Microbial Electrosynthesis of Isobutyric, Butyric ... - ACS Publications

Subscriber access provided by University of Winnipeg Library

Article

Microbial Electrosynthesis of Isobutyric, Butyric, Caproic acids and corresponding Alcohols from Carbon dioxide Igor Vassilev, Paula Andrea Hernandez, Pau Batlle Vilanova, Stefano Freguia, Jens Olaf Krömer, Jürg Keller, Pablo Ledezma, and Bernardino Virdis ACS Sustainable Chem. Eng., Just Accepted Manuscript • Publication Date (Web): 04 Jun 2018 Downloaded from http://pubs.acs.org on June 4, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sustainable Chemistry & Engineering

1

Microbial Electrosynthesis of Isobutyric, Butyric,

2

Caproic acids and corresponding Alcohols from

3

Carbon dioxide

4

Igor Vassilev,† Paula A. Hernandez,† Pau Batlle-Vilanova,‡ Stefano Freguia,† Jens O.

5

Krömer,§ Jürg Keller,† Pablo Ledezma† and Bernardino Virdis*†

6

†

7

Laboratories Building (60), Brisbane, QLD 4072, Australia

8

‡

9

Catalonia E-17071, Spain

Advanced Water Management Centre, The University of Queensland, Gehrmann

LEQUiA, Institute of the Environment, University of Girona, Campus Montilivi, Girona,

10

§

11

Permoserstraße 15, Leipzig 04318, Germany

12

*E-mail: [email protected]

13

KEYWORDS: Microbial electrosynthesis, Carbon-capture and utilisation, Mixed reactor

14

microbiome, Carbon chain elongation, Acetogenesis, Solventogenesis, Reverse β-oxidation,

15

Bio-isomerisation of butyrate

Department for Solar Materials, Helmholtz Centre for Environmental Research (UFZ),

16 17

ABSTRACT: Microbial electrosynthesis is potentially a sustainable biotechnology for the

18

conversion of the greenhouse gas CO2 into carboxylic acids, thus far mostly limited to acetic

19

acid (C2). In spite of the environmental benefits of recycling CO2 emissions to counter global

20

warming, bioelectrochemical production of acetate is not very attractive from an economic

21

point of view. Conversely, carboxylates and corresponding alcohols with longer C content

22

not only have a higher economical value compared to acetate, but they are also relevant 1 ACS Paragon Plus Environment

ACS Sustainable Chemistry & Engineering 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 31

23

platform chemicals and fuels used on a diverse array of industrial applications. Here, we

24

report on a specific mixed reactor microbiome capable of producing a mixture of C4 and C6

25

carboxylic acids (isobutyric, n-butyric and n-caproic acids) and their corresponding alcohols

26

(isobutanol, n-butanol and n-hexanol) using CO2 as the sole carbon source and reducing

27

power provided by an electrode. Metagenomic analysis supports the hypothesis of a

28

sequential carbon chain elongation process comprising of acetogenesis, solventogenesis and

29

reverse β-oxidation, and that isobutyric acid derived from the isomerisation of n-butyric acid.

30 31

INTRODUCTION

32

The industrial production of numerous chemical commodities is currently unsustainable since

33

it relies upon the use of nonrenewable fossil resources as feedstocks and/or to fulfil energy

34

requirements. Sugar-based production of chemicals from agricultural feedstocks such as corn

35

raises justifiable concerns regarding the appropriateness of using extensive acreages of land

36

resources in competition with food production.1 Therefore, research into the use of more

37

sustainable alternative feedstock is needed to support current demands for chemicals and

38

fuels without impacting on land use or resulting in additional carbon emissions.

39

The production of commodity chemicals and biofuels (mostly short- and medium-

40

chain carboxylic acids and their corresponding alcohols) has been demonstrated from non-

41

conventional resources via gas fermentation and organic waste fermentation.2-4 Gas

42

fermentation requires, in addition to CO2 as a feedstock, supply of hydrogen and/or carbon

43

monoxide as sources of reducing power. On the other hand, organic waste fermentation

44

depends on external supply of specific reduced compounds (e.g., methanol, ethanol or

45

lactate) to sustain conversion of organics to short- and medium-chain carboxylic acids, or

46

alternatively, the use of waste streams already containing these compounds (e.g., brewery

2 ACS Paragon Plus Environment

Page 3 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


47

waste streams containing ethanol).2 Microbial electrochemical technologies such as microbial

48

electrosynthesis (MES) offer potential opportunities for the sustainable production of

49

chemicals using CO2 as the sole carbonaceous feedstock.5-6 In MES, the reducing-power is

50

provided in situ by the cathodic terminal of bioelectrochemical system (BES), poised at a

51

suitable electrochemical potential to supplement the microorganisms (i.e., the biocatalysts)

52

with the required reducing equivalents to drive the bioconversion of CO2 into organic

53

molecules with higher C content (mostly short-chain carboxylic acids and corresponding

54

alcohols).7-8 In a real scale installation, the voltage required to run the process could be

55

provided by renewable sources such as photovoltaics or wind energy. Compared to other

56

biorefinery technologies, turning CO2 emissions into chemicals by MES has multiple

57

benefits. Relatively to syngas fermentation, MES can use streams poor in H2 but rich in CO2,

58

which are vastly available as waste in many industrial processes (e.g., power plants, cement

59

production, petrochemical industry, steel manufacturing),9 and thereby, fixation and

60

conversion of CO2 will have the added benefit of reducing the carbon footprint of these

61

industrial processes.

62

In spite of the environmental benefits, production rates and titers of chemicals

63

attainable by MES are still low and limited to short-chain carboxylic acids, mostly acetic

64

acid.10 Unfortunately, with a market value of ca. 500 € t-1, MES of acetate is not very

65

attractive from the economic point of view.6,

66

diversification and selection towards chemicals with a higher value is critical for the

67

economic viability of MES. Recently, the MES of isopropanol, propionate and butyrate from

68

CO2 by a mixed culture was reported, proving the potential of MES to produce longer chain

69

carboxylates with a value higher than acetate.12-15 Pure cultures have also been shown as able

70

to produce longer chain carboxylates (e.g., 2-oxobutyrate) from CO2 by MES.16

11

Therefore, research targeting product



Page 4 of 31

71

Here, we report on a specific mixed MES culture, enriched under particular

72

conditions that trigger the production of industrially relevant carboxylic acids, notably n-

73

butyric acid (widely used in pharmaceutical and chemical industries), and n-caproic acid (C6)

74

(broadly used as feed additive, flavour enhancer, antimicrobial agent and feedstock for

75

chemical and biofuel industries).2,

76

industrially relevant platform chemical isobutyric acid, a precursor in the production of

77

methacrylic acid, a valuable commodity chemical priced at 1,820 € t-1 and used in the

78

production of the transparent thermoplastic poly(methyl methacrylate), also known as acrylic

79

glass.11, 17 Furthermore, isobutyric acid is required for the synthesis of 2,2,4,4-Tetramethyl-

80

1,3-cyclobutanediol, a commercialised co-polyester for production of clear and bisphenol A-

81

free plastics with various applications (e.g., water bottles, kitchen- and house-wares, medical

82

equipment).18 Other uses of isobutyric acid include the manufacturing of resins, inks, paints,

83

beverage additives, coatings, adhesive agents, plasticizers, amongst others.19-20 Currently,

84

isobutyric acid is produced by the acid-catalysed Koch carbonylation reaction of propene.21

85

However, this production process has two important drawbacks: it is dependent on the use of

86

non-renewable fossil fuels like petroleum or natural gas to obtain propene. In addition, the

87

chemical synthesis route is an energy-intensive process and includes the use of

88

environmentally harmful compounds such as sulfuric acid, hydrogen fluoride and boron

89

fluoride.17 In order to find a sustainable alternative production route for isobutyric acid

90

production, research into microbial biosynthesis of this compound has been conducted.

91

Isobutyrate is an intermediate in the Ehrlich pathway, which was successfully engineered in

92

Escherichia coli and Pseudomonas taiwanensis VLB120∆C to produce isobutyric acid from

93

glucose with production titers of 4.8 g L-1 and 2.4 g L-1, respectively.20,

94

pathway was also engineered in the chemolithoautotroph Acidithiobacillus ferrooxidans,

95

thereby enabling the production of trace amounts (99.9%, BOC, Australia) was periodically sparged in the reactor to

136

ensure the supply of inorganic carbon (vide infra).

137

At the end of each SB, approximately 95% of the cathodic medium was drained and

138

replaced by fresh medium to start the following new SB. Each SB was considered finished

139

when no further increase in the production of C4 and C6 carboxylic acids were observed, as

140

per liquid-phase sample analysis.

141

The pH was monitored continuously by a pH/ORP transmitter (Liquisys CPM253,

142

Endress+Hauser, Germany). In fact, due to the non-optimal proton diffusion from the anodic

143

to the cathodic chamber as result of imperfect selectivity of the cation exchange membrane

144

towards protons, the pH in the anodic chamber was 1.9, while the pH in the cathodic chamber


Page 7 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


145

tended to increase. However, periodic CO2 gassing of the cathodic chamber enabled the pH

146

to decrease since dissolved CO2 dissociates in water into carbonic acid (H2CO3). CO2 was not

147

only used to regulate the pH in the cathodic chamber, but also to provide the microorganisms

148

with the carbon source required throughout the experiments. At the pH of around 5, most of

149

the inorganic carbon was in the form of dissolved CO2, thus available to the microbial

150

metabolism. In SB-I to -III, the CO2 was provided approximately every one to three days by

151

routinely sparging the catholyte for 15 minutes at a rate of 50 mL min-1. In SB-IV to -VI, a

152

CO2 mass-flow controller (El-FLOW Select, Bronkhorst, Netherlands) was installed and

153

connected to the pH/ORP transmitter which activated the gas sparging at a flow of 10 mL

154

min-1 when the pH of the catholyte reached values above the set-point of 5.2. Continuous

155

availability of CO2 as a substrate was confirmed by measuring the gas in the reactor

156

headspace (Text S1 and Figure S2). In addition, in SB-IV to -VI a peristaltic pump was

157

connected to the pH/ORP transmitter for automatic addition of a 1 M NaOH solution if the

158

pH dropped below 4.9 due to high production of organic acids.

159

Abiotic electrochemical control and biotic open circuit control. To gain

160

information on the biological nature of the process, an abiotic electrochemical control test

161

was performed by operating an identical BES under the same electrochemical conditions and

162

using the same anodic and cathodic mediums as for the main BES described above, but under

163

sterile conditions. In addition, to gain information on the bioelectrochemical nature of the

164

process, a biotic open-circuit control test was performed by inoculating and operating another

165

identical BES at open circuit, that is without reducing equivalents being provided to the

166

microorganisms by the cathode.

167

Analytics. Samples of the liquid-phase were taken routinely from the cathodic broth

168

(approximately every two to four days) and filtered immediately through a 0.22 µm pore

169

filter for analysis of volatile fatty acids (VFAs) and alcohols via gas chromatography (GC)



Page 8 of 31

170

and high performance liquid chromatography (HPLC).27-28 The methods are described in

171

detail in the supporting information (Text S1). At the end of each SB, the filtered samples

172

were analysed for total organic carbon (TOC) via a TOC analyser as described in Text S1.27

173

Anaerobic conditions, methanogenesis inhibition and CO2 availability were confirmed by

174

routine analysis of gas samples taken with a glass syringe from the reactor headspace via GC

175

(for method details see Text S1).29

176

Characterisation of the microbiome. At the end of SB-III to -VI, samples of

177

planktonic cells (i.e., cells floating in the liquid medium) and biofilm (i.e., growing on the

178

surface of graphite granules) were harvested and prepared for metagenomic analysis as

179

following: For planktonic cells, 60 mL of the reactor broth was centrifuged (4000 rpm, 20

180

min) and the resulted pellet was used for DNA extraction. For the biofilm, ca. 20 granules

181

were taken from the reactor, carefully washed with fresh medium and kept for five minutes in

182

an ultrasonic bath (Unisonics, Australia). After vortexing, the granules were removed, the

183

liquid sample was centrifuged in the same way as the planktonic cells and the obtained pellet

184

was used for DNA extraction.

185

The cells were lysed via bead-beating for 5 minutes with 0.1 mm zirconia beads (Mo-

186

Bio PowerLyzer, USA). The DNA was extracted from swabs (Copan, 4N6FLOQ) using a

187

Faecal Extraction Kit (Chemagic, PerkinElmer) following the manufacturer’s protocol. DNA

188

was quantitated with a Qubit broad range assay (ThermoFisher Scientific; Qubit 3.0

189

Fluorometer and Qubit dsDNA BR Assay Kit) as per the manufacturer’s protocol.

190

The libraries were prepared following the procedures described in the manufacturer’s

191

protocol using Nextera XT Library Preparation Kit (Illumina, USA). For quality control and

192

quantification, each library was assessed using an Agilent D5000 HS tap. Then, successful

193

libraries were pooled in equimolar amounts for sequencing. On completion, pooled libraries

194

were quantified using qPCR on the ViiA7 platform (Applied Biosystems, USA) and analysed


Page 9 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


195

with ViiA7 v1.2 software. Library QC was performed using the Agilent HS Bioanalyzer kit

196

following Agilent’s protocol. Successfully pooled libraries were denatured for sequencing on

197

the NextSeq500 system using the manufacturer’s protocol and a High Output 2 x 150bp PE

198

Sequencing reagent cartridge (Illumina, USA). Following that procedure, 2 gigabases of data

199

per sample were generated.

200

The metagenome was annotated to establish functional gene orthologs via KEGG

201

(Kyoto Encyclopedia of Genes and Genomes) Orthology database to prove if required

202

enzymes for the putative pathways could be potentially expressed. The reads were aligned

203

with DIAMOND (https://ab.inf.uni-tuebingen.de/software/diamond) to Uniref100,30 and the

204

top hits were mapped to the KEGG Orthology database (http://www.genome.jp/kegg/) using

205

UniProt ID mapping service (http://www.uniprot.org/).31-32 In addition, DIAMOND was used

206

to blast open reading frames (ORFs) against UniProt100 (containing UniRef100_D2BR82

207

Alpha-ketoisovalerate decarboxylase n=1 Tax=Lactococcus lactis subsp. lactis (strain

208

KF147), TaxID=684738, RepID=D2BR82_LACLK). The taxonomy of the mixed culture

209

was defined to the genus level via CommunityM, a tool for identifying, classifying and

210

assembling 16S reads in metagenomic data (https://github.com/dparks1134/CommunityM).33

211 212

Calculations. All calculations used in this study are detailed in the Supporting Information, Text S1.

213 214

RESULTS AND DISCUSSION

215

The BES was operated for a total of 462 days during which a fixed potential of -0.8 V was

216

applied to the cathode electrode. As a result, cathodic current production (i.e., negative in

217

value) was observed. Figure 1A depicts the current vs. time trace measured during the reactor

218

operations. While a large variability could be observed, especially in the first three SBs, an

219

average current density of -0.9 ± 0.3 mA cm-2 was delivered. 9 ACS Paragon Plus Environment


Page 10 of 31

220

221 222 223 224 225 226 227 228 229

Figure 1. Microbial electrosynthesis by an open reactor microbiome. (A) Profiles of current density, charge transferred at the electrode and charge recovered as multi-carbon products (calculations detailed in Text S1) vs. the time. (B) Concentration vs. time of volatile fatty acids, VFAs, and (C) alcohols. The bioelectrochemical system was operated for a period of 462 days during which six semi-batch (SB) tests were performed. The potential of the working electrode (the cathode) was poised to -0.8 V vs. SHE and CO2 was provided as the sole carbon source.

230

During MES, the supplied electrons can be taken up by the microbial community

231

either in the form of hydrogen (electrochemically or bioelectrochemically produced), or

232

potentially through direct electron transfer from the cathode (Figure 2A, pink shaded area),34


Page 11 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


233

although our data do not provide evidence of direct electron transfer pathway. In fact, H2 was

234

continuously produced, which was confirmed by analysis of the composition of the reactor

235

head-space (Figure S2), and H2 acted putatively as an electron carrier, suggesting indirect

236

electron transfer took place. In this way, H2 provided by the cathode allowed the microbial

237

community to transform CO2 to multi-carbon compounds (C2 to C6 carboxylates and

238

corresponding alcohols) with a charge recovery in products increasing from 0.08 MC

239

(megacoulomb) in SB-I to 0.21 MC in SB-VI (Figure 1A), equivalent to a charge efficiency

240

of 7.1% and 44.0% (Table S1), respectively (see Text S1 for calculations). Recovery of

241

electrons into biomass could not be determined, as the biomass growth could not be

242

quantified in our system. Most likely, the supplied electrons not recovered in multi-carbon

243

products were vented out of the system as unused hydrogen (>50%).15 Another potential

244

inefficiency could derive from the presence of oxygen in the cathode chamber, which might

245

act as alternative electron sink thereby diverting electrons from CO2 reduction. In our system,

246

oxygen was produced through electrolysis of water in the anodic chamber and potentially, it

247

could reach the cathode by diffusion through the cation exchange membrane.6 While no

248

oxygen was detected in the headspace of the cathode chamber (Figure S2), it is possible that

249

direct electroreduction at the electrode and/or microbial metabolism have kept its levels

250

below detection during the reactor operations.

251

Figure 1B and 1C detail the concentration profiles of the products observed in the

252

BES. The measured multi-carbon products covered 97 ± 14% of the total organic carbon

253

(Table S1), indicating that most biosynthesised products as shown in Figure 1B and C were

254

taken into account. The concentration of all summed products increased by approximately 1.3

255

times in each SB over the course of the reactor operations reaching a maximum level of 422.6

256

mM-C (mM of carbon) in SB VI. We assume that the increase of the concentration of the

257

products in each SB is due to the adaptation of the microorganisms over the long-term



Page 12 of 31

258

operations to the given conditions and became more tolerant to a higher concentration of

259

VFAs and alcohols.35 No organic products were measured in control experiments performed

260

in the absence of energy supply or catalytic biomass, confirming the role of energy supply in

261

the production process as well as its biological nature (Figure S3).

262

Figure 1B shows that acetate was the main product of microbial electrosynthesis, with

263

titers reaching a maximum of 98.5 ± 6.2 mM-C (equivalent to 49.3 ± 3.1 mM, 2.9 ± 0.2 g L-1)

264

across SBs-I to -V. In SB-VI, acetate concentration increased to a maximum of 165.0 mM-C

265

(equivalent to 82.5 mM, 4.9 g L-1), at a net volumetric production rate of 4.8 mM-C d-1.

266

While production of carboxylates with longer chain was minimal in SB-I, with only up to

267

30.0 mM-C of isobutyrate (equivalent to 7.5 mM, 0.7 g L-1) and 16.0 mM-C of butyrate

268

(equivalent to 4.0 mM, 0.4 g L-1), a steady increase in C4 compounds was observed starting

269

from SB-II (Figure 1B). Interestingly, production of butyrate and isobutyrate occurred

270

simultaneously, with butyrate becoming the dominant C4 carboxylate from SB-III onwards.

271

A maximum butyrate titer of 142.8 mM-C (equivalent to 35.7 mM, 3.1 g L-1) was observed in

272

SB-VI, and a maximum isobutyrate titer of 74.1 mM-C (equivalent to 18.5 mM, 1.6 g L-1)

273

was measured in the same SB. The net production rates for C4 carboxylates in SB-VI was

274

determined as 3.3 mM-C d-1 and 1.9 mM-C d-1 for butyrate and isobutyrate, respectively.

275

Caproic acid (C6) started to be detected in the cathodic broth in SB-III when a maximum of

276

27.9 mM-C (equivalent to 4.7 mM, 0.5 g L-1) was measured. Interestingly, its production

277

increased as reactor operations progressed, and in SB-VI up to 64.2 mM-C (equivalent to

278

10.7 mM, 1.2 g L-1) was measured, at a net volumetric production rate of 2.0 mM-C d-1.

279

These are some of the highest titers of butyrate, isobutyrate, and caproate observed in a BES

280

fed with CO2 as the sole C source.

281

Alongside the production of carboxylic acids, the production of their corresponding

282

alcohols was also observed during reactor operations (Figure 1C). Ethanol was the main


Page 13 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


283

alcohol produced during the initial SB, with levels reaching 56.1 mM-C (equivalent to 28.1

284

mM, 1.3 g L-1). Only small titers of butanol were detected in SB-I (< 10 mM-C).

285

Interestingly, the levels of ethanol decreased to a maximum of 18.6 mM-C in SB-II

286

(equivalent to 9.3 mM, 0.4 g L-1), and similar levels were maintained in the following SBs

287

(Figure 1C), while at the same time, production of butanol increased to 20.8 mM-C

288

(equivalent to 5.2 mM, 0.4 g L-1), hence doubling the levels observed in the first batch, and

289

remained similar in the following SB-III and -IV. An increase in butanol was measured in

290

SB-V and -VI, with a maximum titer of 44.0 mM-C (equivalent to 11.0 mM, 0.8 g L-1)

291

observed at day 444, one of the highest titers reported by an electrosynthesis microbiome.

292

Small, although increasing levels of isobutanol and hexanol were also observed during

293

reactor operations, with titers reaching respectively 11.3 mM-C (equivalent to 2.8 mM, 0.2 g

294

L-1) and 12.8 mM-C (equivalent to 2.1 mM, 0.2 g L-1) in SB-VI.

295

Notably, when at the end of the SB-VI we transferred some of the biomass

296

(granules-associated biomass and planktonic cells) from the primary reactor to a sterile BES

297

operated under identical conditions, after 42 days, the new BES displayed a product spectrum

298

comparable to that of the primary reactor in SB-IV (Figure S4), thus proving not only the

299

reproducibility of the performance, but also that the specific microbiome enriched under the

300

operating conditions applied was responsible for the range of products observed.

301 302

Analysis of the metagenome in samples taken from the planktonic and biofilm cells at

303

the end of SB-III, -IV, -V, and VI revealed that in spite of some temporal variability,

304

organisms belonging to the genus Clostridium dominated both the biofilm (in SB-III to -VI)

305

and the planktonic (in SB-V and -VI) microbial communities (Figure S5). The prevalence of

306

Clostridium spp. in our reactor microbiome is not surprising, and might explain the

307

production of acetate in our reactor (Figure 1B). In fact, several Clostridium spp. such as C.



Page 14 of 31

308

ljungdahlii, have been frequently reported in BES microbiomes as responsible for converting

309

CO2 to acetic acid.10

310

A putative metabolic pathway network based on the observed product spectrum

311

linked with the genomic information we collected on our system is compiled in Figure 2. The

312

metagenome data was used to establish functional gene orthologs via KEGG (Kyoto

313

Encyclopedia of Genes and Genomes) Orthology database (http://www.kegg.jp) to prove if

314

the enzymes required for the proposed pathways could potentially be expressed by our

315

microbiome (Text S3).

316 317 318 319 320 321 322 323 324 325 326 327 328 329

Figure 2. Putative metabolic pathways for microbial electrosynthesis identified in the reactor microbiome. (A) Cathodically-generated reducing power in a bioelectrochemical system via electrochemically or bioelectrochemically produced H2, or potentially via direct microbial electron uptake. (B) Potential pathways converting CO2 (substrate circled in green) into short and medium carbon chains (products circled in orange). Blue highlighted pathway shows Wood-Ljungdahl pathway for fixation of CO2 and synthesis of acetyl-CoA, which can be converted to acetate or ethanol (yellow highlighted). Green highlighted pathway represents the Ehrlich pathway for isobutyrate/isobutanol synthesis. Acetyl-CoA can also be used for butyrate/butanol and caproate/hexanol production via reverse β-oxidation, highlighted in pink and grey, respectively. Enzymes, which ortholog genes were confirmed in the microbial community genome, are highlighted with a yellow border. Dashed yellow border indicates that the gene associated with that particular enzyme was not confirmed in the genome. ACD, acyl-CoA dehydrogenase; ACK, acetate kinase; ACS, acetyl-CoA synthase; ADH, alcohol 14 ACS Paragon Plus Environment

Page 15 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


330 331 332 333 334 335 336 337 338 339 340 341

dehydrogenase; ADHE, aldehyde-alcohol dehydrogenase; ALDH, aldehyde dehydrogenase; ALS, acetolactate synthase; AOR, aldehyde ferredoxin oxidoreductase; ATO, acetyl-CoA Cacetyltransferase; BCD, butyryl-CoA dehydrogenase; BDH, butanol dehydrogenase; BUK, butyrate kinase; CoAT, acetate CoA/acetoacetate CoA transferase; CODH, carbon monoxide dehydrogenase; CRT, crotonase; Fd, ferrodoxin; FDH, formate dehydrogenase; FTS, formylTHF synthase; HAD, hydroxyacyl-CoA dehydrogenase; HBD, hydroxybutyryl-CoA dehydrogenase; HYD, hydrogenase; ICM, isobutyryl-CoA mutase; ILVC, ketol-acid reductoisomerase; ILVD, ketoisovalerate decarboxylase; KIVD, alpha-ketoisovalerate decarboxylase; MTC, methenyl-THF cyclohydrolase; MTD, methylene-THF dehydrogenase; MTR, methylene-THF reductase; ox, oxidised; PFOR, pyruvate ferredoxin oxidoreductase; PTA, phosphate acetyltransferase; PTB, phosphate butyryltransferase; red, reduced.

342

Production of acetate is accomplished via acetogenesis through a multi-step metabolic

343

pathway starting with the reduction of two CO2 molecules and the synthesis of the central

344

metabolite acetyl-Coenzyme A (CoA) via the Wood-Ljungdahl pathway (Figure 2B,

345

blue-shaded area).10, 36 Since acetate has a higher degree of reduction than CO2 (+4 vs. 0), the

346

reducing equivalents to drive acetogenesis are most likely provided by the cathode in form of

347

electrochemically or biologically/enzymatically produced hydrogen, or potentially direct

348

microbial electron uptake (Figure 2A, pink-shaded area).34 Acetyl-CoA is then converted to

349

acetate via phosphorylation for ATP generation (Figure 2B, yellow-shaded area).10, 36 Genes

350

encoding for this pathway have all been detected in the metagenome associated with our

351

reactor (Text S3), and corroborate the results by Marshall et al., who performed a detailed

352

analysis of the metagenome of a MES microbiome dominated by Acetobacterium,

353

Sulfurospirillum, and Desulfovibrio, producing mostly acetate from CO2.37 However, in

354

contrast to the results of Marshall et al., the product spectrum observed in our reactor

355

included C2, C4 and C6 compounds. We hypothesise, as the acetate production progressed,

356

accumulation of the undissociated acetic acid in the culturing broth at low pH may reach

357

toxic levels and inhibit cellular metabolism.38-39 To prevent further accumulation of

358

undissociated acetic acid, Clostridium spp. can alter their metabolism from acetogenesis to

359

solventogenesis and re-assimilate acetate into its corresponding alcohol ethanol (Figure 2B,

360

yellow-shaded area).38,

40

In our experiment the onset of ethanol production was observed 15 ACS Paragon Plus Environment


Page 16 of 31

361

shortly after high levels of acetate were reached (close to the maximum), after which point,

362

acetate levels started to decrease while ethanol concentration increased (Figure 1B and 1C).

363

Genome annotation analysis confirmed the presence in the reactor microbiome of the genes

364

required to perform solventogenesis (Text S3).

365

The simultaneous presence of acetate and ethanol in cultures presenting Clostridium

366

spp. (e.g. C. kluyveri) can lead to longer chain carboxylates via the reverse β-oxidation chain

367

elongation pathway,41 which adds an acetyl-CoA derived from ethanol to acetate (C2),

368

thereby elongating to n-butyrate (C4), and then to n-butyrate to obtain n-caproate (C6).2

369

Production of n-butyrate was observed in the cathodic broth even during the SB-I (Figure

370

1B), with maximum titers increasing by 1.6 times each SB over the course of the reactor

371

operations. Caproic acid was also observed (Figure 1B), with maximum levels across the

372

batch tests increasing by 12.1 times from SB-II to the last SB. Genes encoding for enzymes

373

required for the reverse β-oxidation pathway were detected in the reactor microbiome, thus

374

suggesting that this pathway of production of C4 and C6 compounds was possible (Figure 2,

375

pink- and grey-shaded areas, and Text S3). Alternatively, a direct pathway for n-butyrate

376

(and n-caproate) production from CO2 via acetyl-CoA (and butanoyl CoA) without

377

intermediate production of acetate and ethanol (Figure 2B) should also be considered. In fact,

378

it was previously suggested that the strong reducing conditions provided in BES environment

379

may favour this direct pathway (the authors operated their system at a p(H2) of 1.2 atm).13

380

However, 1) the fact that acetate and ethanol are both consumed during the SB tests; 2) the

381

presence of a certain lag usually observed between the production of acetate (and ethanol)

382

and the production of C4-C6 products; and 3) the fact that our system was always operated at

383

p(H2)

Microbial Electrosynthesis of Isobutyric, Butyric ... - ACS Publications

Recommend Documents