response of methanogen community to elevation of cathode … · 2020. 11. 24. · 4 running title:...
TRANSCRIPT
1
Response of methanogen community to elevation of cathode 1
potentials in the presence of magnetite 2
3
Running title: Electromethanogenesis of paddy soil community 4
5
Kailin Gao, † Xin Wang,
‡ Junjie Huang,
† Xingxuan Xia,
† Yahai Lu
*, † 6
7
†College of Urban and Environmental Sciences, Peking University, Beijing, 100871, 8
China 9
‡College of Environmental Science and Engineering, Nankai University, Tianjin 10
300350, China 11
12
*Corresponding author: 13
Yahai Lu, Peking University, College of Urban and Environmental Sciences, No. 5, 14
Yiheyuan Road, Haidian District, Beijing 100871, China. 15
Phone/fax: 0086 10 62750669 16
E-mail address: [email protected] (Y. Lu) 17
18
Keywords Magnetite; Electromethanogenesis; Methanogens; Paddy soil; 19
Methanospirillum 20
21
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
2
ABSTRACT 22
Electromethanogenesis refers to the process where methanogens utilize electrons 23
derived from cathodes for the reduction of CO2 to CH4. Setting of low cathode 24
potentials is essential for this process. In this study, we test if magnetite, an iron oxide 25
mineral widespread in environment, can facilitate the adaption of methanogen 26
community to the elevation of cathode potentials in electrochemical reactors. 27
Two-chamber electrochemical reactors were constructed with inoculants obtained 28
from a paddy field soil. We elevated cathode potentials stepwise from the initial -0.6 29
V vs standard hydrogen electrode (SHE) to -0.5 V and then to -0.4 V over the 120 30
days acclimation. Only weak current consumption and CH4 production were observed 31
in the reactors without magnetite. But biocathodes were firmly developed and 32
significant current consumption and CH4 production were recorded in the magnetite 33
reactors. The robustness of electro-activity in the magnetite reactors was not affected 34
with the elevation of cathode potentials from -0.6 V to -0.4 V. But, the current 35
consumption and CH4 production were virtually halted in the reactors without 36
magnetite when cathode potential was elevated to -0.4 V. Methanogens related to 37
Methanospirillum were enriched on cathode surface of the magnetite reactors at -0.4 38
V, while Methanosarcina relatively dominated in the reactors without magnetite. 39
Methanobacterium also increased in the magnetite reactors but stayed off electrodes 40
in the culture medium at -0.4 V. Apparently, magnetite greatly facilitates the 41
development of biocathodes, and it appears that with the aid of magnetite 42
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
3
Methanospirillum spp. can adapt to high cathode potentials performing the efficient 43
electromethanogenesis. 44
45
46
IMPORTANCE 47
Converting CO2 to CH4 through bioelectrochemistry is a promising approach for 48
development of green energy biotechnology. This process however requires setting 49
the low cathode potentials, which takes cost. In this study, we test if magnetite, a 50
conductive iron mineral, can facilitate the adaption of methanogens to the elevation of 51
cathode potentials. In the two-chamber reactors constructed using inoculants obtained 52
from a paddy field soil, biocathodes were firmly developed in the presence of 53
magnetite, whereas only weak electro-activity was observed in the reactors without 54
magnetite. The elevation of cathode potentials did not affect the robustness of 55
electro-activity in the magnetite reactors over the 120 days acclimation. 56
Methanospirillum was identified as the key methanogens associated with cathode 57
surface during the operation at relatively high potentials. The findings reported in this 58
study shed a new light on the adaption of methanogen community to the elevated 59
cathode potentials in the presence of magnetite. 60
61
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
4
INTRODUCTION 62
Bioelectrochemical technology has been developed rapidly in recent decades with one 63
of attracting applications being the conversion of CO2 to CH4 (1-3). 64
Electromethanogenesis refers to the process where methanogens utilize electrons 65
derived from cathodes for the reduction of CO2 to CH4 (4). Two plausible 66
mechanisms have been proposed for electron transfer from cathodes to methanogens. 67
One is the H2-mediated electron transfer, which assumes that H2 is electrochemically 68
produced either abiotically or biotically that is then used by hydrogenotrophic 69
methanogens for reduction of CO2 to CH4 (5, 6) and the other, though not fully 70
proven, is the direct electron transfer from cathodes to methanogens (7-10). 71
Cathode potential is the critical factor controlling either chemical or biological H2 72
evolution. Theoretically, chemical H2 production from proton reduction can occur at 73
-0.414 V (all potentials reported here are relative to standard hydrogen electrode, 74
SHE), but in practice the cathode potential must be set substantially lower due to the 75
overpotential in electrochemical operation (4, 11-13). Consequently, cathode 76
potentials of -0.5 V to -0.8 V were usually applied in electromethanogenic reactors 77
where H2-mediated electron transfer was assumed as the key process (4, 14-16). H2 78
evolution can occur through different bioelectrochemical mechanisms. For instances, 79
it has been proposed that the high rate of hydrogen-mediated electromethanogenesis 80
in Methanococcus maripaludis was due to the release of hydrogenases from living 81
and dead cells, which attached to cathode surface and catalyzed H2 production (5). In 82
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
5
mixed culture electrochemical systems, organisms other than methanogens may 83
produce H2 by taking up cathode electrons directly or indirectly and then channel H2 84
to hydrogenotrophic methanogens for CO2 reduction to CH4 (6). In a defined 85
coculture it was demonstrated that the Fe(0)-corroding sulfate-reducing strain IS4 86
performed direct cathode electron uptake during sulfate reduction with active H2 87
production at -0.4 V and -0.6 V, and H2 was consumed by M. maripaludis for CO2 88
reduction to CH4 (17). 89
Methane production from CO2 reduction can occur at a redox potential of -0.244 V 90
under standard conditions. A few studies have suggested that some methanogens can 91
operate the direct electron uptake from cathodes for catabolic metabolism, thus 92
bypassing the H2-mediated processes. In this case it becomes less critical to apply a 93
low potential as compared to the H2-dependent reactors. For instances, a 94
marine-origin Methanobacterium-like strain IM1, which grew on iron specimen but 95
hardly on H2, could sustain electromethanogenesis at -0.4 V, while the typical 96
hydrogenotrophic M. maripaludis failed the operation at this potential (18). Similarly, 97
Methanosarcina barkeri, a methanogen known to conduct direct interspecies electron 98
transfer (DIET) with Geobacter metallireducens, can perform electromethanogenesis 99
at -0.4 V (9, 10, 19). The hydrogenase-independent electron uptake by the M. barkeri 100
mutant lacking hydrogenases has been observed at a potential of -0.484 V (8). 101
Therefore, while H2-mediated electromethanogenesis generally requires low cathode 102
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
6
potentials, H2-independent or DET-associated electromethanogenesis can work at 103
relatively high potentials. 104
The performance of electromethanogenesis could be improved by introducing 105
supplemental materials that are optimized for cathode electron transfer or that can 106
efficiently catalyze H2 evolution at low potentials (20-22). Iron oxide minerals such as 107
magnetite are common in soils (23-27). Recently, it has been demonstrated that the 108
presence of magnetite nanoparticles (MNP) greatly promotes methanogenesis from 109
oxidation of short-chain fatty acids in rice paddy soils and anaerobic sludges, and the 110
stimulatory effect has been attributed to the facilitation of DIET (25, 27-30). 111
Magnetite is a mixed-valent iron oxide mineral containing Fe(II) and Fe(III) in a ratio 112
of 1:2. The edge-sharing of the octahedral Fe(II) and Fe(III) in magnetite facilitates 113
the electron hopping or rapid electron exchange along the octahedral sublattice 114
resulting in high electrical conductivity and redox activity (23, 31). It is tempting to 115
explore the influence of MNP on electromethanogenesis. 116
Previous studies on electromethanogensis often collected inoculants from anaerobic 117
digesters. Anaerobic digesters specialized in industry may limit the diversity of 118
methanogens (1, 14, 15, 21, 22, 32), while natural environments such as paddy soil or 119
wetlands can harbor a wider variety of methanogens (33) and hence offer an 120
opportunity to screen methanogens capable of efficient electromethanogenesis. In the 121
present study, electromethanogenic reactors were constructed using an inoculum 122
obtained from a rice paddy soil. To test the effect of magnetite and explore how 123
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
7
methanogens response to varying cathode potentials, the cathode potentials of reactors 124
were elevated stepwise from -0.6 V to -0.5 V and finally to -0.4 V. Over four months 125
of electromethanogenesis acclimation, we continuously monitored the CH4 126
production, the electrochemical performance and the microbial population dynamics. 127
Systematic analysis of the combined data about the adaption of methanogens to the 128
increasing cathode potential with and without magnetite sheds the new light on 129
electro-active methanogens community originated from paddy soil. 130
131
RESULTS 132
Magnetite promoted electromethanogenesis 133
Three batches of reactors were constructed, namely the reactors with MNP (MNP 134
reactors), the reactors without MNP (no-MNP reactors), and the reactors without 135
inoculants but with MNP (the abiotic MNP control). All reactors were operated 136
continuously except two breaks. In the initial phase, cathode potential was set at -0.6 137
V. Neither current consumption nor CH4 production were detected in the abiotic 138
control. For the inoculated reactors, electromethanogenesis initiated after over a 139
month acclimation. Magnetite exerted a strong influence. In the MNP reactors, rapid 140
current consumption was detected at 33 d with concomitant sharp increase of CH4 141
production at 40 d (Fig. 2). The current consumption and CH4 production, however, 142
were much lower in the no-MNP reactors where obvious electromethanogenesis 143
occurred only after 50 d. When the current consumption in the MNP reactors did not 144
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
8
increase further, the first break was applied. After the CV tests of cathodes and the 145
exchange of catholytes with fresh culture medium, the circuits were reconnected and 146
cathode potentials were elevated to -0.5 V for the second stage of operation. Methane 147
production and current consumption occurred without delay in the MNP reactors, 148
indicating that microbial populations retained on electrode surfaces (microbes in the 149
catholytes were disposed during medium exchange) were well adapted to the 150
electrochemical environment (Fig. 2). The current consumption and CH4 production 151
also occurred in the no-MNP reactors, but the activity remained much lower as 152
compared with the MNP reactors. The total CH4 accumulation was 22.68 mmol L-1
in 153
the MNP reactors and 7.71 mmol CH4 L-1
in the no-MNP reactors, respectively, over 154
the period at -0.5 V operation (Fig. 2a). When the current consumption stopped 155
increasing again, the second break was implemented. After the CV tests of cathodes 156
and culture medium exchange repeated, the cathode potentials were elevated to -0.4 V 157
for the third stage of operation. The current consumption and CH4 production kept 158
highly active in the MNP reactors. The total CH4 accumulation over 30 days 159
operation actually exceeded those observed in the earlier stages at -0.6 V and -0.5 V. 160
However, CH4 production and current consumption in the no-MNP reactors were 161
virtually halted at -0.4 V (Fig. 2). 162
H2 accumulated to 0.02 mmol L-1
in the abiotic control during the initial operation at 163
-0.6 V. H2 was occasionally detected in the MNP reactors at concentrations close to 164
detection limit (0.01 mmol L-1
). Otherwise H2 concentrations were below the 165
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
9
detection limit in most cases, especially during the operations at -0.5 V and -0.4 V. To 166
verify if addition of MNP influenced redox potentials (Eh) in culture medium, we 167
measured Eh under open circuit conditions. The results showed no difference in the 168
presence (-0.340 V) and absence (-0.338 V) of MNP. 169
At each break and at the end of experiment, CV tests were conducted to evaluate the 170
catalytic activity of cathodes. At the break after -0.6 V operation, the no-MNP 171
reactors revealed the lowest redox activity of cathodes, followed by the abiotic 172
reactors while the highest redox activity was recorded for the MNP reactors (Fig. 3a). 173
The activity of the abiotic reactors indicates that addition of MNP improved the 174
conductivity of electrodes and electrolytes compared to the inoculated reactors 175
without MNP. CV tests at the break after -0.5 V operation and at the end of 176
experiment exhibited similar results and confirmed the significantly enhanced 177
catabolic activity of cathodes in the MNP reactors compared with the no-MNP 178
reactors (Fig. 3b, c). The voltammograms showed some distinct redox peaks and 179
inflection points in the current profiles (Fig. 3a-c). For instances, at the break after 180
-0.6 V operation, the current generation in the MNP reactors plateaued in the potential 181
range from -0.23 V to -0.51 V, followed by an increase and then flattened again at 182
-0.6 V. The inflection point occurred at -0.48 V for the MNP reactors at the break 183
after -0.5 V operation and at -0.41 V after -0.4 V operation. These reflection points 184
illustrated the robust development of biocathodes in the MNP reactors. 185
Response of archaeal and bacterial communities 186
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
10
Amplicon sequencing was used to analyze community compositions of archaea and 187
bacteria attached on the cathode surfaces and living in the culture mediums (catholyte 188
solution). The sequence summary for all archaea amplicons was given in Table S1. 189
Archaea were composed exclusively of methanogen populations. Methanogens in the 190
original inoculum obtained from rice paddy soil were dominated by Methanospirillum 191
accounting for 90% of total archaeal sequences (Fig. 4a). The rest mainly affiliated to 192
Methanosarcina and Methanoregula. Over the operation of electromethanogenesis, 193
Methanobacterium replaced Methanoregula, arising as the third dominant 194
methanogen (Fig. 4a; Fig. S1). The methanogen compositions in reactors were 195
markedly influenced by the magnetite treatment, the potential elevation and the 196
sample location (cathode surface vs catholyte medium). 197
In the no-MNP reactors, the relative abundance of Methanospirillum and 198
Methanobacterium decreased, while that of Methanosarcina increased from 15% to 199
60% with the elevation of cathode potentials from -0.6 V to -0.4 V (Fig. 4a). There 200
was no significant difference in methanogen composition between the cathode surface 201
and the culture medium (Fig. 4a). The MNP reactors showed a significantly different 202
pattern. After the operation at -0.6 V, Methanospirillum remained dominant while 203
Methanobacterium slightly increased compared with the original inoculum. When 204
cathode potential was shifted to -0.5 V, Methanobacterium greatly increased and 205
surpassed Methanospirillum both on cathode surface and in catholyte medium. 206
Methanosarcina also increased relatively in the culture medium. When the cathode 207
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
11
potential was further elevated to -0.4 V, Methanospirillum returned as the most 208
dominant methanogen on cathode surface while Methanobacterium kept dominant 209
only in catholyte medium (Fig. 4a). Apparently, Methanospirillum were selected 210
against Methanobacterium and Methanosarcina on cathode surfaces at -0.4 V (Fig. 211
4a). 212
The shift of methanogen community was also illustrated by the nonmetric 213
multidimensional scale (NMDS) analysis (Fig. 4b). Methanogen communities under 214
electromethanogenesis were all distinct from original inoculum. At the break after 215
-0.6 V operation, communities from all samples were clustered (green oval). At the 216
break after -0.5 V operation, communities from the MNP and no-MNP reactors were 217
separated (two purple ovals), while at the end after -0.4 V operation, a more 218
significant divergence was revealed (three red ovals), with the separations not only 219
depending on magnetite treatments but also on sample sites (cathode surface vs 220
culture medium). 221
The phylogenetic relationship of top 10 archaeal OTUs was depicted in Fig. 5a. 222
Clone_Mspi were closely related to Methanospirillum psychrodurum X-18 and 223
Methanospirillum lacunae Ki8-1. Clone_Mba1 and clone_Mba2 were related to 224
Methanobacterium formicicum MF and Methanobacterium flexile GH. Clone_Msar 225
was related to Methanosarcina horonobensis HB-1 and Methanosarcina mazei DSM 226
2053. Given that Methanospirillum were predominant and selected by cathode surface 227
in the MNP reactors at -0.4 V, we further analyzed the relative abundances of 15 228
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
12
OTUs affiliated to Methanospirillum (Fig. 5b). A significant shift was revealed for 229
these Methanospirillum phylotypes. Clone_Mspi2 was dominated in the original 230
inoculum. At the first break after -0.6 V operation, Clone_Mspi1, 4 and 14 increased, 231
while clone_Mspi2 dropped. At the second break after -0.5 V operation, all the 232
Methanospirillum OTUs declined. But at the end after -0.4 V operation, ten other 233
OTUs including clone_Mspi3, 6-13 and 15 were substantially enriched (Fig. 5b). 234
These results implied that clone_Mspi3, 6-13 and 15 instead of the dominant 235
clone_Mspi2 in the original inoculum were enriched during electromethanogenesis at 236
-0.4 V. 237
The sequence summary for the bacteria amplicons was listed in Table S2. Bacteria 238
community consisted mainly of Acetobacterium, Anaerolineaceae, Sulfuricurvum, 239
Geobacter, Bacteroidales, and Desulfovibrio (Fig. S2). Acetobacterium and 240
Anaerolineaceae, the first and second most abundant bacteria across reactors, were 241
more enriched in the MNP reactors compared to the no-MNP reactors (Fig. 6). By 242
comparison, the other four bacteria mentioned above exhibited higher relative 243
abundances in the no-MNP reactors than in the MNP reactors. The relative abundance 244
of Acetobacterium however decreased sharply with the elevation of cathode potentials, 245
from 54% to 10% in the MNP reactors and from 32.5% to 0.06% in the no-MNP 246
reactors, respectively. Though relatively higher abundances of Acetobacterium on 247
cathode surface than in catholyte medium of the MNP reactors at -0.6 V and -0.5 V, 248
this difference was diminished at -0.4 V (Fig. 6). Anaerolineaceae showed the 249
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
13
opposite tendency. Their relative abundances were low at -0.6 V and -0.5 V, but 250
markedly increased with the elevation of cathode potentials to -0.4 V where they were 251
the most dominant bacteria both on cathode surface and catholyte medium. 252
Actinobacteria showed low relative abundance across all samples except on cathode 253
surface in the MNP-reactors at -0.4 V where they became the second dominant 254
bacteria after Anaerolineaceae. The NMDS analysis revealed that the samples from 255
the MNP reactors at -0.6 and -0.5 V formed a cluster while those at -0.4 V formed 256
another cluster (Fig. S4). The samples from the no-MNP reactors were separated into 257
three clusters depending on cathode potentials and sample locations. 258
259
DISCUSSION 260
In the present study we demonstrated that electromethanogenesis derived from a 261
paddy soil community was substantially promoted by magnetite during the 120 days 262
operation with the elevation of cathode potentials from -0.6 V to -0.4 V (Fig. 2). 263
Granular active carbon and zero-valent iron have been employed to accelerate the 264
start-up of biocathodes and upgrade electromethanogenesis efficiency (20-22). One 265
such a study however showed that electromethanogenesis was not affected by 266
magnetite amendment in comparison with magnetite-free reactors (20). Thus, the 267
stimulating effect appears dependent on reactor conditions and especially microbial 268
identities in operations (34). Nevertheless, it has been demonstrated that in natural 269
systems the syntrophic oxidation of short chain fatty acids and CH4 production were 270
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
14
significantly stimulated by magnetite nanoparticles (25-27, 35). Magnetite was also 271
shown to facilitate syntrophic interaction between a phototrophic bacterium 272
Rhodopseudomonas palustris and an iron-reducing Geobacter sulfurreducens (36, 273
37). More recently, it was demonstrated that magnetite accelerated the aceticlastic 274
methanogenesis by a pure culture Methanosarcina mazei zm-15 and its corresponding 275
environment enrichment (38). The mechanisms are considered to be related to DIET 276
in syntrophic interaction (25) or serving as an environmental battery facilitating 277
electron transfer among microbes (37). In the present experiment, we lifted cathode 278
potentials from the initial -0.6 V to the final -0.4 V over the 120 days acclimation. 279
The electromethanogenesis remained robust in the presence of magnetite whereas the 280
activity was substantially inhibited in the reactors without magnetite. Albeit mixture 281
community in our reactors hindered the elucidation of exact mechanism, we assume 282
that magnetite amendment facilitated the electron transfer from cathodes to 283
methanogens either directly or indirectly. 284
Microbial communities in our reactors were markedly influenced by the magnetite 285
treatment, the potential elevation and the sample location (Fig. 4, Fig. 6 and Fig. 286
S3-S4). In the no-MNP reactors, the relative abundance of Methanosarcina gradually 287
increased with the elevation of cathode potentials (Fig. 4a). Methanosarcina barkeri 288
and Methanosarcina mazei have been demonstrated to be capable of DIET with 289
Geobacter metallireducens or perform electromethanogenesis at -0.4 V (8-10, 19). It 290
is possible that Methanosarcina spp. contributed to electromethanogenesis in the 291
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
15
no-MNP reactors albeit their weak electrochemical performance. The relative 292
abundance of Methanosarcina, however, substantially declined in the reactors with 293
magnetite. Apparently, Methanosarcina did not respond to the stimulatory effect of 294
magnetite or were less competitive during the development of biocathodes in the 295
MNP reactors. Methanobacterium on the other hand significantly increased in the 296
reactors with magnetite, especially at -0.5 V, reaching the relative abundance of about 297
50% on cathode surface as well as in catholyte medium. These methanogens however 298
wane from cathode surface when cathode potentials were further elevated to -0.4 V. 299
As the replacement, Methanospirillum were dominated on cathode surface at -0.4 V. 300
The concomitant CH4 production and current consumption indicated that biocathodes 301
were well developed in the MNP reactors (Fig. 2). The facts that the culture medium 302
exchange at two breaks and the elevation of cathode potentials from -0.6 V to -0.4 V 303
did not influence the electrochemical performance indicate that the biocathodes 304
remained robust in the MNP reactors over the 120 days operation. This robustness of 305
biocathode development was further illustrated with the CV tests at two breaks and at 306
the end of experiment (Fig. 3). Accordingly, we assumed that microbial populations 307
associated with cathode surface played the key role for electromethanogenesis in the 308
MNP reactors. Many of previous studies have found that Methanobacterium were 309
enriched in the cathode chambers of electrosynthesis reactors (13-16, 39). We show 310
here that Methanospirillum were dominated on cathode surface at -0.4 V in the 311
presence of magnetite (Fig. 4a). Methanospirillum have been underappreciated in the 312
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
16
previous studies on electromethanogenesis. Recently, it has been reported that the 313
archaellum of Methanospirillum hungatei is electrically conductive, making it an 314
excellent candidate for the research of extracellular electron transfer in archaea and 315
the application in electrochemical systems (40). It shall warrant a further investigation 316
for whether the Methanospirillum in our reactors contain the electrically conductive 317
archaellum, which facilitates the extracellular electron transfer in the biocathode 318
ecosystems. 319
For the mixed culture reactors, the accompanying bacteria can play important roles in 320
electromethanogenesis. Electro-acetogenesis refers to biological production of acetate 321
from CO2 with electrons derived from cathodes (41, 42). Acetobacterium was the 322
most dominant bacteria across all samples and were relatively enriched in the MNP 323
reactors (Fig. 6), indicating their possible involvement in electro-acetogenesis. But the 324
relative abundance of these acetogens significantly declined in the reactors at -0.4 V, 325
suggesting they were not tolerant to the elevation of cathode potentials. The role of 326
bacteria in the electromethanogenic reactors can be multifold. Firstly, acetogens 327
produce acetate electrochemically which is then used by aceticlastic methanogens like 328
Methanosarcina in the reactors. Second, some electrically-active bacteria may 329
generate H2 by taking up electrons from electrodes and then channel H2 to 330
hydrogenotrophic methanogens, forming syntrophic interaction (17). Third, it is 331
plausible that DIET is established between the electrically-active bacteria and 332
methanogens without the involvement of intermediate H2 production. Except 333
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
17
Acetobacterium, some of potentially electrically-active organisms like Geobacter, 334
Sulfuricuvum and Desulfovibrio were present as major bacteria members (Fig. S2). 335
These organisms however exhibited relatively high abundances only in the no-MNP 336
reactors, and thus did not play a significant role in the MNP reactors. 337
Anaerolineaceae were the second dominant bacteria next to Acetobacterium (Fig. 6). 338
These bacteria were significantly enriched in the MNP reactors at -0.4 V. 339
Anaerolineaceae species have long filamentous structure (43, 44), and hence have an 340
advantage of forming microbial aggregates or biofilms by attaching to electrode 341
surface. Some Anaerolineaceae are known to ferment sugars and grow better in 342
coculture with H2-consuming methanogens (44), implying their syntrophic life style. 343
Therefore, it is plausible that Anaerolineaceae are involved in electromethanogenesis 344
through forming syntrophic associations or improving methanogenic aggregate 345
formation in the present experiment. 346
In summary, this study demonstrates that MNP greatly facilitates the adaption of 347
methanogens in electrochemical reactors with the stepwise elevation of cathode 348
potentials from -0.6 V to -0.4 V. Methanospirillum was identified as the key 349
methanogens associated with cathode surface at -0.4 V. Given that the archaellum of 350
Methanospirillum hungatei is known to be electrically conductive, it is worthwhile to 351
further explore if Methanospirillum can perform direct electron transfer in the 352
electrochemical reactors. Though a set potential of -0.4 V is considered to be 353
sufficiently high to limit electrochemical H2 production under pure culture conditions 354
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
18
(1, 8-10, 18), we are not able to dissect if the H2-mediated or H2-independent 355
methanogenesis prevails in our reactors. Further research is necessary to determine if 356
syntrophy is involved in electromethanogenesis, in which some bacteria may fetch 357
electrons from electrodes forming H2 which is then utilized by the hydrogenotrophic 358
Methanospirillum and Methanobacterium. It also remains unclear if the redox-active 359
enzymes including hydrogenases are released from living and dead cells, which are 360
then deposited on electrode surface facilitating H2-mediated processes (5, 45). These 361
open questions await a further research in the future. 362
363
MATERIALS AND METHODS 364
Enrichment preparation for cathode inoculation 365
To focus on methanogen community and exclude carbon sources from soil, 366
methanogenic enrichment from a rice paddy soil was prepared as below. The 367
water-saturated soil samples were collected from a paddy field located in the 368
northeastern China close to Heihe city of Heilongjiang province, China (127.36°E, 369
49.90°N). Four successive transfers of anaerobic incubation were conducted. For the 370
first transfer, fresh soil samples were suspended in autoclaved degassed water at a 371
water to soil ratio of 5:1 (soil mass in dry weight). Aliquots (50 ml) of the 372
homogenized soil slurry were then dispensed into 125 ml sterile serum bottles. 373
Glucose was added into bottles at a final concentration of 5 mM in slurries. 374
Headspace of serum bottles were flushed with N2 thoroughly. All the bottles were 375
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
19
sealed with butyl stoppers and aluminum crimp caps, and put in the dark at room 376
temperature (27 ℃). When the rate of daily CH4 production reached to the 377
quasi-steady state, the soil slurries were used as inoculum (5% v/v) for the next three 378
transfers where acetate (10 mM) was spiked into 50 ml HEPES-buffered (30 mM, pH 379
7) culture medium and the headspace of serum bottles were flushed with H2/CO2 380
(80:20; v/v). Both acetate and H2/CO2 served as the carbon and energy source for 381
methanogen enrichment (Fig. 1). The basal medium consisted of MgCl2·6H2O (0.4 382
g/L), CaCl2·2H2O (0.1 g/L), NH4Cl (0.1 g/L), KH2PO4 (0.2 g/L), KCl (0.5 g/L). 383
Supplements of vitamin, and trace element solutions were applied. Resazurin (46) and 384
cysteine were omitted to avoid the possible effect of electron shuttle molecules (47). 385
All the bottles for the enrichment cultivation were sealed and put in the dark at room 386
temperature (27 ℃). When acetate was used up in the medium, the enrichment 387
cultures were used to inoculate (10% v/v) the cathode chambers of bioelectrochemical 388
reactors. Abiotic control reactors were established without inoculation. 389
Setup of bioelectrochemical systems 390
The bioelectrochemical reactors consisted of two chambered borosilicate gastight 391
H-type microbial electrolysis cells, in which the 250 ml anode and cathode chambers 392
were separated by a Nafion 117 proton exchange membrane (surface area 5 cm2, 393
DuPont, Wilmington, DE, USA) (Fig. 1). Prior to use, membranes were successively 394
treated one hour each with H2O2 (3.5 wt.%), distilled water, H2SO4 (5 wt.%), and 395
distilled water. In each chamber, there are an upper and a lower openings, which were 396
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
20
sealed with butyl stoppers and aluminum crimp caps. The upper openings were used 397
to collect gas samples (Fig. 1). Carbon fiber brush (volume 20 cm3) acted as the work 398
electrode in the cathode chambers and platinum foil was used as the counter electrode 399
in the anode chambers. An Ag/AgCl reference electrode (+0.2046 V vs SHE, 25℃) 400
was placed in the cathode chamber as a reference electrode. Magnetite nanoparticles 401
(MNP) was synthesized via a conventional aqueous co-precipitation method (48) as 402
described previously (27). 403
Bioelectrochemical measurements and cathode potential elevation 404
Nitrogen was flushed up to one hour to make the anaerobic condition. After 405
autoclaving, the anaerobic basal medium was added into two chambers of reactors. 406
The composition of basal medium was same as the enrichment cultivation but without 407
the addition of acetate. 125 ml medium was dispensed into anode chambers. And 408
112.5 ml of basal medium and 12.5 ml of inoculum were dispensed into cathode 409
chambers to bring the total volume to 125 ml. The abiotic control was prepared with 410
addition of 125 ml same medium in both chambers without inoculation. The 411
headspace of each reactor had a volume of 125 ml and was flushed with N2/CO2 412
(80:20; v/v) thoroughly. CO2 was the sole carbon source in reactors. Reactors were 413
put in the dark at room temperature (27℃). The experiment was divided into three 414
groups. For the first group, MNP was added into the cathode chambers with a final 415
concentration of 10 mM in Fe atom. The second group was prepared similarly but 416
without MNP in chambers. The third group was set up as abiotic control with addition 417
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
21
of MNP in cathodes but without inoculation of enrichment. All reactors were 418
connected into the eight channel potentiostat (CHI 1000C, CH Instruments Inc., 419
Shanghai, China) which recorded current every 100 seconds automatically. The 420
reactors were operated in a continuous mode except two breaks for the shifting of 421
cathode potential. The initial cathode potential was set at -0.6 V vs standard hydrogen 422
electrode (SHE, all potentials below, unless specified, were all versus standard 423
hydrogen electrode). When the current density of MNP bioreactors showed no further 424
change or started falling, all the reactors were disconnected from the potentiostat for 425
the first break. The cyclic voltammetry (CV) was conducted immediately with a scan 426
rate of 5 mV/s and scan range of -0.8 V to 0 V vs SHE to characterize biocathodes. 427
Microbial samples were also collected immediately from the cathodes and catholyte 428
medium (see details below). The reactors were then reconnected to the potentiostat 429
with the cathode potential elevated to -0.5 V and operated until the current density of 430
MNP bioreactors started falling again. The second break was then applied for CV 431
measurement and microbial sampling. Finally, the cathode potential was raised to -0.4 432
V for the last round of electrochemical operation. At the end, the last CV tests and 433
microbial sampling were performed again. 434
During each break, the chambers were rinsed thoroughly to remove remaining 435
microbes and magnetite, then replenished with the fresh medium. Cathode chambers 436
were inoculated by 10% (v/v) of the used medium and the remaining used medium 437
was discarded. The electrodes with biofilm remained unchanged. Magnetite 438
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
22
nanoparticles (MNP) were re-supplemented in the MNP reactors with the same 439
concentration as before. Microbial sampling and medium refreshing were performed 440
in an anaerobic glove box. There were nine reactors in total at the beginning, i.e. three 441
each for MNP bio-reactors, no-MNP bio-reactors and MNP abiotic reactors. But 442
during the first-round operation at -0.6 V, one of the no-MNP biotic reactors failed to 443
produce CH4 after 50 days and hence this reactor was suspended. Thus, there were 444
eight functioning reactors in operation throughout the experiment. 445
Chemical analyses 446
Methane and hydrogen were monitored throughout experiments. Gas samples (200 447
µL) were regularly taken from the headspace of cathode chambers with a 448
Pressure-Lok precision analytical syringe (Bation Rouge, LA, USA). The 449
concentration of CH4 and H2 were analyzed using GC-7890B gas chromatograph 450
(Agilent Technologies, Santa Clara, CA, USA) equipped with flame ionization 451
detector. The detection limits are 5 Pa for both CH4 and H2. Every time after gas 452
sampling, chambers were shaken vigorously for one minute. 453
Microbial community analysis 454
At the end of CV tests, an aliquot of culture medium was collected, while small pieces 455
of carbon brushes were cut off from electrodes (without influencing the electrode 456
integrity). Carbon brush samples were washed twice with sterile demineralized water 457
to remove the loosely attached microbial cells. Microbial DNA were extracted from 458
both culture medium and carbon brush samples. FastDNATM
SPIN Kit (MP 459
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
23
Biomedicals, Irvine, CA, USA) was used for DNA extraction according to the 460
manufacturer’s instructions. Archaeal DNA was amplified using the 16S rRNA gene 461
primers 1106F (TTWAGTCAGGCAACGAGC) and 1378R 462
(TGTGCAAGGAGCAGGGAC) (49). The bacterial DNA was amplified using the 463
16S rRNA gene primers 515F (GTGCCAGCMGCCGCGGTAA) and 806R 464
(GGACTACHVGGGTWTCTAAT). Amplicon sequencing was completed by 465
Novogene (Beijing, China) using Ion S5XL platform. About 80000 reads were 466
obtained for each sample and were clustered into operational taxonomic units (OTUs) 467
with a similarity threshold of 97%. Nonmetric multidimensional scale (NMDS) was 468
conducted using R and vegan community ecology package (50). The closest matching 469
sequences were identified by searching with the BLAST program in the NCBI 470
database (51). Neighbor-joining phylogenetic trees were constructed using MEGA-X 471
and presented with ggtree package in R (52, 53). The nucleotide sequences generated 472
from this study has been deposited in SAR database under the accession numbers of 473
SAMN13136352–SAMN13136413. 474
475
ACKNOWLEDGEMENTS 476
This study was financially supported by the National Natural Science Foundation of 477
China (No. 41630857; 91951206). 478
CONFLICT OF INTEREST 479
The authors declare that they have no conflict of interest. 480
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
24
SUPPLEMENTAL MATERIAL 481
Sequence summary for amplicon sequencing of microbes, total abundance of 482
methanogens and bacteria in electromethanogenic reactors, relative abundance of 483
bacterial community, nonmetric multidimensional scale (NMDS) plot of bacteria, 484
CH4 production, current density generation and CV tests in replicate or triplicate 485
reactors. 486
REFERENCES 487
1. Blasco-Gomez R, Batlle-Vilanova P, Villano M, Balaguer MD, Colprim J, 488
Puig S. 2017. On the Edge of Research and Technological Application: A 489
Critical Review of Electromethanogenesis. Int J Mol Sci 490
18.doi:10.3390/ijms18040874 491
2. Holmes DE, Smith JA. 2016. Biologically Produced Methane as a Renewable 492
Energy Source, p 1-61. In Sariaslani S, Gadd GM (ed), Adv Appl Microbiol, 493
Vol 97, vol 97. Elsevier Academic Press Inc, San Diego. 494
3. Logan BE, Rabaey K. 2012. Conversion of Wastes into Bioelectricity and 495
Chemicals by Using Microbial Electrochemical Technologies. Science 496
337:686-690.doi:10.1126/science.1217412 497
4. Cheng SA, Xing DF, Call DF, Logan BE. 2009. Direct Biological Conversion 498
of Electrical Current into Methane by Electromethanogenesis. Environ Sci 499
Technol 43:3953-3958.doi:10.1021/es803531g 500
5. Deutzmann JS, Sahin M, Spormann AM. 2015. Extracellular Enzymes 501
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
25
Facilitate Electron Uptake in Biocorrosion and Bioelectrosynthesis. Mbio 502
6.doi:10.1128/mBio.00496-15 503
6. Dykstra CM, Pavlostathis SG. 2017. Methanogenic Biocathode Microbial 504
Community Development and the Role of Bacteria. Environ Sci Technol 505
51:5306-5316.doi:10.1021/acs.est.6b04112 506
7. Lohner ST, Deutzmann JS, Logan BE, Leigh J, Spormann AM. 2014. 507
Hydrogenase-independent uptake and metabolism of electrons by the archaeon 508
Methanococcus maripaludis. ISME J 8:1673-1681.doi:10.1038/ismej.2014.82 509
8. Rowe AR, Xu S, Gardel E, Bose A, Girguis P, Amend JP, El-Naggar MY. 510
2019. Methane-Linked Mechanisms of Electron Uptake from Cathodes by 511
Methanosarcina barkeri. Mbio 10:12.doi:10.1128/mBio.02448-18 512
9. Yee MO, Rotaru A-E. 2020. Extracellular electron uptake in 513
Methanosarcinales is independent of multiheme c-type cytochromes. Sci Rep 514
10:372-372.doi:10.1038/s41598-019-57206-z 515
10. Yee MO, Snoeyenbos-West OL, Thamdrup B, Ottosen LDM, Rotaru A-E. 516
2019. Extracellular Electron Uptake by Two Methanosarcina Species. Front 517
Energy Res 7.doi:10.3389/fenrg.2019.00029 518
11. Batlle-Vilanova P, Puig S, Gonzalez-Olmos R, Vilajeliu-Pons A, Baneras L, 519
Dolors Balaguer M, Colprim J. 2014. Assessment of biotic and abiotic 520
graphite cathodes for hydrogen production in microbial electrolysis cells. Int J 521
Hydrogen Energ 39:1297-1305.doi:10.1016/j.ijhydene.2013.11.017 522
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
26
12. Mitov M, Chorbadzhiyska E, Rashkov R, Hubenova Y. 2012. Novel 523
nanostructured electrocatalysts for hydrogen evolution reaction in neutral and 524
weak acidic solutions. Int J Hydrogen Energ 525
37:16522-16526.doi:10.1016/j.ijhydene.2012.02.102 526
13. Villano M, Aulenta F, Ciucci C, Ferri T, Giuliano A, Majone M. 2010. 527
Bioelectrochemical reduction of CO2 to CH4 via direct and indirect 528
extracellular electron transfer by a hydrogenophilic methanogenic culture. 529
Bioresource Technol 101:3085-3090.doi:10.1016/j.biortech.2009.12.077 530
14. Batlle-Vilanova P, Puig S, Gonzalez-Olmos R, Vilajeliu-Pons A, Balaguer 531
MD, Colprim J. 2015. Deciphering the electron transfer mechanisms for 532
biogas upgrading to biomethane within a mixed culture biocathode. Rsc Adv 533
5:52243-52251.doi:10.1039/c5ra09039c 534
15. Saheb-Alam S, Persson F, Wilen BM, Hermansson M, Modin O. 2019. A 535
variety of hydrogenotrophic enrichment cultures catalyse cathodic reactions. 536
Sci Rep 9:13.doi:10.1038/s41598-018-38006-3 537
16. Siegert M, Yates MD, Spormann AM, Logan BE. 2015. Methanobacterium 538
Dominates Biocathodic Archaeal Communities in Methanogenic Microbial 539
Electrolysis Cells. ACS Sustain Chem Eng 540
3:1668-1676.doi:10.1021/acssuschemeng.5b00367 541
17. Deutzmann JS, Spormann AM. 2017. Enhanced microbial electrosynthesis by 542
using defined co-cultures. ISME J 11:704-714.doi:10.1038/ismej.2016.149 543
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
27
18. Beese-Vasbender PF, Grote JP, Garrelfs J, Stratmann M, Mayrhofer KJJ. 2015. 544
Selective microbial electrosynthesis of methane by a pure culture of a marine 545
lithoautotrophic archaeon. Bioelectrochemistry 546
102:50-55.doi:10.1016/j.bioelechem.2014.11.004 547
19. Rotaru AE, Shrestha PM, Liu F, Markovaite B, Chen S, Nevin KP, Lovley DR. 548
2014. Direct Interspecies Electron Transfer between Geobacter 549
metallireducens and Methanosarcina barkeri. Appl Environ Microb 550
80:4599-4605.doi:10.1128/aem.00895-14 551
20. Dykstra CM, Paylostathis SG. 2017. Zero-Valent Iron Enhances Biocathodic 552
Carbon Dioxide Reduction to Methane. Environ Sci Technol 553
51:12956-12964.doi:10.1021/acs.est.7b02777 554
21. Kim KR, Kang J, Chae KJ. 2017. Improvement in methanogenesis by 555
incorporating transition metal nanoparticles and granular activated carbon 556
composites in microbial electrolysis cells. Int J Hydrogen Energ 557
42:27623-27629.doi:10.1016/j.ijhydene.2017.06.142 558
22. Liu DD, Roca-Puigros M, Geppert F, Caizan-Juanarena L, Ayudthaya SPN, 559
Buisman C, ter Heijne A. 2018. Granular Carbon-Based Electrodes as 560
Cathodes in Methane-Producing Bioelectrochemical Systems. Front Bioeng 561
Biotech 6:10.doi:10.3389/fbioe.2018.00078 562
23. Cornell RM, Schwertmann U. 2004. Electronic, Electrical and Magnetic 563
Properties and Colour, p 111-137, The Iron Oxides. 564
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
28
24. Hazen RM, Papineau D, Leeker WB, Downs RT, Ferry JM, McCoy TJ, 565
Sverjensky DA, Yang H. 2008. Mineral evolution. Am Mineral 566
93:1693-1720.doi:10.2138/am.2008.2955 567
25. Li HJ, Chang JL, Liu PF, Fu L, Ding DW, Lu YH. 2015. Direct interspecies 568
electron transfer accelerates syntrophic oxidation of butyrate in paddy soil 569
enrichments. Environ Microbiol 17:1533-1547.doi:10.1111/1462-2920.12576 570
26. Xia XX, Zhang JC, Song TZ, Lu YH. 2019. Stimulation of 571
Smithella-dominating propionate oxidation in a sediment enrichment by 572
magnetite and carbon nanotubes. Env Microbiol Rep 573
11:236-248.doi:10.1111/1758-2229.12737 574
27. Zhang JC, Lu YH. 2016. Conductive Fe3O4 Nanoparticles Accelerate 575
Syntrophic Methane Production from Butyrate Oxidation in Two Different 576
Lake Sediments. Front in Microbiol 7.doi:10.3389/fmicb.2016.01316 577
28. Kato S, Hashimoto K, Watanabe K. 2012. Methanogenesis facilitated by 578
electric syntrophy via (semi)conductive iron-oxide minerals. Environ 579
Microbiol 14:1646-1654.doi:10.1111/j.1462-2920.2011.02611.x 580
29. Viggi CC, Rossetti S, Fazi S, Paiano P, Majone M, Aulenta F. 2014. Magnetite 581
Particles Triggering a Faster and More Robust Syntrophic Pathway of 582
Methanogenic Propionate Degradation. Environ Sci Technol 583
48:7536-7543.doi:10.1021/es5016789 584
30. Zhuang L, Ma JL, Yu Z, Wang YQ, Tang J. 2018. Magnetite accelerates 585
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
29
syntrophic acetate oxidation in methanogenic systems with high ammonia 586
concentrations. Microb Biotechnol 11:710-720.doi:10.1111/1751-7915.13286 587
31. Skomurski FN, Kerisit S, Rosso KM. 2010. Structure, charge distribution, and 588
electron hopping dynamics in magnetite (Fe3O4) (100) surfaces from first 589
principles. Geochim Cosmochim Ac 590
74:4234-4248.doi:10.1016/j.gca.2010.04.063 591
32. Sanchez O, Garrido L, Forn I, Massana R, Ignacio Maldonado M, Mas J. 592
2011. Molecular characterization of activated sludge from a 593
seawater-processing wastewater treatment plant. Microb Biotechnol 594
4:628-642.doi:10.1111/j.1751-7915.2011.00256.x 595
33. Conrad R. 2009. The global methane cycle: recent advances in understanding 596
the microbial processes involved. Env Microbiol Rep 597
1:285-292.doi:10.1111/j.1758-2229.2009.00038.x 598
34. Ren GP, Chen P, Yu J, Liu JB, Ye J, Zhou SG (2019) Recyclable 599
magnetite-enhanced electromethanogenesis for biomethane production from 600
wastewater. Water Res 166. Doi:10.1016/j.watres.2019.115095 601
35. Fu L, Song TZ, Zhang W, Zhang J, Lu YH. 2018. Stimulatory Effect of 602
Magnetite Nanoparticles on a Highly Enriched Butyrate-Oxidizing 603
Consortium. Front Microbiol 9.doi:10.3389/fmicb.2018.01480 604
36. Byrne JM, van der Laan G, Figueroa AI, Qafoku O, Wang C, Pearce CI, 605
Jackson M, Feinberg J, Rosso KM, Kappler A. 2016. Size dependent 606
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
30
microbial oxidation and reduction of magnetite nano- and micro-particles. Sci 607
Rep 6.doi:10.1038/srep30969 608
37. Byrne JM, Klueglein N, Pearce C, Rosso KM, Appel E, Kappler A. 2015. 609
Redox cycling of Fe(II) and Fe(III) in magnetite by Fe-metabolizing bacteria. 610
Science 347:1473-1476.doi:10.1126/science.aaa4834 611
38. Wang H, Byrne JM, Liu PF, Liu J, Dong X, Lu YH. 2020. Redox cycling of 612
Fe(II) and Fe(III) in magnetite accelerates aceticlastic methanogenesis by 613
Methanosarcina mazei. Env Microbiol Rep 614
12:97-109.doi:10.1111/1758-2229.12819 615
39. Van Eerten-Jansen MCAA, Veldhoen AB, Plugge CM, Stams AJM, Buisman 616
CJN, Ter Heijne A. 2013. Microbial Community Analysis of a 617
Methane-Producing Biocathode in a Bioelectrochemical System. Archaea 618
2013.doi:10.1155/2013/481784 619
40. Walker DJF, Martz E, Holmes DE, Zhou ZM, Nonnenmann SS, Lovley DR. 620
2019. The Archaellum of Methanospirillum hungatei Is Electrically 621
Conductive. Mbio 10.doi:10.1128/mBio.00579-19 622
41. Nevin KP, Woodard TL, Franks AE, Summers ZM, Lovley DR. 2010. 623
Microbial Electrosynthesis: Feeding Microbes Electricity To Convert Carbon 624
Dioxide and Water to Multicarbon Extracellular Organic Compounds. Mbio 625
1.doi:10.1128/mBio.00103-10 626
42. Marshall CW, Ross DE, Fichot EB, Norman RS, May HD. 2012. 627
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
31
Electrosynthesis of Commodity Chemicals by an Autotrophic Microbial 628
Community. Appl Environ Microb 78:8412-8420.doi:10.1128/aem.02401-12 629
43. Yamada T, Sekiguchi Y, Hanada S, Imachi H, Ohashi A, Harada H, Kamagata 630
Y. 2006. Anaerolinea thermolimosa sp nov., Levilinea saccharolytica gen. 631
nov., sp nov and Leptolinea tardivitalis gen. nov., so. nov., novel filamentous 632
anaerobes, and description of the new classes Anaerolineae classis nov and 633
Caldilineae classis nov in the bacterial phylum Chloroflexi. Int J Syst Evol 634
Micr 56:1331-1340.doi:10.1099/ijs.0.64169-0 635
44. Yamada T, Imachi H, Ohashi A, Harada H, Hanada S, Kamagata Y, Sekiguchi 636
Y. 2007. Bellilinea caldifistulae gen. nov., sp nov and Longilinea arvoryzae 637
gen. nov., sp nov., strictly anaerobic, filamentous bacteria of the phylum 638
Chloroflexi isolated from methanogenic propionate-degrading consortia. Int J 639
Syst Evol Micr 57:2299-2306.doi:10.1099/ijs.0.65098-0 640
45. Yates MD, Siegert M, Logan BE. 2014. Hydrogen evolution catalyzed by 641
viable and non-viable cells on biocathodes. Int J Hydrogen Energ 642
39:16841-16851.doi:10.1016/j.ijhydene.2014.08.015 643
46. Lian J, Tian XL, Guo JB, Guo YK, Song YY, Yue L, Wang YJ, Liang XH. 644
2016. Effects of resazurin on perchlorate reduction and bioelectricity 645
generation in microbial fuel cells and its catalysing mechanism. Biochem Eng 646
J 114:167-175.doi:10.1016/j.bej.2016.06.028 647
47. Kaden J, Galushko AS, Schink B. 2002. Cysteine-mediated electron transfer in 648
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
32
syntrophic acetate oxidation by cocultures of Geobacter sulfurreducens and 649
Wolinella succinogenes. Arch Microbiol 650
178:53-58.doi:10.1007/s00203-002-0425-3 651
48. Kang YS, Risbud S, Rabolt JF, Stroeve P. 1996. Synthesis and characterization 652
of nanometer-size Fe3O4 and gamma-Fe2O3 particles. Chem Mater 653
8:2209-&.doi:10.1021/cm960157j 654
49. Zhang JW, Tang HY, Zhu JG, Lin XG, Feng YZ. 2016. Divergent responses of 655
methanogenic archaeal communities in two rice cultivars to elevated 656
ground-level O-3. Environ Pollut 657
213:127-134.doi:10.1016/j.envpol.2016.01.062 658
50. Dixon P. 2003. VEGAN, a package of R functions for community ecology. J 659
Veg Sci 14:927-930.doi:10.1658/1100-9233(2003)014[0927:vaporf]2.0.co;2 660
51. Lyu Z, Lu YH. 2018. Metabolic shift at the class level sheds light on 661
adaptation of methanogens to oxidative environments. ISME J 662
12:411-423.doi:10.1038/ismej.2017.173 663
52. Yu GC, Smith DK, Zhu HC, Guan Y, Lam TT-Y. 2017. GGTREE: an R 664
package for visualization and annotation of phylogenetic trees with their 665
covariates and other associated data. Methods Ecol Evol 666
8(1):28-36.doi:10.1111/2041-210x.12628 667
53. Yu GC, Lam TT-Y, Zhu HC, Guan Y. 2018. Two Methods for Mapping and 668
Visualizing Associated Data on Phylogeny Using Ggtree. Mol Biol Evol 669
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
33
35(12):3041-3043.doi:10.1093/molbev/msy194 670
671
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
34
FIGURE CAPTIONS 672
673
Fig. 1 The schematic figure for enrichment cultivation of methanogens and the set of 674
electrochemical reactors. 675
676
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
35
677
Fig. 2 The production of CH4 and current density generation. The initial potential was 678
set at -0.6 V and then was elevated to -0.5 V and -0.4 in two steps (yellow arrows). 679
Data shown is a representative example of replicate or triplicate experiments (n=2 or 680
3), data for replicate or triplicate experiments could be found in Fig. S5 and S6 681
682
0 70 80 90Time (Days)
Cu
rre
nt
de
ns
ity
(A
m)
Abiotic
MNP
0
5
10
15
20
25
30
35
0 10 20 30 40 50 60 70 80 90 100 110 120 130
CH
4 (m
mo
l L
-1)
MNP added
0
5
10
15
20
25
30
35
0 10 20 30 40 50 60 70 80 90 100 110 120 130
CH
4 (m
mo
l L
-1)
MNP added
a
b
-0.6 V -0.5 V -0.4 V
-0.6 V -0.5 V -0.4 V
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
36
683
Fig. 3 Cyclic voltammogram determined at the end of electromethanogenic operations 684
under three successively elevated cathodic potentials (a, b, c separately). Red, MNP 685
reactors; black, no-MNP reactors; blue, abiotic control reactors, which contained 686
MNP but without inoculum. L, M, H denote -0.6 V, -0.5 V and -0.4 V; n, magnetite 687
nanoparticles; A, abiotic control. Data shown is a representative example of replicate 688
experiments (n=2 or 3), data for replicate experiments could be found in Fig. S7 689
690
-0.8 -0.6 -0.4 -0.2 0.0
-150
-100
-50
0
50
100
150
200 LnLAnL
Cu
rre
nt
de
ns
ity
(A
m-3)
-0.6 V
E (V vs SHE)
-0.8 -0.6 -0.4 -0.2 0.0-600
-400
-200
0
200
400MnM
-0.5 V
E (V vs SHE)
-0.8 -0.6 -0.4 -0.2 0.0-400
-300
-200
-100
0
100
200 -0.4 V
E (V vs SHE)
HnH
a b c
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
37
691
Fig. 4 Composition dynamics of methanogens during the operation of 692
electromethanogenesis. The relative abundance of methanogen community at genera 693
level (a). Nonmetric multidimensional scale (NMDS) plot of methanogen community 694
across samples (b). L, no-MNP biocathode at -0.6 V; nL, MNP biocathode at -0.6 V; 695
cL, no-MNP catholyte at -0.6 V; ncL, MNP catholyte at -0.6 V; M, no-MNP 696
biocathode at -0.5 V; nM, MNP biocathode at -0.5 V; cM, no-MNP catholyte at -0.5 697
V; ncM, MNP catholyte at -0.5 V; H, no-MNP biocathode at -0.4 V; nH, MNP 698
biocathode at -0.4 V; cH, no-MNP catholyte at -0.4 V; ncH, MNP catholyte at -0.4 V. 699
Error bars represent mean values ± one standard deviation, n=2 or 3 700
701
a b
Inocu
lum
L cL M cM H cH nL ncL nM ncM nH ncH
Re
lati
ve
ab
un
da
nc
eMethanospirillum
Methanobacterium
Methanosarcina
Methanoregula
Methanosphaerula
Methanocella
Methanobrevibacter
Methanosaeta
Methanolobus
Methanosphaera
Methanimicrococcus
Others0.0
0.2
0.4
0.6
0.8
1.0
No-MNP reactors MNP reactors
NMDS1
H
cH
Inoculum
nH
nM
M
cM
nL
L
cL
-0.4
V
-0.5
V
-0.6
V
0.4
0.2
0.0
-0.2
ncM
ncH
ncL
NM
DS
2
-0.50 -0.25 0.00 0.25
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
38
702
Fig. 5 Neighbor-joining phylogenetic relationship of methanogens (a) and heatmap 703
showing the change of phylotypes within Methanospirillum with the elevation of 704
cathode potential (b). nL, MNP biocathode at -0.6 V; nM, MNP biocathode at -0.5 V; 705
nH, MNP biocathode at -0.4 V 706
707
clone_Mspi1
clone_Mspi2
clone_Mspi3
clone_Mspi4
clone_Mspi5
clone_Mspi6
clone_Mspi7
clone_Mspi8
clone_Mspi9
clone_Mspi10
clone_Mspi11
clone_Mspi12
clone_Mspi13
clone_Mspi14
clone_Mspi15
0
1
2
-2
-1
Inoculu
mnL1
nL2
nL3
nM
1nM
2nM
3nH
1nH
2nH
3
a b
Methanospirillum hungatei JF-1 (NR 074177.1)
Methanospirillum stamsii ps (NR 117705.1)
clone_Msph
Methanosphaerula palustris E1-9c (NR 074167.1)
Methanoregula boonei 6A8 (NR 074180.1) clone_Mre
Methanoregula formicica SMSP (NR 102441.1)
Methanobacterium formicicum MF (NR 115168.1) clone_Mba1
Methanobacterium palustre F (NR 041713.1)
Methanobacterium bryantii MOH (NR 042781.1)
clone_Mba2
Methanobacterium flexile GH (NR 116276.1)
clone_Msar Methanosarcina horonobensis HB-1 JCM 15518 (NR 112648.1)
Methanosarcina mazei DSM 2053 OCM 26 (NR 041956.1)
Methanosarcina acetivorans C2A (NR 074110.1)
Methanosarcina barkeri MS (NR 118371.1) clone_Mce
Methanocella arvoryzae
Methanocella conradii HZ254 (NR 118245.1)
Methanocella paludicola SANAE (NR 074192.1)
0.050
Methanospirillum hungatei JF-1 i (NR 074177.1)
Methanospirillum stamsii ps (NR 117705.1)
clone_MsarMethanosarcina horonobensis HB-1 JCM 15518 (NR 112648.1)
Methanosarcina mazei DSM 2053 OCM 26 (NR 041956.1)
Methanosarcina acetivorans C2A (NR 074110.1)A
Methanosarcina barkeri MS (NR 118371.1)
Methanobacterium formicicum MF (NR 115168.1)clone_Mba1
Methanobacterium palustre F (NR 041713.1)
Methanobacterium bryantii MOH (NR 042781.1)
clone_Mba2
Methanobacterium flexile GH (NR 116276.1)
clone_Mspi
(including 15 clones)
MRE50 (NR074232.1)
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
39
708
Fig. 6 Bacterial community composition and relative abundance. Error bars represent 709
mean values ± one standard deviation, n=2 or 3 710
711
712
0.00
0.25
0.50
0.75
1.00
Inoculum L cL M cM H cH nL ncLnM ncM nH ncH
Re
lati
ve
Ab
un
dan
ce
Others
Syntrophobacter
Desulfomicrobium
Verrucomicrobia
Actinobacteria
Gracilibacter
Alphaproteobacteria
Anaerolineaceae
Anaerovorax
Lentimicrobium
Desulfovibrio
Spirochaetes
Bacteroidales
Geobacter
Sulfuricurvum
Gammaproteobacteria
Acetobacterium
No-MNP reactors MNP reactors
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
Counter electrodeWork electrode Reference
electrode
Gas
sample
Liquid
sample
Anode
Cathode
!"##$%&'()
*)+,'-.%/0%123
4! (5%67.%7."#-8",.
95',+)"6.
:,.6"6.%/;<%123
=!>?@! (5%67.%7."#-8",.:,.6"6.%/;<%123
=!>?@! (5%67.%7."#-8",.
:,.6"6.%/;<%123=!>?@! (5%67.%7."#-8",.
95',+)"6. 95',+)"6. 95',+)"6.%
67.%,"67'#.%
,7"1A.B-
C5B(,7%67.%1.67"5'D.5-
/E7B..%-+,,.--(F.%6B"5-G.B3
:,6(F"6.%6'6")%1(,B'A(")%",6(F(6$%
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
0 70 80 90Time (Days)
Cu
rre
nt
de
ns
ity
(A
m)
Abiotic
MNP
0
5
10
15
20
25
30
35
0 10 20 30 40 50 60 70 80 90 100 110 120 130
CH
4 (m
mo
l L
-1)
MNP added
0
5
10
15
20
25
30
35
0 10 20 30 40 50 60 70 80 90 100 110 120 130
CH
4 (m
mo
l L
-1)
MNP added
a
b
-0.6 V -0.5 V -0.4 V
-0.6 V -0.5 V -0.4 V
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
-0.8 -0.6 -0.4 -0.2 0.0
-150
-100
-50
0
50
100
150
200 LnLAnL
Cu
rren
t d
en
sit
y (
A m
-3)
-0.6 V
E (V vs SHE)
-0.8 -0.6 -0.4 -0.2 0.0-600
-400
-200
0
200
400MnM
-0.5 V
E (V vs SHE)
-0.8 -0.6 -0.4 -0.2 0.0-400
-300
-200
-100
0
100
200 -0.4 V
E (V vs SHE)
HnH
a b c
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
a b
Inoculu
m L cL M cM H cH nL ncL nM ncM nH ncH
Re
lati
ve
ab
un
da
nc
e
Methanospirillum
Methanobacterium
Methanosarcina
Methanoregula
Methanosphaerula
Methanocella
Methanobrevibacter
Methanosaeta
Methanolobus
Methanosphaera
Methanimicrococcus
Others0.0
0.2
0.4
0.6
0.8
1.0
No-MNP reactors MNP reactors
NMDS1
H
cH
Inoculum
nH
nM
M
cM
nL
L
cL
-0.4
V
-0.5
V
-0.6
V
0.4
0.2
0.0
-0.2
ncM
ncH
ncL
NM
DS
2-0.50 -0.25 0.00 0.25
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
clone_Mspi1
clone_Mspi2
clone_Mspi3
clone_Mspi4
clone_Mspi5
clone_Mspi6
clone_Mspi7
clone_Mspi8
clone_Mspi9
clone_Mspi10
clone_Mspi11
clone_Mspi12
clone_Mspi13
clone_Mspi14
clone_Mspi15
0
1
2
-2
-1
Inoculu
mnL1
nL2
nL3
nM
1nM
2nM
3nH
1nH
2nH
3
a b
Methanospirillum hungatei JF-1 (NR 074177.1)
Methanospirillum stamsii ps (NR 117705.1)
clone_Msph
Methanosphaerula palustris E1-9c (NR 074167.1)
Methanoregula boonei 6A8 (NR 074180.1) clone_Mre
Methanoregula formicica SMSP (NR 102441.1)
Methanobacterium formicicum MF (NR 115168.1) clone_Mba1
Methanobacterium palustre F (NR 041713.1)
Methanobacterium bryantii MOH (NR 042781.1)
clone_Mba2
Methanobacterium flexile GH (NR 116276.1)
clone_Msar Methanosarcina horonobensis HB-1 JCM 15518 (NR 112648.1)
Methanosarcina mazei DSM 2053 OCM 26 (NR 041956.1)
Methanosarcina acetivorans C2A (NR 074110.1)
Methanosarcina barkeri MS (NR 118371.1) clone_Mce
Methanocella arvoryzae
Methanocella conradii HZ254 (NR 118245.1)
Methanocella paludicola SANAE (NR 074192.1)
0.050
Methanospirillum hungatei JF-1 ei (NR 074177.1)
Methanospirillum stamsii ps (NR 117705.1)
clone_MsarMethanosarcina horonobensis HB-1 JCM 15518 (NR 112648.1)
Methanosarcina mazei DSM 2053 OCM 26 (NR 041956.1)
Methanosarcina acetivorans C2A (NR 074110.1) C2A
Methanosarcina barkeri MS (NR 118371.1)
Methanobacterium formicicum MF (NR 115168.1) clone_Mba1
Methanobacterium palustre F (NR 041713.1)
Methanobacterium bryantii MOH (NR 042781.1)
clone_Mba2
Methanobacterium flexile GH (NR 116276.1)
clone_Mspi
(including 15 clones)
MRE50 (NR074232.1)
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint
0.00
0.25
0.50
0.75
1.00
Inoculu
m L cL M cM H cH nL ncLnM ncM nH ncH
Re
lati
ve
Ab
un
da
nc
e
Others
Syntrophobacter
Desulfomicrobium
Verrucomicrobia
Actinobacteria
Gracilibacter
Alphaproteobacteria
Anaerolineaceae
Anaerovorax
Lentimicrobium
Desulfovibrio
Spirochaetes
Bacteroidales
Geobacter
Sulfuricurvum
Gammaproteobacteria
Acetobacterium
No-MNP reactors MNP reactors
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 25, 2020. ; https://doi.org/10.1101/2020.11.24.397190doi: bioRxiv preprint