Effect of model methanogens on the electrochemical activity, stability, and microbial community structure of Geobacter spp. dominated biofilm anodes

Combining anaerobic digestion (AD) and microbial electrochemical technologies (MET) in AD-MET holds great potential. Methanogens have been identified as one cause of decreased electrochemical activity and deterioration of Geobacter spp. biofilm anodes. A better understanding of the different interactions between methanogenic genera/species and Geobacter spp. biofilms is needed to shed light on the observed reduction in electrochemical activity and stability of Geobacter spp. dominated biofilms as well as observed changes in microbial communities of AD-MET. Here, we have analyzed electrochemical parameters and changes in the microbial community of Geobacter spp. biofilm anodes when exposed to three representative methanogens with different metabolic pathways, i.e., Methanosarcina barkeri, Methanobacterium formicicum, and Methanothrix soehngenii. M. barkeri negatively affected the performance and stability of Geobacter spp. biofilm anodes only in the initial batches. In contrast, M. formicicum did not affect the stability of Geobacter spp. biofilm anodes but caused a decrease in maximum current density of ~37%. M. soehngenii induced a coloration change of Geobacter spp. biofilm anodes and a decrease in the total transferred charge by ~40%. Characterization of biofilm samples after each experiment by 16S rRNA metabarcoding, whole metagenome nanopore sequencing, and shotgun sequencing showed a higher relative abundance of Geobacter spp. after exposure to M. barkeri as opposed to M. formicicum or M. soehngenii, despite the massive biofilm dispersal observed during initial exposure to M. barkeri.

Methanothrix soehngenii (acetotrophic methanogen), on the electrochemical activity, stability, and microbial community structure of pregrown Geobacter spp.biofilm anodes.To achieve this goal, the electrochemical activity and stability of pre-grown Geobacter spp.biofilm anodes were monitored during exposure to axenic cultures of each methanogen.To support the electrochemical results, the bacterial and archaeal community of Geobacter spp.biofilm samples and selected planktonic samples collected at the end of each experiment were analyzed at the DNA level using 16S rRNA metabarcoding, whole metagenome nanopore sequencing, and whole metagenome shotgun sequencing.

Results and discussion
Activity of methanogens in Geobacter spp.media During the biological control experiment, the acetate consumption of M. barkeri and M. soehngenii, the formate consumption of M. formicicum as well as the headspace gas composition were measured (Supplementary Fig. 1).The acetate and formate consumption profiles (Supplementary Fig. 1a), as well as the CH 4 production (Supplementary Fig. 1b) provide evidence that all three methanogenic archaea remained active in their respective media.Therefore, it was concluded that also mixing 50:50, v/v of acetate-based medium with methanogenic cultures grown in BFS01 medium to study specific interactions with Geobacter spp.biofilms in the single chamber MEC is not detrimental to methanogens.For a more detailed discussion of the results, please refer to the supplementary results, section 2.
Electrochemical performance of Geobacter spp.biofilms in a 50:50, v/v mixture of acetate-based medium and abiotic BFS01 medium The electrochemical activity and stability of Geobacter spp.biofilms in the 50:50, v/v mixture of acetate-based medium and BFS01 medium were monitored (Fig. 1a).The control batches using only acetate-based medium (C1-C3, see Fig. 4) already showed a slight decrease in the mean values of transferred charge (Q) and maximum current density (j max ) over time.This is related to the natural maturation of Geobacter spp.biofilms 44 .A similar slight decrease in the mean values of Q and j max was also observed during exposure to BFS01 medium without methanogens (B1* to B4*, see Fig. 4) due to ongoing maturation.Acetate utilization by Geobacter spp.biofilm and other microorganisms embedded in the biofilm structure decreased by ~23% from C1 to B4*, as also shown by the ΔCOD.Comparison of successive control batch cycles and exposure batch cycles (C1-C3 and B1*-B4*) revealed an insignificant decrease of all parameters studied (i.e., Q, j max , and ΔCOD), as indicated by overlapping confidence interval (CI) in Fig. 1a.This is consistent with the coulombic efficiency (CE) values which show no significant variations from C1 to B4* (Supplementary Table 2, Electrochemical control).
In mixed species biofilms, Geobacter spp.have been found to inhabit the inner biofilms near the electrode surface, while other microorganisms grow on the outer layers 45 .This indicates that with maturation, more nonelectroactive microorganisms form the outer layers of the biofilm where they consume substrates and reduce diffusion of substrates into the inner layers where they are oxidized by EAM.This results in a slight decrease in electrochemical biofilm performance, as observed from C1 to B4*.Therefore, we conclude that the 50:50, v/v mixture of acetate-based medium and BFS01 medium used from B1* to B4*, has no negative effect on the performance of Geobacter spp.biofilms.
Electrochemical activity and stability of Geobacter spp.biofilm upon exposure to M. barkeri The electrochemical activity and stability of Geobacter spp.biofilms in the 50:50, v/v mixture of acetate-based medium in presence of M. barkeri were monitored (Fig. 1b).It was observed that the mean values of Q and j max gradually but insignificantly decreased with increasing number of control batches (C1 to C3), while ΔCOD already showed a significant decrease during the control batches.The latter observations can be directly linked to the maturation of the biofilm leading to the formation of non-electroactive microorganisms in the outer layers of the biofilm that could decrease acetate diffusion 44 .Compared to the last control batch (C3), the first exposure batch (B1), showed no significant decrease in the mean values of Q and ΔCOD, as indicated by overlapping CI.This is in contrast to the mean value of j max which dropped by ~41%.From the first to the second exposure batch (B1 to B2), the mean values of Q, j max , and ΔCOD significantly decreased.Unexpectedly, biofilm performance reversed in the third and fourth exposure batch (B3 & B4), as indicated by a significant increase in the mean values of Q and j max .An increase was also observed in the mean values of ΔCOD, but this was only significant from B3 to B4.Comparison of the last control batch (C3) with the last exposure batch (B4), indicated no significant difference in the mean values of Q and ΔCOD, while j max values remained slightly lower in B4 than in C3.
To better understand the increase in biofilm performance observed from B3 when consistently using 3-week-old M. barkeri cultures from B1 to B4 (Fig. 1b), a similar experiment was conducted using M. barkeri cultures aged 3 to 6 weeks from B1 to B4, respectively (Fig. 2).It was observed that the previous decrease in biofilm performance in B2 (Fig. 1b), already occurred in B1 for j max and ΔCOD.However, subsequent batches showed an increase in Q, j max , and ΔCOD, with values in the last exposure batch B4 overlapping (j max and ΔCOD) or exceeding (Q) those in the last control batch C3 (Fig. 2).Thus, it seems that the age of the M. barkeri cultures used here (i.e., 3-weekold or older) had no direct influence on the interactions with Geobacter spp.biofilms.
The results in Figs.1b and 2 show that Geobacter spp.biofilms were significantly affected by M. barkeri during the first two exposure batches.The observed decrease in the performance of Geobacter spp.biofilms was evidenced by a visually perceptible dispersal (Supplementary Fig. 3b), showing the detachment of the outer biofilm layers.Visual examination of the biofilm structure in successive batches (B3 to B4) also indicated dispersal of outer biofilm layers, coupled to the formation of new biofilm layers underneath the dispersing layers (data not shown).This is in line with studies suggesting that Geobacter spp.dominate in the near-electrode portion of the biofilm 45 .The formation of new biofilm layers during biofilm dispersal may indicate that M. barkeri mainly affected the outer layers of anodic biofilms, allowing acetate to easily reach the inner layers resulting in increased Geobacter spp.activity.The latter assumption is supported by the mean CE values (Supplementary Table 2, M. barkeri + Geobacter spp.biofilms), which increased significantly in the batch cycles showing stronger biofilm dispersal (i.e., B2 in Fig. 1b and B1 in Fig. 2).The obtained overestimated CE values (CE > 100%) may also have been induced by H 2 produced at the cathode that can be consumed by Geobacter spp 46,47 .as well as  M. barkeri and other microorganisms that may cross-feed Geobacter spp 15,48 .In the batch cycles following stronger dispersal of the outer biofilm layers (i.e., from B3 in Fig. 1b and B2 in Fig. 2), the mean CE values remained nearly constant or were slightly decreased, respectively.This is consistent with the work of, e.g., Korth et al. 47 , who reported that newly formed Geobacter spp.biofilms utilize a large proportion of electrons from acetate oxidation for biomass production during growth, mainly resulting in decreased CE.
For a better understanding of the previously described variations in the electrochemical activity and stability of Geobacter spp.biofilm when exposed to M. barkeri, it is also important to consider the metabolic properties of M. barkeri.Compared to enriched biofilms of Geobacter spp. in single chamber MEC batch reactor, M. barkeri cultured at 37 °C is reported to have a lower affinity for acetate (Michaelis-Menten constant, K M of 0.67 mM for Geobacter spp. vs. 5 mM for M. barkeri) 4,49,50 .Therefore, we assume that the newly formed biofilm layers by Geobacter spp.may have outcompeted M. barkeri in substrate utilization under the experimental conditions used, resulting in the observed increase in biofilm performance from B3 in Fig. 1b and B2 in Fig. 2.
Members of the Methanosarcinaceae are reported to be able to establish DIET with EAM 27,38 .It has been proposed that DIET between Methanosarcinaceae and EAM is facilitated by multiheme cytochrome c.Yet, cytochromes are not essential for DIET, as some species, e.g., M. barkeri, lacking multiheme complexes can still perform DIET 27,38 .Referring to Fig. 1b and 2, it can be deduced that, during B1 and B2, DIET may have occurred, resulting in biofilm dispersal.This indicates that DIET-related genes are likely expressed in the biofilm outer layers that mainly dispersed in B1 and B2.Therefore, we strongly advocate follow-up studies shedding light on the expression of DIET-related genes in the Geobacter spp.biofilm structure.Furthermore, advanced techniques such as scanning electron microscopy and/or optical coherence tomography (which enables direct investigation of biofilms inside the cultivation device) would be very useful in future studies for in-depth investigation of changes in biofilm morphology 51,52 .
Electrochemical activity and stability of Geobacter spp.biofilm upon exposure to M. formicicum The electrochemical activity and stability of Geobacter spp.biofilms in the 50:50, v/v mixture of acetate-based medium and M. formicicum cultures were monitored (Fig. 1c).It was observed that the mean values of Q, j max , and ΔCOD decreased gradually but insignificantly with increasing number of control batches (C1 to C3).This is in line with the observation for Geobacter spp.biofilms in a 50:50, v/v mixture of acetate-based medium and abiotic BFS01 (Fig. 1a).Between the last control batch (C3) and the subsequent first exposure batch (B1), no significant differences were observed for all monitored parameters, as indicated by overlapping CI.Compared to B1, the mean value of j max in the second exposure batch (B2) dropped significantly by ~37% and remained nearly constant until the last exposure batch (B4).In contrast to j max , the mean values of Q and ΔCOD remained nearly constant from B1 to B4 and overlapped with the last control batch (C3).In general, the mean values of Q and ΔCOD upon biofilm exposure to M. formicicum (Fig. 1c) showed a similar trend to the control experiment without methanogens (Fig. 1a).This was evidenced by a gradual but insignificant decrease of both parameters in successive control and exposure batch cycles (C1 to B4).Similarly, CE values did not vary significantly when comparing C3 and each exposure batch cycle (Supplementary Table 2, M. formicicum + Geobacter spp.biofilms), which is indicating a limited effect of M. formicicum on the performance of Geobacter spp.biofilms.
M. formicicum is known as a hydrogenotrophic methanogen that uses H 2 and formate as electron source for the reduction of CO 2 to CH 4 , following the HIT pathway 41,[53][54][55] .H 2 can be produced in MEC via the reduction of protons at the cathode, and then used as an electron donor by M. formicicum and Geobacter spp 47,56 .However, M. formicicum is known to have a high growth rate and doubling time of only a few hours under H 2 -CO 2 feeding 53,57 .Therefore, we hypothesize that M. formicicum cells consumed a large share of the H 2 produced at the cathode.Putative H 2 consumption by M. formicicum may explain the decrease in j max observed in B2 (Fig. 1c), although CE did not vary significantly from B1 to B2 (Supplementary Table 2).Similar biofilm performance was observed by Korth et al. 47 who estimated the contribution of H 2 to biofilm performance, using a two-chamber MEC with and without additional H 2 as electron donor in the anode compartment.Whereas j max in B1 was nearly identical to j max in C3, suggesting that M. formicicum needed time to adapt to the growth conditions, Q and ΔCOD were not significantly affected in any of the batches of Geobacter spp.biofilm exposure to M. formicicum due to the inability of the latter to metabolize acetate.Furthermore, Methanobacteriaceae (e.g., M. formicicum) have been shown to be unable to perform DIET, meaning that they cannot accept electrons from acetate oxidation by Geobacter spp. to reduce CO 2 to CH 4   41,58   .Visual examination of Geobacter spp.biofilm after exposure to M. formicicum indicated no or limited biofilm dispersal to the planktonic phase (Supplementary Fig. 3c).
Electrochemical activity and stability of Geobacter spp.biofilm upon exposure to M. soehngenii The electrochemical activity and stability of Geobacter spp.biofilms in the 50:50, v/v mixture of acetate-based medium and M. soehngenii cultures were monitored (Fig. 1d).It was observed that the mean values of Q and j max decreased slightly but insignificantly with increasing number of control batches (C1 to C3).This is also in line with Q and j max for Geobacter spp.biofilms in the 50:50, v/v mixture of acetate-based medium and abiotic BFS01 (Fig. 1a).A minor but insignificant decrease in ΔCOD was observed only in C3.Compared to the last control batch (C3), the mean value of Q in the first exposure batch to M. soehngenii (B1) dropped by ~40%, unlike the mean value of j max which decreased insignificantly.From B1 to the last exposure batch (B4), both Q and j max remained nearly constant, while COD was always completely removed.The CE also dropped by ~34% from C3 to B1 and afterwards remained nearly constant (see Supplementary Table 2, M. soehngenii + Geobacter spp.biofilms).
M. soehngenii is an obligate acetoclastic methanogen 38,[59][60][61] .Like other members of the Methanosaetaceae, M. soehngenii lacks genes involved in the previously described HIT pathway 40 .Although lacking multiheme cytochrome c, certain members of the Methanosaetaceae, such as M. soehngenii have been reported to perform DIET, as they can accept electrons from acetate/ethanol oxidation by Geobacter spp.(e.g., G. metallireducens/G.sulfurreducens), for the reduction of CO 2 to CH 4   38,40,41   .However, the mean value of j max in the last exposure batch (B4) (Fig. 1d) is nearly identical to the mean value of j max in the last batch of the control experiment without methanogens (B4) (Fig. 1a).Therefore, we speculate that DIET between Geobacter spp.biofilm and M. soehngenii does not occur or occurs only at a limited rate and does not induce significant dispersal of the biofilm outer layers.
Chronoamperograms obtained upon exposure of Geobacter spp.biofilms to M. soehngenii showed no current production from day 4 to 5 of each batch cycle (B1-B4), indicating acetate limitation (Supplementary Fig. 4).The acetate limitation for Geobacter spp.biofilm during the last 2-3 days of each batch cycle, may also explain the significant drop in the mean value of Q (Fig. 1d) and CE (Supplementary Table 2) observed from C3 to B1.
After exposure to M. soehngenii, outer layers of Geobacter spp.biofilms appeared dark brownish in contrast to the inner biofilm layers which appeared reddish (Supplementary Fig. 3d), almost similar to the control biofilm (Supplementary Fig. 3a).M. soehngenii is well known for its high affinity for acetate (K M of 0.4-0.8mM) 60,62,63 , which is similar to that of Geobacter spp.(K M of 0.67 mM) 4 , and can consume acetate at very low concentrations (7-70 µmol L −1 ) 64 making it a strong competitor for acetate.The competition for acetate and the high affinity of M. soehngenii for acetate may have led to severe starvation of Geobacter spp., especially in the outer biofilm layers, during the last 2-3 days of each batch cycle from B1 to B4.Following the latter hypothesis, we speculate that cytochromes in the outer membranes of Geobacter spp.cells were oxidized, which could lead to the observed change in biofilm color.However, as M. soehngenii does not metabolize H 2 , we assume that j max values observed from B1 to B4 are the contribution of electrons from acetate oxidation by Geobacter spp.biofilm and part of the H 2 produced at the cathode.Why the outer layers of the Geobacter spp.biofilm did not disperse in the presence of M. soehngenii but in the presence of M. barkeri, albeit both species can perform DIET, cannot be answered by our data.Answering this question would require, e.g., a much longer exposure of Geobacter spp.biofilm to both methanogens, a change ratio of the mixture acetate-based medium and methanogens, or knock out/down mutants of specific DIET related genes.

Molecular biology analysis
To trace possible changes in the microbial community of Geobacter spp.biofilms and to relate those changes to the electrochemical biofilm performance (i.e., Q, j max , ΔCOD and CE), three biological replicates of biofilm samples at the end of each experiment and selected planktonic samples were analyzed by 16S rRNA metabarcoding, whole metagenome shotgun sequencing, and whole metagenome nanopore sequencing for subsequently comparison.Control biofilms (i.e., biofilms after exposure to the 50:50, v/v mixture of acetate-based medium and abiotic BFS01 medium) were used as a benchmark.Biofilm samples were analyzed with emphasis on the relative abundance of Geobacter spp.and methanogenic communities in the biofilms.
The 16S rRNA metabarcoding analysis at the genus level indicated a relative abundance of Geobacter spp. of ~58% in control biofilms, ~42% after exposure to M. barkeri, ~24% after exposure to M. formicicum, and ~17% after exposure to M. soehngenii (Fig. 3a).Other bacteria such as the genus Aminiphilus and the family Porphyromonadaceae occurred inconsistently in all biofilm samples.In contrast, the genus Bacteroides, not detected in control biofilms, showed a relative abundance of ~5%, ~2%, and ~12%, in biofilms after exposure to M. barkeri, M. formicicum and M. soehngenii, respectively.The genus Thermoanaerobacterium with a relative abundance of ~12% in control biofilms, decreased to ~5% in biofilms after exposure to M. barkeri and M. formicicum, and was barely detected in biofilms after exposure to M. soehngenii.The metabarcoding analysis showed that the main dominant methanogenic community in all biofilm samples was the genus Methanobacterium with a relative abundance of ~3% in control biofilms, ~11% in biofilms after exposure to M. formicicum, and ~5% in biofilms after exposure to M. barkeri and M. soehngenii.The methanogenic community in Geobacter spp.dominated biofilms grown in acetate-based medium is reported to be ~90% dominated by the genus Methanobacterium 47 .This may explain the overall relative abundance of ~3% Methanobacterium spp.detected in the control biofilm metagenome.Overall, it seems that, in the order of their appearance, M. barkeri, M. formicicum, and M. soehngenii induce a gradual decrease in the relative abundance of Geobacter spp. in anodic biofilms.
As previously described, biofilm outer layers of Geobacter spp.were mostly dispersed initially (B1-B2 in Figs.1b, 2, and Supplementary Fig. 3b) upon exposure to M. barkeri while new biofilm layers were formed during the following batch cycles leading to a recovery of electrochemical activity (B4 in Fig. 1b).The nearly identical j max in the last batches of the control experiment (B4* in Fig. 1a) and of the biofilm exposure to M. barkeri (B4 in Fig. 1b) is consistent with the similar relative abundance of Geobacter spp.observed in both experiments (Fig. 3a).The similar j max observed in B4* and B4 also indicate a comparable rate of acetate oxidation and thereby, electrochemical activity of Geobacter spp. in these two batches.Furthermore, similar relative abundance of Geobacter spp. was observed in the biofilms after exposure to M. barkeri aged 3 to 6 weeks from B1 to B4 (Supplementary Fig. 5) and in the biofilms after exposure to M. barkeri aged 3 weeks over all exposure batches.This confirms the hypothesis that the age of the M. barkeri cultures used here (i.e., 3-week-old or older) had no direct influence on the interactions with Geobacter spp.biofilms Metabarcoding of the planktonic phase in B4 after biofilm exposure to M. barkeri indicated a relative abundance of ~16% of Methanosarcina spp.and ~11% of Geobacter spp.(Supplementary Fig. 6).We assume that the ~11% of Geobacter spp.detected in the planktonic phase was caused by biofilm dispersal.
The observed lower relative abundance of Geobacter spp. in biofilms after exposure to M. formicicum compared to the control biofilms may be related to the variation in the relative abundance of other genera, for example, Methanobacterium and Eubacterium.Methanobacterium spp.are well known to perform HIT with CH 4 as final product 65 .The relative abundance of Methanobacterium spp. in biofilms after exposure to M. formicicum was ~3 times higher than in control biofilms indicating an active incorporation of M. formicicum in the biofilm outer layers.Therefore, Geobacter spp.were most likely limited in supply with cathodic H 2 due to H 2 consumption by M. formicicum, leading not only to a lower j max in B4 (Fig. 1c), but also to a lower relative abundance of Geobacter spp.compared to the control biofilms (Fig. 3a).
The relative abundance of the genus Eubacterium after biofilm exposure to M. formicicum was ~11% in biofilm samples and ~20% in the planktonic phase, unlike in the control biofilm where it was not detected.We assume that exposure of Geobacter spp.biofilm to M. formicicum enabled growth of Eubacterium spp.both in the biofilm and the planktonic phase.Further research on the specific interactions between Eubacterium spp., Geobacter spp.and M. formicicum is needed to shed light on the significant increase in relative abundance of Eubacterium spp.and the decrease in relative abundance of Geobacter spp.upon exposure to M. formicicum.
The relative abundance of Geobacter spp. in biofilms after exposure to M. soehngenii was ~3 times lower than in the control biofilms (Fig. 3a).This may be directly related to M. soehngenii cells in the planktonic phase that were incorporated in the biofilm, which comprised ~3% Methanothrix (Methanosaeta) spp.versus 0% in the control biofilm.Planktonic samples in B4 after biofilm exposure to M. soehngenii were not sequenced due to low DNA recovery.However, the high affinity of M. soehngenii to acetate may have caused acetate limitation for Geobacter spp. in the biofilms and therefore, triggered the observed decrease in the relative abundance of Geobacter spp.Furthermore, the relative abundance of the genus Thermoanaerobacterium in control biofilms was ~12% versus 0% in biofilms after exposure to M. soehngenii.Thermoanaerobacterium spp.have been described as fermenting bacteria involved in the conversion of VFA to CO 2 and H 2

66
. Thus, it appears that the high affinity of M. soehngenii for acetate (K M of 0.4-0.8mM) 60,62,63 , not only restricted the utilization of acetate by Geobacter spp.(K M of 0.67 mM) 4 but also hindered or reduced the electrochemical activity of other bacteria.Other microorganisms such as the order Bacteroidales (including the genus Bacteroides), the families Rikenellaceae, Enterobacteriaceae and Dethiosulfovibrionaceae, which were not detected in the control biofilms, showed significant relative abundance in biofilm after exposure to M. soehngenii.However, the effect (positive, negative or neutral) of each of the aforementioned microorganism groups on the biofilm performance needs further investigation in future research.
As the metabarcoding only covers the V3-V4 region of the 16S rRNA gene, which does not allow for sufficient differentiation down to the species level, but is limited to the genus or in some cases only the family level, biofilm samples taken after each experiment were analyzed by whole metagenome shotgun sequencing (Fig. 3b).Special attention was paid to sequence reads mapping to reference genomes of G. anodireducens, G. sulfurreducens, M. barkeri, M. formicicum, and M. soehngenii and the comparison with the relative metabarcoding abundance of the corresponding genera in Fig. 3a.Unmapped reads, i.e., Geobacter spp.sequences not found in the genomes used as references (given that the Geobacter spp.biofilm was from a mixed origin and not based on a pure strain) and other organisms, were not analyzed due to the unknown nature of their functional involvement in biofilm performance.G. anodireducens and G. sulfurreducens are known as two of the main electroactive species of the genus Geobacter 45 forming electroactive biofilms.The raw data mapping showed that ~76% and ~11% of the control sample reads mapped against G. anodireducens and G. sulfurreducens, respectively.The relative mapping abundance of both species observed in the biofilms after exposure to M. formicicum and M. soehngenii were reduced to ~57-53% and ~7.8-7.3%, respectively.In contrast the relative abundance of Geobacter spp.did not significantly change upon exposure to M. barkeri, assumingly due to the observed biofilm recovery after dispersal (vide supra, B3-B4 in Fig. 1b).However, the combined read percentages of the two Geobacter spp. in Fig. 3b indicates a clear decrease in the relative abundance of Geobacter spp. in all biofilm samples for both methods, i.e., metabarcoding (Fig. 3a) and whole metagenome shotgun sequencing (Fig. 3b).The lower relative abundance measured by metabarcoding compared to shotgun sequencing may be explained by different priming efficiency of the primers used to target archaea and bacteria in the PCR amplification process 67 .In addition, it is worth noting that the copy number of the 16S rRNA gene shows a high variation, where 1-15 copies have been reported in bacterial genomes, while it is typically lower in archaea with 1-4 copies 68,69 .This might, however, not be a major contributor in the current study since it was found that the copy number in the published reference genomes were 2 for M. soehngenii (NC_015416), M. formicicum (NZ_CP006933), and G. sulfurreducens (NZ_CP078092.1),while M. barkeri (NZ_CP009528) had 3 copies.No closed genome has been published for any verified G. anodireducens and it likely encodes more than the single copy covered by the published draft genome (NZ_JADBFD000000000). Thus, except for M. barkeri, which will be over-estimated based on 16S compared to shotgun sequencing, the other organisms (possibly with the exception of G. anodireducens) likely show a representative coverage in the amplicon sequencing study.Meanwhile abundance estimates based on read mapping percentages will also be affected by differences in genome size, where larger genomes like those of M. barkeri (4.5 Mbp), G. sulfurreducens (3.8 Mbp), and G. anodireducens (3.7 Mbp) will be overestimated compared to the smaller ones of M. formicicum (2.4 Mbp) and M. soehngenii (3.0 Mbp), given that the cell lysis otherwise has been complete for all species during DNA extraction.Thus, if the aim were to present a "cell-count" representation of abundance rather than visualizing how the biofilm microbiota composition changed over time, more normalization steps would have had to be included.
Mapping of sequence reads to the reference genomes of M. barkeri, M. formicicum, and M. soehngenii indicated that none of them were present in the control biofilms.Surprisingly, no sequence reads corresponding to the reference genome of M. barkeri were identified for biofilms after exposure to M. barkeri.This may indicate that M. barkeri cells did not integrate into the biofilm structure, which could be one of the reasons why the electrochemical activity of Geobacter spp.biofilm could recover after the dispersal of the biofilm outer layers (e.g., B3-B4 in Fig. 1b).Read percentages of ~1.3% mapping to the reference genome of M. formicicum was identified for biofilms after exposure to M. formicicum.This indicates that M. formicicum incorporated into the biofilm structure without affecting its integrity.Furthermore, since M. formicicum uses exclusively H 2 as energy source, we can reinforce our previous hypothesis that the inner layers of Geobacter spp.biofilms were deprived of H 2 , which may explain the observed decrease in j max and relative abundance of Geobacter spp.Read percentages of ~1% mapping to the reference genome of M. soehngenii was identified for biofilms after exposure to M. soehngenii.M. soehngenii is well known to have high affinity for acetate (K M of 0.4-0.8mM) 60,62,63 , and to perform DIET 38,40,41 .This may have induced strong competition for acetate with Geobacter spp.(K M of 0.67 mM) 4 , explaining the lower relative abundance of Geobacter spp.observed upon biofilm exposure to M. soehngenii compared to control biofilms.In general, it can be assumed that the higher read percentages not matching any reference genome in biofilms after exposure to M. formicicum and M. soehngenii compared to control biofilms and biofilms after exposure to M. barkeri may be related to the incorporation of M. formicicum and M. soehngenii, in the biofilm structure, probably allowing other species that already exist in the Geobacter spp.biofilms to thrive, e.g., via syntrophic interactions.Furthermore, this increase in unmapped reading could be the result of opportunistic microorganisms (having different effects on biofilm performance, or being totally neutral in terms of interaction) growing periodically in mixed cultures, which may have been promoted by either condition.
Similar to whole metagenome shotgun sequencing, whole metagenome nanopore sequencing results indicate slight incorporation of M. formicicum and M. soehngenii in the biofilms after exposure (6-and 4-fold incorporation of M. formicicum and M. soehngenii into biofilms compared to M. barkeri).Furthermore, read percentages not matching any reference genome in biofilms after exposure to M. formicicum and M. soehngenii slightly increase compared to control biofilms and biofilms after exposure to M. barkeri.The lower unmapped read observed using nanopore sequencing, as opposed to shotgun sequencing, can be explained by the high probability of sequence similarities, but with distinct differences when studying metagenomes.Thus, distinguishing between sequencing errors and real sequence differences becomes difficult when generating assemblies using nanopore sequencing (single-base accuracy ~80-90%).Overall, it seems that no major change occurred in the microbial community of the Geobacter spp.biofilm regardless of the experimental conditions.However, the electrochemical results and the results from paired-end amplicon sequencing as well as shotgun sequencing showed significant interactions, e.g., variation in j max when comparing control biofilms with biofilms after exposure to M. soehngenii.This leads to the question of the suitability of using nanopore sequencing alone when studying metagenomes.Shotgun sequencing produce short reads with a high accuracy and thus is not as sensitive to DNA fragmentation prior to library preparation compared to nanopore sequencing, which can produce ultralong sequence reads, but at a lower single base accuracy.The nanopore reads were filtered, removing all reads shorter than 1000 bp, while for the Miseq shotgun data the average read-length was typically in the range of 150-200 bp (data not shown).Thus, it cannot be excluded that the results can be affected by DNA extraction efficiency and associated DNA shearing, which might be different in one sample compared to the next.Although shotgun sequencing results appear to be more consistent with electrochemical results compared to nanopore sequencing results, further research may be needed to shed light on which of the two sequencing techniques is best suited to characterize metagenomes in MET.However, it is worth pointing out that this is no general philosophical discussion but a point of pragmatic application to our research.

General remarks
All microbial experiments were conducted under strictly anoxic conditions.All reported potentials refer to the Ag/AgCl sat.KCl reference electrode ( + 0.197 V vs. SHE).All chemicals were of analytical or biochemical grade.Electrochemical experiments were performed as independent biological replicates with n ≥ 3 and results are presented as mean values with corresponding confidence interval (CI) at 95% confidence level 70 .
Experimental setup for growing axenic methanogen cultures and Geobacter spp.dominated biofilms The experimental setup for culturing each methanogenic strain consisted of 200 mL serum bottles closed with butyl rubber stoppers and aluminium crimp seals (LABSOLUTE, Th.Geyer GmbH).The bottles were prepared, rendered anoxic and sterile according to Dzofou et al. 71 .The experimental setup for growing Geobacter spp.biofilms and performing electrochemical experiments consisted of a three-electrode arrangement, integrated into 250 mL three-neck round bottom flasks that were used as single-chamber MEC, i.e., no separation of anodic and cathodic compartment.The working and the counter electrodes were made of graphite rods (anode: d = 10 mm, L = 20 mm, A = 7.1 cm 2 , cathode: d = 10 mm, L = 30 mm, A = 10.2 cm 2 , quality CP-2200, CP-Graphitprodukte GmbH).Stainless steel wires (d = 0.5 mm, Goodfellow GmbH) and epoxy glue (Toolcraft, Conrad Electronic SE) were used to fabricate the electrodes as previously described 32 .Stainless steel wires were always isolated using a shrink tube made of modified polyolefin (ABB Ltd).The assembled MEC, with the exception of the reference electrodes (Xylem Analytics, Sensortechnik Meinsberg), were autoclave-sterilized at 121 °C for 20 min.Reference electrodes were sterilized in Beckman solution 72 for at least 2 h and then rinsed with sterile water before use.The three-neck round bottom flasks were closed with polychloroprene stoppers.To avoid overpressure in the MEC during each batch cycle, hollow needles (with attached 0.2 µm syringe filter) connected to tygon®-tubes (E 3603, inner d: 1.6 mm, Saint -Gobain Performance Plastics) were inserted in the stoppers under sterile conditions.The produced gas was released continuously into 50 mL serum bottles, half filled with autoclave-sterilized saturated solution of KCl serving as a water lock.

Microbial strains, media and cultivation
The Methanobacterium formicicum MF (DSM 1535) and Methanosarcina barkeri MS (DSM 800) strains, originally purchased from the German Collection of Microorganisms and Cell cultures (DSMZ), were provided by the working group microbiology of anaerobic systems at the department of Environmental Microbiology of the Helmholtz Centre for Environmental Research, Leipzig, Germany.The Methanothrix soehngenii GP6 (DSM 3671) strain was kindly provided by the microbiology laboratory at the department for agrotechnology and food sciences of Wageningen University and Research, Wageningen, The Netherlands.All three methanogens were cultured in the simplified methanogen medium BFS01 containing per liter: 0.348 g K 2 HPO 4 , 0.227 g KH 2 PO 4 , 0.5 g NH 4 Cl, 0.406 g MgCl 2 × 6 H 2 O, 0.25 g CaCl 2 × 2 H 2 O, 2.25 g NaCl, 1.42 mg FeCl 2 × 4H 2 O, 0.85 g NaHCO 3 , 0.3 g L-Cysteine hydrochloride monohydrate, 1 mL trace element solution SL-10, and 1 mL Wolin's vitamin solution-10 71 .Before each inoculation, the BFS01 medium was supplemented with carbon sources and electron donors as indicated in Table 1.All three methanogens were cultured separately in 200 mL serum bottles containing 120 mL of each fully supplemented BFS 01 medium (see Table 1) and 10 mL (~7.7% v/v) methanogenic inoculum in the late logarithmic phase and/or early stationary phase.M. barkeri and M. formicicum were cultured for 3 weeks and M. soehngenii for 4 weeks due to its slow growth rate 61,64 .Nearly stable CH 4 concentration in the culture headspaces over several days was used as an indicator of the maximum cell density achieved for each methanogenic culture.Due to their cell morphology: 1) M. barkeri, which aggregated as irregular-sized, interconnected cell clumps 71 , 2) M. formicicum, which formed cell colony resembling a sponge-like structure 71 , and 3) M. soehngenii, which had a rod-shaped filament morphology 71 , optical density measurement (using OD 600 ) and cell counting (using Multisitzer 3 Coulter Counter, Beckmann Coulter TM) were attempted, but regarded as inappropriate.5 mL of each methanogenic culture were used for process monitoring and the remaining 125 mL were used for the electrochemical experiments, to ensure comparable methanogenic biomass in the replicates of each single experiment (e.g., M. barkeri + Geobacter spp.biofilm).See specifications and additional information on the BFS01 medium used, growth monitoring and physiological parameters for each methanogen in Dzofou et al. 71 .
Table 1 summarizes the different substrates, gas phase composition, pH, and incubation time during the growth of the different methanogens.
Geobacter spp.biofilms were initially grown according to Gimkiewicz et al. 73 using wastewater from a primary clarifier of a local wastewater treatment plant (AZV Parthe, 04551 Borsdorf, Germany).Biofilms were subsequently enriched using a simple electrochemical enrichment procedure according to Liu et al. 74 .The growth medium used during biofilm enrichment and growth consisted of 50 mmol L −1 phosphate buffer, amended with 10 mmol L −1 sodium acetate, 12.5 mL L −1 vitamins and 12.5 mL L −1 trace elements 73,75 (for simplicity referred to as acetate-based medium).For inoculation of a fresh graphite anode, the biofilms were scraped off the anode with a sterile spatula and resuspended in fresh growth medium by vortexing.Several repetitions of this electrochemical enrichment procedure were performed until the turbidity of the planktonic phase after each batch cycle decreased significantly and a strong reddish biofilm was obtained, showing that biofilms dominated by Geobacter spp.were received 73,75,76 .
For electrochemical experiments, previously enriched Geobacter spp.biofilms were scraped off the anode using a sterile spatula and resuspended in autoclave-sterilized acetate-based medium.Subsequently, the inoculum was used in the MEC to grow new biofilm anodes under strict anoxic conditions.
The MEC were sparged with sterile N 2 gas (Nitrogen 5.0, Linde AG) for at least 30 min to maintain anoxic conditions prior to each batch cycle and operated at 38 °C using an incubator hood (Unihood 650, UniEquip).To maintain homogeneity and reduce mass transfer limitations, the media were stirred using a magnetic stirrer (Variomag Poly 15, Thermo Scientific) at 250 rpm.All MEC were connected to a multipotentiostat (PARSTAT MC, AMETEK Inc.) for biofilm growth and for the experiments.
Biofilm formation and maturation was performed in MEC using consecutive and repeated cycles of chronoamperometry (CA) for ~23 h at 0.2 V. CA was followed by three cycles of cyclic voltammetry (CV) with vertex potentials at −0.5 V and 0.3 V and a scan rate of 1 mV s −1 .Geobacter spp.biofilms were pre-grown over three successive batch cycles with one batch cycle lasting always one week.The nearly constant maximum current density (j max , normalized maximum current point in each batch cycle to the projected surface area of the working electrode) measured during the batch cycles before exposure of Geobacter spp.biofilms to the methanogens, was used as an indicator of biofilm maturation or steady-state.

Experiments
To study the interactions between Geobacter spp.biofilms and the three methanogens, five sets of experiments (four electrochemical and one biological) including two control experiments were conducted (Supplementary Table 1).Figure 4 gives an overview of the electrochemical experiments that were always conducted in batch mode with one batch cycle lasting one week.Geobacter spp.biofilms were initially pre-grown in acetate-based medium over three successive batch cycles denoted as control batches C1, C2, and C3.0.5 mmol L −1 of the autoclave-sterilized methanogen inhibitor, 2-bromoethanesulfonate (2-BES, Sigma-Aldrich) was added to the acetatebased medium during C1 and C2 to promote anode-respiring bacteria and limit methanogenesis from acetate 77 .During C3, Geobacter spp.biofilm was depleted in remaining 2-BES by using only acetate-based medium.
To verify that there was no inhibition of the three methanogens during the electrochemical experiments described below, a biological control experiment was performed to test the activity of M. barkeri, M. soehngenii, and M. formicicum in the media used to grow Geobacter spp.biofilms, i.e., acetate-based medium.In case of M. formicicum, acetate was replaced by 50 mmol L -1 formate.200 mL serum bottles containing 120 mL of acetatebased medium were inoculated with 10 mL (~7.7% v/v inoculum) of actively growing M. barkeri and M. soehngenii cultures and monitored for 42 days.A similar setup containing 120 mL of formate-based medium (i.e., 50 mmol L −1 phosphate buffer, supplemented with 50 mmol L −1 formate, 12.5 mL L −1 vitamins, and 12.5 mL L -1 trace elements) was inoculated with 10 mL (~7.7% v/v inoculum) of actively growing M. formicicum culture and monitored for 42 days.The gas phase compositions in each control were identical to those in Table 1.The gas phase in the M. formicicum culture was renewed once a week to avoid underpressure due to H 2 and CO 2 consumption for CH 4 production.
An electrochemical control experiment was performed to ensure that Geobacter spp.biofilms were not inhibited by the BFS01 medium and to subsequently compare methanogen-induced changes in the electrochemical activity, stability, and microbial community of Geobacter spp.During the electrochemical control experiment, the activity of pre-grown Geobacter spp.biofilms was monitored over four successive exposure batch cycles, denoted B1*, B2*, B3*, and B4*, by immersing biofilms in a 50:50, v/v mixture of acetate-based medium and abiotic anaerobic BFS01 medium.
To study the specific interactions between Geobacter spp.biofilms and each methanogen, three electrochemical experiments were performed.Pre-grown Geobacter spp.biofilms were immersed in anaerobic and freshly prepared 50:50, v/v mixtures of acetate-based medium and each methanogenic culture grown in the BFS01 medium (see Supplementary Table 1).Interactions between Geobacter spp.biofilms and methanogens were monitored over four successive exposure batch cycles, denoted B1, B2, B3, and B4.Prior to each experiment, the remaining acetate concentration in the methanogenic cultures (except those of M. formicicum) was measured one day in advance as previously described 32 .The final acetate concentration in the 50:50, v/v mixtures was subsequently adjusted to ~10 mmol L −1 .

Chemical and electrochemical analysis
The growth of each methanogenic culture was monitored weekly using chemical oxygen demand removal (ΔCOD), consumption of acetate and formate measured using high performance liquid chromatography (HPLC), and headspace gas composition measured using a gas chromatograph (GC).
ΔCOD was measured using COD cuvette tests (LCK 014, Hach-Lange), according to the manufacturer's instructions.The COD removal efficiency was calculated as follows: Where COD t 0 is the initial COD concentration at t o and COD t the COD concentration at the sampling point t 78 .
For measuring the acetate and formate concentrations in the methanogenic cultures and the acetate concentration in the MEC over the experiments, 1 mL aliquots were sampled from the serum bottles and the MEC at inoculation and then once a week.The samples were filtered using 0.2 µm syringe filters (Nylon, VWR) and stored at −20 °C or analysed immediately using HPLC (Shimadzu Scientific Instruments) equipped with a refractive index detector RID 10 A, a prominence diode array detector SPD.M20A, and a HiPlex H column (300 × 7.7 mm, 8 mm pore size, Agilent Technolgies) with a pre-column HiPlex H column (50 mm × 7.7 mm, 8 mm pore size, Agilent Technolgies).The sample volume for HPLC measurement was 200 µL and the injection volume was 20 µL. 5 mmol L −1 sulfuric acid was used as isocratic mobile phase with a flow rate of 0.7 mL min −1 at 60 °C, over a total run time of 60 min.
CH 4 , H 2 , and CO 2 concentrations in the headspace of each serum bottle and MEC were determined weekly.Therefore, two replicates of 1 mL gas samples were taken from the headspace of each serum bottle and MEC using sterile needle-syringe arrangements rendered anoxic by flushing with sterile oxygen-free N 2 gas.Each gas sample was injected into glass vials preflushed for 30 min with argon (Argon 4.8, Linde AG).Gas composition was analyzed using a GC equipped with an autosampler (Perkin Elmer Inc, Waltham).The GC was equipped with HayeSep N/Mole Sieve 13X columns and a thermal conductivity detector.The oven and detector temperatures were 60 °C and 200 °C, respectively.The carrier gas was argon.Each gas sample was analysed within 4 h after sampling.
The electrochemical activity of Geobacter spp.biofilms during electrochemical experiments was measured by consecutive and repeated CA and CV cycles, as during biofilm formation and maturation.CA data were analyzed for 1) maximum current density (j max ) 32,79 , 2) total transferred amount of charge (QÞ 5,32 , 3) ΔCOD 32 and 4) coulombic efficiency (CE, percentage of the electrons present in the substrate acetate that is recovered as current) 79,80 .The CE during the electrochemical experiments was determined for each batch cycle using Eq. 2.
M Ac = 59.04 g mol −1 is the molar mass of acetate, V = 250 mL is the medium volume in the MEC, F is the Faraday constant (F = 96,485.34C mol −1 ), z = 8 is the number of released electrons during oxidation of acetate, Δc = c 0 -c 1 is the difference in acetate concentration that is acetate consumption in g L −1 measured by HPLC and R idt is the total transferred amount of charge (Q), calculated by integrating the current over time 73 .
CV measurements were only performed for quality control (no data is reported).Furthermore, each electrochemical experiment was monitored weekly by measuring pH (pH 3310, WTW) and conductivity of the media (Cond 3110, WTW).

Microbial community analysis
Samples for microbial community analysis were taken at the end of each electrochemical experiment from the biofilm anodes and from the planktonic phase of each replicate.Biofilm samples were scratched off from the anodes using a sterile spatula.Both, planktonic and biofilm samples were centrifuged at 10,000 × g for 10 min and the pellets were stored in sterile 2 mL microcentrifugation tubes at −20 °C.
Genomic DNA was extracted from biofilm samples and selected planktonic samples using the Quick-DNA™ Fecal/Soil Microbe Miniprep Kit (ZYMO RESEARCH).DNA concentrations were measured by fluorescence quantification using the Qubit TM dsDNA BR Assay Kit (Thermo Fisher Scientific) and a Qubit TM 4 Fluorometer (Thermo Fisher Scientific).The purity of the DNA was evaluated by microvolume absorbance on a DS-11 FX+ (Denovix), where A260/A280 and A260/A230 values around 1.8 and 2.0, respectively, are considered as pure nucleic acid.
The bacterial and archaeal diversity of the samples was determined by metabarcode sequencing, as previously described 71 .
Since metabarcoding only covers the V3-V4 region of the 16S rRNA gene, which does not allow for good differentiation down to the species level, but is limited to the genus or in some cases only the family level, MiSeq whole metagenome shotgun sequencing of biofilm samples was performed and the relative read-mapping results compared with the MiSeq amplicon sequencing results.Sequencing libraries were prepared according to the Nextera XT DNA Library Prep Kit (Illumina) and sequenced on an Illumina MiSeq with the MiSeq Reagent Kit v3 in a paired-end mode and read lengths of 300 bp.Raw sequencing data was demultiplexed and converted to fastq-files in Local Run manager (Illumina).The data were further processed in the CLC Genomics Workbench 22.0.1 (Qiagen).The raw reads were filtered using Trim Reads 2.6, and mapped against the 5 reference genomes (G.anodireducens, RefSeq GCF_014883105.1;G. sulfurreducens, RefSeq GCF_019904315.1;M. soehngenii, RefSeq GCF_000204415.1;M. formicicum, RefSeq GCF_000762265.1, and M. barkeri RefSeq GCF_000970025.1) using Map Reads to Reference 1.8 with default parameters.The archaea reference genomes were previously found to be suitable references for these 3 species 71 .

Reporting summary
Further information on research design is available in the Nature Research Reporting Summary linked to this article.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material.If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Fig. 2 |
Fig. 2 | Exploring the effect of M. barkeri age on Geobacter spp.biofilm.Transferred charge (Q), current density (j max ) and chemical oxygen demand removal (ΔCOD) at the end of each batch cycle upon exposure of Geobacter spp.biofilm to a 50:50, v/v mixture of acetate-based medium and M. barkeri cultures in BFS01 medium.C1-C3: control batches with only acetate-based medium, B1-B4: four successive batch cycles with Geobacter spp.biofilms exposed to M. barkeri cultures aged 3-6 weeks, respectively, n = 4, error bars indicate confidence interval CI.

Fig. 4 |
Fig. 4 | Overview of electrochemical experiments.Each batch cycle always lasts one week."C1 to C3" indicate the three successive control batch cycles during the pregrowth of Geobacter spp.biofilms with 100% acetate-based medium, "Exposure batch cycles" indicate the successive batch cycles during the electrochemical control experiment without methanogens, i.e., B1*-B4* with the 50:50, v/v mixture of acetate-based medium and abiotic BFS01 medium (top) as well as the electrochemical experiment with methanogens, i.e., B1-B4 with the 50:50, v/v mixture of acetate-based medium and each methanogenic culture (bottom).

Table 1 |
Growth conditions for methanogenic cultures