One-Carbon Feedstocks for Sustainable Bioproduction (Advances in Biochemical Engineering/Biotechnology, 180) 303106853X, 9783031068539

Citation preview

Advances in Biochemical Engineering/Biotechnology  180 Series Editors: Thomas Scheper · Roland Ulber

An-Ping Zeng Nico J. Claassens   Editors

One-Carbon Feedstocks for Sustainable Bioproduction

180 Advances in Biochemical Engineering/Biotechnology Series Editors Thomas Scheper, Hannover, Germany Roland Ulber, Kaiserslautern, Germany Editorial Board Members Shimshon Belkin, Jerusalem, Israel Thomas Bley, Dresden, Germany Jörg Bohlmann, Vancouver, Canada Man Bock Gu, Seoul, Korea (Republic of) Wei Shou Hu, Minneapolis, MN, USA Bo Mattiasson, Lund, Sweden Lisbeth Olsson, Göteborg, Sweden Harald Seitz, Potsdam, Germany Ana Catarina Silva, Porto, Portugal An-Ping Zeng, Hamburg, Germany Jian-Jiang Zhong, Shanghai, Minhang, China Weichang Zhou, Shanghai, China

Aims and Scope This book series reviews current trends in modern biotechnology and biochemical engineering. Its aim is to cover all aspects of these interdisciplinary disciplines, where knowledge, methods and expertise are required from chemistry, biochemistry, microbiology, molecular biology, chemical engineering and computer science. Volumes are organized topically and provide a comprehensive discussion of developments in the field over the past 3–5 years. The series also discusses new discoveries and applications. Special volumes are dedicated to selected topics which focus on new biotechnological products and new processes for their synthesis and purification. In general, volumes are edited by well-known guest editors. The series editor and publisher will, however, always be pleased to receive suggestions and supplementary information. Manuscripts are accepted in English. In references, Advances in Biochemical Engineering/Biotechnology is abbreviated as Adv. Biochem. Engin./Biotechnol. and cited as a journal.

An-Ping Zeng • Nico J. Claassens Editors

One-Carbon Feedstocks for Sustainable Bioproduction With contributions by J. Bae
B. Blombach
T. Brautaset
V. R. Bysani
F.-X. Chang
T. H. T. Chau
B.-K. Cho
S. Cho
N. J. Claassens
B. Dronsella
B. J. Eikmanns
F. Enzmann
Q. Fei
F. Guo
S. Guo
S. Heux
D. Holtmann
Y. Hong
M. Jiang
Y. Jiang
S. Jin
S. Kang
G. Kosec
E. Y. Lee
H. Lee
S. N. Lindner
D. T. N. Nguyen
J. Nie
E. Orsi
I. Parera Olm
M. Pfitzer
V. Rainaldi
J. Ren
A. Satanowski
J. Shin
D. Siebert
Y. Song
D. Z. Sousa
M. Stöckl
J. Sun
W. Wang
Y. Wang
V. F. Wendisch
S. Wenk
D. Weuster-Botz
F. Xin
H. Xu
S. Yilmaz
Y.-C. Yong
W. Yuan
A.-P. Zeng
S. Zhang
W. Zhang
P. Zheng
F. Zhu
L. Zou

Editors An-Ping Zeng Institute for Bioprocess and Biosystems Engineering Hamburg University of Technology Hamburg, Germany

Nico J. Claassens Laboratory of Microbiology Wageningen University WAGENINGEN, The Netherlands

Center of Synthetic Biology and Integrated Bioengineering Westlake University Hangzhou, China

ISSN 0724-6145 ISSN 1616-8542 (electronic) Advances in Biochemical Engineering/Biotechnology ISBN 978-3-031-06853-9 ISBN 978-3-031-06854-6 (eBook) https://doi.org/10.1007/978-3-031-06854-6 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

In Memory of Arren Bar-Even

Arren Bar-Even was a scientist who was able to do chemistry in his head – who knew thermodynamics by heart. In his short but remarkable career, he (re-)designed metabolic pathways by combining a profound understanding of metabolism and its underlying chemical and thermodynamic principles with a high degree of creativity. This reflects in the titles of his theoretical publications with phrasings like “daring metabolic designs” or “rethinking glycolysis.” The chemical and mathematical logic he rigorously followed there later transformed into the synthetic, new-to-nature metabolic networks that came alive in his lab. By the time Arren had finished his Ph.D., he had already made a significant contribution to our basic understanding of the general features of enzymes and metabolic pathways, in a series of insightful meta-analyses on the design principles of metabolism. This deep grasp of the fundamentals of how metabolism operates and evolves was the basis for his extraordinary contributions to metabolic engineering and synthetic biology. The most representative examples of his achievements are the many carbon fixation pathways and photorespiration bypasses he designed, offering engineering blueprints for decades to come. Not only were his synthetic pathways more efficient than natural metabolism, they were also both elegant and esthetically beautiful, a powerful demonstration of human creativity. Among them, the most striking was the reductive glycine pathway for the assimilation of formate, which earned him the title “formate-guy” at many conferences. Researchers around the globe were inspired by his idea to engineer synthetic formatotrophy via this pathway and continue to work on the establishment of a circular one-carbon based bioeconomy. Arren’s legacy is not only his scientific achievements, but also his way of looking at and discussing things. Wherever he was, there was an open dialogue between everyone in the room without restrictions or hierarchies. If the seed he was planting fell on fertile ground, he would constantly challenge one while teaching his way of thinking science and specifically metabolism. His interpretation of being a PI was to be always there for you, at most, only a phone call away.

v

vi

In Memory of Arren Bar-Even

When Arren died on September 18, 2020 at the age of 40, we lost a person who was inspiring in many different ways, someone to look up to and ask for advice. Someone who always gave you the feeling to work with him, even when he was running miles ahead. He left a lab of nearly 40 scientists and students who found a mentor, supervisor, friend, and father figure in him. We, the Bar-Even Lab will continue to live and spread his way of thinking, teaching, designing, and doing science. Arren will not and can never be forgotten. Systems and Synthetic Metabolism Group (Bar-Even Lab) Max Planck Institute of Molecular Plant Physiology Potsdam-Golm, Germany

Jan Lukas Krüsemann Steffen N. Lindner

Introduction: The Future Belongs to the One-Carbons

The most recent report of the Intergovernmental Panel on Climate Change (IPCC) further emphasizes the rapid climate change our planet is facing and the limited time window left to prevent worse consequences. The promise of sustainable and biotechnological production for reduced greenhouse emissions has been around for a while. However, concerns about the reliance of typical biotechnological processes on agricultural substrates (mostly sugars), which compete with food production and further threaten global biodiversity, have raised criticism and doubts about conventional biotech solutions to reduce climate change. In the past decade, this issue has driven scientists and entrepreneurs to explore alternative feedstocks for biotechnological production. Obvious waste-based feedstocks, other than relatively limitedly available and hard-to-process lignocellulosic biomass residues, include waste gas streams consisting of one-carbon (C1) gases and their mixtures: methane (CH4), carbon monoxide (CO), and carbon dioxide (CO2) as well as hydrogen (H2). Apart from these waste streams, atmospheric CO2 is the only truly abundant form of “sustainable carbon” that humanity can unlimitedly harvest to satisfy its large needs for carbon-based chemicals, food, and fuels. Chemical technologies to capture waste gases and atmospheric CO2, and possibly to upgrade them, have been developed substantially in recent years. However, chemical catalysis has its limitations in the range and specificity of products that can be formed at high efficiency. The field of electrochemistry has, for example, demonstrated highly efficient conversion of CO2 with (renewable) electricity to more reduced C1-compounds, such as CO and formic acid (CHOOH) by direct electrocatalysis, as well as methanol (CH2OH) via two-step catalysis from H2. Yet, the formation of two-carbon and especially longer chain compounds by chemical catalysis is still limited to low conversion efficiencies and rates. To convert these C1waste gas streams and C1-carriers from electrochemical production into desired multi-carbon products, biological conversion seems to be the most promising way forward. Extensive research in the microbial utilization of C1-substrates were already performed in the second half of the last century, when a great range of

vii

viii

Introduction: The Future Belongs to the One-Carbons

fascinating microorganisms were discovered that can grow on C1-substrates. In the last decade, the C1 bioconversion field has regained momentum, fueled by the urgent need to develop more sustainable solutions for biotechnological production. An early and promising success story of the field is the industrial-scale conversion of steel mill off-gas (containing CO, H2, and CO2) into ethanol by LanzaTech. Yet, the full potential of C1-feedstocks is by far not realized, considering atmospheric CO2, the sky is literally the limit! In this book, we aim to provide a state-of-the-art overview of C1-based biotechnology, which is a broad field on its own, covering diverse disciplines, and it includes some contrasting approaches. There is, for example, a high interest in gaseous C1-substrates (CH4, CO, CO2), but also the interest in the use of soluble C1-substrates (methanol and formate) is rapidly increasing. Several chapters cover gaseous substrates (e.g., the chapters of Siebert et al., Guo et al.), while others focus more on the soluble feedstocks (e.g., the chapter by Wendisch et al.). Siebert et al. discuss the conversion of the gaseous substrate carbon monoxide by relatively limitedly explored aerobic carboxydotrophs. This chapter highlights the advent of genetic tools for a few of these promising organisms. The chapter of Guo et al. focuses on methanotrophs for the use of methane gas. In particular, recent advances in engineering methanotrophs using synthetic biology approaches, systematic manipulation, metabolic modeling, and carbon flux enhancement are summarized and discussed. The use of the promising soluble substrate methanol is extensively discussed in the chapter of Wendisch et al. This chapter provides an extensive overview of aerobic organisms and pathways for the conversion of methanol. This chapter also covers different products that are or could be made with aerobic methylotrophs and challenges and opportunities of growth on methanol at the bioprocess level. Apart from different substrate types, there are different types of metabolism and organisms that can be employed for C1-bioconversion. A clear distinction is the anaerobic versus aerobic conversion of substrates, which both come with their own opportunities and challenges. For example, the chapter of Song et al. discusses anaerobic acetogenic organisms. It summarizes efforts in understanding the cellular responses and regulation mechanisms during C1 feedstock fermentation using a systems biology approach. Other previously introduced chapters (e.g., the chapters of Siebert et al. and Wendisch et al.) focus on diverse types of aerobic organisms for different substrates. A relatively recent, exciting dichotomy in the field is the emergence of engineered “synthetic” C1-utilizers (see, e.g., chapters of Wang et al., Guo et al., and Claassens et al.) versus naturally isolated C1-converting organisms (see, e.g., chapters of Siebert et al., Guo et al., Wendisch et al.). In the chapter of Guo et al., both natural and designed pathways for methanol assimilation are discussed and their potential to engineer them for “synthetic methylotrophy” in a range of attractive hosts. In the chapter of Wang et al., the implementation of several methanol assimilation pathways in non-methylotrophic hosts is discussed. This chapter also has a specific focus on the use of adaptive laboratory evolution, which was essential in most recently demonstrated examples of synthetic methylotrophy.

Introduction: The Future Belongs to the One-Carbons

ix

For both natural and synthetic C1-consuming hosts, the increased number of genetic engineering and synthetic biology tools has allowed major leaps forward to generate more efficient cell factories for C1-to-product conversion. Also, there are a number of pathways available for C1-conversion, including canonical natural pathways (such as the Calvin Cycle, Wood–Ljungdahl pathway, and the ribulose monophosphate pathway; see, e.g., Siebert et al., Guo et al., and Wendisch et al.), as well as designed pathways (see, e.g., Guo et al., Ren et al., and Claassens et al.). One of the highly promising pathways for C1-assimilation is the reductive glycine pathway (rGlyP), which was designed by Arren Bar-Even and also engineered successfully by his lab recently. This “synthetic” pathway was more recently even also found in nature as the 7th CO2 fixation pathway. Claassens et al. provide a comprehensive overview of the structure and variants of the rGlyP, as well as strategies and progress in engineering this pathway for assimilation of several C1substrates. Ren et al. discuss the catalytic core of this pathway, the glycine synthase/ cleavage system. In this chapter, the fundamental, enzymatic mechanisms for the different enzymes in this system are discussed, as well as strategies to improve their catalytic performance for applications in the rGlyP. The realization of C1-based biotechnology requires many disciplines, definitely going beyond microbial physiology and metabolic engineering. This includes process engineering and electrochemistry. The chapters of Weuster-Botz, Enzmann et al., and Zou et.al. discuss these process and electrochemical aspects of the field and show how they need to be integrated with microbial disciplines, to truly realize sustainable and economically feasible C1-biotechnological processes. In particular, Weuster-Botz summarizes and compares the process engineering characteristics and economic constraints of (syn)gas fermentation with acetogenic bacteria and photobioprocesses with microalgae. Enzmann et al. compare different possibilities to combine electrochemistry and biology to convert C1 compounds into useful products. Zou et.al. discuss the rational construction of electrodes and genetic engineering of producing strains as important efforts to solve bottlenecks encountered in electrobiosynthesis. Furthermore, these authors also discuss artificial photosynthetic biohybrid system (PBS) as a promising means for solar-driven CO2 conversion to chemicals. The broadness and some apparent dichotomies in the field of C1-biotechnology in practice provide us with a wide portfolio of opportunities to realize sustainable bioproduction for different types of substrates, products, and processes. Hence, most of the strategies deserve further studies, and also sometimes synergies between strategies lead to new promising solutions. This book considers synergetic developments in the field, including the use of C1-substrates along other substrates, such as sugars. In the chapter of Hong and Zeng, the advantages and interesting substrate combinations for C1-mixotrophy are discussed. Combining of different types of organisms in C1-converting co-cultures is elaborated on in chapter of Olm and Sousa. Co-cultures, including open mixed cultures and synthetic co-cultures, are of special interest for transforming C1-carbons, especially impure syngas, into a broad spectrum of products beyond the classic C2 compounds such as medium-chain carboxylic acids and higher alcohols.

x

Introduction: The Future Belongs to the One-Carbons

We are proud of this broad up-to-date overview of the C1-biotechnology field, to which many leading scientists and young researchers in the field have contributed. We are especially grateful to Arren Bar-Even, who was one of the initiators of this book volume. This remarkable scientist quickly built an impressive scientific record, career, and vision on the design, biochemistry, and metabolic engineering of C1metabolism. In the summer of 2020, Arren started together with An-Ping Zeng as an editor composing the topics and approaching authors for this book. Sadly, he suddenly passed away in September 2020, not being able to finish this job and his planned chapter for this book (see Obituary by his group members Steffen N. Lindner and Jan L. Krüsemann). The loss of the uniquely talented scientist and remarkable person Arren came as a huge shock throughout the C1-community and beyond. Yet, we are grateful for the unusual scientific and mentoring legacy he left behind. We pay tribute to him in an obituary included in this book and several of his group members contributed a chapter on his signature design: the reductive glycine pathway. We hope, also in the spirit of Arren and his vision that the diverse, collaborative C1-research community will further expand and be successful in realizing the promise of biotechnology in the near future: truly sustainable C1-based bioproduction of chemicals, food, and fuels at industrial scale. Hangzhou, China Wageningen, The Netherlands March 2022

An-Ping Zeng Nico J. Claassens

Contents

Exploiting Aerobic Carboxydotrophic Bacteria for Industrial Biotechnology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Daniel Siebert, Bernhard J. Eikmanns, and Bastian Blombach

1

Process Engineering Aspects for the Microbial Conversion of C1 Gases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dirk Weuster-Botz

33

Systems Biology on Acetogenic Bacteria for Utilizing C1 Feedstocks . . . Yoseb Song, Jiyun Bae, Jongoh Shin, Sangrak Jin, Seulgi Kang, Hyeonsik Lee, Suhyung Cho, and Byung-Kwan Cho Systems Metabolic Engineering of Methanotrophic Bacteria for Biological Conversion of Methane to Value-Added Compounds . . . . Shuqi Guo, Diep Thi Ngoc Nguyen, Tin Hoang Trung Chau, Qiang Fei, and Eun Yeol Lee

57

91

Developing Synthetic Methylotrophs by Metabolic Engineering-Guided Adaptive Laboratory Evolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 Yu Wang, Ping Zheng, and Jibin Sun Bioconversion of Methanol by Synthetic Methylotrophy . . . . . . . . . . . . . 149 Feng Guo, Shangjie Zhang, Yujia Jiang, Huixin Xu, Fengxue Xin, Wenming Zhang, and Min Jiang Aerobic Utilization of Methanol for Microbial Growth and Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169 Volker F. Wendisch, Gregor Kosec, Stéphanie Heux, and Trygve Brautaset Empower C1: Combination of Electrochemistry and Biology to Convert C1 Compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213 Franziska Enzmann, Markus Stöckl, Marc Pfitzer, and Dirk Holtmann

xi

xii

Contents

Extracellular Electrons Powered Microbial CO2 Upgrading: Microbial Electrosynthesis and Artificial Photosynthesis . . . . . . . . . . . . . . . . . . . . 243 Long Zou, Fei Zhu, Fu-Xiang Chang, and Yang-Chun Yong Understanding and Engineering Glycine Cleavage System and Related Metabolic Pathways for C1-Based Biosynthesis . . . . . . . . . . . . . . . . . . . 273 Jie Ren, Wei Wang, Jinglei Nie, Wenqiao Yuan, and An-Ping Zeng Engineering the Reductive Glycine Pathway: A Promising Synthetic Metabolism Approach for C1-Assimilation . . . . . . . . . . . . . . . . . . . . . . . 299 Nico J. Claassens, Ari Satanowski, Viswanada R. Bysani, Beau Dronsella, Enrico Orsi, Vittorio Rainaldi, Suzan Yilmaz, Sebastian Wenk, and Steffen N. Lindner Biosynthesis Based on One-Carbon Mixotrophy . . . . . . . . . . . . . . . . . . . 351 Yaeseong Hong and An-Ping Zeng Conversion of Carbon Monoxide to Chemicals Using Microbial Consortia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 373 Ivette Parera Olm and Diana Z. Sousa

Adv Biochem Eng Biotechnol (2022) 180: 1–32 https://doi.org/10.1007/10_2021_178 © The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 Published online: 12 December 2021

Exploiting Aerobic Carboxydotrophic Bacteria for Industrial Biotechnology Daniel Siebert, Bernhard J. Eikmanns, and Bastian Blombach

Contents 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Carboxydotrophic Candidates for Aerobic Gas Fermentation Approaches . . . . . . . . . . . . . . . . . 2.1 Afipia Carboxidovorans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Hydrogenophaga Pseudoflava . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 Other Carboxydotrophic Bacteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4 Synthetic Carboxydotrophy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 The Aerobic Carboxydotrophic Metabolism and Biochemical Traits of Key Enzymes . . . . 3.1 Oxygen-Insensitive CO Dehydrogenase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Hydrogenases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3 Branched Respiratory Chain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4 Calvin-Benson-Bassham (CBB) Cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Genomic Organization of Genes Relevant for the Autotrophic Lifestyle of Aerobic Carboxydotrophs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 Genetic Organization of Genes Coding for CO Dehydrogenase . . . . . . . . . . . . . . . . . . . . . . 4.2 Genetic Organization of Genes Coding for Hydrogenase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 Genetic Organization of Genes Coding for Proteins of the CBB Cycle . . . . . . . . . . . . . . 5 Genetic and Metabolic Engineering Tools of Carboxydotrophic Bacteria . . . . . . . . . . . . . . . . . . 6 Conclusion and Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2 4 4 5 8 10 11 11 15 15 16 17 17 18 20 20 24 24

Abstract Aerobic carboxydotrophic bacteria are a group of microorganisms which possess the unique trait to oxidize carbon monoxide (CO) as sole energy source with D. Siebert and B. Blombach (*) Microbial Biotechnology, Campus Straubing for Biotechnology and Sustainability, Technical University of Munich, Straubing, Germany SynBiofoundry@TUM, Technical University of Munich, Straubing, Germany e-mail: [email protected]; [email protected] B. J. Eikmanns Institute of Microbiology and Biotechnology, Ulm University, Ulm, Germany e-mail: [email protected]

2

D. Siebert et al.

molecular oxygen (O2) to produce carbon dioxide (CO2) which subsequently is used for biomass formation via the Calvin-Benson-Bassham cycle. Moreover, most carboxydotrophs are also able to oxidize hydrogen (H2) with hydrogenases to drive the reduction of carbon dioxide in the absence of CO. As several abundant industrial off-gases contain significant amounts of CO, CO2, H2 as well as O2, these bacteria come into focus for industrial application to produce chemicals and fuels from such gases in gas fermentation approaches. Since the group of carboxydotrophic bacteria is rather unknown and not very well investigated, we will provide an overview about their lifestyle and the underlying metabolic characteristics, introduce promising members for industrial application, and give an overview of available genetic engineering tools. We will point to limitations and discuss challenges, which have to be overcome to apply metabolic engineering approaches and to utilize aerobic carboxydotrophs in the industrial environment. Graphical Abstract

Keywords Afipia carboxidovorans (formerly known as Oligotropha carboxidovorans), Carbon dioxide, Carbon monoxide, Carboxydotrophic bacteria, Gas fermentation, Hydrogen, Hydrogenophaga pseudoflava, Industrial biotechnology

1 Introduction Due to dramatic climate change, depletion of fossil resources, and pollution of the global environment, it is most important to establish a bio-based circular economy. Industrial biotechnology is regarded as key technology to convert renewable resources toward bulk and fine chemicals as well as fuels. Currently, most biotechnological production processes are sugar-based which leads to ethical and ecological concerns [1]. Therefore, the valorization of non-food carbon sources as well as industrial or agricultural side or waste streams is of particular interest for the future

Exploiting Aerobic Carboxydotrophic Bacteria for Industrial Biotechnology

3

Table 1 Examples of industrial off-gases containing significant amounts of oxygen Source Cement plant flue gasa Coal-fired flue gasc Gas-fired flue gasc

O2 (mol %) 7.5

H2O (mol%) 18.2

CO2 (mol%) 17.8

N2 (mol %) 56.5

CO (ppm) 1,200

SO2 (ppm) 9

NOx (ppm) 120b

4.4

6.2

68–77

50

420

420

4.5

14.6

12.5– 12.8 7.4–7.7

73–74

200– 300

–

60–70

a

[7] Assuming the molecular weight of NO2 (46 g mol1), as recommended by [8] for NOx c [9]

b

economy and contributes to mitigate the negative impact of humans on the environment. Gas fermentation is a promising biotechnological approach to convert C1 containing gases into chemicals and fuels. Processes under complete oxygen exclusion with strictly anaerobic acetogenic bacteria (acetogens) are widely described [2– 5] and the production of ethanol from steel mill flue gas with such microbes reached industrial maturity [6]. Typical applied gas mixtures are comprised of CO, CO2, and H2 (often referred to as synthesis gas or syngas) and are generated by, e.g., steel mills, power stations, refineries or can be produced via gasification of biomass or municipal waste [3, 5]. A drawback for application of anaerobic organisms is their intrinsic sensitivity toward molecular oxygen, which might represent, besides N2, SO2 and NOx, a significant portion of industrial off-gases (Table 1). Thus, the application of strict anaerobic production strains requires an additional gas cleanup step to remove oxygen [4, 10]. Especially CO2 emission by the global cement production with approximately 1.5 Gt annually is one of the biggest anthropogenic sources of this greenhouse gas [11]. Since the off-gas of cement plants contains relatively high amounts of oxygen as well as carbon monoxide, this gas stream represents an available carbon source for microbial gas fermentation approaches (Table 1; [7]). The currently applied acetogenic bacteria use the Wood–Ljungdahl (WL) pathway for CO2 fixation. Although this pathway is the most efficient one with regard to ATP and H2/electron requirements, this group of anaerobic bacteria is considered to be energy-limited which prevents the efficient production of longer-chain or ATP-demanding molecules [5, 12]. To overcome the above-mentioned energy limitations, aerobic carboxydotrophs, such as Afipia carboxidovorans (formerly known as Oligotropha carboxidovorans) and Hydrogenophaga pseudoflava, have recently been proposed for industrial gas fermentation. The aerobic oxidation of CO provides a higher free energy (ΔG0 ¼ 514 kJ) and thus, a higher ATP-yield when compared to the anaerobic route (ΔG0 ¼ 174 kJ) [5, 13–16]. Aerobic carboxidotrophic bacteria are able to tolerate a high concentration of CO (up to 90 vol%; [17, 18]) and utilize this gas as energy source to eventually use the reductive pentose-phosphate or Calvin-Benson-Bassham (CBB) cycle for CO2

4

D. Siebert et al.

fixation. In contrast, the so-called carboxidovores are also able to oxidize CO, but are incapable to proliferate with CO as sole carbon and energy source [19, 20]. “Knallgas” bacteria such as Cupriavidus necator (also known as Ralstonia eutropha) as well as aerobic carboxydotrophs can oxidize molecular hydrogen with oxygen to gain energy for autotrophic growth on CO2. However, Knallgas bacteria do natively not possess a CO dehydrogenase and therefore, are unable to oxidize CO and to grow on this C1 compound [21]. Thus, aerobic carboxydotrophic bacteria have unique traits and therefore represent promising candidates to convert CO- and O2-containing gas streams into higher value products as niche application in industrial biotechnology.

2 Carboxydotrophic Candidates for Aerobic Gas Fermentation Approaches With very few exceptions, carboxydotrophic bacteria are non-pathogenic (biosafety level 1 or risk group 1), mesophilic (Topt about 30 C) and neutralophilic (pHopt about 7) and belong to the phylum of Gram-negative proteobacteria (α or β). Carboxydotrophs are obligately aerobic and facultatively autotrophic [22].

2.1

Afipia Carboxidovorans

Afipia carboxidovorans is probably the best-studied aerobic, carboxydotrophic bacterium. This Gram-negative, facultative chemolithoautotrophic, obligate aerobic α-proteobacterium [22, 23] was formerly known as Pseudomonas carboxydovorans (Kistner) comb. nov. [24], but was recently transferred into the genus Afipia from its former designation Oligotropha [25]. While there are different A. carboxidovorans strains described, namely OM2 to OM5 [22], only strain OM5 is commercially available as Afipia carboxidovorans DSM 1227 at the DSMZ (German Collection of Microorganisms and Cell Cultures GmbH, Brunswick, Germany). Cells of A. carboxidovorans OM5, re-isolated from a wastewater treatment plant in Göttingen (Germany) in the 1970s, are slightly curved rods, form star-shaped aggregates under certain conditions, and are motile via a subpolar flagellum [22, 23]. This bacterium is able to grow lithoautotrophically with CO as sole carbon and energy source, as well as with CO2 plus H2, under aerobic conditions in mineral medium. Organoheterotrophic growth can be observed with several organic acids (e.g., pyruvate, lactate or acetate), but not with sugars, such as glucose or fructose [22, 23]. The genome comprises the bacterial chromosome (3.595 Mbp) and two megaplasmids, pOC167 (167 kbp) and pHCG3 (133 kbp), and the complete sequence is accessible [26]. Furthermore, proteome and transcriptome data are available regarding the autotrophic growth on syngas mixtures [16, 27]. These

Exploiting Aerobic Carboxydotrophic Bacteria for Industrial Biotechnology

5

studies underlined the importance of the megaplasmid pHCG3 for the ability of A. carboxidovorans strain OM5 to grow autotrophically. All genes encoding the enzymes and proteins necessary for autotrophic growth are located on this plasmid and they are strongly transcribed when CO and/or CO2 and H2 are the only carbon and energy sources. This is especially evident for the cox gene cluster, encoding CO dehydrogenase and the proteins necessary for its functional assembly, the hox/hup/ hyp genes, encoding the functional hydrogenase and proteins for its assembly, and the genes for the enzymes of the CBB cycle [16, 26, 27]. Furthermore, the transcriptome data indicated that all CBB cycle gene homologues located on the bacterial chromosome are more specific for utilization of heterotrophic substrates. The megaplasmid pOC167 might be more relevant for utilization of gaseous compounds since it harbors genes encoding proteins annotated for metal ion homeostasis and translocation which might be required for maturation of the metal ion-dependent CO dehydrogenase and hydrogenase [16]. Besides, the oxygen-insensitive CO dehydrogenase of A. carboxidovorans OM5 has been intensely studied and biochemically characterized [28–30]. Also its hydrogenase activity was precisely analyzed [31] and the membrane composition was described in detail [27, 32, 33]. Interestingly, while most aerobic organisms are usually strongly inhibited by CO, A. carboxidovorans can withstand 90 vol% CO due to its specific, CO-insensitive respiratory chain [18, 34]. With regard to industrial biotechnology it is important to note that the doubling time of A. carboxidovorans under autotrophic conditions is rather low (Table 2). While earlier studies presented doubling times of 20 h with CO and 7 h with CO2 and H2 under aerobic conditions [23], more recently, doubling times of 40–50 h with syngas mixtures were shown [16, 27]. For aerobic, heterotrophic conditions, doubling times of 6–10 h with different organic acids, e.g. acetate, pyruvate and lactate, have been published [16, 23, 27].

2.2

Hydrogenophaga Pseudoflava

For Hydrogenophaga pseudoflava, formerly known as Pseudomonas pseudoflava or Pseudomonas carboxydoflava [41], there are at least three strains commercially available, namely the type strain DSM 1034, corresponding to strain ATCC 33668 (American Type Culture Collection, Manassas, Virginia, USA) or GA3, strain DSM 1084 (Z-1107), and strain ATCC 700892 (PALL 6266-34 3X(6569-4). Strain ATCC 700892 is mostly described as organism with the ability to penetrate membrane filters for sterilization and therefore used as test system for such filters [42–44]. The type strain H. pseudoflava DSM 1034 was isolated as P. pseudoflava strain GA3 from water samples of the River Weende, Göttingen, Germany in the 1970s [41, 45] and it was characterized by its ability to produce polyhydroxyalkanoates (PHAs), especially polyhydroxybutyrate (PHB), [45–48]. H. pseudoflava strain GA3 shows fast autotrophic growth with CO2, H2, and O2 (doubling time of 2.5 h) and heterotrophic growth with sucrose (doubling time of 1.5 h) and other sugars [49]. However, the type strain is unable to oxidize CO and growth is inhibited to some extent by CO

B. japonicum 110spc4

P. carboxydohydrogena

H. pseudoflava DSM 1084

A. carboxidovorans OM2–OM4 A. carboxidovorans OM5

Organism

i

h

a

g

a

f

d

c

c

a

Ref.

0.02

0.10

0.04

0.09

0.06

0.06

0.01

0.10

0.03– 0.04 0.03

μ (h1)

40

7

18.5

8

12.3

11.6

49.5

7

20

16.2–25.8

Doubling time (h)

n.a.

n.a.

2.3b

n.a.

51.3b

73.9

13.9e

n.a.

n.a.

37.9 (OM3)b

q(CO) (mmol gCDW1 h1)

14.2

56.2

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

2.2e

23.0e

n.a.

n.a.

n.a.

n.a.

q(H2) (mmol gCDW1 h1)

n.a.

n.a.

n.a.

q(CO2) (mmol gCDW1 h1)

n.a.

n.a.

n.a.

n.a.

n.a.

31.4

4.0e

n.a.

n.a.

n.a.

q(O2) (mmol gCDW1 h1) 50% CO, 50% air 40% CO, 55% N2, 5% O2 10% CO2, 85% H2, 5% O2 40% CO, 10% CO2, 40% H2, 8% N2, 2% O2 40% CO, 10% CO2, 40% H2, 8% Ar, 2% O2 50% CO, 50% air 10% CO2, 80% H2, 10% O2 50% CO, 50% air 10% CO2, 80% H2, 10% O2 50% CO, 50% air

Gas mixture (vol%)

17

n.a.

255

n.a.

570

n.a.

n.a.

n.a.

223

275–641

CO oxidizing activity (nmol gas mg protein1 min1)

n.a.

n.a.

1,785

n.a.

12,035

n.a.

n.a.

n.a.

818

1,341–1,875

H2 oxidizing activity (nmol gas mg protein1 min1)

Table 2 Examples of different aerobic, carboxydotrophic organisms, regarding relevant process parameters and biochemical properties with gaseous substrates

6 D. Siebert et al.

j

0.01

50.4

0.054

n.a.

n.a.

a

Ref. ¼ reference; q (mmol gCDW1 h1) ¼ biomass specific consumption rate; n.a. ¼ not available [35] b Assuming approx. 50% of CDW is protein [36] c [23] d [16] e Assuming conversion factor of CDW ¼ 0.55 g L1  OD600 (unpublished data) f [15] g [37] h [38] i [39] j [40]

C. necator H16 (pBBR1MCS-3::PL:: coxMSLDEFGOc)

n.a.

13% CO, 3% CO2, 13% H2, 56% N2, 15% O2

n.a.

n.a.

Exploiting Aerobic Carboxydotrophic Bacteria for Industrial Biotechnology 7

8

D. Siebert et al.

[17]. The raw genome sequence of P. pseudoflava DSM 1034 is available, but so far only divided into 42 unassembled contigs (ASM159228v1 [50]). H. pseudoflava DSM 1084 was isolated in the 1970s as P. carboxydoflava strain Z-1107 from mud and soil of the River Moskva, Russia [41, 51]. Similar to H. pseudoflava DSM 1034, strain DSM 1084 is able to grow aerobically rather fast on several sugars and also autotrophically with CO2 and H2 (doubling time of 7 h). Moreover, strain DSM 1084 can utilize CO as sole carbon and energy source (doubling time of 12.3 h), tolerates up to 90 vol% CO without growth or respiration impairment, and shows high biomass specific CO consumption (qCO) rates of 73.9 mmol gCDW1 h1 (Table 2; [15, 18, 22, 35]). The growth rate of strain DSM 1084 on CO is in good accordance with recently published data applying a syngas mixture providing a doubling time of 11.6 h (Table 2; [15]). As described above for A. carboxidovorans, the CO dehydrogenase and the hydrogenase of the strain DSM 1084 were intensively studied [37, 52, 53]. Compared to other carboxydotrophs, strain DSM 1084 shows relatively high hydrogenase activities (Table 2; [35]). The cells of the Gram-negative, β-proteobacterium H. pseudoflava strain DSM 1084 are non-motile, rod-shaped and show a yellow pigmentation [22]. Recently, the genome of H. pseudoflava DSM 1084 was fully sequenced. It consists of a chromosome of 4.681 Mbp and a megaplasmid of 45 kbp. The genes of all enzymes required for chemolithoautotrophic growth on CO, for CO2-fixation and H2 oxidation are located on the chromosome [15]. Other omics data are not available for this organism yet and genome sequences of the other H. pseudoflava strains are still not available. With regard to application for industrial biotechnology, this organism shows promising traits including a high growth rate and a broad substrate spectrum, ranging from gaseous compounds over organic acids to sugars and sugar alcohols such as xylose or glycerol, respectively [15, 22, 35, 45, 49]. Furthermore, H. pseudoflava strain DSM 1084 has been already engineered for the production of (E)-α-bisabolene from syngas (s. below; [15]).

2.3

Other Carboxydotrophic Bacteria

While A. carboxidovorans and H. pseudoflava received considerable scientific attention, several other representatives of this group are hardly characterized, however, might also be promising candidates for biotechnological approaches. Pseudomonas carboxydohydrogena (DSM 1083, ATCC 29978) was isolated from a sewage treatment plant in Russia in the early 1970s as Seliberia carboxydohydrogena (strain Z-1062) and later proposed to be included to the genus Afipia, thus being an α-proteobacterium [38, 51, 54]. It was characterized as organism with a limited substrate spectrum, being able to metabolize several organic acids as well as sugars such as fructose and sucrose. P. carboxydohydrogena shows doubling times of 18.5 h with CO and 7 h with CO2 and H2 and the CO uptake rate as well as the oxidizing activity for CO and H2 is comparable to that of A. carboxidovorans OM5 (Table 2). Noteworthy, cultures of

Exploiting Aerobic Carboxydotrophic Bacteria for Industrial Biotechnology

9

P. carboxydohydrogena reach relatively high cell densities when compared to other carboxydotrophs [35, 38]. The soluble CO dehydrogenase of P. carboxydohydrogena was purified and biochemically characterized [55] and the hydrogenase is membrane bound [35, 38]. While earlier literature describes prototrophic growth in mineral medium [22, 23, 35, 38], the DSMZ recommends supplementation of a complex vitamin solution (“Wolin’s vitamin solution”) and addition of hydrogen carbonate for chemoorganotrophic growth [56, 57]. Recently, a bioprocess was published, elucidating the ability of this organism to produce PHA/PHB autotrophically from a syngas mixture (CO2, H2, and CO) in a 10 L-bioreactor under aerobic conditions [58]. Zavarzinia compransoris, isolated in the 1970s as Comamonas compransoris (strain Z-1155) from mud of River Moskva, Russia [51], is also known as Pseudomonas compransoris [22, 24] and is commercially available as strain DSM 1231 or ATCC 51430 and proposed to be an α-proteobacterium [54]. When growing autotrophically with 50 vol% CO and 50 vol% air, the CO uptake rate and oxidizing activity are comparable to those of H. pseudoflava DSM 1084, while the growth rate and CO/H2-oxidizing activities are similar to those of P. carboxydohydrogena [35]. Beside the genome sequence (ASM317305v1 and ASM436299v1 [50]), the available literature for this organism is very limited, but describes it as being thiamineauxotroph, with a quite small substrate spectrum and being strongly inhibited by low concentrations of CO (39% growth inhibition at 20 vol% CO; [17, 22]), rendering this organism only applicable for gas streams with a low CO content. Methyloversatilis universalis, isolated as Pseudomonas gazotropha (strain Z-1156) from mud of River Moskva, Russia [22, 51], is available as strain DSM 1085 and classified as β-proteobacterium [59]. While the available information about this organism is also very limited, the limitations for an application are quite similar to Z. compransoris with the exception of M. universalis being vitamin B12auxotrophic. Interestingly, it is described that thiamine for Z. compransoris is provided by M. universalis and vitamin B12 vice versa in co-cultures of these organisms [22, 51]. Aminobacter carboxidus was isolated as Achromobacter carboxydus (strain Z-1171) from soil near a stream at Neskuchny Garden (Moscow, Russia) in the 1970s, meanwhile reclassified as Alcaligenes carboxydus and Carbophilus carboxidus, and recently introduced into the genus Aminobacter (thus classified as α-proteobacterium) and thereafter proposed to be identical with Aminobacter lissarensis [22, 24, 25, 51, 60]. It is commercially available as strain DSM 1086 or ATCC 51424. A. carboxidus grows rather slowly (42 h doubling time) with CO as sole carbon and energy source under aerobic conditions (50 vol% CO and 50 vol% air). This organism has the rare characteristic for an aerobic carboxydotroph to be unable to grow with CO2 and H2, although a measurable, but very low H2-oxidizing activity is present [24, 35]. While showing CO-oxidizing activities comparable to that of P. carboxydohydrogena, growth of A. carboxidus is inhibited at low CO concentrations (20% inhibition at 20 vol% CO; [17, 35]). Recently, a draft genome sequence was presented and identified the presence of forms I and II CO dehydrogenase systems [61].

10

D. Siebert et al.

Bradyrhizobium japonicum, classified as α-proteobacterium [62], is a Gramnegative soil bacterium and associated with soybeans as plant symbiont. As other rhizobia have been considered to be usually chemoorganotrophs, there are chemolithoautrophic strains described, even with the capability to utilize CO as sole carbon and energy source under aerobic conditions (Table 2; [39]). In the case of B. japonicum 110spc4 a rather slow growth (with a doubling time of around 35 h) and CO oxidizing activity under autotrophic conditions was described (Table 2; [39]). Besides of all aforementioned bacteria being mesophilic, there are also some species described with the ability to oxidize CO under thermophilic conditions. An example is Hydrogenibacillus schlegelii (strain MA-48/DSM 2000), formerly known as Bacillus schlegelii, which is a Gram-variable bacterium and can grow rapidly (doubling time of 3–4.5 h) with CO at around 70 C [22, 63]. A moderately thermophilic carboxydotroph is Pseudomonas thermocarboxydovorans (strain C2/DSM 13294), which was isolated from raw sewage in the United Kingdom in the 1980s. P. thermocarboxydovorans grows optimally at around 50 C and autotrophically only with CO, but not with CO2 and H2. It shows no H2-oxidizing activity, while its CO dehydrogenase has the same subunit composition and significant sequence homologies to the one of A. carboxidovorans [64, 65].

2.4

Synthetic Carboxydotrophy

Escherichia coli is natively not able to grow autotrophically and is highly susceptible toward CO toxicity [17]. However, E. coli was recently engineered to form biomass exclusively from CO2 and formate by installing a CBB cycle and a formate dehydrogenase to provide reducing power for CO2-fixation [66]. Furthermore, Kaufmann et al. [30] heterologously expressed a fully functional CO dehydrogenase from A. carboxidovorans in E. coli and purified the enzyme complex. To elucidate the importance of the different accessory proteins for full activity of CO dehydrogenase, different gene combinations of the cox cluster were assembled and expressed [30]. Cupriavidus necator (also known as Ralstonia eutropha) strain H16, classified as β-proteobacterium, is the best-studied “Knallgas” bacterium, which is able to grow autotrophically on CO2 and H2, but is unable to utilize CO [21]. C. necator is known for its ability to produce high amounts of PHB and its potential for the CO2- and H2-based production of biofuels has recently been reviewed [67]. Since the growth of C. necator is only weakly inhibited by CO (66% growth inhibition at 90 vol% of CO; [17]), Heinrich et al. [40] implemented a functional CO dehydrogenase from A. carboxidovorans OM5 into C. necator H16 and showed that for the resulting strain CO served as sole carbon source for growth and PHB production (Table 2). However, the authors stated that this recombinant strain was not able to utilize CO as energy source, still being dependent on H2 supplementation. Therefore, Heinrich et al. [40] termed this engineered strain as “carboxyhydrogenotrophic” rather than carboxydotrophic C. necator [40].

Exploiting Aerobic Carboxydotrophic Bacteria for Industrial Biotechnology

11

3 The Aerobic Carboxydotrophic Metabolism and Biochemical Traits of Key Enzymes Aerobic carboxydotrophic bacteria are able to grow chemolithoautotrophically in the presence of O2 on CO or on H2, CO2 and CO-containing gas mixtures (synthesis gas). They all obtain energy (and carbon) for growth from oxidation of CO using O2 as terminal electron acceptor, according to the overall equation
2CO þ O2 ! 2CO2 ΔG0 ¼ 514 kJ : In accordance, all so far studied carboxydotrophic bacteria are strictly oxidative, withstand the toxicity of up to 90 vol% CO, and possess (1) an oxygen-insensitive CO dehydrogenase (complex) for CO oxidation, (2) a CO-insensitive respiratory chain for electron transport and thus, for energy conservation, and (3) the CBB cycle for CO2 fixation using NAD(P)H generated most probably by reverse electron transfer from CO or H2 oxidation. In the absence of CO and also in its presence, most carboxydotrophic bacteria can also use H2 and CO2 for lithoautotrophic growth, H2 being oxidized by hydrogenases for the reduction of CO2 and for energy conservation by electron transport phosphorylation. The growth of carboxydotrophic bacteria with H2 and CO2 is even faster than with CO alone [22]. Several of the carboxydotrophic bacteria grow also organoheterotrophically with organic acids and/or with sugars (see above; [15, 16, 22] and thus, in fact these are facultatively lithoautotrophic organisms. Figure 1 gives an overview of the actual understanding about the autotrophic metabolism and key enzymes of carboxydotrophic bacteria. In the following, we will discuss in more detail what is known about CO dehydrogenase, the hydrogenases, the respiratory chain, and the CBB cycle of carboxydotrophs. However, we will mainly focus on the enzymes/proteins and their biochemical traits of A. carboxidovorans OM5 and H. pseudoflava DSM 1084 since these show relatively high growth rates (up to 0.06 h1) under autotrophic conditions (see above) and since they have been characterized in most detail.

3.1

Oxygen-Insensitive CO Dehydrogenase

In general, the first step of CO utilization is catalyzed by a CO dehydrogenase that oxidizes CO to CO2 with concomitant generation of H2 or reduction of electron acceptors. The most important element for carboxydotrophy under aerobic conditions is the presence of an O2-insensitive CO dehydrogenase (also designated as CO: acceptor oxidoreductase or as CO oxidase; [22]). This enzyme is assumed to be very similar or identical in all carboxydotrophic bacteria (Fig. 2; [22]) and has been extensively studied in A. carboxidovorans and H. pseudoflava [28–30, 52, 53]. CO dehydrogenase in these and other carboxydotrophic bacteria is associated with the

12

D. Siebert et al.

Fig. 1 Overview of the autotrophic metabolism of aerobic carboxydotrophs based on the actual knowledge of A. carboxidovorans OM5, H. pseudoflava DSM 1084, and C. necator H16. CM cytoplasmic membrane, CODH CO dehydrogenase, ETC electron transport chain (respiratory chain), CBB Calvin-Benson-Bassham cycle, MBH membrane-bound hydrogenase, SH soluble hydrogenase. Proteins: CoxLMS ¼ large/medium/small subunit of CO dehydrogenase; b561, b563 ¼ cytochrome b561/563; Q10 ¼ ubiquinone10; HoxSL ¼ large/small subunit of MBH; HoxZ ¼ di-heme cytochrome b; HoxFUYH ¼ α-/γ-/β-/δ-subunit of SH as found in C. necator H16. Genes encoding: cbbSL ¼ small and large subunits of ribulose-1,5-bisphosphate carboxylase-oxygenase (RuBisCo); pgk ¼ phosphoglycerate kinase; gap ¼ glyceraldehyde-3-phosphate dehydrogenase; tpiA ¼ triose-phosphate isomerase; cbbA ¼ fructose-1,6-/sedoheptulose-1,7-bisphosphate aldolase; cbbF ¼ fructose-1,6-/sedoheptulose-1,7-bisphosphatase; cbbT ¼ transketolase; cbbE ¼ pentose-5phosphate 3-epimerase; rpiA ¼ ribose-5-phosphate isomerase; cbbP ¼ phosphoribulokinase. Metabolites and co-factors: GAP ¼ glyceraldehyde 3-phosphate; MCD ¼ molybdopterin cytosine dinucleotide; FenSn ¼ iron-sulfur clusters; FAD ¼ flavin adenine dinucleotide; FMN ¼ Flavin mononucleotide; NAD(P)+/NAD(P)H ¼ oxidized/reduced form of nicotinamide adenine dinucleotide (phosphate); ATP ¼ adenosine triphosphate; ADP ¼ adenosine diphosphate; Pi ¼ inorganic phosphate. This figure was adapted and compiled using the following references: [15, 26, 28, 31, 34, 68–73]

inner plasma membrane [74] and (1) feeds electrons into the respiratory chain, (2) provides electrons for generation of NAD(P)H, and (3) releases CO2 for the CCB cycle (Fig. 1). A minor part (4–16%) of the CO2 generated by CO dehydrogenase enters the CBB cycle and thus serves CO2 assimilation and energy conservation. The major part of CO oxidation serves energy conservation by respiration and generation of reducing equivalents for carbon reduction in the CBB cycle [22, 23], releasing the excess of CO2 into the environment (e.g., observable as

Exploiting Aerobic Carboxydotrophic Bacteria for Industrial Biotechnology

13

negative biomass specific consumption rates in Table 2/[15]). Whereas the CO dehydrogenases in aerobic bacteria use oxygen as terminal electron acceptor and need Mo and Cu as cofactors (see below), the completely different O2-sensitive CO dehydrogenases present in anaerobic organisms use various other electron acceptors and are dependent on Ni [75, 76]. CO oxidation by the O2-insensitive CO dehydrogenase in carboxydotrophic bacteria is represented by the equation CO þ H2 O þ XðoxÞ ! CO2 þ XðredÞ , water being the source of O2 in CO2 and X(ox) being the electron acceptor [74]. The CO dehydrogenase of A. carboxidovorans does not directly reduce NAD+, NADP+, FAD+, FMN+ (all with a redox potential (E00 ) of < 200 mV) nor cytochrome c or O2. Under anaerobic conditions it reduces methylene blue and pyocyanin, both substances with a redox potential E00 > 100 mV, and it reduces with a very low rate phyllochinon (vitamin K1) [77]. Ubiquinone 8 and 10 and also menaquinones have been identified in carboxydotrophic bacteria, however, there are reports arguing against and also for a functioning of these substances as the physiological electron acceptors [74, 78]. Cypionka and Meyer [18, 34] and Rohde et al. [79] suggested cytochrome b561 as physiological electron acceptor and cytochrome o (¼ b563) as CO-insensitive terminal oxidase (Fig. 1; [34, 79]). Aside from CO oxidation activity, the CO dehydrogenase of A. carboxidovorans has been reported to exhibit hydrogenase activity, oxidizing H2 to protons and transferring electrons to artificial oxidants [80] and this was substantiated by Wilcoxen and Hille [81] [81], however, a physiological function of this activity remains unclear. While CO dehydrogenases of different carboxydotrophic bacteria seem to be rather similar (reviewed in [74] and cf. Fig. 2), the best studied O2-insensitive CO dehydrogenase is that of A. carboxidovorans. The structure and recent advances in the understanding of the reaction mechanism have been recently reviewed by Hille et al. [28] and by Kalimuthu et al. [29]. The enzyme is a heterotrimer (αβγ)2 with a MW of 277,000 Da [82] consisting of a small subunit (CoxS) containing two [2Fe-2S] iron-sulfur clusters, a medium-size FAD-containing subunit (CoxM), and a large subunit (CoxL) bearing a binuclear Mo-S-Cu moiety at the active site (Fig. 1; [83]). Both, the CO dehydrogenase of A. carboxidovorans and that of H. pseudoflava have been crystallized and characterized and the structures have been found to be very comparable [84, 85]. This conclusion is in agreement with the highly similar sizes (17.8 and 17.7 kDa for CoxS, 30.2 and 30.5 kDa for CoxM, and 88.7 and 87.0 kDa for CoxL) and the high identity of the amino acid sequences (determined via BLASTP 2.12.0+) of the three subunits, being 69% (CoxS), 53% (CoxM) and 69% (CoxL) [15, 26, 86].

14

D. Siebert et al.

Fig. 2 (a) Overview of the genetic organization of the cox cluster (14,538 bps) on megaplasmid pHCG3 in A. carboxidovorans OM5, showing coxMSL, coding for structural proteins (blue), coxBCHK, coding for membrane associated proteins (green), coxFI, coding for XdhC-like proteins (gray), coxD, coding for a membrane associated AAA+ ATPase (black), and coxEG, coding for proteins with yet unknown function (white). Promoters are indicated by “P” (adapted from [26, 28, 72]). (b–d) Genetic organization of CO dehydrogenase coding genes and similarities (represented by grey gradients) of gene products in the aerobic carboxydotrophs (b) H. pseudoflava DSM 1084 [15], (c) A. carboxidus DSM 1086 [61] and (d) Z. compransoris DSM 1231 (ASM436299v1 [50]) compared to the cox cluster of A. carboxidovorans OM5 (gene names s. (A); [26]).

Exploiting Aerobic Carboxydotrophic Bacteria for Industrial Biotechnology

3.2

15

Hydrogenases

Hydrogenases in carboxydotrophic bacteria catalyze H2 oxidation and serve energy generation via electron phosphorylation and provision of NAD(P)H, the latter similar to CO dehydrogenase via a reverse electron transfer [22]. The main hydrogenases in carboxydotrophs are membrane-bound, they do not directly reduce NAD (P) and are present in cells grown with H2 and O2 and CO2, with CO and O2, and surprisingly, also with pyruvate and O2 [23, 87]. The enzyme of A. carboxidovorans has been purified and characterized by Santiago and Meyer [31]. It is a heterodimeric NiFe-protein, consisting of HoxL (MW ¼ 40,100 Da) and HoxS (MW ¼ 67,100 Da) subunits and assumed to be attached to the membrane via a membrane anchor protein HoxZ (Fig. 1). A single membrane-bound hydrogenase was also described for H. pseudoflava. H2-oxidizing activity was found in autotrophically grown cells and the respective hox genes are located on the chromosome [15, 35]. The subunits HoxK and HoxG show 81% and 76% identity (determined via BLASTP 2.12.0+) compared to the respective subunits of A. carboxidovorans [15, 26, 86]. Genome sequencing revealed evidence for an additional soluble hydrogenase in H. pseudoflava (see below; [15]), however, functional expression as well as operation of the enzyme remains to be elucidated.

3.3

Branched Respiratory Chain

Carboxydotrophic bacteria rely on ATP synthesis by oxidative phosphorylation, using CO or H2 as initial origin of reducing equivalents during chemolithoautotrophic growth and organic substrates during organoheterotrophic growth. Different cytochromes (b-, c-, a, and o-type), quinones and iron-sulfur proteins have been found in these bacteria and proposed to be constituents of the respiratory chain [22]. Cypionka and Meyer [18, 34] firstly obtained experimental evidence for a branched respiratory chain in A. carboxidovorans, branched off at the level of b-type cytochromes or at ubiquinone 10 [34]. The electrons from oxidation of organic substrates (and thus, from NADH and/or FADH) and also those from H2 oxidation in the absence of CO are channeled into the “heterotrophic branch,” consisting of cytochromes b558, c, and a1, the latter being the CO-sensitive terminal oxidase. The electrons from oxidation of CO and of H2 are channeled into the autotrophic branch, consisting of cytochromes b561 and b563 (¼ cytochrome o), the latter being the CO-insensitive terminal oxidase (Fig. 1). In contrast to the cytochrome a1 oxidase, the cytochrome b563 oxidase has a quite high affinity toward O2 which might be the reason for the CO-insensitivity [88]. It should be noted here that the CO-insensitive branch of the respiratory chain conserves less energy than the CO-sensitive branch, due to the loss of one coupling site [89].

16

D. Siebert et al.

The redox potentials of the CO/CO2 and H2/H+ couples are negative enough to allow the direct reduction of NAD(P). However, both CO dehydrogenase and hydrogenase do not directly reduce NAD(P) [77] and use electron acceptors with a higher midpoint potential, indicating that carboxydotrophic bacteria use a proton-motive-force-driven reverse electron transport for NAD(P)H generation. This suggestion was substantiated by the finding that protonophores, but not ATPase inhibitors, hamper NAD(P)H formation (reviewed in [74]).

3.4

Calvin-Benson-Bassham (CBB) Cycle

When growing in the presence of CO, the CO dehydrogenase in carboxydotrophic bacteria provides CO2 for the assimilation of this compound by ribulose-1,5bisphosphate carboxylase in the CBB cycle. Up to 16% of the CO2 generated by CO dehydrogenase in carboxydotrophic bacteria enters the CBB cycle and thus serves CO2 assimilation [22, 23, 87]. When these bacteria grow with synthesis gas or with H2 and CO2, the latter is directly assimilated via the CBB cycle. The presence and functionality of this cycle during growth with CO and/or with H2 and CO2 has been reported for many of the carboxydotrophic bacteria [22, 65]. In several carboxydotrophs, the specific activities of key enzymes of the CBB cycle, ribulose-1,6-bisphosphate carboxylase and phosphoribulokinase have been shown to be high during growth in the presence of CO, whereas they were low during heterotrophic growth (summarized in [22]). This finding is corroborated by a recent RNASeq analysis of A. carboxidovorans OM5. The cbb genes encoding ribulose1,6-bisphosphate carboxylase and phosphoribulokinase as well as for all other CBB cycle enzymes are highly induced (up to 2,000-fold) during growth with synthesis gas compared to heterotrophic growth with acetate ([16]; see also above). Furthermore, strong indications for the operation of the CBB cycle in carboxydotrophic bacteria are (1) that all carboxydotrophic bacteria tested possess active ribulose-1,6bisphosphate carboxylase activity or the respective cbb genes (Fig. 4; [15, 16, 22]), (2) that dilution of 14CO in the gas phase of growing cultures with CO2 resulted in a significant decrease of label in the biomass and (3) that immediately after labelling with 14CO and 14CO2 the early fixation products in the biomass were identified as phosphoglycerate (40% of the incorporated label) and as phosphorylated sugars (reviewed in [22]). To the best of our knowledge, there is not much information about the biochemistry of CBB cycle key enzymes ribulose-1,5-bisphosphate carboxylase and ribulokinase in aerobic carboxydotrophs. AMP and phosphoenolpyruvate are considered to be general inhibitors of the ribulokinase in many lithotrophic bacteria [87]. The holoenzyme of ribulose-1,5-bisphosphate carboxylase from the carboxydotrophic Arthrobacter strain 11/x has been reported to have oxygenase activity, to be regulated/activated by NADH, to have a molecular weight of 571,000 Da, and to consist of eight small (MW about 15,000 Da) and eight large subunits (MW about 56,000 Da) [22, 87]. These subunit sizes are consistent with the

Exploiting Aerobic Carboxydotrophic Bacteria for Industrial Biotechnology

17

sizes of the small and large subunits deduced from the respective cbbS and cbbL genes identified by genome sequencing of A. carboxidovorans (MWs of 16,600 Da and 53,800 Da, respectively) and H. pseudoflava (13,600 Da and 52,700 Da, respectively) (see also below; [15, 26]) and in general with those of microbial form I enzymes [91]. As deduced from the cbbS and cbbL genes from the two organisms via BLASTP 2.12.0+ [15, 26, 86], the respective subunits show 35% and 56% identity on amino acid sequence level.

4 Genomic Organization of Genes Relevant for the Autotrophic Lifestyle of Aerobic Carboxydotrophs For most of the above-mentioned aerobic carboxydotrophs an annotated genome map is still missing. Only for A. carboxidovorans OM5 and OM4 [26] and for H. pseudoflava DSM 1084 [15] a closed and annotated genome map has been published, while the genome sequences of A. carboxidus DSM 1086 [61] and Z. compransoris DSM 1231 (ASM317305v1 and ASM436299v1 [50]) are only available as draft shotgun sequences divided into single but annotated contigs. For H. pseudoflava DSM 1034 only a non-annotated shotgun sequence is available (ASM159228v1 [50]). In the following, the genomic organization of genes encoding CO dehydrogenase and hydrogenase as well as CBB cycle genes will be described.

4.1

Genetic Organization of Genes Coding for CO Dehydrogenase

Because the oxygen-insensitive CO dehydrogenase of A. carboxidovorans OM5 was analyzed in detail [28–30], the according cox gene cluster is most suitable as reference to investigate the organization of CO dehydrogenase genes in other aerobic carboxydotrophs. The cox cluster of A. carboxidovorans is 14,538 bps long, contains 12 genes, and differs only in 1 bp in strains OM4 and OM5 (Fig. 2a; [26]). In the past, the gene coxH was considered to be around 1,200 bp in size [72], however, more recently it has been ”divided” in two open reading frames (ORFs), with coxH2 and coxH1 being the functional gene and a pseudogene, respectively (Fig. 2a; [26, 92, 93]. While A. carboxidovorans harbors only (genes for) a form I CO dehydrogenase, H. pseudoflava DSM 1084 additionally contains (genes for) a form II CO dehydrogenase (Fig. 2b; [15]). Previously, it was proposed that some aerobic CO oxidizing bacteria possess both forms of CO dehydrogenase and that form I of the so-called OMP (from Oligotropha, Mycobacterium, and Pseudomonas) group, and form II of the BMS (from Bradyrhizobium, Mesorhizobium, and Sinorhizobium) group, have different affinities toward CO, enabling the organisms

18

D. Siebert et al.

to adapt to different CO concentrations in the environment [94, 95]. The structural genes of form I in H. pseudoflava DSM 1084 are arranged as coxMSL, typical for the OMP group, while the structural genes for form II are organized as coxSLM, typical for the BMS group (Fig. 2b; [95]. Interestingly, H. pseudoflava DSM 1084 has earlier been attributed to the OMP group with genes encoding form I CO dehydrogenase to be located on a megaplasmid [94, 96]. However, the recently published genome sequence revealed the presence of separated gene clusters for a form I and a form II CO dehydrogenase, both being located on the chromosome [15]. Accordingly, Kang and Kim [52] identified previously the structural genes cutMSL on the chromosome of H. pseudoflava. Compared to A. carboxidovorans OM5, the gene cluster encoding form I CO dehydrogenase of H. pseudoflava DSM 1084 lacks homologues for coxB, coxG, coxH2, coxI, and coxK, while the cluster of form II CO dehydrogenase lacks a homologue of coxC, but harbors a homologue for coxG (Fig. 2b; [15]). It was already described that coxH is not essential for maturation of an active CO dehydrogenase. Moreover, some cox genes of A. carboxidovorans OM5 share homology (e.g., coxI and coxF, coxC, and coxH) and might be able to functionally substitute the corresponding homologue which might be the reason for the more complex cluster architecture in this organism [28, 30, 40, 72]. Homologs for coxB and coxK are also not present in A. carboxidus DSM 1086 and Z. compransoris DSM 1231 (Fig. 2c, d) and seem generally to be absent in other carboxydotrophs [72]. In case of A. carboxidus DSM 1086, also gene clusters for both forms of CO dehydrogenases can be found. In this organism, the form I gene cluster shows homologues for almost all A. carboxidovorans OM5 cox genes, with the exception of coxB and coxK, while the gene cluster for form II harbors only the structural genes coxSLM (Fig. 2c; [61]). Z. compransoris DSM 1231 seems to have only the form I CO dehydrogenase, while the cox cluster of this organism contains homologues for nearly all cox genes from A. carboxidovorans OM5, except for deduced sequences of coxB and coxK (Fig. 2d; ASM436299v1 [50]), rendering these genes quite exclusive for A. carboxidovorans.

4.2

Genetic Organization of Genes Coding for Hydrogenase

Hydrogenase activities in aerobic carboxydotrophs such as A. carboxidovorans OM5 and H. pseudoflava DSM 1084 have been analyzed in detail [23, 35, 37], however, the underlying genomic architecture has not been well studied so far. In contrast, the best-studied reference for hydrogenases represents the “Knallgas” bacterium C. necator H16 [70]. This organism harbors a membrane-bound, a soluble and a regulatory hydrogenase and the encoding genes are spread over different clusters located on the megaplasmid pHG1 [70, 90]. In contrast, in A. carboxidovorans OM5 and OM4 the genes necessary for expression of the membrane-bound hydrogenase are arranged in a single cluster (differing only in two bp between both strains) on the megaplasmid pHCG3 and pHCG3B, respectively (Fig. 3a; [26]). The organization of this cluster differs considerably from the

Exploiting Aerobic Carboxydotrophic Bacteria for Industrial Biotechnology

19

Fig. 3 (a) the α-proteobacteria A. carboxidus DSM 1086 [61] and (b) Z. compransoris DSM 1231 (ASM436299v1 [50]) compared to A. carboxidovorans OM5 [26] and (c) the β-proteobacterium H. pseudoflava DSM 1084 [15] compared to C. necator H16 [90]

corresponding one in C. necator H16. While the accessory hyp genes are only available as single copies in A. carboxidovorans OM5, up to three homologs of each hyp gene can be found in C. necator H16 [26, 90]. Until now, only the presence of a membrane-bound hydrogenase was described for A. carboxidovorans OM5 [23, 31], however, tBLASTx 2.12.0+ analysis [86] identified homologs (namely, hoxA, hoxB, OCA5_pHCG300670 and hoxV) of the hoxABCJ genes encoding a regulatory hydrogenase in C. necator H16 [26, 70, 90]. Furthermore, the α-proteobacteria A. carboxidovorans OM5 and Z. compransoris DSM 1231 as well as the β-proteobacteria C. necator H16 and H. pseudoflava DSM 1084 share a very similar architecture of the hydrogenase encoding clusters (Fig. 3; ASM436299v1; [15, 50, 58, 90]). However, in Z. compransoris DSM 1231 the hyp genes are not clustered with the other genes necessary for functional expression of a hydrogenase (Fig. 3a; ASM436299v1 [50]). In H. pseudoflava DSM 1084, besides of the membrane-bound hydrogenase, homologs of the structural genes hoxFUYH coding for a soluble hydrogenase in C. necator H16 are present [15, 70, 86, 90]. While it was mentioned earlier that H. pseudoflava DSM 1084 seems to lack genes of a regulatory hydrogenase [15], a homolog of hoxA (HPF_01490) can be found, as well as a gene cluster (HPF_01520–HPF_01535) coding for proteins with high similarities to HoxBCJ from C. necator H16 [15, 86, 90]. For A. carboxidus DSM 1086 we did not identify genes for a hydrogenase [26, 61, 86]. Accordingly, Cypionka et al. [35] determined only neglectable hydrogenase activity in this

20

D. Siebert et al.

organism which might be an artifact of the side-activity of the aerobic CO dehydrogenase toward hydrogen [81].

4.3

Genetic Organization of Genes Coding for Proteins of the CBB Cycle

In the α-proteobacterium A. carboxidovorans OM5 the CBB cycle genes are organized as a single cluster on the megaplasmid pHCG3 and the strains OM5 and OM4 show the same nucleotide sequence for the whole cluster [26]. A. carboxidus DSM 1086 and Z. compransoris DSM 1231 have an almost identical architecture of the CBB cycle cluster compared to A. carboxidovorans OM5 (Fig. 4a, b; [26, 61]). In Z. compransoris DSM 1231 the gph gene (DES42_104441) is located outside the cluster and the cbbY gene is absent (Fig. 4b; ASM436299v1; [26, 50]). In the β-proteobacterium H. pseudoflava DSM 1084 one copy of each CBB cycle gene is present, organized in two separated clusters. One harboring the ribulose-1,5bisphosphate carboxylase genes cbbLS, and a second cluster with all other CBB cycle genes, except for a homologue of a gene encoding CbbX in C. necator H16 (Fig. 4c). Both clusters in H. pseudoflava DSM 1084 provide a gene encoding a putative transcriptional regulator (Fig. 4c; [15, 90]). For all the above-mentioned bacteria the genes encoding the regulatory proteins are divergently orientated to the appropriate cluster (Fig. 4). The genes tpiA and rpiA are required for a functional CBB cycle as well as for the non-autotrophic central metabolism and therefore are located outside of the CBB clusters or even missing in the case of rpiA of Z. compransoris DSM 1231 (ASM436299v1; [15, 26, 50, 61, 68, 90]). The latter finding is surprising since RpiA is essential for functionality of the pentose phosphate pathway and the CBB cycle [97].

5 Genetic and Metabolic Engineering Tools of Carboxydotrophic Bacteria A major prerequisite to efficiently study and manipulate microbial metabolism is the accessibility toward genetic engineering tools. While for model organisms such as E. coli or Saccharomyces cerevisiae a huge library for methods of molecular biology is available [98], a versatile toolbox for aerobic carboxydotrophs is still pending and might be challenging to develop [99]. To the best of our knowledge, genetic engineering was only shown for H. pseudoflava strain DSM 1084 and A. carboxidovorans strain OM5 so far [15, 16]. As a prerequisite for genetic engineering, it is essential to efficiently transfer DNA into the host cell. For the α-proteobacterial A. carboxidovorans strain OM5 transformation via electroporation applying a modified E. coli protocol has recently

Exploiting Aerobic Carboxydotrophic Bacteria for Industrial Biotechnology

21

Fig. 4 Genetic organization of genes coding for proteins of the CBB cycle and similarities (represented by grey gradients) of gene products in (a) the β-proteobacteria H. pseudoflava DSM 1084 [15] compared to C. necator H16 [90], (b) in α-proteobacteria A. carboxidus DSM 1086 [61], and (c) Z. compransoris DSM 1231 (ASM436299v1 [50]) compared to A. carboxidovorans OM5 [26]

been established [16]. For H. pseudoflava strain DSM 1084 electroporation was not successful so far. However, transconjugation applying E. coli S17-1 as donor strain can be applied to transform H. pseudoflava strain DSM 1084 with pBBR1MCS and pK19mobsacB derivatives [15]. Also transconjugation with A. carboxidovorans OM5 and E. coli S17-1 was briefly mentioned [100]. While literature describing electroporation protocols of β-proteobacteria is only available to a very limited extent, an approach has been described for the chemolithoautotrophic, ammoniaoxidizing, β-proteobacterium Nitrosomonas europaea [101–103]. This might be a promising starting point for the development of electroporation protocols regarding the β-proteobacterial carboxydotrophs. The broad-host-range vector pBBR1MCS and its derivatives [104], which have already been utilized in many α- and β-proteobacteria such as C. necator H16, B. japonicum, Gluconobacter species, Paracoccus species or bacteria of the Roseobacter clade [40, 105–109], were also successfully employed to drive gene expression (e.g. gfp, mCherry) in H. pseudoflava DSM 1084 and A. carboxidovorans OM5 [15, 16]. To allow IPTG-inducible gene expression, Ptac and lacIq were introduced into pBBR1MCS-2, resulting in pOCEx1 [15]. However, in A. carboxidovorans OM5 this plasmid system showed instability since mCherry was not homogeneously expressed [16]. Another limitation might be the low-copy number of the pBBR1 replicon, as in E. coli only 5–10 copies per cell of plasmids carrying this replicon can be found [110]. Thus, increasing the copy number and improving the functionality of such plasmids as well as establishing novel plasmid systems have to be addressed in future studies to improve (heterologous) gene expression in these carboxydotrophs. A suitable source might be a plasmid toolbox

22

D. Siebert et al.

for proteobacteria which has recently been generated in a systematic approach [111] or the Standard European Vector Architecture (SEVA) which has been successfully applied for C. necator H16 [67, 112]. A suitable promoter library for aerobic carboxydotrophs, which provides constitutively active and inducible promoters, is indispensable for future metabolic engineering approaches. Especially when native promoters should be applied, its functionality might be strongly impacted by the cultivation condition (e.g., heterotrophic vs. autotrophic). For H. pseudoflava DSM 1084, it was shown that induced mCherry expression from Ptac resulted in significantly lower expression levels under heterotrophic and autotrophic conditions compared to expression from endogenous PgapA1, PgapA2 and PaceE2 promoters. Moreover, expression from PgapA2 was significantly reduced under autotrophic conditions [15]. Recently, RNAseq data were employed to select promoters for metabolic flux control in the methylotrophic thermophile Bacillus methanolicus based on transcript abundances (represented by the normalized “reads per kilobase of transcript per million mapped reads” (RPKM) [113]. We used a comparable data set available for A. carboxidovorans OM5 to identify promising promoter regions by the abundances of different gene transcripts under hetero- and autotrophic conditions (Table 3; [16]). Table 3 provides an overview of candidate promoters, which might be constitutively or primarily active under auto- and heterotrophic conditions, respectively. However, the distinct promoter elements upstream of these genes, while already available for the cox cluster in A. carboxidovorans OM5 [72], have to be identified in future studies. Genome engineering via homologous recombination using the plasmid pK19mobsacB [114] was successfully carried out with H. pseudoflava DSM 1084 and A. carboxidovorans strain OM5 [15, 16]. While this system uses kanamycin as selection marker for successful integration into the genome, the desintegration of the plasmid backbone is mediated by the toxicity of levansucrase which is synthesized from sucrose by levansucrase, encoded by sacB from Bacillus subtilis [114]. The functional application of this plasmid in A. carboxidovorans OM5 is interesting, because this strain was explicitly characterized by its inability to utilize sugars such as fructose or glucose [22, 23]. Thus, this recombination system is not only applicable for the mentioned α- and β-proteobacterium, but even a functional sugar utilization system is no prerequisite for this system [16]. Perspectively, the utilization of mRNA-silencing methods based on the RNA-binding protein Hfq might be promising for carboxydotrophic bacteria. This tool has already been applied to engineer model organisms such as E. coli [115, 116], Corynebacterium glutamicum, which lacks Hfq [117], as well as unconventional hosts such as Pseudomonas putida [118]. Hfq is widespread in proteobacteria [119, 120], and gene copies are also present in the chromosome of H. pseudoflava DSM 1084 and A. carboxidovorans OM5 [15, 26]. Another, nowadays well-studied and potent genome editing as well as transcription-silencing tool, are CRISPR- and CRISPRi-based systems [121], which should be considered as well for engineering of aerobic carboxydotrophs. The recent developments in this field, especially regarding H. pseudoflava DSM 1084, led to the first described metabolic engineering of an aerobic carboxydotroph,

Exploiting Aerobic Carboxydotrophic Bacteria for Industrial Biotechnology

23

Table 3 Candidate genes for promoter selection to control gene expression in A. carboxidovorans OM5 based on RNAseq data [16] Locus tag (gene symbol) Product/function Primarily active under autotrophic conditionsd OCA5_pHCG300300 CO-dehydrogenase small/large/ OCA5_pHCG300310 medium chain CoxS/L/M OCA5_pHCG300290 (coxS/coxL/coxM) OCA5_pHCG300640 Uptake hydrogenase small/large OCA5_pHCG300630 subunit HoxS/L (hoxS/hoxL) OCA5_pHCG300480 Ribulose bisphosphate carboxylOCA5_pHCG300470 ase small/large chain CbbS/L (cbbS/cbbL) Primarily active under heterotrophic conditionse OCA5_c06660 Hypothetical proteins OCA5_c06650 (OCA5_c06660/ OCA5_c06650) OCA5_c06680 ABC transporter, phosphonate (phnD) transport protein PhnD Constitutively activef OCA5_c22100 Hypothetical protein (OCA5_c22100) OCA5_c17650 tatA/E: Sec-independent protein (OCA5_c17650) translocase OCA5_c12030 Hypothetical protein (OCA5_c12030) OCA5_c14430 Hypothetical protein (OCA5_c14430) OCA5_c26050 Hypothetical protein, flagellin(OCA5_c26050) like protein

Mean RPKMa (autotrophicb)

Mean RPKMa (heterotrophicc)

63,376 29,174 26,111

315 120 222

13,726 6,263

8 10

10,274 8,349

49 23

41 35

8,004 6,337

46

4,727

169,184

276,430

11,992

18,402

15,374

16,081

10,993

15,714

16,389

12,953

Genes in close vicinity to each other are concluded; High abundancy means at least RPKM of 4,000 for autotrophic/heterotrophic and 10,000 for constitutive a Mean value of RPKM (reads per kilobase of transcript per million mapped reads) of all four replicates from data set b Syngas mixture c Sodium acetate as sole carbon and energy source d Inducible by autotrophic conditions ¼ high transcript abundancy only under autotrophic growth conditions e Inducible by heterotrophic conditions ¼ high transcript abundancy only under heterotrophic growth conditions f Constitutive ¼ high transcript abundancy under both growth conditions

resulting in the production of (E)-α-bisabolene from syngas by heterologous expression of the (E)-α-bisabolene synthase gene agBIS from Abies grandis [15]. Furthermore, genetic engineering tools and heterologous expression systems are now available for A. carboxidovorans OM5 [16], laying out the fundament for future metabolic engineering studies with this organism.

24

D. Siebert et al.

6 Conclusion and Outlook Carboxydotrophic bacteria have unique traits which render these organisms promising candidates for industrial biotechnology to convert CO and O2 containing gas streams into higher value chemicals and fuels. Although in the 1970s several representatives have been isolated and well characterized [17, 22, 35, 51], there is a huge gap in the analysis of the genomic organization, the metabolism and its regulation under heterotrophic but especially under autotrophic conditions. Systems level analysis applying omics-technologies and metabolic modelling will provide a comprehensive understanding as prerequisite to engineer such microbes for autotrophic or mixotrophic processes. However, to realize efficient genetic engineering procedures, a more sophisticated toolbox has to be established for this group of bacteria. Even more challenging is the scale-up from lab to industrial scale. Gas fermentation approaches in general face mass transfer issues due to the low solubility of H2 and CO [5] and aerobic processes in addition have to provide sufficient oxygen. Besides the toxicity of CO, aerobic gas fermentation with H2 and/or CO led to explosion hazard, which requires production plants with explosion prevention. These points must be carefully addressed and advanced bioprocesses have to be realized in the future. One promising solution to overcome such limitations might be the application of aerobic carboxydotrophs in bioelectrochemical systems, which in principle can provide the gaseous compounds in situ and on demand [122]. Besides the here presented candidates, nature might provide more so far unexplored carboxydotrophs with even improved intrinsic properties which are waiting to be explored. Additional Information The yet unpublished data presented in Figs. 2, 3, and 4 were all created by applying Easyfig 2.2.5 [123] for visualization, the tBLASTx algorithm of NCBI BLAST+ 2.12.0 [86] and Clone Manager Professional edition, Version 10.0 (Scientific & Educational Software, Westminster, Colorado, USA). All pictures were compiled with the software Inkscape 0.92.3 (https://inkscape.org/). Authors Contributions DS, BE and BB conceived and wrote the paper. All authors read and approved the manuscript. Funding This work was funded by the Deutsche Forschungsgemeinschaft (grant BL1408/3-1). Conflict of Interests The authors declare no competing interests.

References 1. Kiefer D, Merkel M, Lilge L, Henkel M, Hausmann R (2021) From acetate to bio-based products: underexploited potential for industrial biotechnology. Trends Biotechnol 39:397– 411. https://doi.org/10.1016/j.tibtech.2020.09.004

Exploiting Aerobic Carboxydotrophic Bacteria for Industrial Biotechnology

25

2. Bengelsdorf FR, Beck MH, Erz C, Hoffmeister S, Karl MM, Riegler P, Wirth S, Poehlein A, Weuster-Botz D, Dürre P (2018) Bacterial anaerobic synthesis gas (Syngas) and CO2+H2 fermentation. Adv Appl Microbiol 103:143–221. https://doi.org/10.1016/bs.aambs.2018. 01.002 3. Köpke M, Simpson SD (2020) Pollution to products: recycling of ‘above ground’ carbon by gas fermentation. Curr Opin Biotechnol 65:180–189. https://doi.org/10.1016/j.copbio.2020. 02.017 4. Molitor B, Richter H, Martin ME, Jensen RO, Juminaga A, Mihalcea C, Angenent LT (2016) Carbon recovery by fermentation of CO-rich off gases – turning steel mills into biorefineries. Bioresour Technol 215:386–396. https://doi.org/10.1016/j.biortech.2016.03.094 5. Takors R, Kopf M, Mampel J, Bluemke W, Blombach B, Eikmanns B, Bengelsdorf FR, Weuster-Botz D, Dürre P (2018) Using gas mixtures of CO, CO2 and H2 as microbial substrates: the do’s and don’ts of successful technology transfer from laboratory to production scale. J Microbial Biotechnol 11:606–625. https://doi.org/10.1111/1751-7915.13270 6. Nesbitt ER (2020) Using waste carbon feedstocks to produce chemicals. https://www.usitc. gov/publications/332/working_papers/using_waste_carbon_feedstocks_to_produce_ chemicals_0.pdf. Accessed 11 August 2021 7. Schakel W, Hung CR, Tokheim L-A, Strømman AH, Worrell E, Ramírez A (2018) Impact of fuel selection on the environmental performance of post-combustion calcium looping applied to a cement plant. Appl Energy 210:75–87. https://doi.org/10.1016/j.apenergy.2017.10.123 8. United States Environmental Protection Agency (2021) Air emissions inventories: how are oxides of nitrogen (NOx) defined in the NEI? https://www.epa.gov/air-emissions-inventories/ how-are-oxides-nitrogen-nox-defined-nei. Accessed 5 May 2021 9. Xu X, Song C, Wincek R, Andresen JM, Miller BG, Scaroni AW (2003) Separation of CO2 from power plant flue gas using a novel CO2 “Molecular Basket” adsorbent. ACS Div Fuel Chem Preprints 48:162–163 10. Liew F, Martin ME, Tappel RC, Heijstra BD, Mihalcea C, Köpke M (2016) Gas fermentationA flexible platform for commercial scale production of low-carbon-fuels and chemicals from waste and renewable feedstocks. Front Microbiol 7:694. https://doi.org/10.3389/fmicb.2016. 00694 11. Andrew RM (2019) Global CO2 emissions from cement production, 1928–2018. Earth Syst Sci Data 11:1675–1710. https://doi.org/10.5194/essd-11-1675-2019 12. Fast AG, Papoutsakis ET (2012) Stoichiometric and energetic analyses of non-photosynthetic CO2-fixation pathways to support synthetic biology strategies for production of fuels and chemicals. Curr Opin Chem Eng 1:380–395. https://doi.org/10.1016/j.coche.2012.07.005 13. Diender M, Stams AJM, Sousa DZ (2015) Pathways and bioenergetics of anaerobic carbon monoxide fermentation. Front Microbiol 6:1275. https://doi.org/10.3389/fmicb.2015.01275 14. Emerson DF, Stephanopoulos G (2019) Limitations in converting waste gases to fuels and chemicals. Curr Opin Biotechnol 59:39–45. https://doi.org/10.1016/j.copbio.2019.02.004 15. Grenz S, Baumann PT, Rückert C, Nebel BA, Siebert D, Schwentner A, Eikmanns BJ, Hauer B, Kalinowski J, Takors R, Blombach B (2019) Exploiting Hydrogenophaga pseudoflava for aerobic syngas-based production of chemicals. Metab Eng 55:220–230. https://doi.org/10.1016/j.ymben.2019.07.006 16. Siebert D, Busche T, Metz AY, Smaili M, Queck BAW, Kalinowski J, Eikmanns BJ (2020) Genetic engineering of Oligotropha carboxidovorans strain OM5-A promising candidate for the aerobic utilization of synthesis gas. ACS Synth Biol 9:1426–1440. https://doi.org/10.1021/ acssynbio.0c00098 17. Cypionka H, Meyer O (1982) Influence of carbon monoxide on growth and respiration of carboxydobacteria and other aerobic organisms. FEMS Microbiol Lett 15:209–214. https:// doi.org/10.1111/j.1574-6968.1982.tb00069.x 18. Cypionka H, Meyer O (1983) The cytochrome composition of carboxydotrophic bacteria. Arch Microbiol 135:293–298. https://doi.org/10.1007/BF00413484

26

D. Siebert et al.

19. Cordero PRF, Bayly K, Man Leung P, Huang C, Islam ZF, Schittenhelm RB, King GM, Greening C (2019) Atmospheric carbon monoxide oxidation is a widespread mechanism supporting microbial survival. ISME J 13:2868–2881. https://doi.org/10.1038/s41396-0190479-8 20. DePoy AN, King GM, Ohta H (2020) Anaerobic carbon monoxide uptake by microbial communities in volcanic deposits at different stages of successional development on O-yama Volcano, Miyake-jima, Japan. Microorganisms 9. https://doi.org/10.3390/ microorganisms9010012 21. Brigham C (2019) Perspectives for the biotechnological production of biofuels from CO2 and H2 using Ralstonia eutropha and other ‘Knallgas’ bacteria. Appl Microbiol Biotechnol 103: 2113–2120. https://doi.org/10.1007/s00253-019-09636-y 22. Meyer O, Schlegel HG (1983) Biology of aerobic carbon monoxide-oxidizing bacteria. Annu Rev Microbiol 37:277–310. https://doi.org/10.1146/annurev.mi.37.100183.001425 23. Meyer O, Schlegel HG (1978) Reisolation of the carbon monoxide utilizing hydrogen bacterium Pseudomonas carboxydovorans (Kistner) comb. nov. Arch Microbiol 118:35–43. https://doi.org/10.1007/BF00406071 24. Meyer O, Stackebrandt E, Auling G (1993) Reclassification of ubiquinone Q-10 containing carboxidotrophic bacteria: transfer of “[Pseudomonas] carboxydovorans” OM5T to Oligotropha, gen. nov., as Oligotropha carboxidovorans, comb. nov., transfer of “[Alcaligenes] carboxydus” DSM 1086T to Carbophilus, gen. nov., as Carbophilus carboxidus, comb. nov., transfer of “[Pseudomonas] compransoris” DSM 1231T to Zavarzinia, gen. nov., as Zavarzinia compransoris, comb. nov., and amended descriptions of the new genera. Syst Appl Microbiol 16:390–395. https://doi.org/10.1016/S0723-2020(11) 80271-7 25. Hördt A, López MG, Meier-Kolthoff JP, Schleuning M, Weinhold L-M, Tindall BJ, Gronow S, Kyrpides NC, Woyke T, Göker M (2020) Analysis of 1,000+ type-strain genomes substantially improves taxonomic classification of Alphaproteobacteria. Front Microbiol 11: 468. https://doi.org/10.3389/fmicb.2020.00468 26. Volland S, Rachinger M, Strittmatter A, Daniel R, Gottschalk G, Meyer O (2011) Complete genome sequences of the chemolithoautotrophic Oligotropha carboxidovorans strains OM4 and OM5. J Bacteriol 193:5043. https://doi.org/10.1128/JB.05619-11 27. Paul D, Kumar R, Nanduri B, French T, Pendarvis K, Brown A, Lawrence ML, Burgess SC (2011) Proteome and membrane fatty acid analyses on Oligotropha carboxidovorans OM5 grown under chemolithoautotrophic and heterotrophic conditions. PLoS One 6:e17111. https://doi.org/10.1371/journal.pone.0017111 28. Hille R, Dingwall S, Wilcoxen J (2015) The aerobic CO dehydrogenase from Oligotropha carboxidovorans. J Biol Inorg Chem 20:243–251. https://doi.org/10.1007/s00775-014-1188-4 29. Kalimuthu P, Petitgenet M, Niks D, Dingwall S, Harmer JR, Hille R, Bernhardt PV (2020) The oxidation-reduction and electrocatalytic properties of CO dehydrogenase from Oligotropha carboxidovorans. Biochim Biophys Acta Bioenerg 1861:148118. https://doi.org/10.1016/j. bbabio.2019.148118 30. Kaufmann P, Duffus BR, Teutloff C, Leimkühler S (2018) Functional studies on Oligotropha carboxidovorans molybdenum-copper CO dehydrogenase produced in Escherichia coli. Biochemistry 57:2889–2901. https://doi.org/10.1021/acs.biochem.8b00128 31. Santiago B, Meyer O (1997) Purification and molecular characterization of the H2 uptake membrane-bound NiFe-hydrogenase from the carboxidotrophic bacterium Oligotropha carboxidovorans. J Bacteriol 179:6053–6060. https://doi.org/10.1128/jb.179.19.6053-6060. 1997 32. Choma A, Zamłyńska K, Mazur A, Pastuszka A, Kaczyński Z, Komaniecka I (2020) Lipid A from Oligotropha carboxidovorans lipopolysaccharide that contains two galacturonic acid residues in the backbone and malic acid A tertiary acyl substituent. Int J Mol Sci 21. https:// doi.org/10.3390/ijms21217991

Exploiting Aerobic Carboxydotrophic Bacteria for Industrial Biotechnology

27

33. Komaniecka I, Choma A, Zamlynska K, Sroka-Bartnicka A, Sowinski P (2017) Structure of O-specific polysaccharide of Oligotropha carboxidovorans OM5 – a wastewater bacterium. Carbohydr Res 439:30–34. https://doi.org/10.1016/j.carres.2017.01.001 34. Cypionka H, Meyer O (1983) Carbon monoxide-insensitive respiratory chain of Pseudomonas carboxydovorans. J Bacteriol 156:1178–1187. https://doi.org/10.1128/JB.156.3.1178-1187. 1983 35. Cypionka H, Meyer O, Schlegel HG (1980) Physiological characteristics of various species of strains of carboxydobacteria. Arch Microbiol 127:301–307. https://doi.org/10.1007/ BF00427208 36. Stouthamer AH (1973) A theoretical study on the amount of ATP required for synthesis of microbial cell material. Antonie Van Leeuwenhoek 39:545–565. https://doi.org/10.1007/ BF02578899 37. Kiessling M, Meyer O (1982) Profitable oxidation of carbon monoxide or hydrogen during heterotrophic growth of Pseudomonas carboxydoflava. FEMS Microbiol Lett 13:333–338. https://doi.org/10.1111/j.1574-6968.1982.tb08283.x 38. Meyer O, Lalucat J, Schlegel HG (1980) Pseudomonas carboxydohydrogena (Sanjieva and Zavarzin) comb. nov., a monotrichous, nonbudding, strictly aerobic, carbon monoxideutilizing hydrogen bacterium previously assigned to Seliberia. Int J Syst Bacteriol 30:189– 195. https://doi.org/10.1099/00207713-30-1-189 39. Lorite MJ, Tachil J, Sanjuán J, Meyer O, Bedmar EJ (2000) Carbon monoxide dehydrogenase activity in Bradyrhizobium japonicum. Appl Environ Microbiol 66:1871–1876. https://doi. org/10.1128/aem.66.5.1871-1876.2000 40. Heinrich D, Raberg M, Steinbüchel A (2018) Studies on the aerobic utilization of synthesis gas (syngas) by wild type and recombinant strains of Ralstonia eutropha H16. J Microbial Biotechnol 11:647–656. https://doi.org/10.1111/1751-7915.12873 41. Willems A, Busse J, Goor M, Pot B, Falsen E, Jantzen E, Hoste B, Gillis M, Kersters K, Auling G, de Ley J (1989) Hydrogenophaga, a new genus of hydrogen-oxidizing bacteria that includes Hydrogenophaga flava comb. nov. (formerly Pseudomonas flava), Hydrogenophaga palleronii (formerly Pseudomonas palleronii), Hydrogenophaga pseudoflava (formerly Pseudomonas pseudoflava and “Pseudomonas carboxydoflava”), and Hydrogenophaga taeniospiralis (formerly Pseudomonas taeniospiralis). Int J Syst Bacteriol 39:319–333. https://doi.org/10.1099/00207713-39-3-319 42. Haake G, Kaesler-Neumann I, Hennig H, Meltzer TH, Jornitz MW (2012) The importance of accurate microorganism identification in microbial challenge tests of membrane filters. Part II. The comparison of Hydrogenophaga pseudoflava ATTC 33668 and curvibacter sp. ATCC 700892 by microbial challenge tests with membrane filters. PDA J Pharm Sci Technol 66:346– 353. https://doi.org/10.5731/pdajpst.2012.00876 43. Lee A, McVey J, Faustino P, Lute S, Sweeney N, Pawar V, Khan M, Brorson K, Hussong D (2010) Use of Hydrogenophaga pseudoflava penetration to quantitatively assess the impact of filtration parameters for 0.2-micrometer-pore-size filters. Appl Environ Microbiol 76:695– 700. https://doi.org/10.1128/AEM.01825-09 44. Sundaram S, Lewis M, Eisenhuth J, Howard G, Larson B (2002) Method for qualifying microbial removal performance of 0.1 micron rated filters. Part IV: Retention of Hydrogenophaga pseudoflava (ATCC 700892) and Ralstonia pickettii (ATCC 700591) by 0.2 and 0.22 micron rated filters. PDA J Pharm Sci Technol 56:150–171 45. Auling G, Reh M, Lee CM, Schlegel HG (1978) Pseudomonas pseudoflava, a new species of hydrogen-oxidizing bacteria: its differentiation from Pseudomonas flava and other yellowpigmented, gram-negative, hydrogen-oxidizing species. Int J Syst Bacteriol 28:82–95. https:// doi.org/10.1099/00207713-28-1-82 46. Choi MH, Yoon SC, Lenz RW (1999) Production of poly(3-hydroxybutyric acid-co-4hydroxybutyric acid) and poly(4-hydroxybutyric acid) without subsequent degradation by Hydrogenophaga pseudoflava. Appl Environ Microbiol 65:1570–1577. https://doi.org/10. 1128/AEM.65.4.1570-1577.1999

28

D. Siebert et al.

47. Koller M, Hesse P, Bona R, Kutschera C, Atlić A, Braunegg G (2007) Potential of various archae- and eubacterial strains as industrial polyhydroxyalkanoate producers from whey. Macromol Biosci 7:218–226. https://doi.org/10.1002/mabi.200600211 48. Ramsay BA, Lomaliza K, Chavarie C, Dubé B, Bataille P, Ramsay JA (1990) Production of poly-(beta-hydroxybutyric-co-beta-hydroxyvaleric) acids. Appl Environ Microbiol 56:2093– 2098. https://doi.org/10.1128/AEM.56.7.2093-2098.1990 49. Lee CM, Schlegel HG (1981) Physiological characterization of Pseudomonas pseudoflava GA3. Curr Microbiol 5:333–337. https://doi.org/10.1007/BF01566744 50. Assembly (2012) [Internet]: Bethesda (MD): National Library of Medicine (US), National Center for Biotechnology Information. https://www.ncbi.nlm.nih.gov/assembly/. Accessed 23 June 2021 51. Zavarzin GA, Nozhevnikova AN (1977) Aerobic carboxydobacteria. Microb Ecol 3:305–326. https://doi.org/10.1007/BF02010738 52. Kang BS, Kim YM (1999) Cloning and molecular characterization of the genes for carbon monoxide dehydrogenase and localization of molybdopterin, flavin adenine dinucleotide, and iron-sulfur centers in the enzyme of Hydrogenophaga pseudoflava. J Bacteriol 181:5581– 5590. https://doi.org/10.1128/JB.181.18.5581-5590.1999 53. Meyer O, Gremer L, Ferner R, Ferner M, Dobbek H, Gnida M, Meyer-Klaucke W, Huber R (2000) The role of Se, Mo and Fe in the structure and function of carbon monoxide dehydrogenase. Biol Chem 381:865–876. https://doi.org/10.1515/BC.2000.108 54. Anzai Y, Kim H, Park JY, Wakabayashi H, Oyaizu H (2000) Phylogenetic affiliation of the pseudomonads based on 16S rRNA sequence. Int J Syst Evol Microbiol 50(Pt 4):1563–1589. https://doi.org/10.1099/00207713-50-4-1563 55. Kim YM, Hegeman GD (1981) Purification and some properties of carbon monoxide dehydrogenase from Pseudomonas carboxydohydrogena. J Bacteriol 148:904–911. https://doi.org/ 10.1128/jb.148.3.904-911.1981 56. DSMZ GmbH (2009) Medium 133: carbon monoxide oxidizer medium. https://www.dsmz. de/microorganisms/medium/pdf/DSMZ_Medium133.pdf. Accessed 16 June 2021 57. DSMZ GmbH (2017) Medium 141: ethanogenium medium (H2/CO2). https://www.dsmz.de/ microorganisms/medium/pdf/DSMZ_Medium141.pdf. Accessed 16 June 2021 58. Volova T, Zhila N, Shishatskaya E (2015) Synthesis of poly(3-hydroxybutyrate) by the autotrophic CO-oxidizing bacterium Seliberia carboxydohydrogena Z-1062. J Ind Microbiol Biotechnol 42:1377–1387. https://doi.org/10.1007/s10295-015-1659-9 59. Kalyuzhnaya MG, de Marco P, Bowerman S, Pacheco CC, Lara JC, Lidstrom ME, Chistoserdova L (2006) Methyloversatilis universalis gen. nov., sp. nov., a novel taxon within the Betaproteobacteria represented by three methylotrophic isolates. Int J Syst Evol Microbiol 56:2517–2522. https://doi.org/10.1099/ijs.0.64422-0 60. Artuso I, Turrini P, Pirolo M, Lucidi M, Tescari M, Visaggio D, Mansi A, Lugli GA, Ventura M, Visca P (2021) Phylogenomic analysis and characterization of carbon monoxide utilization genes in the family Phyllobacteriaceae with reclassification of Aminobacter carboxidus (Meyer et al. 1993, Hördt et al. 2020) as Aminobacter lissarensis comb. nov. (McDonald et al. 2005). Syst Appl Microbiol 44:126199. https://doi.org/10.1016/j.syapm. 2021.126199 61. Turrini P, Artuso I, Tescari M, Lugli GA, Frangipani E, Ventura M, Visca P (2020) Draft genome sequence of the carboxydotrophic alphaproteobacterium Aminobacter carboxidus type strain DSM 1086. Microbiol Resour Announc 9. https://doi.org/10.1128/MRA.01170-20 62. Kuykendall LD (2015) Bradyrhizobium. In: Whitman WB, Rainey F, Kämpfer P, Trujillo M, Chun J, DeVos P, Hedlund B, Dedysh S (eds) Bergey’s manual of systematics of archaea and bacteria, vol 60. Wiley, pp 1–11 63. Kämpfer P, Glaeser SP, Busse H-J (2013) Transfer of Bacillus schlegelii to a novel genus and proposal of Hydrogenibacillus schlegelii gen. nov., comb. nov. Int J Syst Evol Microbiol 63: 1723–1727. https://doi.org/10.1099/ijs.0.045146-0

Exploiting Aerobic Carboxydotrophic Bacteria for Industrial Biotechnology

29

64. Hänzelmann P, Hofmann B, Meisen S, Meyer O (1999) The redox centers in the molybdo iron-sulfur flavoprotein CO dehydrogenase from the thermophilic carboxidotrophic bacterium Pseudomonas thermocarboxydovorans. FEMS Microbiol Lett 176:139–145. https://doi.org/ 10.1111/j.1574-6968.1999.tb13654.x 65. Lyons CM, Justin P, Colby J, Williams E (1984) Isolation, characterization and autotrophic metabolism of a moderately thermophilic carboxydobacterium, Pseudomonas thermocarboxydovorans sp. nov. Microbiology 130:1097–1105. https://doi.org/10.1099/ 00221287-130-5-1097 66. Gleizer S, Ben-Nissan R, Bar-On YM, Antonovsky N, Noor E, Zohar Y, Jona G, Krieger E, Shamshoum M, Bar-Even A, Milo R (2019) Conversion of Escherichia coli to generate all biomass carbon from CO2. Cell 179:1255–1263.e12. https://doi.org/10.1016/j.cell.2019. 11.009 67. Panich J, Fong B, Singer SW (2021) Metabolic engineering of Cupriavidus necator H16 for sustainable biofuels from CO2. Trends Biotechnol 39:412–424. https://doi.org/10.1016/j. tibtech.2021.01.001 68. Berg JM, Tymoczko JL, Gatto GJ, Stryer L (2018) Stryer biochemie. Springer, Berlin 69. Brigham CJ, Gai CS, Lu J, Speth DR, Worden RM, Sinskey AJ (2013) Engineering Ralstonia eutropha for production of isobutanol from CO2, H2, and O2. In: Lee JW (ed) Advanced biofuels and bioproducts, vol 3. Springer, New York, pp 1065–1090 70. Burgdorf T, Lenz O, Buhrke T, van der Linden E, Jones AK, Albracht SPJ, Friedrich B (2005) NiFe-hydrogenases of Ralstonia eutropha H16: modular enzymes for oxygen-tolerant biological hydrogen oxidation. J Mol Microbiol Biotechnol 10:181–196. https://doi.org/10.1159/ 000091564 71. Mörsdorf G, Frunzke K, Gadkari D, Meyer O (1992) Microbial growth on carbon monoxide. Biodegradation 3. https://doi.org/10.1007/BF00189635 72. Santiago B, Schübel U, Egelseer C, Meyer O (1999) Sequence analysis, characterization and CO-specific transcription of the cox gene cluster on the megaplasmid pHCG3 of Oligotropha carboxidovorans. Gene 236:115–124. https://doi.org/10.1016/S0378-1119(99)00245-0 73. The UniProt Consortium (2021) UniProt: the universal protein knowledgebase in 2021. Nucleic Acids Res 49:D480–D489. https://doi.org/10.1093/nar/gkaa1100 74. Meyer O, Jacobitz S, Krüger B (1986) Biochemistry and physiology of aerobic carbon monoxide-utilizing bacteria. FEMS Microbiol Lett 39:161–179. https://doi.org/10.1111/j. 1574-6968.1986.tb01858.x 75. Inoue M, Nakamoto I, Omae K, Oguro T, Ogata H, Yoshida T, Sako Y (2018) Structural and phylogenetic diversity of anaerobic carbon-monoxide dehydrogenases. Front Microbiol 9: 3353. https://doi.org/10.3389/fmicb.2018.03353 76. Ragsdale SW, Kumar M (1996) Nickel-containing carbon monoxide dehydrogenase/acetylCoA synthase. Chem Rev 96:2515–2540. https://doi.org/10.1021/cr950058+ 77. Meyer O, Schlegel HG (1979) Oxidation of carbon monoxide in cell extracts of Pseudomonas carboxydovorans. J Bacteriol 137:811–817. https://doi.org/10.1128/jb.137.2.811-817.1979 78. Wilcoxen J, Zhang B, Hille R (2011) Reaction of the molybdenum- and copper-containing carbon monoxide dehydrogenase from Oligotropha carboxydovorans with quinones. Biochemistry 50:1910–1916. https://doi.org/10.1021/bi1017182 79. Rohde M, Mayer F, Jacobitz S, Meyer O (1985) Attachment of CO dehydrogenase to the cytoplasmic membrane is limiting the respiratory rate of Pseudomonas carboxydovorans. FEMS Microbiol Lett 28:141–144. https://doi.org/10.1111/j.1574-6968.1985.tb00780.x 80. Santiago B, Meyer O (1996) Characterization of hydrogenase activities associated with the molybdenum CO dehydrogenase from Oligotropha carboxidovorans. FEMS Microbiol Lett 136:157–162. https://doi.org/10.1111/j.1574-6968.1996.tb08042.x 81. Wilcoxen J, Hille R (2013) The hydrogenase activity of the molybdenum/copper-containing carbon monoxide dehydrogenase of Oligotropha carboxidovorans. J Biol Chem 288:36052– 36060. https://doi.org/10.1074/jbc.M113.522441

30

D. Siebert et al.

82. Meyer O, Schlegel HG (1980) Carbon monoxide:methylene blue oxidoreductase from Pseudomonas carboxydovorans. J Bacteriol 141:74–80. https://doi.org/10.1128/jb.141.1.74-80. 1980 83. Dobbek H, Gremer L, Kiefersauer R, Huber R, Meyer O (2002) Catalysis at a dinuclear [CuSMo(¼O)OH] cluster in a CO dehydrogenase resolved at 1.1-Å resolution. Proc Natl Acad Sci U S A 99:15971–15976. https://doi.org/10.1073/pnas.212640899 84. Dobbek H, Gremer L, Meyer O, Huber R (1999) Crystal structure and mechanism of CO dehydrogenase, a molybdo iron-sulfur flavoprotein containing S-selanylcysteine. Proc Natl Acad Sci U S A 96:8884–8889. https://doi.org/10.1073/pnas.96.16.8884 85. Hänzelmann P, Dobbek H, Gremer L, Huber R, Meyer O (2000) The effect of intracellular molybdenum in Hydrogenophaga pseudoflava on the crystallographic structure of the selenomolybdo-iron-sulfur flavoenzyme carbon monoxide dehydrogenase. J Mol Biol 301:1221– 1235. https://doi.org/10.1006/jmbi.2000.4023 86. Altschul SF, Madden TL, Schäffer AA, Zhang J, Zhang Z, Miller W, Lipman DJ (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25:3389–3402. https://doi.org/10.1093/nar/25.17.3389 87. Bowien B, Schlegel HG (1981) Physiology and biochemistry of aerobic hydrogen-oxidizing bacteria. Annu Rev Microbiol 35:405–452. https://doi.org/10.1146/annurev.mi.35.100181. 002201 88. Cypionka H, Reijnders WNM, Wielink JE, Oltmann LF, Stouthamer AH (1985) Half reduction potentials and oxygen affinity of the cytochromes of Pseudomonas carboxydovorans. FEMS Microbiol Lett 27:189–193. https://doi.org/10.1111/j.1574-6968.1985.tb00665.x 89. Cypionka H, Verseveld HW, Stouthamer AH (1984) Proton translocation coupled to carbon monoxide-insensitive and -sensitive electron transport in Pseudomonas carboxydovorans. FEMS Microbiol Lett 22:209–213. https://doi.org/10.1111/j.1574-6968.1984.tb00728.x 90. Pohlmann A, Fricke WF, Reinecke F, Kusian B, Liesegang H, Cramm R, Eitinger T, Ewering C, Pötter M, Schwartz E, Strittmatter A, Voss I, Gottschalk G, Steinbüchel A, Friedrich B, Bowien B (2006) Genome sequence of the bioplastic-producing “Knallgas” bacterium Ralstonia eutropha H16. Nat Biotechnol 24:1257–1262. https://doi.org/10.1038/ nbt1244 91. Watson GM, Tabita FR (1997) Microbial ribulose 1,5-bisphosphate carboxylase/oxygenase: a molecule for phylogenetic and enzymological investigation. FEMS Microbiol Lett 146:13–22. https://doi.org/10.1111/j.1574-6968.1997.tb10165.x 92. Gene (2004) [Internet]: Bethesda (MD): National Library of Medicine (US), National Center for Biotechnology Information. https://www.ncbi.nlm.nih.gov/gene/. Accessed 17 July 2021 93. Kanehisa M, Goto S (2000) KEGG: kyoto encyclopedia of genes and genomes. Nucleic Acids Res 28:27–30. https://doi.org/10.1093/nar/28.1.27 94. King GM (2003) Molecular and culture-based analyses of aerobic carbon monoxide oxidizer diversity. Appl Environ Microbiol 69:7257–7265. https://doi.org/10.1128/AEM.69.12. 7257-65 95. King GM, Weber CF (2007) Distribution, diversity and ecology of aerobic CO-oxidizing bacteria. Nat Rev Microbiol 5:107–118. https://doi.org/10.1038/nrmicro1595 96. Kraut M, Hugendieck I, Herwig S, Meyer O (1989) Homology and distribution of CO dehydrogenase structural genes in carboxydotrophic bacteria. Arch Microbiol 152:335–341. https://doi.org/10.1007/BF00425170 97. Zhang R-G, Andersson CE, Savchenko A, Skarina T, Evdokimova E, Beasley S, Arrowsmith CH, Edwards AM, Joachimiak A, Mowbray SL (2003) Structure of Escherichia coli ribose-5phosphate isomerase: a ubiquitous enzyme of the pentose phosphate pathway and the Calvin cycle. Structure 11:31–42. https://doi.org/10.1016/s0969-2126(02)00933-4 98. Green MR, Sambrook J (2012) Molecular cloning: a laboratory manual.4th edn. Cold Spring Harbor Laboratory Press, Cold Spring Harbor

Exploiting Aerobic Carboxydotrophic Bacteria for Industrial Biotechnology

31

99. Riley LA, Guss AM (2021) Approaches to genetic tool development for rapid domestication of non-model microorganisms. Biotechnol Biofuels 14:30. https://doi.org/10.1186/s13068020-01872-z 100. Pelzmann AM, Mickoleit F, Meyer O (2014) Insights into the posttranslational assembly of the Mo-, S- and Cu-containing cluster in the active site of CO dehydrogenase of Oligotropha carboxidovorans. J Biol Inorg Chem 19:1399–1414. https://doi.org/10.1007/s00775-0141201-y 101. Hommes NG, Sayavedra-Soto LA, Arp DJ (1996) Mutagenesis of hydroxylamine oxidoreductase in Nitrosomonas europaea by transformation and recombination. J Bacteriol 178: 3710–3714. https://doi.org/10.1128/jb.178.13.3710-3714.1996 102. Koops H-P, Purkhold U, Pommerening-Röser A, Timmermann G, Wagner M (2006) The lithoautotrophic ammonia-oxidizing bacteria. In: Dworkin M, Falkow S, Rosenberg E, Schleifer K-H, Stackebrandt E (eds) The prokaryotes, vol 39. Springer, New York, pp 778–811 103. Sayavedra-Soto LA, Stein LY (2011) Genetic transformation of ammonia-oxidizing bacteria. Methods Enzymol 486:389–402. https://doi.org/10.1016/B978-0-12-381294-0.00017-1 104. Kovach ME, Elzer PH, Steven Hill D, Robertson GT, Farris MA, Roop RM, Peterson KM (1995) Four new derivatives of the broad-host-range cloning vector pBBR1MCS, carrying different antibiotic-resistance cassettes. Gene 166:175–176. https://doi.org/10.1016/03781119(95)00584-1 105. Bartosik D, Baj J, Sochacka M, Piechucka E, Wlodarczyk M (2002) Molecular characterization of functional modules of plasmid pWKS1 of Paracoccus pantotrophus DSM 11072. Microbiology 148:2847–2856. https://doi.org/10.1099/00221287-148-9-2847 106. Fricke PM, Klemm A, Bott M, Polen T (2021) On the way toward regulatable expression systems in acetic acid bacteria: target gene expression and use cases. Eur J Appl Microbiol Biotechnol 105:3423–3456. https://doi.org/10.1007/s00253-021-11269-z 107. Johnson AO, Gonzalez-Villanueva M, Tee KL, Wong TS (2018) An engineered constitutive promoter set with broad activity range for Cupriavidus necator H16. ACS Synth Biol 7:1918– 1928. https://doi.org/10.1021/acssynbio.8b00136 108. Noh J-G, Jeon H-E, So J-S, Chang W-S (2015) Effects of the Bradyrhizobium japonicum waaL (rfaL) gene on hydrophobicity, motility, stress tolerance, and symbiotic relationship with soybeans. Int J Mol Sci 16:16778–16791. https://doi.org/10.3390/ijms160816778 109. Piekarski T, Buchholz I, Drepper T, Schobert M, Wagner-Doebler I, Tielen P, Jahn D (2009) Genetic tools for the investigation of Roseobacter clade bacteria. BMC Microbiol 9:265. https://doi.org/10.1186/1471-2180-9-265 110. Jahn M, Vorpahl C, Hübschmann T, Harms H, Müller S (2016) Copy number variability of expression plasmids determined by cell sorting and droplet digital PCR. Microb Cell Fact 15: 211. https://doi.org/10.1186/s12934-016-0610-8 111. Schuster LA, Reisch CR (2021) A plasmid toolbox for controlled gene expression across the proteobacteria. Nucleic Acids Res. https://doi.org/10.1093/nar/gkab496 112. Martínez-García E, Goñi-Moreno A, Bartley B, McLaughlin J, Sánchez-Sampedro L, Pascual Del Pozo H, Prieto Hernández C, Marletta AS, de Lucrezia D, Sánchez-Fernández G, Fraile S, de Lorenzo V (2020) SEVA 3.0: an update of the standard European vector architecture for enabling portability of genetic constructs among diverse bacterial hosts. Nucleic Acids Res 48: D1164–D1170. https://doi.org/10.1093/nar/gkz1024 113. Irla M, Hakvåg S, Brautaset T (2021) Developing a riboswitch-mediated regulatory system for metabolic flux control in thermophilic Bacillus methanolicus. Int J Mol Sci 22. https://doi.org/ 10.3390/ijms22094686 114. Schäfer A, Tauch A, Jäger W, Kalinowski J, Thierbach G, Pühler A (1994) Small mobilizable multi-purpose cloning vectors derived from the Escherichia coli plasmids pK18 and pK19: selection of defined deletions in the chromosome of Corynebacterium glutamicum. Gene 145: 69–73. https://doi.org/10.1016/0378-1119(94)90324-7

32

D. Siebert et al.

115. Na D, Yoo SM, Chung H, Park H, Park JH, Lee SY (2013) Metabolic engineering of Escherichia coli using synthetic small regulatory RNAs. Nat Biotechnol 31:170–174. https://doi.org/10.1038/nbt.2461 116. Yoo SM, Na D, Lee SY (2013) Design and use of synthetic regulatory small RNAs to control gene expression in Escherichia coli. Nat Protoc 8:1694–1707. https://doi.org/10.1038/nprot. 2013.105 117. Sun D, Chen J, Wang Y, Li M, Rao D, Guo Y, Chen N, Zheng P, Sun J, Ma Y (2019) Metabolic engineering of Corynebacterium glutamicum by synthetic small regulatory RNAs. J Ind Microbiol Biotechnol 46:203–208. https://doi.org/10.1007/s10295-018-02128-4 118. Apura P, Saramago M, Peregrina A, Viegas SC, Carvalho SM, Saraiva LM, Arraiano CM, Domingues S (2020) Tailor-made sRNAs: a plasmid tool to control the expression of target mRNAs in Pseudomonas putida. Plasmid 109:102503. https://doi.org/10.1016/j.plasmid. 2020.102503 119. Feliciano JR, Seixas AMM, Pita T, Leitão JH (2020) Comparative genomics and evolutionary analysis of RNA-binding proteins of Burkholderia cenocepacia J2315 and other members of the B. cepacia complex. Genes (Basel):11. https://doi.org/10.3390/genes11020231 120. Sun X, Zhulin I, Wartell RM (2002) Predicted structure and phyletic distribution of the RNA-binding protein Hfq. Nucleic Acids Res 30:3662–3671. https://doi.org/10.1093/nar/ gkf508 121. Adli M (2018) The CRISPR tool kit for genome editing and beyond. Nat Commun 9:1911. https://doi.org/10.1038/s41467-018-04252-2 122. Utesch T, Zeng A-P (2018) A novel all-in-one electrolysis electrode and bioreactor enable better study of electrochemical effects and electricity-aided bioprocesses. Eng Life Sci 18: 600–610. https://doi.org/10.1002/elsc.201700198 123. Sullivan MJ, Petty NK, Beatson SA (2011) Easyfig: a genome comparison visualizer. Bioinformatics 27:1009–1010. https://doi.org/10.1093/bioinformatics/btr039

Adv Biochem Eng Biotechnol (2022) 180: 33–56 https://doi.org/10.1007/10_2021_172 © The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 Published online: 22 July 2021

Process Engineering Aspects for the Microbial Conversion of C1 Gases Dirk Weuster-Botz

Contents 1 C1-Gases as Microbial Carbon Source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Gas Fermentations with Acetogenic Microorganisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Autotrophic Growth and Low Gas Solubilities in Water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Gas–Liquid Mass Transfer at Low Volumetric Power Input . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 Autotrophic Growth Kinetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4 Autotrophic Gas Fermentation in Bubble Column Reactors . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5 Continuous Gas Fermentation Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.6 Enlarging the Product Spectrum of Gas Fermentation by Co-cultivation . . . . . . . . . . . . 2.7 Requirements for Syngas Purification in Gas Fermentation . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Conversion of CO2 with Microalgae . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Open Photobioreactors for Microalgae Mass Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Requirements for Combustion Gas Purification for Photoautotrophic Processes . . . . . 4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

34 35 35 37 38 39 40 44 45 46 46 49 50 52

Abstract Industrially applied bioprocesses for the reduction of C1 gases (CO2 and/or CO) are based in particular on (syn)gas fermentation with acetogenic bacteria and on photobioprocesses with microalgae. In each case, process engineering characteristics of the autotrophic microorganisms are specified and process engineering aspects for improving gas and electron supply are summarized before suitable bioreactor configurations are discussed for the production of organic products under given economic constraints. Additionally, requirements for the purity of C1 gases are summarized briefly. Finally, similarities and differences in microbial CO2 valorization are depicted comparing gas fermentations with acetogenic bacteria and photobioprocesses with microalgae.

D. Weuster-Botz (*) Institute of Biochemical Engineering, Technical University of Munich, Garching, Germany e-mail: [email protected]

34

D. Weuster-Botz

Graphical Abstract

H2

vvvvvvvvv

CO2

CO2

Keywords Acetogen, Bubble column reactor, Chain elongation, Co-cultivation, Combustion gas impurities, Continuous processes, Gas-lift reactor, Kinetics, Membrane reactor, Microalgae, Open photobioreactor, Raceway pond, Syngas fermentation, Syngas impurities, Thin-layer cascade reactor, Trickle-bed biofilm reactor

1 C1-Gases as Microbial Carbon Source Reducing the emissions of greenhouse gases like CO2 is a major demand for mankind to embank the effect of climate change. In this context great interest has emerged in biological CO2-fixing processes which are able to effectively convert CO2 emissions into multi-carbon organic chemicals and may therefore open the door for the establishment of a circular carbon economy [1]. Interestingly, microorganisms have a whole range of metabolic possibilities for providing electrons for CO2 reduction and the biosynthesis of organic products: – Strictly anaerobic, acetogenic microorganisms can use hydrogen gas (H2) as an electron source under autotrophic conditions. These bacteria reduce CO2 with H2 via the so-called Wood-Ljungdahl pathway to the metabolic intermediate acetylCoA. Since the natural product spectrum starting from acetyl-CoA is essentially limited to the C2 products acetic acid and ethanol with these bacteria, ethanol is produced on an industrial scale from steel mill exhaust gases using acetogenic bacteria since 2018 [2]. – A possibility to use electrons directly for microbial CO2 reduction was described first in 2010 [3]. Here, the provision of electrons takes place at a cathode covered with a biofilm of acetogenic bacteria. As with industrial production processes using microalgae, this can be considered as a “two-dimensional process” because direct transfer of electrons is only possible in thin biofilms and thus large cathode

Process Engineering Aspects for the Microbial Conversion of C1 Gases

35

areas must be installed in a production reactor. Industrial applications are not known so far. – Aerobic, carboxidotrophic microorganisms can provide electrons for CO2 reduction under chemolithoautotrophic conditions [4] by metabolically reacting H2 with oxygen O2 to form water, and using some of the electrons made available to reduce CO2. Although probably the best known carboxidotrophic bacterium Cupriavidus necator (formerly Ralstonia eutropha) has been known for many decades and studied as an efficient producer of polyhydroxyalkanoates (PHA), no industrial exploitation using CO2 as a carbon source has been reported so far. A crucial aspect preventing the industrial use of these aerobic bacteria is the fact that a non-limiting supply of H2 and O2 to the microorganisms is difficult in bioreactors from a safety point of view [5]. – Like plants, photoautotrophic microalgae (prokaryotic cyanobacteria and eukaryotic unicellular organisms) use energy generated by sunlight to reduce and fix CO2. Photoautotrophic microalgae are already being used industrially to produce antioxidants (e.g., ß-carotene), pigments (e.g., astaxanthin), and proteins (e.g., phycocyanin) as intracellular products from CO2 using large-scale, open photobioreactors [6]. Here, large-area photobioreactors are required due to the shallow penetration of sunlight into aqueous algal suspensions (“two-dimensional processes”). The electrochemical CO2 reduction via electrolysis is an alternative to the direct microbial consumption and different companies already established demonstration reactors or pilot plants [7]. Generally, CO2 electrolysis has two main commercially desirable scenarios. First, the production of syngas (CO/H2O) for subsequent methanol production and second, direct electrosynthesis of value-added organic chemicals such as acetate or alcohols [8]. Alternatively, an artificial photosynthesis approach of coupling CO2 electrolysis with bacterial fermentation, in which bacteria are tolerant to different reduction equivalents (CO, H2) is promising as both can be used stoichiometric independently. In a collaborative work, Siemens, Evonik and Covestro recently reported that this approach is already at the current technological status relevant for industrial purposes shown on the example of 1-butanol and 1-hexanol bioelectrosynthesis from CO2 and H2O [9].

2 Gas Fermentations with Acetogenic Microorganisms 2.1

Autotrophic Growth and Low Gas Solubilities in Water

Acetogenic bacteria are able to produce CO2-based chemicals in aqueous media by autotrophic conversion of CO2 with high energetic efficiency of up to 70–90% [10]. Natural products can be acetate/ethanol, butyrate/butanol, 2,3-butanediol, and hexanoate/hexanol. Recombinant acetogens have already been engineered to

36

D. Weuster-Botz cX,max

μ

YAcetate,C

CO CO2 + H2 qAcetate

YX,C

cAcetate,max

Fig. 1 Comparison of autotrophic batch processes with Clostridium aceticum in a stirred-tank bioreactor with continuous gas supply (T ¼ 30 C, pH 8.0, P/V ¼ 6.8 W L
1, FGas ¼ 0.083 vvm). Blue: CO2 + H2 with pCO2,in ¼ 120 mbar and pH 2,in ¼ 480 mbar. Green: CO as sole carbon and energy source with pCO,in ¼ 100 mbar (Symbols: cX,max ¼ maximum dry cell mass concentration, μ ¼ maximum cell-specific growth rate, YX,C ¼ Biomass yield on C-mol basis, cAcetate,max ¼ maximum acetate concentration, qAcetate ¼ maximum cell specific acetate formation rate, YAcetate,C ¼ acetate yield on C-mol basis). Data from Mayer and Weuster-Botz [13]

produce acetone, isopropanol, 3-hydroxypropionate, mevalonate, isoprene, farnesene, butanoic acid butyl ester, methylethylketone, and isobutanol [11]. The autotrophic growth of acetogenic microorganisms with CO2 or CO as carbon source is strongly energy-limited, since no ATP can be obtained via the WoodLjungdahl pathway even with acetate as metabolic end product. Acetogenic microorganisms therefore depend on providing ATP chemiosmotically. The ATP yield for acetogenic bacteria varies widely under autotrophic conditions with – 4.5 mol ATP (mol product)
1 to +4.4 mol ATP (mol product)
1 depending on the final product, metabolic pathway, and electron donor (H2 or CO) [12]. Figure 1 shows how growth rate, biomass yield, acetate formation rate, and acetate yield vary in batch processes as function of the electron donor on the example of Clostridium aceticum. Products with ATP demand (negative ATP yields) can therefore only be produced autotrophically if other products with ATP gain are simultaneously produced by the cells. Nevertheless, energy-limited growth and low biomass yields under autotrophic conditions can result in high cell-specific CO2/H2 or CO uptake and product formation rates if, as in acetogenic microorganisms, the ATP yield per product molecule formed is low. In such a case, high carbon throughput may be beneficial for providing sufficient ATP and reduction equivalents to build up microbial biomass. However, high cell-specific conversion rates often cannot be achieved in gas fermentations due to the low solubilities of the gaseous substrates H2 and CO in water. For example, the water solubility of pure hydrogen (H2) under standard

Process Engineering Aspects for the Microbial Conversion of C1 Gases

37

conditions (1 bar, 20 C) is 65% lower on molar basis compared to pure oxygen (O2). For pure carbon monoxide gas (CO), the water solubility is reduced by 75% on molar basis compared to O2. Compared to typical aerobic heterotrophic bioprocesses with air supply, the maximum microbial cell concentrations in the bioreactor without cell retention are therefore reduced by a factor of 10–100 in gas fermentations with acetogenic bacteria, resulting in about 0.5–2 g L
1 dry cell mass, which basically leads to comparatively low volumetric productivities (space-time yields).

2.2

Gas–Liquid Mass Transfer at Low Volumetric Power Input

Mass transfer rates between gas and liquid phases are very often limiting in bioreactors due to the poor water solubilities of the gaseous substrates H2 and CO. CO2 limitations usually do not occur because CO2 solubility in water is comparatively high due to the carbonic acid equilibrium, especially at a pH in the neutral range (6.0 < pH < 8.0), which many acetogenic microorganisms require for growth. H2 and CO transfer into the water phase can generally be improved in bioreactors with dispersed gas phase by increasing volumetric power input and/or by increasing the partial pressures of these gas components in the gas phase. However, increasing the volumetric power input, for example by increasing the stirrer speed in stirred tank bioreactors, quickly reaches economic limits, especially when low value-added products such as short-chain organic acids or alcohols are produced. Increasing the partial pressures of H2 and CO in the dispersed gas phase is therefore the method of choice, which is also applied industrially, using bioreactors with a liquid height of 30 m and above. The hydrostatic pressure of the water column above the gas distributor is used to increase the partial pressures in the dispersed gas phase at the bottom of the bioreactor. Therefore, bubble column or gas-lift reactors are preferred industrially for gas fermentations with suspended acetogenic microorganisms. Bubble column reactors represent simple mass transfer and reaction apparatus in which a gas phase is brought into contact with a liquid phase. These reactors are characterized by their simple design and, in contrast to the stirred tank reactor, by the complete absence of mechanically moving internals. On an industrial scale, the power input of bubble column reactors is achieved solely by the isothermal expansion of the gas phase dispersed at the bottom of the reactor. As a result, the power input increases linearly with increasing gas flow rate and nonlinearly with the height of the water column above the gas distributor [8]. Increasing partial pressures and increasing power input by an increase of the water column thus provide, in principle, improved mass transfer between the gas and liquid phases (in the lower part) of bubble column or gas-lift reactors.

38

2.3

D. Weuster-Botz

Autotrophic Growth Kinetics

However, improving the gas–liquid mass transfer needs acetogens able to survive at elevated H2 or CO partial pressures, which can reach several bar at the bottom of industrial bubble column reactors depending on the gas mixture supplied. Unfortunately, no or less quantitative data has been published so far on substrate inhibition kinetics of acetogenic microorganisms [14–17]. The kinetics of acetogenic microorganisms as a function of gaseous substrates can differ greatly (Fig. 2). For example, the acetogenic bacterium Clostridium aceticum shows strong CO excess inhibition with a very low inhibition constant of KI ¼ 87 mbar CO and a growth optimum at 5.4 mbar CO [18], whereas Clostridium carboxidivorans has a much higher inhibition constant of KI ¼ 1,259 mbar CO with a growth optimum at 565 mbar CO, which is two orders of magnitude higher than Clostridium aceticum [19]. The inhibition constant indicates when the half-maximal growth rate is reached with increasing substrate concentration. The partial pressures of CO, H2, and CO2 of the gas bubbles rising in the bubble column reactor change not only due to the decreasing hydrostatic pressure of the remaining water column, but also due to the locally different consumption or formation of gas components by microbial metabolic processes. Many acetogenic microorganisms are unable to simultaneously utilize H2 and CO2, especially at higher CO partial pressures. Furthermore, some acetogenic bacteria show a typical two-phase process behavior: If the pH in the aqueous phase is high enough (usually > pH 6), the anaerobic

0.25

(5.4 mbar)

Clostridium carboxidivorans

Growth rate, h-1

0.20

(565 mbar)

0.15

0.10

0.05

Clostridium aceticum

0.00 0

200

400

600

800

1000

CO pressure, mbar

Fig. 2 CO inhibition kinetics of Clostridium aceticum (T ¼ 30 C, pH 8.0) [18] and Clostridium carboxidivorans (T ¼ 37 C, pH ¼ 6.0) [19] as function of the mean CO partial pressures measured in autotrophic batch processes with continuous gas supply at low dry cell mass concentrations (P/V ¼ 6.8–15.1 W L
1, FGas ¼ 0.083 vvm). The optimum mean CO partial pressures are indicated (5.4 mbar CO for Clostridium aceticum and 565 mbar CO for Clostridium carboxidivorans

Process Engineering Aspects for the Microbial Conversion of C1 Gases

39

bacteria can grow and organic acids, preferably acetic acid and, with significantly lower concentrations, butyric acid or even hexanoic acid, are formed initially. In this “acetogenic phase,” the pH in the aqueous solution decreases due to the formation of the acids. As a consequence, the growth of the acetogenic bacteria continuously decreases until a pH minimum is reached, usually below pH 5.0. At this pH, alcohol formation starts, whereby not only the formation of alcohols from the gaseous substrates takes place, but also, in particular, the previously formed organic acids are reduced to the corresponding alcohols. In this “solventogenic phase,” growth no longer occurs and, due to the microbial consumption of the acids, the pH in the aqueous medium rises again.

2.4

Autotrophic Gas Fermentation in Bubble Column Reactors

Axial gradients are therefore unavoidable in bubble column reactors on industrial scale, both in the gas phase with varying partial pressures of the gas components and in the liquid phase with varying concentrations of the solved gas components, product concentrations, and pH as function of the reactor height. One way of reducing strongly pronounced gradients in bubble column reactors is to use gas-lift reactors, combining gas bubbles and liquid flow upwards through the riser, the gassed part of the reactor, and liquid circulation through the downcomer, which can be realized as external or internal loop. This results in a defined circulation of the liquid phase in the gravity field due to the average density differences between the non-gassed and gassed parts of the gas-lift reactor. The more homogeneous mixing in gas-lift reactors can be an advantage compared to simple bubble column reactors. However, the mean hydraulic residence time in the downcomer should not be too high, since there is no gas supply here and substrate limitations may occur within a very short time in the microbial suspension (e.g., within a few seconds depending on cell density) due to the low H2 or CO solubility in the water phase, as a result of which the acetogenic bacteria in these reactor zones show less or hardly any metabolic activity. While mass transfer and mixing in bubble column or gas-lift reactors can be well simulated by known modeling approaches, there are so far hardly any quantitative data on growth, substrate uptake, and product formation kinetics of acetogenic microorganisms as a function of substrate (H2, CO) and product (acetic acid, ethanol, ...) concentrations, as well as pH. One approach to compensate for missing kinetic information could be the use of genome-wide stoichiometric models of acetogenic bacteria. Combined with simple uptake kinetics for CO and H2, the stoichiometric model can be used to estimate growth and product formation rates via metabolic flux analysis applying appropriate assumptions [20, 21]. Alternatively, biothermodynamics and mass transfer can be combined to a hybrid large-scale bubble column model for simulating microbial syngas fermentations [22].

40

D. Weuster-Botz

These are always estimates that have to be verified experimentally. Stirred tank reactors that are ideally mixed and as completely controlled as possible are particularly suitable on a laboratory scale for kinetic studies of the metabolic performance of acetogenic microorganisms under autotrophic conditions. Controlling the gas supply with individually adjustable partial pressures and volume flows of H2, CO, and CO2 and on-line waste gas analysis to determine the consumption rates of the individual gas components play an important role in addition to the control of the standard reaction conditions like pH, temperature, and power input, respectively. Gas fermentations at controlled conditions provide basic information on the metabolic performance of acetogenic microorganisms without gas–liquid mass transfer limitations, since high volumetric power inputs are possible with lab-scale stirred tank reactors (see, for example, [13, 23–26]). Growth and product formation kinetics can thus be identified as a function of dissolved gas concentrations or gas partial pressures at low cell concentrations [18, 19]. For the identification of growth kinetics, steady-state investigations in a continuously operated stirred tank reactor (chemostat) are in principle more suitable, even for gas fermentations, since arbitrary growth rates can be specified by setting different hydraulic residence times (see, for example, [27]). However, the low growth rates of acetogenic microorganisms under autotrophic conditions represent a major obstacle, especially under unfavorable reaction conditions (pH, temperature, ...), since, on the one hand, very long experimental times are required until steady-state conditions are reached, and, on the other hand, achieving stable operating states in the chemostat is only possible to a very limited extent in the case of substrate inhibition. In 2018, the first industrial gas fermentation plant started operation for ethanol production from steel mill waste gases with a capacity of 46,000 metric tons per year in Caofeidian, China [2]. The gas fermentation process developed by LanzaTech, a biotech company headquartered in Chicago, Illinois/USA, uses Clostridium autoethanogenum in bubble column reactors. In the pilot plant in Shougang (China), LanzaTech was able to convert 70–75% of the CO contained in the steel mill off-gas to ethanol [28]. The first industrial gas fermentation plant in Europe is currently under construction at the world’s largest steel producer ArcelorMittal in Ghent (Belgium). The plant is designed to handle 50,000 Nm3 h
1 of steel mill waste gas and is expected to have an annual capacity of 60,000 m3 of ethanol [29]. Aspects on future industrial applications of gas fermentations with acetogenic microorganisms considering existing industrial infrastructure and market requirements are summarized in Takors et al. [30].

2.5

Continuous Gas Fermentation Processes

Continuous processes are essential to achieve sufficient high volumetric productivities (space-time yields) in gas fermentations, which can be economically significant especially in the production of mass products with low added value such as organic acids and alcohols. In the case of suspended microorganisms, cell retention in the

Process Engineering Aspects for the Microbial Conversion of C1 Gases

41

continuously operated bioreactor can be achieved by submerged microfiltration membranes, as already used on a large scale in biological wastewater treatment [31]. Using the example of gas fermentation (CO2/H2) with Acetobacterium woodii, it was indeed possible to achieve increased cell densities and thus also significantly increased volumetric acetate productivities of more than 147 g L
1 d
1 in the continuously operated stirred tank reactor with submerged microfiltration membranes compared with 19 g L
1 d
1 in the batch process or in the continuous process without cell retention. However, the product concentrations were significantly reduced from 59 g L
1 acetate in the batch process to 17.6–23.5 g L
1 acetate due to the continuous withdrawal of the products solved in the aqueous phase in the continuous process [32]. In two-phase gas fermentation processes, higher volumetric productivities can also be achieved than in the batch process by connecting two continuously operated bioreactors in series. For this purpose, growth and acid formation take place in the first reactor at a higher pH (“acetogenic phase”), while in the second reactor the acetogenic microorganisms originating from the first reactor reduce the organic acids from the first reactor to the corresponding alcohols at a now lower pH (“solventogenic phase”). Furthermore, they form additional alcohols from the gaseous substrates, as shown in the example of continuous ethanol production with Clostridium ljungdahlii [33, 34]. In the continuous conversion of CO to produce alcohols (ethanol, 1-butanol, 1-hexanol) with Clostridium carboxidivorans in a stirred tank cascade operated at pH 6 in the first reactor and pH 5 in the second reactor (Fig. 3), higher space-time yields were obtained as expected compared with the batch process (increase by a factor of 6 for the alcohols ethanol, 1-butanol, and 1-hexanol). Surprisingly, however, the product concentrations in the continuous cascade process were also increased by a factor of 3 on average [33]. This is due to the unusual property of Clostridium carboxidivorans to form about as much ethanol and 1-butanol as acetate and butyrate already in the “acetogenic phase” in the first reactor during growth. Trickle-bed biofilm reactors with immobilized acetogenic bacteria represent an interesting alternative for continuous gas fermentations. In the classical design, a carrier material is filled in a cylindrical vessel, the (inner) surface of which is covered with a biofilm. The liquid phase is evenly distributed on the surface of the filling and flows over the carrier material with the immobilized microorganisms as a thin trickling film. At the bottom of the trickle-bed biofilm reactor, the liquid phase is collected and discharged (Fig. 4). In contrast to bubble column reactors, the gas phase is not dispersed in the liquid phase, but represents the continuous phase in trickle-bed biofilm reactors. The pressure drop of the gas phase transported either cocurrently or countercurrently through the trickle-bed biofilm reactor is thus negligible, and the power input of a trickle-bed biofilm reactor is determined solely by the transport of the liquid phase to the head of the reactor. Despite the low volumetric power input, the mass transfer rates of CO and H2 from the gas phase to the liquid film and to the biofilm surface are usually sufficiently high due to the thin liquid film. Due to the low operating costs at high mass transfer rates, trickle-bed biofilm reactors and special designs, such as the so-called submerged trickle-bed reactors

42

D. Weuster-Botz

Biomass, g

30

20

20

10

10 2

4

6

8

10

0

2

4

6

8

0 10 150 0

100

100

50

50

0

0

-50

C4-Products, g

20 0

Cascade

40

30

0 150 0

C2 Products, g

Reactor 2 (pH 5)

Reactor 1 (pH 6)

40

2

4

6

8

2

4

6

8

10

0

2

4

6

8

10

-50 0

2

4

6

8

10

2

4

6

8

10

0

2

4

6

8

10

0

2

4

6

8

10

15 10 5 0 0

Process time, d

Process time, d

Process time, d

Fig. 3 Cumulative production of acids and alcohols from CO in a continuously operated stirredtank reactor cascade with C. carboxidivorans with a dilution rate of 0.12 h
1 in the first reactor (V ¼ 1 L) and 0.08 h
1 in the second reactor (V ¼ 1.5 L), pH 6.0 in the first reactor and pH 5.0 in the second reactor, continuous gassing with CO ¼ 600 mbar, CO2 ¼ 400 mbar, FGas ¼ 0.083 vvm, T ¼ 37 C. Biomass and alcohols are shown in full circles, acids are shown in open circles. Standard deviations result from two independent experiments performed in the bioreactor cascade. The batch phase is shaded in gray. Data from Doll et al. [33]

(horizontally rotating fixed-bed reactors that are partially “submerged” in the liquid phase) are widely used on a technical scale in biological wastewater treatment. A limiting factor is the diffusive mass transport in the biofilm due to the low gas solubility in water, so that theoretically, depending on cell-specific gas uptake rates and gas partial pressures, only a small biofilm thickness of a few 10 μm can be supplied with the gaseous substrates. Very few studies have investigated the use of trickle-bed biofilm reactors in laboratory-scale gas fermentations [36–41]. Biofilm formation seems to be easily achievable with most acetogenic microorganisms, although no systematic studies on this exist to date. High CO conversions of more than 90% were observed when Clostridium ragsdalei was used in a trickle-bed biofilm reactor with simple, non-porous glass beads of 6 mm in diameter [38]. Hydrogen conversion of more than 80% was obtained when Clostridium carboxidivorans was used in a submerged trickle-bed biofilm reactor with non-porous HDPE support material with a volume-specific surface area of 500 m2 m
3. Compared to a gas fermentation in a continuously operated stirred tank reactor

Process Engineering Aspects for the Microbial Conversion of C1 Gases

43

Gas-lift reactor

Packed-bed biofilm reactor

Trickle-bed biofilm reactor Liquid phase

Gas phase

Fig. 4 Reversible retrofitting of a standard steam sterilizable stirred-tank bioreactor for operation as a gas-lift reactor, a packed-bed biofilm reactor and a trickle-bed biofilm reactor with the liquid distributor shown in more detail. Left: computer-aided design (CAD) drawings; right: photos of the bioreactors with acrylic glass tube as reactor wall for visualization. The carrier material is made of porous sintered glass in form of hollow cylinders (inner diameter 0.9 cm, outer diameter 1.4 cm, height 1.4 cm) and a porosity of up to 70% [35, 36]

under comparable operating conditions, the volumetric ethanol productivity was increased by a factor of 3.3 [40]. Axial gradients of pH and product concentrations in the aqueous trickling film are unavoidable as a function of trickle-bed biofilm reactor height on an industrial scale (10–15 m). Axial gradients could be reduced by recirculating the aqueous phase, but this would result in an increase in volumetric power input. Compared to bubble column or gas-lift reactors, the liquid volume (reaction or working volume) in trickle-bed biofilm reactors is reduced to about 20% of the total volume due to the volume of the inert support material and the gas volume, compared to about 80% of the total volume in bubble column reactors with a gas content of 20%. The reduced working volume of trickle-bed biofilm reactors is partially balanced by improved mass transfer at high gas conversion and by higher biocatalyst concentrations in the working volume due to immobilization. In addition to solving fundamental questions about scale-up, a technical application of trickle-bed biofilm reactors in gas fermentation will only become possible if stable operation over long process times is made possible, i.e., if a stable biofilm can be formed, especially in the future use of

44

D. Weuster-Botz

recombinant acetogenic microorganisms for improved production of natural, but also non-natural products. To date, no studies have been published on this subject.

2.6

Enlarging the Product Spectrum of Gas Fermentation by Co-cultivation

The value of the fermentation product increases with the number of carbon atoms, e.g. hexanoate has a two times higher market value than that of ethanol [42]. In addition to the higher product value, the recovery of middle chain fatty acids (>C6) is less costly due to their lower solubility in water in contrast to ethanol and shortchain fatty acids [43, 44]. But the production rates and concentrations of butyrate/ butanol or hexanoate/hexanol are low in gas fermentations with acetogens [45]. Microbial formation of middle chain carboxylic acids can be achieved from acetate and ethanol by chain elongation via the reversed β-oxidase pathway and has attracted interest in the recent years [46–49]. Clostridium kluyveri is an organism converting a 1:2 mixture of acetate and ethanol to butyrate, hexanoate, hydrogen gas (H2) and small amounts of octanoate [50]. Carbon dioxide is necessary for growth of Clostridium kluyveri [51]. C14labeling studies showed that more than 25% of the cellular carbon comes from carbon dioxide [52] and the availability of CO2 was reported to be necessary for chain elongation and resulted in highest hexanoate concentration of 21.4 g L
1 in a batch process applying a stirred tank bioreactor with controlled pH [53, 54]. Chain elongation of a syngas fermentation effluent with pure cultures of Clostridium kluyveri was studied in continuously operated bioreactors. Carbon conversion efficiencies of 90% were possible and hexanoate was produced with a space-time yield of 4.6 g L
1 d
1. Even octanoate was produced, but concentrations were below 0.3 g L
1 [48]. Instead of operating two bioreactors in series, co-cultivation of the acetogen Clostridium ljungdahlii with the chain elongating Clostridium kluyveri in a continuously operated bioreactor system gassed with 60% CO, 35% H2, and 5% CO2, respectively, resulted in a hexanoate production rate of 4.2 g L
1 d
1. Two-third of the ethanol produced by the acetogen was converted by Clostridium kluyveri into hexanoate, which was reduced again by Clostridium ljungdahlii to 1-hexanol with a volumetric productivity of 3.2 g L
1 d
1 [55]. The correct pH was found to be an important parameter for this co-culture to function (pH 6). This can be expected, because the proportion of ethanol to acetate formed by C. ljungdahlii is pH dependent [56] and chain elongation as well as growth of C. kluyveri is not possible at pH 5.5 or less [57, 58]. The partial pressure of carbon monoxide was shown to be another critical factor for successful co-cultivation as the metabolism of a pure culture of C. kluyveri was already inhibited at 50 kPa CO in the gas phase [59], whereas in co-cultures with acetogens high CO partial pressures are necessary to avoid CO-limitations and reduced acetate/ethanol production rates.

Process Engineering Aspects for the Microbial Conversion of C1 Gases

45

Although co-cultivation processes with acetogens and chain elongating bacteria seem to be very promising approaches for the microbial production of middle chain carboxylic acids/alcohols from C1-gases, the small operating windows (e.g., pH, CO partial pressure and others) for optimal interaction of a synthetic co-culture needs further research for better understanding and simulation of the mutual microbial interactions in bioreactors on an industrial scale.

2.7

Requirements for Syngas Purification in Gas Fermentation

Industrial residual gas streams containing CO2 and CO, but also synthesis gases from the gasification of residual biomass or non-recyclable plastic waste, contain a whole range of gaseous impurities. In addition to the non-critical inert gas N2, the gasification of residual biomass can produce, for example, methane, various gaseous hydrocarbons, such as ethene and ethane, and various gaseous sulfur and nitrogen compounds (such as H2S, COS, SO2 or HCN, NH3, NOX) in addition to CO2, CO and H2 [60]. In general, it is found that the requirements for syngas purification in gas fermentation are significantly lower than in the classical Fischer-Tropsch process for the catalytic production of hydrocarbons (see, for example, [29, 61]). However, quantitative studies on inhibitory effects of individual impurity components of synthesis gases are few. For Clostridium ljungdahlii, no effect on growth rate was observed at H2S concentrations of up to 5.2% (v/v) in the syngas [62]. In Clostridium ragsdalei, the effect of NH3 on metabolism was investigated and, as a result, it was postulated that medium costs could be saved if the nitrogen supply of acetogenic microorganisms could be at least partially covered by syngas contamination NH3 [63]. Furthermore, it has already been shown that increased NO concentrations in the gas phase inhibit the hydrogenase of Clostridium carboxidivorans, thereby inhibiting growth and eventually affecting the product spectrum. Complete inhibition of the hydrogenase occurs at 160 ppm (0.288 μM) NO content in the gas phase [64]. The syngas contaminants NH3 and H2S had a positive effect on both autotrophic growth and alcohol formation (ethanol, 1-butanol, 1-hexanol) in batch processes with C. carboxidivorans, but NOx contaminants in syngas should be selectively removed from the syngas [65]. Even “low” HCN concentrations in syngas from biomass gasification were the reason of the failed start-up of a syngas fermentation plant for ethanol production at the former Ineos Bio Indian River Biorefinery in Vero Beach, Florida, USA [66]. In contrast, steel mill waste gases, such as those used at LanzaTech, are characterized by very low impurities of less than 1 ppm HCN [29]. Syngas purification prior to gas fermentation seems necessary, but details are unknown due to the lack of quantitative data on inhibitory effects of individual impurity components of synthesis gases on acetogenic microorganisms.

46

D. Weuster-Botz

3 Conversion of CO2 with Microalgae Photoautotrophic microalgae (prokaryotic cyanobacteria and eukaryotic protozoa) use water and sunlight as an energy source for biomass formation from CO2. Either the microalgal biomass produced photoautotrophically from CO2 is used directly as feed or for human nutrition or intracellular products such as antioxidants (e.g., ß-carotene), pigments (e.g., astaxanthin), and proteins (e.g., phycocyanin) are extracted from the algal biomass and used industrially as food additives [6]. Since some microalgae are able to accumulate lipids with up to 70% (w/w) of their dry cell weight as carbon reservoir, research has been conducted for many years on the use of microalgae for the production of biofuels (biodiesel, biokerosene) [67]. Suspension reactors are required to enable easy harvesting and further processing of microalgal biomass. Sunlight irradiation at the liquid surface is attenuated by absorption in the microalgal suspension. The resulting light attenuation with increasing distance from the liquid surface can usually be described using Lambert-Beer’s law (exponential decrease in light intensity with increasing liquid layer thickness and increasing biomass concentration) despite its inability to differentiate between light absorption, scattering, and reflection. The lower the liquid layer thickness of an algae suspension, the higher the biomass concentrations and thus higher intracellular product concentrations can be achieved. For this reason, photobioreactor facilities with huge surface areas and with low liquid layer thickness (ideally a few centimeters) are needed for mass production of microalgae. Basically, photobioprocesses with microalgae are therefore “two-dimensional processes.” Consequently, they are characterized by the biomass productivity per unit area (g dry cell mass m
2 d
1).

3.1

Open Photobioreactors for Microalgae Mass Production

In principle, a distinction is made between open and closed photobioreactors. Closed photobioreactors with a liquid layer thickness of a few centimeters (tubular reactors with power input via pumps or plate reactors with power input via isothermal gas expansion, i.e. gas-lift reactors) are less suitable for mass production of microalgae due to the low added value even when using very inexpensive materials, such as simple tear-resistant and UV-resistant plastic bags [6, 67]. Therefore, the industrial photoautotrophic production of microalgal products, such as ß-carotene, astaxanthin or phycocyanin, is carried out in open photobioreactors, mainly in the so-called raceway ponds. Raceway ponds, usually made of pond liner, have surfaces of up to 8,000 m2 and consist of meandering flat channels of a few meters width with a water depth of 10–30 cm with closed recirculation of the liquid phase via paddle wheels. Due to the relatively high water depth (layer thickness of the algal suspension), usually only algal dry mass concentrations of 0.3–1.0 g L
1 are achieved [68–70]. In production plants, individual raceway ponds are interconnected over areas of many 100 ha (scale-up by “numbering-up”). In close proximity to the paddle wheel for

Process Engineering Aspects for the Microbial Conversion of C1 Gases

47

Fig. 5 Photograph of an open thin-layer cascade photobioreactor made of pond liner with an illuminated surface area of 50 m2 and a liquid volume of 330 L operated at the TUM-AlgaeTec Center (Ottobrunn, Germany). Two 12.5 m  2 m channels with a decline of 1% were each connected with an open retention tank, a centrifugal pump, and an inlet module for continuous circulation of the microalgal suspension with a mean suspension layer thickness of ~0.6 cm. The open photobioreactor is placed in a greenhouse with application of a dynamic climate simulation technology (light, air temperature and humidity) integrating natural sunlight and multi-color LED arrays for a highly realistic reproduction of the sunlight spectrum to enable the physical indoor simulation of dynamic outdoor environmental conditions [74, 75]

homogenization of the algae suspension, there are devices for CO2 input, pH measurement, and supply of nutrient salts (nitrogen, phosphate, ...) and fresh water for compensation of evaporation. The liquid phase is usually in the transition zone from laminar to turbulent flow at a mechanical power input of 1 W m
3 or 0.25 W m
2 (electrical power ~ 0.5 W m
2). At suitable sites, the areal productivity reaches an annual average of 10–15 g dry cell mass m
2 d
1, corresponding to a photo-conversion efficiency of 1–3% (energy of absorbed light related to the combustion enthalpy of the algal biomass) [71, 72]. Thin-layer cascade photobioreactors, on the other hand, are installed on an inclined plane with slopes of about 1% and represent channels of a few meters width that are gravity-driven with a liquid layer thickness of 1 cm or less [73] (Fig. 5). Power input is provided by pumps that transport the algal suspension from the bottom of the collecting tank to the top of the photobioreactor, where uniform distribution of the algal suspension across the entire channel width occurs. CO2 input, pH control, and supply of nutrient salts (nitrogen, phosphate, ...) and fresh water for compensation of evaporation take place in the collecting tank. Due to the low liquid layer thickness of less than 1 cm, algal dry mass concentrations of up to 50 g L
1 can be achieved [74]. It makes sense to select operating points with high

48

D. Weuster-Botz

areal productivity (20–30 g dry cell mass m
2 d
1), where lower cell densities of around 20 g L
1 algal dry mass are achieved. This corresponds to a photoconversion efficiency of 4–5% (energy of absorbed light related to the combustion enthalpy of the algal biomass). The hydraulic power input is twice as high in thinlayer cascade photobioreactors as in raceway ponds, exceeding 0.5 W m
2. However, due to the higher areal productivity in thin-layer cascade photobioreactors, the hydraulic power input in terms of algal biomass produced is comparable for both open photobioreactor systems. Fluid flow is usually turbulent or transitional in thinlayer cascade photobioreactors assuring sufficient vertical mixing of the microalgae [76]. Open photobioreactors are uncontrolled except for pH control via CO2 input, supply of nutrient salts, and setting of defined mean hydraulic residence times during continuous operation. This means that growth and product formation kinetics of the microalgae as function of dynamically changing light intensities and temperatures in suspension (day and night cycles, as well as seasons) must be known to enable reliable scale-up. Furthermore, due to the continuous water evaporation of open photobioreactors, only (hyper)saline microalgae are suitable as production organisms especially if brackish water or sea water is used and recycling of the process water is applied to reduce nutrient costs [77]. The microalgae should be able to grow at highly fluctuating temperatures and at high pH (pH > 8.5) to minimize CO2 losses to the atmosphere and reduce the risk of contamination. With sufficiently high pH and suitable CO2 input devices, CO2 yields of nearly 90% can then be achieved if the alkalinity of the aqueous phase is controlled [78]. The kinetic data for scale-up can be determined in closed, fully controllable flatplate gas-lift photobioreactors at known liquid layer thickness in the lab [79–82]. In these gas-lift reactors, power input is achieved by isothermal gas expansion. First, the light absorption in the suspension must be determined for each microalgae strain and described by a suitable physical model. Assuming turbulent liquid flow, an average light intensity (photon flux density) can thus be determined as a function of the light intensity at the surface of the suspension, the liquid layer thickness, and the algal concentration. The estimation of the growth rate of microalgae is usually carried out performing batch processes at different constant incident light intensities. In the exponential growth phase, the mean photon flux densities, which depend on cell density and liquid layer thickness, are additionally averaged as a function of process time resulting in the mean integral photon flux density. The measured exponential growth rates are plotted as a function of these mean integral photon flux densities in order to identify the model parameters of suitable kinetic approaches to describe light saturation and light inhibition kinetics. Similarly, model parameters describing the effect of temperature on microalgal growth can be identified. Growth kinetics identified with flat-plate gas-lift reactors can also be used with completely different photobioreactors, if a defined liquid layer thickness can be assured, as was shown on the example of thin-layer cascade photobioreactors [83]. However, in order to be able to determine realistic growth kinetics, a fixed day and night cycle must be applied in the investigations, otherwise the growth rates of microalgae are significantly underestimated. In addition, increasing the salinity as function of

Process Engineering Aspects for the Microbial Conversion of C1 Gases

49

process time is necessary in the closed lab-scale photobioreactors to simulate the evaporation of open photobioreactors [84]. Coupling mass balances with the kinetics of microalgal growth and product formation enables the simulation of batch and continuous phototrophic production processes with microalgae under dynamic diurnal variations of incident light and temperature in thin-layer cascade photobioreactors [75].

3.2

Requirements for Combustion Gas Purification for Photoautotrophic Processes

Combustion gases from fossil power plants are commonly used as a source of CO2 for photoautotrophic production plants. Combustion exhaust gases can contain over 100 other substances in addition to CO2. Essentially, these are nitrogen oxides (NOX), sulfur oxides (SOX), carbon monoxide (CO), and incompletely oxidized hydrocarbons, in addition to N2, O2, and H2O [85]. The nitrogen oxides dissolved in water during the absorption of combustion gases are ultimately present as nitrite or nitrate in the algae suspension and can thus be directly utilized as nitrogen sources, so that improved growth of microalgae is often observed when NOX-containing combustion gases are used [86]. Sulfur oxides dissolve well in water. The sulfates ultimately formed in water in the presence of O2 can be used by microalgae as a source of sulfur for growth. Microalgae show different sensitivities to bisulfite (SO32
), whose concentration in water is pH-dependent due to the dissociation equilibrium. This means that high SO2 concentrations in combustion gases (>100 ppm) can lead to growth inhibition at pH < 6 [67]. Since pH > 8.5 is requested in open photobioreactors to minimize CO2 losses and contamination, growth inhibition by SO2 in combustion gases usually does not occur. In principle, however, the pH reduction due to nitrogen oxides and sulfur oxides dissolved in water during absorption must be taken into account. CO is poorly soluble in water. The CO fraction dissolved in water, as well as hydrocarbons originating from combustion exhaust gases, could in principle be utilized by some microalgae as a carbon source in addition to CO2 [85, 87, 88]. Therefore, purification of combustion exhaust gases to recycle CO2 with microalgae is not necessary in most cases from a biological point of view. If CO2 from combustion flue gases has to be transported over longer distances via pipelines to a microalgae production plant, flue gas treatment for CO2 enrichment and removal of sulfur oxides and nitrogen oxides is necessary for economic and, in many countries, emission control reasons, as well as for corrosion protection. Most economic studies have been published to date on the production of lipids as a feedstock for biofuels. These can be summarized as follows: The lowest production costs are achieved with open photobioreactors applying combustion exhaust gases without gas purification and the largest possible production plants (economy of scale) with recycling of nutrients (nitrogen, phosphate) using seawater. In addition,

50

D. Weuster-Botz

light availability should be high, so that particularly non-agricultural areas with brackish or seawater availability in dry, subtropical climates are preferred. In many studies, the use of nitrogen and phosphate rich wastewater as nutrient sources for microalgae is considered to be particularly economically viable [70, 72].

4 Conclusions What do gas fermentations with acetogenic microorganisms and photobioprocesses with microalgae have in common and how do they differ? Or put another way, which bulk products can be produced more efficiently with acetogenic microorganisms from CO2 and which with microalgae? The following comparison (see Table 1) is based on bubble column reactors, on the one hand, and open photobioreactors (raceway ponds/thin-layer cascade reactors), on the other, and is limited to CO2 utilization. Gas fermentations and photoautotrophic processes are characterized by comparable biocatalyst concentrations due to substrate limitations. In gas fermentations these are the low hydrogen gas solubilities in the liquid phase and in photoautotrophic processes the limited light transfer into microalgal suspensions. In gas fermentations, substrate limitation can be counteracted by improving gas–liquid mass Table 1 Comparison of gas fermentations with acetogenic microorganisms and photobioprocesses with microalgae (RP raceway pond, TLC thin-layer cascade)

Electron source Biocatalyst concentration Reactor height Space requirement Power input (P/V) Products Reaction conditions Maximum growth rate Product concentrations Space-time yield CO2 uptake rate Gas purification requirement (CO2) Economy of scale

Gas fermentations (acetogens) H2 (ideally renewable) 1–3 g dry mass L
1 (H2 input is limiting) Up to 30 m (due to low H2 solubility) Low 300 W m
3 C2–C6 (extracellular) Fully controlled

Yes

Photobioprocesses (microalgae) Sunlight (suitable climate zones) 1 (RP) – 25 (TLC) g dry mass L
1 (light input is limiting) Up to 0.3 m (due to light attenuation) Very high 1 W m
3 >C40 (intracellular) Not controlled, with the exception of pH ~1 d
1 30 mM formate, the evolved strain could grow up to 100 mM formate, and in the presence of bicarbonate (probably due to pH buffering) even up to 300 mM. Growth on higher formate concentrations is beneficial for bioprocesses. However, given the relatively high toxicity of formate some kind of fed-batch or continuous process will be required to achieve sufficiently high biomass concentrations and product titers on formate. The only other study so far that showed full growth via an engineered rGlyP was done in C. necator [28]. Here, the expression levels of the C1-module and GCS-module were first optimized in a glycine+C1-auxotroph. Then, growth of C. necator on glycine was tested, which after short-term evolution was established, demonstrating a good potential for C. necator to perform glycine assimilation. However, transcriptomics and knockouts of this glycine growth phenotype revealed C. necator employed wasteful glycine oxidation to glyoxylate and glyoxylate

Engineering the Reductive Glycine Pathway: A Promising Synthetic. . .

331

assimilation via the glycerate pathway. Still, it was attempted to integrate the C1 + GCS modules with this native glycine assimilation capacity. This was tested in a strain that was knocked out for the CBB cycle, which is the native formatotrophic pathway of C. necator, by removing both copies of the Rubisco genes. This already led to formatotrophic growth, but only at low growth rate (~60 h doubling time) and low yield, likely caused by the native inefficient glycine assimilation via the glycine oxidase route. After overexpression of native enzymes for the more efficient serine deaminase route, as well as short-term ALE, this led to relatively fast formatotrophic growth (~12 h doubling time) and reasonably high yield (2.6 gCDW/mol formate), albeit still slightly lower than yield of formatotrophic growth via the CBB cycle (2.9 gCDW/mol formate (Table 2). One of the interesting mutations found after ALE in C. necator led to the inactivation of the CBB cycle regulator gene (CbbR), which probably lowered the expression of remaining CBB cycle enzymes, which were not required in this strain, thus reducing burden. In addition, a mutation was found in the strong constitutive promoter integrated in the genome to regulate the GCS, which was shown to further upregulate its genomic expression. Interestingly, in an earlier C. necator rGlyP strain harboring a plasmid-based GCS expression a mutation lowering GCS expression was found after short-term ALE. This underlines that fine-tuning GCS expression to a sweet spot between limiting flux and expression burden is important for efficient rGlyP operation and can be achieved by ALE.

3.3

Other Selection Schemes and Considerations for Future Engineering

For the engineering of rGlyP with an alternative module for glycine assimilation rather than the canonical serine deaminase variant, other selection schemes can be used. For several of the glycine conversion variant modules, P-glycerate will be a key intermediate. A possible selection scheme for the selection of such pathways could be a strain in which glycolysis is disrupted below P-glycerate by an enolase knockout (Δeno) (Fig. 4h). In this case modules of a rGlyP variant making P-glycerate could supply upper metabolism (~28% of biomass) and the rest of metabolism could, for example, be fed with pyruvate as helper substrate. This selection scheme is very similar to a scheme used to evolve the full functionality of an engineered CBB cycle in E. coli [99]. More selection schemes can be envisioned for different variants and in different hosts. Suitable selection schemes will also be partly host-dependent. To be able to create specific auxotrophs in a host several conditions need to be met. For example, all genes towards the auxotroph metabolite need to be known, the knockout of the genes should be technically and biologically possible and the host should be able to transport the auxotrophic metabolites inside the cell from the medium. This may lead to the selection of specific selections schemes when engineering the rGlyP in a different host. In

332

N. J. Claassens et al.

general, based on the experience of engineering the rGlyP in a few host, one can start with already somewhat “high selection” strains covering multiple modules (e.g., C1-module and GCS module in the glycine +C1 auxotroph) at once. If this directly works one will be able to proceed faster. In case no growth can be achieved, one needs to fall back on lower level selection strains and first optimize an individual module.

3.4

Pathway Module Confirmation by 13C-Labelling Studies

Even though the growth of a selection strain is a strong indication for the functionality of the rGlyP, the confirmation of module or pathway functionality can be obtained more conclusively by 13C-labelling studies. In such a study, labelled substrates, in the case of rGlyP typically 13C-formate or 13CO2, are fed to the strain operating a module or the full pathway. Then, a convenient way to prove functionality is the analysis of steady-state labelling patterns in the proteinogenic amino acids. Proteinogenic amino acids are abundant and can be easily hydrolyzed from culture samples and analyzed by LC-MS. For more details on related protocols, we refer to Wenk et al. [94]. Distinct labelling patterns can determine if, for example, methylene-THF (via the labelling of the C1-derived carbon in methionine and histidine), glycine, serine or pyruvate (via labelling of derived amino acids such as alanine and valine) are indeed synthesized as expected by the rGlyP modules (see also Table 2). Full operation of the rGlyP should lead to full labelling of all proteinogenic amino acids when grown on 13C-formate plus 13CO2. Growing strains with the full rGlyP on either 13C-formate or 13CO2 can be potentially used to prove the operation of certain variants of the pathway. In this way the use of the glycine reductase route could be demonstrated in the natural rGlyP auxotroph D. desulfuricans [30]. However, neither 13C-formate nor 13CO2 labelling can differentiate between the other glycine to biomass conversion pathways. In addition, labelling with either 13-formate or 13CO2 during growth via the full rGlyP can be used to determine how high the “wasteful” TCA cycle flux is. Labelling of proteinogenic amino acids derived from TCA cycle metabolites can be used to determine if fluxes to these metabolites primarily comes from anaplerosis or from full TCA cycling. This method was shown for E. coli and C. necator with an engineered rGlyP, for both it was found that the anaplerotic flux is higher than the “wasteful” TCA cycling [27, 28].

3.5

Non-growth Coupled Engineering Efforts

Some studies have expressed heterologous enzymes for the rGlyP without following a growth-coupled approach. In some of the E. coli studies on engineering the rGlyP, modules were tested without coupling them (fully) to growth [36, 100]. Even though

Engineering the Reductive Glycine Pathway: A Promising Synthetic. . .

333

this led to partially functional modules, it is harder to assess module functionality in these studies. And once the modules were confirmed without selection, it was realized that a serine auxotroph strain could be beneficial to force more flux to serine and even downstream to pyruvate [100]. In some other hosts, initial steps were taken to implement the rGlyP, such as the overexpression of the GCS enzymes in the anaerobic glycerol fermenter Clostridium pasteurianum, with the goal to create mixotrophic fermentation of glycerol and formate [69]. However, the overexpression of these genes was not coupled to growth via a selection scheme, and high expression burden of the GCS overexpression was observed. This suggests optimization of expression levels of the GCS genes is needed for which a growth-coupled selection could be helpful and allow for expression optimization via evolution, as also demonstrated in C. necator. In this study still some physiological impacts of the GCS were observed in the form of increased formate consumption in the presence of glucose, but this work could not yet prove high flux through rGlyP modules. In another study FTL was overexpressed to establish modules of the rGlyP as a pathway for formate assimilation to support carbon fixation by the CBB cycle in the cyanobacterium Synechocystis sp. PCC6803. In this non-growth coupled approach some physiological effects were observed that could potentially be linked to flux through the rGlyP [101]. However, labelling and transcriptomics revealed that the overexpression of FTL led to unanticipated changes in the glycine, serine, C1 and nitrogen metabolism. Also here, a growth-coupled approach could potentially further force flux through the modules towards establishing the rGlyP in this host in the future. A seemingly more successful establishment of the full rGlyP was achieved in the acetogen E. limosum [42]. Here, the GCS from the rGlyP-operating acetogen C. drakei was heterologously introduced on a plasmid. This was done without any modular or growth-coupling based approaches. Direct introduction of the GCS led to improved growth and acetate biosynthesis on H2/CO2. However, it must be noted that the rGlyP here did not lead to full autotrophic growth, as the pathway operates in concert with the native rAcP. Also, no labelling data on the engineered pathway activity were provided.

4 Comparing the rGlyP to Other C1-Pathways Apart from the rGlyP there are several other pathways for assimilation of C1-substrates and CO2 in nature, as well as hundreds of theoretical pathways that have been proposed and analyzed in silico [19, 23, 81, 103–105]. Only for a subset of these pathways, experimental studies have attempted to implement them in vitro or in vivo [52, 104, 106]. For even fewer pathways, full functional operation has been realized so far in engineered hosts. Apart from the rGlyP, successful implementation was recently shown for the RuMP cycle for synthetic methylotrophy in E. coli and for the CBB cycle for synthetic formatotrophy in E. coli and

334

N. J. Claassens et al.

methylotrophy in Pichia pastoris [88, 107]. In this section, we provide a comparison of the rGlyP versus other key pathways for the assimilation of formate, methanol and/or CO2 that are present in nature, have been engineered or have been proposed as promising pathways by in silico analyses.

4.1

Assessing the rGlyP for Formatotrophic Growth

Some natural and many more theoretical formate-assimilating pathways have been described in literature [23, 33] (Table 3). Briefly, these can be distinguished into two groups based on their formate utilization strategy. To the first cluster belong those pathways which oxidize formate to CO2, exploiting the extracted electrons as reducing power for CO2 fixation routes. Examples of such pathways are the naturally occurring CBB cycle, reductive tricarboxylic acid cycle, and synthetic CO2 fixations pathways such as the CETCH cycle and GED (6-phosphogluconate dehydrogenase (Gnd) – Entner–Doudoroff) cycle (Table 3). To the second group belong those metabolic routes which directly assimilate formate. The latter category includes the rGlyP and other major natural formate-assimilation pathways, the rAcP and Serine cycle, which all use FTL for the direct assimilation and activation of formate. A key parameter to compare different pathways is their theoretical yield, which is mostly determined by the ATP investments (and in some cases also by “wasteful” reducing equivalent investments) required to produce a certain metabolite (e.g., pyruvate) or biomass. Another key parameter to compare pathways is their kinetic performance, as the overall pathway kinetics determine potential growth rates, productivities, and enzyme burden. Due to limited, trustworthy estimates for the kinetic parameters that accurately reflect in vivo conditions, this performance is harder to quantify and compare. Still, several studies have compared different formate natural and synthetic assimilation pathways [19, 23, 33] including a recent detailed assessment taking into account thermodynamic driving force and kinetic parameters [81]. The general picture that emerges is that the rGlyP outcompetes most other formate-assimilation pathways – as well as CO2 fixation pathways – in terms of theoretical yields, due to the lower ATP consumption for pyruvate and biomass biosynthesis. The only pathways that can rival the rGlyP in terms of theoretical yield are the rAcP and rTCA, which have a lower ATP consumption than most variants of the rGlyP (Table 3). However, the rAcP and rTCA usually require anaerobic conditions due to oxygen-sensitive enzymes [108], and relatedly have ATP-limitations and a limited product spectrum. When comparing the rGlyP further to other oxygen-tolerant formate-pathways, the ATP efficiency is higher, leading to small yield advantages over most pathways [19, 81]. Also in terms of kinetics based on available parameter estimates, the rGlyP is suggested to be among the best performing pathways. Another advantage of the rGlyP is its architecture as a linear pathway. We note here that some studies refer to the rGlyP as a circular pathway, since the co-factor THF is used as an acceptor at the start of the pathway and released later in the GCS

RuMP (FBA-TAL variant) cycle Serine cycle (natural) Serine cycle (modified serine cycle) Serine cycle (homoserine cycle) Formolase pathway (dihydroxyacetone)

rTCA cycle

CBB cycle

rGlyP (glycine reductasea) rAcP

Pathway name (variant name) rGlyP (serine deaminase)

Partly engineered: formate [106] Theoretical: CO2, CO, methanol

Naturally present, (partly) engineered or only theoretically proposed per C1-substrate [refs.] Engineered: formate, methanol [27–29] Theoretical: CO2, CO Natural: CO2, formate [30] Theoretical: CO, methanol Natural: CO2, CO, formate, methanol Natural: CO2, CO, formate, methanol engineered: CO2, formate, methanol [107, 116] Natural: CO2 [117, 118] Theoretical: formate, CO methanol Natural: methanol Engineered: methanol [88] Theoretical: CO, formate Natural: formate, methanol Theoretical: CO2, CO Partially engineered: formate, methanol Theoretical: CO2, CO Partially engineered: methanol Theoretical: CO2, CO, formate ? +

100%). While the same thermodynamic and kinetic boundary rules govern both heterotrophic and autotrophic bioconversion processes, the combination of them enables compensation by complementary modules. Further, the choice of cellular hosts is broadened, enabling usage of heterotrophs with artificial C1 fixation as demonstrated by various examples of synthetic C1 fixation pathways which are integrated into heterotrophs with varying efficiencies – even to the extent of utilization of C1 compounds as sole carbon source [35–38]. Microbial conversion of C1 compounds to target products strongly differs in requirement for energy and reducing equivalents depending on the fixation pathway

354

Y. Hong and A.-P. Zeng

as well as on the oxidation state of employed C1 compounds [6, 8]. Thus, for the design of (mixotroph) carbon fixation pathways or reactions and for the prediction of its efficiency, the thermodynamic requirements need individual consideration. For instance, acetate biosynthesis by acetogens from CO alone generates excess ATP than from CO2 and H2 (Gibbs free energy of 175 kJ mol1 vs. 95 kJ mol1) [39]. Further, Redl et al. [40] estimated the reduction of Gibbs free energy (from 322.8 kJ mol1 to 269.7 kJ mol1) for acetone biosynthesis from CO by Moorella thermoacetica under non-standard conditions using realistic process temperature and concentration-related correction terms. Closely linked to the C1 fixation module, the product biosynthesis module and cellular growth constitute the counter parts of consumption of energy in form of ATP and reducing equivalents. In addition to the reaction-stoichiometry of ATP and NAD(P)H (and other reducing equivalents) demands, overcoming several energetic barriers for thermodynamically unfavorable reactions poses key challenges in carbon fixation pathways as discussed by Bar-Even et al. [41] and computationally analyzed by Löwe et al. [9]. If the energy supply of acetogens is regarded as “thermodynamic limit” [11], mixotrophy enables potential compensation by its co-consumption of organic substrate(s). Even for hosts which are not fully adapted toward energy-depriving synthetic carbon fixation pathways, co-supply of energy and reducing equivalents for the desired biosynthesis pathway via mixotrophy can result in compensation for energy and redox balances without shortage in other essential cellular processes. Especially for the energyintensive biomass generation, co-supply via secondary substrate makes the utilization of slow growing (synthetic) autotrophs for bioproduction possible. From a bioprocess engineering point of view, the formation of sufficient quantity of biomass and thus the achievement of a certain volumetric productivity of a target product (normally in the range of several grams per liter per hour) is essential for the industrial relevance of a bioprocess [42]. In this chapter, several mixotrophic examples studied in recent years are assessed to demonstrate the potentials and to identify challenges of using mixotrophy for biosynthesis (Fig. 1). Research needed are discussed for further development.

2 Acetogenic Mixotrophy Among microorganisms with native mixotrophic metabolism, acetogenic microorganisms are promising candidates for mixotroph bioprocess as proposed by Fast et al. [29]. Acetogens gained high emphasis in the past years for the utilization of gaseous C1 compounds (CO2 and CO) with H2 for biochemical production [27, 43]. Their native capability of autotrophic growth on syngas via reductive acetyl-CoA pathway (Wood-Ljungdahl pathway) enables direct recycling of CO2 to organic substances coupled to ATP conserving pathway through generation of H+/ Na+ gradient for ATPase by using electron transfer from Fdred to NAD [11, 44– 46]. In combination with an electrochemical system, electrons from electrodes can be incorporated as additional energy source for the synthesis of biochemicals from

Biosynthesis Based on One-Carbon Mixotrophy

355

Fig. 1 Major concepts and metabolic interplays of mixotrophy: (a) Acetogenic mixotrophy consists of fixation of C1 compounds from syngas via reductive acetyl-CoA pathway (rAP) and conversion of organic substrates to cellular intermediates. Hydrogen from syngas is utilized for generation of reducing equivalents ([H]) required for the reduction of C1 compounds. (b) In methylotrophic mixotrophy, carbons from methanol are oxidized to formaldehyde (HCHO), formic acid (HCOOH), and CO2 that can be assimilated over ribulose monophosphate (RuMP) cycle, xylulose monophosphate (XuMP) cycle, serine cycle or ribulose bisphosphate (RuBP) cycle respectively, in addition to conversion from organic substrates. Over the oxidation of C1 compounds, reducing equivalents are generated, which are used for the serine and RuBP cycles. In all the cases, compensation of energy (ATP) and reducing equivalents occurs between product and biomass biosynthesis modules, C1 fixation modules, and assimilation of organic substrates. Bold lines represent carbon fluxes and thin lines represent fluxes of reducing equivalents and ATP. Dashed lines depict multi-step conversion of intermediates

CO2 [47, 48]. Also, nitrate as an inexpensive electron acceptor was suggested for acetogens [49]. Matching the increasing interest on utilization of acetogens with ongoing commercialization [50, 51], tools for acetogens are being rapidly developed, including metabolic engineering tools for chromosomal editing, such as double-crossover homologous recombination [52], genome editing utilizing Cas9 or Cas12a [53, 54] or phage serine integrase [55], and genome-scale model for metabolism and gene expression [56, 57]. Genetic engineering, engineering tools, and system-level analysis for acetogens were recently summarized by Jin et al. [58]. Acetogenic mixotrophy constitutes combinatory use of heterotrophic catabolism and reductive acetyl-CoA pathway as C1 fixation module, in which carbons are supplied from CO2/CO and H2 delivers reducing power. Mixotrophy utilizing acetogens is not limited to accomplishing negative carbon emission by incorporation of C1 compounds. Instead, carbon yield exceeding calculated theoretical yield of heterotrophy can be archived by minimizing loss of carbons as CO2. In general, all

356

Y. Hong and A.-P. Zeng

Table 1 Selected examples of acetogenic mixotrophy Microorganism Clostridium ljungdahlii

Substrates Fructose, H2, CO2, (CO)

Product Acetone, 2,3-butanediol, 3-hydroxybutyrate, ethanol, acetate

Mevalonate, isoprene 17 different acetogens (incl. non-model strains)

Hexose glycerol, CO2, CO, H2

Acetate, ethanol, lactate, 1,3-propanediol, butyrate, butanol

Feature Deletion of secondary alcohol dehydrogenase (CLJU_c24860), continuous fermentation reaching high cell density, H2 supplementation for enhanced carbon utilization Heterogeneous expression of mevalonic acid and isoprene biosynthesis genes Syngas-enhanced mixotrophy for Blautia producta, Clostridium scatologenes, and Thermoanaerobacter kivui leading to near stoichiometric carbon conversion, engineering of C. ljungdahlii for glucose utilization

Source [60]

[64]

[65]

biosynthesis pathways requiring acetyl-CoA qualify for possible improvement by mixotrophy using acetogens. The fixation of evolved CO2 of glycolysis from the decarboxylation of pyruvate results in carbon-loss free conversion to acetyl-CoA, which provides additional carbon building blocks for the desired biochemical [59]. Further, the nature of diverse metabolic versatilities of acetogens remains a unique feature for mixotrophy. In addition to the native C1 fixation machinery, utilization of electron donors such as H2, hexose, alcohol, and organic acid [29, 60– 63] broadens the range of substrates for mixotrophy. As a primary example (s. Table 1), Jones et al. [60] demonstrated with an engineered Clostridium ljungdahlii strain (deletion of secondary alcohol dehydrogenase gene) the biosynthesis of acetone with a mass yield exceeding the theoretical yield under heterotrophic conditions (34% vs. 32%). In addition, continuous fermentation with high cell density was performed, which reached optical densities up to 60 (approx. 18 g L1) and overcame thus the primary limitation of low biomass concentration of autotrophs. Additional experiments verified uptake of exogenous hydrogen (at 20% in the headspace) for mixotrophic growth, which led to fixation of nearly all evolved CO2 and shift in the metabolite spectrum to reduced compounds: lower acetate and acetone concentrations and elevated ethanol titer showed incorporation of exogenous reducing equivalents. Maru et al. [65] screened 17 different acetogens for non-photosynthetic mixotrophy and engineered C. ljungdahlii for the utilization of non-native substrate glucose. Strains with heterogeneous expression of glucose phosphotransferase reached carbon yields of 115% and 125% with acetate as the main product. As an alternative strategy, glycerol mixotrophy was demonstrated with Clostridium scatologenes by the same authors: with glycerol as a more reduced feedstock in comparison with conventional sugar, co-supply of reducing equivalents enhanced the reduction of CO2. Highly improved carbon yield in comparison with

Biosynthesis Based on One-Carbon Mixotrophy

357

heterotrophic growth with N2 sparge (99.8% vs. 88.9%) was achieved with 1,3-propanediol, butyrate, and acetate production from consumed carbons. The overall demonstration of mixotrophy of non-model acetogens shows the broad availability and flexibility as candidates for acetogenic mixotrophy. Applicability for the biosynthesis via mixotrophy of non-native products, such as mevalonate and isoprene, was demonstrated by Diner et al. [64].

3 Methylotrophic Mixotrophy Research on synthetic methylotrophy has greatly expanded in the past years to target methanol as an exceptionally interesting substrate for C1 utilization due to its high energy content, high degree of reduction (6 vs. 4 of glucose), and decreasing price via synthesis from methane fraction of natural gases [66–68]. Thus, methylotrophic mixotrophy focuses on co-utilization of highly reduced C1 substrates, such as methanol, as carbon and reducing power supply. The limited number of genetic tools applicable for native methylotrophs has impeded extensive strain development. Thus, the primary research focus was shifted to synthetic methylotrophy using platform microorganisms [69–71]. The current progress in methylotrophy is well summarized in the reviews of Wang et al. [72] and Chen and Lan [73]. Since detailed analysis of the underlying principles and methods will exceed the scope of this work, the main aspect of mixotrophy for methylotrophs is to be discussed in the following from a few selected studies (Table 2). Gonzalez et al. [74] screened 25 different co-substrates (amino acids, organic acids, and monosaccharides) for methanol assimilation in methylotrophic E. coli. Only for a few natively degradable amino acids, acetate, pyruvate, succinate, xylose, and glucose, biomass doublings >1 in the presence of methanol were observed. In experiments employing 13C-methanol labeling, threonine was identified as a promising candidate with relatively high abundance of labeled amino acids, leading to the finding of regulatory involvement by a leucine-responsive regulatory protein for methanol assimilation. In addition to the improvement of methanol utilization, the performed screening elucidates the potential of mixotrophy regarding versatility of applicable co-substrate. Another example of methanol mixotrophy worth mentioning is the work of Zhang et al. [75] demonstrating yield improvement of succinate from 91% to 98% through methanol assimilation and reaching succinate titer up to 68.54 g L1. Further, additional potential improvement was hypothesized by the authors through enhanced methanol assimilation and interconversion of NADH and NAD. Although only a small fraction of consumed methanol was fixated into the final product (1.45% of succinate), the yield improvement at competitive titers highlights the potential of mixotrophic approach for biosynthesis. The applicability for bioproduction of fine chemicals, such as flavanone naringenin, was demonstrated for methanol mixotrophy with 18% of the final product with at least one carbon originated from methanol [76]. In addition, solventogenic methylotrophy for methanol conversion to ethanol and n-butanol was demonstrated in vitro by Bogorad et al.

358

Y. Hong and A.-P. Zeng

Table 2 Selected examples of methylotrophic mixotrophy Microorganism Methylotrophic Escherichia coli

Substrates Organic acids, amino acids, xylose, glucose, methanol

Product Biomass

Glucose, methanol

Succinate

Glucose, xylose, tryptone, yeast extract, methanol

Naringenin

Xylose, formate, methanol, bicarbonate

Biomass, ethanol

Methylotrophic Corynebacterium glutamicum

Glucose, ribose, methanol

Cadaverine

Methylotrophic Saccharomyces cerevisiae

Yeast extract, methanol

Biomass, pyruvate

Engineered Pichia pastoris

Glycerol, methanol

Monacolin J, lovastatin

Sorbitol, glycerol, methanol

Recombinant human erythropoietin

Feature Evaluation of 25 co-substrates for methanol assimilation, identification of L-threonine as methanol assimilation enhancing co-substrate Introduction of methanol assimilation module to succinate production strain, demonstration of incorporation of C1 compounds with yield increase and redox influence Introduction of methanol assimilation module, 13 C-labeling detection of central carbon intermediates from methanol, first demonstration of specialty chemical biosynthesis from C1 compounds Introduction of modified serine cycle utilizing methanol, formate, and bicarbonate, ethanol biosynthesis from C1 compounds demonstrated via labeling experiments Multi-step engineering for synthetic methylotrophy for methanol assimilation over ribulose monophosphate pathway, demonstration of mixotrophic production of non-native product cadaverine Introduction of methanol assimilation module, elevation of methanol consumption and biomass generation via supplementation of yeast extract Pathway splitting and co-cultivation of mutants for biosynthesis of monacolin J and lovastatin, glycerolfeeding for enhanced biomass generation Analysis of methanol feeding rate in relation to sorbitol consumption rate, fed-batch high cell density operation

Source [74]

[75]

[76]

[77]

[78]

[79]

[80]

[81]

Biosynthesis Based on One-Carbon Mixotrophy

359

[82]. In addition to the synthetic ribulose monophosphate pathway for methanol assimilation, (modified) serine cycle for C1 assimilation to ethanol was demonstrated by Yu and Liao [77]. Synthetic methylotrophy is not limited to E. coli and alternative model-microorganisms, such as Corynebacterium glutamicum and Saccharomyces cerevisiae, were likewise engineered for methylotrophic mixotrophy: Leßmeier et al. [78] engineered C. glutamicum for the diamine compound cadaverine reaching up to 1.5 g L1 in shake flask cultivations, which was confirmed via labeling experiments. Dai et al. [79] reported introduction of methanol assimilation genes from Pichia pastoris into S. cerevisiae, resulting in increase of biomass generation as well as pyruvate secretion. Detailed strategies and analysis of strain engineering for synthetic methylotrophy can be found in the review of Bennet et al. [69]. In contrast to the advances of synthetic methylotrophy, native methylotrophs offer unique traits comparable to acetogens: the native capability of methylotrophy is accompanied by highly adapted and efficient operation and cellular regulatory mechanisms of C1 fixation that eliminate additional metabolic fine-tuning, which is crucial for synthetic methylotrophy. For instance, engineered Bacillus methanolicus strain was reported to produce L-glutamate at a titer of 69 g L1 solely from methanol as demonstrated by Brautaset et al. [83]. Mixotroph examples can be also found for native methylotrophs. In the work of Liu et al. [80], different strategies were applied for the native methylotroph yeast Pichia pastoris aiming biosynthesis of pharmaceuticals, monacolin J and lovastatin. Strain engineering and optimization to introduce genes from Aspergillus terreus strains led to the production of both compounds. To further improve the biosynthesis on methanol, co-culture strategy was applied with different mutants of P. pastoris in an up-scaled bioreactor system (5 L), reaching titers up to 593.9 mg L1 and 250.8 mg L1 of monacolin J and lovastatin, respectively. Although simultaneous co-utilization of substrates was not applied, prior glycerol-feeding for enhanced biomass generation provides an alternative strategy for mixotrophy overcoming limitations due to limited quantity of cellular machinery. Similar batch-wise co-substrate feeding of sorbitol and methanol was reported in the work of Celik et al. [81]. In addition, engineering of P. pastoris to an autotroph was recently reported by Gassler et al. [84], broadening the applicability of C1 feedstocks.

4 Mixotrophy via Anaplerotic and Naturally Occurring Carbon Fixating Reactions One of the major advantages of mixotrophy is the broad applicability. In contrast to sole utilization of C1 compounds for all cellular processes, the proportion of organic substrates in mixotrophy can vary according to the intended design – from supplementary additive for complex cellular intermediates to streamlined proportion of the desired product, so that even single anaplerotic carbon fixating reaction can be used

360

Y. Hong and A.-P. Zeng

as C1 fixation module for elevation of carbon yields. In fact, C1 fixating reaction can be already found in the central carbon metabolism of heterotrophs [85]. Inclusion of multiple sequential fixation reactions results in a fixation pathway. However, depending on the applicability, incomplete fixation cycle can still be employed to draw up the potential of biological C1 fixation. The large repertoire of known C1 fixating enzymes enables finding ideal candidates for the desired biosynthetic pathway, as previously approx. 5,000 known enzymes were employed for analysis of synthetic carbon fixation pathways [8]. With the development of artificial enzymes and engineering of enzymes [86– 88], an even broader spectrum of enzymes is available. The most crucial criterion for the selection of appropriate C1 fixating enzyme lies in the reversible nature of enzymatic reaction. Elimination of undesired thermodynamic condition plays an essential role as exemplarily described for the potential condensation reaction of formyl-CoA due to its similarity to acetyl-CoA, which is hampered by its poor reaction equilibrium [89]. Further, pathway optimization is crucial as demonstrated in the following examples. In vivo utilization of pyruvate formate-lyase (PFL) examined by Zelcbuch et al. [90] for the fixation of formate in mixotrophy employing acetate shows the importance of metabolic connectivity of key metabolites: condensation of acetyl-CoA and formate to pyruvate (C2 + C1 ! C3) was demonstrated through deletion of isocitrate lyase and overexpression of PFL and PFL-activating enzyme in E. coli (s. Fig. 2a). Another example constitutes the utilization of phosphoenolpyruvate (PEP) carboxylase (PEPC), PEP carboxykinase (PEPCK), pyruvate carboxylase (PC), and reverse operating malate dehydrogenases (C3 + C1 ! C4) that are of high importance for succinate and L-malate biosynthesis as discussed by Zhu and Tang [91] and Dong et al. [92]: depending on the desired C1 fixation strategy, fine-tuning of metabolic fluxes for parallel or competing reactions of the interplay between TCA cycle and glycolysis is required. Further, reversely operating decarboxylases constitutes another potential candidate for such mixotrophic application. Shifting the reaction condition to the carboxylation direction was applied for in vitro biosynthesis of amino acids [93], phenol and styrene derivatives [94] demonstrating C1 fixation for C  5 molecules. By rerouting native metabolic pathways, Satanowski et al. [95] demonstrated CO2 fixation pathway by 6-phosphogluconate dehydrogenase as key enzyme of reductive carboxylation, in which only endogenous enzymes in E. coli were used. Deep analysis of carboxylases for CO2 fixation can be found elsewhere, Erb [85], Schada von Borzskowski et al. [96]. Recently, utilization of aldolases for C1 fixations was demonstrated for formaldehyde condensation to pyruvate via 2-keto-4-hydroxybutyrate aldolase (KHBA) or to acetaldehyde via deoxyribose-5-phosphate aldolase (DERA) [97–99], highlighting applicability for carbonyl group containing precursors. Re-purposing natural catalysts not only for stoichiometric conversion of organic substrates, but also for concatenating incorporation of low-value feedstock and C1 fixation as demonstrated by Meng et al. [99] for biosynthesis of 1,3-propanediol from ethanol and formaldehyde opens up new possibilities. Among the various C1 fixation systems, glycine synthase (reversely operating glycine cleavage system, Fig. 2b) obtained in the past years special attention as core

Biosynthesis Based on One-Carbon Mixotrophy

361

Fig. 2 Examples of anaplerotic and naturally occurring carbon fixation reactions. (a): Anaplerotic reactions of the central carbon metabolism are potential candidates for mixotrophic C1 fixationenhanced bioconversion. C1 compounds such as formaldehyde (HCHO), formic acid (HCOOH), CO2 or bicarbonate (HCO3) can be carboxylated/condensed to C2, C3 compounds to provide carbon intermediates for biomass generation or biochemical synthesis. (b): Another example constitutes glycine synthase reaction that is the reversely operating glycine cleavage. Close cyclic reaction cascade of the four glycine synthase proteins is generating glycine, a C2 amino acid, assimilating CO2 and methylene-5,10-tetrahydrofolate (CH2-THF). Glycine can be further converted to central carbon intermediates, such as pyruvate via serine hydroxymethyltransferase and serine dehydratase or acetyl-CoA via glycine reductase, phosphate acetyltransferase, and PFL under thioredoxin (TRX) oxidation. Bold lines represent carbon fixating reactions. Dashed lines depict multiple conversion reactions

reaction for the reductive glycine pathway for formate use as proposed by Bar-Even et al. [100]. Natively, the four proteins (P: glycine decarboxylase, T: aminomethyltransferase, L: dihydrolipoyl dehydrogenase, H: carrier protein) of the glycine cleavage system (GCS) plays a crucial role in one-carbon and folate metabolism [101, 102]. Although the full glycine synthesis reaction including return to the reactive state of the H-protein is comprised of multiple reactions, the balanced reaction sequence constitutes glycine synthesis from CO2, 5,10-methylenetetrahydrofolate, ammonia, and a reducing equivalent. With recent demonstrations of the reductive glycine pathway for artificial formatotrophic (and methylotrophic) growth, several studies with repurposed glycine cleavage system were reported [38, 103–109]. For engineered E. coli strains, several engineering targets were addressed simultaneously: among others, construction of auxotroph strains, improving formate condensation and reduction, overexpressing glycine synthase genes, optimizing supply of energy, and reducing equivalents (by methanol oxidation) and re-wiring of central carbon metabolism. However, for less genetically accessible strains, such as Clostridium pasteurianum, the C1-fixating capability of glycine synthase was demonstrated as well [110]: solely the integration of glycine synthase

362

Y. Hong and A.-P. Zeng

of Gottschalkia acidurici led to formate consumption exceeding its native production for glucose mixotrophy and up to 46% reduced native C1 production in glycerol fermentation. In the latter case, the total carbon yield was improved from 84.8% to 92.1% consisting of biomass, solvents, and organic acids. Accompanied by the targeted utilization of glycine synthase for C1 fixation, further knowledge regarding the reaction mechanism and dynamics of the shuttle protein H was recently gained. Using molecular dynamic simulations Zhang et al. [111] characterized in silico the major steps of the interaction between H and T proteins and identified the key residues responsible for release of the aminomethyl moiety carrying lipoate arm from a cavity of protein H, a step governing the GCS activity. Engineering of a key residue, Ser-67 in protein H, led to a bidirectional tuning of the release process, thus opening ways to target C1 metabolism for the utilization of formate and CO2 for biosynthesis. Interestingly, Xu et al. [112] discovered stand-alone catalytic activity of lipoylated H protein in an in vitro study for glycine synthesis and cleavage without the involvement of other GCS proteins. These recent findings and improved understanding of reaction mechanisms may help to more efficiently engineer GCS (glycine synthase) for utilization of formate and CO2 in formatotrophic or mixotrophic biosynthesis.

5 The Case of Mixotrophic Ethanol Biosynthesis Ethanol as the simplest and most common biofuel is being produced predominantly (>96% in 2020) from crops/sugar, while utilization of alternative biomass (lignocellulosic and algal biomass) is being intensively studied and devolved to the level of technical application and demonstration [113, 114] in addition to use of gaseous C1 compounds [115]. The market size for bioethanol is estimated to be 140 billion liters in 2022 [113]. Independently from the source of sugar for ethanol fermentation, its anaerobic conversion nearly occurs with a molar ratio of 1:1 for ethanol and CO2 [116]. In addition to direct usage of CO2 from ethanol fermentation [117], recycle of CO2 waste streams for the fixation into ethanol can be achieved by capture and fixation in photoreactor, gas fermentation system or via Fischer-Tropsch process coupled with electrolysis for hydrogen supply [118–120]. Alternatively, follow-up fixation of gaseous CO2 to alternative biochemicals can be applied, as reported via mixotrophy aiming at succinic acid production from recycled CO2 streams by Actinobacillus succinogenes with a CO2 fixation rate of 16.2 g L1 h1 [121]. However, simultaneous bioconversion of organic substrates and CO2 into the same product as ethanol in one reactor setup requires less equipment and operation costs and simplifies the overall fermentation process in addition to improved carbon efficiency. Metabolic engineering of S. cerevisiae for introduction of CalvinBenson-Bassham cycle (RuBP cycle) generated various examples for such a mixotrophic process enabling direct fixation of evolved CO2 and its bioconversion to ethanol [122–125]. The use of CO2 as a sink for reducing power partially replaced reducing power consumption through biosynthesis of glycerol with simultaneously

Biosynthesis Based on One-Carbon Mixotrophy

363

improved ethanol formation [122]. Improved integration of the RuBP cycle and carbon flux channeling via deletion of the gene coding for glycerol-3-phosphate dehydrogenase was reported to result in 15% higher ethanol yield [124]. Moreover, in the utilization of mixed-substrates (maltose and xylose) targeting bioconversion of lignocellulosic biomass, the engineered strain from Li et al. [123] fixated CO2 with up to 0.436 g L1 h1, exceeding CO2 fixation rates of photo- or chemoautotrophic microorganisms. A comparable engineering strategy was also reported for E. coli with total yield of ethanol, acetate, and lactate exceeding the theoretical heterotrophy molar yield (2.2 vs. 2.0) [126]. Although zero carbon emission was not reached, these examples demonstrate yield improvement over heterotrophy via incorporation of CO2 or avoiding of CO2 formation for well-established bioprocesses.

6 Concluding Remarks and Outlook Biosynthesis with negative or zero carbon emission is the ultimate goal of bioprocesses for chemical production. The research of the past years in microbial C1 fixation with various examples of acetogens, methylotrophs, and heterotrophs with integrated mixotrophic pathways shows the broad range of employable cellular machineries. Smart design of mixotrophic growth pathways in combination with suitable biosynthesis pathways is essential for full exploitation of mixotrophic biosynthesis. In addition, mixotrophy utilizing complex feedstocks such as hydrolyzed lignocellulose or other abundant organic waste streams in combination with C1 compounds enables streamlined bioconversion to biochemicals with minimal ecological footprint [95, 127]. However, feasibility in large-scale operations and competitiveness to conventional bioprocesses or chemical processes should still be demonstrated. Although mixotrophic biosynthesis can broaden the spectra of employable strains, fixation reactions, thermodynamic and kinetic boundaries, it comes with additional challenges. A major hurdle comes from potential catabolite repression as observed for acetogens, C. ljungdahlii, M. thermoacetica, or Eubacterium limosum, or similar metabolic reflexes, which can hamper co-utilization of a secondary substrate [29, 45, 128–130]. Comparably, catabolite repression was also hypothesized to limit methanol co-consumption for synthetic methylotrophic E. coli [74]. Furthermore, conditional metabolic regulations, such as formate-dependent acetogenic growth of Clostridium bovifaecis and Syntrophococcus sucromutans [131, 132], need to be addressed for more efficiently channeling carbon fluxes into target products. Currently, fine-tuning of cellular metabolism regulation restricts the choice of cellular hosts that can be subjected to genetic manipulation with established knowledge of genetic regulations or tools of molecular biology. In addition, unexpected metabolic imbalances due to cellular compensation mechanisms may occur, for which careful holistic metabolic engineering and functional genomics (system biology) are required. So far, there is only limited omics data available for understanding cellular behavior and biosynthesis on mixotrophy

364

Y. Hong and A.-P. Zeng

[29]. Independent from the starting point of native or synthetic C1 fixating hosts, we believe that identification of potential reciprocal influences of regulations in the underlying heterotrophy and autotrophy pathways is crucial for practical application and industrial relevance of mixotrophy. Potentially, cellular hurdles that are accompanied by C1 assimilation constitute a balancing act for the economical aspect: although abundant presence of C1 substrates (as waste streams) and successful incorporation into the desired product would reduce the required amount of carbon source and thus pure material costs, any compromises or sacrifices in volumetric productivity and cell growth due to metabolic bottlenecks associated with C1 assimilation would increase the investment or operational costs. In comparison with purely heterotrophic or autotrophic bioprocesses, however, the convenience of mixotrophic adjustability between both modes can be used to bridge unfavorable operation phases that facilitate large-scale operation (e.g., heterotrophy for the buildup phase of biomass and gradual transition to mixotrophy for the production phase). Further, emerging techniques to assimilate non-gaseous C1 substrates alleviate mass-transfer-related concerns for large-scale operation. For a commercial industrial bioprocess for bulk chemicals or fuels, the following process parameters are normally needed: product concentration nearly or above 10%, productivity higher than 1–2 g L1 h1, and yield nearly theoretical maximum. Currently, autotrophy is far away from these targets with the exception of syngas conversion to ethanol and acetate and alternative strategies are urgently required for a widespread application of C1 fixation in industrial bioprocesses [133]. For instance, stepwise and modular engineering of heterotrophic strains into a mixotroph, while maintaining the desired process parameters, may pose a realistic approach. With the knowledge gained from recent efforts in exploitation of C1 fixation, one can choose from various native or synthetic pathways or even single anaplerotic reaction for such a purpose. For the design and implementation of C1-fixation-enhanced bioprocesses, more efforts are needed to get beyond the stage of proof-of-concept.

References 1. Choi KR, Jiao S, Lee SY (2020) Metabolic engineering strategies toward production of biofuels. Curr Opin Chem Biol 59:1–14. https://doi.org/10.1016/j.cbpa.2020.02.009 2. Almquist J, Cvijovic M, Hatzimanikatis V et al (2014) Kinetic models in industrial biotechnology - improving cell factory performance. Metab Eng 24:38–60. https://doi.org/10.1016/j. ymben.2014.03.007 3. Folch PL, Bisschops MMM, Weusthuis RA (2021) Metabolic energy conservation for fermentative product formation. Microb Biotechnol. https://doi.org/10.1111/1751-7915.13746 4. Woodley JM (2017) Bioprocess intensification for the effective production of chemical products. Comput Chem Eng 105:297–307. https://doi.org/10.1016/j.compchemeng.2017. 01.015 5. Cueto-Rojas HF, van Maris AJA, Wahl SA et al (2015) Thermodynamics-based design of microbial cell factories for anaerobic product formation. Trends Biotechnol 33:534–546. https://doi.org/10.1016/j.tibtech.2015.06.010

Biosynthesis Based on One-Carbon Mixotrophy

365

6. Bar-Even A, Noor E, Milo R (2012) A survey of carbon fixation pathways through a quantitative lens. J Exp Bot 63:2325–2342. https://doi.org/10.1093/jxb/err417 7. Katsyv A, Müller V (2020) Overcoming energetic barriers in acetogenic C1 conversion. Front Bioeng Biotechnol 8:621166. https://doi.org/10.3389/fbioe.2020.621166 8. Bar-Even A, Noor E, Lewis NE et al (2010) Design and analysis of synthetic carbon fixation pathways. Proc Natl Acad Sci U S A 107:8889–8894. https://doi.org/10.1073/pnas. 0907176107 9. Löwe H, Kremling A (2021) In-depth computational analysis of natural and artificial carbon fixation pathways. bioRxiv. https://doi.org/10.1101/2021.01.05.425423 10. Bertsch J, Müller V (2015) Bioenergetic constraints for conversion of syngas to biofuels in acetogenic bacteria. Biotechnol Biofuels 8:210. https://doi.org/10.1186/s13068-015-0393-x 11. Schuchmann K, Müller V (2014) Autotrophy at the thermodynamic limit of life: a model for energy conservation in acetogenic bacteria. Nat Rev Microbiol 12:809–821. https://doi.org/10. 1038/nrmicro3365 12. Teixeira LV, Moutinho LF, Romão-Dumaresq AS (2018) Gas fermentation of C1 feedstocks: commercialization status and future prospects. Biofuels Bioprod Biorefin 12:1103–1117. https://doi.org/10.1002/bbb.1912 13. Drzyzga O, Revelles O, Durante-Rodríguez G et al (2015) New challenges for syngas fermentation: towards production of biopolymers. J Chem Technol Biotechnol 90:1735– 1751. https://doi.org/10.1002/jctb.4721 14. Wainaina S, Horváth IS, Taherzadeh MJ (2018) Biochemicals from food waste and recalcitrant biomass via syngas fermentation: a review. Bioresour Technol 248:113–121. https://doi.org/ 10.1016/j.biortech.2017.06.075 15. Mohammadi M, Najafpour GD, Younesi H et al (2011) Bioconversion of synthesis gas to second generation biofuels: a review. Renew Sust Energ Rev 15:4255–4273. https://doi.org/ 10.1016/j.rser.2011.07.124 16. Kundiyana DK, Huhnke RL, Wilkins MR (2010) Syngas fermentation in a 100-L pilot scale fermentor: design and process considerations. J Biosci Bioeng 109:492–498. https://doi.org/ 10.1016/j.jbiosc.2009.10.022 17. Redl S, Diender M, Jensen TØ et al (2017) Exploiting the potential of gas fermentation. Ind Crop Prod 106:21–30. https://doi.org/10.1016/j.indcrop.2016.11.015 18. Gong F, Liu G, Zhai X et al (2015) Quantitative analysis of an engineered CO2-fixing Escherichia coli reveals great potential of heterotrophic CO2 fixation. Biotechnol Biofuels 8: 86. https://doi.org/10.1186/s13068-015-0268-1 19. Stoll IK, Boukis N, Sauer J (2020) Syngas fermentation to alcohols: reactor technology and application perspective. Chemie Ingenieur Technik 92:125–136. https://doi.org/10.1002/cite. 201900118 20. Roy P, Dutta A, Chang S (2016) Development and evaluation of a functional bioreactor for CO fermentation into ethanol. Bioresour Bioprocess 3. https://doi.org/10.1186/s40643-0160082-z 21. Yasin M, Jang N, Lee M et al (2019) Bioreactors, gas delivery systems and supporting technologies for microbial synthesis gas conversion process. Bioresour Technol Rep 7: 100207. https://doi.org/10.1016/j.biteb.2019.100207 22. Kantzow C, Mayer A, Weuster-Botz D (2015) Continuous gas fermentation by Acetobacterium woodii in a submerged membrane reactor with full cell retention. J Biotechnol 212:11–18. https://doi.org/10.1016/j.jbiotec.2015.07.020 23. Yasin M, Jeong Y, Park S et al (2015) Microbial synthesis gas utilization and ways to resolve kinetic and mass-transfer limitations. Bioresour Technol 177:361–374. https://doi.org/10. 1016/j.biortech.2014.11.022 24. Mao W, Yuan Q, Qi H et al (2020) Recent progress in metabolic engineering of microbial formate assimilation. Appl Microbiol Biotechnol 104:6905–6917. https://doi.org/10.1007/ s00253-020-10725-6

366

Y. Hong and A.-P. Zeng

25. Claassens NJ, He H, Bar-Even A (2019) Synthetic methanol and formate assimilation via modular engineering and selection strategies. Curr Issues Mol Biol 33:237–248. https://doi. org/10.21775/cimb.033.237 26. Gomaa MA, Al-Haj L, Abed RMM (2016) Metabolic engineering of cyanobacteria and microalgae for enhanced production of biofuels and high-value products. J Appl Microbiol 121:919–931. https://doi.org/10.1111/jam.13232 27. Zhang L, Zhao R, Jia D et al (2020) Engineering Clostridium ljungdahlii as the gas-fermenting cell factory for the production of biofuels and biochemicals. Curr Opin Chem Biol 59:54–61. https://doi.org/10.1016/j.cbpa.2020.04.010 28. Emerson DF, Stephanopoulos G (2019) Limitations in converting waste gases to fuels and chemicals. Curr Opin Biotechnol 59:39–45. https://doi.org/10.1016/j.copbio.2019.02.004 29. Fast AG, Schmidt ED, Jones SW et al (2015) Acetogenic mixotrophy: novel options for yield improvement in biofuels and biochemicals production. Curr Opin Biotechnol 33:60–72. https://doi.org/10.1016/j.copbio.2014.11.014 30. Subashchandrabose SR, Ramakrishnan B, Megharaj M et al (2013) Mixotrophic cyanobacteria and microalgae as distinctive biological agents for organic pollutant degradation. Environ Int 51:59–72. https://doi.org/10.1016/j.envint.2012.10.007 31. Wang J, Yang H, Wang F (2014) Mixotrophic cultivation of microalgae for biodiesel production: status and prospects. Appl Biochem Biotechnol 172:3307–3329. https://doi.org/ 10.1007/s12010-014-0729-1 32. Chen C-Y, Yeh K-L, Aisyah R et al (2011) Cultivation, photobioreactor design and harvesting of microalgae for biodiesel production: a critical review. Bioresour Technol 102:71–81. https://doi.org/10.1016/j.biortech.2010.06.159 33. Chen Y, Xu C, Vaidyanathan S (2018) Microalgae: a robust “green bio-bridge” between energy and environment. Crit Rev Biotechnol 38:351–368. https://doi.org/10.1080/07388551. 2017.1355774 34. Patel AK, Choi YY, Sim SJ (2020) Emerging prospects of mixotrophic microalgae: way forward to sustainable bioprocess for environmental remediation and cost-effective biofuels. Bioresour Technol 300:122741. https://doi.org/10.1016/j.biortech.2020.122741 35. Chen FY-H, Jung H-W, Tsuei C-Y et al (2020) Converting Escherichia coli to a synthetic methylotroph growing solely on methanol. Cell 182:933–946.e14. https://doi.org/10.1016/j. cell.2020.07.010 36. Gleizer S, Ben-Nissan R, Bar-On YM et al (2019) Conversion of Escherichia coli to generate all biomass carbon from CO2. Cell 179:1255–1263.e12. https://doi.org/10.1016/j.cell.2019. 11.009 37. Bang J, Hwang CH, Ahn JH et al (2020) Escherichia coli is engineered to grow on CO2 and formic acid. Nat Microbiol 5:1459–1463. https://doi.org/10.1038/s41564-020-00793-9 38. Kim S, Lindner SN, Aslan S et al (2020) Growth of E. coli on formate and methanol via the reductive glycine pathway. Nat Chem Biol 16:538–545. https://doi.org/10.1038/s41589-0200473-5 39. Tracy BP, Jones SW, Fast AG et al (2012) Clostridia: the importance of their exceptional substrate and metabolite diversity for biofuel and biorefinery applications. Curr Opin Biotechnol 23:364–381. https://doi.org/10.1016/j.copbio.2011.10.008 40. Redl S, Sukumara S, Ploeger T et al (2017) Thermodynamics and economic feasibility of acetone production from syngas using the thermophilic production host Moorella thermoacetica. Biotechnol Biofuels 10:150. https://doi.org/10.1186/s13068-017-0827-8 41. Bar-Even A, Flamholz A, Noor E et al (2012) Thermodynamic constraints shape the structure of carbon fixation pathways. Biochim Biophys Acta 1817:1646–1659. https://doi.org/10. 1016/j.bbabio.2012.05.002 42. Zeng A-P (2021) Fermentation and bioprocess development. In: Kirk RE, Othmer DF (eds) Encyclopedia of chemical technology. Wiley. In print 43. Schiel-Bengelsdorf B, Dürre P (2012) Pathway engineering and synthetic biology using acetogens. FEBS Lett 586:2191–2198. https://doi.org/10.1016/j.febslet.2012.04.043

Biosynthesis Based on One-Carbon Mixotrophy

367

44. Köpke M, Held C, Hujer S et al (2010) Clostridium ljungdahlii represents a microbial production platform based on syngas. Proc Natl Acad Sci U S A 107:13087–13092. https:// doi.org/10.1073/pnas.1004716107 45. Tremblay P-L, Zhang T, Dar SA et al (2012) The Rnf complex of Clostridium ljungdahlii is a proton-translocating ferredoxin:NAD+ oxidoreductase essential for autotrophic growth. mBio 4:e00406-12. https://doi.org/10.1128/mBio.00406-12 46. Biegel E, Müller V (2010) Bacterial Na+translocating ferredoxin:NAD+ oxidoreductase. Proc Natl Acad Sci U S A 107:18138–18142. https://doi.org/10.1073/pnas.1010318107 47. Nevin KP, Hensley SA, Franks AE et al (2011) Electrosynthesis of organic compounds from carbon dioxide is catalyzed by a diversity of acetogenic microorganisms. Appl Environ Microbiol 77:2882–2886. https://doi.org/10.1128/AEM.02642-10 48. Igarashi K, Kato S (2017) Extracellular electron transfer in acetogenic bacteria and its application for conversion of carbon dioxide into organic compounds. Appl Microbiol Biotechnol 101:6301–6307. https://doi.org/10.1007/s00253-017-8421-3 49. Emerson DF, Woolston BM, Liu N et al (2019) Enhancing hydrogen-dependent growth of and carbon dioxide fixation by Clostridium ljungdahlii through nitrate supplementation. Biotechnol Bioeng 116:294–306. https://doi.org/10.1002/bit.26847 50. Daniell J, Köpke M, Simpson S (2012) Commercial biomass syngas fermentation. Energies 5: 5372–5417. https://doi.org/10.3390/en5125372 51. de Tissera S, Köpke M, Simpson SD et al (2019) Syngas biorefinery and syngas utilization. Adv Biochem Eng Biotechnol 166:247–280. https://doi.org/10.1007/10_2017_5 52. Leang C, Ueki T, Nevin KP et al (2013) A genetic system for Clostridium ljungdahlii: a chassis for autotrophic production of biocommodities and a model homoacetogen. Appl Environ Microbiol 79:1102–1109. https://doi.org/10.1128/AEM.02891-12 53. Huang H, Chai C, Li N et al (2016) CRISPR/Cas9-based efficient genome editing in Clostridium ljungdahlii, an autotrophic gas-fermenting bacterium. ACS Synth Biol 5:1355–1361. https://doi.org/10.1021/acssynbio.6b00044 54. Zhao R, Liu Y, Zhang H et al (2019) CRISPR-Cas12a-mediated gene deletion and regulation in Clostridium ljungdahlii and its application in carbon flux redirection in synthesis gas fermentation. ACS Synth Biol 8:2270–2279. https://doi.org/10.1021/acssynbio.9b00033 55. Huang H, Chai C, Yang S et al (2019) Phage serine integrase-mediated genome engineering for efficient expression of chemical biosynthetic pathway in gas-fermenting Clostridium ljungdahlii. Metab Eng 52:293–302. https://doi.org/10.1016/j.ymben.2019.01.005 56. Nagarajan H, Sahin M, Nogales J et al (2013) Characterizing acetogenic metabolism using a genome-scale metabolic reconstruction of Clostridium ljungdahlii. Microb Cell Factories 12: 118. https://doi.org/10.1186/1475-2859-12-118 57. Liu JK, Lloyd C, Al-Bassam MM et al (2019) Predicting proteome allocation, overflow metabolism, and metal requirements in a model acetogen. PLoS Comput Biol 15:e1006848. https://doi.org/10.1371/journal.pcbi.1006848 58. Jin S, Bae J, Song Y et al (2020) Synthetic biology on acetogenic bacteria for highly efficient conversion of C1 gases to biochemicals. Int J Mol Sci 21. https://doi.org/10.3390/ ijms21207639 59. Fast AG, Papoutsakis ET (2012) Stoichiometric and energetic analyses of non-photosynthetic CO2-fixation pathways to support synthetic biology strategies for production of fuels and chemicals. Curr Opin Chem Eng 1:380–395. https://doi.org/10.1016/j.coche.2012.07.005 60. Jones SW, Fast AG, Carlson ED et al (2016) CO2 fixation by anaerobic non-photosynthetic mixotrophy for improved carbon conversion. Nat Commun 7:12800. https://doi.org/10.1038/ ncomms12800 61. Drake HL, Daniel SL, Küsel K et al (1997) Acetogenic bacteria: what are the in situ consequences of their diverse metabolic versatilities? Biofactors 6:13–24. https://doi.org/10. 1002/biof.5520060103

368

Y. Hong and A.-P. Zeng

62. Schuchmann K, Schmidt S, Martinez Lopez A et al (2015) Nonacetogenic growth of the acetogen Acetobacterium woodii on 1,2-propanediol. J Bacteriol 197:382–391. https://doi. org/10.1128/JB.02383-14 63. Schuchmann K, Müller V (2016) Energetics and application of heterotrophy in acetogenic bacteria. Appl Environ Microbiol 82:4056–4069. https://doi.org/10.1128/AEM.00882-16 64. Diner BA, Fan J, Scotcher MC et al (2018) Synthesis of heterologous mevalonic acid pathway enzymes in Clostridium ljungdahlii for the conversion of fructose and of syngas to mevalonate and isoprene. Appl Environ Microbiol 84. https://doi.org/10.1128/AEM.01723-17 65. Maru BT, Munasinghe PC, Gilary H et al (2018) Fixation of CO2 and CO on a diverse range of carbohydrates using anaerobic, non-photosynthetic mixotrophy. FEMS Microbiol Lett 365. https://doi.org/10.1093/femsle/fny039 66. Whitaker WB, Sandoval NR, Bennett RK et al (2015) Synthetic methylotrophy: engineering the production of biofuels and chemicals based on the biology of aerobic methanol utilization. Curr Opin Biotechnol 33:165–175. https://doi.org/10.1016/j.copbio.2015.01.007 67. Fei Q, Guarnieri MT, Tao L et al (2014) Bioconversion of natural gas to liquid fuel: opportunities and challenges. Biotechnol Adv 32:596–614. https://doi.org/10.1016/j. biotechadv.2014.03.011 68. Schrader J, Schilling M, Holtmann D et al (2009) Methanol-based industrial biotechnology: current status and future perspectives of methylotrophic bacteria. Trends Biotechnol 27:107– 115. https://doi.org/10.1016/j.tibtech.2008.10.009 69. Bennett RK, Steinberg LM, Chen W et al (2018) Engineering the bioconversion of methane and methanol to fuels and chemicals in native and synthetic methylotrophs. Curr Opin Biotechnol 50:81–93. https://doi.org/10.1016/j.copbio.2017.11.010 70. Zhang W, Zhang T, Wu S et al (2017) Guidance for engineering of synthetic methylotrophy based on methanol metabolism in methylotrophy. RSC Adv 7:4083–4091. https://doi.org/10. 1039/C6RA27038G 71. Chistoserdova L (2018) Applications of methylotrophs: can single carbon be harnessed for biotechnology? Curr Opin Biotechnol 50:189–194. https://doi.org/10.1016/j.copbio.2018. 01.012 72. Wang Y, Fan L, Tuyishime P et al (2020) Synthetic methylotrophy: a practical solution for methanol-based biomanufacturing. Trends Biotechnol 38:650–666. https://doi.org/10.1016/j. tibtech.2019.12.013 73. Chen AY, Lan EI (2020) Chemical production from methanol using natural and synthetic methylotrophs. Biotechnol J 15:e1900356. https://doi.org/10.1002/biot.201900356 74. Gonzalez JE, Bennett RK, Papoutsakis ET et al (2018) Methanol assimilation in Escherichia coli is improved by co-utilization of threonine and deletion of leucine-responsive regulatory protein. Metab Eng 45:67–74. https://doi.org/10.1016/j.ymben.2017.11.015 75. Zhang W, Zhang T, Song M et al (2018) Metabolic engineering of Escherichia coli for high yield production of succinic acid driven by methanol. ACS Synth Biol 7:2803–2811. https:// doi.org/10.1021/acssynbio.8b00109 76. Whitaker WB, Jones JA, Bennett RK et al (2017) Engineering the biological conversion of methanol to specialty chemicals in Escherichia coli. Metab Eng 39:49–59. https://doi.org/10. 1016/j.ymben.2016.10.015 77. Yu H, Liao JC (2018) A modified serine cycle in Escherichia coli coverts methanol and CO2 to two-carbon compounds. Nat Commun 9:3992. https://doi.org/10.1038/s41467-018-06496-4 78. Leßmeier L, Pfeifenschneider J, Carnicer M et al (2015) Production of carbon-13-labeled cadaverine by engineered Corynebacterium glutamicum using carbon-13-labeled methanol as co-substrate. Appl Microbiol Biotechnol 99:10163–10176. https://doi.org/10.1007/s00253015-6906-5 79. Dai Z, Gu H, Zhang S et al (2017) Metabolic construction strategies for direct methanol utilization in Saccharomyces cerevisiae. Bioresour Technol 245:1407–1412. https://doi.org/ 10.1016/j.biortech.2017.05.100

Biosynthesis Based on One-Carbon Mixotrophy

369

80. Liu Y, Tu X, Xu Q et al (2018) Engineered monoculture and co-culture of methylotrophic yeast for de novo production of monacolin J and lovastatin from methanol. Metab Eng 45: 189–199. https://doi.org/10.1016/j.ymben.2017.12.009 81. Celik E, Calik P, Oliver SG (2009) Fed-batch methanol feeding strategy for recombinant protein production by Pichia pastoris in the presence of co-substrate sorbitol. Yeast 26:473– 484. https://doi.org/10.1002/yea.1679 82. Bogorad IW, Chen C-T, Theisen MK et al (2014) Building carbon-carbon bonds using a biocatalytic methanol condensation cycle. Proc Natl Acad Sci U S A 111:15928–15933. https://doi.org/10.1073/pnas.1413470111 83. Brautaset T, Jakobsen ØM, Degnes KF et al (2010) Bacillus methanolicus pyruvate carboxylase and homoserine dehydrogenase I and II and their roles for L-lysine production from methanol at 50 degrees C. Appl Microbiol Biotechnol 87:951–964. https://doi.org/10.1007/ s00253-010-2559-6 84. Gassler T, Sauer M, Gasser B et al (2020) The industrial yeast Pichia pastoris is converted from a heterotroph into an autotroph capable of growth on CO2. Nat Biotechnol 38:210–216. https://doi.org/10.1038/s41587-019-0363-0 85. Erb TJ (2011) Carboxylases in natural and synthetic microbial pathways. Appl Environ Microbiol 77:8466–8477. https://doi.org/10.1128/AEM.05702-11 86. Siegel JB, Smith AL, Poust S et al (2015) Computational protein design enables a novel one-carbon assimilation pathway. Proc Natl Acad Sci U S A 112:3704–3709. https://doi.org/ 10.1073/pnas.1500545112 87. Lu X, Liu Y, Yang Y et al (2019) Constructing a synthetic pathway for acetyl-coenzyme A from one-carbon through enzyme design. Nat Commun 10:1378. https://doi.org/10.1038/ s41467-019-09095-z 88. Gong F, Zhu H, Zhang Y et al (2018) Biological carbon fixation: from natural to synthetic. J CO2 Util 28:221–227. https://doi.org/10.1016/j.jcou.2018.09.014 89. Bar-Even A (2016) Formate assimilation: the metabolic architecture of natural and synthetic pathways. Biochemistry 55:3851–3863. https://doi.org/10.1021/acs.biochem.6b00495 90. Zelcbuch L, Lindner SN, Zegman Y et al (2016) Pyruvate formate-lyase enables efficient growth of Escherichia coli on acetate and formate. Biochemistry 55:2423–2426. https://doi. org/10.1021/acs.biochem.6b00184 91. Zhu L-W, Tang Y-J (2017) Current advances of succinate biosynthesis in metabolically engineered Escherichia coli. Biotechnol Adv 35:1040–1048. https://doi.org/10.1016/j. biotechadv.2017.09.007 92. Dong X, Chen X, Qian Y et al (2017) Metabolic engineering of Escherichia coli W3110 to produce L-malate. Biotechnol Bioeng 114:656–664. https://doi.org/10.1002/bit.26190 93. Martin J, Eisoldt L, Skerra A (2018) Fixation of gaseous CO2 by reversing a decarboxylase for the biocatalytic synthesis of the essential amino acid l-methionine. Nat Catal 1:555–561. https://doi.org/10.1038/s41929-018-0107-4 94. Wuensch C, Glueck SM, Gross J et al (2012) Regioselective enzymatic carboxylation of phenols and hydroxystyrene derivatives. Org Lett 14:1974–1977. https://doi.org/10.1021/ ol300385k 95. Satanowski A, Dronsella B, Noor E et al (2020) Awakening a latent carbon fixation cycle in Escherichia coli. Nat Commun 11:5812. https://doi.org/10.1038/s41467-020-19564-5 96. Schada von Borzyskowski L, Rosenthal RG, Erb TJ (2013) Evolutionary history and biotechnological future of carboxylases. J Biotechnol 168:243–251. https://doi.org/10.1016/j.jbiotec. 2013.05.007 97. Wang C, Ren J, Zhou L et al (2019) An aldolase-catalyzed new metabolic pathway for the assimilation of formaldehyde and methanol to synthesize 2-keto-4-hydroxybutyrate and 1,3-propanediol in Escherichia coli. ACS Synth Biol 8:2483–2493. https://doi.org/10.1021/ acssynbio.9b00102

370

Y. Hong and A.-P. Zeng

98. He H, Höper R, Dodenhöft M et al (2020) An optimized methanol assimilation pathway relying on promiscuous formaldehyde-condensing aldolases in E. coli. Metab Eng 60:1–13. https://doi.org/10.1016/j.ymben.2020.03.002 99. Meng H, Wang C, Yuan Q et al (2021) An aldolase-based new pathway for bioconversion of formaldehyde and ethanol into 1,3-propanediol in Escherichia coli. ACS Synth Biol 10:799– 809. https://doi.org/10.1021/acssynbio.0c00597 100. Bar-Even A, Noor E, Flamholz A et al (2013) Design and analysis of metabolic pathways supporting formatotrophic growth for electricity-dependent cultivation of microbes. Biochim Biophys Acta 1827:1039–1047. https://doi.org/10.1016/j.bbabio.2012.10.013 101. Kikuchi G (1973) The glycine cleavage system: composition, reaction mechanism, and physiological significance. Mol Cell Biochem 1:169–187. https://doi.org/10.1007/ BF01659328 102. Brosnan ME, Brosnan JT (2016) Formate: the neglected member of one-carbon metabolism. Annu Rev Nutr 36:369–388. https://doi.org/10.1146/annurev-nutr-071715-050738 103. Yishai O, Bouzon M, Döring V et al (2018) In vivo assimilation of one-carbon via a synthetic reductive glycine pathway in Escherichia coli. ACS Synth Biol 7:2023–2028. https://doi.org/ 10.1021/acssynbio.8b00131 104. Bang J, Lee SY (2018) Assimilation of formic acid and CO2 by engineered Escherichia coli equipped with reconstructed one-carbon assimilation pathways. Proc Natl Acad Sci U S A 115:E9271–E9279. https://doi.org/10.1073/pnas.1810386115 105. Gonzalez L, de Cruz J, Machens F, Messerschmidt K et al (2019) Core catalysis of the reductive glycine pathway demonstrated in yeast. ACS Synth Biol 8:911–917. https://doi. org/10.1021/acssynbio.8b00464 106. Sánchez-Andrea I, Guedes IA, Hornung B et al (2020) The reductive glycine pathway allows autotrophic growth of Desulfovibrio desulfuricans. Nat Commun 11:5090. https://doi.org/10. 1038/s41467-020-18906-7 107. Claassens NJ, Bordanaba-Florit G, Cotton CAR et al (2020) Replacing the calvin cycle with the reductive glycine pathway in Cupriavidus necator. Metab Eng 62:30–41. https://doi.org/ 10.1016/j.ymben.2020.08.004 108. Tashiro Y, Hirano S, Matson MM et al (2018) Electrical-biological hybrid system for CO2 reduction. Metab Eng 47:211–218. https://doi.org/10.1016/j.ymben.2018.03.015 109. Song Y, Lee JS, Shin J et al (2020) Functional cooperation of the glycine synthase-reductase and Wood-Ljungdahl pathways for autotrophic growth of Clostridium drakei. Proc Natl Acad Sci U S A 117:7516–7523. https://doi.org/10.1073/pnas.1912289117 110. Hong Y, Arbter P, Wang W et al (2020) Introduction of glycine synthase enables uptake of exogenous formate and strongly impacts the metabolism in Clostridium pasteurianum. Biotechnol Bioeng. https://doi.org/10.1002/bit.27658 111. Zhang H, Li Y, Nie J et al (2020) Structure-based dynamic analysis of the glycine cleavage system suggests key residues for control of a key reaction step. Commun Biol 3:756. https:// doi.org/10.1038/s42003-020-01401-6 112. Xu Y, Li Y, Zhang H et al (2021) Stand-alone lipoylated H-protein of the glycine cleavage system enables glycine cleavage and the synthesis of glycine from one-carbon compounds in vitro. bioRxiv. https://doi.org/10.1101/2021.03.28.437365 113. Sharma B, Larroche C, Dussap C-G (2020) Comprehensive assessment of 2G bioethanol production. Bioresour Technol 313:123630. https://doi.org/10.1016/j.biortech.2020.123630 114. Harun R, Yip JWS, Thiruvenkadam S et al (2014) Algal biomass conversion to bioethanol - a step-by-step assessment. Biotechnol J 9:73–86. https://doi.org/10.1002/biot.201200353 115. Richter H, Martin M, Angenent L (2013) A two-stage continuous fermentation system for conversion of syngas into ethanol. Energies 6:3987–4000. https://doi.org/10.3390/en6083987 116. Bai FW, Anderson WA, Moo-Young M (2008) Ethanol fermentation technologies from sugar and starch feedstocks. Biotechnol Adv 26:89–105. https://doi.org/10.1016/j.biotechadv.2007. 09.002

Biosynthesis Based on One-Carbon Mixotrophy

371

117. Xu Y, Isom L, Hanna MA (2010) Adding value to carbon dioxide from ethanol fermentations. Bioresour Technol 101:3311–3319. https://doi.org/10.1016/j.biortech.2010.01.006 118. Huang Z, Grim G, Schaidle J et al (2020) Using waste CO2 to increase ethanol production from corn ethanol biorefineries: techno-economic analysis. Appl Energy 280:115964. https:// doi.org/10.1016/j.apenergy.2020.115964 119. Liew F, Martin ME, Tappel RC et al (2016) Gas fermentation-a flexible platform for commercial scale production of low-carbon-fuels and chemicals from waste and renewable feedstocks. Front Microbiol 7:694. https://doi.org/10.3389/fmicb.2016.00694 120. Seth JR, Wangikar PP (2015) Challenges and opportunities for microalgae-mediated CO2 capture and biorefinery. Biotechnol Bioeng 112:1281–1296. https://doi.org/10.1002/bit.25619 121. Zhang Q, Nurhayati CC-L et al (2017) Carbon capture and utilization of fermentation CO2: integrated ethanol fermentation and succinic acid production as an efficient platform. Appl Energy 206:364–371. https://doi.org/10.1016/j.apenergy.2017.08.193 122. Guadalupe-Medina V, Wisselink HW, Luttik MA et al (2013) Carbon dioxide fixation by Calvin-cycle enzymes improves ethanol yield in yeast. Biotechnol Biofuels 6:125. https://doi. org/10.1186/1754-6834-6-125 123. Li Y-J, Wang M-M, Chen Y-W et al (2017) Engineered yeast with a CO2-fixation pathway to improve the bio-ethanol production from xylose-mixed sugars. Sci Rep 7:43875. https://doi. org/10.1038/srep43875 124. Papapetridis I, Goudriaan M, Vázquez Vitali M et al (2018) Optimizing anaerobic growth rate and fermentation kinetics in Saccharomyces cerevisiae strains expressing Calvin-cycle enzymes for improved ethanol yield. Biotechnol Biofuels 11:17. https://doi.org/10.1186/ s13068-017-1001-z 125. Xia P-F, Zhang G-C, Walker B et al (2017) Recycling carbon dioxide during xylose fermentation by engineered Saccharomyces cerevisiae. ACS Synth Biol 6:276–283. https://doi.org/ 10.1021/acssynbio.6b00167 126. Tseng I-T, Chen Y-L, Chen C-H et al (2018) Exceeding the theoretical fermentation yield in mixotrophic rubisco-based engineered Escherichia coli. Metab Eng 47:445–452. https://doi. org/10.1016/j.ymben.2018.04.018 127. Vees CA, Neuendorf CS, Pflügl S (2020) Towards continuous industrial bioprocessing with solventogenic and acetogenic clostridia: challenges, progress and perspectives. J Ind Microbiol Biotechnol 47:753–787. https://doi.org/10.1007/s10295-020-02296-2 128. Huang H, Wang S, Moll J et al (2012) Electron bifurcation involved in the energy metabolism of the acetogenic bacterium Moorella thermoacetica growing on glucose or H2 plus CO2. J Bacteriol 194:3689–3699. https://doi.org/10.1128/JB.00385-12 129. Lebloas P, Lindley ND, Loubiere P (1996) Regulation of carbon and energy metabolism during the linear growth phase in batch fermentations of the acetogenic methylotroph Eubacterium limosum on methanol/CO2. Enzym Microb Technol 19:187–195. https://doi.org/10. 1016/0141-0229(95)00230-8 130. Tan Y, Liu J, Chen X et al (2013) RNA-seq-based comparative transcriptome analysis of the syngas-utilizing bacterium Clostridium ljungdahlii DSM 13528 grown autotrophically and heterotrophically. Mol BioSyst 9:2775–2784. https://doi.org/10.1039/c3mb70232d 131. Yao Y, Fu B, Han D et al (2020) Formate-dependent acetogenic utilization of glucose by the fecal acetogen Clostridium bovifaecis. Appl Environ Microbiol 86. https://doi.org/10.1128/ aem.01870-20 132. Doré J, Bryant MP (1990) Metabolism of one-carbon compounds by the ruminal acetogen Syntrophococcus sucromutans. Appl Environ Microbiol 56:984–989. https://doi.org/10.1128/ AEM.56.4.984-989.1990 133. Zeng A-P (2019) New bioproduction systems for chemicals and fuels: needs and new development. Biotechnol Adv 37:508–518. https://doi.org/10.1016/j.biotechadv.2019.01.003

Adv Biochem Eng Biotechnol (2022) 180: 373–408 https://doi.org/10.1007/10_2021_180 © The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 Published online: 23 November 2021

Conversion of Carbon Monoxide to Chemicals Using Microbial Consortia Ivette Parera Olm and Diana Z. Sousa

Contents 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 Syngas Fermentation for a Circular Economy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 Microbes Using Carbon Monoxide for Growth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3 The Microbial Consortia Approach for Syngas Fermentation . . . . . . . . . . . . . . . . . . . . . . . 2 CO Conversion by Open Mixed Cultures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Anaerobic Sludges as Biocatalysts for Syngas Fermentation . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Syngas Biomethanation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 Production of Ethanol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4 Production of Carboxylic Acids and Higher Alcohols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 CO Conversion by Synthetic Co-cultures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Synthetic Co-cultures: A Win-Win . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Production of Methane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3 Production of Carboxylic Acids and Alcohols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4 Production of Other Value-Added Chemicals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 CO Conversion via Sequential Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Challenges and Opportunities of Syngas-Fermenting Microbial Communities . . . . . . . . . . . 6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

374 374 375 376 378 378 384 386 387 389 389 390 393 394 395 396 398 398

Abstract Syngas, a gaseous mixture of CO, H2 and CO2, can be produced by gasification of carbon-containing materials, including organic waste materials or lignocellulosic biomass. The conversion of bio-based syngas to chemicals is foreseen as an important process in circular bioeconomy. Carbon monoxide is also produced as a waste gas in many industrial sectors (e.g., chemical, energy, steel). Often, the purity level of bio-based syngas and waste gases is low and/or the ratios of syngas components are not adequate for chemical conversion (e.g., by I. Parera Olm and D. Z. Sousa (*) Laboratory of Microbiology, Wageningen University and Research, Wageningen, The Netherlands e-mail: [email protected]; [email protected]

374

I. Parera Olm and D. Z. Sousa

Fischer-Tropsch). Microbes are robust catalysts to transform impure syngas into a broad spectrum of products. Fermentation of CO-rich waste gases to ethanol has reached commercial scale (by axenic cultures of Clostridium species), but production of other chemical building blocks is underexplored. Currently, genetic engineering of carboxydotrophic acetogens is applied to increase the portfolio of products from syngas/CO, but the limited energy metabolism of these microbes limits product yields and applications (for example, only products requiring low levels of ATP for synthesis can be produced). An alternative approach is to explore microbial consortia, including open mixed cultures and synthetic co-cultures, to create a metabolic network based on CO conversion that can yield products such as medium-chain carboxylic acids, higher alcohols and other added-value chemicals. Graphical Abstract

Keywords Acetogens, C1 feedstocks, Cross-feeding, Gas fermentation, Microbial consortia, Microbial interactions, Syngas

1 Introduction 1.1

Syngas Fermentation for a Circular Economy

As the worldwide population grows and the consumption of fossil resources increases, there is the need to develop new technologies to produce commodity chemicals from renewable resources. By 2050, chemicals may no longer be synthesised from fossil fuels, according to targets established after the Paris agreement and the European Green Deal [1, 2]. Lignocellulosic biomass and wastes (agricultural, industrial and municipal) have been identified as priority feedstocks for a bio-based industry [3, 4]. These are inedible materials, and their use does not compete with human or animal nutrition or with the utilisation of arable land, therefore circumventing ethical concerns. Wastes in particular are heavily underutilised materials, especially in developing countries [5]. The conversion of biomass and wastes through hydrolysis-fermentation is very attractive, but bottlenecks of the process are the low biodegradability of lignin (which represents 10–25% of plant

Conversion of Carbon Monoxide to Chemicals Using Microbial Consortia

375

biomass) and the costly pre-treatment steps [6, 7]. An alternative that gets increasing attention is the gasification of biomass and wastes followed by the chemical or biological conversion of the generated synthesis gas (also known as syngas) [8, 9]. Syngas is a gas mixture of mainly carbon monoxide (CO), hydrogen (H2) and carbon dioxide (CO2) that can be generated from solid carbonaceous feedstocks (e.g., coal, lignocellulosic biomass) and carbon-containing wastes (e.g., agricultural waste). Chemical conversion of syngas by, e.g., Fischer-Tropsch (FT) process is a mature technology used for the conversion of mainly coal-generated syngas into hydrocarbons, alcohols and organic acids [10]. FT processes use metal catalysts under high temperature and pressures and require high H2:CO molar ratios. Chemical catalysts are highly sensitive to syngas impurities such as ammonia [11], sulphur species [12], alkali ions [13] or water [11], which makes them less suitable for the treatment of biomass/waste-generated syngas. Gas clean-up treatments can reduce the concentration of most impurities, but complete removal is hindered by the cost of these technologies and the inherent variability of the feedstock [14]. Biological conversion of syngas involves its fermentation by microorganisms, which are in general more resistant to impurities in the gas and, in addition, do not require a fix H2:CO molar ratio [8, 15–17]. The biological route operates under mild temperature and pressure conditions, and overall has higher mass and energy conversion efficiencies compared to chemical catalysis [9, 18]. Furthermore, microbial processes result in higher product selectivity with the formation of fewer by-products. Syngas fermentation technology can also be applied for the treatment of CO-containing waste gases from heavy industry such as steelmaking. Often, CO-rich off-gases gases from steel mills are burned leading to CO2 emissions; in 2019, on average 1.83 tonnes of CO2 were emitted per every ton of steel produced [19], contributing to approximately 8% of global emissions. This is a serious environmental problem with impact on climate change. Other opportunities are emerging to use gas fermentation technology associated to CO2 capture technology. For example, the production of CO by electrochemical reduction of CO2 has been proved feasible and with high Faraday efficiencies (>80%) [20–22].

1.2

Microbes Using Carbon Monoxide for Growth

Microbes have exploited CO as sustenance for much of their evolutionary history. Proof of that are the different ways in which CO may be involved in microbial metabolism, which makes it necessary to define some terms. Microorganisms that can use CO as carbon and energy source are denominated carboxydotrophs, to be distinguished from carboxydovores, which may use electrons from CO but require organic carbon for growth [23]. At the same time, CO metabolism can take two forms: respiratory and fermentative [24]. The former relies either on O2 (aerobic) or other external electron acceptors (anaerobic). In this review, the focus is on the latter: fermentative CO metabolism, which is, by definition, anaerobic. Microorganisms that ferment CO are carboxydotrophic. Therefore, the term carboxydotroph is used

376

I. Parera Olm and D. Z. Sousa

in this text to refer to the ability to use CO anaerobically but, in another context, it may refer to both aerobic and anaerobic microorganisms. The fermentation of CO/syngas is carried out by acetogens, a specialised group of anaerobic bacteria able to use CO and H2/CO2 as sole carbon and energy sources via the reductive acetyl-CoA pathway, also known as the Wood-Ljungdahl pathway (WLP) [8]. Acetogenesis is not a phylogenetic trait; it is widely represented in at least 23 bacterial genera [25, 26]. Most known acetogens belong to the genera Clostridium and Acetobacterium, within the Clostridia class. The WLP results in acetyl-CoA as end-product of CO and H2/CO2 fermentation. Since autotrophy via the WLP is energetically limited, most acetyl-CoA is directed towards acetate production to generate ATP. Thus, the majority of acetogens produce acetic acid as sole metabolic end-product. Some microorganisms can derive other chemicals from acetyl-CoA as intermediate. For example, Clostridium autoethanogenum, Clostridium ljungdahlii, Clostridium ragsdalei and Alkalibaculum bacchi are able to produce ethanol; C. autoethanogenum, C. ljungdahlii and C. ragsdalei can also produce 2,3-butanediol (2,3-BDO); Eubacterium limosum and Butyribacterium methylotrophicum are able to produce butyrate; and Clostridium carboxidivorans can produce butyrate, butanol, caproate and hexanol [27]. The key enzyme of CO oxidation to CO2, carbon monoxide dehydrogenase, is present in other anaerobic microorganisms that harbour variations of the WLP. Besides acetogenic bacteria, CO can be used as electron donor and/or carbon source by some methanogenic archaea and sulphate-reducing bacteria [28]. However, compared to acetogens, methanogens and sulphate-reducing bacteria are more sensitive to elevated levels of CO. Syngas fermentation processes can be implemented with pure cultures of acetogens or with microbial communities. This chapter focuses on the latter: undefined and defined consortia of microorganisms that convert syngas to biochemicals of interest. For an overview of monoculture-based processes, we refer to recent reviews [17, 29, 30].

1.3

The Microbial Consortia Approach for Syngas Fermentation

The fermentation of syngas has been most studied and implemented in industry using pure cultures of acetogens [8, 16]. From a process perspective, monocultures are easy to control and predict, since optimal conditions for growth are well-defined. Process conditions can be tuned to target a product of interest with high selectivity and yields. As an example, the highest ethanol concentration (48 g L
1) and volumetric productivity (369 g L
1 day
1) from syngas were achieved with pure cultures of C. ljungdahlii [31, 32]. However, a major limitation of the use of monocultures for syngas conversion is the energetic constraint in the formation of products other than acetate and ethanol [33]. Genetic and metabolic engineering

Conversion of Carbon Monoxide to Chemicals Using Microbial Consortia

377

of Clostridia strains has advanced remarkably in the last decade, paving the way towards the expression of heterologous products and enhanced yields [34, 35]. For instance, industrially relevant titres have recently been achieved for acetone, iso-propanol and 2,3-BDO [36]. Yet, important hurdles remain to be addressed in genetic engineering of Clostridia to further expand the product portfolio of syngas fermentation, such as low DNA transformation efficiencies, insufficient highthroughput recombineering tools and, in general, the need for a better understanding of acetogenic platforms at the molecular level [37]. Another disadvantage of monoculture-based strategies is the lack of robustness against process fluctuations. This is of particular relevance in the case of syngas fermentation, since the gas composition varies depending on the source or gasification method [38]. Mixed cultures are less affected than monocultures of carboxydotrophs by changes in the syngas composition [39], and are also expected to be more robust against syngas impurities (e.g., nitrogen oxides, tars), which have been shown to inhibit cell growth or interfere with product distribution in monocultures of acetogens [40]. Microbial consortia are emerging as a promising strategy aimed at overcoming the limitations of monocultures and taking syngas fermentation a step forward [30, 41, 42]. In nature, microbes rarely thrive alone; instead, cooperation and communication with other microorganisms are extremely important for survival [43, 44]. Communities can perform complicated functions that individual populations cannot, for instance the production of energy-demanding products. Compared to monocultures, microbial consortia can convert much more complex substrates and have better robustness, both because of a highly diverse community structure and a capacity to evolve. The capabilities of microbial consortia have long been exploited in bioremediation, wastewater treatment and the production of fermented foods [45]. In the last decade, advances in -omics approaches and a greater understanding of microbial interactions have driven forward the fields of microbiome engineering and synthetic ecology, aimed at unlocking the full potential of microbial communities for biotechnological applications [46–50]. The use of microbial consortia in syngas fermentation has specific advantages. For one thing, communities composed of multiple carboxydotrophic microbes with different CO tolerance can handle syngas streams with variable composition. This functional redundancy may enhance gas consumption and mitigate the detrimental effect of syngas contaminants on individual populations. Moreover, provided that carboxydotrophic populations keep CO levels low, CO-sensitive microbes can thrive in an environment that would otherwise be hostile. In addition, the co-culture capabilities can be extended by syntrophic interactions between species in the consortia, such as cross-feeding of intermediates (e.g., acetate, ethanol) or exchange of essential nutrients (e.g., amino acids, vitamins). An example of this is the co-culture of Citrobacter amalonaticus Y19 and the acetogen Sporomusa ovata. The latter has been reported to produce acetate from CO, but at rather low rates [51, 52]. On the other hand, C. amalonaticus Y19 is unable to use autotrophic substrates but it can oxidise CO to H2 and CO2 [53], which may be used as substrates by the acetogen. A study found that co-cultures of S. ovata and C. amalonaticus Y19 produced almost double the amount of acetate than monocultures of S. ovata, from

378

I. Parera Olm and D. Z. Sousa

the same amount of CO [52]. In addition, growth of both microorganisms and CO consumption was higher in the co-culture than in monocultures. This example is just one of many that demonstrate the relevance of mutualistic interactions in microbial consortia [42, 54]. There are, on the other hand, potential downsides when using consortia of microorganisms compared to monocultures. Some examples are the occurrence of competing or inhibiting pathways, the generation of side products that reduce product selectivity or incompatible cultivation conditions (i.e., pH, temperature, etc.) between species in the community. These issues can be tackled through rational microbial consortia engineering, multi-species metabolic modelling and bioreactor/bioprocess design [47, 55, 56]. Two types of microbial consortia can be distinguished: open mixed cultures (also referred to as ‘open cultures’, ‘mixed cultures’ or ‘microbiomes’) and synthetic co-cultures (Fig. 1). The former consist of self-assembled, highly diverse microbial communities naturally occurring in defined habitats, in which the populations are mostly unspecified. Synthetic co-cultures are consortia of specified microbial strains that engage in interaction under aseptic and controlled conditions. Most synthetic co-cultures reported in literature are composed of two or three microbial species, with a few including up to five [57]. The following sections summarise the main developments regarding the implementation of open mixed cultures (Sect. 2) and synthetic co-cultures (Sect. 3) in syngas fermentation processes.

2 CO Conversion by Open Mixed Cultures 2.1

Anaerobic Sludges as Biocatalysts for Syngas Fermentation

The main components of syngas (CO, H2 and CO2) can sustain anaerobic growth of a number of microbial groups: acetogens and hydrogenogenic bacteria, carboxydotrophic and hydrogenotrophic methanogens and sulphate-reducing microorganisms [24]. In turn, the products of CO/syngas fermentation (mainly H2, CO2, acetate and ethanol) can support growth of acetoclastic methanogens, chainelongating bacteria, ethanol oxidisers, syntrophic acetate-oxidising bacteria and propionibacteria, among others [16, 58]. The range of final products that can be obtained via syngas fermentation by open mixed cultures (in the absence of external electron acceptors) therefore includes short- and medium-chain carboxylic acids, simple and higher alcohols and methane. When sulphate is available and sulphatereducing microorganisms are present, sulphide is also produced. An overview of some of the works on syngas/CO conversion by mixed cultures is shown in Table 1. Open mixed cultures for syngas fermentation are based on inocula from anaerobic natural or engineered environments that harbour a high microbial diversity. Typical inocula include anaerobic digester sludges [66, 70–76], wastewater treatment

Conversion of Carbon Monoxide to Chemicals Using Microbial Consortia

379

Fig. 1 Overview of the types of microbial communities that can be applied in syngas fermentation to produce biochemicals. Syngas (CO, H2 and CO2) can be obtained via the gasification of organic wastes or lignocellulosic residues. Off-gases from the steel and iron industries are also sources of CO-rich gas. Open mixed cultures (from, e.g., sludges) enriched in carboxydotrophic, methanogenic and/or Clostridia species can be used to produce methane, alcohols or mediumchain carboxylic acids from syngas. Alternatively, synthetic co-cultures can be used, composed of non-engineered microorganisms alone or in combination with genetically engineered platform organisms such as Escherichia coli. The latter option allows to expand the range of products that can be obtained from syngas to compounds of added-value, such as 3-hydroxypropionic acid or itaconic acid. Acronym: GM, genetically modified

7

7

8.2

60

TBR coupled to FBG 60

55

35–37 5.8–6.7

55

35

TBR (7.5L)

Floating MBR

MOBB

HfMBR

HfMBR

Biofilm: Clostridium (86.3%) A. bacchi (56%), A. propionicum (34%), Clostridium sp. (10%)

4.5 7

HfMBR CSTR with cell-recycle

35 37

Acinetobacter, Alcaligenes, Rhodobacter, 7 to 5.8–6.5a Methanobacterium, Methanoseaeta

Biofilm: Clostridium (41.6%), undefined (42%)

Biofilm: Thermoanaerobacterium (92.8%)

N.A.

N.A.

Derived from Asimakopoulos et al. [59]

Enriched microbial taxa (relative abundance) Biofilm: Methanothermobacter (30%), Therminicola (16%), Coprothermobacter (23%) Liquid: undefined (30%), Therminicola (23%), Methanothermobacter (15%) Derived from Asimakopoulos et al. [59]

Column reactor filled 35 with porous pad

6

6.5

T ( C) pH 60 7

Cultivation system TBR (lab-scale)

Table 1 Overview of syngas fermentation processes using open mixed cultures

CO/H2 (60:40) CO/H2/N2 (28: 60:12)

CO/N2 (60:40)b

CO/H2 (40:60)

CO/H2/CO2/N2 (20:45:25:10) CO/H2/CO2/N2 (11:35:44:10) CO/H2/CO2 (55: 20:10) CO/H2/CO2 (60: 30:10) CO/H2 (40:60)

Syngas composition (% v/v) CO/H2/CO2/N2 (20:45:25:10)

Methane 17.6 mmol Lbed
1 h
1 Methane 14.4 mmol Lbed-1 h
1 Methane 1.43 mmol L-1 h
1 Methane 3.04 mmol L
1 h
1 Acetate 24.6 g L
1 16.4 g L
1 day
1 Butyrate 1.4 g L
1 Caproate 0.88 g L
1 Caprylate 0.53 g L
1 Caproate 0.22 g L
1 Heptanoate 0.21 g L
1 Caprylate 0.15 g L
1 Ethanol 16.9 g L
1 Ethanol 8 g L
1; Propanol 6 g L
1 Butanol 1 g L
1

Product titre/ productivity Methane 8.49 mmol Lbed
1 h
1

Wang et al. [66–68] Liu et al. [69]

He et al. [65]

Shen et al. [64]

Shen et al. [64]

Asimakopoulos et al. [60, 61] Asimakopoulos et al. [60, 61] Chandolias et al. [62] [63]

Reference Asimakopoulos et al. [59]

380 I. Parera Olm and D. Z. Sousa

37

33

STR

STR

N.A.

C. ljungdahlii, C. carboxidivorans, C. kluyveri

6–4.8a

4.9

N.A.

7–4.3a CO/H2/CO2/N2 (32:32:8:28) CO/H2/CO2/N2 (32:32:8:28) CO (100)

Butyrate 2.17 g L
1 Butanol 0.43 g L
1 Butanol 1.1 g L
1 Hexanol 0.6 g L
1 Ethanol 11.1 g L
1 Butanol 1.8 g L
1 Hexanol 1.5 g L
1 Chakraborty et al. [72]

Ganigué et al. [71]

Ganigué et al. [70]

TBR trickle-bed reactor, FBG fluidised bed gasifier, MBR membrane bioreactor, MOBB multi-orifice baffled bioreactor, HfMBR hollow-fibre membrane bioreactor, STR stirred-tank reactor (batch), N.A. not available data a The pH was initially set and not controlled afterwards b Percentage at the end of the process, with an inflow syngas composition of 60 kPa CO and N2 as make-up gas, assuming a total pressure of 101 kPa

37

STR

Conversion of Carbon Monoxide to Chemicals Using Microbial Consortia 381

382

I. Parera Olm and D. Z. Sousa

granules [76, 77] and faeces of herbivores [78–82]. Sludges from anaerobic digesters employed in traditional wastewater treatment processes have been proposed as most suitable syngas biocatalysts [73, 76, 77]. These cultures have a high adaptation capacity, essential to treat a vast range of organic and inorganic substrates. Most importantly for syngas fermentation, the ability to oxidise CO seems to be a ubiquitous property across anaerobic sludges [76]. An acclimation period is generally required to obtain a microbiome capable of efficiently converting CO. The inoculum largely determines the duration of this acclimation period, which can be as short as a few days [76, 77, 79, 81] or weeks [67, 71, 76], and in some cases lasts several months [65, 73, 75]. Because different microbes have different optimal conditions for growth, the outcome of the process depends not only on the inoculum source but also on the environmental conditions applied to enrich and maintain the culture. Acetogenic bacteria are more tolerant to CO than other microorganisms present in anaerobic sludges, therefore dominating enriched cultures exposed to moderate or high levels of CO [64, 66, 67, 75, 83– 85]. For example, during the CO-enrichment process of an anaerobic sludge, the relative abundance of members from the Clostridiales order, which includes many acetogenic species, increased from 5% in the inoculum to 66–95% in enriched cultures (pCO ¼ 20–61 kPa) [75]. Similarly, a recent study showed that mixed cultures exposed to high pCO (96 kPa) were dominated by members of the Firmicutes phylum, to which many acetogens belong, while low pCO (35 kPa) shifted the community towards Proteobacteria, a phylum that includes hydrogenogenic carboxydotrophs [83]. Methanogens, on the other hand, are generally inhibited at moderate CO pressures starting from 30 to 80 kPa [73, 75, 77]. In a recent study, Duan et al. [83] revealed the crucial role of under-characterised taxa in CO-enriched communities. Authors identified novel bacterial genera and species which may participate in CO oxidation to end-products and maintain fundamental metabolism (e.g., citric cycle, amino acid biosynthesis), extending the functional redundancy of the communities and overall increasing their stability. Besides CO, the presence of other gases in the mixture has an impact on the performance of syngas-converting communities. With few exceptions, pure cultures of carboxydotrophs can rarely consume H2 and CO simultaneously, since almost all hydrogenases are inhibited by CO [86–91]. In contrast, mixed cultures can metabolise H2 along with CO, since H2 can be used by hydrogenotrophic microorganisms that might be present in the community [39]. Overall, the addition of CO2 and H2 has been shown to increase the microbial diversity of CO-enriched cultures and promote a higher acetate/ethanol ratio [39, 75]. Temperature and pH are two operational parameters with a major influence on the evolution of CO enrichments. Grimalt-Alemany et al. [92] showed that mesophilic syngas-enrichments (37 C) are characterised by a higher microbial diversity and a more intricate metabolic network compared to thermophilic syngas-enrichments (60 C). Another finding of that study was that the maximum specific growth rates of microbes were significantly higher (twofold) in thermophilic conditions. Similar

Conversion of Carbon Monoxide to Chemicals Using Microbial Consortia

383

findings were reported by Alves et al. [73], who observed a rapid decrease in microbial diversity in long-term CO/syngas enrichments of anaerobic sludges at 55 C. The pH is perhaps the most crucial parameter determining the structure and product composition of syngas-converting communities. Several authors have emphasised its critical effect on the regulation between acetogenesis and solventogenesis, requiring a tight control of acidic conditions [71, 72, 84, 85], while, for methanogenesis, neutral or slightly alkaline conditions are required [84, 85]. Finally, the addition of medium supplements (e.g., yeast extract, reducing agents) and inhibitors of specific types of metabolism is also common practice to alter the structure of microbial communities and stimulate the production of target products [72, 92, 93]. In this regard, the addition of methanogenic inhibitors is an extended practice in syngas fermentation processes using open mixed cultures to supress methane production. Inocula from anaerobic sludges are likely to harbour an active methanogenic population. This can be a hurdle when products other than methane are targeted. Three approaches are commonly used to inhibit methane production by open mixed cultures used in bioprocesses, namely (1) operation under (mildly) acidic conditions [94, 95], (2) heat-shock treatment of the inoculum [95, 96], and (3) the addition of methanogenic inhibitors such as 2-bromoethanesulphonate (2-BES) [70, 71, 95, 97, 98]. The latter, in concentrations ranging 10–50 mM, has proven very efficient and is therefore a popular choice; however, it can certainly contribute to increasing process costs at industrial scale since periodical addition is necessary in continuous operation. Moreover, 2-BES can lose efficacy during longterm operation [65] and can be metabolised by dehalogenating and sulphatereducing bacteria present in microbial communities [95, 99]. Eventually, moderate to high concentrations of CO should inhibit most methanogenic activity and avoid the addition of specific inhibitors. One of the big advantages of using open cultures in bioprocesses is that these do not require operation under aseptic conditions. In addition, long-term reproducibility of mixed cultures can be ensured by using suitable cryopreservation methods [100]. On the other hand, these systems are highly dependent on microbial interactions, which are very difficult to predict and, to a great extent, unknown. Other drawbacks are the long times required to achieve steady-state conditions and the challenging product recovery due to the presence of many by-products at low concentrations. Large-scale continuous processes using open mixed cultures are well-established in industry (e.g., in wastewater treatment and food fermentation), but not yet applied for syngas fermentation. Decades of research have shed light on the structure of anaerobic sludges, their governing microbial interactions and the influence of operational parameters, although mainly in the context of conversion of organic compounds [68, 101–103]. However, developments in the last decade are driving forward the syngas fermentation platform for mixed cultures at a rapid pace. The next sections relate these advances, centred on the production of methane, carboxylic acids and alcohols.

384

2.2

I. Parera Olm and D. Z. Sousa

Syngas Biomethanation

Conversion of CO-rich waste gases by open mixed cultures is a popular method to produce methane [84, 85, 104]. Traditional wastewater treatment processes rely on microbial communities that perform methanogenesis as ultimate step of anaerobic digestion. These cultures are suited to produce methane from syngas, provided they have or acquire a sufficient carboxydotrophic potential [77]. Biomethanation of syngas presents several advantages over its analogous catalytic process [84, 85, 105]. Microbes are less sensitive than metal catalysts to impurities and to the ratio C/H in syngas. Biocatalysts are cheap, self-replicating, and can yield high methane contents in a single step. Using microbes, higher methane selectivity can be obtained, in contrast to the use of metal catalysts that result in the production of higher hydrocarbons as by-products. However, production rates of biomethanation are lower compared to the chemical process. Recent years have witnessed increased efforts to improve the efficiency of syngas biomethanation; these include the development of novel reactor configurations, insights on the impact of operational parameters and improved knowledge on the microbial community structure and interactions [84, 85]. Methanogenesis from CO can take place via three routes: (1) direct conversion by carboxydotrophic methanogens, (2) via acetate as intermediate by acetoclastic methanogens, and (3) via H2/CO2 as intermediates by hydrogenotrophic methanogens. Direct methanation of CO is rather infrequent due to complete inactivation of methanogens in the presence of moderate concentrations of CO [76, 77]. Four methanogenic species have been demonstrated to use CO for growth: Methanothermobacter thermautotrophicus [106], Methanosarcina barkeri [107], Methanosarcina acetivorans [108, 109], and Methanobacter marburgensis [110]. However, all of them require rather long periods of adaptation to CO and growth is significantly slower than on their typical substrates. Consequently, CO conversion to methane in microbial communities is highly dependent on bacterial– archaeal interactions. Several studies have demonstrated the preferential use of certain pathways in anaerobic sludges used for syngas biomethanation. In this regard, the incubation temperature plays a determining role. Experiments with vancomycin, an inhibitor of acetogenic activity, have shown that, under mesophilic conditions, methanogenesis occurs primarily via acetate as intermediate [74, 93]. Mesophilic conditions are favourable to acetogenic bacteria, which provide acetate to acetoclastic methanogens; in contrast, higher temperatures generally shift the microbial community towards H2-producing carboxydotrophs, which favour the hydrogenotrophic methanogenic route [76, 77, 92]. In both environments, when the H2 pressure is kept sufficiently low, acetate can be converted to H2/CO2 by syntrophic acetate-oxidising bacteria, which can compete with acetoclastic methanogens and create a niche for hydrogenotrophic methanogens [78, 93]. The effect of temperature on the microbial composition of communities performing syngas biomethanation extends to the process performance. Several studies have reported the positive impact of thermophilic over mesophilic conditions on

Conversion of Carbon Monoxide to Chemicals Using Microbial Consortia

385

conversion rates [59, 61, 76, 77, 92]. For example, Grimalt-Alemany et al. [92] observed an 18-fold higher methane productivity from enrichments incubated at 60 C compared to enrichments incubated at 37 C. Thermophilic conditions are therefore the preferred mode of operation for syngas biomethanation processes [84, 85]. Two other operational conditions, the pH and the pCO, also influence the performance of methanogenic communities using syngas. Since most methanogens grow optimally around neutrality, syngas biomethanation processes are generally operated at pH values between 7 and 7.6 [84, 85]. Carboxydotrophic bacteria can also proliferate in this pH range [27], thus providing intermediates for methanogenesis. The effect of CO levels has been extensively studied. Alves et al. [73] reported no methane production in thermophilic enrichments from anaerobic sludge incubated with solely CO as substrate (35 kPa). Methane was detected in enrichments incubated with CO/H2/CO2 (pCO ¼ 18 kPa), but production ceased in subsequent transfers. Instead, both syngas- and CO-enriched cultures produced acetate. Elimination of methanogens in the enrichments could be due to the low growth rate of methanogens or their higher susceptibility to CO [77, 109, 110]. Similarly, Luo et al. [111] reported 50% lower methanogenic activity by an anaerobic sewage sludge exposed to a pCO of 51 kPa, compared to the control in the absence of CO. These observations are in line with those of Guiot et al. [77], who reported inhibition of methanogenesis in enriched granular sludge at pCO between 30 and 83 kPa. Nevertheless, some archaeal genus, such as Methanobacterium, have been shown to tolerate CO levels up to 96 kPa in microbial communities [83]. In addition, strategies are in place to enhance CO utilisation by methanogenic cultures. For instance, a system with gas recirculation enabled a CO conversion efficiency of 75% and a methane yield of 95% under a pCO of 60 kPa [77]. In some cases, an acclimation phase has enabled methane production by anaerobic sludges exposed to 100 kPa CO [93, 112]. Syngas biomethanation has been investigated in a variety of process configurations with the aim to improve gas-to-liquid transfer and cell concentrations. Besides the use of traditional stirred-tank reactors (CSTRs), tested designs include bubble columns, gas-lift reactors, trickle-bed reactors (TBRs) and multi-orifice baffled bioreactors (MOBBs) [84, 85]. The most promising results so far have been obtained with the use of TBRs. Recently, Asimakopoulos et al. [59] reported a CH4 productivity from syngas of 8.49 mmolLbed
1h
1 in a lab-scale TBR operated in continuous mode at 60 C. The inoculum used was an enriched mixture of two anaerobic sludges, intended to increase microbial diversity. Interestingly, the methanogenic population was more abundant in the biofilm of the TBR, while carboxydotrophic bacteria were mostly found in the liquid phase. In a follow-up study, the process was scaled up; a 7.5 L TBR, that used the same enrichment and syngas mixture and was operated in the same conditions as the lab-scale reactor, achieved a maximum CH4 productivity of 17.6 mmolLbed
1 h
1, the highest reported so far for syngas biomethanation [60]. At this rate, H2 and CO conversion efficiencies were 97% and 76%, respectively, and CH4 selectivity was 99%. The higher performance of the scaled-up system was attributed to an improved gas-liquid mass transfer due to a

386

I. Parera Olm and D. Z. Sousa

more efficient sparging component and a much higher height/diameter ratio, in addition to a more accurate pH control. To further test the process, the TBR was coupled to a gasifier that generated syngas from wood pellets; the generated gas contained 11–27% CO and was fed into the reactor at atmospheric pressure. The microbial consortium produced CH4 at a maximum rate of 14.4 mmolLbed
1 h
1 without any inhibitory effects [60]. Few other studies have demonstrated continuous operation of bioreactors for syngas biomethanation by open cultures with syngas (CO/H2/CO2) as sole substrate [60, 84, 85]. Pereira [63] studied syngas conversion to methane by a mesophilic sludge in a 10.6 L MOBB operated in continuous mode. The system produced CH4 at a maximum rate of 73 mmol L
1 day
1 with negligible amounts of by-products in the liquid and a YCH4/CO of 0.6–0.8 (mol/mol), higher than reported in similar works [84, 85]. Yet, the conversion efficiency eventually dropped due to the prolonged high flow rates applied [63]. In a recent study, Chandolias et al. [62] tested a novel configuration consisting of a floating membrane in a membrane bioreactor and achieved a maximum CH4 productivity of 34 mmol L
1 day
1, in this case, using a thermophilic digester sludge. Overall, CH4 productivity in syngas biomethanation processes is very dependent on the process configuration and specific process conditions, which affect gas-toliquid mass transfer and cell concentrations. Considerable progress over the last years and successful examples of scale-up cases such as that of Asimakopoulos et al. [60] offer good perspectives for syngas biomethanation in the future. A key aspect to bring this technology to commercial application will be to ensure its economic feasibility by, e.g., combining syngas biomethanation with existing gasification plants and improving reactor design to increase productivities [84, 85].

2.3

Production of Ethanol

Ethanol is undoubtedly the most common target product of syngas fermentation due to its commercial use as biofuel [16, 104, 113]. Despite high productivities have been achieved with pure cultures of acetogens, the robustness of open mixed culture operation has driven an interest for its production in these systems. Singla et al. [80] were the first to demonstrate ethanol production by microbial communities using syngas. In their study, a mesophilic enriched consortium obtained from faeces produced up to 2.2 g L
1 ethanol in semi-continuous mode (adding fresh syngas to serum bottles every 24 h). Liu et al. [69] tested continuous fermentation of syngas to ethanol by mixed culture in a CSTR including a cell recirculation unit. Authors reported the production of up to 8 g L
1 ethanol from syngas at 37 C and pH 7. Consumption of ethanol was followed by the accumulation of propanol and butanol, with peak concentrations of 6 g L
1 and 1 g L
1, respectively. The microbial community was composed of the alkaliphilic acetogen Alkalibaculum bacchi (56%), the propionibacterium Anaerotignum propionicum (formerly, Clostridium propionicum; [114]) (34%) and other Clostridia species (10%). A follow-up study concluded that the mixed culture could convert 50% more carboxylic acids into their

Conversion of Carbon Monoxide to Chemicals Using Microbial Consortia

387

respective alcohols compared to monocultures of A. bacchi [115], evidencing the positive effect of synergistic microbial interactions in syngas-fermenting communities. A limitation of the fermentation process of Liu et al. [69] was the rather low CO and H2 utilisation (20–60%), a common problem due to the low solubility of these gases and low gas mass transfer rates in CSTRs. Novel reactor configurations can help overcome this issue [116]. In a recent study, Wang et al. [67] reported a relatively high ethanol production from syngas by mixed culture in a hollow-fibre membrane biofilm reactor (HfMBR). HfMBRs, most popular in the field of gas and wastewater treatment, have recently attracted the attention of researchers in the field of syngas fermentation [64, 67, 116, 117]. In a HfMBR, a gaseous substrate flows through the lumen of a hollow-fibre membrane and is consumed by the biofilm formed on the outer surface of the membrane. The high surface area allows a high volumetric gas transfer rate which, in turn, translates into high production rates [118]. A non-acclimated sludge used by Wang et al. [67] produced up to 16.9 g L
1 ethanol from CO/H2 (60:40) in a HfMBR operated in consecutive batch at pH 4.5 and 35 C. Ethanol was the only soluble product of CO/H2 fermentation. Interestingly, a similar HfMBR-based process operating at pH 6.5 and 55 C converted CO/H2 (40:60) to mostly acetate (98.6%) [64]. While the different temperature, gas composition and sludge characteristics could have contributed to the divergent product profile observed in these two studies, pH is most likely the determining factor. Several studies have reported that an acidic pH is key to promote ethanol production in syngas fermentation cultures [71, 72, 84, 85, 119]. Yet, no ethanol (but acetate) was produced by a sludge-derived culture in a HfMBR operated at pH 4.5 using H2/CO2 as substrates [120], highlighting that CO, which is a stronger reductant than H2, is also essential to promote alcohol production by acetogens.

2.4

Production of Carboxylic Acids and Higher Alcohols

Acetate is the simplest carboxylic acid that can be produced from syngas. Titres in the range of 20–30 g L
1 have been obtained for sludge-derived consortia utilising syngas in continuous fermentation [64, 96]. Continuous operation of a thermophilic HfMBR reached a maximum acetate production rate of 16.4 g L
1 day
1 with high product selectivity [64]. However, acetate production by mixed cultures has not received much attention due to the significantly higher production rates that can be obtained by pure cultures and the rather low economic value of this product. Instead, over the last decade, increased attention has been given to the production of medium-chain carboxylic acids (MCCAs) via anaerobic fermentation processes, a type of biorefinery referred to as the carboxylate platform [121–124]. The carboxylate platform relies mostly on sludges used in classical anaerobic digestion and aims at revalorising wastewater streams. Recently, several studies have extended this platform to the revalorisation of CO-rich gases. The products of syngas fermentation, acetate and ethanol, can be used by microorganisms that perform ethanol-based

388

I. Parera Olm and D. Z. Sousa

chain-elongation, producing MCCAs such as butyrate and caproate as end-products. Different processes are devised to convert syngas into MCCAs by microbial communities, including the use of synthetic co-cultures, discussed in Sect. 3.3, and multiple-step processes, summarised elsewhere [58]. Here, the focus is on one-step conversions by open mixed cultures. Ethanol chain-elongating communities are present in both natural and engineered environments and are dominated by relatives of Clostridium kluyveri, the beststudied ethanol chain-elongating microorganism and only isolate to date [122, 124, 125]. Mesophilic conditions are preferred to C. kluyveri and most acetogens. However, the two isolated strains of C. kluyveri grow optimally at a pH of 6.8 and 7.6 [126, 127], while acetogenic bacteria generally thrive in mildly acidic conditions [27, 128]. Therefore, pH is critical in determining the outcome of syngasfermenting chain-elongating communities [70, 71, 129]. Ganigué et al. [70] used a carboxydotrophic enrichment from sludge in a syngas fermentation process without pH control. Acetogenesis was dominant at the initial pH of 7, while production of C4-C6 compounds prevailed at the mid/end of the fermentation, when the pH dropped to 4.3. At the end of the process, the C4-C6 products represented 75–90% of the total, with butyrate (2.17 g L
1) as main product. In a follow-up study, it was determined that pH values around 4.8 favoured a sustained production of higher alcohols [71]. In a semi-continuous process without pH control (initial pH 6), the mixed culture converted syngas to a maximum of 1.1 g L
1 butanol and 0.6 g L
1 hexanol. To favour the synthesis of C6 compounds, attributable to C. kluyveri, it was critical to prevent pH to decrease below 4.5–5. In a recent study, He et al. [65] used a novel reactor configuration to promote gas transfer in a chain-elongating process. The system consisted of a reactor filled with a porous sponge pad and with a gas recirculation line. CO was used as sole carbon and energy source, and the partial pressure was gradually increased through the fermentation, from 15 to 61 kPa CO. Similar to the studies of Ganigué et al. [70, 71], operation was done at mesophilic conditions and the pH, initially set at 7, was not controlled. In contrast to those studies, though, the culture did not produce alcohols but a mixture of odd- and even-chain carboxylic acids including caprylate (C8), detected for the first time in a syngas fermentation process by a mixed culture. Production of C6-C8 carboxylates only began at the end of the fermentation, when CO pressure was 61 kPa. Maximum concentrations of caproate, heptanoate and caprylate were 0.22 g L
1, 0.21 g L
1 and 0.15 g L
1, respectively. The production of C5-C8 carboxylates halted in the last phase, likely due to product inhibition. The different product profile compared to the studies of Ganigué and co-workers could be explained by the different inoculum source and enrichment process, which resulted in quite different microbial compositions of the enriched cultures. The consortium used by Ganigué et al. [71] was mainly composed of C. ljughdalii, C. carboxidivorans and C. kluyveri, while He et al. [65] enriched a microbial community dominated by species of Acinetobacter, Alcaligenes, Rhodobacteraceae and a low abundance of Clostridium spp. On the other hand, product concentrations and production rates achieved by He et al. [65] were not higher than those reported in similar studies. This could be due to (1) the use of a non-acclimated inoculum, (2) the

Conversion of Carbon Monoxide to Chemicals Using Microbial Consortia

389

use of CO (not syngas) as sole carbon and energy source and (3) CO provided (60 kPa CO) being insufficient. Regardless, this work demonstrated that up to C8 carboxylic acids can be produced from CO in one-pot cultivation. Similar to ethanol production, the use of HfMBR has proven very promising for the production of carboxylic acids from syngas. Shen et al. [64] demonstrated production of MCCAs from CO/H2 for the first time in a HfMBR. Cultivation of the sludge-derived culture was studied at pH 6 under mesophilic (35 C) and thermophilic (55 C) conditions. In both scenarios, utilisation of CO and H2 exceeded 95%. Mesophilic cultivation in sequential batch mode produced caproate (0.88 g L
1) and up to 0.53 g L
1 caprylate, the highest caprylate concentration reported for a CO-fermenting system using mixed cultures. In contrast, thermophilic batch cultivation yielded a high acetate concentration (27.9 g L
1) and product specificity (96.7%), with butyrate (50 kPa CO in monocultures, it was sustained under a headspace of 130 kPa CO in co-cultivation with C. autoethanogenum. This co-culture illustrated how mutualistic interactions can be exploited to establish robust synthetic co-cultures. C. autoethanogenum produces the substrates for C. kluyveri (acetate and ethanol), at the same time that keeps dissolved CO levels low enough to allow growth of its partner. In addition, it was observed that ethanol production in monocultures of C. autoethanogenum would not have been sufficient to support growth of C. kluyveri. A follow-up study suggested that C. kluyveri enhanced solventogenic metabolism of C. autoethanogenum by removing ethanol from the environment [138]. Gene transcription of the central metabolism of C. autoethanogenum did not change in co-culture compared to monoculture conditions, indicating that the metabolic shift in the presence of C. kluyveri was thermodynamically driven. This is in line with related studies supporting that acetogenesis/ solventogenesis in gas-fermenting microorganisms is controlled at the thermodynamic level [119, 146]. In continuous fermentation, the co-culture of C. autoethanogenum and C. kluyveri produced butyrate and caproate at rates of 0.55 g L
1 day
1 and 0.41 g L
1 day
1, respectively. This work, established as proof-of-concept, was recently picked up by industry in a joint project of the corporations Evonik and Siemens, demonstrating the great potential of syngasfermenting synthetic co-cultures [147]. Richter and colleagues upgraded the synthetic co-culture approach of Diender et al. [137] with a continuous bioprocess that included cell-recycling and in-line product extraction [129]. The co-culture was established with C. kluyveri and C. ljungdahlii, a close relative of C. autoethanogenum with excellent ethanol productivities from syngas [128]. Similar to the co-culture of Diender et al. [137], in this consortium C. ljungdahlii produced acetate and ethanol from syngas, which were used by C. kluyveri to produce longer-chain carboxylates via chain-elongation. In

394

I. Parera Olm and D. Z. Sousa

addition, at a narrow pH range of 5.7–6.4, the elongated carboxylates were reduced by C. ljungdahlii to their respective alcohols, n-butanol, n-hexanol and, for the first time in a syngas-fermenting system, n-octanol was detected (up to 0.78 g L
1 in the condensate of the gas stripping system) [129]. Operation of their bioreactor at pH values higher than 6.4 gradually reduced and eventually crashed the population of C. ljungdahlii, which requires mildly acidic conditions to support growth. Without the acetogen, ethanol production halts and the co-culture crashes. On the other hand, mildly acidic conditions are detrimental to C. kluyveri, which grows in a pH range of 6–7.5 [126, 127]. In addition, acidic pH values result in the accumulation of undissociated acids (pKa  4.7), which are toxic to microorganisms [148]. This discrepancy between optimum pH for solventogenesis and acid production has been reported in similar studies [71, 137]. As Richter and co-workers stated, there is a need to isolate chain-elongating microorganisms with an optimum pH of growth of 5–5.5, a more favourable environment for acetogens to produce ethanol [129]. Alternatively, carboxydotrophic strains could be employed that can thrive at a pH range around neutrality. While the most prominent carboxydotrophic strains thrive in mildly acidic conditions [128], acetogens have been isolated with optimal pH values ranging from 5.4 to 9.8 [27]. An example is Acetobacterium wieringae strain JM, a novel carboxydotroph that grows optimally at pH 7 [112]. The authors of that study speculated that A. wieringae strain JM played a crucial role in syngas-enriched communities by providing substrate to ethanol-consuming propionibacteria, which grow optimally at neutral pH [149]. To test this hypothesis, Moreira et al. [139] cultivated A. wieringae strain JM with Anaerotignum neopropionicum, a propionibacterium unique for its ability to grow on ethanol [150]. The synthetic co-culture was capable of converting CO to propionate (1.78 g L
1) via crossfeeding of ethanol at pH 7. In addition, isovalerate was detected in low amounts in the co-culture, while not in monocultures. Isovalerate could be produced by A. neopropionicum from amino acids; therefore, authors hypothesised that amino acid transfer took place between A. wieringae and A. neopropionicum in co-culture. Interestingly, proteomic analysis of the co-culture revealed sign of stress response in both strains, such as increased abundance of sporulation and antibiotic resistance proteins. It remains a question whether this would negatively affect the stability and functionality of the co-culture in the long-term or, on the contrary, the two populations would eventually come to a beneficial deal.

3.4

Production of Other Value-Added Chemicals

So far, this chapter has discussed various case studies of synthetic co-cultures and open mixed cultures employed in the conversion of syngas/CO to commodity chemicals such as methane, MCCAs and simple alcohols. The production of biochemicals of higher complexity by syngas-consuming cultures is still challenging, for example due to metabolic limitations of the microorganisms or to the requirement of different environmental conditions. A strategy that has been used to overcome this

Conversion of Carbon Monoxide to Chemicals Using Microbial Consortia

395

issue is the introduction of engineered strains of genetically accessible microorganisms. The majority of studies on the use of synthetic co-cultures applied to syngas fermentation have relied on the abilities of wild-type microbial strains [52, 105, 110, 129, 136]. Engineering of Clostridia strains to produce heterologous compounds has advanced rapidly in the last decade, but limitations remain and further understanding of molecular mechanisms is required [37]. Model microorganisms such as Escherichia coli, in contrast, have been widely engineered for the production of a vast range of biochemicals. If a suitable strategy is in place, such strains could be co-cultivated with acetogens to facilitate the production of valuable biochemicals from syngas. Cha and co-workers followed this approach with two co-cultures of the acetogen E. limosum grown with two genetically engineered E. coli strains [140]. In both co-cultures, E. limosum converted CO into acetate, which was used as carbon source by E. coli. The two engineered strains of E. coli used acetate to produce 3-hydroxypropionic acid (3-HP) and itaconic acid (ITA), respectively. At the end of batch cultivations (72 h), the co-cultures produced a maximum of 45.7 mg L
1 3-HP and 25.8 mg L
1 ITA. Consumption of CO increased 10% in co-cultivation compared to monocultures of E. limosum, evidencing that the mutualistic interaction enhanced carbon flux. This study demonstrated for the first time the production of value-added chemicals (3-HP and ITA) from syngas using co-cultures. However, several issues need to be addressed, as noted by the authors. First, the concentration of CO dissolved had to be minimised to allow growth of E. coli, thus a condition of mass-transfer limitation was sought. This was initially achieved by inoculating E. limosum and E. coli at high ratios, up to 150:1 (based on OD600). Over the course of cultivations, though, CO consumption rates and cell concentrations decreased, pointing to the need to improve process stability. Another major problem is that acetate assimilation by E. coli in anaerobic conditions requires the addition of an electron acceptor. Trimethylamine N-oxide (TMAO) was chosen since it yielded the highest acetate assimilation rate and it did not significantly affect CO conversion rates. The fact that TMAO should be supplied proportionally to the desired amount of product would significantly reduce co-culture efficiency and increase process costs, making it unfeasible to implement this strategy at industrial scale. Nevertheless, the work of Cha et al. [140] is a first step towards modular pathway engineering of synthetic co-cultures to facilitate the production of high-value chemicals from syngas.

4 CO Conversion via Sequential Processes Sequential processes can also be used to produce value-added biochemicals from syngas. While these are not one-pot strategies, they may comply with the feature of modularity, characteristic of synthetic cultures. The greatest advantage of this approach compared to one-pot cultures is that it circumvents cultivation divergences between partners in a consortium, by growing each partner in a separate bioreactor

396

I. Parera Olm and D. Z. Sousa

(e.g., combinations aerobic/anaerobic, low/high pH or temperature, etc.). A few proof-of-concept studies have shown the potential of this strategy applied to the conversion of CO-rich gases. Hu et al. [151] designed a two-stage process to produce microbial oil from syngas. In the first reactor (60 C, pH 6), the acetogen Moorella thermoacetica converted CO/H2/CO2 to acetate. Acetate was then fed into the second reactor (28 C, pH 7.3) for its aerobic conversion to C16-C18 triacylglycerides by an engineered strain of the yeast Yarrowia lipolytica. The integrated process produced 18 g L
1 lipids at a rate of 0.19 g L
1 h
1. Acetate as intermediate was also used in the two-step process established by Oswald et al. [91] to produce malic acid. In this case, C. ljungdahlii was first grown in a batch reactor converting syngas to acetate. Subsequently, the reactor was adapted for aerobic cultivation of Aspergillus oryzae, which was inoculated on top of the existing culture. Malic acid was produced to a maximum concentration of 2.02 g L
1. However, the second-stage process was not reproducible in triplicate reactors. Recently, the production of biopolymers has also been demonstrated using sequential processes [152, 153]. In one study, effluent from syngas fermentation by C. autoethanogenum (containing acetate, ethanol and 2,3-BDO) was fed in pulses into a second reactor for the production of polyhydroxyalkanoates (PHAs) [153]. The second reactor contained an enriched mixed culture used in a previous process adapted to PHA production. Only acetate was consumed by the mixed culture, which accumulated a maximum of 24% PHA. Hwang et al. [152] designed a two-stage process for polyhydroxybutyrate (PHB) production differing from the rest in that formate, instead of acetate, was used as intermediate. In the first stage, the acetogen Acetobacterium woodii was used for conversion of syngas under optimised conditions for 100% formate selectivity. The formate solution was concentrated and supplied in fed-batch mode into the second reactor, where formate was converted to PHB by genetically modified Methylobacterium extorquens AM1. All these studies require further improvements, mostly related to medium optimisation in the second reactor. Inadequate composition of ions and certain (toxic) components in the syngas fermentation effluent can have a detrimental effect on the non-acetogenic partner. Nonetheless, these works show that integrated bioprocesses are a feasible platform to convert gaseous substrates to biochemicals of added-value.

5 Challenges and Opportunities of Syngas-Fermenting Microbial Communities By discussing syngas fermentation by mixed cultures, it becomes clear that there are many challenges, but also many opportunities, for future developments in the field. Open mixed cultures are very robust and resilient, offering good prospects for the production of methane and short-chain fatty acids, such as acetate and butyrate. The main challenge with open mixed cultures is product selectivity. A better understanding of microbial compositions and interactions, and the effect of varying process

Conversion of Carbon Monoxide to Chemicals Using Microbial Consortia

397

parameters, is necessary. Knowledge on complex microbial communities converting syngas may also source inspiration for the creation of synthetic co-cultures as recently exemplified by Moreira et al. [139], where a co-culture producing propionate was constructed based on the microbial composition of an enriched culture. Compared to the enriched culture, the co-culture produced higher amounts of propionate, and side products (like methane) were eliminated. Open mixed cultures and laboratory enrichments may also lead to the isolation of new microorganisms, carboxydotrophs or others, with better characteristics for the construction of synthetic co-cultures. For example, Richter et al. [129] observed suboptimal performance of a co-culture of C. ljungdahlii and C. kluyveri due to a mismatch in the optimal pH of the two species. Isolation of (ethanol-driven) chain-elongators with lower optimal pH for growth would be useful for pairing with solventogenic acetogens during syngas fermentation. Such microorganisms are currently not available in culture collections. Currently, there are also only a limited amount of thermophilic carboxydotrophs isolated, and most of them exhibit a hydrogenogenic metabolism. Studying high-temperature adapted microbiomes (e.g., thermophilic anaerobic sludges, hydrothermal vents, etc.) could lead to discovering novel microbes and metabolisms. Thermophilic organisms could be used to produce volatile compounds, allowing their separation in the gas-phase and reducing streaming costs. Other environments, such as high salinity sediments, are also not well studied in regard to their potential to convert CO/syngas [154]. The first steps for co-cultivation of microbes for syngas fermentation are taken. Now, work can be done in two fronts: improvement of current co-cultures for the production of, e.g., MCFA and alcohols (higher titres, higher yields, higher carbon fixation, etc.) or the development of new co-cultures for the diversification of products. The improvement of co-culture systems can be aided by genome-based models (GEMs). These models describe the set of possible reactions by the microbes in the co-culture (based on their genomic content), including extracellular exchange of metabolites [131, 134, 155–157]. The GEM constructed by Benito-Vaquerizo et al. [131] describes growth of the syngas-converting co-culture comprising C. autoethanogenum and C. kluyveri, and predicts that succinate addition would improve the production of MCFAs. Experimental testing needs to be conducted to ascertain this, but this is an example of how GEMs can aid in the generation of hypothesis and eventually result in accelerated optimisation of co-cultures. Computational models can also be used to predict novel microbial interactions. In a recent study, Li and Henson [134] performed in silico simulations on 170 combinations of acetogen and butyrate-producing bacteria pairings. This led to the discovery of highly performing co-culture designs for syngas fermentation that could guide future experimental studies. Yet, reconstruction of GEMs, and especially their manual curation and experimental validation is still a time-consuming procedure, and applications of GEMs to co-cultivation are so far scarce [158]. Co-cultures are also suitable for the introduction of genetically engineered strains, e.g. to supress/overexpress the expression of certain genes [159], engineer symbiosis [160], create ‘artificial’ division of labour [161] or control populations of different strains [162]. Regarding syngas fermentations, up to date only wild-type acetogens

398

I. Parera Olm and D. Z. Sousa

have been used to establish synthetic co-cultures but this could change soon with the recent developments on genetic engineering of autotrophic Clostridia strains able to convert CO [34, 35]. Carboxydotrophy can also be engineered in solventogenic Clostridia, as shown by heterologous expression of a carbon monoxide dehydrogenase in Clostridium acetobutylicum [163]. The urge to produce high-value chemicals from syngas is also putting the focus on engineering pathways for the assimilation of one-carbon compounds (e.g., glycine pathway) in E. coli strains that can natively produce value-added chemicals [164, 165].

6 Conclusion Microbial communities have a tremendous potential in syngas fermentation processes, broadening the product spectrum beyond acetate and ethanol. Open mixed cultures are sustained by decades of research and industrial experience on the field of anaerobic digestion (of wastes/wastewaters), which can be transferred to the conversion of CO-rich gases to methane, MCCAs and alcohols. Synthetic co-cultures can enhance product selectivity, offer modularity and allow the use of kinetic and genome-scale metabolic models for further optimisation. Recently, genetically engineered strains of model organisms have been co-cultured with acetogens, enabling the production of added-value chemicals from C1 substrates. Industrial implementation of syngas-fermenting microbial consortia for the production of valuable biochemicals is not yet a reality. However, growing interest on the utilisation of C1-gases nurtured by major efforts undertaken over the last few years might certainly drive this platform forward faster than anticipated. Acknowledgements The authors would like to acknowledge the Netherlands Science Foundation (NWO) (Project NWO-GK-07), the NWO Applied and Engineering Sciences (AGS) (Perspectief Programma P16-10), and the Netherlands Ministry of Education, Culture and Science (Project 024.002.002) for financial support.

References 1. European Commission (2018) A clean planet for all – a European long-term strategic vision for a prosperous, modern, competitive and climate neutral economy (COM(2018) 773). https:// eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri¼CELEX:52018DC0773&from¼EN 2. European Commission (2019) The European Green Deal (COM(2019) 640). https://eur-lex. europa.eu/legal-content/EN/TXT/PDF/?uri¼CELEX:52019DC0640&from¼EN 3. Cherubini F (2010) The biorefinery concept: using biomass instead of oil for producing energy and chemicals. Energ Conver Manage 51(7):1412–1421. https://doi.org/10.1016/j.enconman. 2010.01.015 4. Popp J, Lakner Z, Harangi-Rákos M, Fári M (2014) The effect of bioenergy expansion: food, energy, and environment. Renew Sustain Energy Rev 32:559–578. https://doi.org/10.1016/j. rser.2014.01.056

Conversion of Carbon Monoxide to Chemicals Using Microbial Consortia

399

5. Tripathi N, Hills CD, Singh RS, Atkinson CJ (2019) Biomass waste utilisation in low-carbon products: harnessing a major potential resource. NPJ Clim Atmos Sci 2(1):35. https://doi.org/ 10.1038/s41612-019-0093-5 6. Kawaguchi H, Hasunuma T, Ogino C, Kondo A (2016) Bioprocessing of bio-based chemicals produced from lignocellulosic feedstocks. Curr Opin Biotechnol 42:30–39. https://doi.org/10. 1016/j.copbio.2016.02.031 7. Sims REH, Mabee W, Saddler JN, Taylor M (2010) An overview of second generation biofuel technologies. Bioresour Technol 101(6):1570–1580. https://doi.org/10.1016/j.biortech.2009. 11.046 8. Daniell J, Köpke M, Simpson S (2012) Commercial biomass syngas fermentation. Energies 5(12):5372–5417. https://doi.org/10.3390/en5125372 9. Griffin DW, Schultz MA (2012) Fuel and chemical products from biomass syngas: a comparison of gas fermentation to thermochemical conversion routes. Environ Prog Sustain Energy 31(2):219–224. https://doi.org/10.1002/ep.11613 10. Wood DA, Nwaoha C, Towler BF (2012) Gas-to-liquids (GTL): a review of an industry offering several routes for monetizing natural gas. J Nat Gas Sci Eng 9:196–208. https://doi. org/10.1016/j.jngse.2012.07.001 11. Christensen JM, Jensen PA, Jensen AD (2011) Effects of feed composition and feed impurities in the catalytic conversion of syngas to higher alcohols over alkali-promoted cobalt–molybdenum sulfide. Ind Eng Chem Res 50(13):7949–7963. https://doi.org/10.1021/ie200235e 12. Bambal AS, Guggilla VS, Kugler EL, Gardner TH, Dadyburjor DB (2014) Poisoning of a silica-supported cobalt catalyst due to presence of sulfur impurities in syngas during Fischer– Tropsch synthesis: effects of chelating agent. Ind Eng Chem Res 53(14):5846–5857. https:// doi.org/10.1021/ie500243h 13. Balonek CM, Lillebø AH, Rane S, Rytter E, Schmidt LD, Holmen A (2010) Effect of alkali metal impurities on Co–Re catalysts for Fischer–Tropsch synthesis from biomass-derived syngas. Catal Lett 138(1–2):8–13. https://doi.org/10.1007/s10562-010-0366-4 14. Abdoulmoumine N, Adhikari S, Kulkarni A, Chattanathan S (2015) A review on biomass gasification syngas cleanup. Appl Energy 155:294–307. https://doi.org/10.1016/j.apenergy. 2015.05.095 15. Abubackar HN, Veiga MC, Kennes C (2011) Biological conversion of carbon monoxide: rich syngas or waste gases to bioethanol. Biofuels Bioprod Biorefin 5(1):93–114. https://doi.org/ 10.1002/bbb.256 16. Liew F, Martin ME, Tappel RC, Heijstra BD, Mihalcea C, Köpke M (2016) Gas fermentation – a flexible platform for commercial scale production of low-carbon-fuels and chemicals from waste and renewable feedstocks. Front Microbiol 7. https://doi.org/10.3389/fmicb.2016.00694 17. Redl S, Diender M, Jensen TØ, Sousa DZ, Nielsen AT (2017) Exploiting the potential of gas fermentation. Ind Crop Prod 106:21–30. https://doi.org/10.1016/j.indcrop.2016.11.015 18. Wei L, Pordesimo LO, Igathinathane C, Batchelor WD (2009) Process engineering evaluation of ethanol production from wood through bioprocessing and chemical catalysis. Biomass Bioenergy 33(2):255–266. https://doi.org/10.1016/j.biombioe.2008.05.017 19. World Steel Association (2020) Sustainability indicators 2003 to 2019 and 2019 participation details. https://www.worldsteel.org/steel-by-topic/sustainability/sustainability-indicators.html 20. Ma S, Luo R, Gold JI, Yu AZ, Kim B, Kenis PJA (2016) Carbon nanotube containing Ag catalyst layers for efficient and selective reduction of carbon dioxide. J Mater Chem A 4(22):8573–8578. https://doi.org/10.1039/C6TA00427J 21. Messias S, Sousa MM, Nunes da Ponte M, Rangel CM, Pardal T, Reis Machado AS (2019) Electrochemical production of syngas from CO2 at pressures up to 30 bar in electrolytes containing ionic liquid. React Chem Eng 4(11):1982–1990. https://doi.org/10.1039/ C9RE00271E 22. Verma S et al (2018) Insights into the low overpotential electroreduction of CO2 to CO on a supported gold catalyst in an alkaline flow electrolyzer. ACS Energy Lett 3(1):193–198. https://doi.org/10.1021/acsenergylett.7b01096

400

I. Parera Olm and D. Z. Sousa

23. King GM, Weber CF (2007) Distribution, diversity and ecology of aerobic CO-oxidizing bacteria. Nat Rev Microbiol 5(2):107–118. https://doi.org/10.1038/nrmicro1595 24. Diender M, Stams AJM, Sousa DZ (2015) Pathways and bioenergetics of anaerobic carbon monoxide fermentation. Front Microbiol 6:1–18. https://doi.org/10.3389/fmicb.2015.01275 25. Drake HL, Gössner AS, Daniel SL (2008) Old acetogens, new light. Ann N Y Acad Sci 1125(1):100–128. https://doi.org/10.1196/annals.1419.016 26. Schuchmann K, Müller V (2014) Autotrophy at the thermodynamic limit of life: a model for energy conservation in acetogenic bacteria. Nat Rev Microbiol 12(12):809–821. https://doi. org/10.1038/nrmicro3365 27. Bengelsdorf FR et al (2018) Bacterial anaerobic synthesis gas (syngas) and CO2 + H2 fermentation. Adv Appl Microbiol:143–221. https://doi.org/10.1016/bs.aambs.2018.01.002 28. Sipma J, Henstra AM, Parshina SN, Lens PNL, Lettinga G, Stams AJM (2006) Microbial CO conversions with applications in synthesis gas purification and bio-desulfurization. Crit Rev Biotechnol 26(1):41–65. https://doi.org/10.1080/07388550500513974 29. Sun X, Atiyeh HK, Huhnke RL, Tanner RS (2019) Syngas fermentation process development for production of biofuels and chemicals: a review. Bioresour Technol Rep 7:100279. https:// doi.org/10.1016/j.biteb.2019.100279 30. Wainaina S, Horváth IS, Taherzadeh MJ (2018) Biochemicals from food waste and recalcitrant biomass via syngas fermentation: a review. Bioresour Technol 248(Pt A):113–121. https://doi. org/10.1016/j.biortech.2017.06.075 31. Gaddy JL, Arora DK, Ko CW, Phillips JR, Basu R, Wikstrom CV, Clausen EC (2007) Methods for increasing the production of ethanol from microbial fermentation. Patent US7285402 32. Phillips JR, Klasson KT, Clausen EC, Gaddy JL (1993) Biological production of ethanol from coal synthesis gas. Appl Biochem Biotechnol 39–40(1):559–571. https://doi.org/10.1007/ BF02919018 33. Katsyv A, Müller V (2020) Overcoming energetic barriers in acetogenic C1 conversion. Front Bioeng Biotechnol 8:621166. https://doi.org/10.3389/fbioe.2020.621166 34. Humphreys CM, Minton NP (2018) Advances in metabolic engineering in the microbial production of fuels and chemicals from C1 gas. Curr Opin Biotechnol 50:174–181. https:// doi.org/10.1016/j.copbio.2017.12.023 35. Joseph RC, Kim NM, Sandoval NR (2018) Recent developments of the synthetic biology toolkit for Clostridium. Front Microbiol 9:154. https://doi.org/10.3389/fmicb.2018.00154 36. Köpke M, Simpson SD (2020) Pollution to products: recycling of “above ground” carbon by gas fermentation. Curr Opin Biotechnol 65:180–189. https://doi.org/10.1016/j.copbio.2020. 02.017 37. Charubin K, Bennett RK, Fast AG, Papoutsakis ET (2018) Engineering Clostridium organisms as microbial cell-factories: challenges & opportunities. Metab Eng 50:173–191. https:// doi.org/10.1016/j.ymben.2018.07.012 38. Couto N, Rouboa A, Silva V, Monteiro E, Bouziane K (2013) Influence of the biomass gasification processes on the final composition of syngas. Energy Procedia 36:596–606. https://doi.org/10.1016/j.egypro.2013.07.068 39. Esquivel-Elizondo S, Delgado AG, Rittmann BE, Krajmalnik-Brown R (2017) The effects of CO2 and H2 on CO metabolism by pure and mixed microbial cultures. Biotechnol Biofuels 10(1):220. https://doi.org/10.1186/s13068-017-0910-1 40. Xu D, Tree DR, Lewis RS (2011) The effects of syngas impurities on syngas fermentation to liquid fuels. Biomass Bioenergy 35(7):2690–2696. https://doi.org/10.1016/j.biombioe.2011. 03.005 41. Cui Y, Yang K-L, Zhou K (2020) Using co-culture to functionalize Clostridium fermentation. Trends Biotechnol 39(9):914–926. https://doi.org/10.1016/j.tibtech.2020.11.016 42. Diender M, Parera Olm I, Sousa DZ (2021) Synthetic co-cultures: novel avenues for bio-based processes. Curr Opin Biotechnol 67:72–79. https://doi.org/10.1016/j.copbio.2021.01.006

Conversion of Carbon Monoxide to Chemicals Using Microbial Consortia

401

43. West SA, Diggle SP, Buckling A, Gardner A, Griffin AS (2007) The social lives of microbes. Annu Rev Ecol Evol Syst 38(1):53–77. https://doi.org/10.1146/annurev.ecolsys.38.091206. 095740 44. Zengler K, Zaramela LS (2018) The social network of microorganisms – how auxotrophies shape complex communities. Nat Rev Microbiol 16(6):383–390. https://doi.org/10.1038/ s41579-018-0004-5 45. Bader J, Mast-Gerlach E, Popović MK, Bajpai R, Stahl U (2010) Relevance of microbial coculture fermentations in biotechnology. J Appl Microbiol 109(2):371–387. https://doi.org/ 10.1111/j.1365-2672.2009.04659.x 46. Brenner K, You L, Arnold FH (2008) Engineering microbial consortia: a new frontier in synthetic biology. Trends Biotechnol 26(9):483–489. https://doi.org/10.1016/j.tibtech.2008. 05.004 47. Johns NI, Blazejewski T, Gomes AL, Wang HH (2016) Principles for designing synthetic microbial communities. Curr Opin Microbiol 31:146–153. https://doi.org/10.1016/j.mib.2016. 03.010 48. Lawson CE et al (2019) Common principles and best practices for engineering microbiomes. Nat Rev Microbiol 17(12):725–741. https://doi.org/10.1038/s41579-019-0255-9 49. McCarty NS, Ledesma-Amaro R (2019) Synthetic biology tools to engineer microbial communities for biotechnology. Trends Biotechnol 37(2):181–197. https://doi.org/10.1016/j. tibtech.2018.11.002 50. Strous M, Sharp C (2018) Designer microbiomes for environmental, energy and health biotechnology. Curr Opin Microbiol 43:117–123. https://doi.org/10.1016/j.mib.2017.12.007 51. Balk M, Mehboob F, van Gelder AH, Rijpstra WIC, Damsté JSS, Stams AJM (2010) (Per)chlorate reduction by an acetogenic bacterium, Sporomusa sp., isolated from an underground gas storage. Appl Microbiol Biotechnol 88(2):595–603. https://doi.org/10.1007/s00253-0102788-8 52. Lee CR et al (2018) Co-culture-based biological carbon monoxide conversion by Citrobacter amalonaticus Y19 and Sporomusa ovata via a reducing-equivalent transfer mediator. Bioresour Technol 259:128–135. https://doi.org/10.1016/j.biortech.2018.02.129 53. Yeol Jung G, Rae Kim J, Ok Jung H, Park J-Y, Park S (1999) A new chemoheterotrophic bacterium catalyzing water-gas shift reaction. Biotechnol Lett 21(10):869–873. https://doi.org/ 10.1023/a:1005599600510 54. Morris JJ, Lenski RE, Zinser ER (2012) The Black Queen hypothesis: evolution of dependencies through adaptive gene loss. MBio 3(2):e00036–e00012. https://doi.org/10.1128/ mBio.00036-12 55. Ben Said S, Or D (2017) Synthetic microbial ecology: engineering habitats for modular consortia. Front Microbiol 8:1125. https://doi.org/10.3389/fmicb.2017.01125 56. De Vrieze J, Christiaens MER, Verstraete W (2017) The microbiome as engineering tool: manufacturing and trading between microorganisms. N Biotechnol 39:206–214. https://doi. org/10.1016/j.nbt.2017.07.001 57. Goers L, Freemont P, Polizzi KM (2014) Co-culture systems and technologies: taking synthetic biology to the next level. J R Soc Interface 11(96):20140065. https://doi.org/10. 1098/rsif.2014.0065 58. Baleeiro FCF, Kleinsteuber S, Neumann A, Sträuber H (2019) Syngas-aided anaerobic fermentation for medium-chain carboxylate and alcohol production: the case for microbial communities. Appl Microbiol Biotechnol 103(21–22):8689–8709. https://doi.org/10.1007/ s00253-019-10086-9 59. Asimakopoulos K et al (2020) Temperature effects on syngas biomethanation performed in a trickle bed reactor. Chem Eng J 393:124739. https://doi.org/10.1016/j.cej.2020.124739 60. Asimakopoulos K et al (2021) Scale up study of a thermophilic trickle bed reactor performing syngas biomethanation. Appl Energy 290:116771. https://doi.org/10.1016/j.apenergy.2021. 116771

402

I. Parera Olm and D. Z. Sousa

61. Asimakopoulos K, Grimalt-Alemany A, Lundholm-Høffner C, Gavala HN, Skiadas IV (2021) Carbon sequestration through syngas biomethanation coupled with H2 supply for a clean production of natural gas grade biomethane. Waste Biomass Valori. https://doi.org/10.1007/ s12649-021-01393-2 62. Chandolias K, Pekgenc E, Taherzadeh M (2019) Floating membrane bioreactors with high gas hold-up for syngas-to-biomethane conversion. Energies 12(6):1046. https://doi.org/10.3390/ en12061046 63. Pereira FM (2014) Intensified bioprocess for the anaerobic conversion of syngas to biofuels. University of Minho. http://hdl.handle.net/1822/34834 64. Shen N, Dai K, Xia X-Y, Zeng RJ, Zhang F (2018) Conversion of syngas (CO and H2) to biochemicals by mixed culture fermentation in mesophilic and thermophilic hollow-fiber membrane biofilm reactors. J Clean Prod 202:536–542. https://doi.org/10.1016/j.jclepro. 2018.08.162 65. He P, Han W, Shao L, Lü F (2018) One-step production of C6–C8 carboxylates by mixed culture solely grown on CO. Biotechnol Biofuels 11(1):4. https://doi.org/10.1186/s13068017-1005-8 66. Wang H-J, Dai K, Wang Y-Q, Wang H-F, Zhang F, Zeng RJ (2018) Mixed culture fermentation of synthesis gas in the microfiltration and ultrafiltration hollow-fiber membrane biofilm reactors. Bioresour Technol 267:650–656. https://doi.org/10.1016/j.biortech.2018.07.098 67. Wang H-J, Dai K, Xia X-Y, Wang Y-Q, Zeng RJ, Zhang F (2018) Tunable production of ethanol and acetate from synthesis gas by mesophilic mixed culture fermentation in a hollow fiber membrane biofilm reactor. J Clean Prod 187:165–170. https://doi.org/10.1016/j.jclepro. 2018.03.193 68. Wang P, Wang H, Qiu Y, Ren L, Jiang B (2018) Microbial characteristics in anaerobic digestion process of food waste for methane production – a review. Bioresour Technol 248 (Pt A):29–36. https://doi.org/10.1016/j.biortech.2017.06.152 69. Liu K, Atiyeh HK, Stevenson BS, Tanner RS, Wilkins MR, Huhnke RL (2014) Continuous syngas fermentation for the production of ethanol, n-propanol and n-butanol. Bioresour Technol 151:69–77. https://doi.org/10.1016/j.biortech.2013.10.059 70. Ganigué R, Ramió-Pujol S, Sánchez P, Bañeras L, Colprim J (2015) Conversion of sewage sludge to commodity chemicals via syngas fermentation. Water Sci Technol 72(3):415–420. https://doi.org/10.2166/wst.2015.222 71. Ganigué R, Sánchez-Paredes P, Bañeras L, Colprim J (2016) Low fermentation pH is a trigger to alcohol production, but a killer to chain elongation. Front Microbiol 7:1–11. https://doi.org/ 10.3389/fmicb.2016.00702 72. Chakraborty S, Rene ER, Lens PNL, Veiga MC, Kennes C (2019) Enrichment of a solventogenic anaerobic sludge converting carbon monoxide and syngas into acids and alcohols. Bioresour Technol 272:130–136. https://doi.org/10.1016/j.biortech.2018.10.002 73. Alves JI, Stams AJM, Plugge CM, Madalena Alves M, Sousa DZ (2013) Enrichment of anaerobic syngas-converting bacteria from thermophilic bioreactor sludge. FEMS Microbiol Ecol 86(3):590–597. https://doi.org/10.1111/1574-6941.12185 74. Arantes AL, Alves JI, Stams AJM, Alves MM, Sousa DZ (2018) Enrichment of syngasconverting communities from a multi-orifice baffled bioreactor. J Microbial Biotechnol 11(4):639–646. https://doi.org/10.1111/1751-7915.12864 75. Esquivel-Elizondo S, Delgado AG, Krajmalnik-Brown R (2017) Evolution of microbial communities growing with carbon monoxide, hydrogen, and carbon dioxide. FEMS Microbiol Ecol 93(6). https://doi.org/10.1093/femsec/fix076 76. Sipma J, Lens PNL, Stams AJM, Lettinga G (2003) Carbon monoxide conversion by anaerobic bioreactor sludges. FEMS Microbiol Ecol 44(2):271–277. https://doi.org/10.1016/S01686496(03)00033-3 77. Guiot SR, Cimpoia R, Carayon G (2011) Potential of wastewater-treating anaerobic granules for biomethanation of synthesis gas. Environ Sci Technol 45(5):2006–2012. https://doi.org/10. 1021/es102728m

Conversion of Carbon Monoxide to Chemicals Using Microbial Consortia

403

78. Li C, Zhu X, Angelidaki I (2020) Carbon monoxide conversion and syngas biomethanation mediated by different microbial consortia. Bioresour Technol 314:123739. https://doi.org/10. 1016/j.biortech.2020.123739 79. Park S et al (2013) Rapid enrichment of (homo)acetogenic consortia from animal feces using a high mass-transfer gas-lift reactor fed with syngas. J Ind Microbiol Biotechnol 40(9):995–1003. https://doi.org/10.1007/s10295-013-1292-4 80. Singla A, Verma D, Lal B, Sarma PM (2014) Enrichment and optimization of anaerobic bacterial mixed culture for conversion of syngas to ethanol. Bioresour Technol 172:41–49. https://doi.org/10.1016/j.biortech.2014.08.083 81. Tang Y, Huang Y, Gan W, Xia A, Liao Q, Zhu X (2021) Ethanol production from gas fermentation: rapid enrichment and domestication of bacterial community with continuous CO/CO2 gas. Renew Energy 175:337–344. https://doi.org/10.1016/j.renene.2021.04.134 82. Xu S, Fu B, Zhang L, Liu H (2015) Bioconversion of H2/CO2 by acetogen enriched cultures for acetate and ethanol production: the impact of pH. World J Microbiol Biotechnol 31(6):941–950. https://doi.org/10.1007/s11274-015-1848-8 83. Duan H, He P, Shao L, Lü F (2021) Functional genome-centric view of the CO-driven anaerobic microbiome. ISME J. https://doi.org/10.1038/s41396-021-00983-1 84. Grimalt-Alemany A, Łężyk M, Lange L, Skiadas IV, Gavala HN (2018) Enrichment of syngas-converting mixed microbial consortia for ethanol production and thermodynamicsbased design of enrichment strategies. Biotechnol Biofuels 11(1):198. https://doi.org/10.1186/ s13068-018-1189-6 85. Grimalt-Alemany A, Skiadas IV, Gavala HN (2018) Syngas biomethanation: state-of-the-art review and perspectives. Biofuels Bioprod Biorefin 12(1):139–158. https://doi.org/10.1002/ bbb.1826 86. Adams MWW (1990) The structure and mechanism of iron-hydrogenases. Biochim Biophys Acta 1020(2):115–145. https://doi.org/10.1016/0005-2728(90)90044-5 87. Chen J-S, Blanchard DK (1978) Isolation and properties of a unidirectional H2-oxidizing hydrogenase from the strictly anaerobic N2-fixing bacterium Clostridium pasteurianum W5. Biochem Biophys Res Commun 84(4):1144–1150. https://doi.org/10.1016/0006-291X(78) 91703-5 88. Fox JD, Kerby RL, Roberts GP, Ludden PW (1996) Characterization of the CO-induced, CO-tolerant hydrogenase from Rhodospirillum rubrum and the gene encoding the large subunit of the enzyme. J Bacteriol 178(6):1515–1524. https://doi.org/10.1128/JB.178.6. 1515-1524.1996 89. Hess V, Oyrik O, Trifunović D, Müller V (2015) 2,3-Butanediol metabolism in the acetogen Acetobacterium woodii. Appl Environ Microbiol 81(14):4711–4719. https://doi.org/10.1128/ AEM.00960-15 90. Kim BH, Bellows P, Datta R, Zeikus JG (1984) Control of carbon and electron flow in Clostridium acetobutylicum fermentations: utilization of carbon monoxide to inhibit hydrogen production and to enhance butanol yields. Appl Environ Microbiol 48(4):764–770. https://doi. org/10.1128/AEM.48.4.764-770.1984 91. Oswald F et al (2016) Sequential mixed cultures: from syngas to malic acid. Front Microbiol 7: 891. https://doi.org/10.3389/fmicb.2016.00891 92. Grimalt-Alemany A, Łężyk M, Kennes-Veiga DM, Skiadas IV, Gavala HN (2019) Enrichment of mesophilic and thermophilic mixed microbial consortia for syngas biomethanation: the role of kinetic and thermodynamic competition. Waste Biomass Valori 11(2):465–481. https://doi.org/10.1007/s12649-019-00595-z 93. Sancho Navarro S, Cimpoia R, Bruant G, Guiot SR (2016) Biomethanation of syngas using anaerobic sludge: shift in the catabolic routes with the CO partial pressure increase. Front Microbiol 7:1188. https://doi.org/10.3389/fmicb.2016.01188 94. Pereira FM, Alves MM, Sousa DZ (2013) Effect of pH and pressure on syngas fermentation by anaerobic mixed cultures. International Water Association. http://hdl.handle.net/1822/26374

404

I. Parera Olm and D. Z. Sousa

95. Steinbusch KJJ, Arvaniti E, Hamelers HVM, Buisman CJN (2009) Selective inhibition of methanogenesis to enhance ethanol and n-butyrate production through acetate reduction in mixed culture fermentation. Bioresour Technol 100(13):3261–3267. https://doi.org/10.1016/j. biortech.2009.01.049 96. Nam CW, Jung KA, Park JM (2016) Biological carbon monoxide conversion to acetate production by mixed culture. Bioresour Technol 211:478–485. https://doi.org/10.1016/j. biortech.2016.03.100 97. Luo G, Jing Y, Lin Y, Zhang S, An D (2018) A novel concept for syngas biomethanation by two-stage process: focusing on the selective conversion of syngas to acetate. Sci Total Environ 645:1194–1200. https://doi.org/10.1016/j.scitotenv.2018.07.263 98. Zinder SH, Anguish T, Cardwell SC (1984) Selective inhibition by 2-bromoethanesulfonate of methanogenesis from acetate in a thermophilic anaerobic digestor. Appl Environ Microbiol 47(6):1343–1345. https://doi.org/10.1128/AEM.47.6.1343-1345.1984 99. Chiu PC, Lee M (2001) 2-Bromoethanesulfonate affects bacteria in a trichloroethenedechlorinating culture. Appl Environ Microbiol 67(5):2371–2374. https://doi.org/10.1128/ AEM.67.5.2371-2374.2001 100. Grimalt-Alemany A, Łężyk M, Asimakopoulos K, Skiadas IV, Gavala HN (2020) Cryopreservation and fast recovery of enriched syngas-converting microbial communities. Water Res 177:115747. https://doi.org/10.1016/j.watres.2020.115747 101. Amani T, Nosrati M, Sreekrishnan TR (2010) Anaerobic digestion from the viewpoint of microbiological, chemical, and operational aspects – a review. Environ Rev 18:255–278. https://doi.org/10.1139/A10-011 102. Leng L et al (2018) A review on the bioenergetics of anaerobic microbial metabolism close to the thermodynamic limits and its implications for digestion applications. Bioresour Technol 247:1095–1106. https://doi.org/10.1016/j.biortech.2017.09.103 103. Novaes RFV (1986) Microbiology of anaerobic digestion. Water Sci Technol 18(12):1–14. https://doi.org/10.2166/wst.1986.0159 104. Gavala HN, Grimalt-Alemany A, Asimakopoulos K, Skiadas IV (2021) Gas biological conversions: the potential of syngas and carbon dioxide as production platforms. Waste Biomass Valori. https://doi.org/10.1007/s12649-020-01332-7 105. Diender M, Uhl PS, Bitter JH, Stams AJMM, Sousa DZ (2018) High rate biomethanation of carbon monoxide-rich gases via a thermophilic synthetic co-culture. ACS Sustain Chem Eng 6(2):2169–2176. https://doi.org/10.1021/acssuschemeng.7b03601 106. Daniels L, Fuchs G, Thauer RK, Zeikus JG (1977) Carbon monoxide oxidation by methanogenic bacteria. J Bacteriol 132(1):118–126. https://doi.org/10.1128/JB.132.1. 118-126.1977 107. O’Brien JM, Wolkin RH, Moench TT, Morgan JB, Zeikus JG (1984) Association of hydrogen metabolism with unitrophic or mixotrophic growth of Methanosarcina barkeri on carbon monoxide. J Bacteriol 158(1):373–375. https://doi.org/10.1128/JB.158.1.373-375.1984 108. Kliefoth M, Langer JD, Matschiavelli N, Oelgeschläger E, Rother M (2012) Genetic analysis of MA4079, an aldehyde dehydrogenase homolog, in Methanosarcina acetivorans. Arch Microbiol 194(2):75–85. https://doi.org/10.1007/s00203-011-0727-4 109. Rother M, Metcalf WW (2004) Anaerobic growth of Methanosarcina acetivorans C2A on carbon monoxide: an unusual way of life for a methanogenic archaeon. Proc Natl Acad Sci 101(48):16929–16934. https://doi.org/10.1073/pnas.0407486101 110. Diender M, Pereira R, Wessels HJCT, Stams AJM, Sousa DZ (2016) Proteomic analysis of the hydrogen and carbon monoxide metabolism of Methanothermobacter marburgensis. Front Microbiol 7:1049. https://doi.org/10.3389/fmicb.2016.01049 111. Luo G, Wang W, Angelidaki I (2013) Anaerobic digestion for simultaneous sewage sludge treatment and CO biomethanation: process performance and microbial ecology. Environ Sci Technol 47(18):130904143045005. https://doi.org/10.1021/es401018d 112. Arantes AL et al (2020) Enrichment of anaerobic syngas-converting communities and isolation of a novel carboxydotrophic Acetobacterium wieringae strain JM. Front Microbiol 11:58. https://doi.org/10.3389/fmicb.2020.00058

Conversion of Carbon Monoxide to Chemicals Using Microbial Consortia

405

113. Teixeira LV, Moutinho LF, Romão-Dumaresq AS (2018) Gas fermentation of C1 feedstocks: commercialization status and future prospects. Biofuels Bioprod Biorefin 12(6):1103–1117. https://doi.org/10.1002/bbb.1912 114. Ueki A, Goto K, Ohtaki Y, Kaku N, Ueki K (2017) Description of Anaerotignum aminivorans gen. nov., sp. nov., a strictly anaerobic, amino-acid-decomposing bacterium isolated from a methanogenic reactor, and reclassification of Clostridium propionicum, Clostridium neopropionicum and Clostridium lactatifermentans. Int J Syst Evol Microbiol 67(10):4146–4153. https://doi.org/10.1099/ijsem.0.002268 115. Liu K, Atiyeh HK, Stevenson BS, Tanner RS, Wilkins MR, Huhnke RL (2014) Mixed culture syngas fermentation and conversion of carboxylic acids into alcohols. Bioresour Technol 152: 337–346. https://doi.org/10.1016/j.biortech.2013.11.015 116. Asimakopoulos K, Gavala HN, Skiadas IV (2018) Reactor systems for syngas fermentation processes: a review. Chem Eng J 348:732–744. https://doi.org/10.1016/j.cej.2018.05.003 117. Shen Y, Brown R, Wen Z (2014) Syngas fermentation of Clostridium carboxidivorans P7 in a hollow fiber membrane biofilm reactor: evaluating the mass transfer coefficient and ethanol production performance. Biochem Eng J 85:21–29. https://doi.org/10.1016/j.bej.2014.01.010 118. Martin KJ, Nerenberg R (2012) The membrane biofilm reactor (MBfR) for water and wastewater treatment: principles, applications, and recent developments. Bioresour Technol 122:83–94. https://doi.org/10.1016/j.biortech.2012.02.110 119. Back H et al (2016) Achieving long-term stable perovskite solar cells via ion neutralization. Energ Environ Sci 9(4):1258–1263. https://doi.org/10.1039/C6EE00612D 120. Zhang F et al (2013) In situ hydrogen utilization for high fraction acetate production in mixed culture hollow-fiber membrane biofilm reactor. Appl Microbiol Biotechnol 97(23):10233–10240. https://doi.org/10.1007/s00253-013-5281-3 121. Agler MT, Wrenn BA, Zinder SH, Angenent LT (2011) Waste to bioproduct conversion with undefined mixed cultures: the carboxylate platform. Trends Biotechnol 29(2):70–78. https:// doi.org/10.1016/j.tibtech.2010.11.006 122. Angenent LT et al (2016) Chain elongation with reactor microbiomes: open-culture biotechnology to produce biochemicals. Environ Sci Technol 50(6):2796–2810. https://doi.org/10. 1021/acs.est.5b04847 123. Moscoviz R, Trably E, Bernet N, Carrère H (2018) The environmental biorefinery: state-ofthe-art on the production of hydrogen and value-added biomolecules in mixed-culture fermentation. Green Chem 20(14):3159–3179. https://doi.org/10.1039/C8GC00572A 124. Spirito CM, Richter H, Rabaey K, Stams AJ, Angenent LT (2014) Chain elongation in anaerobic reactor microbiomes to recover resources from waste. Curr Opin Biotechnol 27: 115–122. https://doi.org/10.1016/j.copbio.2014.01.003 125. Candry P, Huang S, Carvajal-Arroyo JM, Rabaey K, Ganigué R (2020) Enrichment and characterisation of ethanol chain elongating communities from natural and engineered environments. Sci Rep 10(1):3682. https://doi.org/10.1038/s41598-020-60052-z 126. Barker HA, Taha SM (1942) Clostridium kluyverii, an organism concerned in the formation of caproic acid from ethyl alcohol. J Bacteriol 43(3):347–363. https://doi.org/10.1128/JB.43.3. 347-363.1942 127. Weimer PJ, Stevenson DM (2012) Isolation, characterization, and quantification of Clostridium kluyveri from the bovine rumen. Appl Microbiol Biotechnol 94(2):461–466. https://doi. org/10.1007/s00253-011-3751-z 128. Martin ME, Richter H, Saha S, Angenent LT (2016) Traits of selected Clostridium strains for syngas fermentation to ethanol. Biotechnol Bioeng 113(3):531–539. https://doi.org/10.1002/ bit.25827 129. Richter H, Molitor B, Diender M, Sousa DZ, Angenent LT (2016) A narrow pH range supports butanol, hexanol, and octanol production from syngas in a continuous so-culture of Clostridium ljungdahlii and Clostridium kluyveri with in-line product extraction. Front Microbiol 7: 1773. https://doi.org/10.3389/fmicb.2016.01773

406

I. Parera Olm and D. Z. Sousa

130. Straathof AJJ, Wahl SA, Benjamin KR, Takors R, Wierckx N, Noorman HJ (2019) Grand research challenges for sustainable industrial biotechnology. Trends Biotechnol 37(10):1042–1050. https://doi.org/10.1016/j.tibtech.2019.04.002 131. Benito-Vaquerizo S et al (2020) Modeling a co-culture of Clostridium autoethanogenum and Clostridium kluyveri to increase syngas conversion to medium-chain fatty-acids. Comput Struct Biotechnol J:3255–3266. https://doi.org/10.1016/j.csbj.2020.10.003 132. Foster C, Charubin K, Papoutsakis ET, Maranas CD (2021) Modeling growth kinetics, interspecies cell fusion, and metabolism of a Clostridium acetobutylicum/Clostridium ljungdahlii syntrophic co-culture. mSystems 6(1). https://doi.org/10.1128/mSystems. 01325-20 133. Li X, Henson MA (2019) Metabolic modeling of bacterial co-culture systems predicts enhanced carbon monoxide-to-butyrate conversion compared to monoculture systems. Biochem Eng J 151:107338. https://doi.org/10.1016/j.bej.2019.107338 134. Li X, Henson MA (2021) Dynamic metabolic modeling predicts efficient acetogen-gut bacterium cocultures for CO-to-butyrate conversion. J Appl Microbiol. https://doi.org/10. 1111/jam.15155 135. Klasson KT, Cowger JP, Ko CW, Vega JL, Clausen EC, Gaddy JL (1990) Methane production from synthesis gas using a mixed culture of R. rubrum, M. barkeri, and M. formicicum. Appl Biochem Biotechnol 24–25(1):317–328. https://doi.org/10.1007/BF02920256 136. Kohlmayer M, Huber R, Brotsack R, Mayer W (2018) Simultaneous CO2 and CO methanation using microbes. bioRxiv:326561. https://doi.org/10.1101/326561 137. Diender M, Stams AJM, Sousa DZ (2016) Production of medium-chain fatty acids and higher alcohols by a synthetic co-culture grown on carbon monoxide or syngas. Biotechnol Biofuels 9(1):82. https://doi.org/10.1186/s13068-016-0495-0 138. Diender M et al (2019) Metabolic shift induced by synthetic co-cultivation promotes high yield of chain elongated acids from syngas. Sci Rep:1–11. https://doi.org/10.1038/s41598019-54445-y 139. Moreira JPC et al (2021) Propionate production from carbon monoxide by synthetic co-cultures of Acetobacterium wieringae spp. and propionigenic bacteria. Appl Environ Microbiol 87(14):e0283920. https://doi.org/10.1128/AEM.02839-20 140. Cha S, Lim HG, Kwon S, Kim D, Kang CW, Jung GY (2021) Design of mutualistic microbial consortia for stable conversion of carbon monoxide to value-added chemicals. Metab Eng 64: 146–153. https://doi.org/10.1016/j.ymben.2021.02.001 141. Westman S, Chandolias K, Taherzadeh M (2016) Syngas biomethanation in a semi-continuous reverse membrane bioreactor (RMBR). Fermentation 2(4):8. https://doi.org/10.3390/ fermentation2020008 142. Slepova TV et al (2006) Carboxydocella sporoproducens sp. nov., a novel anaerobic CO-utilizing/H2-producing thermophilic bacterium from a Kamchatka hot spring. Int J Syst Evol Microbiol 56(4):797–800. https://doi.org/10.1099/ijs.0.63961-0 143. Sokolova TG (2002) Carboxydocella thermautotrophica gen. nov., sp. nov., a novel anaerobic, CO-utilizing thermophile from a Kamchatkan hot spring. Int J Syst Evol Microbiol 52(6):1961–1967. https://doi.org/10.1099/ijs.0.02173-0 144. Phillips JR et al (2015) Butanol and hexanol production in Clostridium carboxidivorans syngas fermentation: medium development and culture techniques. Bioresour Technol 190: 114–121. https://doi.org/10.1016/j.biortech.2015.04.043 145. Chang IS, Kim BH, Lovitt RW, Bang JS (2001) Effect of CO partial pressure on cell-recycled continuous CO fermentation by Eubacterium limosum KIST612. Process Biochem 37(4):411–421. https://doi.org/10.1016/S0032-9592(01)00227-8 146. Valgepea K et al (2018) H2 drives metabolic rearrangements in gas-fermenting Clostridium autoethanogenum. Biotechnol Biofuels 11(1):55. https://doi.org/10.1186/s13068-018-1052-9 147. Haas T, Krause R, Weber R, Demler M, Schmid G (2018) Technical photosynthesis involving CO2 electrolysis and fermentation. Nat Catal 1(1):32–39. https://doi.org/10.1038/s41929-0170005-1

Conversion of Carbon Monoxide to Chemicals Using Microbial Consortia

407

148. Kim SA, Rhee MS (2013) Marked synergistic bactericidal effects and mode of action of medium-chain fatty acids in combination with organic acids against Escherichia coli O157: H7. Appl Environ Microbiol 79(21):6552–6560. https://doi.org/10.1128/AEM.02164-13 149. Piwowarek K, Lipińska E, Hać-Szymańczuk E, Kieliszek M, Ścibisz I (2018) Propionibacterium spp. – source of propionic acid, vitamin B12, and other metabolites important for the industry. Appl Microbiol Biotechnol 102(2):515–538. https://doi.org/10. 1007/s00253-017-8616-7 150. Tholozan JL, Touzel JP, Samain E, Grivet JP, Prensier G, Albagnac G (1992) Clostridium neopropionicum sp. nov., a strict anaerobic bacterium fermenting ethanol to propionate through acrylate pathway. Arch Microbiol 157(3):249–257. https://doi.org/10.1007/ BF00245158 151. Hu P et al (2016) Integrated bioprocess for conversion of gaseous substrates to liquids. Proc Natl Acad Sci 113(14):3773–3778. https://doi.org/10.1073/pnas.1516867113 152. Hwang HW et al (2020) Two-stage bioconversion of carbon monoxide to biopolymers via formate as an intermediate. Chem Eng J 389:124394. https://doi.org/10.1016/j.cej.2020. 124394 153. Lagoa-Costa B, Abubackar HN, Fernández-Romasanta M, Kennes C, Veiga MC (2017) Integrated bioconversion of syngas into bioethanol and biopolymers. Bioresour Technol 239:244–249. https://doi.org/10.1016/j.biortech.2017.05.019 154. Sorokin DY et al (2020) Natranaerofaba carboxydovora gen. nov., sp. nov., an extremely haloalkaliphilic CO -utilizing acetogen from a hypersaline soda lake representing a novel deep phylogenetic lineage in the class “Natranaerobiia”. Environ Microbiol 23(7):3460–3476 https://doi.org/10.1111/1462-2920.15241 155. Hanemaaijer M, Olivier BG, Röling WFM, Bruggeman FJ, Teusink B (2017) Model-based quantification of metabolic interactions from dynamic microbial-community data. PLoS One 12(3):e0173183. https://doi.org/10.1371/journal.pone.0173183 156. Pacheco AR, Moel M, Segrè D (2019) Costless metabolic secretions as drivers of interspecies interactions in microbial ecosystems. Nat Commun 10(1):103. https://doi.org/10.1038/ s41467-018-07946-9 157. Stolyar S et al (2007) Metabolic modeling of a mutualistic microbial community. Mol Syst Biol 3(1):92. https://doi.org/10.1038/msb4100131 158. Gu C, Kim GB, Kim WJ, Kim HU, Lee SY (2019) Current status and applications of genomescale metabolic models. Genome Biol 20(1):121. https://doi.org/10.1186/s13059-019-1730-3 159. Wen Z et al (2017) Enhanced solvent production by metabolic engineering of a twin-clostridial consortium. Metab Eng 39:38–48. https://doi.org/10.1016/j.ymben.2016.10.013 160. Wen Z, Wu M, Lin Y, Yang L, Lin J, Cen P (2014) Artificial symbiosis for acetone-butanolethanol (ABE) fermentation from alkali extracted deshelled corn cobs by co-culture of Clostridium beijerinckii and Clostridium cellulovorans. Microb Cell Fact 13(1):92. https:// doi.org/10.1186/s12934-014-0092-5 161. Zhang H, Li Z, Pereira B, Stephanopoulos G (2015) Engineering E. coli – E. coli cocultures for production of muconic acid from glycerol. Microb Cell Fact 14(1):134. https://doi.org/10. 1186/s12934-015-0319-0 162. Liao MJ, Din MO, Tsimring L, Hasty J (2019) Rock-paper-scissors: engineered population dynamics increase genetic stability. Science 365(6457):1045–1049. https://doi.org/10.1126/ science.aaw0542 163. Carlson ED, Papoutsakis ET (2017) Heterologous expression of the Clostridium carboxidivorans CO dehydrogenase alone or together with the acetyl coenzyme a synthase enables both reduction of CO2 and oxidation of CO by Clostridium acetobutylicum. Appl Environ Microbiol 83(16). https://doi.org/10.1128/AEM.00829-17 164. Gleizer S et al (2019) Conversion of Escherichia coli to generate all biomass carbon from CO2. Cell 179(6):1255–1263.e12. https://doi.org/10.1016/j.cell.2019.11.009 165. Kim S et al (2020) Growth of E. coli on formate and methanol via the reductive glycine pathway. Nat Chem Biol 16(5):538–545. https://doi.org/10.1038/s41589-020-0473-5