiothermus (mesophilic) and Thermomonas (thermophilic) genera harbour iron-oxidising species [7, 36] and were significantly found in HEM2_15 being considered putative pioneering colonisers in microfouling of the heat exchanger [22]. Sediminibacterium was also abundant in HEM2_15, and it was revealed by Wang et al. [37] as a corrosion-inducing bacterium.

Fig. 1.

Abundance of iron-oxidising bacteria (FeOB) distributed at Irapé hydroelectric power plant (HPP-Irapé). The colour intensity represents the quantity of recovered sequences of FeOB genera obtained from metagenomes of sampling sites. RS, rock sediment; CS, cave sediment; FS, filter sediment; HEM, heat exchanger microfouling. Numbers 1 and 2 after the acronyms represent the sampling campaigns

Indeed, the construction of a phylogenetic tree to predict taxonomic affiliation from clustered sequences is an approach already recommended by other studies as an efficient tool to complement the metagenomic findings [38–40]. The five described genera affiliated to Gallionellaceae are represented in the phylogenetic tree, Gallionella, Ferriphaselus, Sideroxydans, Ferrigenium, and Candidatus Nitrotoga, being four of them FeOB and one a nitrite-oxidising bacterium (Ca. Nitrotoga) (Figs. 1 and 2).

Fig. 2.

Phylogenetic tree showing recovered sequences at Irapé hydroelectric power plant (HPP-Irapé) with Gallionellaceae-related genera. RS, rock sediment; CS, cave sediment; FS, filter sediment; HEM, heat exchanger microfouling. Numbers 1 and 2 after the acronyms represent the sampling campaigns

According to the phylogenetic tree, Ferrigenium was the most abundant and ubiquitous genus in the power plant, being found in CS1, FS1, FS2, and HEM2_15 and exclusive in FS1, FS2, and HEM2_15 samples, while Sideroxydans, Ferriphaselus, and Ca. Nitrotoga were found in CS1 (Fig. 2 and Table S1). Otherwise, Gallionellaceae sequences recovered from metagenomes analyses identified Gallionella as the most ubiquitous genus, being found in CS1, FS2, HEM1_30, and HEM2_15. Possibly, the ability of Gallionella to deal successfully in environments with the presence of potentially toxic metals and metalloids, due to its resistance gene repertoire [1, 10], favoured its ubiquity in the power plant. However, Ferrigenium was characterised by not utilising hydrogen, thiosulfate, sulphide, nitrite, Mn(II), glucose, acetate, pyruvate, or citrate as an energy source [2], presenting a lifestyle more restrictive than Gallionella, despite being also ubiquitous in this environment.

Microscopic analysis detected the presence of stalks and sheaths, which are biosignatures of Betaproteobacteria occurrence, only in the sediments from CS1 (Fig. 3). As reported by Chan et al. [12, 41], stalks are carboxyl-rich polysaccharide structures that may prevent cell encrustation with iron oxides. Otherwise, sheaths are described as an empty tube, initially made of a complex of a cysteine-rich peptide and a polysaccharide composed of uronic acids and galactosamine [42], which is followed by a later deposition of iron and manganese previously oxidised by the cells [43]. This empty tube surrounds a chain of cells enabling these bacteria to attach to solid surfaces where it cannot be consumed by protozoa or attacked by predators such as Bdellovibrio bacteriovorus, and it can also work as a physical barrier against other microorganisms [43]. These biosignatures may also contribute to the maintenance and structuring of bacterial biofilms, since as reported by Kato et al. [13], stalks can work like an anchor to attach to solid structures preventing the cell from being disconnected from it. Surprisingly, although stalk-forming bacteria [15, 32] were also detected in HEM (Gallionella and Ferriphaselus) (Fig. 3A–C), stalks were not found in the heat exchanger samples, where microbial slime (biofilm and sediment) was formed (Fig. 3J) [22].

Fig. 3.

Scanning electron microscopy (SEM) of cave sediment (CS). A, B Sheaths (yellow arrows) and stalks (pink arrowhead) found in CS. C Zoom highlighting stalk in the centre of the micrograph. D, E Zoom highlighting sheaths on micrographs. F, G SEM of sheaths impregnated with sediment. H Sheath extension measurement in SEM. I Sheath diameter measurements. J Sediment of heat exchanger without stalks and sheaths

Sheath-forming bacteria are often associated with Leptothrix [44] despite this not being the unique genus that produces it. Sphaerotilus, Crenothrix polyspora, Calothrix, and Lieskeella bifida have also been described as iron-oxidising sheath-forming bacteria [45, 46]. From these bacteria, only Crenothrix was recovered by in the previous metagenomic survey [22]. Indeed, it is not possible to identify the producer of the sheaths based on the microscopy (Fig. 3B, D, E–I). However, in spite of these, two hypotheses were formed: (1) the sheaths were produced by Crenothrix recovered in CS1 and (2) since these structures are recognised as microfossils due to their persistence in the environment and thermostability (600–1100 °C) [12, 47, 48], they could be produced by other sheath-forming bacteria that are no longer present in CS, such as Leptothrix.

Otherwise, Ca. Nitrotoga is related to nitrite oxidation, an important process for removing nitrogen (N) from wastewater [49]. Candidatus Nitrotoga fabula, the first genus isolated from activated sludge, previously considered an environment entirely dominated by Nitrospira, develops at higher temperatures (20 °C; optimum, 24 to 28 °C) and adapts to low-oxygen or potentially anoxic conditions when compared to Ca. Nitrotoga arctica, which are cold-adapted nitrite oxidisers isolated from permafrost and that grow optimally at temperatures as low as 10 °C [19, 20, 50, 51]. Thus, Ca. Nitrotoga supports psychrophilic and mesophilic conditions and is widespread in engineered and natural ecosystems [20, 21]. Besides, it has been demonstrated that Sideroxydans is able to occupy niches depleted in N, due to its three clusters of nif genes [1, 10]. Thus, the presence of both bacteria in CS1 could be related to the metabolism of nitrogen in this environment, with the first removing nitrogen by nitrite oxidation and the second benefiting itself from the nitrogen-lacking environment.

HPP-Irapé is located in an iron- and sulphur-rich region, probably favouring the abundance of bacteria with a key role in these biogeochemical cycles. In particular, CS and RS samples presented large amounts of sulphur (Fig. S1). Kato et al. [13] unveiled genes related to sulphur oxidation in strains of Ferriphaselus, reinforcing the massive occurrence of this genus in the cave.

In conclusion, our data revealed the distribution and diversity of Gallionellaceae-related sequences, previously recovered from a metagenomic survey and also analysed herein, by a phylogenetic tree construction. Differences between the Gallionellaceae classification were observed from both analyses, allowing us to deepen the discussion about its distribution. After the phylogenetic analysis was detected a Gallionellaceae diversity more complex than those presented by Reis et al. (2021). The ubiquity of Gallionella and Ferrigenium reinforced their plasticity to adapt to different locations of the power plant. Besides, Gallionella, Ferriphaselus, Sideroxydans, and Ca. Nitrotoga were also found in the cave, indicating that this sample harbours more Gallionellaceae diversity, being a hotspot for neutrophilic FeOB.

Moreover, microscopic biosignatures such as stalks and sheaths were found only in CS1, Ferriphaselus possibly being responsible for producing the stalks and Crenothrix likely having produced the sheaths. This is the most likely explanation as pieces of evidence of both of these bacteria were found to have previously inhabited the cave. Further studies are required to provide more insights on Gallionellaceae distribution and diversity patterns in hydroelectric power plants.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contribution

Conceptualization and methodology: Rayan Silva de Paula, Clara Carvalho e Souza, and Mariana de Paula Reis; formal analysis and investigation: Rayan Silva de Paula, Clara Carvalho e Souza, Carlos Alberto Xavier Gonçalves, Marcelo Victor de Holanda Moura, Anna Carolina Paganini Guañabens, Gabriela Rabelo Andrade, and Mariana de Paula Reis; writing—original draught preparation: Rayan Silva de Paula and Clara Carvalho e Souza; writing—review and editing: Andréa Maria Amaral Nascimento, Antonio Valadão Cardoso, Mariana de Paula Reis, and Erika Cristina Jorge; funding acquisition and resources: Antonio Valadão Cardoso; supervision: Mariana de Paula Reis and Erika Cristina Jorge.

Funding

Erika C. Jorge received a scholarship from the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq). This work was supported by the Companhia Energética de Minas Gerais (CEMIG) R&Ds ANEEL GT-0604.

Data availability

The datasets generated during the current study are available in the Sequence Read Archive (SRA)-NCBI under submission number PRJNA643245, and the other two datasets were previously characterised by metagenomics published by Reis et al. [22] (submission number on SRA-NCBI: PRJNA628802).

Declarations

Conflict of interest

The authors declare no competing interests.