Enrichment of anaerobic and facultative anaerobic bacteria from a graphite mine sample on graphene oxide

Graphene oxide (GO) is progressively synthesized and applied in various industrial fields, such as healthcare, medicine, gas transport, electric power industry, etc. Thus, its distribution in the environment increases, which leads to changes in various microbial communities. Mechanisms of interaction between bacterial communities and novel carbon-based nanomaterials, such as GO, are still to be elucidated. We developed stable consortia in order to identify bacteria with the ability to thrive in the presence of GO. Here, we show the results of metagenomic analyses of the graphite mine sample, a consortium with GO as a sole source of carbon and a consortium with fumarate as an additional carbon source. The aim of this study was to establish anaerobic enrichment cultures from a graphite mine sample and GO, and to identify their diversity/community composition.


Introduction
GO is a two-dimensional nanomaterial, comprised of a single layer of sp 2 -hybridized carbon atoms in a honeycomb structure and oxygen-containing functional groups, such as carboxyl, carbonyl, epoxide, hydroxyl and ketone [1].It has great tensile strength, electric and thermal conductivity, which decrease as the degree of oxidation increases [2].These properties make it applicable in many industrial fields, which undoubtedly leads to its distribution in the environment, exerting an unclear impact on organisms [3].Studies show that nanocomposites with GO have good biocompatibility with different cell lines [4].However, GO alone shows in vitro, in vivo and in ovo cytotoxicity, the grade of which depends on its concentration and particle size [4].GO's effects on pure bacterial cultures and bacterial communities depend mainly on its concentration and size particles.In 2015 and 2018 the first studies over the influence of GO on natural environmental bacterial communities have been published [5,6].In a recent review article were summarized the effects of graphene-based materials, such as GO and microbial communities in controlled environments, like anaerobic digesters, activated sludges and soil [3].It was shown that GO's effect on alpha diversity was inconclusive, as it was often transitory or inconsistent.Beta diversity was also shown to be affected, although there was no specific pattern observed [3].
So far, concerning the degradation of GO, was proposed that it can be degraded to oxidized derivatives of polycyclic aromatic hydrocarbons (PAHs) during a photo-Fenton reaction [7].In addition, it was found that bacterial enzymes, such as manganese peroxidase and lignin peroxidase might be able to degrade GO by attacking the oxygen-containing groups, resulting in graphene quantum dots, further degraded to PAHs [8].
PAHs and GO are similar in structure, given the shared conformation of the carbon atoms that form molecules, made up of many benzene rings.PAHs, in turn, can be degraded from anaerobic bacteria by several known and proposed mechanisms such as methylation, carboxylation and hydroxylation [9].Given the structural similarity between PAHs and GO, these mechanisms could be a stable basis for the nature of the bacterial interaction with GO.The aim of this study was to create enrichment cultures of anaerobic bacterial communities, which grow in the presence of GO and identify the OTUs they are composed of.

Materials and methods
The analyzed mine sample (S5) was obtained from Chundu Island in Africa.The sample was taken using the conventional method, put in plastic bags, transported by flight, grinded with a sterilized mortar and pestle under aseptic conditions and kept in a fridge.The landowner provided permission for the field study conducted on private land.

Properties of the GO dispersion
Graphene oxide water dispersion (1 wt%) was obtained from Graphenea Inc. (Spain, EU) with monolayer content over 95% and particles less than 10 µm in size.

Culture medium
Fresh water medium (SW) [10], supplemented with one of the three trace element solutions -SL9 [11], SL10 [12] or SL12 [13] was used for the preparation of enrichment cultures.Resazurin was added to monitor the absence of oxygen in the medium.After autoclaving, the medium was saturated with N 2 /CO 2 (90:10, v/v) gas mixture and was supplemented with selenite-tungstate solution, and 7 vitamins solution.The medium was dispensed in 100 mL infusion bottles, sealed with rubber septa under the same gas head space.

Enrichment cultures and consortia
Two types of enrichment cultures were prepared, according to the final concentration of GO, containing: 0.01 g.L -1 and 0.001 g.L -1 GO, 0.1 g of the mine sample in 10 mL infusion bottles with N 2 /CO 2 (90:10, v/v) head space.The work set-up is schematically presented in Figure 1.
Enrichment cultures with both concentrations of GO were prepared with 20 mM nitrate as a final electron acceptor.Half of them were supplemented with 10 mM fumarate.All enrichment cultures were cultivated at 28°C for 42 days.For all types of enrichment cultures were prepared positive and negative controls, where in the positive controls was added 0.1 g of the mine sample and in the negative controls the sample was omitted.Both controls were autoclaved at 121°C for 30 min.The most well growing enrichment cultures were chosen for downstream applications and studies.After 2 passages it was suggested to have obtained stable consortia growing under the given conditions.Enrichment cultures with 0.01 g.L -1 GO and obtained from them bacterial consortia were named SWNS5.1,SWNS5.2,SWNS5.3,SWNFS5.4,SWNFS5.5 and SWNFS5.6 (Table 1).The presence of GO particles interfered with bacteria, when growth was asserted by measurements of the optical density of the cultures, therefore to prove the growth of bacteria it was decided to use microscopy methods.

Microscopy methods
Scanning spectrophotometry was used to check the optimal absorbance of the cultures, however, it gave ambivalent results.Thus, to prove the growth of the cultures and the presence of bacteria, we decided to apply microscopy methods.Light microscopy was implemented to monitor the bacterial growth, motility, and morphology in the enriched cultures and bacterial consortia at the 21 st and 42 nd day of cultivation.Additionally, Gram staining was done for the selected most well grown stable consortia.The staining procedure was based on the ASM Gram Stain Protocol [14], where instead of 95% ethanol, 96% was used.Leica DM500 microscope with an oil immersion objective Plan 100×/1,25 for routine observations was used.The pictures were taken with camera SM-M127F.Furthermore, scanning electron microscopy (SEM) was applied to obtain bacterial morphology and further confirm the bacterial growth.Samples were prepared by transferring 2 mL of each culture into a microcentrifuge tube, followed by centrifuging for 10 minutes at 10 000 rpm, at room temperature.This procedure was done in duplicate.The obtained cell suspension of 500 μL was washed in 0.1% cacodylate buffer (CB) and fixed for 2 hours at 4°C in 4% glutaraldehyde in CB.After threefold washing in CB, the samples were treated with 1% OsO 4 for 1 hour at 4°C and dehydrated by immersing in solutions with increasing ethanol concentrations from 40% to 100%.So prepared cells were then coated with gold in a vacuum evaporator (Edwards).The scanning electron microscopy was performed with JEOL JSM-5510.

DNA extraction and metagenomic libraries
Bacterial consortia were cultivated at the same conditions and nutrient media, as the enrichment cultures.They were established after two consecutive passages of the initial enrichment cultures under the same conditions.To obtain enough biomass for DNA extraction, best growing bacterial consortia were cultivated in 50 mL bottles.Cells were pelleted by centrifugation at 9000 rpm for 20 minutes at room temperature and the supernatants were discarded.Total DNA from the mine sample and the two consortia with and without fumarate was obtained with the kit Genematrix Soil DNA Purification Kit (EurX, Poland) as per manufacturer`s instructions.Two fractions of 0.5 g and 1.0 g from the original mine sample were proceeded simultaneously.Each preparation was eluted in 120 µL elution buffer.DNA from the consortia was eluted in 100 µL elution buffer each.DNA concentrations were measured in triplicate for each sample with Quawell UV Spectrophotometer Q3000.
Metagenomic libraries were performed by amplification of the V 3 -V 4 region of the 16S rRNA gene at Macrogen Inc. (Korea).The amplification was done with LightCycle qPCR using the degenerate primer pair Bact_341F (5'-CCT ACG GGN GGC WGC AG -3') and Bact_805R (5'-GAC TAC HVG GGT ATC TAA TCC -3') [15].The length of the amplicons was proved with TapeStation D1000 Screen Tape and their concentration was measured by using the Picogreen quantitation reagent.The amplicons were sequenced by the Illumina MiSeq Platform.

Bioinformatics and identification of OTUs
The obtained nucleotide sequences of operational taxonomic units (OTUs) were compared in BLAST (Basic Local Alignment Search Tool, National Center for Biotechnology Information, USA) and RDP (Ribosomal Database Project) to identify the microorganisms presented in mine sample and the consortia.OTUs from the mine sample and the consortia were further compared.

Enrichment cultures
Enrichment cultures with concentration of 0.01 g.L -1 GO were cultivated with NO 3 -as a final electron acceptor to enable the growth of facultative anaerobic bacteria.Based on studies, proposing fumarate addition as an activation step for the anaerobic bacterial degradation of the benzene ring and PAHs after methylation of the respective molecule [16,17], we chose to supplement half of the enrichments with sodium fumarate (Figure 1, Table 1).All of the enrichment cultures with concentration of 0.01 g.L -1 GO IOP Publishing doi:10.1088/1755-1315/1305/1/0120144 showed better growth after 2 passages, hence the formation of stable bacterial consortia.Out of them enrichment cultures SWNS5.2 (or SW5.2) and SWNFS5.5 (or SW5.5) were chosen as most well growing.On figure 1 is shown the work set-up and in Table 1 are described the types of enrichment cultures grown on 0.01 g.L -1 GO. Figure 1.A schematic representation of the enrichments and consortia, the conditions they were cultivated at, and the methods used to confirm bacterial growth.

Table 1.
Characteristics of the enriched cultures, grown on 0.01 g.L -1 GO (SW5.1 to SW5.3), and 0.01 g.L -1 GO, with 10 mM fumarate (SW5.4 to SW5.6) based on light microscopy observations.The number of '+' signs corresponds to the rate of observable bacterial growth.The underlined cultures were chosen for downstream applications and studies.Microscopic observations without staining revealed abundant bacterial growth in 5 of the 6 cultures.In only one culture -SW5.2, grown on GO without fumarate, were observed motile bacteria, instead of two cultures grown on GO, supplemented with 10 mM fumarate, as shown in Table 1.The abundant bacterial growth in the enrichments, supplemented with sodium fumarate, may correlate with the studies showing activation of PAHs degradation with fumarate [16,17].All enrichment cultures were cultivated on Fresh water medium, containing either SL9, SL10 or SL12, and supplemented with vitamins as described above.Visible growth, respectively high turbidity, was observed on the 17 th day of cultivation of enrichment culture SW5.2, as shown on figure 2. Gram-staining of enrichment cultures SW5.2 and SW5.5 (Figure 3) showed that the increased turbidity was undoubtedly caused by bacterial growth.The images in figure 3 also showed wide variety of shapes of Gram-negative bacteria and Grampositive cocci in SW5.2, whereas in SW5.5 only Gram-positive cocci were found.
Enrichment cultures on 0.001 g.L -1 GO and 0.001 g.L -1 GO supplemented with 10 mM fumarate were proceeded identically with those grown on 0.01 g.L -1 GO.Observations of consortia on GO 0.001 g.L -1 show scarce growth even in the presence of fumarate, therefore we did not proceed to work with them.

Consortia growth
Stable consortia SW5.2 and SW5.5, obtained from the respective enrichment cultures were cultivated as follows: SW5.2 was grown in the absence of fumarate, whereas SW5.5 was grown on GO with fumarate.Bacterial growth in the consortia was confirmed with Gram staining (Figure 4).The pictures show apparent variety of shapes in SW5.5.The growth in the cultures and consortia with GO of 0.01 g.L -1 and fumarate show the positive effects of the addition of fumarate, after first transfer.These results were confirmed with SEM observations (Figure 5).The images show GO structure (A) and scarce bacterial growth in SW5.2 (B), whereas in SW5.5 (C) many short rod-shaped bacteria were found.The abundant bacterial growth in the consortium, supplemented with sodium fumarate, correlates with results from studies, proposing fumarate addition as an activation step for the anaerobic bacterial degradation of the benzene ring and PAHs after methylation of the respective molecule [16,17].The images of the negative control and SW5.5 at the same magnification differed.The rigid layered GO particles in the negative control were not apparent in SW5.5.Instead of them, a structure, similar to exopolysaccharides was observed, which indicated biofilm formation.

Amplification and metagenomic analyses
The DNA concentration and volume obtained are shown in table 2. The choice of primers to use in the amplification of 16S rDNA is known to be crucial for the variety of OTUs, which can be identified.The primer pair used in the amplification of the V 3 -V 4 rDNA region was chosen, while taking into account the annealing temperatures of the primers and their coverage and phylum spectrum, i.e. high percentage of maximum number of phyla matched [18].The obtained V 3 -V 4 amplicons from samples S5 and SW5.2 were 611 bp long, whereas the amplicon from SW5.5 was 620 bp in length.This amplicon length includes the length of the targeted region (464 bp), the primers and the adapter overhang of the nucleotide sequences [19].Further, the amplicons were trimmed from the nucleotide sequences of primers and overhangs and proceeded in databases.
Graphical representation through Venn diagrams of the obtained results from the bioinformatic analysis is shown in figure 6 (https://bioinformatics.psb.ugent.be/webtools/Venn/):A) B) Figure 6.Venn diagrams, showing the number of shared OTUs between the samples according to A) BLAST and B) RDP.
The results from BLAST showed 162 OTUs detected in the mine sample (S5), 18 of which were unidentified.The consortium without fumarate SW5.2 had overall 31 OTUs, all of which were identified in BLAST.In consortium SW5.5 were identified all 10 OTUs in BLAST.The Venn diagram shows that all samples shared 4 OTUs, identified by accession number as Staphylococcus taiwanensis, Nocardioides aquiterrae, Cupriavidus oxalaticus NBRC 13593 and Neobacillus endophyticus.The mine sample shared 3 common OTUs with consortium SW5.2, identified as Marileptolyngbya sina, Microvirga pudoricolor and Pseudomonas paracarnis, and 2 OTUs with consortium SW5.5 identified as Massilia agri and Microvirga tunisiensis.The presence of aerobic species could be related to the presence of oxygen and hydroxyl groups in the structure of GO, which they may use as final electron acceptor.The discrepancies between number of identified species in SW5.2 and original mine sample S5 are highly likely due to incomplete desorption of the cells from the mine particles, whereas same cells in liquid medium were detached from the particles and transferred into the consortium, where they can proliferate and be identified.
Results from RDP (Figure 6B) are as follows: S5 contained 232 OTUs, 3 of which were unassigned.The OTUs we received were labelled as 'unknown 5.x' if they were unnamed and 'S5/x' if they were named as uncultured.All the samples shared 4 OTUs, which corresponds with the result from BLAST IOP Publishing doi:10.1088/1755-1315/1305/1/0120148 database.They were identified as 'unknown 5.29', 'unknown 5.47', 'S5/7' and 'S5/73' in RDP.Two of the accession numbers, corresponding to 'unknown 5.29' led to different strains of the genus Cupriavidus.The 'unknown 5.47' was identified as different strains of Nocardioides sp.These two genera matched two of the genera of the OTUs, obtained by BLAST.The other two OTUs were not identified at specific taxonomic ranks in RDP.The mine sample shared 2 OTUs with the sample from consortium SW5.2.One of them 'unknown 5.34' was identified as Pseudomonas sp. by one accession number.The other 'S5/8' was not identified at a specific taxonomic rank.The mine sample also shared 2 OTUs with the sample from consortium SW5.5: 'unknown 5.13' was identified as a bacterium from the order Hyphomicrobiales by three accession numbers, two of which led to strains of Microvirga sp.The identification of OTU 'S5/83' was inconclusive.
An interesting positive result was obtained with consortium SW5.5, where addition of fumarate initially suppressed a number of cells in enrichment cultures, plausibly the organic content was too high.Nonetheless, straight after the first transfer in fresh medium, fumarate supported better proliferation of the cells and accumulation of biomass (see Figure 5), however only of limited number of bacterial cells, as shown on Figure 6.
The identified OTUs of interest are those representing bacteria, able to grow on the conditions we provided, and are shared by the graphite mine sample and one or both of the consortia.The OTU, which meets all the criteria for being a subject of interest, showed highest similarity to Cupriavidus oxalaticus NBRC 13593.It is a facultative anaerobe and possesses genes for enzymes, related to the anaerobic degradation of substituted benzene rings, such as (E)-benzylidenesuccinyl-CoA hydratase (EC:4.2.1.180),involved in phenol degradation, flavin prenyltransferase (EC 2.5.1.129),which converts phenol into 4-hydroxybenzoate and 2-hydroxycyclohexanecarboxyl-CoA dehydrogenase (EC 1.1.1.-)and 2-ketocyclohexanecarboxyl-CoA hydrolase (EC 3.1.2.-), which participate in the degradation of benzoyl-CoA to pimeloyl-CoA [20][21][22][23].Given the fact that the genome of C. oxalaticus NBRC 13593 is not complete, more research is needed to conclude whether this strain possesses the ability to express the enzymes needed for the utilization of PAHs and other compounds, such as GO.Nevertheless, further experiments with pure cultures are needed to prove the identity of the bacteria in the consortia, respectively, their potential to degrade GO.

Conclusion
This study shows the potential of bacteria to grow in anoxic conditions in the presence of GO with and without fumarate.The results reveal that GO in concentration of 1.0 g.L -1 is toxic for bacteria.However, the concentration of 0.01 g.L -1 GO is favorable to formation of consortia.In comparison to the cultures with 0.01 g.L -1 GO, growth of bacterial cultures with 0.001 g.L -1 GO was slower and weaker.Meanwhile, supplementation with sodium fumarate as an additional carbon source promotes bacterial growth in both cases.The number of OTUs, detected in the mine sample is low, nonetheless it was reached a high level of enrichment of the bacterial consortia.From all identified OTUs, OTU identified as Cupriavidus oxalaticus is the most promising for further studies.More research is needed to conclude whether the consortia possess the ability to utilize GO as a carbon source and what are the degradation products.

Figure 2 .
Figure 2. Visualization of difference in turbidity between A) an enrichment culture SW5.2 and B) a negative control of enrichment culture SW.

Figure 3 .
Figure 3.Light microscopy images of Gram stained enrichment cultures: A) SW5.2 and B) SW5.5.The images in figure3also showed wide variety of shapes of Gram-negative bacteria and Grampositive cocci in SW5.2, whereas in SW5.5 only Gram-positive cocci were found.Enrichment cultures on 0.001 g.L -1 GO and 0.001 g.L -1 GO supplemented with 10 mM fumarate were proceeded identically with those grown on 0.01 g.L -1 GO.Observations of consortia on GO 0.001 g.L -1 show scarce growth even in the presence of fumarate, therefore we did not proceed to work with them.