ABSTRACT ZELDES, BENJAMIN. Leveraging Extreme

ABSTRACT

ZELDES, BENJAMIN. Leveraging Extreme Thermoacidophily for Archaeal Metabolic Engineering (Under the direction of Dr. Robert Kelly).

Recent improvements in molecular genetics tools for extreme thermophiles mean that microbial metabolic engineering is now possible at temperatures in excess of 70°C.

Thermophilic organisms have had a dramatic impact in both science and industry based on the utility of their thermostable, thermoactive enzymes. Extreme thermophile metabolic engineering means that more complex bio-transformations involving multi-enzyme pathways are now possible. Among the many promising microorganisms for industrial biotechnology are members of the thermoacidophilic (Topt > 75°C, pHopt < 3) archaeal order Sulfolobales, many of which are chemolithoautotrophs. As such, they contain pathways for acquiring energy from inorganic chemical sources, such as metal ores and elemental sulfur, and a carbon fixation cycle for taking up CO2. Portions of the carbon fixation cycle expressed in another extreme thermophile,

Pyrococcus furiosus, have produced the bioplastic precursor 3-hydroxypropionate (3HP), where one-third of the carbon in the final product is derived from CO2. Expression of chemical production pathways within a chemolithoautotrophic species of Sulfolobales would allow for production of carbon chemicals entirely from carbon dioxide, using inorganic chemical energy sources which are plentiful and inexpensive.

Metabolic engineering also has the potential to provide insights into aspects of thermophilic metabolism that remain poorly understood. Co-expression of additional enzymes alongside those for carbon fixation in P. furiosus determined that carbonic anhydrase plays an important role in CO2 uptake in the Sulfolobales, and a biotinylating maturation enzyme dramatically improved function of the first enzyme in the cycle. Similar insights into the process of sulfur oxidation in Sulfolobales were obtained by cloning two sulfur oxidation enzymes into Sulfolobus acidocaldarius, a species in which lithoautotrophic sulfur oxidation has been lost.

While the sulfur oxygenase reductase (SOR) and thiosulfate quinone oxidoreductase (TQO) had been characterized individually, their co-expression revealed cooperative effects as a full sulfur oxidation pathway. Sulfur was toxic to the strain expressing SOR alone, but adding TQO led to robust growth in the presence of sulfur and significant sulfur oxidation. Transcriptomic analysis revealed that S. acidocaldarius retains mechanisms to detect and respond to the presence of sulfur, but exhibits minimal response to CO2. Future work will focus on a carbon-fixation associated regulatory system, and the new model species for sulfur oxidation, Acidianus brierleyi, for which a genome sequence has just become available, and sulfur-transcriptome data is pending.

Continued progress in extreme thermophile metabolic engineering will depend on fully realizing the unique advantages found at high temperatures. One of the most promising is the potential for facilitated purification and continuous removal of a volatile chemical as it is produced by a thermophilic host, termed “bio-reactive distillation” (BRD). Acetone has both the requisite volatility, and value as a chemical product. Moderately thermophilic acetone production has been demonstrated, but BRD requires temperatures in excess of 70°C. Despite a dearth of thermophilic native acetone producers, enzyme candidates were identified, including the first thermophilic acetoacetyl-CoA-transferases to be characterized, and an unusually thermostable enzyme from the mesophile Clostridium acetobutylicum. The isolated subunits of CoA transferase exhibit dramatically different thermostabilities, but in complex alpha protects the more labile beta. Together with a previously characterized thermophilic thiolase, these enzymes function as an in vitro synthetic pathway to produce acetone from acetyl-CoA at 70°C. The work reported here provides improved understanding of chemolithoautotrophic energy and carbon fixation pathways in the Sulfolobales, as well as thermostable enzymatic routes for production of 3HP and acetone. Together with rapidly improving molecular genetics techniques, these results constitute the first steps towards creation of a metabolically engineered

Sulfolobus strain for production of volatile bio-based chemicals from inorganic carbon and energy sources.

By Benjamin Zeldes

A dissertation submitted to the Graduate Faculty of North Carolina State University in partial fulfillment of the requirements for the degree of Doctor of Philosophy

Chemical Engineering

Raleigh, North Carolina

2018

APPROVED BY:

______Dr. Robert M. Kelly Dr. Jason Haugh Committee Chair

______Dr. Rodolphe Barrangou Dr. Chase Beisel

BIOGRAPHY

Benjamin Monroe Zeldes was born in North Carolina to Steven Zeldes and Dru Monroe, but spent most of his childhood growing up in Arizona with his younger sister, Kristin. After high school, Ben attended the University of Pittsburgh, where he met his future wife Jennifer

Huling. After graduating with degrees in Bioengineering (BS) and Political science (BA) in

2012, Ben decided to continue his education in the Department of Chemical and Biomolecular

Engineering at North Carolina State University. Ben has spent the past several years studying a variety of exotic extremophiles under the advisement of Dr. Robert Kelly. After completing his

PhD in June 2018, Ben plans to continue doing research with an eye towards an R&D career in biotechnology.

ACKNOWLEDGMENTS

I would like to thank my advisor, Dr. Robert Kelly, for his guidance and mentorship throughout graduate school. Whenever challenges come up during your PhD it is nice to hear from someone who has seen a quite a few of them and can give you some perspective. I would also like to thank past and present members of the Kelly lab, both for their emotional support and technical assistance throughout this process. Being surrounded by people who are passionate about research has made graduate school a rewarding and mostly fun experience. The NIH

Molecular Biology Training program has provided excellent training, and afforded me opportunities to explore potential future career paths.

Of course I would like to acknowledge my parents, for their loving support and encouragement throughout my life, and for inspiring me to be curious from an early age. My whole extended family has been supportive of my decision to pursue higher education. Many have graduate degrees themselves, and encouraged me to consider this path, either directly through conversations and advice, or simply by example. Finally, I would like to thank my wife,

Dr. Huling, for her love and support, and for leading the way through this whole PhD process.

iii

TABLE OF CONTENTS

LIST OF TABLES ...... vii LIST OF FIGURES ...... viii CHAPTER 1 Extremely thermophilic microorganisms as metabolic engineering platforms for production of fuels and industrial chemicals ...... 1 1.1. Abstract ...... 2 1.2. Introduction ...... 3 1.3. Genetics in extreme thermophiles ...... 5 1.4. Candidates for high-temperature metabolic engineering ...... 8 Thermococcus kodakarensis ...... 8 Pyrococcus furiosus ...... 10 Sulfolobus species ...... 14 Thermus thermophilus ...... 17 1.5. Early stage genetic systems for extreme thermophiles ...... 18 Metallosphaera sedula:...... 18 Thermoanaerobacter mathranii: ...... 19 Caldicellulosiruptor bescii: ...... 19 Thermotoga:...... 20 1.6. Overview of current state of industrial bioprocessing ...... 21 1.7. Future of extremely thermophilic metabolic engineering: challenges and promise ...... 22 1.8. References ...... 38 CHAPTER 2 Ancillary contributions of heterologous biotin protein ligase and carbonic

anhydrase for CO2 incorporation into 3-hydroxypropionate by metabolically engineered Pyrococcus furiosus ...... 50 2.1. Abstract ...... 51 2.2. Introduction ...... 52 2.3. Materials and Methods ...... 55 P. furiosus strain construction ...... 55 Bioreactor growth of P. furiosus strains ...... 56 Quantification of metabolites in bioreactors ...... 57 Heterologous expression of M. sedula genes in E. coli ...... 58

Purification of recombinant His-tagged M. sedula proteins ...... 58 Expression and purification of recombinant M. sedula ACC from P. furiosus ...... 59 Biochemical assays ...... 60 2.4. Results ...... 62 Identification and characterization of the M. sedula carbonic anhydrase (CA) ...... 62 Production and characterization of recombinant carboxylase ACC ...... 64 Effect of recombinant M. sedula CA and BPL on 3HP production in P. furiosus ...... 66 2.5. Discussion ...... 68 2.6. Acknowledgments ...... 71 2.7. References ...... 77 CHAPTER 3 Recovering chemolithoautotrophy in Sulfolobus acidocaldarius: Insights

into restoring CO2 fixation and sulfur oxidation processes ...... 80 3.1. Abstract ...... 81 3.2. Introduction ...... 82 3.3. Carbon dioxide fixation in the Sulfolobales ...... 84 3.4. Sulfur oxidation in the Sulfolobales ...... 84 3.5. Methods ...... 87 Cultivation of Sulfolobales ...... 87 Creation of recombinant S. acidocaldarius strains ...... 88 Monitoring sulfur oxidation ...... 89 RNA extraction ...... 89 Oligonucleotide microarray ...... 90 qPCR analysis ...... 91 RNA sequencing ...... 91 3.6. Results ...... 92 Growth of recombinant strains ...... 92 Transcriptomic analysis of effect of sulfur on parent and mutant S. acidocaldarius strains ...... 94 Verification of cDNA microarray results ...... 98 Tests for autotrophic and mixotrophic growth ...... 98 Acidianus brierleyi autotrophy and preliminary transcriptomic data ...... 100

3.7. Discussion ...... 101 3.8. Acknowledgements ...... 105 3.9. References ...... 116 CHAPTER 4 Synthetic enzymatic pathway for thermophilic acetone production incorporating an unusually thermostable enzyme from mesophilic Clostridium acetobutylicum ...... 119 4.1. Abstract ...... 120 4.2. Introduction ...... 121 4.3. Methods ...... 125 Identification of thermophilic gene candidates ...... 125 Protein expression ...... 125 Protein purification ...... 125 Enzyme assays ...... 126 Other methods ...... 127 4.4. Results ...... 128 Thermophilic acetone pathway candidates ...... 128 Expression of acetone pathway enzymes ...... 129 Characterization of acetone pathway enzymes ...... 131 Full in vitro pathway ...... 135 4.5. Conclusions ...... 135 4.6. Acknowledgements ...... 138 4.7. References ...... 146

LIST OF TABLES

Table 1.1: Selective pressures for genetic manipulations in extreme thermophiles ...... 30 Table 1.2: Extreme thermophiles with functional genetic systems, and successful metabolic engineering efforts ...... 31 Table 1.3: Commodity prices of fuel feedstocks ...... 33 Table 1.4: Commercial scale biochemical production ...... 34 Table 2.1: P. furiosus strains used in this study ...... 72 Table 2.2: Recombinant M. sedula enzymes used in this study ...... 73 Table 2.3: Maximum 3PH titers reached during bioreactor runs with recombinant strains ...... 73 Table 2.4 Activity of ACC measured in cell extract of bioreactor grown cells ...... 74 Table 3.1: Saci Strongly responding genes in microarray ...... 106 Supplemental Table 3.2: Primers used in this study ...... 114 Table 4.1: Highest temperatures for solvent production natively and in metabolically engineered hosts ...... 139 Table 4.2: SCooP melting temperatures of Adc structures ...... 139 Table 4.3: Enzyme candidates for thermophilic acetone pathway ...... 139

vii

LIST OF FIGURES

Figure 1.1: Genetically tractable extreme thermophiles and their optimum growth temperatures ...... 35 Figure 1.2: General strategy for chromosomal gene insertion ...... 36 Figure 1.3: Temperature-shift strategy involving a hyperthermophilic host expressing more moderately thermostable recombinant enzymes ...... 37 Figure 2.1: 3-hydroxypropionate formation and accessory enzymes ...... 74 Figure 2.2: Amino acid sequence alignment of BCCP around the canonical lysine residue (K) . 75 Figure 2.3: In vitro biotinylation of recombinant Msed ACC-β with recombinant Msed BPL .... 75 Figure 2.4: Representative bioreactor runs of metabolically engineered P. furiosus strains ...... 76 Figure 3.1: 16S tree of Sulfolobales and their lithotrophic metabolisms ...... 108 Figure 3.2: 3-Hydroxypropionate/4-hydroxybutyraate carbon fixation cycle in Sulfolobales ... 108 Figure 3.3: Enzymes for sulfur oxidation and their distribution within the Sulfolobales...... 109 Figure 3.4: Experimental design for microarray experiment ...... 110 Figure 3.5: Phenotype of recombinant Saci strains on sulfur ...... 110 Figure 3.6: Transcriptional differences within strains under sulfur and non-sulfur conditions.. 111 Figure 3.7: Transcriptional differences between strains MW001 and RK34 ...... 112 Figure 3.8: qPCR results for key genes in modified strains ...... 113 Figure 3.9: Transcript levels of 3HP/4HB genes in Saci vs. Msed as a percentile of all genes in array ...... 113 Figure 4.1: Vapor-liquid equilibria for acetone and ethanol ...... 140 Figure 4.2: Sequence alignments of acetone enzymes ...... 141 Figure 4.3: Structure of E. coli Ctf and multispecies alignment of active site sequences ...... 142 Figure 4.4: SDS-PAGE and Blue-native PAGE of acetone enzymes ...... 143 Figure 4.5: Thermal inactivation of Tmel Ctf ...... 143 Figure 4.6: Hanes plot of Tmel Ctf kinetics ...... 144 Figure 4.7: Thermal stability of Cace Adc ...... 144 Figure 4.8: In vitro function of full acetone pathway ...... 145 Figure 4.9: Three enzyme pathway to acetone production ...... 145

viii

CHAPTER 1

Extremely thermophilic microorganisms as metabolic engineering platforms for

production of fuels and industrial chemicals

Benjamin M. Zeldes1, Matthew W. Keller2, Andrew J. Loder1, Christopher T. Straub1, Michael

W.W. Adams2, and Robert M. Kelly1,*

1Department of Chemical and Biomolecular Engineering, North Carolina State University,

Raleigh, NC 27695-7905

2Department of Biochemistry and Molecular Biology, University of Georgia, Athens, GA 30602

Published in: Frontiers in Microbiology (2015), issue 6, article 1209

1.1. Abstract

Enzymes from extremely thermophilic microorganisms have been of technological interest for some time because of their ability to catalyze reactions of industrial significance at elevated temperatures. Thermophilic enzymes are now routinely produced in recombinant mesophilic hosts for use as discrete biocatalysts. Genome and metagenome sequence data for extreme thermophiles provide useful information for putative biocatalysts for a wide range of biotransformations, albeit involving at most a few enzymatic steps. However, in the past several years, unprecedented progress has been made in establishing molecular genetics tools for extreme thermophiles to the point that the use of these microorganisms as metabolic engineering platforms has become possible. While in its early days, complex metabolic pathways have been altered or engineered into recombinant extreme thermophiles, such that the production of fuels and chemicals at elevated temperatures has become possible. Not only does this expand the thermal range for industrial biotechnology, it also potentially provides biodiverse options for specific biotransformations unique to these microorganisms. The list of extreme thermophiles growing optimally between 70 and 100°C with genetic toolkits currently available includes archaea and bacteria, aerobes and anaerobes, coming from genera such as Caldicellulosiruptor,

Sulfolobus, Thermotoga, Thermococcus and Pyrococcus. These organisms exhibit unusual and potentially useful native metabolic capabilities, including cellulose degradation, metal solubilization, and RuBisCO-free carbon fixation. Those looking to design a thermal bioprocess now have a host of potential candidates to choose from, each with its own advantages and challenges that will influence its appropriateness for specific applications. Here, the issues and opportunities for extremely thermophilic metabolic engineering platforms are considered with an eye towards potential technological advantages for high temperature industrial biotechnology.

1.2. Introduction

Microorganisms have been utilized for millennia in the production of food and beverages.

With the advent of the industrial revolution, microbes were used to produce ethanol for fuel

(Songstad et al., 2009), then acetone and butanol as chemical feedstocks during World War I

(Jones and Woods, 1986). However, the discovery of easily accessible petroleum deposits, coupled with improvements in oil refineries, placed the biological routes at an insurmountable disadvantage for decades. The rising economic and environmental costs of petroleum-based products have renewed interest in biological production of commodity fuels, as well as specialty chemicals not easily synthesized via petrochemical routes. Most research in this area has focused on microbes growing in the mesophilic temperature range (25 - 37°C). However, high temperature fermentations, closer to the temperatures used in chemical refineries, are possible through the use of extremely thermophilic (Topt > 70°C) microbial hosts, offering a number potential of advantages over mesophilic biorefineries.

The enzymes of extreme thermophiles have been of considerable interest in biotechnology ever since the development of the polymerase chain reaction (Bartlett and Stirling,

2003). Given the usefulness of thermostable enzymes in molecular biological laboratory methods, it is not surprising that they have been proposed as powerful tools for industrial catalysis as well (Zamost, 1991;Vieille and Zeikus, 2001;Atomi et al., 2011). It is also becoming increasingly possible to improve the thermostability of mesophilic enzymes, either through protein engineering or techniques such as enzyme immobilization (Lehmann and Wyss,

2001;Harris et al., 2010;Steiner and Schwab, 2012;Singh et al., 2013), so that thermally-based bioprocessing can be considered.

High-temperature bioprocessing has a number of advantages, including reduced risk of contamination as compared to mesophilic hosts such as E. coli and S. cerevisiae, lowered chances of phage infection, improved solubility of substrates such as lignocellulosic biomass, continuous recovery of volatile chemical products directly from fermentation broth, and reduced cooling costs due to the greater temperature differential between the fermenter and the ambient air, which is the ultimate heat acceptor (Frock and Kelly, 2012;Keller et al., 2014). Increasing temperature can also make reactions that would be unfavorable in mesophiles thermodynamically feasible. In the industrial production of fructose from corn syrup via glucose isomerase, higher temperatures favor the fructose side of the reaction, creating a final product with better sweetening power (Bhosale et al., 1996). Hydrogen production becomes more favorable at high temperatures, leading to increased hydrogen productivities in thermophiles

(Verhaart et al., 2010), and bioleaching of highly refractory ores such as chalcopyrite is more favorable under thermophilic conditions (du Plessis, 2007). Methane production actually yields less energy at high temperature (Amend and Shock, 2001), but this leads to improved methane production in thermophilic methanogens (Ahring, 1995) since more methane must be generated to provide the same amount of cellular energy.

Thermostable enzymes can be used for in vitro single-step reactions, such as hydrolyzing large biopolymers into smaller components by proteinases, chitinases, cellulases and other carbohydrate-degrading enzymes (Vieille and Zeikus, 2001). However, more complex multistep chemical conversions require an intact cellular host (Ladkau et al., 2014). Many interesting and potentially industrially relevant pathways require multiple enzyme steps, regeneration of cofactors, energy conservation via coupling to a transmembrane gradient, or input of additional chemical energy. The potential for extreme thermophiles to serve as intact platforms for

metabolic engineering and whole-cell biocatalysts has been considered (M P Taylor, 2011;Frock and Kelly, 2012), but the field remains in its infancy (Figure 1.1).

1.3. Genetics in extreme thermophiles

A major challenge to genetic modification in extreme thermophiles is establishing selective pressure for obtaining positive transformants. The antibiotics typically used in mesophiles often target cell components specific to bacteria, and so are ineffective against the archaeal species that dominate at high temperatures. Even in cases where antibiotics are effective, both the antimicrobial compound and the gene product that confers resistance to it must be stable at elevated temperatures. Because of the challenges with antibiotics, nutritional selection techniques such as those initially established in yeast genetics (Romanos et al., 1992), predominate in the genetic systems currently available for extreme thermophiles. Selective markers that have been successfully used in extreme thermophiles are summarized in Table 1.1.

The most frequently used nutritional selection method in extreme thermophiles is based on uracil prototrophic selection from an auxotrophic parental strain. Synthesis of uracil involves enzymes encoded by the pyrE (orotate phosphoribosyltransferase) and pyrF (orotidine-5'- phosphate decarboxylase) genes that, besides their role in uracil production, also convert the synthetic chemical 5-fluoroorotic acid (5-FOA) into the cytotoxic fluorodeoxyuridine (Jund and

Lacroute, 1970;Krooth et al., 1979). Therefore, growth of strains with functional uracil pathways on media containing 5-FOA selects for natural mutants with disruptions in pyrE or pyrF (Krooth et al., 1979;Worsham and Goldman, 1988).

While nutritional selection currently dominates the genetics of thermophiles, antibiotics have played a critical role in the development of these genetic systems. Simvastatin represents an

important exception to the rule that antibiotics are only effective against mesophilic bacteria. It affects thermophilic archaea because it targets the production of archaeal membranes (Lam and

Doolittle, 1989;Matsumi et al., 2007), and displays no detectable degradation at 100°C in the absence of oxygen (Simões et al., 2014). These important characteristics have allowed simvastatin to play a major role in the development of genetic systems in the hyperthermophilic archaea Thermococcus kodakarensis (Matsumi et al., 2007) and Pyrococcus furiosus (Waege et al., 2010).

Several other antibiotics are stable at high temperatures for durations adequate to provide selective pressure. Resistance genes that confer protection from these antibiotics at lower temperature can be evolved into more thermostable mutants for use at high temperature. This strategy has been enacted by cloning mutagenized kanamycin nucleotidyltryansferase into the thermophilic bacterium Bacillus stearothermophilus. The result is an increase in the temperature limit of kanamycin resistance from 47°C up to 70°C, which is near the 72°C limit of the antibiotic itself (Matsumura and Aiba, 1985;Liao et al., 1986). These thermostable mutants were further evolved to increase activity and thermostability to facilitate use in the thermophile

Thermus (Ts) thermophilus (Hoseki et al., 1999). Bleomycin, a more thermostable antibiotic, was used to develop a hyperthermophilic (Topt >80°C) selectable marker; the engineered bleomycin- binding protein denatures at 85°C in the absence of bleomycin and 100°C in its presence (Brouns et al., 2005). Adding to the growing list of thermophilic selectable markers, E. coli hygromycin

B phospotransferase was evolved for use in Ts. thermophilus at a maximum temperature of 67°C

(Nakamura et al., 2005).

Natural competence – the ability to uptake and incorporate foreign DNA without exogenously applied pressure (such as divalent cations or an electric field) – is prevalent among

extreme thermophiles, and has been proposed to play an important role in adaptation to extreme temperature (Averhoff and Muller, 2010). Some of the simplest and most frequently utilized genetic systems in the extreme thermophiles are in the species that exhibit natural competence:

Ts. thermophilus (Koyama et al., 1986), P. furiosus strain COM1 (Lipscomb et al., 2011) and T. kodakarensis (Hileman and Santangelo, 2012). The benefit of natural competence is the simplicity of the transformation protocol, which involves simply mixing DNA with a dilute culture prior to selective plating. Transformation in other species is considerably more complex.

Even though electroporation may be effective for introducing DNA constructs into the cell, once there, cellular defenses designed to destroy foreign DNA are problematic. Restriction enzymes in C. bescii degrade DNA that is not properly methylated, creating an added challenge for genetic manipulations in this organism (Chung et al., 2012).

Most genetic manipulations in extreme thermophiles are chromosome-based rather than plasmid-based. This is due primarily to a lack of reliable replicating vectors in many organisms, and partly to the higher prevalence of nutritional selection, which is less amenable to long-term maintenance of plasmids. Chromosomal modifications are preferable for generating an industrial host, because they provide greater strain stability and eliminate the need for continued selective pressure. Chromosomal insertions and deletions are directed by, and dependent on, homologous flanking regions. These homologous flanking regions guide the maker along with or to the target gene and transform the chromosome. This can occur via either single or double homologous recombination, after which the selective marker can be removed in another recombination event using counter-selection (Figure 1.2). Several extreme thermophiles, including P. furiosus,

Sulfolobus species, and T. kodakarensis, can be transformed using either single crossover

(circular DNA) or double crossover (linear DNA) homologous recombination, while others, such as C. bescii, are limited to single crossover events.

1.4. Candidates for high-temperature metabolic engineering

While the list of extremely thermophilic microorganisms available in pure culture has expanded considerably over the past four decades, only a small subset have been characterized in detail from physiological and genomic perspectives. This group has been the focus of efforts to develop molecular genetics tools, driven mostly by a desire to study basic physiological issues.

However, the tools that have been developed open up opportunities for metabolic engineering, although only limited results are available to date on this aspect. Table 1.2 summarizes the key features of extreme thermophiles for which genetics systems have been described. These microorganisms are discussed in detail below.

Thermococcus kodakarensis

Originally isolated from a marine solfatara and named Pyrococcus kodakaraensis, this euryarchaeon was reported to have an optimum growth temperature of 95°C (Morikawa et al.,

1994). However, it was later shown to grow optimally at 85°C and was re-classified as

Thermococcus kodakaraensis (Atomi et al., 2004) (now spelled T. kodakarensis). T. kodakarensis is an anaerobic heterotroph that grows well on carbohydrates and it can also utilize peptides if elemental sulfur (S0) is present (Atomi et al., 2004). It is well established as a source for archaeal and thermophilic proteins, with nearly 100 genes characterized by the early 2000s

(Imanaka and Atomi, 2002). In the past decade, a genetic system has been developed, optimized, and used to investigate a variety of questions about the basic biology of thermophiles and archaea. Indeed, while the development of genetic tools for organisms that can grow above 80°C

was pioneered in T. kodakarensis by Atomi, Imanaka and coworkers (as outlined below), so far little work has been done to create metabolically engineered T. koodakarensis strains of industrial interest.

T. kodakarensis genetics: The first demonstration of targeted gene disruption used a uracil auxotrophic strain created by exposure to 5-FOA, allowing the pyrF gene to be used as a selectable marker for disruption of another gene (trpE) by homologous recombination (Sato et al., 2003). Soon after this initial report, a complete genome sequence of T. kodakarensis was published (Fukui et al., 2005). Over the next decade extensive work was carried out to optimize genetic techniques for T. kodakarensis (Santangelo and Reeve, 2011;Hileman and Santangelo,

2012). In summary, T. kodakarensis is naturally competent for transformation by homologous recombination with either linear or circular DNA, which can be introduced via E. coli shuttle vector. Selection can be accomplished with nutrient auxotrophy for uracil, tryptophan, arginine

(citrulline), and agmatine, and antibiotic sensitivity to mevinolin/simvastatin, while counter- selection is possible with 5-FOA and the adenine analog 6-methylpurine (6-MP). More recent studies involving genetic manipulation of T. kodakarensis have continued to rely on methods laid out in the reviews referenced above, which have also been successfully applied in related species such as Thermococcus onnurineus (Kim et al., 2013). The focus has now shifted away from development of the genetic system to applying these mature techniques to answer a variety of scientific questions.

Examples of the extent to which genetic modification is possible in T. kodakarensis include one investigation that generated 13 different deletion strains, including double-deletion mutants which required the use of 6-MP counter-selection (Santangelo et al., 2011), and another

that involved 14 different strains and included one with four rounds of pyrF/5-FOA selection/counter-selection (Harnvoravongchai et al., 2014). Additional metabolic engineering tools available in T. kodakarensis include the strong constitutive promoters of glutamate dehydrogenase (Matsumi et al., 2007) and cell surface glycoprotein (Mueller et al., 2009) for protein overexpression, and a signal peptide that enables protein secretion (Takemasa et al.,

2011).

T. kodakarensis metabolic engineering: T. kodakarensis has been metabolically engineered for improved hydrogen production, resulting in a strain capable of producing several times more hydrogen than wild-type (Santangelo et al., 2011). Efforts to further improve hydrogen production are ongoing (Kanai et al., 2015), and T. kodakarensis shows promise as a bio- hydrogen production platform. Strains optimized for overexpression and secretion of enzymes have also been developed (Takemasa et al., 2011). There are currently no reports of expressing full heterologous metabolic pathways in T. kodakarensis, but it has been used to produce recombinant versions of proteins from other thermophiles for study when expression in E. coli yielded inactive protein (Mueller et al., 2009). T. kodakarensis expresses a natural viral defense system in the form of several CRISPR cassettes (Grissa et al., 2007). Strains have been engineered to use the CRISPR system to target specific sequences of foreign DNA (Elmore et al., 2013), opening up the potential to protect industrial strains from problematic viruses.

Pyrococcus furiosus

Robust and fast-growing, Pyrococcus furiosus is convenient to work with, and also happens to be the highest temperature organism for which a versatile genetic system is available.

Able to grow on a variety of peptide and carbohydrate substrates, P. furiosus is a natural

hydrogen producer, and is also capable of reducing elemental sulfur (S0). It has already been genetically engineered to express a variety of heterologous metabolic pathways, despite the fact that the first reports of a functional genetic system did not appear until 2010.

P. furiosus was isolated from a shallow marine solfatara and found to have an optimum temperature of 100°C (Fiala and Stetter, 1986). It is a euryarchaeal heterotroph capable of growth on peptides and some oligo- and polysaccharides. Sulfur (S0) is required for peptide utilization, resulting in the production of organic acids and H2S as byproducts. Carbohydrates

0 can be utilized either with S or without, with reducing equivalents are disposed of as H2 (Adams et al., 2001). P. furiosus and T. kodakarensis appear to share a virtually identical system of reductant disposal, where the highly homologous multi-subunit membrane complex proteins

Mbh and Mbx are responsible for H2 and H2S production, respectively (Schut et al., 2013).

P. furiosus genetics: The genetic system in P. furiosus has benefited from successes and developments made previously in T. kodakarensis. In fact, the first successful genetic manipulation in P. furiosus, which used simvastatin selection to allow for protein expression from a shuttle vector (Waege et al., 2010), borrowed a transformation protocol (using divalent cations and heat-shock) developed for T. kodakarensis (Sato et al., 2003). A subsequent effort applied counter-selection via 6-MP to perform a single nucleotide deletion (Kreuzer et al., 2013).

These are the only reports of genetic transformations in wild-type P. furiosus, the majority of genetic manipulations have been based on the discovery of a naturally competent strain COM1

(Lipscomb et al., 2011).

Compared to the wild-type P. furiosus strain, the COM1 strain has undergone several large-scale chromosomal rearrangements. However, less than 2% of genes (36 ORFs)

experienced changes that reduced sequence identity with their WT homologues below 90%, and no differences in metabolic phenotype were found (Bridger et al., 2012). The original report of

COM1 transformation used uracil/5-FOA selection to generate a double-deletion mutant. The uracil auxotroph for this transformation was generated by simvastatin selection, while agmatine

(Hopkins et al., 2011) and tryptophan (Farkas et al., 2012) prototrophic selection have also been demonstrated. Most recent work has continued to rely on the simpler uracil/5-FOA system.

Promoters available for protein overexpression include the strong constitutive promoter of the S- layer protein (Pslp) (Hopkins et al., 2011), and the cold-inducible promoter from cold-shock protein A (PcipA) (Basen et al., 2012). Recently, entire operons up to 17 kilobases long have been transformed into P. furiosus using bacterial artificial chromosomes (Basen et al., 2014;Lipscomb et al., 2014).

P. furiosus metabolic engineering: As with most new genetic systems, the first demonstrations of genetic modification in P. furiosus involved gene knockouts. The genes encoding two soluble hydrogenases, believed to be essential to P. furiosus metabolism, were deleted in the original report of COM1, although there was no change in phenotype under the usual laboratory conditions (Lipscomb et al., 2011). The ancestral ability of P. furiosus to utilize chitin was recovered by a single base deletion (Kreuzer et al., 2013). But the most impressive examples of metabolic engineering in this organism have been in the expression of heterologous pathways.

The core metabolic strategies available at the extreme temperatures where P. furiosus grows best appear to be rather limited. For example, alcohol production is very rare among extreme thermophiles, possibly because the archaea, which predominate at high temperatures, are not known to produce significant amounts of any alcohol naturally (Basen et al., 2014).

Therefore, heterologous pathways to be inserted into P. furiosus must be taken from less thermophilic organisms, requiring that they be expressed under what amounts to cold-shock conditions for P. furiosus, which maintains some degree of growth down to temperatures as low as 70°C (Weinberg et al., 2005). For example, addition of a lactate dehydrogenase from

Caldicellulosiruptor bescii (Topt 75°C), under the control of PcipA, allowed production of the non- native product lactate to a titer of 0.3 g/L at 72oC (Basen et al., 2012). Another single-enzyme insertion gave P. furiosus the ability to produce ethanol via a novel pathway also incorporating native enzymes, reaching titers of 2 g/L (Basen et al., 2014).

More complex multi-enzyme metabolic engineering has also been demonstrated in P. furiosus. A three-enzyme pathway (involving five genes) constituting part of the carbon fixation cycle of Metallosphaera sedula (Topt ~75°C) was expressed heterologously in P. furiosus, yielding titers of the industrially relevant chemical 3-hydroxypropionate (3HP) of 0.05 g/L at

72oC (Keller et al., 2013). Deletion of a competing pathway roughly doubled 3HP titers over the parent strain (Thorgersen et al., 2014), while additional accessory enzymes and improved bioreactor conditions increased 3HP titers to 0.3 g/L (Hawkins et al., 2015). A synthetic butanol pathway consisting of six genes from three thermophilic bacteria enabled P. furiosus to produce butanol at 60°C, but titers remained low (0.07 g/L) even in 200x concentrated cell-suspensions

(Keller et al., 2015). Heterologous expression of the massive 17 kb, 18-gene formate dehydrogenase operon from Thermococcus onnurineus allowed P. furiosus to generate H2 from formate (Lipscomb et al., 2014). Insertion of another large operon from T. onnurineus, encoding the carbon monoxide dehydrogenase complex, conferred the ability of P. furiosus to use CO as a source of reductant (Basen et al., 2014).

With an incredible diversity of functional engineered pathways, P. furiosus is the greatest success story so far in metabolic engineering of extreme thermophiles. However, additional improvements will be necessary to turn current strains, which often produce only trace amounts of the target chemicals, into viable industrial hosts. Some progress has already been made in this area, particularly in 3HP production, where additional enzymes and improved growth conditions increased titers nearly ten-fold (Hawkins et al., 2015).

Sulfolobus species

Members of the archaeal genus Sulfolobus are found in a variety of acidic freshwater hot springs with water temperature around 80°C and pH below 3, making them extreme thermoacidophiles. Three Sulfolobus species, S. acidocaldarius, S. solfataricus, and S. islandicus, have functional genetic systems. All three species are obligate aerobes, grow well on rich media, and single colonies can be isolated on solid substrates. Combined with their genetic tractability, these traits make them excellent model organisms. Sulfolobus species have been used extensively to elucidate the mechanisms and cellular machinery of transcription in archaea, as a model host for archaeal viruses, and as a source of easily crystallized thermophilic proteins. So far, no member of Sulfolobus has been metabolically engineered to produce a commercially- desirable chemical product.

Following the initial isolation of S. acidocaldarius (Brock et al., 1972) and S. solfataricus

(De Rosa, 1975), species were regularly re-isolated independently (Zillig et al., 1980). This, combined with difficulties obtaining pure cultures, led to the use of mixed and misidentified strains during the early stages of Sulfolobus research (Grogan, 1989). S. islandicus was isolated more recently (Zillig et al., 1993), but has suffered from similar confusion: fresh isolates are reported so frequently that no single strain has been dominant enough to be thoroughly

characterized; in fact, the species name islandicus is not yet officially recognized (Zuo et al.,

2014).

The first isolation report for a Sulfolobus species described its ability to grow autotrophically by oxidation of sulfur (Brock et al., 1972); subsequently, autotrophic Sulfolobus sp. have been re-isolated from the environment (Wood et al., 1987;Nixon and Norris, 1992).

However, current laboratory strains of S. acidocaldarius and S. solfataricus appear to have lost this ability, and only grow if organic substrates are present (Berkner and Lipps, 2008). A comparison of the two strains during heterotrophic growth indicates that, while both grow well on complex protein sources and starch, S. solfataricus can utilize a much more diverse set of mono- and disaccharides than S. acidocaldarius (Grogan, 1989). The original S. islandicus isolates were determined to be obligate heterotrophs, growing well on complex media containing tryptone or yeast extract (Zillig et al., 1993), but no detailed characterization of preferred energy sources has been done.

Sulfolobus genetics: The first genetic modifications in Sulfolobus relied on nutrient selection.

Lactose selection (Worthington et al., 2003) is limited to S. solfataricus and S. islandicus, because S. acidocaldarius cannot be cultured on lactose-based minimal media. Uracil selection, which has been used in all three species (Albers et al., 2006;She et al., 2009;Wagner et al.,

2012), allows for richer media and better cell growth than lactose, but suffers from high background due to residual uracil, especially in S. solfataricus and S. islandicus (Berkner and

Lipps, 2008). Genetic transformations have also been accomplished in S. islandicus with simvastatin selection (Zhang and Whitaker, 2012). Electroporation is the standard means of transformation across all three species. DNA methylation is used in some methods involving S.

solfataricus (Albers and Driessen, 2008) and S. acidocaldarius (Wagner et al., 2012), but does not seem to be necessary for S. islandicus.

Genome sequences are available for the type strain of S. acidocaldarius (DSM 639)

(Chen et al., 2005) and S. solfataricus strain P2 (DSM 1616) (She et al., 2001), while nearly 20 different strains of S. islandicus have been sequenced. None of the islandicus strains are available from culture collections and no type strain has been designated (Zuo et al., 2014).

Sulfolobus species, particularly S. islandicus, have been isolated in conjunction with viruses

(Guo et al., 2011), and have played an important role in understanding the diversity and host interactions of archaeal viruses (Zillig et al., 1993;Greve et al., 2004;Bize et al.,

2009;Prangishvili, 2013). As a result, a variety of viral-based vectors are available for genetic manipulation (Aucelli et al., 2006;Berkner et al., 2007;Berkner and Lipps, 2008). Inducible promoter systems are available (Berkner et al., 2010;Peng et al., 2012), and a β-galactosidase based reporter system (Jonuscheit et al., 2003) has been utilized in all three species.

Sulfolobus metabolic engineering: While production of novel products has yet to be demonstrated in Sulfolobus, an S. sulfataricus strain that utilizes cellulose more effectively has been created by overproduction of a native endoglucanase (Girfoglio et al., 2012). Therefore, if product pathways are developed, S. sulfataricus could be engineered to utilize desirable complex biomass sources. There are many other examples of homologous or heterologous overexpression of proteins in Sulfolobus, sometimes because the protein cannot be expressed in functional form otherwise, but often simply as demonstration of a new method of genetic modification. Genetic system development in Sulfolobus has been underway for over a decade, but remains a research focus.

Thermus thermophilus

First isolated from a Japanese hot spring in 1974, Thermus thermophilus is an aerobic bacterium that grows best between 70°C (Oshima and Imahori, 1974;Williams et al., 1995) and 80°C

(Swarup et al., 2014). Like other thermophiles, Ts. thermophilus has served as a source of crystallizable proteins, with structures available for over 20% of the proteins represented in its genome (Swarup et al., 2014). Ts. thermophilus is especially capable for natural DNA uptake

(Koyama et al., 1986;Hidaka et al., 1994), and as such has been used to study the cellular machinery involved in natural competence (Schwarzenlander and Averhoff, 2006;Salzer et al.,

2015a;Salzer et al., 2015b).

Ts. thermophilus grows aerobically on sugars and peptides (Oshima and Imahori, 1974), and is also capable of utilizing lipids and triglycerides (Leis et al., 2014). Additionally, strain

HB8 can grow anaerobically by denitrification (Cava et al., 2009). Ts. thermophilus appears to have a metabolic efficiency comparable to E. coli, achieving similar biomass yields during growth on minimal glucose medium, albeit with a longer doubling time of approximately 3 hours. (Swarup et al., 2014).

Ts. thermophilus genetics: Genome sequences are available for Ts. thermophilus strains HB8 and HB27, consisting of a slightly less than 2 Mb chromosome, a 200 kb megaplasmid, and a second 9 kb plasmid in HB8 (Henne et al., 2004;Bruggemann and Chen, 2006). The HB8 strain appears to be polyploid, like the closely related Deinococcus radiodurans, with cells able to maintain two different antibiotic resistance genes at the same location on the chromosome

(Ohtani et al., 2010). The presence of multiple chromosome copies was proposed to present a

significant challenge to genetic manipulation, but more recent reports of genetic modification in strain HB27 (Leis et al., 2014;Carr et al., 2015) make no mention of it.

Early genetics work in Ts. thermophilus relied on thermostable kanamycin resistance genes for plasmid cloning and gene knockouts, but thermostable genes for resistance to the antibiotics hygromycin and bleomycin have since been developed (Cava et al., 2009).

Markerless deletions can be accomplished by the uracil/5-FOA method (Tamakoshi et al., 1999), or by a recently reported method entailing use of a phenylalanine analog and a mutated tRNA gene that confers sensitivity to it (Carr et al., 2015).

Ts. thermophilus metabolic engineering: The genetic system of Ts. thermophilus is facile enough to generate quadruple-knockout strains (Leis et al., 2014), and has been used to overexpress active tagged versions of its own proteins (Hidalgo et al., 2004;Moreno et al., 2005).

So far, the only metabolic engineering reported in Ts. thermophilus involved transferring the ability to grow anaerobically by denitrification to previously obligately aerobic strains (Ramirez-

Arcos et al., 1998).

1.5. Early stage genetic systems for extreme thermophiles

Metallosphaera sedula:

M. sedula, a relative of the Sulfolobus species discussed above, is an extreme thermoacidophile, growing optimally at 75°C and pH of around 3 (Huber et al., 1989). It is an aerobe, capable of autotrophic growth by oxidizing sulfidic ores or hydrogen, heterotrophic growth on peptides, or a combination of the two (Auernik and Kelly, 2010). The genome sequence has been published (Auernik et al., 2008). So far there is only one report on

development of a genetic system, which used uracil/5-FOA and electroporation to knockout a gene involved in M. sedula’s surprisingly high tolerance to heavy metal cations (Maezato et al.,

2012). Despite current limitations to the genetic system, M. sedula shows promise as an industrially relevant strain because of its highly versatile native metabolism. Its uniquely archaeal carbon fixation pathway (Kockelkorn and Fuchs, 2009), coupled with the ability to grow on hydrogen gas, makes M. sedula a promising candidate for eventual production of electrofuels (Hawkins et al., 2013), while the ability to solubilize highly refractory chalcopyrite ores makes it a candidate for use in high-temperature bioleaching operations (Zhu et al., 2013).

Thermoanaerobacter mathranii:

T. mathranii is an extremely thermophilic anaerobic bacterium that was isolated from an

Icelandic hot spring. It grows optimally at 70-75°C on xylose and produces primarily ethanol

(~20 mM) and acetate (~13 mM) (Larsen et al., 1997). T. mathranii garnered biotechnological interest when it was found to grow on, and produce ethanol from, lignocellulosic biomass at high temperatures (Ahring et al., 1999). While this ability to use lignocellulosic biomass is appealing, improving its ethanol yield is necessary to exploit T. mathranii for biofuel production. To accomplish this, glycerol dehydrogenase from Thermotoga maritima was expressed in T. mathranii, while knocking out lactate dehydrogenase using kanamycin resistance as a selectable marker. The result is ablation of lactate production and an increase of 19% in ethanol yield (Yao and Mikkelsen, 2010b).

Caldicellulosiruptor bescii:

Recently re-classified from its original designation as Anaerocellum thermophilum, C. bescii grows optimally at 75°C, and is capable of utilizing a variety of cellulosic substrates

(Svetlichnyi et al., 1990;Yang et al., 2010). It has a published genome sequence (Kataeva et al.,

2009). The development of a genetic system in this organism is in the beginning stages, based on uracil/5-FOA and electroporation (Chung et al., 2013). One added challenge has been the presence of restriction enzymes, necessitating either the use of methylated transformation constructs, or strains that lack the relevant enzymes (Chung et al., 2012;Chung et al., 2013).

Despite its recent development, the genetic system has already been used to metabolically engineer C. bescii strains that produce ethanol (0.6 g/L) (Chung et al., 2014), and to increase hydrogen production by deleting an alternative reductant disposal pathway (Cha et al., 2013). In addition, a heterologous gene encoding an archaeal tungsten-containing enzyme was successfully expressed in C. bescii, thereby demonstrating that the organism could assimilate tungsten, a metal rarely used in biological systems (Scott et al., In press).

Thermotoga:

Members of the genus Thermotoga, including T. maritima (Huber et al., 1986) and T. neapolitana (Jannasch et al., 1988), have been isolated from various marine geothermal vents.

With optimum growth temperatures around 80°C, Thermotoga species are among the most thermophilic bacteria known. Members of the genus contain a remarkable number of sugar utilization genes, allowing for growth on a wide variety of carbohydrates (Chhabra et al.,

2003;Conners et al., 2005;Frock et al., 2012). The early publication of the genome sequence of

T. maritima (Nelson et al., 1999) cemented its status as a model hyperthermophile. A plasmid isolated from Thermotoga strain RQ7 was used to transform T. maritima and T. neapolitana with antibiotic resistance to chloramphenicol or kanamycin, but the plasmids were gradually lost even with continued selective pressure (Yu et al., 2001). Ten years later, T. maritima and T. sp. RQ7 were finally stably transformed with an E. coli shuttle vector conferring kanamycin resistance

(Han et al., 2012). Transformation has been accomplished with liposomes (Yu et al., 2001) and

electroporation (Han et al., 2012), but some strains are also naturally competent (Han et al.,

2014). The only demonstration of metabolic engineering thus far involved recombinant expression of cellulases from Caldicellulosiruptor saccharolyticus, which were fused with native signal peptides for export, leading to cellulolytic activity in the cell supernatant; however, expression was not stable long term (Xu et al., (in press)).

1.6. Overview of current state of industrial bioprocessing

In order to understand the potential of extreme thermophiles for bioprocessing, as well as how far they still have to go, it is worth considering the current state of the art, which is still reliant on mesophilic hosts. While many chemicals and fuels of interest have been successfully created in microorganisms, production has been hindered by a few key factors. One of the most notable can be attributed to the high proportion of water necessary for all biological processes.

Thus, separating low concentrations of the target molecule from massive quantities of water requires significant energy inputs for heating, cooling, distillation, and transportation. A second consequence of water is the high capital costs for reactors, tanks, piping, pumps, and other equipment that is being utilized to move and store process streams that are often more than 80% water. Another hurdle to renewable chemical production is the high cost of feedstocks, although, as shown in Table 1.3, U.S. corn and Brazilian sugar are now available at rates competitive with crude oil. Both corn and crude oil provide energy at less than $10/GJ. Various other feedstocks with potential for use by biological organisms, such as hydrogen and natural gas, are also available at competitive rates.

In the landscape of chemicals being targeted for production via biological organisms, ethanol stands alone as the only chemical to have rivaled an industrial commodity on a volume

and economic basis. Over 14 billion gallons of ethanol were produced in the United States in

2014, nearly all from corn starch, with Brazil adding over six billion gallons from sugar (RFA,

2014). No other biologically produced chemical or fuel has reached the one billion gallon mark.

Ethanol benefits from a number of advantages, especially the natural ability of yeast to metabolize sugars to ethanol at high titers and with high efficiency, thus avoiding the need for extensive genetic engineering. However, the success of ethanol proves that bio-based chemical production at scale is possible, and recent progress by industry startups generating a variety of other chemicals (Table 1.4) confirms this.

A joint venture between Dupont and Tate & Lyle was the first to achieve a scale in the thousands of metric tons per year of a commodity chemical. Production of 1,3-propanediol from corn starch commenced in 2006 and, nearly a decade later, the company reports progress on expanding production. Recent years have brought more commercial facilities onto the scene

(Table 1.4), representing billions of dollars in capital investment. The move from demonstration to commercial-scale shows that the investment community is optimistic about the future prospects of biological chemical production. While none of the examples in Table 1.4 uses a thermophilic host, most of the research on these commercial scale projects was started a decade ago, before significant tools were available for genetic manipulation of thermophiles. The experiences gained in these initial projects may provide further evidence of the advantages of thermophilic processes.

1.7. Future of extremely thermophilic metabolic engineering: challenges and promise

The genetic systems for several species of extreme thermophiles are now advanced enough to begin developing metabolically engineered strains to produce industrially relevant

chemicals, but so far only P. furiosus has seen significant progress in this area. Thermophiles research has historically been focused on answering the basic questions of how life functions at high temperature: how do transcription and translation proceed, how do protein molecules fold, how can metabolism proceed through thermolabile intermediates, and could the earliest life have evolved at high temperatures? However, the emerging sector of industrial biotechnology has added the additional focus of producing fuels and chemicals by exploiting thermophilicity. Some thoughts on this subject follow.

Process control: A number of advantages become available when working with a thermophilic host. One that has already been applied in P. furiosus (Basen et al., 2012;Keller et al., 2013) is the use of temperature to regulate expression of recombinant enzymes, allowing for a shift from growth phase (where substrate goes to increase cell biomass) to production phase, where substrate is primarily being converted to product, and host metabolism is minimized (Figure

1.3).

A similar temperature-shift can performed with E. coli, where cells are cooled to 10-15°C prior to expression to improve recombinant protein solubility (Qing et al., 2004). The method is common enough at lab-scale that kits are commercially available, including ArcticExpress

(Agilent Technologies, Santa Clara, USA) and pCold (TaKaRa, Otsu, Japan), but the costs of refrigeration make it infeasible for an industrial process. In contrast, a cold-shock expression in thermophiles could be accomplished using ambient water or air as a heat acceptor.

Contamination: By far the most significant benefit of thermophilic production is expected to be minimized contamination risk. Biorefineries experience two main types of contamination.

Chronic, low level infections can reduce yields, but at large scale they are so difficult to prevent that they have been accepted as a fact of life at many corn ethanol plants. Severe infections, which crop up unpredictably, can compromise the viability of the production host, leading to a

“stuck fermentation” and requiring a complete shutdown (Skinner and Leathers, 2004).

The harsh pre-treatments sometimes used to solubilize biomass would be expected to eliminate wild-type bacteria that might compete with the production host, but contamination remains a major industrial problem. The load of bacterial contaminants was found to be comparable between two dry-grind ethanol plants that incorporated a high-temperature saccharification pre-treatment, and a wet-mill plant which did not (Skinner and Leathers, 2004).

Bacterial load was also not significantly affected by differing degrees of antibiotic use among the three processing plants, although the diversity of bacterial contaminants declined with increased antibiotics. Therefore, having a sterile feedstock or antibiotics in the bioreactor is not enough to prevent opportunistic bacteria from taking advantage of the abundant nutrients and mild conditions found in a bioreactor operated under mesophilic conditions. In contrast, using a thermophilic host ensures that all parts of the plant can be kept at elevated temperatures, so there are no ‘safe’ corners for a persisting reservoir of contaminating microorganisms. The benefits of a contamination-free thermophilic process are difficult to quantitate; production losses due to chronic contamination in ethanol plants have been estimated at anywhere between 2 and 22%

(Beckner et al., 2011). The cost of severe contamination events (leading to ‘stuck’ fermentations) is even more difficult to estimate, since there is little information about their frequency in the literature.

The use of an extremely thermophilic host is expected to reduce the risk of bacterial contamination in biorefineries, but the possibility of viral infections remains. Viruses affecting

both bacterial and archaeal hyperthermophiles have been identified, although interestingly the archaeal species infected are limited to the crenarchaeoata (Pina et al., 2011;Pietilä et al., 2014).

No virus has yet been reported for a member of the hyperthermophilic euryarchaeota, although the presence of a CRISPR system in both T. kodakarensis and P. furiosus (Grissa et al., 2007) suggests that they do experience viral infections. The native CRISPR system of T. kodakarensis was engineered to target specific sequences of foreign DNA, successfully preventing transformation with plasmids containing the target sequences (Elmore et al., 2013). This indicates that T. kodakarensis’s CRISPR system could be used to “immunize” industrial strains against problematic viruses in a manner similar to that already implemented in commercially available cheese and yogurt cultures (DuPont/Danisco, 2012;Grens).

Energy requirements: One of the oft-cited concerns for the use of thermophiles in industrial fermentations is the energy required to heat the process. There are two major thermal energy requirements for an industrial fermentation: sterilization and fermentor temperature maintenance.

As discussed above, sterilization is required for mesophilic fermentations to avoid contamination, and often achieved by heating the fermentation medium to 121°C for a short period (up to 60 min), then cooled back down to the target fermentation temperature. For a thermophilic fermentation, the sterilization process heat inputs would be no higher, and could potentially be reduced; hardy bacterial spores may be able to survive exposure to high temperatures, but they cannot grow at them.

Fermentor temperature maintenance actually provides an opportunity for energy savings if a thermophile is used. All organisms produce heat as a byproduct of metabolic processes. At large scales, the metabolic heat produced outweighs heat lost to the environment through the

fermentor walls or evaporation (Yang et al., 2008). As a result, cooling is required to maintain the fermentor at a constant temperature. The cooling duty for a large bioreactor with a high- density culture is extremely high, and can be one of the limiting factors for fermentation scale-up

(Yang, 2010). A further difficulty for mesophilic fermentations is the small thermal driving force between the fermentation temperature and the ambient environment, which limits heat removal, often making refrigeration necessary, a further energy cost (Abdel-Banat et al., 2010). The metabolic heat generated, and thus the heat removal required, are primarily dependent on the metabolic activity of the culture and not on the fermentor temperature or organism used (Blanch and Clark, 1997). Thus, a thermophilic fermentation would require a similar amount of heat removal as a mesophilic fermentation. Furthermore, this heat would be much easier to remove because of the large temperature differential between a thermophilic fermenation and the environment, providing the possibility for substantial cost savings (Abdel-Banat et al., 2010).

An additional opportunity for energy savings in thermophilic industrial fermentations is product separation, which can be the most energy intensive part of a process, since it is often carried out at elevated temperatures. In particular, thermophilic production of volatile products, such as fuel alcohols, allows for the possibility of facilitated product removal. The use of thermophilic organisms would be a favorable match for most separations processes recovering volatile products from fermentation broth, including distillation, gas stripping, and pervaporation

(Vane, 2008).

Metabolic engineering potential: While metabolic engineering opens up the possibility of engineering desired chemical pathways into any genetically tractable host, it is worth remembering that S. cerevisiae, the current workhorse of bio-ethanol production, came to

dominate the field because it was already an excellent ethanol producer. Just because the appropriate enzymes can be inserted into an organism does not mean the resulting mutant will be industrially useful. Therefore, desirable hosts should be selected not only for what they can be engineered to do, but also for what they already do well. Fortunately, the group of extreme thermophiles discussed above exhibit a variety of desirable properties natively. Many are capable of metabolizing a diverse set of sugar polymers and monomers, and C. bescii can even degrade unpretreated lignocellulosic biomass (Yang et al., 2009). Coupled with production of ethanol as either a minor or major natural metabolite, this makes them promising candidates for bio-ethanol production from non-food feedstocks. M. sedula and the Sulfolobus species grow well at low pH, a significant advantage for production of acidic products such as lactic and 3-hydroxypropionic acids, which are easier to purify in their protonated forms (Maris et al., 2004). M. sedula’s ability to solubilize metals by oxidizing them has applications in bioleaching, while its novel carbon- fixation pathway (Berg et al., 2007) offers a potential alternative to the RuBisCo-dependent

Calvin Cycle for carbon-capture applications.

Fuels are typically highly reduced organic molecules, so various efforts to maximize biofuel titers have focused on tuning the redox pathways within mesophilic hosts to favor the production of reduced end products (Liu et al., 2015). More oxidized products, such as lactic acid for production of biodegradable plastics, can be selected for by shifting metabolism in the other direction (Lee et al., 2015). Substantial progress towards redox-tuning in thermophiles has already been made. The redox pathways of several extreme thermophiles are well understood

(Schut et al., 2013), and extensive manipulation of these pathways has been demonstrated in T. kodakarensis (Santangelo et al., 2011). The ability to select for the production of alcohols, rather

than organic acids, has been demonstrated in recombinant P. furiosus by modulating the external redox environment (Basen et al., 2014).

While anaerobes like P. furiosus are potentially good producers of fermentation products, it should be remembered that they are able to extract only a fraction of the energy that aerobes obtain from the same substrates. This causes fermentative anaerobes to exist in a constant state of energy limitation, where even seemingly minor energetic costs, such as export of a final product, can be problematic (Maris et al., 2004). Therefore, expression in anaerobic hosts should be focused on pathways that are either energy-neutral or energy-yielding.

Promise of thermophiles: It is now apparent that fossil fuels cannot continue to be used at their current rate without causing irreparable environmental harm. The shift away from petroleum will necessitate dramatic changes to the current motor-fuel regime, but will also significantly alter the production of plastics, solvents, and other specialty chemicals that are currently generated in chemical refineries. Extreme thermophiles, because of their unique advantages and the recent expansion of genetic systems allowing for metabolic engineering, are perfectly positioned to fill the need for massive chemical production from renewable feedstocks. They are able to survive the high temperatures that can result from heat generated in large-scale bioreactors, and when operated at these temperatures are less likely to be contaminated by ambient microorganisms and phages. Many also exhibit unique metabolic properties as a product of their extreme environment. Much work remains to be done before the promise of using thermophilic hosts to produce large quantities of renewable fuels and chemicals can be realized, but the genetic tools are now in place to allow that work to be carried out.

1.8. Acknowledgements

This work was supported in part by the US National Science Foundation (CBET-

1264052, CBET-1264053), the US Air Force Office of Scientific Research (AFOSR) (FA9550-

13-1-0236), and the BioEnergy Science Center (BESC), a U.S. Department of Energy Bioenergy

Research Center supported by the Office of Biological and Environmental Research in the DOE

Office of Science. AJ Loder and BM Zeldes acknowledge support from an NIH Biotechnology

Traineeship (NIH T32GM008776-11), and CT Straub acknowledges support from a U.S.

Department of Education GAANN Fellowship (P200A100004-12).

Table 1.1: Selective pressures for genetic manipulations in extreme thermophiles

Selection Notes

Effective in almost all cases, acceptor strain easily Uracil prototrophy generated by growth on 5-FOA. Can suffer from high background due to contaminating uracil

Less background than uracil, another selection

mechanism required to generate acceptor strain, but Tryptophan prototrophy can be used in conjunction with uracil prototrophic

based selection.

Less background than uracil, another selection mechanism required to generate acceptor strain, but

Nutrient Agmatine prototrophy can be used in conjunction with uracil prototrophic selection.

Only effective for species capable of growth on Lactose utilization lactose minimal media. Slow due to nutrient limitations.

Kanamycin Bacteria specific

Bleomycin Bacteria specific

Hygromycin Bacteria specific

Antibiotics Simvastatin Archaea specific

pyrF counterselection, requires uracil. Can also be 5-Fluoroorotic acid used to generate the initial acceptor strain

Analogous to 5-FOA: counterselection requires

Counter Counter selection 6-Methyl purine adenine.

Table 1.2: Extreme thermophiles with functional genetic systems, and successful metabolic engineering efforts

Organism Topt Genetic tools Metabolism Metabolic Sources (Domain) engineering (titer) Thermoanaerobacter 70°C Kanamycin Anaerobe Ethanol (2.3 g/L). (Larsen et al., 1997)a, (Yao and mathranii Electroporation Heterotroph (carbohydrates) Mikkelsen, 2010a)b,c Bacterial Thermus thermophilus 70°C Kanamycin, uracil/5- Aerobe Protein (Oshima and Imahori, 1974;Cava et Bacterial FOA Heterotroph (carbohydrates overexpression. al., 2009)a, (Tamakoshi et al., Natural competence or peptides) 1999;Hashimoto et al., 2001)b, Denitrification, some (Moreno et al., 2005)c strains) Metallosphaera sedula 75°C Uracil/5-FOA Aerobe (Huber et al., 1989)a, (Maezato et Crenarchaeal Electroporation Heterotroph (peptides) al., 2012)b - Autotroph (S0, sulfidic ores, or H2) Caldicellulosiruptor 75°C Uracil/5-FOA Anaerobe Ethanol (0.6 g/L). (Svetlichnyi et al., 1990;Yang et a b bescii Electroporation, Heterotroph (cellulose, Increased H2. al., 2010) , (Chung et al., 2013) , Bacterial restriction enzyme hexose and pentose sugars) (Cha et al., 2013;Chung et al., deletion 2014)c Sulfolobus acidocaldarius 80°C# Uracil/5-FOA Aerobe (Brock et al., 1972;Grogan, 1989)a, Crenarchaeal Sugar inducible Heterotroph (starch, (Berkner et al., 2007;Berkner et al., promoters peptides, some - 2010;Wagner et al., 2012)b Shuttle vector, uracil, monosaccharides) β-gal screen Sulfolobus islandicus 78°C# Uracil/5-FOA, Aerobe (Zillig et al., 1993)a, (Deng et al., Crenarchaeal agmatine, simvastatin Heterotroph 2009;Peng et al., 2012;Zheng et al., - Sugar inducible 2012;Zhang et al., 2013)b promoters Sulfolobus solfataricus 80°C# Lactose, agmatine Aerobe Cellulose degradation. (Zillig et al., 1980;Grogan, 1989)a, Crenarchaeal Sugar inducible Heterotroph (peptides, Protein (Worthington et al., 2003;Berkner promoters many mono and overexpression. et al., 2007;Zhang et al., 2013)b, polysaccharides) (Albers et al., 2006;Girfoglio et al., 2012)c Thermotoga maritima, T. 80°C Kanamycin Anaerobe Cellulase expression (Huber et al., 1986)a, (Han et al., sp. RQ7, RQ2 Shuttle vector, Heterotroph (carbohydrates) in RQ2. 2012;Han et al., 2014)b, (Xu et al., Bacterial electroporation, (in press))c liposomes RQ7 natural competence

Table 1.2 (continued) Thermococcus 85°C Uracil/5-FOA, Anaerobe Protein expression (Atomi et al., 2004)a, (Sato et al., kodakarensis tryptophan/6-MP, Heterotroph (carbohydrates, and secretion. 2005;Matsumi et al., 0 b Euryarchaeal simvastatin peptides with S reduction) Increased H2. 2007;Santangelo et al., 2010) , Natural competence (Santangelo et al., 2011;Takemasa β-glycosidase screen et al., 2011)c Pyrococcus furiosus 100°C Shuttle vector, Anaerobe Lactate (0.3 g/L). (Fiala and Stetter, 1986)a, (Waege Euryrchaeal simvastatin, chemical Heterotroph (carbohydrates, 3HP (0.3 g/L). et al., 2010;Lipscomb et al., 2011)b, competence (CaCl) and peptides with S0 reduction) Ethanol (2 g/L). (Basen et al., 2012;Basen et al., cold shock Butanol (0.1 g/L). 2014;Hawkins et al., 2015;Keller et Uracil/5-FOA, natural al., 2015)c. competence # Topt values from (Grogan, 1989) and (Albers and Siebers, 2014). Sources: a) Isolation/metabolism, b) genetic methods, c) metabolic engineering.

Table 1.3: Commodity prices of fuel feedstocks Common biomass and fossil-derived sources. Energy content as lower heating value, except electricity.

Energy Source Cost Energy Density ΔHc (Standard (Standard ($/GJ) Units) Units) Powder River $11.55/tona 8800 BTU/lba $1.01 Basin Coal (Wyoming) Natural Gas $2.71/MMBTUb 1020 BTU/scfb $2.71 Central $49.95/tona 12500 BTU/lba $3.07 Appalachian Coal Crude Oil (WTI) $43.22/bblc 45.54 MJ/kgj $7.52 Corn Stover $75/mtd 7560 BTU/lbj $8.61 Corn Starch $3.72/bue 16.5 MJ/kgk $9.48 Hydrogen Gas $1470/mtf 121 MJ/kgk $12.25 Brazilian Sugar $262/mt TRSg 16.5 MJ/kgk $15.90 Electricity $0.0747/kWhh - $20.75 Carbon Monoxide $240/mti 5.08 MJ/kgl $23.76 a) Spot prices from EIA website, August 2015 b) Henry Hub spot prices from EIA website, August 2015 c) Cushing, OK spot prices from EIA website, August 2015 d) Prices received for stover delivered to POET-DSM plant, Emmetsburg, IA, Sept 2014 e) Yellow dent corn spot price per Chicago Board of Trade – (Assumptions: Corn starch 75wt% of bushel & 20% discount for DDG credit) f) Hydrogen production from natural gas (Clean Energy States Alliance) g) World Bank Sugar Monthly Price (Index Mundi, July 2015, no by-product credit) h) EIA June 2015 reported industrial electricity costs, West North Central average (IA, KS, MN, MO, NE, ND, SD) i) Estimate (production cost) from moderate scale on-site Calcor Process (2001 Report by DNV) j) Oak Ridge National Lab list of heating values for gases, liquids, and solids k) Values of heats of combustion l) Heating Value of Gases, EIA

Table 1.4: Commercial scale biochemical production Excluding 1st generation ethanol. Actual plant startup production typically falls well below nameplate capacity. NR = not reported. Nameplate capacity Startup Company Chemical Organism (1000 metric tons) (yr) Facility location Beta Renewable Ethanol (cellulosic) NR 40 2013 Crescentino, Italy s DSM-Poet Ethanol (cellulosic) Yeast 60 2014 Emmetsburg, IA Abengoa Ethanol (cellulosic) Yeast 75 2014 Hugoton, KS Zymomonas Dupont Ethanol (cellulosic) 82 2015 Nevada, IA mobilus Ethanol (from waste Clostridium Lanzatech 47 2017 Ghent, Belgium gas) autoethanogenum Cargill a Lactic acid NR 180 2002 Blair, NE Dupont Tate & 1,3-propanediol Escherichia coli 63 2006 Loudon, TN Lyle Genomatic 1,4-butanediol Escherichia coli 30 2015 Adria, Italy a Butamax Isobutanol Yeast 134 2015 Gevo Isobutanol Yeast 55 2012 Luverne, MN Lake Providence, Myriant Succinic acid Escherichia coli 14 2013 LA Sarnia, Ontario, BioAmber Succinic acid Yeast 30 2015 Canada C10-C18 oils, acid, Elevance Abiotic catalyst 180 2013 Gresik, Indonesia olefins Amyris Farnesene, farnesane Yeast 33 2012 Brotas, Brazil Solazyme Custom oils Microalgae 20 2014 Clinton, IA a) (Vink et al., 2004) Other sources: company websites, press releases, and patents

Figure 1.1: Genetically tractable extreme thermophiles and their optimum growth temperatures Any organism with an optimum temperature above 45°C is classified as a thermophile, but this range is extremely broad and extends 80 degree units up to the upper limit of life. Therefore, thermophiles have been further subdivided into moderate thermophiles that grow optimally between 45 and 70°C, extreme thermophiles are that grow optimally at 70°C and above, and hyperthermophiles grow optimally at 80°C and above.

Figure 1.2: General strategy for chromosomal gene insertion Used in most of the organisms discussed here. Upstream flanking regions (blue) and downstream flanking regions (red) are used to direct DNA for the insertion of a target gene (green). A marker gene (purple) provides resistance to a selective pressure, such as the addition of an antibiotic or the absence of an essential nutrient. In the case of a single first-crossover (from a circular plasmid), counter-selection results in a second crossover with the other set of homologous regions, resulting the loss of the plasmid backbone. With double-crossover, short homologous regions flanking the marker allow its removal. In either case, counter-selection recovers a markerless acceptor strain that can be used for subsequent insertions. This method can also be used for gene knockouts.

Figure 1.3: Temperature-shift strategy involving a hyperthermophilic host expressing more moderately thermostable recombinant enzymes Reduced temperatures result in a transition from growth to production phase. The hosts enzymes, naturally optimized for higher temperatures, become less active, and the cell growth rate stalls. Meanwhile, the recombinant enzymes from less thermophilic sources then re-fold and begin producing chemicals. Enzyme production can furthermore be coupled to temperature shift through the use of cold-induced promoter.

1.9. References

Abdel-Banat, B.A., Hoshida, H., Ano, A., Nonklang, S., and Akada, R. (2010). High-temperature fermentation: how can processes for ethanol production at high temperatures become superior to the traditional process using mesophilic yeast? Appl Microbiol Biotechnol 85, 861-867. doi: 10.1007/s00253-009-2248-5. Adams, M.W., Holden, J.F., Menon, A.L., Schut, G.J., Grunden, A.M., Hou, C., Hutchins, A.M., Jenney, F.E., Jr., Kim, C., Ma, K., Pan, G., Roy, R., Sapra, R., Story, S.V., and Verhagen, M.F. (2001). Key role for sulfur in peptide metabolism and in regulation of three hydrogenases in the hyperthermophilic archaeon Pyrococcus furiosus. J Bacteriol 183, 716-724. doi: 10.1128/JB.183.2.716-724.2001. Ahring, B.K. (1995). Methanogenesis in thermophilic biogas reactors. Antonie van Leeuwenhoek 67, 91-102. doi: 10.1007/BF00872197. Ahring, B.K., Licht, D., Schmidt, A.S., Sommer, P., and Thomsen, A.B. (1999). Production of ethanol from wet oxidised wheat straw by Thermoanaerobacter mathranii. Bioresource Technol 68, 3-9. doi: 10.1016/S0960-8524(98)00077-7. Albers, S.-V., and Siebers, B. (2014). "The Family Sulfolobaceae," in The Prokaryotes, ed. E.E.A. Rosenberg. (Berlin: Springer-Verlag), 323-346. Albers, S.V., and Driessen, A.J. (2008). Conditions for gene disruption by homologous recombination of exogenous DNA into the Sulfolobus solfataricus genome. Archaea 2, 145-149. Albers, S.V., Jonuscheit, M., Dinkelaker, S., Urich, T., Kletzin, A., Tampe, R., Driessen, A.J., and Schleper, C. (2006). Production of recombinant and tagged proteins in the hyperthermophilic archaeon Sulfolobus solfataricus. Appl Environ Microbiol 72, 102- 111. doi: 10.1128/AEM.72.1.102-111.2006. Amend, J.P., and Shock, E.L. (2001). Energetics of overall metabolic reactions of thermophilic and hyperthermophilic archaea and bacteria. FEMS Microbiol Rev 25, 175-243. Atomi, H., Fukui, T., Kanai, T., Morikawa, M., and Imanaka, T. (2004). Description of Thermococcus kodakaraensis sp. nov., a well studied hyperthermophilic archaeon previously reported as Pyrococcus sp. KOD1. Archaea 1, 263-267. Atomi, H., Sato, T., and Kanai, T. (2011). Application of hyperthermophiles and their enzymes. Curr Opin Microbiol 22, 618-626. doi: DOI 10.1016/j.copbio.2011.06.010. Aucelli, T., Contursi, P., Girfoglio, M., Rossi, M., and Cannio, R. (2006). A spreadable, non- integrative and high copy number shuttle vector for Sulfolobus solfataricus based on the genetic element pSSVx from Sulfolobus islandicus. Nucleic Acids Res 34, e114. doi: 10.1093/nar/gkl615. Auernik, K.S., and Kelly, R.M. (2010). Physiological versatility of the extremely thermoacidophilic archaeon Metallosphaera sedula supported by transcriptomic analysis of heterotrophic, autotrophic, and mixotrophic growth. Appl Environ Microbiol 76, 931- 935. doi: 10.1128/AEM.01336-09. Auernik, K.S., Maezato, Y., Blum, P.H., and Kelly, R.M. (2008). The genome sequence of the metal-mobilizing, extremely thermoacidophilic archaeon Metallosphaera sedula provides insights into bioleaching-associated metabolism. Appl Environ Microbiol 74, 682-692. doi: 10.1128/AEM.02019-07.

Averhoff, B., and Muller, V. (2010). Exploring research frontiers in microbiology: recent advances in halophilic and thermophilic extremophiles. Res Microbiol 161, 506-514. doi: 10.1016/j.resmic.2010.05.006. Bartlett, J.M., and Stirling, D. (2003). A short history of the polymerase chain reaction. Methods Mol Biol 226, 3-6. doi: 10.1385/1-59259-384-4:3. Basen, M., Schut, G.J., Nguyen, D.M., Lipscomb, G.L., Benn, R.A., Prybol, C.J., Vaccaro, B.J., Poole, F.L., Kelly, R.M., and Adams, M.W.W. (2014). Single gene insertion drives bioalcohol production by a thermophilic archaeon. Proc Natl Acad Sci U S A 111, 17618- 17623. doi: 10.1073/pnas.1413789111. Basen, M., Sun, J., and Adams, M.W. (2012). Engineering a hyperthermophilic archaeon for temperature-dependent product formation. mBio 3, e00053-00012. doi: 10.1128/mBio.00053-12. Beckner, M., Ivey, M.L., and Phister, T.G. (2011). Microbial contamination of fuel ethanol fermentations. Lett Appl Microbiol 53, 387-394. doi: 10.1111/j.1472-765X.2011.03124.x. Berg, I.A., Kockelkorn, D., Buckel, W., and Fuchs, G. (2007). A 3-hydroxypropionate/4- hydroxybutyrate autotrophic carbon dioxide assimilation pathway in archaea. Science 318, 1782-1786. doi: 10.1126/science.1149976. Berkner, S., Grogan, D., Albers, S.-V., and Lipps, G. (2007). Small multicopy, non-integrative shuttle vectors based on the plasmid pRN1 for Sulfolobus acidocaldarius and Sulfolobus solfataricus, model organisms of the (cren-)archaea. Nucleic Acids Res 35, e88-e88. doi: 10.1093/nar/gkm449. Berkner, S., and Lipps, G. (2008). Genetic tools for Sulfolobus spp.: vectors and first applications. Arch Microbiol 190, 217-230. doi: 10.1007/s00203-008-0392-4. Berkner, S., Wlodkowski, A., Albers, S.-V., and Lipps, G. (2010). Inducible and constitutive promoters for genetic systems in Sulfolobus acidocaldarius. Extremophiles 14, 249-259. doi: 10.1007/s00792-010-0304-9. Bhosale, S.H., Rao, M.B., and Deshpande, V.V. (1996). Molecular and industrial aspects of glucose isomerase. Microbiol Rev 60, 280-300. Bize, A., Karlsson, E.A., Ekefjard, K., Quax, T.E., Pina, M., Prevost, M.C., Forterre, P., Tenaillon, O., Bernander, R., and Prangishvili, D. (2009). A unique virus release mechanism in the Archaea. Proc Natl Acad Sci U S A 106, 11306-11311. doi: 10.1073/pnas.0901238106. Blanch, H.W., and Clark, D.S. (1997). Biochemical Engineering. New York, NY: Marcel Dekker, Inc. Bridger, S.L., Lancaster, W.A., Poole, F.L., Schut, G.J., and Adams, M.W.W. (2012). Genome sequencing of a genetically tractable Pyrococcus furiosus strain reveals a highly dynamic genome. J Bacteriol 194, 4097-4106. doi: 10.1128/jb.00439-12. Brock, T.D., Brock, K.M., Belly, R.T., and Weiss, R.L. (1972). Sulfolobus: A new genus of sulfur-oxidizing bacteria living at low pH and high temperature. Arch Mikrobiol 84, 54- 68. doi: 10.1007/BF00408082. Brouns, S.J., Wu, H., Akerboom, J., Turnbull, A.P., De Vos, W.M., and Van Der Oost, J. (2005). Engineering a selectable marker for hyperthermophiles. J Biol Chem 280, 11422-11431. doi: 10.1074/jbc.M413623200. Bruggemann, H., and Chen, C. (2006). Comparative genomics of Thermus thermophilus: Plasticity of the megaplasmid and its contribution to a thermophilic lifestyle. J Biotechnol 124, 654-661. doi: 10.1016/j.jbiotec.2006.03.043.

Carr, J.F., Danziger, M.E., Huang, A.L., Dahlberg, A.E., and Gregory, S.T. (2015). Engineering the genome of Thermus thermophilus using a counter-selectable marker. J Bacteriol, JB.02384-02314. doi: 10.1128/JB.02384-14. Cava, F., Hidalgo, A., and Berenguer, J. (2009). Thermus thermophilus as biological model. Extremophiles 13, 213-231. doi: 10.1007/s00792-009-0226-6. Cha, M., Chung, D.W., Elkins, J.G., Guss, A.M., and Westpheling, J. (2013). Metabolic engineering of Caldicellulosiruptor bescii yields increased hydrogen production from lignocellulosic biomass. Biotechnol Biofuels 6:85. Chen, L., Brugger, K., Skovgaard, M., Redder, P., She, Q., Torarinsson, E., Greve, B., Awayez, M., Zibat, A., Klenk, H.P., and Garrett, R.A. (2005). The genome of Sulfolobus acidocaldarius, a model organism of the Crenarchaeota. J Bacteriol 187, 4992-4999. doi: 10.1128/JB.187.14.4992-4999.2005. Chhabra, S.R., Shockley, K.R., Conners, S.B., Scott, K.L., Wolfinger, R.D., and Kelly, R.M. (2003). Carbohydrate-induced differential gene expression patterns in the hyperthermophilic bacterium Thermotoga maritima. J Biol Chem 278, 7540-7552. doi: 10.1074/jbc.M211748200. Chung, D., Cha, M., Guss, A.M., and Westpheling, J. (2014). Direct conversion of plant biomass to ethanol by engineered Caldicellulosiruptor bescii. Proc Natl Acad Sci U S A 111, 8931-8936. doi: 10.1073/pnas.1402210111. Chung, D., Farkas, J., Huddleston, J.R., Olivar, E., and Westpheling, J. (2012). Methylation by a unique α-class N4-cytosine methyltransferase is required for DNA transformation of Caldicellulosiruptor bescii DSM6725. PLOS One 7, e43844. doi: 10.1371/journal.pone.0043844. Chung, D.W., Farkas, J., and Westpheling, J. (2013). Overcoming restriction as a barrier to DNA transformation in Caldicellulosiruptor species results in efficient marker replacement. Biotechnol Biofuels 6:82. Conners, S.B., Montero, C.I., Comfort, D.A., Shockley, K.R., Johnson, M.R., Chhabra, S.R., and Kelly, R.M. (2005). An expression-driven approach to the prediction of carbohydrate transport and utilization regulons in the hyperthermophilic bacterium Thermotoga maritima. J Bacteriol 187, 7267-7282. doi: 10.1128/JB.187.21.7267-7282.2005. De Rosa, M., Gambacorta, A., Bu'lock, J. (1975). Extremely thermophilic acidophilic bacteria convergent with Sulfolobus acidocaldarius. J Gen Microbiol 86, 156-164. doi: 10.1099/00221287-86-1-156. Deng, L., Zhu, H.J., Chen, Z.J., Liang, Y.X., and She, Q.X. (2009). Unmarked gene deletion and host-vector system for the hyperthermophilic crenarchaeon Sulfolobus islandicus. Extremophiles 13, 735-746. doi: DOI 10.1007/s00792-009-0254-2. Du Plessis, C.A., Batty, J.D., Dew, D.W. (2007). "Commercial applications of thermophile bioleaching," in Biomining, ed. D.E. Rawlings, Johnson, D.B. (Berlin Heidelberg: Springer-Verlag). Dupont/Danisco (2012). DuPont nutrition & health hurdles the process challenges for pizza cheese [Online]. Available: http://www.danisco.com/about-dupont/news/news- archive/2012/dupont-hurdles-the-process-challenges-for-pizza-cheese/ [Accessed August 2015]. Elmore, J.R., Yokooji, Y., Sato, T., Olson, S., Claiborne V.C. Glover, Iii, Graveley, B.R., Atomi, H., Terns, R.M., and Terns, M.P. (2013). Programmable plasmid interference by the

CRISPR-Cas system in Thermococcus kodakarensis. RNA Biol 10, 828-840. doi: 10.4161/rna.24084. Farkas, J., Stirrett, K., Lipscomb, G.L., Nixon, W., Scott, R.A., Adams, M.W.W., and Westpheling, J. (2012). Recombinogenic properties of Pyrococcus furiosus strain COM1 enable rapid selection of targeted mutants. Appl Environ Microbiol 78, 4669-4676. doi: 10.1128/aem.00936-12. Fiala, G., and Stetter, K. (1986). Pyrococcus furiosus sp. nov. represents a novel genus of marine heterotrophic archaebacteria growing optimally at 100°C. Arch Microbiol 145, 56-61. doi: 10.1007/bf00413027. Frock, A.D., Gray, S.R., and Kelly, R.M. (2012). Hyperthermophilic Thermotoga species differ with respect to specific carbohydrate transporters and glycoside hydrolases. Appl Environ Microbiol 78, 1978-1986. doi: 10.1128/AEM.07069-11. Frock, A.D., and Kelly, R.M. (2012). Extreme thermophiles: moving beyond single-enzyme biocatalysis. Curr Opin Chem Eng 1, 363-372. doi: 10.1016/j.coche.2012.07.003. Fukui, T., Atomi, H., Kanai, T., Matsumi, R., Fujiwara, S., and Imanaka, T. (2005). Complete genome sequence of the hyperthermophilic archaeon Thermococcus kodakaraensis KOD1 and comparison with Pyrococcus genomes. Genome Res 15, 352-363. doi: 10.1101/gr.3003105. Girfoglio, M., Rossi, M., and Cannio, R. (2012). Cellulose degradation by Sulfolobus solfataricus requires a cell-anchored endo-beta-1-4-glucanase. J Bacteriol 194, 5091- 5100. doi: 10.1128/JB.00672-12. Grens, K. (2015). There's CRISPR in your yogurt. The Scientist 29, 1. Greve, B., Jensen, S., Brugger, K., Zillig, W., and Garrett, R.A. (2004). Genomic comparison of archaeal conjugative plasmids from Sulfolobus. Archaea 1, 231-239. Grissa, I., Vergnaud, G., and Pourcel, C. (2007). The CRISPRdb database and tools to display CRISPRs and to generate dictionaries of spacers and repeats. BMC Bioinformatics 8, 172. doi: 10.1186/1471-2105-8-172. Grogan, D.W. (1989). Phenotypic characterization of the archaebacterial genus Sulfolobus - Comparison of 5 wild-type strains. Journal of Bacteriology 171, 6710-6719. Guo, L., Brugger, K., Liu, C., Shah, S.A., Zheng, H., Zhu, Y., Wang, S., Lillestol, R.K., Chen, L., Frank, J., Prangishvili, D., Paulin, L., She, Q., Huang, L., and Garrett, R.A. (2011). Genome analyses of Icelandic strains of Sulfolobus islandicus, model organisms for genetic and virus-host interaction studies. J Bacteriol 193, 1672-1680. doi: 10.1128/JB.01487-10. Han, D., Norris, S.M., and Xu, Z. (2012). Construction and transformation of a Thermotoga-E. coli shuttle vector. BMC Biotechnology 12, 2-2. doi: 10.1186/1472-6750-12-2. Han, D., Xu, H., Puranik, R., and Xu, Z. (2014). Natural transformation of Thermotoga sp. strain RQ7. BMC Biotechnology 14, 39-39. doi: 10.1186/1472-6750-14-39. Harnvoravongchai, P., Kobori, H., Orita, I., Nakamura, S., Imanaka, T., and Fukui, T. (2014). Characterization and gene deletion analysis of four homologues of group 3 pyridine nucleotide disulfide oxidoreductases from Thermococcus kodakarensis. Extremophiles 18, 603-616. doi: 10.1007/s00792-014-0643-z. Harris, J.M., Epting, K.L., and Kelly, R.M. (2010). N-terminal fusion of a hyperthermophilic chitin-binding domain to xylose isomerase from Thermotoga neapolitana enhances kinetics and thermostability of both free and immobilized enzymes. Biotechnol Prog 26, 993-1000. doi: 10.1002/btpr.416.

Hashimoto, Y., Yano, T., Kuramitsu, S., and Kagamiyama, H. (2001). Disruption of Thermus thermophilus genes by homologous recombination using a thermostable kanamycin- resistant marker. FEBS Lett 506, 231-234. Hawkins, A.B., Lian, H., Zeldes, B.M., Loder, A.J., Lipscomb, G.L., Schut, G.J., Keller, M.W., Adams, M.W.W., and Kelly, R.M. (2015). Bioprocessing analysis of Pyrococcus furiosus strains engineered for CO2-based 3-hydroxypropionate production. Biotechnol Bioeng 112, 1533-1543. doi: 10.1002/bit.25584. Hawkins, A.S., Mcternan, P.M., Lian, H., Kelly, R.M., and Adams, M.W.W. (2013). Biological conversion of carbon dioxide and hydrogen into liquid fuels and industrial chemicals. Curr Opin Biotech 24, 376-384. doi: 10.1016/j.copbio.2013.02.017. Henne, A., Bruggemann, H., Raasch, C., Wiezer, A., Hartsch, T., Liesegang, H., Johann, A., Lienard, T., Gohl, O., Martinez-Arias, R., Jacobi, C., Starkuviene, V., Schlenczeck, S., Dencker, S., Huber, R., Klenk, H.P., Kramer, W., Merkl, R., Gottschalk, G., and Fritz, H.J. (2004). The genome sequence of the extreme thermophile Thermus thermophilus. Nat Biotechnol 22, 547-553. doi: 10.1038/nbt956. Hidaka, Y., Hasegawa, M., Nakahara, T., and Hoshino, T. (1994). The entire population of Thermus thermophilus cells is always competent at any growth phase. Biosci Biotech Biochem 58, 1338-1339. doi: 10.1271/bbb.58.1338. Hidalgo, A., Betancor, L., Moreno, R., Zafra, O., Cava, F., Fernandez-Lafuente, R., Guisan, J.M., and Berenguer, J. (2004). Thermus thermophilus as a cell factory for the production of a thermophilic Mn-dependent catalase which fails to be synthesized in an active form in Escherichia coli. Appl Environ Microbiol 70, 3839-3844. doi: 10.1128/AEM.70.7.3839-3844.2004. Hileman, T.H., and Santangelo, T.J. (2012). Genetics techniques for Thermococcus kodakarensis. Front Microbiol 3, 195-195. doi: 10.3389/fmicb.2012.00195. Hopkins, R.C., Sun, J., Jenney, F.E., Jr., Chandrayan, S.K., Mcternan, P.M., and Adams, M.W.W. (2011). Homologous expression of a subcomplex of Pyrococcus furiosus hydrogenase that interacts with pyruvate ferredoxin oxidoreductase. PLOS ONE 6, e26569. doi: 10.1371/journal.pone.0026569. Hoseki, J., Yano, T., Koyama, Y., Kuramitsu, S., and Kagamiyama, H. (1999). Directed evolution of thermostable kanamycin-resistance gene: a convenient selection marker for Thermus thermophilus. J Biochem 126, 951-956. Huber, G., Spinnler, C., Gambacorta, A., and Stetter, K.O. (1989). Metallosphaera sedula gen, and sp. nov. represents a new genus of aerobic, metal-mobilizing, thermoacidophilic archaebacteria. Syst Appl Microbiol 12, 38-47. doi: 10.1016/S0723-2020(89)80038-4. Huber, R., Langworthy, T.A., König, H., Thomm, M., Woese, C.R., Sleytr, U.B., and Stetter, K.O. (1986). Thermotoga maritima sp. nov. represents a new genus of unique extremely thermophilic eubacteria growing up to 90 C. Arch Microbiol 144, 324-333. Imanaka, T., and Atomi, H. (2002). Catalyzing “hot” reactions: enzymes from hyperthermophilic archaea. Chem Rec 2, 149-163. doi: 10.1002/tcr.10023. Jannasch, H.W., Huber, R., Belkin, S., and Stetter, K.O. (1988). Thermotoga neapolitana new species of the extremely thermophilic eubacterial genus Thermotoga. Arch Microbiol 150, 103-104. Jones, D.T., and Woods, D.R. (1986). Acetone-butanol fermenation revisited. Microbiol Rev 50, 484-524.

Jonuscheit, M., Martusewitsch, E., Stedman, K.M., and Schleper, C. (2003). A reporter gene system for the hyperthermophilic archaeon Sulfolobus solfataricus based on a selectable and integrative shuttle vector. Mol Microbiol 48, 1241-1252. doi: 10.1046/j.1365- 2958.2003.03509.x. Jund, R., and Lacroute, F. (1970). Genetic and physiological aspects of resistance to 5- fluoropyrimidines in Saccharomyces cerevisiae. J Bacteriol 102, 607-615. Kanai, T., Simons, J.-R., Tsukamoto, R., Nakajima, A., Omori, Y., Matsuoka, R., Beppu, H., Imanaka, T., and Atomi, H. (2015). Overproduction of the membrane-bound [NiFe]- hydrogenase in Thermococcus kodakarensis and its effect on hydrogen production. Front Microbiol 6, article 847. doi: 10.3389/fmicb.2015.00847. Kataeva, I.A., Yang, S.J., Dam, P., Poole, F.L., 2nd, Yin, Y., Zhou, F., Chou, W.C., Xu, Y., Goodwin, L., Sims, D.R., Detter, J.C., Hauser, L.J., Westpheling, J., and Adams, M.W. (2009). Genome sequence of the anaerobic, thermophilic, and cellulolytic bacterium "Anaerocellum thermophilum" DSM 6725. J Bacteriol 191, 3760-3761. doi: 10.1128/JB.00256-09. Keller, M., Loder, A.J., Basen, M., Izquierdo, J., Kelly, R.M., and Adams, M.W. (2014). Production of lignofuels and electrofuels by extremely thermophilic microbes. Biofuels 5, 499-515. Keller, M.W., Lipscomb, G.L., Loder, A.J., Schut, G.J., Kelly, R.M., and Adams, M.W. (2015). A hybrid synthetic pathway for butanol production by a hyperthermophilic microbe. Metab Eng 27, 101-106. Keller, M.W., Schut, G.J., Lipscomb, G.L., Menon, A.L., Iwuchukwu, I.J., Leuko, T.T., Thorgersen, M.P., Nixon, W.J., Hawkins, A.S., Kelly, R.M., and Adams, M.W.W. (2013). Exploiting microbial hyperthermophilicity to produce an industrial chemical, using hydrogen and carbon dioxide. Proc Natl Acad Sci U S A 110, 5840-5845. doi: 10.1073/pnas.1222607110. Kim, M.S., Bae, S.S., Kim, Y.J., Kim, T.W., Lim, J.K., Lee, S.H., Choi, A.R., Jeon, J.H., Lee, J.H., Lee, H.S., and Kang, S.G. (2013). CO-dependent H2 production by genetically engineered Thermococcus onnurineus NA1. Appl Environ Microbiol 79, 2048-2053. doi: 10.1128/AEM.03298-12. Kockelkorn, D., and Fuchs, G. (2009). Malonic Semialdehyde Reductase, Succinic Semialdehyde Reductase, and Succinyl-Coenzyme A Reductase from Metallosphaera sedula: Enzymes of the Autotrophic 3-Hydroxypropionate/4-Hydroxybutyrate Cycle in Sulfolobales. J Bacteriol 191, 6352-6362. doi: 10.1128/jb.00794-09. Koyama, Y., Hoshino, T., Tomizuka, N., and Furukawa, K. (1986). Genetic transformation of the extreme thermophile Thermus thermophilus and of other Thermus spp. J Bacteriol 166, 338-340. Kreuzer, M., Schmutzler, K., Waege, I., Thomm, M., and Hausner, W. (2013). Genetic engineering of Pyrococcus furiosus to use chitin as a carbon source. BMC Biotechnol 13, 9. Krooth, R., Hsiao, W.-L., and Potvin, B. (1979). Resistance to 5-fluoroorotic acid and pyrimidine auxotrophy: A new bidirectional selective system for mammalian cells. Somat Cell Genet 5, 551-569. doi: 10.1007/BF01542694. Ladkau, N., Schmid, A., and Bühler, B. (2014). The microbial cell — functional unit for energy dependent multistep biocatalysis. Curr Opin Biotech 30, 178-189. doi: 10.1016/j.copbio.2014.06.003.

Lam, W.L., and Doolittle, W.F. (1989). Shuttle vectors for the archaebacterium Halobacterium volcanii. Proc Natl Acad Sci U S A 86, 5478-5482. Larsen, L., Nielsen, P., and Ahring, B.K. (1997). Thermoanaerobacter mathranii sp. nov., an ethanol-producing, extremely thermophilic anaerobic bacterium from a hot spring in Iceland. Arch Microbiol 168, 114-119. Lee, J.Y., Kang, C.D., Lee, S.H., Park, Y.K., and Cho, K.M. (2015). Engineering cellular redox balance in Saccharomyces cerevisiae for improved production of L-lactic acid. Biotechnol Bioeng 112, 751-758. doi: 10.1002/bit.25488. Lehmann, M., and Wyss, M. (2001). Engineering proteins for thermostability: the use of sequence alignments versus rational design and directed evolution. Curr Opin Biotech 12, 371-375. Leis, B., Angelov, A., Li, H., and Liebl, W. (2014). Genetic analysis of lipolytic activities in Thermus thermophilus HB27. J Biotechnol 191, 150-157. doi: 10.1016/j.jbiotec.2014.07.448. Liao, H., Mckenzie, T., and Hageman, R. (1986). Isolation of a thermostable enzyme variant by cloning and selection in a thermophile. Proc Natl Acad Sci U S A 83, 576-580. Lipscomb, G.L., Schut, G.J., Thorgersen, M.P., Nixon, W.J., Kelly, R.M., and Adams, M.W.W. (2014). Engineering hydrogen gas production from formate in a hyperthermophile by heterologous production of an 18-subunit membrane-bound complex. J Biol Chem 289, 2873-2879. doi: 10.1074/jbc.M113.530725. Lipscomb, G.L., Stirrett, K., Schut, G.J., Yang, F., Jenney, F.E., Scott, R.A., Adams, M.W.W., and Westpheling, J. (2011). Natural competence in the hyperthermophilic archaeon Pyrococcus furiosus facilitates genetic manipulation: construction of markerless deletions of genes encoding the two cytoplasmic hydrogenases. Appl Environ Microbiol 77, 2232- 2238. doi: 10.1128/aem.02624-10. Liu, J., Qi, H., Wang, C., and Wen, J. (2015). Model-driven intracellular redox status modulation for increasing isobutanol production in Escherichia coli. Biotechnol Biofuels 8, 108. doi: 10.1186/s13068-015-0291-2. M P Taylor, L.V.Z. (2011). Genetic tool development underpins recent advances in thermophilic whole-cell biocatalysts. J Microbial Biotech 4, 438-448. doi: 10.1111/j.1751- 7915.2010.00246.x. Maezato, Y., Johnson, T., Mccarthy, S., Dana, K., and Blum, P. (2012). Metal resistance and lithoautotrophy in the extreme thermoacidophile Metallosphaera sedula. J Bacteriol 194, 6856-6863. doi: 10.1128/JB.01413-12. Maris, A.J.a.V., Konings, W.N., Dijken, J.P.V., and Pronk, J.T. (2004). Microbial export of lactic and 3-hydroxypropanoic acid: implications for industrial fermentation processes. Metab Eng 6, 245-255. doi: 10.1016/j.ymben.2004.05.001. Matsumi, R., Manabe, K., Fukui, T., Atomi, H., and Imanaka, T. (2007). Disruption of a sugar transporter gene cluster in a hyperthermophilic archaeon using a host-marker system based on antibiotic resistance. J Bacteriol 189, 2683-2691. doi: Doi 10.1128/Jb.01692- 06. Matsumura, M., and Aiba, S. (1985). Screening for thermostable mutant of kanamycin nucleotidyltransferase by the use of a transformation system for a thermophile, Bacillus stearothermophilus. J Biol Chem 260, 15298-15303.

Moreno, R., Haro, A., Castellanos, A., and Berenguer, J. (2005). High-level overproduction of His-tagged Tth DNA polymerase in Thermus thermophilus. Appl Environ Microbiol 71, 591-593. doi: 10.1128/AEM.71.1.591-593.2005. Morikawa, M., Izawa, Y., Rashid, N., Hoaki, T., and Imanaka, T. (1994). Purification and characterization of a thermostable thiol protease from a newly isolated hyperthermophilic Pyrococcus sp. Appl Environ Microbiol 60, 4559-66 Mueller, M., Takemasa, R., Schwarz, A., Atomi, H., and Nidetzky, B. (2009). “Short-chain” α- 1,4-glucan phosphorylase having a truncated N-terminal domain: Functional expression and characterization of the enzyme from Sulfolobus solfataricus. BBA-Proteins Proteom 1794, 1709-1714. doi: 10.1016/j.bbapap.2009.08.006. Nakamura, A., Takakura, Y., Kobayashi, H., and Hoshino, T. (2005). In vivo directed evolution for thermostabilization of Escherichia coli hygromycin B phosphotransferase and the use of the gene as a selection marker in the host-vector system of Thermus thermophilus. J Biosci Bioeng 100, 158-163. doi: 10.1263/jbb.100.158. Nelson, K.E., Clayton, R.A., Gill, S.R., Gwinn, M.L., Dodson, R.J., Haft, D.H., Hickey, E.K., Peterson, J.D., Nelson, W.C., Ketchum, K.A., Mcdonald, L., Utterback, T.R., Malek, J.A., Linher, K.D., Garrett, M.M., Stewart, A.M., Cotton, M.D., Pratt, M.S., Phillips, C.A., Richardson, D., Heidelberg, J., Sutton, G.G., Fleischmann, R.D., Eisen, J.A., White, O., Salzberg, S.L., Smith, H.O., Venter, J.C., and Fraser, C.M. (1999). Evidence for lateral gene transfer between archaea and bacteria from genome sequence of Thermotoga maritima. Nature 399, 323-329. doi: 10.1038/20601. Nixon, A., and Norris, P.R. (1992). Autotrophic growth and inorganic sulphur compound oxidation by Sulfolobus sp. in chemostat culture. Arch Microbiol 157, 155-160. doi: 10.1007/BF00245284. Ohtani, N., Tomita, M., and Itaya, M. (2010). An extreme thermophile, Thermus thermophilus, is a polyploid bacterium. J Bacteriol 192, 5499-5505. doi: 10.1128/JB.00662-10. Oshima, T., and Imahori, K. (1974). Description of Thermus thermophilus (Yoshida and Oshima) comb. nov., a nonsporulating thermophilic bacterium from a Japanese thermal spa. Int J Syst Bacteriol 24, 102-112. doi: 10.1099/00207713-24-1-102. Peng, N., Deng, L., Mei, Y., Jiang, D., Hu, Y., Awayez, M., Liang, Y., and She, Q. (2012). A synthetic arabinose-inducible promoter confers high levels of recombinant protein expression in hyperthermophilic archaeon Sulfolobus islandicus. Appl Environ Microbiol 78, 5630-5637. doi: 10.1128/AEM.00855-12. Pietilä, M.K., Demina, T.A., Atanasova, N.S., Oksanen, H.M., and Bamford, D.H. (2014). Archaeal viruses and bacteriophages: comparisons and contrasts. Trends Microbiol 22, 334-344. doi: 10.1016/j.tim.2014.02.007. Pina, M., Bize, A., Forterre, P., and Prangishvili, D. (2011). The archeoviruses. FEMS Microbiol Rev 35, 1035-1054. doi: 10.1111/j.1574-6976.2011.00280.x. Prangishvili, D. (2013). The wonderful world of archaeal viruses. Annu Rev Microbiol 67, 565- 585. doi: 10.1146/annurev-micro-092412-155633. Qing, G., Ma, L.C., Khorchid, A., Swapna, G.V., Mal, T.K., Takayama, M.M., Xia, B., Phadtare, S., Ke, H., Acton, T., Montelione, G.T., Ikura, M., and Inouye, M. (2004). Cold-shock induced high-yield protein production in Escherichia coli. Nat Biotechnol 22, 877-882. doi: 10.1038/nbt984.

Ramirez-Arcos, S., Fernandez-Herrero, L.A., Marin, I., and Berenguer, J. (1998). Anaerobic growth, a property horizontally transferred by an Hfr-like mechanism among extreme thermophiles. J Bacteriol 180, 3137-3143. Rfa (2014). World Fuel Ethanol Production [Online]. Renewable Fuels Association. Available: http://ethanolrfa.org/pages/World-Fuel-Ethanol-Production [Accessed 19 August 2015]. Romanos, M.A., Scorer, C.A., and Clare, J.J. (1992). Foreign gene expression in yeast: a review. Yeast 8, 423-488. doi: 10.1002/yea.320080602. Salzer, R., Joos, F., and Averhoff, B. (2015a). Different effects of MglA and MglB on pilus- mediated functions and natural competence in Thermus thermophilus. Extremophiles 19, 261-267. doi: 10.1007/s00792-014-0711-4. Salzer, R., Kern, T., Joos, F., and Averhoff, B. (2015b). The Thermus thermophilus comEA/comEC operon is associated with DNA binding and regulation of the DNA translocator and type IV pili. Environ Microbiol. doi: 10.1111/1462-2920.12820. Santangelo, T.J., Cubonova, L., and Reeve, J.N. (2010). Thermococcus kodakarensis genetics: TK1827-encoded beta-glycosidase, new positive-selection protocol, and targeted and repetitive deletion technology. Appl Environ Microbiol 76, 1044-1052. doi: Doi 10.1128/Aem.02497-09. Santangelo, T.J., Čuboňová, L.U., and Reeve, J.N. (2011). Deletion of alternative pathways for reductant recycling in Thermococcus kodakarensis increases hydrogen production. Mol Microbiol 81, 897-911. doi: 10.1111/j.1365-2958.2011.07734.x. Santangelo, T.J., and Reeve, D.J.N. (2011). "Genetic tools and manipulations of the hyperthermophilic heterotrophic archaeon Thermococcus kodakarensis," ed. K. Horikoshi. Springer, Japan, 567-582. Sato, T., Fukui, T., Atomi, H., and Imanaka, T. (2003). Targeted gene disruption by homologous recombination in the hyperthermophilic archaeon Thermococcus kodakaraensis KOD1. J Bacteriol 185, 210-220. doi: Doi 10.1128/Jb.185.1.210-220.2003. Sato, T., Fukui, T., Atomi, H., and Imanaka, T. (2005). Improved and versatile transformation system allowing multiple genetic manipulations of the hyperthermophilic archaeon Thermococcus kodakaraensis. Appl Environ Microbiol 71, 3889-3899. doi: Doi 10.1128/Aem.71.7.3889-3899.2005. Schut, G.J., Boyd, E.S., Peters, J.W., and Adams, M.W.W. (2013). The modular respiratory complexes involved in hydrogen and sulfur metabolism by heterotrophic hyperthermophilic archaea and their evolutionary implications. FEMS Microbiol Rev 37, 182-203. doi: 10.1111/j.1574-6976.2012.00346.x. Schwarzenlander, C., and Averhoff, B. (2006). Characterization of DNA transport in the thermophilic bacterium Thermus thermophilus HB27. FEBS J 273, 4210-4218. doi: 10.1111/j.1742-4658.2006.05416.x. Scott, I.M., Rubinstein, G.M., Lipscomb, G.L., Basen, M., Schut, G.J., Rhaesa, A.M., Lancaster, W.A., Poole, F.L., 2nd, Kelly, R.M., and Adams, M.W. (In press). A new class of tungsten-containing oxidoreductase in the genus of the plant biomass-degrading, thermophilic bacteria Caldicellulosiruptor. Appl Environ Microbiol. 81(20), 7339-47 doi: 10.1128/AEM.01634-15. She, Q., Singh, R.K., Confalonieri, F., Zivanovic, Y., Allard, G., Awayez, M.J., Chan-Weiher, C.C., Clausen, I.G., Curtis, B.A., De Moors, A., Erauso, G., Fletcher, C., Gordon, P.M., Heikamp-De Jong, I., Jeffries, A.C., Kozera, C.J., Medina, N., Peng, X., Thi-Ngoc, H.P., Redder, P., Schenk, M.E., Theriault, C., Tolstrup, N., Charlebois, R.L., Doolittle, W.F.,

Duguet, M., Gaasterland, T., Garrett, R.A., Ragan, M.A., Sensen, C.W., and Van Der Oost, J. (2001). The complete genome of the crenarchaeon Sulfolobus solfataricus P2. Proc Natl Acad Sci U S A 98, 7835-7840. doi: 10.1073/pnas.141222098. She, Q., Zhang, C., Deng, L., Peng, N., Chen, Z., and Liang, Y.X. (2009). Genetic analyses in the hyperthermophilic archaeon Sulfolobus islandicus. Biochem Soc Trans 37, 92-96. doi: 10.1042/BST0370092. Simões, R.G., Diogo, H.P., Dias, A., Oliveira, M.C., Cordeiro, C., Bernardes, C.E.S., and Minas Da Piedade, M.E. (2014). Thermal stability of simvastatin under different atmospheres. J Pharm Sci 103, 241-248. doi: 10.1002/jps.23787. Singh, R.K., Tiwari, M.K., Singh, R., and Lee, J.K. (2013). From protein engineering to immobilization: promising strategies for the upgrade of industrial enzymes. Int J Mol Sci 14, 1232-1277. doi: 10.3390/ijms14011232. Skinner, K.A., and Leathers, T.D. (2004). Bacterial contaminants of fuel ethanol production. J Ind Microbiol Biotechnol 31, 401-408. doi: 10.1007/s10295-004-0159-0. Songstad, D.D., Lakshmanan, P., Chen, J., Gibbons, W., Hughes, S., and Nelson, R. (2009). Historical perspective of biofuels: learning from the past to rediscover the future. In Vitro Cell Dev Biol Plant 45, 189-192. doi: 10.1007/s11627-009-9218-6. Steiner, K., and Schwab, H. (2012). Recent advances in rational approaches for enzyme engineering. Comput Struct Biotechnol J 2, e201209010. doi: 10.5936/csbj.201209010. Svetlichnyi, V.A., Svetlichnaya, T.P., Chernykh, N.A., and Zavarzin, G.A. (1990). Anaerocellum thermophilum gen. nov. sp. nov.: An extreme thermophilic celluloytic eubacterium isolated from hot springs in the valley of geysers Mikrobiologiya 59, 871-879. Swarup, A., Lu, J., Dewoody, K.C., and Antoniewicz, M.R. (2014). Metabolic network reconstruction, growth characterization and 13C-metabolic flux analysis of the extremophile Thermus thermophilus HB8. Metab Eng 24, 173-180. doi: 10.1016/j.ymben.2014.05.013. Takemasa, R., Yokooji, Y., Yamatsu, A., Atomi, H., and Imanaka, T. (2011). Thermococcus kodakarensis as a host for gene expression and protein secretion. Appl Environ Microbiol 77, 2392-2398. doi: 10.1128/aem.01005-10. Tamakoshi, M., Yaoi, T., Oshima, T., and Yamagishi, A. (1999). An efficient gene replacement and deletion system for an extreme thermophile, Thermus thermophilus. FEMS Microbiol Lett 173, 431-437. Thorgersen, M.P., Lipscomb, G.L., Schut, G.J., Kelly, R.M., and Adams, M.W.W. (2014). Deletion of acetyl-CoA synthetases I and II increases production of 3-hydroxypropionate by the metabolically-engineered hyperthermophile Pyrococcus furiosus. Metab Eng 22, 83-88. doi: 10.1016/j.ymben.2013.12.006. Vane, L.M. (2008). Separation technologies for the recovery and dehydration of alcohols from fermentation broths. Biofuels, Bioproducts and Biorefining 2, 553-588. doi: 10.1002/bbb.108. Verhaart, M.R.A., Bielen, A.a.M., Oost, J.V.D., Stams, A.J.M., and Kengen, S.W.M. (2010). Hydrogen production by hyperthermophilic and extremely thermophilic bacteria and archaea: mechanisms for reductant disposal. Environ Technol 31, 993-1003. doi: 10.1080/09593331003710244. Vieille, C., and Zeikus, G.J. (2001). Hyperthermophilic enzymes: sources, uses, and molecular mechanisms for thermostability. Microbiol Mol Biol R 65, 1-43. doi: 10.1128/MMBR.65.1.1-43.2001.

Vink, E.T.H., Rábago, K.R., Glassner, D.A., Springs, B., O'connor, R.P., Kolstad, J., and Gruber, P.R. (2004). The sustainability of NatureWorks™ polylactide polymers and Ingeo™ polylactide fibers: an update of the future. Macromol Biosci 4, 551-564. doi: 10.1002/mabi.200400023. Waege, I., Schmid, G., Thumann, S., Thomm, M., and Hausner, W. (2010). Shuttle vector-based transformation system for Pyrococcus furiosus. Appl Environ Microbiol 76, 3308-3313. doi: 10.1128/AEM.01951-09. Wagner, M., Van Wolferen, M., Wagner, A., Lassak, K., Meyer, B.H., Reimann, J., and Albers, S.-V. (2012). Versatile genetic tool box for the crenarchaeote Sulfolobus acidocaldarius. Front Microbiol 3, 214-214. doi: 10.3389/fmicb.2012.00214. Weinberg, M.V., Schut, G.J., Brehm, S., Datta, S., and Adams, M.W.W. (2005). Cold shock of a hyperthermophilic archaeon: Pyrococcus furiosus exhibits multiple responses to a suboptimal growth temperature with a key role for membrane-bound glycoproteins. J Bacteriol 187, 336-348. doi: Doi 10.1128/Jb.187.1.336-348.2005. Williams, R.a.D., Smith, K.E., Welch, S.G., Micallef, J., and Sharp, R.J. (1995). DNA relatedness of thermus strains, description of Thermus brockianus sp. nov., and proposal to reestablish Thermus thermophilus (Oshima and Imahori). Int J Syst Bacteriol 45, 495- 499. doi: 10.1099/00207713-45-3-495. Wood, A.P., Kelly, D.P., and Norris, P.R. (1987). Autotrophic growth of four Sulfolobus strains on tetrathionate and the effect of organic nutrients. Arch Microbiol 146, 382-389. doi: 10.1007/BF00410940. Worsham, P., and Goldman, W. (1988). Selection and characterization of ura5 mutants of Histoplasma capsulatum. Mol Gen Genet 214, 348-352. doi: 10.1007/BF00337734. Worthington, P., Hoang, V., Perez-Pomares, F., and Blum, P. (2003). Targeted disruption of the alpha-amylase gene in the hyperthermophilic archaeon Sulfolobus solfataricus. J Bacteriol 185, 482-488. doi: Doi 10.1128/Jb.185.2.482-488.2003. Xu, H., Han, D., and Xu, Z. ((in press)). Expression of Heterologous Cellulases in Thermotoga sp. Strain RQ2. Biomed Research International, 11. doi: 10.1155/2015/304523. Yang, S.J., Kataeva, I., Hamilton-Brehm, S.D., Engle, N.L., Tschaplinski, T.J., Doeppke, C., Davis, M., Westpheling, J., and Adams, M.W.W. (2009). Efficient degradation of lignocellulosic plant biomass, without pretreatment, by the thermophilic anaerobe "Anaerocellum thermophilum" DSM 6725. Appl Environ Microbiol 75, 4762-4769. doi: Doi 10.1128/Aem.00236-09. Yang, S.J., Kataeva, I., Wiegel, J., Yin, Y.B., Dam, P., Xu, Y., Westpheling, J., and Adams, M.W.W. (2010). Classification of 'Anaerocellum thermophilum' strain DSM 6725 as Caldicellulosiruptor bescii sp. nov. Int J syst Evol Bacteriol 60, 2011-2015. doi: Doi 10.1099/Ijs.0.017731-0. Yang, X. (2010). "Scale-up of Microbial Fermentation Process," in Manual of industrial microbiology and biotechnology, eds. R.H. Baltz, J.E. Davies, A.L. Demain & American Society for Microbiology. (Washington, DC: ASM Press), 669-675. Yang, X., Georgescu, R., Berge, M., and Bhargava, S. (2008). Leveraging fermentation heat transfer data to better understand metabolic activity. BioPharm International 21(4). Yao, S., and Mikkelsen, M. (2010a). Metabolic engineering to improve ethanol production in Thermoanaerobacter mathranii. Appl Microbiol Biot 88, 199-208. doi: 10.1007/s00253- 010-2703-3.

Yao, S., and Mikkelsen, M.J. (2010b). Identification and overexpression of a bifunctional aldehyde/alcohol dehydrogenase responsible for ethanol production in Thermoanaerobacter mathranii. J Mol Microb Biotech 19, 123-133. doi: Doi 10.1159/000321498. Yu, J.S., Vargas, M., Mityas, C., and Noll, K.M. (2001). Liposome-mediated DNA uptake and transient expression in Thermotoga. Extremophiles 5, 53-60. Zamost, B.L., Nielsen, H.K., Starnes, R.L. (1991). Thermostable enzymes for industrial applications. J Ind Microbiol 8, 71-82. doi: 10.1007/BF01578757. Zhang, C., Cooper, T.E., Krause, D.J., and Whitaker, R.J. (2013). Augmenting the genetic toolbox for Sulfolobus islandicus with a stringent positive selectable marker for agmatine prototrophy. Appl Environ Microbiol 79, 5539-5549. doi: 10.1128/AEM.01608-13. Zhang, C., and Whitaker, R.J. (2012). A broadly applicable gene knockout system for the thermoacidophilic archaeon Sulfolobus islandicus based on simvastatin selection. Microbiology 158, 1513-1522. doi: 10.1099/mic.0.058289-0. Zheng, T., Huang, Q., Zhang, C., Ni, J., She, Q., and Shen, Y. (2012). Development of a simvastatin selection marker for a hyperthermophilic acidophile, Sulfolobus islandicus. Appl Environ Microbiol 78, 568-574. doi: 10.1128/AEM.06095-11. Zhu, W., Xia, J.-L., Peng, A.-A., Nie, Z.-Y., and Qiu, G.-Z. (2013). Characterization of apparent sulfur oxidation activity of thermophilic archaea in bioleaching of chalcopyrite. T Nonferr Metal Soc 23, 2383-2388. doi: 10.1016/S1003-6326(13)62745-4. Zillig, W., Kletzin, A., Schleper, C., Holz, I., Janekovic, D., Hain, J., Lanzendörfer, M., and Kristjansson, J.K. (1993). Screening for Sulfolobales, their plasmids and their viruses in icelandic solfataras. Syst Appl Microbiol 16, 609-628. doi: 10.1016/S0723- 2020(11)80333-4. Zillig, W., Stetter, K.O., Wunderl, S., Schulz, W., Priess, H., and Scholz, I. (1980). The Sulfolobus-“Caldariella” group: Taxonomy on the basis of the structure of DNA- dependent RNA polymerases. Arch Microbiol 125, 259-269. doi: 10.1007/BF00446886. Zuo, G., Hao, B., and Staley, J.T. (2014). Geographic divergence of "Sulfolobus islandicus" strains assessed by genomic analyses including electronic DNA hybridization confirms they are geovars. Antonie van Leeuwenhoek 105, 431-435. doi: 10.1007/s10482-013- 0081-4.

CHAPTER 2

Ancillary contributions of heterologous biotin protein ligase and carbonic anhydrase

for CO2 incorporation into 3-hydroxypropionate by metabolically

engineered Pyrococcus furiosus

Hong Lian1 ^, Benjamin M. Zeldes1 ^, Gina L. Lipscomb2, Aaron. B. Hawkins1, Yejun Han1,

Andrew J. Loder1, Declan Nishiyama1, Michael W.W. Adams2, and Robert M. Kelly2*

1Department of Chemical and Biomolecular Engineering, North Carolina State University,

Raleigh, NC 27695-7905

2Department of Biochemistry and Molecular Biology, University of Georgia, Athens, GA 30602

^contributed equally to this work

Published in Biotechnology and Bioengineering, 113:12 2652-60 (2016)

2.1. Abstract

Acetyl-Coenzyme A carboxylase (ACC), malonyl-CoA reductase (MCR), and malonic

- semialdehyde reductase (MRS) convert HCO3 and acetyl-CoA into 3-hydroxypropionate (3HP) in the 3-hydroxypropionate/4-hydroxybutyrate carbon fixation cycle resident in the extremely thermoacidophilic archaeon Metallosphaera sedula. These three enzymes, when introduced into the hyperthermophilic archaeon Pyrococcus furiosus, enable production of 3HP from maltose and CO2. Sub-optimal function of ACC was hypothesized to be limiting for production of 3HP, so accessory enzymes carbonic anhydrase (CA) and biotin protein ligase (BPL) from M. sedula were produced recombinantly in Escherichia coli to assess their function. P. furiosus lacks a native, functional CA, while the M. sedula CA (Msed_0390) has a specific activity comparable to other microbial versions of this enzyme. M. sedula BPL (Msed_2010) was shown to biotinylate the β-subunit (biotin carboxyl carrier protein) of the ACC in vitro. Since the native

BPLs in E. coli and P. furiosus may not adequately biotinylate the M. sedula ACC, the carboxylase was produced in P. furiosus by co-expression with the M. sedula BPL. The baseline production strain, containing only the ACC, MCR, and MSR, grown in a CO2-sparged bioreactor reached titers of approximately 40 mg/L 3HP. Strains in which either the CA or BPL accessory enzyme from M. sedula was added to the pathway resulted in improved titers, 120 or 370 mg/L, respectively. The addition of both M. sedula CA and BPL, however, yielded intermediate titers of 3HP (240 mg/L), indicating that the effects of CA and BPL on the engineered 3HP pathway were not additive, possible reasons for which are discussed. While further efforts to improve

3HP production by regulating gene dosage, improving carbon flux and optimizing bioreactor operation are needed, these results illustrate the ancillary benefits of accessory enzymes for

incorporating CO2 into 3HP production in metabolically engineered P. furiosus, and hint at the important role that CA and BPL likely play in the native 3HP/4HB pathway in M. sedula.

2.2. Introduction

The 3-hydroxypropionate/4-hydroxybutyrate (3HP/4HB) cycle is a carbon fixation pathway unique to the extremely thermoacidophilic archaeal order Sulfolobales, found in species such as Metallosphaera sedula (Berg et al. 2007; Berg et al. 2010). This cycle could be used as a basis for fuel and chemical production in a metabolically engineered host (Hawkins et al. 2013;

Keller et al. 2013). The first part of this cycle, referred to as sub-pathway 1 (SP1), catalyzes the conversion of acetyl-CoA and bicarbonate to 3-hydroxypropionate (see Figure 2.1) and involves three enzymes: the heterotrimeric acetyl-CoA/propionyl-CoA carboxylase (ACC) (Msed_0147,

0148, and 1375) (Hügler et al. 2003); malonyl-CoA reductase (MCR) (Msed_0709) (Alber et al.

2006); and malonic semialdehyde reductase (MSR) (Msed_1993) (Kockelkorn and Fuchs 2009).

The genes encoding the three enzymes of SP1 were inserted into the hyperthermophilic anaerobe

Pyrococcus furiosus, a newly established metabolic engineering platform, to produce 3- hydroxypropionate (3HP) at titers of approximately 60 mg/L from maltose and CO2 (Keller et al.

2013).

The genetic system developed for P. furiosus (Topt 100ºC) is one of the most versatile available for any hyperthermophile, and has allowed for a variety of metabolic engineering efforts in this organism (Zeldes et al. 2015). P. furiosus cells grow quickly, and tolerate sudden shifts to much lower growth temperatures, allowing them to express pathways derived from less extreme thermophiles (Basen et al. 2012), such as the 3HP production pathway from M. sedula

(Topt 75ºC) discussed above. More generally, high temperature fermentations using a

thermophilic host would reduce the risk of contamination, save on cooling costs (which can be considerable for mesophilic fermentations), and improve solubility of likely feedstocks, such as lignocellulose.

The successful demonstration of 3HP production in P. furiosus opens the door for further inquiry into the interactions between native host metabolism and the synthetic pathway to improve production titers. Understanding transcription and enzyme maturation, competing pathways, side reactions, redox and cofactor balance, and catalytic driving forces are all important for optimizing production and performance of any given synthetic pathway (Berríos-

Rivera et al. 2002; Carothers et al. 2009; Shen et al. 2011). To this end, an improvement in overall 3HP production was achieved by the deletion of genes encoding an acetyl-CoA synthetase, a key enzyme for fermentative metabolism in P. furiosus and a competitor for the substrate molecule acetyl-CoA (Thorgersen et al. 2014). Also, through the insertion of two genes encoding accessory enzymes from M. sedula, a biotin protein ligase (BPL) and carbonic anhydrase (CA), P. furiosus strain MW76 produced 3HP titers of 276 mg/L in an agitated bioreactor in which CO2 sparging was used (Hawkins et al. 2015). However, the specific contributions of BPL and CA were not clear. This brings up the question as to what role CO2 incorporation (driven by CA) and carboxylase function (dependent on BPL) played in titer improvement.

Carboxylases, such as ACC, represent the key carbon-incorporating step in autotrophic growth and are typically the primary factor that determines energetic cost and environmental niche of a given carbon fixation cycle (Erb 2011). Therefore, both BPL and CA are expected to improve 3HP titers by increasing the efficiency of ACC in the SP1 production pathway (see

Figure 2.1). ACC is a 560 kDa complex, comprised of biotin carboxylase (α-subunit: ACC-α),

biotin carboxyl carrier protein (BCCP) (β-subunit: ACC-β), and carboxyl transferase (γ-subunit:

ACC-γ), likely in a (αβγ)4 arrangement (Hügler et al. 2003). Biotin, or vitamin H, is an essential prosthetic group that is post-translationally attached at the active site of certain carboxyl- transferring enzymes (Streit and Entcheva 2003). BPL is responsible for attaching biotin to

ACC-β (BCCP) to create active holoenzyme, so it is expected to increase carboxylase activity of the ACC by increasing the fraction of enzyme that is in its active holo-form. CA, on the other hand, catalyzes the interconversion of bicarbonate and dissolved CO2 in aqueous environments

- + (CO2 + H2O  HCO3 + H ), thereby playing an essential role in CO2 transport out of metabolizing aerobic cells (Henry 1996). Its prevalence in autotrophic bacteria has been proposed to mimic CO2 concentrating mechanisms used by plants to facilitate carbon fixation

(Smith et al. 1999). Bicarbonate, rather than CO2, is the substrate of ACC (Hügler et al. 2003), so by accelerating the normally slow equilibration between these two compounds, CA is expected to increase intracellular bicarbonate levels and thereby improve the kinetic rate of CO2 fixation by ACC.

P. furiosus cell extracts lack native carbonic anhydrase activity, so expression of M. sedula CA could improve bicarbonate availability. The native P. furiosus BPL may not biotinylate M. sedula ACC-β efficiently, due to sequence differences around the biotinylation site when compared to the native M. sedula version of this enzyme. To determine their relative contributions in 3HP-producing strains of P. furiosus, the CA and BPL were first characterized in vitro. Then, their in vivo impact on 3HP titers in P. furiosus was analyzed by comparing 3HP production strains containing one or both accessory enzymes. In addition to examining the ancillary contributions of each accessory enzyme to improvements in 3HP titer, prospects were

considered for P. furiosus as a hyperthermophilic metabolic engineering platform for liquid fuels and chemicals production that includes CO2-incorporating steps.

2.3. Materials and Methods

P. furiosus strain construction

P. furiosus was routinely grown in serum bottles using either defined (3.5 g/L cellobiose,

1x amino-acid solution, 1x vitamin mix), or complex (5 g/L yeast extract, 5 g/L maltose) medium, both in a seawater-based medium containing 1x base salts, 1x trace minerals, 10 M sodium tungstate, 0.25 mg/mL resazurin, 0.5 g/L cysteine, 0.5 g/L sodium sulfide, 1g/L sodium bicarbonate, and 1 mM potassium phosphate, all buffered to pH 6.8. For descriptions of stock solutions see (Lipscomb et al. 2011).

All P. furiosus transformations were done using uracil prototrophic selection on defined medium, essentially as previously described (Lipscomb et al. 2011), using linearized plasmids or linear splice overlap extension (SOE) PCR products. Strains were further purified by two consecutive transfers on solid medium, and the final strains were verified by PCR and sequencing of the regions containing the chromosomal insertions. The P. furiosus genome regions (3 and 5) used for chromosomal integration of the constructs contain little to no transcriptional activity, as determined from analysis of tiling array data (Yoon et al. 2011).

Strains used and constructed in this study are listed in Table 2.1.

To construct strain RMK120, the Pgdh-pyrF-Ppep-BPL region, without the CA gene and

Pslp-ACCα locus, were amplified by PCR from MW76 genomic DNA and combined using SOE

PCR to generate the Pgdh-pyrF-Ppep-BPL-Pslp-ACCα construct. The SOE PCR construct was transformed into strain MW60 (Thorgersen et al. 2014), a markerless version of MW56 (Keller

et al. 2013), to construct strain RMK120. To construct strain RMK121, the Pgdh-pyrF-Ppep and

CA-Pslp-ACCα regions were amplified from MW76 genomic DNA, combined using SOE PCR to generate the Pgdh-pyrF-Ppep-CA-Pslp-ACCα construct, and transformed into strain MW60.

To clone and express 6His-tagged ACC in P. furiosus, the artificial operons Ppep-BPL-

CA and Pslp-ACCαβ-his6ACCγ were constructed and transformed into COM1 to generate strain

MW112. PCR products of Ppep-BPL-CA, Pgdh-pyrF, Pslp-ACCαβ, his6ACCγ, and 0.5-kb upstream and downstream flanking regions to genome region 5 (between convergent genes

PF1232 and PF1233) were combined via Gibson assembly (Gibson et al. 2009) (Gibson assembly kit, New England Biolabs, Ipswich, MA) to construct the plasmid pGL033. The Pslp-

ACCαβ-his6ACCγ artificial operon contained a P. furiosus RBS from the gene encoding pyruvate ferredoxin oxidoreductase subunit γ (PF0791, 5'- ggaggtttgaag) upstream of the ACC γ- subunit gene; a 6His tag flanked by two alanine codons (5'-gcacatcaccaccaccatcacgct) was also inserted after the start codon of the ACCγ gene to facilitate protein purification. Linearized pGL033 was transformed into COM1 to construct strain MW112.

Bioreactor growth of P. furiosus strains

Bioreactor media matched routine complex growth media, except for the addition of 0.25 mg/L biotin and omission of sodium bicarbonate. A pH probe was installed in the 3L Applikon glass bioreactors (ADI 1010/1025; Delft, The Netherlands) before autoclaving. Components for

1 L of media were added and the bioreactors were heated to 95ºC, and gas sparging with 20%

CO2, 80% N2 was initiated. Media pH was adjusted to 6.8 with NaOH, based on 10 mL samples taken from the hot bioreactors and quickly cooled to below 30ºC. Probes for pH control were calibrated at room temperature, with pH offset determined using the cooled media samples. Gas flow rates were controlled using Matheson E910 rotameters (FM-1050 series; Basking Ridge,

NJ, USA) and agitation came from Rushton impellers. Reactors were inoculated to 1106 with cells grown for 10 to 12 hours in serum bottles. Growth was monitored by cell counts until density reached 1108 cells/mL, then bioreactors were cooled by passing cold water through a heat exchange-port. During media heating and initial growth phase stir rate, temperature, and gas flow were 250 rpm, 95ºC, and 40 mL/min, then changed to 400 rpm, 72ºC, and 70 mL/min at the temperature switch. Samples consisted of 10 mL separated into a cell pellet, filter-sterilized supernatant, and 1 mL for cell counts. Samples were collected immediately following inoculation, immediately following temperature switch, 5 hours after temperature switch, and then twice daily up to 90 hours.

To assay for ACC activity in cell extracts (CEs), bioreactors of COM1, MW56, and

RMK120 were grown as for the 3HP production experiments, but 24 hours after temperature switch the entire volume of the reactors was harvested. Cells were lysed in low osmotic buffer, and centrifuged at 24,000g for 30 minutes to remove insoluble components. ACC activity in the

CE was determined by monitoring phosphate (described below).

Quantification of metabolites in bioreactors

Acetate, maltose, and 3HP in bioreactor supernatant samples were measured by HPLC using a Rezex-ROA column, 300 mm × 7.8 mm ID (Phenomenex, Torrance, CA) heated to

60°C. The system consisted of a Waters (Milford, MA) 1525 Binary HPLC pump, with detection by Waters 2414 Refractive Index Detector (maltose) and Waters 2487 Dual  Absorbance

Detector (acetate and 3HP). Sulfuric acid was added to cell supernatants to 0.05% v/v, and 40 L injections were run in 5 mM sulfuric acid mobile phase at 0.6 mL/min for 50 minutes. Standards were made in distilled water and sulfuric acid added to 0.05%.

Heterologous expression of M. sedula genes in E. coli

The gene encoding the putative biotin protein ligase in the M. sedula genome

(Msed_2010) was amplified from genomic DNA and cloned into pET-46 Ek/LIC vector with an

N-terminal 6His-tag. The gene encoding the putative carbonic anhydrase in the M. sedula genome (Msed_0390) was amplified from genomic DNA and cloned using primers designed to introduce an XhoI restriction site after the stop codon. Then the PCR product was purified and cloned into pET21b (+) vector with a C-terminal 6His-tag. The M. sedula genes encoding the α-

, β-, and γ-subunits of the carboxylase (Msed_0147, Msed_1048, and Msed_1375) were cloned by a ligation-independent method using pET46 Ek/LIC, pRSF-2 Ek/LIC, and pCDF-2 Ek/LIC

(EMD4 Biosciences), respectively, which are compatible vectors enabling co-expression of these genes in various combinations: αβγ (N-terminal 6His on γ), αβ (N-terminal 6His on β), βγ (N- terminal 6His on γ), or β only (N-terminal 6His) in a single E. coli strain. All vectors containing target genes were transformed into competent E. coli Rosetta 2 (DE3) cells (EMD

Millipore, Darmstadt, Germany). Cells were grown in 1L LB media, containing 100 μg mL-1 ampicillin (for pET46 and pET21b), 30 μg mL-1 kanamycin (for pRSF-2), 50 μg mL-1 streptomycin (for pCDF-2), and 34 μg mL-1 chloramphenicol (to maintain the Rosetta plasmid), according to the different antibiotic resistance markers carried by the plasmids. Cells were harvested by centrifugation 4-5 hours after induction with 0.1-0.2 mM IPTG, and stored at -

20°C.

Purification of recombinant His-tagged M. sedula proteins

E. coli cells containing recombinant M. sedula proteins were re-suspended in 20 mM sodium phosphate, 0.5 M NaCl, pH 7.4, and disrupted by sonication for 10 min at 60%

Amplitude (10 sec on, 10 sec off). Cell extract was incubated at 65°C for 20 min to denature E.

coli native proteins, and centrifuged at 16,000 × g for 15 min. The heat-treated cell extract was loaded onto a 5 mL HiTrap IMAC HP column (GE Healthcare Life Sciences), equilibrated with

20 mM sodium phosphate, 0.5 M NaCl, at pH 7.4. The column was washed with five bed volumes of equilibration buffer and eluted with 300 mM imidazole at a flow rate of 1 mL min-1.

Active fractions eluted with imidazole were pooled and dialyzed in 50 mM Tris-HCl, pH 8.2 to remove imidazole. The pooled, dialyzed fractions were concentrated via Amicon® Ultra-15

Centrifugal Filter Units – 3 kDa (EMD Millipore) and stored at -80°C.

Expression and purification of recombinant M. sedula ACC from P. furiosus

Strain MW112 was routinely grown in bioreactors in growth medium containing 5 g/L maltose, 5 g/L yeast extract, 5 g/L tryptone, and 0.25 mg/L biotin. Two liters of culture were grown at 95°C with 15 mL/min N2 (80%)/CO2 (20%) sparging and 250 rpm agitation to a density of 108 cells/mL. Then, temperature was reduced to 72°C and cells were grown for 18 hours before being harvested by centrifugation.

Cells were lysed by re-suspending in low salt buffer (50 mM Tris-HCl, pH 8.2) and stirring at room temperature for 2 hours. The cell lysate was centrifuged at 10,000 × g for 20 min, and the supernatant was collected and filter-sterilized, before being loaded on a 1 mL

HiTrap IMAC HP column (GE Healthcare Life Sciences) equilibrated with 50 mM Tris, 100 mM NaCl, pH 8.0. The column was washed with five bed volumes of equilibration buffer and eluted with 500 mM imidazole at a flow rate of 1 mL min-1. Fractions were pooled, dialyzed, and concentrated as described above.

The native molecular mass of the enzyme was estimated by gel filtration chromatography. Protein from the IMAC step was applied to a Superdex 75 HR 16/60 gel filtration column (GE Healthcare Life Sciences; volume, 20 mL), which had been equilibrated

with 50 mM sodium phosphate, 100 mM NaCl, pH 7.2. The flow rate was 0.5 mL min-1. The column was calibrated with the following molecular mass standards (Sigma): β-Amylase from sweet potato (200 kDa), alcohol dehydrogenase from yeast (150 kDa), bovine serum albumin (66 kDa), carbonic anhydrase from bovine erythrocytes (29 kDa), cytochrome c from horse heart

(12.4 kDa), and aprotinin from bovine lung (6.5 kDa). Proteins were visualized using sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) 4-12% gels (Invitrogen) and

GelCode Blue stain (ThermoFisher).

Biochemical assays

In vitro biotinylation. Biotin protein ligase (BPL) activity was measured by following biotinylation of purified recombinant E1-β or E1-βγ. Assay conditions were adapted according to literature reports (Chapman-Smith et al. 1999). The assay solution contained 30 mM Tris, pH

8.2, 100 mM KCl, 5 mM MgCl2, 0.5 mM DTT, 0.3 mM ATP, 10 μM biotin and 1 μM purified recombinant putative BPL (Msed_2010). The reaction was initiated by the addition of purified recombinant E1-β or E1-βγ to a final concentration of 2.5 μM and incubated for 30 min at 70°C.

As a control, E1-β or E1-βγ were incubated with BPL storage buffer instead. After incubation, samples were filtered through a 10-kDa molecular weight cutoff filter (YM-10 Thermo

Scientific) to remove excess biotin and re-dissolved in 50 mM Tris-HCl, pH 8.2. A biotin capture method, based on streptavidin sepharose beads (GE Healthcare), was developed to visualize biotinylated proteins on SDS-PAGE. The products from in vitro biotinylation (including control) were loaded onto streptavidin sepharose beads and unbiotinylated protein was washed away.

Bound protein was released by heating the beads at 95°C for 5 min to denature the streptavidin.

Both bound and unbound fractions were then run on SDS-PAGE for visualization.

Acetyl-CoA carboxylase. ACC activity was assayed using a coupled assay with MCR. The assay mixture, containing 80 mM MOPS/NaOH pH 8.0, 0.5 mM acetyl-CoA, 5 mM MgCl2, 5 mM 1,4-dithioerythritol, 5 mM ATP, 10 mM NaHCO3, 0.5 mM NADPH, and purified recombinant MCR, was heated to 70°C. Then, purified recombinant ACC was added, and

NADPH consumption (by MCR acting on the product of ACC, malonyl-CoA) was monitored continuously by absorbance at 340 nm.

A discontinuous assay was also used that relied on a phosphate colorimetric assay kit

(BioVision, Milpitas, CA) to detect free phosphate release by ATP hydrolysis catalyzed by ACC.

The reaction solution contained 75 mM MOPS/NaOH pH 8.0, 0.5 mM acetyl-CoA, 5 mM

MgCl2, 5 mM 1,4-dithioerythritol, 2 mM ATP, and 10 mM NaHCO3, and was incubated at 70°C prior to addition of ACC. Samples were taken regularly for up to 6 minutes and assayed for phosphate.

Carbonic anhydrase. The CO2 hydration assay followed a modified version of the Wilbur-

Anderson method (Wilbur and Anderson 1948). CO2-saturated water was made by subliming dry ice in a glass bottle for 30 min, with the bottle sealed with a rubber stopper to prevent leakage.

The CO2-saturated water (3 mL) was immediately added to 2 mL of Tris–Sulfate buffer (100 mM; pH 8.3) containing 0.5 mL of enzyme solution. The enzyme solution was diluted and the reaction carried out at low temperature (10°C) to promote CO2 solubility and keep reaction rates within measurable limits. The time (t) required for the pH to drop from 8.0 to 6.3, as catalyzed by the enzyme, was measured. The control for the pH change (8.0–6.3) used enzyme storage buffer (50 mM Tris-HCl, pH 8.2) substituted for the enzyme solution. The Wilbur–Anderson

Units were determined from the equation: 1U = (tc-t)/t, where tc is the time for control. Protein

content was determined by Bradford assay and the activity was reported in U/mg protein. All measurements were carried out in triplicate.

Esterase activity was determined by p-nitrophenylacetate hydrolysis at 25°C using a modification of the method of Armstrong et. al. (Armstrong et al. 1966). A 3 mM stock of p- nitrophenylacetate was made by dissolving the compound in acetone and then diluting it 25-fold with water. One mL of this substrate solution was added to 1.9 mL of 15 mM Tris-Sulfate, pH

7.6, and uncatalyzed rate of absorbance change (348 nm, ɛ = 5,000 M−1 cm−1) compared to the rate following addition of 0.1 mL of enzyme. Enzyme storage buffer was added for negative controls, and carbonic anhydrase from bovine erythrocytes (sigma) for positive controls.

2.4. Results

Identification and characterization of the M. sedula carbonic anhydrase (CA)

The M. sedula genome encodes two putative CAs (Msed_0390 and Msed_1618), neither of which has previously been characterized biochemically. Msed_0390 was significantly up- regulated under autotrophic growth compared to heterotrophic growth (Auernik and Kelly 2010), responding in concert with transcription of genes encoding 3HP/4HB carbon fixation cycle enzymes. Msed_1618 did not respond under any growth conditions tested and was transcribed at very low levels; as such, it was not considered further for the purposes of this study. Msed_0390 was annotated as a β-class carbonic anhydrase, most likely required for efficient uptake of inorganic carbon by M. sedula for carbon fixation. This CA was suspected to play a vital role as a complementary enzyme for the M. sedula 3HP/4HB cycle to increase activity of CO2 fixation by rapidly providing bicarbonate to acetyl-CoA carboxylase.

The heat-treated cell extract of recombinant E. coli expressing Msed_0390 exhibited CO2 hydration activity, relative to the control. Furthermore, the IMAC purified recombinant enzyme had a CO2 hydration activity of 292 U/mg at 10⁰C, validating that Msed_0390 encodes a

- functional CA for CO2 and HCO3 interconversion in M. sedula. The purified CA did not show esterase activity, a feature of the α and δ-classes of CAs (Lee et al. 2013), consistent with its annotation as a β-CA. This enzyme is the first CA that has been characterized from extremely thermoacidophilic archaea. The predicted molecular mass of the CA is 22 kDa (including a C- terminal 6×His tag), which is consistent with the Mr of 20.6 kDa determined by size exclusion chromatography, suggesting that this enzyme functions as a monomer (Table 2.2). Thus, the molecular assembly of this M. sedula CA differs from previously characterized β-CAs that typically exist as homodimers, homotetramers and homooctamers, with the fundamental structural unit as a dimer (Rowlett 2010). M. sedula grows at low pH (pH 2) where CO2 has low solubility, but the cytoplasm, where the 3HP/4HB cycle utilizes bicarbonate as a substrate, is neutral (Peeples and Kelly 1995). There is no evidence that the M. sedula -CA is an extracellular or membrane-bound protein (no discernible signal peptide). It most likely functions in the cytoplasm to accelerate the hydration of CO2 to carbonic acid, which is then rapidly deprotonated to bicarbonate at neutral pH (pKa1 = 6.4) (Loerting and Bernard 2010).

P. furiosus does not appear to natively express a functional CA to aid in the conversion of

CO2 to bicarbonate, the substrate of the 3HP/4HB cycle. So far, the only evidence for CA in P. furiosus was the detection of cross-reacting proteins in P. furiosus cell extract to antisera raised from previously characterized mesophilic CAs in Western blots (Smith et al. 1999). Here, the P. furiosus cell extract exhibited no detectable CA activity (< 0.01 Wilbur-Anderson unit at 55°C).

It should be noted that the upper temperature limit of the assay (55°C) is well below the expected

activity optimum for a putative P. furiosus CA, and the decreased solubility of CO2 at higher temperatures under atmospheric pressure impedes determination of CA activity. A crystal structure of a γ-class CA in P. horikoshii has been reported, but no activity was noted for this enzyme (Jeyakanthan et al. 2008). Therefore, the presence of a putative wild-type P. furiosus CA for efficient CO2 uptake and bicarbonate conversion is unlikely.

Assuming that P. furiosus lacks CA activity, the production of the bicarbonate substrate for the M. sedula ACC in the cytoplasm of the engineered strain is limited to the rate of uncatalyzed conversion from CO2, creating a potential bottleneck in the recombinant 3HP production pathway. To overcome this bottleneck, the gene encoding CA from M. sedula was inserted into P. furiosus, generating strain RMK121. This could then be compared to P. furiosus strain MW56, which lacks the M. sedula CA, to determine the role that bicarbonate availability plays in ACC function.

Production and characterization of recombinant carboxylase ACC

Previously, a partially purified version of the native M. sedula ACC was characterized to confirm its function (Hügler et al. 2003). Since a recombinant version of this enzyme was recruited to P. furiosus strains producing 3HP, efforts were made to express the recombinant

ACC in active form to evaluate its biochemical properties. In order to evaluate the in vitro activity of the ACC, the individual subunits were expressed in separate E. coli strains as well as co-expressed on separate plasmids, but in the same strain. In both cases, the recombinant ACC α and γ subunits were recovered only in the insoluble fraction, and furthermore lacked catalytic activity. Hulger et al. expressed only the ACC-γ subunit in E. coli, and saw no change in activity when it was added to the purified native protein (Hügler et al. 2003). The fact that they do not

report a fully recombinant ACC may indicate that the authors encountered similar difficulties in expressing the functional holoenzyme in E. coli.

A biotin quantitation assay indicated that the ACC-β subunit was not being properly biotinylated in E. coli. The biotin cofactor is attached post-translationally to a specific lysine residue on ACC-β by BPL (Clarke et al. 2003). Analysis of the amino acid sequence in the vicinity of the lysine residue targeted by BPL provided some insights into the problem. While similar biotinylated proteins in both E. coli and P. furiosus have a methionine immediately C- terminal to the biotinylated lysine residue, S. tokodaii, M. sedula and the other Sulfolobales have a serine (Li et al. 2006) (Figure 2.2, alignment performed in Geneious 8.4.1 (Kearse et al.

2012)). In S. tokodaii, mutation of the serine to methionine allowed biotinylation by E. coli’s

BPL, and S. tokodaii BPL could target the E. coli enzyme if its methionine was mutated to serine, although still less efficiently than their native substrates (Li et al. 2006; Sueda et al.

2006). Therefore, it seemed likely that the amino acid sequence of M. sedula’s ACC-β would make it difficult for the BPL of either E. coli or P. furiosus to biotinylate it efficiently.

Rather than mutate the M. sedula ACC-β subunit so it could be modified by the native

BPL in P. furiosus, the gene encoding the M. sedula BPL was expressed in P. furiosus. The sequence for BPL from S. tokodaii (Li et al., 2006) was used to query the NCBI database, and

Msed_2010, with 56% identity at the amino acid level, appeared to be a homolog of the S. tokodaii BPL. To confirm, Msed_2010 was expressed in E. coli and incubated with ACC-β expressed in E. coli in the presence of biotin, where it was able to biotinylate ACC-β (Figure

2.3). The predicted molecular mass of the M. sedula BPL is 26 kDa, yet the molecular mass determined by size exclusion chromatography was 57 kDa. Therefore, the M. sedula BPL probably functions as a homodimer (Table 2.2).

Since Msed_2010 was able to biotinylate ACC in vitro, it was cloned into P. furiosus along with his-tagged ACC (N-terminus of γ-subunit) to generate strain MW112. Recombinant

ACC purified from MW112 cell extract (Table 2.2) was functional, although less active than the native enzyme with a specific activity of 0.65 µmol/min/mg compared to 3.2 µmol/min/mg for the native-purified enzyme (Hügler et al. 2003). SDS-PAGE indicated low expression of ACC-β, which may account for some of the loss in activity. The functional production of M. sedula ACC in P. furiosus agrees with previous results; heterologous expression of protein in a more closely related host (in this case another archaeal extreme thermophile) can yield active enzyme when production in E. coli fails (Mueller et al. 2009).

Since P. furiosus strain MW56, which lacks the M. sedula BPL (Msed_2010), was able to produce 3HP, the ACC must be biotinylated to some extent by the native P. furiosus BPL. P. furiosus strain RMK120 was constructed with M. sedula BPL in addition to SP1 for 3HP production, to determine whether production would improve with more effective biotinylation of the ACC. As shown in Table 2.4, ACC activity in cell-extract of RMK120 is approximately double that of MW56, despite being under control of the same constitutive promoter. This confirms that the more efficient biotinylation accomplished by Msed BPL results in increased

ACC activity. Note that the control strain COM1 showed about half as much activity as MW56, which, since it lacks ACC altogether, must be the result of background ATP consumption in the

CE. Therefore, these specific activities are likely overestimates.

Effect of recombinant M. sedula CA and BPL on 3HP production in P. furiosus

To determine the separate and potential complementary roles of CA and BPL in production of 3HP, four metabolically engineered P. furiosus strains and the unmodified parent strain COM1 were grown in 1L bioreactors. All engineered strains, including MW56, contained

the three M. sedula enzymes making up SP1 (ACC, MCR and MRS), while RMK120 also contained BPL, RMK121 also contained CA, and MW76 contained both BPL and CA. Cells were grown at 95°C until they reached a density of 1x108 cells/mL, at which time the temperature was reduced to 72°C to match the optimum temperature of the recombinant M. sedula enzymes. Following the temperature shift, cells went from doubling every hour at 95°C to every 10-20 hours at 72°C (only about one more doubling occurs after temperature shift) (Figure

2.4a). Despite relatively constant cell number, cell size increased dramatically following the switch to 72°C, as has been noted previously (Basen et al. 2012). Cell pellets taken during bioreactor operation show a concomitant increase in size. Production of 3HP becomes evident for the engineered strains within 10 to 20 hours after temperature shift, and levels off once the maltose is consumed (Figure 2.4b). This lag in 3HP production agrees with previous experiments (Hawkins et al. 2015; Keller et al. 2013), and likely is related to the time it takes for the recombinant polypeptides to be synthesized and folded. Maximum 3HP titers were 44 mg/L in MW56 (SP1) and 236 mg/L in MW76 (SP1 + BPL +CA), consistent with previous results, i.e., 60 mg/L found by Keller et al. and 276 mg/L found by Hawkins el al., respectively. As expected, the control strain COM1 produced no detectable 3HP.

Figure 2.4c shows acetate production over the course of cell growth in bioreactors. In the absence of elemental sulfur, P. furiosus obtains all cellular energy from fermenting sugars to acetate, H2, and CO2 (Adams et al. 2001), so acetate levels provide a direct representation of the level of metabolic activity occurring in each strain. The only significant difference in acetate production is in the parent strain (COM1) control, which produced less than half as much acetate as all the engineered strains. COM1 was also the only strain that had residual unconsumed maltose left over at the end of the run (Figure 2.4d). So, it appears that, while the parent strain

goes dormant and reduces metabolic activity at 72°C, the modified strains are forced to maintain active metabolism due to the use of the strong constitutive S-layer protein promoter (Pslp), and the resulting expression of the energy-intensive SP1 enzymes that remain active at the lower temperature. This result agrees with previous transcriptional data indicating genes related to energy metabolism and protein synthesis were up-regulated in MW76 relative to COM1 at 72°C

(Hawkins et al. 2015).

Figure 2.4 shows representative experimental results for each strain, while Table 2.3 gives average maximum 3HP titers of triplicate bioreactor experiments with each strain. The parent strain (COM1) produced no 3HP, while of the engineered strains, RMK120 (SP1 + BPL) produced the most, followed by MW76 (SP1 + CA + BPL), then RMK121 (SP1 + CA), and finally MW56 (SP1), which served as a baseline for the productivity of the SP1 pathway without accessory enzymes. The finding that RMK120 (SP1+ BPL) was the most productive was unexpected, since it had only one of the two accessory enzymes hypothesized to improve productivity.

2.5. Discussion

Two accessory enzymes, important for optimal carboxylase function in the native

3HP/4HB carbon fixation cycle of M. sedula, were each examined in their recombinant forms.

The role of these accessory enzymes in their native context is important to understand when inserting heterologous pathways into new hosts, as essential biochemical modifications may not occur in the host organism. Here, we report the successful production of recombinant M. sedula

ACC using P. furiosus as the host, since no active ACC could be produced in E. coli. The key factor was the co-expression of the M. sedula BPL, which was highly specific for biotinylation

of the M. sedula BCCP (ACC-β); neither the native E. coli nor the native P. furiosus BPL appeared to function as efficiently. The fact that ACC activity doubled in cell extract of the strain containing M. sedula BPL (RMK120) relative to an otherwise identical strain that lacked it

(Table 2.4) confirms that native P. furiosus BPL does a poor job of biotinylating ACC, and helps to explain why the P. furiosus strains lacking the M. sedula BPL (MW56 and RMK121) produced less 3HP than the BPL-expressing strains (MW76 and RMK120) (Table 2.3).

About a 3-fold improvement in 3HP titers could be achieved with the addition of the M. sedula CA to the P. furiosus strain containing the five genes encoding SP1 (MW56). The recombinant M. sedula CA produced in E. coli had an activity of 292 U/mg measured at 10°C.

4 -1 This value is comparable to the 720 U/mg (estimated kcat of 6.1 x 10 s ) at 25°C reported for recombinant CA from another thermophilic archaeon, Methanosarcina thermophila (Alber et al.

1999), especially if consideration is given to assay temperature and enzyme temperature optimum. Carbonic anhydrases are reported to be among the most catalytic enzymes known, often limited by the diffusion rate of their substrate; even less active versions have kcat values on the order of 104 s-1 (Smith and Ferry 2000). Based on the similarity to the Me. thermophila CA,

4 -1. the M. sedula CA would also have a kcat on order of 10 s . Since the native M. sedula ACC has

-1 a reported kcat of 28 s , it is unlikely that the CA-catalyzed production of bicarbonate would be rate-limiting for carbon uptake in M. sedula. However, in P. furiosus, which lacks a functional

- CA, the comparatively slow uncatalyzed conversion of CO2 to HCO3 does appear to impair 3HP production in MW56.

It was interesting that the strain with both CA and BPL (MW76), while yielding 5-fold higher titers than the strain lacking both enzymes (MW56), produced significantly less 3HP as the strain with only BPL (RMK120) that produced 8-fold higher levels of 3HP than MW56.

Clearly, under the conditions tested here, the effects of BPL and CA expression were not additive, and may in fact be antagonistic. There are several possible explanations for this unexpected effect. First, because strong constitutive promoters were used for the expression of

CA and BPL in the recombinant P. furiosus strains, it is possible that the added burden of CA expression reduced BPL production in MW76 relative to RMK120. Improvements in productivity have been considered for other metabolically engineered pathways in extreme thermophiles, where reaction kinetics models suggested adjusting gene dosage can lead to improved product titers, as well as selectivity of desired to undesired products (Loder et al.

2015). A second possible explanation is that protein-protein interactions between the CA and

ACC interfere with biotinylation of ACC by the BPL. In autotrophic bacteria, association of the

CA with the carboxysome facilitates substrate channeling to ribulose-1,5-bisphosphate carboxylase/oxygenase, allowing increased carbon fixation at low CO2 concentrations (Bobik

2006). A similar mechanism may operate in M. sedula, considering the poor solubility of bicarbonate at its low pH growth optimum. If CA associates with ACC, high-level expression of

CA in strain MW76 could cause tighter association, which may interfere with biotinylation of

ACC by the BPL, leading to lower ACC activity than when BPL is expressed without CA, as in strain RMK120. Further work directed at improving titers and yields for 3HP in P. furiosus will examine this possibility and, perhaps, mitigate the deleterious effect by adjusting gene dosage through judicious choice of promoters.

The results demonstrate the importance of ancillary proteins on the function of heterologously expressed pathway components in metabolic engineering efforts. Failure to understand and account for these kinds of deficiencies can significantly compromise product titers and volumetric productivities. This effort also supports the promise of P. furiosus as a

metabolic engineering platform for formation of bio-based products that incorporate CO2. The establishment of extreme thermophiles as widely used industrial microorganisms will depend on further advances that allow for their unique physiological properties to be strategically utilized.

2.6. Acknowledgments

This work was supported with grants to RMK and MWWA by the US Department of

Energy Research ARPA-E Electrofuels Program (DE-AR0000081) and the US National Science

Foundation (CBET-1264052, CBET-1264053). ABH acknowledges support from a US

Department of Education GAANN Fellowship. AJL and BMZ acknowledge support from NIH

Biotechnology Traineeships (2T32GM008776).

Table 2.1: P. furiosus strains used in this study Strain Parent Genotype Source

COM1 DSM3638 ΔpyrF (Lipscomb et al. 2011)

ΔpyrF::PgdhpyrF-Ppep-BPL-CA- MW112 COM1 This work Pslp-ACCαβ-his6ACCγ

ΔpyrF::PgdhpyrF-Pslp-ACCαβγ- MW56 COM1 (Keller et al. 2013) MCR-MSR

ΔpyrF::PgdhpyrF-Ppep-BPL-CA- MW76 COM1 (Hawkins et al. 2015) Pslp-ACCαβγ-MCR-MSR

ΔpyrF::PgdhpyrF-Ppep-BPL-Pslp- This work RMK120 MW60 ACCαβγ-MCR-MSR (Thorgersen et al. 2014)

ΔpyrF::PgdhpyrF-Ppep-CA-Pslp- This work RMK121 MW60 ACCαβγ-MCR-MSR (Thorgersen et al. 2014)

Table 2.2: Recombinant M. sedula enzymes used in this study

Native Subunit Molecular Enzyme Molecular Assembly Comments Mass Mass (kDa) (kDa)

 - 57 Acetyl-CoA Agrees with native form 507  - 18.6 () carboxylase 4 (Hügler et al. 2003)  - 57

Most -CAs are Carbonic 22 22  heteromultimers, but anhydrase 1 Msed_0390 is monomeric

Biotin protein S. tokodaii homolog is 57 26  ligase 2 monomeric (Li et al. 2006)

Table 2.3: Maximum 3PH titers reached during bioreactor runs with recombinant strains

Average N=3 (mg/L) ± Strain Genes SD Titer relative to MW56

MW56 SP1 44 ± 11 1.0

MW76 SP1 + BPL + CA 236 ± 18 5.3

RMK120 SP1 + BPL 366 ± 74 8.2

RMK121 SP1 + CA 119 ± 21 2.7

Table 2.4 Activity of ACC measured in cell extract of bioreactor grown cells

Strain Genes Average N=3 (nmol/min/mg) ± SD

COM1 - 89 ± 20

MW56 SP1 161 ± 55

RMK120 SP1 + BPL 344 ± 29

Figure 2.1: 3-hydroxypropionate formation and accessory enzymes The first three enzymes of the 3HP/4HB cycle (sub-pathway 1 (SP1)) catalyze the formation of 3HP from acetyl-CoA and bicarbonate. Two accessory enzymes are potentially involved in carboxylase function – CA catalyzes the hydration of dissolved CO2 gas to supply bicarbonate and BPL is required for covalently ligating biotin to the B-subunit of ACC. Abbreviations CA – carbonic anhydrase; BPL – biotin protein ligase; ACC – acetyl-CoA carboxylase; MCR – malonyl-CoA reductase; MSR – malonic semialdehyde reductase.

Figure 2.2: Amino acid sequence alignment of BCCP around the canonical lysine residue (K) The conserved -E-X-M-K-M- motif is indicated by triangles. Mutated site in the Sulfolobales is marked by arrow. Msed, Metallosphaera sedula Msed0148, Stok, Sulfolobus tokodaii ST0592, Pfu, Pyrococcus furiosus PF0673, Eco, Escherichia coli ECs4127.

Figure 2.3: In vitro biotinylation of recombinant Msed ACC-β with recombinant Msed BPL Purified M. sedula ACC-β and BPL were combined with biotin cofactor and incubated for 30 min at 70oC. This “reaction mixture” was applied to a streptavidin column and washed. Bound fraction consisting of biotinylated ACC-β was released from the beads by heat denaturation.

Figure 2.4: Representative bioreactor runs of metabolically engineered P. furiosus strains a) cell counts, showing difference in growth rate before and after temperature shift from 95°C to 72°C. Time course of concentrations of b) 3HP, c) acetate, and d) maltose in supernatant. Vertical black line indicates temperature switch from 95oC to 72oC.

2.7. References

Adams MW, Holden JF, Menon AL, Schut GJ, Grunden AM, Hou C, Hutchins AM, Jenney FE, Jr., Kim C, Ma K and others. 2001. Key role for sulfur in peptide metabolism and in regulation of three hydrogenases in the hyperthermophilic archaeon Pyrococcus furiosus. J Bacteriol 183(2):716-724. Alber B, Olinger M, Rieder A, Kockelkorn D, Jobst B, Hügler M, Fuchs G. 2006. Malonyl- coenzyme A reductase in the modified 3-hydroxypropionate cycle for autotrophic carbon fixation in archaeal Metallosphaera and Sulfolobus spp. J Bacteriol 188(24):8551-8559. Alber BE, Colangelo CM, Dong J, Stålhandske CM, Baird TT, Tu C, Fierke CA, Silverman DN, Scott RA, Ferry JG. 1999. Kinetic and spectroscopic characterization of the gamma- carbonic anhydrase from the methanoarchaeon Methanosarcina thermophila. Biochemistry 38(40):13119-13128. Armstrong JM, Myers DV, Verpoorte JA, Edsall JT. 1966. Purification and properties of human erythrocyte carbonic anhydrases. J Biol Chem 241(21):5137-49. Auernik KS, Kelly RM. 2010. Physiological versatility of the extremely thermoacidophilic archaeon Metallosphaera sedula supported by transcriptomic analysis of heterotrophic, autotrophic, and mixotrophic growth. Appl Environ Microbiol 76(3):931-935. Basen M, Sun J, Adams MW. 2012. Engineering a hyperthermophilic archaeon for temperature- dependent product formation. mBio 3(2):e00053-12. Berg IA, Kockelkorn D, Buckel W, Fuchs G. 2007. A 3-Hydroxypropionate/4-Hydroxybutyrate autotrophic carbon dioxide Assimilation pathway in archaea. Science 318(5857):1782- 1786. Berg IA, Kockelkorn D, Ramos-Vera WH, Say RF, Zarzycki J, Hugler M, Alber BE, Fuchs G. 2010. Autotrophic carbon fixation in archaea. Nat Rev Microbiol 8(6):447-60. Berríos-Rivera SJ, Bennett GN, San K-Y. 2002. Metabolic engineering of Escherichia coli: increase of NADH availability by overexpressing an NAD(+)-dependent formate dehydrogenase. Metab Eng 4(3):217-229. Bobik TA. 2006. Polyhedral organelles compartmenting bacterial metabolic processes. Appl Microbiol Biotechnol 70(5):517-525. Carothers JM, Goler JA, Keasling JD. 2009. Chemical synthesis using synthetic biology. Curr Opin Biotechnol 20(4):498-503. Chapman-Smith A, Morris TW, Wallace JC, Cronan JE. 1999. Molecular recognition in a post- translational modification of exceptional specificity. Mutants of the biotinylated domain of acetyl-CoA carboxylase defective in recognition by biotin protein ligase. J Biol Chem 274(3):1449-1457. Clarke DJ, Coulson J, Baillie R, Campopiano DJ. 2003. Biotinylation in the hyperthermophile Aquifex aeolicus. Eur J Biochem 270(6):1277-1287. Erb TJ. 2011. Carboxylases in natural and synthetic microbial pathways. Appl Environ Microbiol 77(24):8466-8477. Gibson DG, Young L, Chuang RY, Venter JC, Hutchison CA, 3rd, Smith HO. 2009. Enzymatic assembly of DNA molecules up to several hundred kilobases. Nat Methods 6(5):343-5. Hawkins AB, Lian H, Zeldes BM, Loder AJ, Lipscomb GL, Schut GJ, Keller MW, Adams MWW, Kelly RM. 2015. Bioprocessing analysis of Pyrococcus furiosus strains engineered for CO2-based 3-hydroxypropionate production. Biotechnol Bioeng 112(8):1533-1543.

Hawkins AS, McTernan PM, Lian H, Kelly RM, Adams MWW. 2013. Biological conversion of carbon dioxide and hydrogen into liquid fuels and industrial chemicals. Curr Opin Biotechnol 24(3):376-384. Henry RP. 1996. Multiple roles of carbonic anhydrase in cellular transport and metabolism. Annu Rev Physiol 58:523-38. Hügler M, Krieger RS, Jahn M, Fuchs G. 2003. Characterization of acetyl-CoA/propionyl-CoA carboxylase in Metallosphaera sedula. Carboxylating enzyme in the 3- hydroxypropionate cycle for autotrophic carbon fixation. Eur J Biochem 270(4):736-744. Jeyakanthan J, Rangarajan S, Mridula P, Kanaujia SP, Shiro Y, Kuramitsu S, Yokoyama S, Sekar K. 2008. Observation of a calcium-binding site in the gamma-class carbonic anhydrase from Pyrococcus horikoshii. Acta Cryst D 64(Pt 10):1012-1019. Kearse M, Moir R, Wilson A, Stones-Havas S, Cheung M, Sturrock S, Buxton S, Cooper A, Markowitz S, Duran C and others. 2012. Geneious Basic: an integrated and extendable desktop software platform for the organization and analysis of sequence data. Bioinformatics 28(12):1647-9. Keller MW, Schut GJ, Lipscomb GL, Menon AL, Iwuchukwu IJ, Leuko TT, Thorgersen MP, Nixon WJ, Hawkins AS, Kelly RM and others. 2013. Exploiting microbial hyperthermophilicity to produce an industrial chemical, using hydrogen and carbon dioxide. Proc Natl Acad Sci U S A 110(15):5840-5845. Kockelkorn D, Fuchs G. 2009. Malonic semialdehyde reductase, succinic semialdehyde reductase, and succinyl-coenzyme a reductase from Metallosphaera sedula: Enzymes of the autotrophic 3-hydroxypropionate/4-hydroxybutyrate cycle in Sulfolobales. J Bacteriol 191(20):6352-6362. Lee RBY, Smith JAC, Rickaby REM. 2013. cloning, expression and characterization of the δ- carbonic anhydrase of Thalassiosira weissflogii (Bacillariophyceae). J Phycol 49(1):170- 177. Li Y-Q, Sueda S, Kondo H, Kawarabayasi Y. 2006. A unique biotin carboxyl carrier protein in archaeon Sulfolobus tokodaii. FEBS Lett 580(6):1536-1540. Lipscomb GL, Stirrett K, Schut GJ, Yang F, Jenney FE, Scott RA, Adams MWW, Westpheling J. 2011. Natural competence in the hyperthermophilic archaeon Pyrococcus furiosus facilitates genetic manipulation: construction of markerless deletions of genes encoding the two cytoplasmic hydrogenases. Appl Environ Microbiol 77(7):2232-2238. Loder AJ, Zeldes BM, Garrison GD, Lipscomb GL, Adams MWW, Kelly RM. 2015. Alcohol selectivity in a synthetic thermophilicn-butanol pathway is driven by biocatalytic and thermostability characteristics of constituent enzymes. Appl Environ Microbiol 81(20):7187-7200. Loerting T, Bernard J. 2010. Aqueous carbonic acid (H2CO3). ChemPhysChem 11(11):2305- 2309. Mueller M, Takemasa R, Schwarz A, Atomi H, Nidetzky B. 2009. “Short-chain” α-1,4-glucan phosphorylase having a truncated N-terminal domain: Functional expression and characterization of the enzyme from Sulfolobus solfataricus. BBA-Proteins Proteom 1794(11):1709-1714. Peeples TL, Kelly RM. 1995. Bioenergetic response of the extreme thermoacidophile Metallosphaera sedula to thermal and nutritional stresses. Appl Environ Microbiol 61(6):2314-2321.

Rowlett RS. 2010. Structure and catalytic mechanism of the β-carbonic anhydrases. Biochim Biophys Acta - Proteins and Proteomics 1804(2):362-373. Shen CR, Lan EI, Dekishima Y, Baez A, Cho KM, Liao JC. 2011. Driving forces enable high- titer anaerobic 1-butanol synthesis in Escherichia coli. Appl Environ Microbiol 77(9):2905-2915. Smith KS, Ferry JG. 2000. Prokaryotic carbonic anhydrases. FEMS Microbiol Rev 24(4):335-66. Smith KS, Jakubzick C, Whittam TS, Ferry JG. 1999. Carbonic anhydrase is an ancient enzyme widespread in prokaryotes. Proc Natl Acad Sci U S A 96(26):15184-15189. Streit WR, Entcheva P. 2003. Biotin in microbes, the genes involved in its biosynthesis, its biochemical role and perspectives for biotechnological production. Appl Microbiol Biotechnol 61(1):21-31. Sueda S, Li Y-Q, Kondo H, Kawarabayasi Y. 2006. Substrate specificity of archaeon Sulfolobus tokodaii biotin protein ligase. Biochem Biophys Res Comm 344(1):155-159. Thorgersen MP, Lipscomb GL, Schut GJ, Kelly RM, Adams MWW. 2014. Deletion of acetyl- CoA synthetases I and II increases production of 3-hydroxypropionate by the metabolically-engineered hyperthermophile Pyrococcus furiosus. Metab Eng 22(0):83- 88. Wilbur KM, Anderson NG. 1948. Electrometric and colorimetric determination of carbonic anhydrase. J Biol Chem 176(1):147-154. Yoon SH, Reiss DJ, Bare JC, Tenenbaum D, Pan M, Slagel J, Moritz RL, Lim S, Hackett M, Menon AL and others. 2011. Parallel evolution of transcriptome architecture during genome reorganization. Genome Res 21(11):1892-1904. Zeldes BM, Keller MW, Loder AJ, Straub CT, Adams MWW, Kelly RM. 2015. Extremely thermophilic microorganisms as metabolic engineering platforms for production of fuels and industrial chemicals. Front Microbiol 6: article 1209

CHAPTER 3

Recovering chemolithoautotrophy in Sulfolobus acidocaldarius: Insights into restoring CO2

fixation and sulfur oxidation processes

Benjamin M. Zeldes, Andrew J. Loder*, James Counts, Mashkurul Haque, Karl A. Widney,

and Robert M. Kelly

Department of Chemical and Biomolecular Engineering, North Carolina State University,

Raleigh, NC 27695-7905

* Present address: Novozymes North America Inc., Franklinton, NC 27525

To be submitted to Applied and Environmental Microbiology

3.1. Abstract

The order Sulfolobales contains thermoacidophilic archaea that thrive in hot acid (T >

75°C, pH < 3), and exhibit remarkable metabolic diversity. Some species exhibit both heterotrophic and chemolithoautotrophic modes of growth, the latter of which is based on a novel CO2 fixation cycle that includes 3-hydroxypropionate and 4-hydroxybutyrate as intermediates. Some species within the Sulfolobales can oxidize elemental sulfur and metals for bioenergetic benefit. While most efforts to produce bio-based chemicals and fuels have relied on the photoautotrophy of plants, chemolithoautotrophic microbes offer the potential to use energy from inorganic chemicals, such as sulfur, to fix carbon dioxide directly into an industrial chemical. While metabolic engineering efforts in the Sulfolobales have thus far been limited, over the past several years molecular genetic tools have become available for the type species

Sulfolobus acidocaldarius, such that in-frame markerless deletions and chromosomal insertions are possible. As a result, thermoacidophilic archaea could be developed as metabolic engineering platforms that expand current options for industrial biotechnology.

Genetically tractable S. acidocaldarius strains do not grow autotrophically, but early isolates apparently did, and genome sequencing confirmed that most of the genes known to be associated with autotrophy are present. We identified two key sulfur oxidation enzymes, Sulfur oxygenase-reductase (SOR) and Thiosulfate-quinone oxidoreductase (TQO), that are missing in the S. acidocaldarius genome, and created recombinant strains containing one or both enzymes.

The single enzyme SOR strain grew poorly on sulfur, likely due to the accumulation of toxic intermediates in the cytosol. With both enzymes present, elemental sulfur was rapidly oxidized to sulfate, with no apparent fitness issues. However, autotrophic growth on CO2 and elemental sulfur was not observed. Transcriptomic analysis of the parent strain and double knock-in

(SOR+TQO) showed that both responded similarly to the presence of sulfur, but that many enzymes associated with carbon fixation were not transcriptionally responsive. Thus, even with a functional sulfur oxidation pathway present, S. acidocaldarius was incapable of the regulatory switch to autotrophic carbon assimilation. To further explore regulatory mechanisms associated with chemolithoautotrophy, the transcriptomes of Acidianus brierleyi, a facultative chemolithoautotrophic member of the Sulfolobales, were examined during heterotrophic and autotrophic growth on sulfur. Growth was faster during autotrophic growth than heterotrophic, indicating that A. brierleyi, while capable of growing on solely organic substrates, is poorly adapted to it. Preliminary qPCR results confirmed 3HP/4HB genes are up-regulated during autotrophy and suggest that A. brierleyi could serve as a strategic metabolic engineering host for chemolithoautotrophy if molecular genetics tools can be established.

3.2. Introduction

Strains of Sulfolobus acidocaldarius from sites in Yellowstone National Park were among the first extreme thermophiles to be characterized. Initial isolates from a sulfur hot spring grew optimally at temperatures above 70oC and pH below 3, and the ability to grow chemolithoautotrophically by oxidizing sulfur was one of the defining characteristics of the then newly established genus Sulfolobus (Brock et al. 1972). The sulfur oxidizing Sulfolobus species belong to the order Sulfolobales, which contains two other closely related autotrophic genera: the metal-oxidizing Metallosphaera and facultatively anaerobic, sulfur oxidizing/reducing Acidianus

(Figure 3.1), although these phenotypic groupings do not always agree with 16S rRNA taxonomy (Albers and Siebers 2014).

Certain species in the Sulfolobales show great promise in the mining industry, where their ability to solubilize otherwise inaccessible metals from low-grade ores (known as biomining or bioleaching) can reduce both the cost and environmental impact of mining operations (Wheaton et al. 2015). The cells also derive substantial energy from the oxidation of ores, allowing many strains to grow autotrophically. Therefore, a metabolically engineered organism could potentially convert this CO2 not only into biomass but also into an organic industrial chemical, providing a second product stream on top of solubilized metal ions. The oxidation of elemental sulfur, widely available as a byproduct from oil refining, provides even more cellular energy than metal oxidation. Therefore, a lithoautotrophic sulfur-oxidizing metabolic host could convert elemental sulfur and carbon dioxide into sulfuric acid and carbon-chemicals of industrial importance.

Despite being the first of the sulfur-utilizing Sulfolobus species to be isolated, the type strain S. acidocaldarius DSM639 present in culture collections does not grow autotrophically on sulfur (Huber et al. 1989; Nixon and Norris 1992). The reasons for this are unknown, but it is likely that this ability was lost during serial laboratory passages on rich medium. There are two key aspects to chemolithoautotrophy: a carbon fixation pathway for incorporating inorganic carbon (CO2) into cell biomass, and a means of extracting energy from inorganic chemicals. A gene loss or mutation in either of these metabolic processes could explain the lack of lithoautotrophic growth in currently studied S. acidocaldarius strains. Recently, there has been substantial progress in the development of genetic tools for S. acidocaldarius (Wagner et al.

2012), thereby opening up the possibility of recovering chemolithoautotrophy in this archaeon, and establishing it as a metabolic engineering platform for production of carbon-neutral chemicals and fuels.

3.3. Carbon dioxide fixation in the Sulfolobales

The Sulfolobales make use of a uniquely archaeal carbon fixation pathway proceeding through 3-hydroxypropionate and 4-hydroxybutyrate (3HP/4HB) chemical intermediates (Berg et al. 2007). Two carboxylation reactions extend acetyl-CoA by two carbons to form succinyl-

CoA, which is further reduced and rearranged to form acetoacetyl-CoA. This is then cleaved to regenerate the starting substrate and produce an additional molecule of acetyl-CoA. Isotopic labeling studies have shown that two-thirds of cellular intermediates are formed via succinate, while the remaining one-third enters metabolism via acetyl-CoA (Estelmann et al. 2011), as shown in Figure3.2. The 3HP/4HB cycle has been best studied in Metallosphaera sedula (Berg et al. 2007; Estelmann et al. 2011; Hawkins et al. 2014; Loder et al. 2016), but all 16 genes

(encoding 13 enzymes) have been identified in the other Sulfolobales, and there is evidence that all share an autotrophy-associated regulatory system (Leyn et al., 2015).

The original S. acidocaldarius DSM639 genome (CP000077) indicates a frame shift mutation in Saci_1149, which encodes the tenth enzyme of the pathway (E10), shortening it from

575 amino acids to 98. Sequencing of the Saci_1149 region in the parent strain as part of this work indicated that the gene was actually intact, and in fact this error in the original genome sequence was reported previously (one of 36 identified in (Mao and Grogan 2012), Table S1 of their supplemental data).

3.4. Sulfur oxidation in the Sulfolobales

Elemental sulfur is minimally soluble in water, making it a difficult growth substrate under any circumstances. For aerobic acidophiles like Sulfolobus species, this is exacerbated by the fact that most of the more soluble sulfur compounds are only created under reducing

conditions (polysulfides) or at neutral to alkaline pH (thiosulfate). In fact, thiosulfate has been a model growth substrate for studying neutral or alkaline sulfur oxidation, but when added to acidic media quickly converts back to inert elemental sulfur (Kletzin 2006).

The lithotrophic sulfur oxidation pathways of Sulfolobales are not fully understood, but several key enzymes have been identified in Acidianus ambivalens (previously Desulfurolobus ambivalens) (Fuchs et al. 1996; Zillig et al. 1986). A. ambivalens overcomes elemental sulfur’s inert nature via a cytoplasmic sulfur oxygenase reductase (SOR), which catalyzes the oxygen- dependent disproportionation of elemental sulfur to sulfite and hydrogen sulfide (Kletzin 1989;

Veith et al. 2011). The cell derives no energy from this enzyme, but is able to use the more reactive sulfite and sulfide products in subsequent energy-conserving reactions. There are two parallel pathways for sulfite oxidation: a series of membrane-associated oxidoreductases, which feed into the electron transport chain, and a cytoplasmic pathway for substrate level phosphorylation which proceeds through an adenylsulfate intermediate (Zimmermann et al.

1999) (Figure 3.3). Most of the genes responsible for these activities have not been identified.

One exception is the membrane-associated thiosulfate-quinone oxidoreductase (TQO), which co- purifies with components of a terminal quinol oxidase (Müller et al. 2004). The TQO and quinol oxidase presumably form a membrane super-complex, where electrons from sulfur oxidation by

TQO are passed via quinones to the oxidase, which uses them to reduce molecular oxygen to water (Purschke et al. 1997).

This suggests that the simplest pathway to energetic sulfur oxidation could consist of just three enzymes or enzyme complexes: SOR to activate sulfur, TQO to produce reduced quinones by oxidizing sulfur compounds, and the terminal oxidase to convert those quinones to a transmembrane proton gradient (Figure 3.3¸ highlighted in red). Since the terminal oxidase is

native to S. acidocaldarius, only three genes need to be expressed recombinantly, i.e., SOR (sor) and both subunits of TQO (doxD and doxA). While the thiosulfate substrate of TQO is not directly produced by SOR, it forms rapidly through an abiotic reaction between the true products sulfide and sulfite.

Other enzymes from A. ambivalens have been characterized that appear to be involved in oxidation of specific sulfur compounds. Sulfide-quinone oxidoreductase (SQO) is another membrane-associated quinone reductase, but it is specific for sulfide (Brito et al. 2009). Direct growth on hydrogen sulfide has been reported in Sulfolobus metallicus (Morales et al. 2011), and may be possible in other Sulfolobales as well, but the primary role of SQO is likely in the recycle of hydrogen sulfide generated by SOR (see Figure 3.3). Hydrogen sulfide will also react abiotically with sulfite (the other product of SOR) to form thiosulfate, which is the substrate for

TQO. Therefore, these two enzymes would seem to be in constant competition for substrate, unless they are expressed under different circumstances. Additionally, many metallic ores are present as metal-sulfides, and the prevalence of SQO homologs in species of Metallosphaera that lack SOR (Figure 3.3) suggests it may play an important role in bioleaching. A tetrathionate hydrolase (TTH) has also been characterized, which appears to be localized to the outer surface of cells (meaning it must function in the extremely acidic external medium), and is only expressed during growth on tetrathionate (Protze et al. 2011). As a more soluble sulfur substrate, tetrathionate is utilized by some species that are incapable of growth on elemental sulfur, and again homologs are present in some species that lack SOR (Figure 3.3).

3.5. Methods

Cultivation of Sulfolobales

Media for all Sulfolobales species consisted of Brock’s salts (DSM medium #88 minus yeast extract) containing per liter: 1.3g (NH4)2SO4, 0.28g KH2PO4, 0.25g MgSO4·7H2O, 0.07g

CaCl2·2H2O, 0.02g FeCl3·6H2O, 1.8mg MnCl2·4H2O, 4.5mg Na2B4O7·10H2O, 0.22mg

ZnSO4·7H2O, 0.05mg CuCl2·2H2O, 0.03mg Na2MoO4·2H2O, 0.03mg VOSO4·2H2O, and

0.02mg CoSO4·7H2O. A low sulfate Brocks salts formulation was used for experiments monitoring sulfate production, which substituted MgCl2·6H2O for MgSO4. Heterotrophic media contained 1g NZ amine (0.1g if sulfur oxidation was being monitored by pH change) and 2g sucrose for S. acidocaldarius, 1g YE for other species. A low carbon formulation of S. acidocaldarius medium used one-tenth as much NZ amine and sucrose (0.1 and 0.2g).

Autotrophic media contained 10g steam-sterilized elemental sulfur or 2.5g sodium tetrathionate supplemented with 0.01 or 0.1g YE (in the absence of sulfur minimal growth was observed at either level of YE). Uracil auxotrophic strains of S. acidocaldarius were supplemented with 0.01 g uracil in all media. Plate media consisted of a 2x concentrated media solution, pre-heated and mixed with an equal volume of 1.2% phytagel. Media was adjusted with concentrated sulfuric acid to pH 3.3 for S. acidocaldarius, and pH 2 for S. metallicus, A. brierleyi, and M. sedula.

Cells were grown at 70 or 75°C. Standard culture was in 150 mL serum bottles with either foam stoppers or needles through rubber stoppers for air exchange. Autotrophic growth was in sealed bottles with air and 5% vol. CO2 headspace replaced daily. Larger cultures used 3-

L glass bioreactors (Applikon Biotechnology, Delft, Netherlands) with a 2-L working volume, sparged with variable flow-rates (10 to 100mL/min) of air supplemented with ~5% CO2.

Bioreactors were inoculated from bottles with identical media, except that sulfur was omitted from bottles going into sulfur-reactors for MW001, RK06, and RK34.

Creation of recombinant S. acidocaldarius strains

To generate the cloning construct for RK06, genes encoding E10 from M. sedula

(Msed_0406) and SOR from S. tokodaii (ST1127) were cloned into strain MW001 under the control of the strong constitutive promoter of the S-layer subunit SlaA (the 201 bp upstream of

Saci_2355), with the SlaB ribosome binding site (12bp upstream of Saci_2354) in between.

Flanking regions of approximately 500bp upstream and downstream of Saci’s E10 homolog

(Saci_1149) were used to insert the synthetic operon at the Saci_1149 locus, with concomitant deletion of Saci_1149. The uracil prototrophy marker pyrBEF (Sso0614-16) was used for selection.

All cloning fragments were amplified from the appropriate genomic DNA using the primers listed in

Supplemental Table 3.2 and assembled onto a plasmid backbone amplified from pUC19 using

Gibson assembly master mix (New England Biolabs). The Gibson assembly reaction was used to transform One Shot Top10 chemically competent E. coli (ThermoFisher Scientific), and the plasmid was sequence-verified.

To generate the cloning construct for strain RK34, genes for the two subunits doxD and doxA from S. tokodaii (ST1855-1856) were assembled into a pUC19 backbone containing the pyrBEF selection cassette under the control of the strong constitutive promoter from the glutamate dehydrogenase (Saci_0155) gene (Berkner et al. 2010). Flanking regions were used to locate the construct immediately after the inserted SOR gene in strain RK06 (so strain RK34 had strain RK06 as a parent, rather than MW001). Since the doxDA genes were amplified from Stok gDNA together, the original operon structure was preserved.

Genetic methods in S. acidocaldarius relied on the uracil auxotrophic mutant MW001, as described in (Wagner et al. 2012). Briefly, electrocompetent S. acidocaldarius parent strains were transformed with methylated cloning plasmids (generated in an E. coli strain expressing the methylase) and plated on uracil-free medium to select for chromosomal integration. Colonies were sub-cultured in uracil-free liquid media and screened by PCR, then plated on medium containing uracil and 0.1g/L 5-fluoroorotic acid to select for removal of the pyrBEF marker.

Colonies were sub-cultured in medium containing uracil, screened by PCR, and the sequence of the Saci_1149 region was verified (Genewiz).

Monitoring sulfur oxidation

Sulfate concentrations in culture supernatant were measured using a turbidimetric assay modified from (Lundquist et al., 1980). First, samples were centrifuged at 15,000g for 5 minutes to remove sulfur and cells, then diluted to fall within range of the standard curve (0-10 mM

Na2SO4). The assay reagent stock solution contained (per liter): 35 g BaCl2·2H2O, 75 g polyethylene glycol (MW 8000), and 20 mL concentrated HCl. Assay reagent was prepared by adding 50 µL of 10 mM Na2SO4 to 10 mL of reagent stock solution. In a 96-well plate, 100 µL sample was mixed with 75 µL of freshly-prepared assay reagent. The absorbance at 600 nm was read on a Synergy MX plate reader (Biotek Instruments, Winooski, VT). A cubic polynomial fit of the standard curve was used to determine sample concentrations.

RNA extraction

To collect RNA for transcriptomic analysis (RNAseq, Microarray, qPCR), 300-600mL of culture was harvested from early-exponential phase bioreactors and rapidly cooled using a dry ice-ethanol bath. Sulfur particles were removed by slow centrifugation (500 × g for 2 min) or allowing to settle on ice for several minutes after chilling, then sulfur-free cell culture was

decanted into new bottles. Cells were pelleted by 4°C centrifugation at 10,000 × g for 20 min.

Cell pellets were stored at -80°C prior to RNA extraction. RNA was extracted using Trizol reagent (Ambion) and the RNeasy RNA isolation kit (Qiagen) according to vendor instructions, and stored at -80°C prior to analysis. RNA quality was confirmed by the presence of intact 16S and 23S bands on an agarose gel.

Oligonucleotide microarray

RNA was reverse-transcribed (Superscript III, Invitrogen) using a dNTP mix with aminoallyl-dUTP in place of some of the dTTP, to provide a reactive site for later dye-labeling.

The resulting cDNA was hydrolyzed and cleaned up in QIAquick PCR purification columns

(Qiagen). At this point cDNA from each condition was separated into two tubes, pooled with a biological replicate, labeled with Cy3 or Cy5 dye (GE Healthcare, Chicago, IL), then mixed with an oppositely labeled tube from a different condition (experimental design shown in Figure 3.4), and hybridized to microarray slides overnight. Spotted whole-genome 60-mer oligonucleotide microarray slides were used, based on 2278 protein-coding open reading frames (ORFs) from the

S. acidocaldarius DSM 639 genome. Spots were printed in five replicates onto SuperChip aminopropylsilane coated slides (ThermoFisher) using a Qarray Mini microarray printer

(Genetix) and UV crosslinked (600mJ).

Slides were washed, and scanned with a GenePix 4000B microarray scanner (Molecular

Devices). Spots were quantified using GenePix Pro v7. JMP Genomics 9 (SAS, Cary, NC) was used for ANOVA and Loess normalization, and a mixed-effects ANOVA model was to determine differential gene-expression. The Bonferroni correction was used to account for multiple comparisons, with alpha set to 0.05 expression differences of -log10(p-value) greater than 5.5 are considered significant.

qPCR analysis

Quantitative reverse-transcription PCR was done using SsoFast Evagreen Supermix (Bio-

Rad, Hercules, CA). Total RNA previously isolated for microarray analysis was reverse- transcribed using iScript Reverse Transcription supermix (Bio-Rad). cDNA corresponding to ~5 ng total RNA was used in each 20-ul qPCR reaction. qPCR reactions were run in triplicate with one no-RT control and one no-template control. Primers are listed in Supplemental Table 3.2.

RNA sequencing

A. brierleyi cells were collected from heterotrophic and autotrophic cultures each grown in triplicate at bioreactor scale, and RNA extracted as described above. Ribosomal RNA was removed using the bacterial Ribo-Zero rRNA removal kit (Illumina), followed by library construction using NEBNext Ultra Directional RNA Library Prep Kit (NEB) and sequencing on an Illumina Hiseq2500 instrument (125 bp read length) by the North Carolina State University

Genomic Sciences Laboratory (Raleigh, NC). Reads were aligned to the A. brierleyi reference genome (CP029289), and the resulting normalized read counts (RPKM) calculated in Geneious version 8.1.9 (http://www.geneious.com, Kearse et al., 2012). Statistical analysis was performed in JMP genomics 9 using the standard Next-Generation sequencing analysis pipeline. Briefly, data were filtered to remove genes where two or more samples had RPKM values below 2, then normalized by Trimmed Mean of M-component (removes large transcript bias inherent in

RPKM). Differences in gene expression between autotrophic and heterotrophic conditions were determined by ANOVA, with statistical significance defined by a false discovery rate (FDR) with α=0.05 to account for multiple comparisons.

3.6. Results

Growth of recombinant strains

A summary of growth, pH change, and specific sulfate production can be seen in Figure

3.5. In heterotrophic media, strain MW001 grew well either with or without elemental sulfur.

When sulfur was present, some sulfur oxidation was observed, as indicated either by increases in the concentration of sulfate, or as a pH drop in low-peptide media (in standard media the breakdown of peptides leads to an increase in pH, which partially masked the pH drop).

Hypothesizing that the E10 mutation could explain the lack of autotrophic growth in S. acidocaldarius MW001, the M. sedula gene Msed_0406, which we previously confirmed catalyzes the relevant 4HB-CoA synthetase activity (Hawkins et al. 2013), was cloned in its place (along with a sulfur-oxidation associated SOR gene – see below), generating strain RK06.

This strain grew identically to the parent on non-sulfur heterotrophic media. The subsequent finding by targeted genome sequencing that the original Saci_1149 gene was actually intact explains the lack of a strong phenotype.

Given that all the genes known to be required for carbon fixation are present in MW001, its inability to grow autotrophically seems likely to stem from a lack of non-heterotrophic energy sources. The sulfur oxygenase reductase (SOR) has been proposed as the first enzymatic step during the oxidation of elemental sulfur by Sulfolobales, since without it sulfur is too inert to be used as a growth substrate (Kletzin 2006). The gene encoding SOR from S. tokodaii (ST1127) was also included in the cloning construct for RK06 (along with Msed_0406). RK06 cells grew poorly when sulfur was added to heterotrophic media. In sulfur-containing media, the cells exhibited higher sulfate productivity that MW001 on a per-cell basis. But, because of the dramatic growth defect, final sulfate concentrations were often lower than for parent strain

cultures. Both the growth and sulfate production phenotypes were somewhat oxygen-dependent, since the differences with parent strain were more pronounced in aerated bioreactors or open-top bottles than in sealed bottles containing only a needle for oxygen transfer. This supports the hypothesis that the RK06 phenotype is due to functional expression of SOR, since both of the substrates for this enzyme (O2 and elemental sulfur) must be present. We can also rule out the possibility of pH stress contributing to poor growth on sulfur, since the phenotype was present even in pH-controlled bioreactors, and strain RK06 grew well in non-sulfur media down to pH 2.

The poor growth of RK06 on sulfur suggested that the sulfur oxidation pathway remained incomplete. Thiosulfate quinone oxidoreductase (TQO) is a well-characterized enzyme in the

Sulfolobales capable of converting sulfur compound oxidation to cellular energy (Müller et al.

2004), and is missing in MW001. We cloned the TQO enzyme from S. tokodaii into RK06 in an effort to address the strain’s poor growth on sulfur. Strain RK34 again showed no growth defect in standard media. It also grew well when sulfur was added to the media and produced substantially more sulfate than both MW001 and RK06, suggesting that the TQO enzyme is expressed and functional.

It is also worth noting some observations made regarding aeration during growth in bioreactors, which offered both greater control in the form of adjustable gas flowrate, and the ability to monitor dissolved oxygen levels continuously with dissolved O2 sensors. Early on, it became apparent that it was possible to over-oxygenate Sulfolobales grown in bioreactors

(especially the facultative anaerobe Acidianus brierleyi), leading to either very long lag phase or failure to grow at all. Since reactors were always inoculated from bottles, this may have been a response to the shock of moving from poorly oxygenated bottles to the gas-sparged reactors. A solution was to start reactors with a very slow gas flow rate (e.g., 10mL/min), and increase

gradually to a max of 100mL/min after cells were actively growing and O2 saturation started to decline. The decline in oxygen saturation was consistently greater in reactors containing sulfur.

This is to be expected considering that sulfur oxidation consumes two molecules of oxygen for each sulfate molecule produced, and (for autotrophic growth) sulfur oxidation provides considerably less cellular energy than the oxidation of organic compounds. This also leads to a potential caveat to experiments comparing heterotrophic and chemolithoautotrophic growth: it is possible that some of the transcriptomic response is actually a result of the dramatically different oxygen levels between the two conditions, and only indirectly related to differences in growth substrate. Oxygen tension is known to affect the ratio of expression between high and low oxygen-affinity cytochrome oxidases in E. coli (Govantes et al. 2000). This issue is especially important for experiments using bottle-grown cells, where monitoring and controlling oxygen levels is more difficult.

Transcriptomic analysis of effect of sulfur on parent and mutant S. acidocaldarius strains

Transcriptional analysis using a cDNA oligonucleotide microarray comparing MW001 and RK34 during growth on sulfur and non-sulfur media identified a total of 303 genes that were differentially transcribed 2-fold or more under at least one of the four comparisons. The largest differences were contrasts within strains when comparing sulfur and non-sulfur growth.

Relatively few genes were differentially transcribed between strains when both were grown on sulfur, possibly because strains shared a common sulfur response masking more subtle strain-to- strain differences under that condition.

The transcriptomic differences can be broadly grouped into three categories:

(1) differences between strains that are not related to sulfur oxidation (most evident in

Figure 3.7 along the line y=x);

(2) differences between strains that are exacerbated by the presence of sulfur (Figure 3.6

along the lines y=0 and x=0); and,

(3) shared responses to sulfur that can be seen when comparing sulfur and non-sulfur grown

cells of both strains (Figure 3.6 along the line y=x).

These categories are indicated by the highlight color of the ORF# in Table 3.1. As mentioned above, strain-to-strain differences are present, but often less dramatic than sulfur differences, so there are several genes where RK34 is different from MW001 only in the non-sulfur comparison, but these are still grouped into category (1). For the purposes of this experiment, the sulfur- induced strain differences (2) are most interesting, because they indicate ways in which the provision of a synthetic sulfur oxidation pathway has changed the recombinant strain’s response to sulfur, and indicate potential downstream genes in sulfur oxidation. However, all the most dramatic transcriptional responses are in category (3), suggesting that S. acidocaldarius natively has a strong transcriptomic response to sulfur, which the presence of the synthetic pathway has not significantly altered. This, perhaps, reflects the legacy of this strain, which is derived from a parent that at one time grew chemolithotrophically on sulfur.

As expected, the single largest transcriptional difference between the two strains

(category 1) is the down-regulation of Saci_1149 in RK34 relative to MW001 under both non- sulfur and sulfur conditions (56-fold and 14-fold, respectively). The actual fold-changes are likely outside the dynamic range of the oligonucleotide microarray used here, but the dramatic difference confirms that the gene in question was successfully knocked out. Other major differences between the strains include strong down-regulation of Saci_0619 in RK34.

Saci_0619 is annotated as a nucleotidyltransferase, and has high homology to a well- characterized S. tokodaii enzyme (ST0452) similar to the enzymes involved in production of

bacterial cell walls and coat lipopolysaccharides (Liao et al. 1986). This response may indicate stress induced by high overexpression level of the membrane-associated TQO, or due to the use of the strong promoter for the S-layer coat protein (potentially titrating S-layer associated transcription factors). Other stress associated enzymes responding in RK34 include: transcription initiation factor IIB 2 (Saci_1341), and two operons encoding genes annotated as universal stress proteins (Saci_1357-58 and Saci_1638-40).

Another difference between the strains, which may be indicative of a sulfur-oxidation response, is the up-regulation of Saci_2101-2102 and Saci_2085-2093 (SoxABCD-L) in response to sulfur in the modified strain, but not parent strain. These genes share good homology to components of membrane complexes which are responsive to sulfur in M. sedula. Saci_2101-

02 is homologous to Msed_0815 and Msed_0814, components of a membrane-associated complex with homology to the A. ambivalens sulfur-reducing enzyme (Laska et al., 2003), but which has been associated with sulfur compound oxidation in M. sedula (Auernik and Kelly

2008). SoxABCD-L is one of several terminal oxidases found in the Sulfolobales, but it appears to be consistently up-regulated during autotrophic growth, including growth on sulfur compounds (Auernik and Kelly 2010). That these genes are only responding to sulfur in the modified strain suggests they are involved in degrading one of the intermediate sulfur compounds generated by SOR or TQO, one which the native S. acidocaldarius strain does not generate even with sulfur present.

The most dramatic transcriptional responses can be seen in the difference between sulfur and non-sulfur grown cells of both strains. One surprising example is the up-regulation on sulfur of the operon Saci_0097-99, which is homologous to the DoxBCE terminal oxidase complex in

A. ambivalens (Purschke et al. 1997). In A. ambivalens, DoxBCE subunits co-purify with the

DoxDA subunits that make up of TQO, suggesting that they form a membrane super-complex that transfers electrons from sulfur oxidation to molecular oxygen. That the expression of

Saci_0097-99 is still sulfur-responsive, even though MW001 lacks DoxDA homologues, indicates that the regulatory architecture for sulfur oxidation remains (at least partially) intact.

Potential mediators of this sulfur-response include several putative transcriptional regulators. Saci_0006 is significantly down-regulated on sulfur in both S. acidocaldarius strains, and is annotated as an AsrR family transcriptional regulator. The broad AsrR family includes the well characterized sulfur-responsive regulators SqrR in Rhodobacter capsulatus (Shimizu et al.

2017) and SurR in Pyrococcus furiosus (Yang et al. 2010). Both SqrR and SurR detect sulfur via its effect on the local redox environment, causing either the formation or cleavage of intramolecular cysteine bonds. But neither set of interacting cysteine residues is present in

Saci_0006. Saci_2219 is a hypothetical protein also strongly down-regulated on sulfur, and with homology to a S. solfataricus gene (Sso2507) annotated as Fis transcriptional regulator.

Additionally, Saci_1289 is annotated as a serine/threonine protein kinase and is strongly up- regulated on both sulfur conditions. Kinases are commonly involved in regulation (activating or inactivating enzymes by phosphorylation), or in signal transduction cascades.

The operon Saci_1459-1464 is strongly up-regulated on sulfur and includes ribosomal proteins, chaperones, and other genes involved in protein translation. Homologous operons are present in M. sedula (Msed_1629-1634) and M. cuprina (Mcup_0595-0600) that both appear to be up-regulated either generally under autotrophic growth (Msed_1629-31, GEO project:

GSE39944), or specifically during growth on sulfur (Mcup_0596) (Jiang et al. 2014). The presence of SecE (Saci_1461) and FtsY (Saci_1462) candidates in this operon, which are essential for the correct formation of membrane-associated proteins (Angelini et al. 2005),

suggests that sulfur growth requires overproduction of specific membrane proteins or that membrane proteins are less stable in the presence of sulfur and must be replenished more rapidly.

Another cluster of ribosome-associated genes (Saci_0763-0769) is also strongly up-regulated, and includes a transcription elongation factor (NusA) candidate.

Verification of cDNA microarray results

The results from qPCR generally agree with the microarray results, confirming that the native E10 gene Saci_1149 is missing from RK34 (Figure 3.8), but also give insights into genes not present in the native S. acidocaldarius genome (and therefore not included in the microarray data). All the inserted genes (SOR, TQO, and Msed-E10) are expressed much more strongly than the reference gene, as would be expected considering the use of strong constitutive promoters.

The operon structure of the construct is also evident, with the Msed-E10 always expressed more strongly than SOR, which follows it in the synthetic operon controlled by the slaA promoter.

TQO, under the control of the gdhA promoter, is even more strongly expressed, suggesting this promoter may be appropriate when very high transcript levels are needed. One possible conflict with the microarray results is for the native Saci-E10 in the parent strain MW001. The qPCR data suggests this gene is several fold up-regulated on sulfur, whereas the microarray indicated that this gene was possibly slightly down on sulfur (not statistically significant). This disagreement is subtle, and may stem from issues with the selection of the qPCR reference gene, which has been known to bias results (Rocha et al. 2015).

Tests for autotrophic and mixotrophic growth

Corroborating previous reports, in our hands S. acidocaldarius strain MW001, which is the DSM 639 strain with a partial knockout of the pyrF gene Saci_1597 (Wagner et al. 2012), failed to grow autotrophically on sulfur, even under a 10% CO2 headspace in media with uracil

and 0.01% yeast-extract as a vitamin supplement. This was in contrast to the known autotrophic strains S. metallicus and A. brierleyi, which showed robust growth on sulfur even without YE, and possibly M. sedula, which had inconsistent growth on sulfur and seemed to require supplemental YE. Of the modified strains, RK06 failed to grow autotrophically on sulfur, which was not surprising given the apparent toxicity of sulfur even in heterotrophic media. Given

RK34’s strong growth and robust sulfur oxidation in sulfur-containing heterotrophic media, it was considered the most likely to grow autotrophically on sulfur, but no growth was observed.

All three S. acidocaldarius strains also failed to grow autotrophically on tetrathionate, which is a common laboratory growth substrate for sulfur oxidizers because its solubility makes it easier to use than elemental sulfur. Thiosulfate, another possible sulfur substrate, was not tested because at low pH it converts spontaneously to sulfite and elemental sulfur (Kletzin 2006).

While fully autotrophic growth was lacking, the combination of SOR and TQO should allow RK34 to obtain energy by oxidizing sulfur, leading to the possibility that even under heterotrophic conditions it might gain some growth benefit. Experiments with M. sedula found that including hydrogen gas in the headspace as a lithotrophic energy source led to faster growth in heterotrophic media than either heterotrophic or lithoautotrophic growth alone (doubling in

3.7 hours on mixotrophic media vs. 5 on heterotrophic or 12 on autotrophic) (Auernik and Kelly

2010). To test this, we created a “low carbon” media formulation with one-tenth the normal heterotrophic growth substrates, and grew cells on this media in bioreactors with CO2- supplemented air, in the hopes that improved growth in sulfur media would be evident for strain

RK34. All three strains grew poorly on the low-carbon media whether sulfur was included or not, reaching final cell densities roughly one-tenth of those observed on normal heterotrophic

media. The previously observed phenotypes of RK06 (poor growth on sulfur) and RK34 (greater sulfate production on sulfur) seemed to be maintained, but were less dramatic.

Acidianus brierleyi autotrophy and preliminary transcriptomic data

A. brierleyi grew well autotrophically on sulfur in bottles. Heterotrophic growth on yeast- extract was also observed in bottles, but under these conditions cells had to be passaged frequently to remain viable. In bioreactors, A. brierleyi grew under both conditions, but exhibited unusually long lag phases, ranging from 25 to 100 hours on both substrates, before growth became evident. This raised the possibility that a contaminant was the actual source of growth, since research in the Sulfolobales has suffered from impure cultures in the past (Grogan 1989).

However, 16S sequences of gDNA extracted from cells harvested at the end of a week of growth on sulfur versus YE matched each other perfectly. Both differed from the published A. brierleyi

16S sequence (X90477) at a single site, which can probably be explained as a sequencing error in the reference gene, given improvements in sequencing technology since the original report

(Fuchs et al. 1996). Therefore, the cells that grew under both conditions are definitively A. brierleyi.

Sulfate production by A. brierleyi in sulfur media was substantially higher than for the S. acidocaldarius strains. Net sulfate production was 2-3 fold higher than for RK34, and A. brierleyi’s lower cell densities mean the specific sulfate production (on a per-cell basis) was even higher. No sulfate production was observed in YE bioreactors. Growth rates indicated a clear preference for autotrophy, with autotrophic sulfur-grown cells doubling in 19 ± 3 hours, compared to 26 ± 2 for heterotrophic growth (N=3). A doubling time of 24.7 ± 4.9 h has been reported for autotrophic growth of A. brierleyi on tetrathionate (heterotrophic growth rates were not reported), substantially slower than the 9.5 h for mixotrophic growth on both tetrathionate

100

and YE (Wood et al. 1987). In contrast, as described above, M. sedula cells grow fastest mixotrophically (3 h), followed by heterotrophically (5 h), and slowest autotrophically (12 h).

These results suggest that A. brierleyi, while capable of heterotrophic growth, grows best autotrophically. In contrast, M. sedula seems to be optimized for heterotrophic growth. That both strains would grow better under mixotrophic conditions is to be expected, since it allows utilization of organic and inorganic sources at whatever ratio is appropriate for their unique growth preferences.

Previous work has indicated that genes in the 3HP/4HB carbon-fixation pathway are up- regulated under autotrophic conditions in a related extreme thermoacidophile, M. sedula

(Auernik and Kelly 2010). Additionally, genes associated with specific lithotrophic energy metabolisms in S. metallicus were differentially transcribed, with SOR in particular being up- regulated during sulfur oxidation (Bathe and Norris 2007). To determine whether this is also the case in A. brierleyi, qPCR primers for the genes encoding SOR and E1α (subunit of the carboxylase) were used to compare expression of these genes to the reference 16S rRNA gene.

The gene for E1α showed the expected up-regulation on sulfur-grown cells, but SOR expression was not significantly different between the two growth conditions. RNAseq data support this finding, with genes encoding the autotrophy regulator hhcR, E5, and both E8 subunits significantly up-regulated on sulfur autotrophy (all but three of the fifteen 3HP/4HB cycle genes appear to be up-regulated 2-fold or more, although some do not quite meet the threshold for statistical significance), while expression of genes implicated in sulfur oxidation (such as SOR and TQO) are unchanged. This suggests that A. brierleyi constitutively expresses sulfur oxidation pathway genes. Another possibility is that sulfur oxidation genes are only repressed if an alternative lithotrophic substrate is present. Note that lithotrophic substrates (particularly metals)

101

can make it difficult to extract high quality RNA for transcriptional response measurements, but here RNA extracted from both conditions exhibited sharp 16S and 23S bands.

Total mapped reads for the RNAseq samples ranged from a low of 18 M to a high of 40

M, and even after ribo-depletion 20-60% of those reads mapped to ribosomal RNA. A total of

247 genes were differentially transcribed (-log10(p-value) greater than 2.35), but of the 92 genes with changes > 2-fold, 51 were up-regulated under autotrophy and 41 up-regulated during heterotrophy. As noted above, genes up-regulated under autotrophic conditions include several associated with the 3HP/4HB CO2 fixation cycle. Genes encoding SOR and TQO, which are known to be associated with sulfur oxidation, were conspicuously unresponsive, though both were transcribed at relatively high levels as indicated by raw read counts.

3.7. Discussion

The absence of evidence for autotrophic or mixotrophic growth in strain RK34, which contains a seemingly intact carbon fixation cycle and a (rudimentary) pathway for sulfur oxidation, begs the question – What could be missing? The initial assumption was that S. acidocaldarius was a former autotrophic sulfur oxidizer, and therefore minimal changes

(provision of two known sulfur-oxidation enzymes that were missing) would be required to recover that ability. The strong transcriptomic response of both strains to sulfur seems to confirm that sulfur oxidation was part of S. acidocaldarius’s ‘recent’ past. The list of strongly sulfur-responsive genes (yellow in Table 3.1) includes transcriptional regulators (Saci_2219,

0006), and ribosome-associated operons (Saci_0763-0769 and 1459-63), suggesting that the cells are gearing up for a major metabolic re-restructuring. That genes associated with autotrophy in other species, such as DoxBCE (Saci_0097-99), SoxABCD-L (Saci_2085-2092), Saci_1459-63, and Saci_2101-2109 are also up-regulated on sulfur, especially in RK34, is further evidence. It

102

is also interesting that, at least in the genomes sequenced so far, SOR is never found without

TQO (Figure 3.3), so the toxicity of SOR alone, as observed in RK06, may be universal. The reverse is not true, as many species express TQO without SOR.

While both SOR and TQO have previously been characterized in vitro, this is the first report confirming that the two enzymes function together as a pathway in vivo, as indicated by the improved sulfur oxidation and alleviation of SOR induced stress by the addition of TQO.

This is also the first report of cloning a gene intended for lithoautotrophic growth into a member of the Sulfolobales. SOR has been expressed recombinantly in S. solfataricus, but only for enzyme purification - no growth in the presence of sulfur is reported (Albers et al. 2006).

Another encouraging result from the recombinant S. acidocaldarius strains is the fact that both were capable of producing sulfuric acid, despite the fact that neither of the recombinant enzymes is expected to produce sulfate directly. Therefore, the cells must already possess the machinery to oxidize intermediate sulfur compounds to sulfate, though the genes responsible for this activity are not known.

Transcriptomic analysis was key in elucidating the complete set of genes responsible for known enzymatic steps in the 3HP/4HB cycle of Sulfolobales (Hawkins et al. 2013). The current understanding of sulfur oxidation within the Sulfolobales is based on the detection of enzymatic activities in cell extracts, but the genes responsible for many of these activities are yet to be identified. Enzyme detection in cell extracts is also unreliable, since TQO activity, which was key to the strong sulfur oxidation by strain RK34, was not even detected in the cell-extract experiment which has served as the basis for the current sulfur oxidation model (Zimmermann et al. 1999), though TQO was later purified from the same organism (Müller et al. 2004).

Therefore, transcriptomic analysis of the facultative autotroph A. brierleyi has the potential to

103

expand our understanding of the genetic basis of sulfur oxidation in thermoacidophiles, much like it did for carbon dioxide fixation. It will also be interesting to compare transcription in S. acidocaldarius, which is a preferential heterotroph but potentially on the verge of recovering autotrophy, with A. brierleyi, a preferential autotroph that is capable of heterotrophy.

In contrast to the strong sulfur-induced response in S. acidocaldarius, the transcriptomes indicated no response from the genes expected to be involved in 3HP/4HB carbon fixation cycle.

Many of these genes display strong differential regulation under heterotrophic and autotrophic conditions in M. sedula (Auernik and Kelly 2010; Hawkins et al. 2014). Does the lack of response in this experiment indicate that S. acidocaldarius never up-regulates the cycle enzymes, or only that the presence of a chemolithotrophic energy substrate is not sufficient? Perhaps a lack of alternative heterotrophic substrates is required for these autotrophy genes to be induced, but since RK34 failed to grow in the absence of those substrates this is a difficult hypothesis to test.

If the problem is regulation of autotrophy genes, one should consider the autotrophy- associated transcriptional regulator hhrR, present throughout the Sulfolobales (Leyn et al. 2015).

HhcR has been found to bind regulatory sequences upstream of most of the genes in the cycle.

However, it lacks an effector binding site and its transcription is not significantly different under autotrophic or heterotrophic conditions. So, how this regulator detects autotrophic conditions and effects a transcriptional response is not clear. To get a better idea of what was occurring, a meta- analysis of transcriptional data for M. sedula and S. acidocaldarius was done, using raw LSmean expression levels from at least three experimental conditions and converting them to percentiles of all expressed genes. The result makes clear that expression levels of 3HP/4HB genes in S. acidocaldarius look more similar to those of M. sedula grown heterotrophically, albeit with some differences (Figure 3.9). Two of the carboxylase (E1) subunits, E7, and E9 all average in

104

the 30th percentile or below, dramatically lower than M. sedula. In contrast, E3, E5, E10, and the regulator hhcR are consistently expressed at very high levels (90th percentile and above) in S. acidocaldarius. The other obligate heterotroph in the Sulfolobales, S. solfataricus, appears to use portions of the 3HP/4HB cycle during heterotrophic growth as a way to replenish TCA cycle intermediates (Wolf et al. 2016), S. acidocaldarius may have a similar heterotrophic use for these three enzymes.

Intriguingly, hhcR is actually more strongly expressed in Saci cells than in Msed under either heterotrophic or autotrophic conditions, suggesting that its expression may actually be to repress autotrophy, and the alleviation of repression is what leads to the up-regulation of

3HP/4HB genes during autotrophic growth. Unlike Msed, in Saci the hhcR gene is not preceded by an HhcR binding site, so the Saci version does not regulate its own expression like in many of the Sulfolobales (Leyn et al. 2015). Therefore, one way to recover proper regulation of the carbon fixation cycle may be to replace the Saci hhcR gene and its upstream regulatory sequence with a homologous region from an autotrophic species. This, along with additional genes for sulfur oxidation identified in the A. brierleyi RNA sequencing experiment (data pending), could facilitate recovery of a truly autotrophic strain of S. acidocaldarius.

3.8. Acknowledgements

This work was supported with a grant to RMK by the US National Science Foundation

(CBET 1264052). BMZ and JAC acknowledge support from an NIH Biotechnology Traineeship

(2T32GM008776).

105

Table 3.1: Saci Strongly responding genes in microarray Response categories: (1) strain, (2) sulfur induced strain, (3) sulfur Fold changes between the two conditions, double underlined values are statistically significant 34- 34- 34S- MW- Notes and references ORF# Product Gene 34S MW MWS MWS Down on sulfur. Other ArsR family transcriptional sulfur-responsive ArsR- Saci_0006 r 11.0 2.0 -1.2 4.5 regulator family proteins are known (Luebke et al., 2014). cytochrome c oxidase doxBCE terminal oxidase Saci_0097 DoxB f -4.4 2.7 -1.4 -16.6 polypeptide I cluster, purifies with doxDA Saci_0098 oxidase DoxC f -6.0 2.3 -1.4 -19.5 subunits of TQO (Müller et al., 2004). Saci_0099 oxidase DoxE f -6.2 2.0 -1.5 -19.3 Up on sulfur Hyp membrane protein, Saci_0301 hypothetical protein r -10.6 1.1 -5.5 -64.0 strongly up on sulfur ST0425 homologue (Zhang Saci_0619 nucleotidyltransferase r 1.4 -7.7 -14.4 -1.3 et al., 2010) Saci_0763 acetolactate synthase r -3.6 4.7 -1.4 -24.3 Grouping of co-regulated genes (not in an operon) Saci_0765 nucleotidyltransferase f -3.4 4.1 -1.5 -20.6 with annotations related to

Saci_0766 50S ribosomal protein L40 r -3.7 5.1 -1.3 -24.1 translation. Strongly up on transcription elongation factor sulfur, and slightly up in 34- Saci_0767 NusA f -3.6 4.1 -1.8 -25.4 NusA non sulfur. Saci_0768 aspC aspC f -2.5 7.0 -2.5 -43.2 This falls into both category tRNA pseudouridine synthase Saci_0769 r -1.9 2.3 -5.6 -24.6 (1) strain, and (3) sulfur D induced differences. Saci’s native E10 enzyme, Saci_1149 acetyl-CoA synthetase (acs-6) hbcS2 -2.0 -55.7 -14.4 2.0 knocked out in RK34 Sulfur responsive regulatory Saci_1289 serine/threonine protein kinase f -9.5 1.0 -1.5 -14.6 kinase? H/ACA RNA-protein complex Pair of oppositely regulated Saci_1340 Gar1 f 1.3 1.5 -2.5 -2.8 protein Gar1 genes in an operon.

transcription initiation factor Saci_1341 f 17.2 4.6 -1.6 2.4 TINF-IIB2 involved in a IIB 2 stress response? Saci_1357 universal stress protein A f 3.4 3.5 -1.9 -2.0 Up in RK34 without sulfur – serine stress response to foreign Saci_1358 r 4.2 5.0 -1.3 -1.6 hydroxymethyltransferase genes? Saci_1415 hypothetical protein r -29.0 -13.3 1.1 -1.9 Operon of translation associated genes, strongly Saci_1459 50S ribosomal protein L11 r -11.3 -1.3 -2.0 -17.8 up-regulated on sulfur. transcription antitermination Saci_1460 NusG r -18.9 -1.3 -1.8 -25.5 protein NusG Mcup_0596 (Saci_1460 preprotein translocase subunit Saci_1461 SecE r -19.7 -1.7 -2.9 -34.3 homologue) up during SecE autotrophic sulfur growth signal recognition particle Saci_1462 -7.3 1.4 -1.6 -16.2 (Jiang et al., 2014) protein Msed_1630-31 (Saci_1462- 63 homologues) up during Saci_1463 prefoldin subunit alpha r -23.9 -1.5 -1.8 -28.2 autotrophic growth (A. Hawkins Thesis)

106

Table 3.1 (continued) alpha/beta hydrolase fold Saci_1509 r 1.2 -4.9 -3.3 1.8 Up on 34 vs MW protein Saci_1638 Zn-dependent protease f 4.0 4.4 2.9 2.6 Operon up on RK34 without sulfur (stress from foreign Saci_1639 hypothetical protein f 4.2 3.6 1.9 2.2 genes?)

Saci_1640 universal stress protein f 3.8 4.4 2.6 2.2 Down in MW-S? Saci_2085 hypothetical protein Saci_2085 SoxL r -4.1 -2.0 1.1 -2.0 Divergent operons encoding SoxABCD-L, up-regulated Saci_2086 hypothetical protein Saci_2086 r -4.8 -2.1 1.1 -2.1 in RK34, especially on

Saci_2087 cytochrome b r -6.6 -2.8 1.0 -2.2 sulfur. cytochrome c and quinol Saci_2088 SoxB r -2.3 -1.9 -1.3 -1.5 oxidase polypeptide I Msed_0286-91 homologous Saci_2089 quinol oxidase subunit 2 SoxA r -2.7 -2.3 -1.1 -1.3 operon up on sulfur compounds (Auernik and Saci_2090 (2Fe-2S)-binding protein SoxC f -4.0 -3.9 -1.3 -1.3 Kelly, 2008) and during Saci_2091 hypothetical protein f -2.9 -3.4 -1.2 -1.0 autotrophic growth on hydrogen (Auernik and Saci_2092 hypothetical protein SoxD f -5.0 -5.4 -1.3 -1.2 Kelly, 2010). Saci_2101 4Fe-4S ferredoxin r -2.1 1.4 3.7 1.2 Msed_0815-14 homologues (Auernik and Kelly, 2008), part of a sulfur oxidizing Saci_2102 oxidoreductase r -2.2 1.8 4.5 1.1 membrane complex? Up in RK34 on sulfur. Homologous to Fis regulator Saci_2219 hypothetical protein r 10.4 2.1 -1.0 4.8 – down on sulfur.

107

Figure 3.1: 16S tree of Sulfolobales and their lithotrophic metabolisms

E # Enzyme name E1α Acetyl-CoA/propionyl-CoA E1β carboxylase E1γ Malonyl-CoA/succinyl-CoA E2 reductase E3 Malonate semialdehyde reductase E4 3-Hydroxypropionate:CoA ligase 3-Hydroxypropionyl-CoA E5 dehydratase E6 Acryloyl-CoA reductase E7 Methylmalonyl-CoA epimerase E8 α Methylmalonyl-CoA mutase E8 β E9 Succinate semialdehyde reductase E10 4-Hydroxybutyrate-CoA synthetase E11 4-Hydroxybutyryl-CoA dehydratase Crotonyl-CoA hydratase/(S)-3- E12 hydroxybutyryl-CoA dehydrogenase E13 Acetoacetyl-CoA β-ketothiolase Figure 3.2: 3-Hydroxypropionate/4-hydroxybutyraate carbon fixation cycle in Sulfolobales Enzymes of the cycle are indicated by number (order in which they first appear in the pathway – note that E1 and E2 are bi-functional). Inputs of CO2, energy and reducing equivalents are shown, along with key chemical intermediates. Succinyl-CoA synthetase (SCS) generates succinate, which is an important cellular building block.

108

X= contains

homologue to Aamb

protein (BLASTp)

lobus solfataricus lobus solfataricus

Acidianus ambivalens Acidianus brierleyi Acidianus sulfidivorans Acidianus Sulfo acidocaldarius Suflolobus DSM639 tokodaii Sulfolobus metallicus Sulfolobus sedula Metallosphaera hackonensis Metallosphaera cuprina Metallospheara New or updated new new update updat genome e Sulfur oxidation Yes Yes Yes Slow Yes SOR X X X X X TQOa X X X X X X X X X TQOb X X X X X X X X X TTH X X X X X X X X SQO X X X X X X X X X

Figure 3.3: Enzymes for sulfur oxidation and their distribution within the Sulfolobales Enzyme abbreviations and references: TTH = tetrathionate hydrolase (Protze et al., 2011), SQO = sulfide:quinone oxidoreductase (Brito et al., 2009), TQO = thiosulfate:quinone oxidoreductase (Müller et al., 2004), SAOR = sulfide:acceptor oxidoreductase, APSR = adenosine-5’- phosphosulfate (APS) reductase, APAT = adenylylsulfate:phosphate adenyltransferase, and AK = adenylate kinase (Zimmermann et al. 1999). The three enzymes of the proposed simplified sulfur oxidation pathway are highlighted.

109

Figure 3.4: Experimental design for microarray experiment Four slide loop comparing parent strain MW001 to RK34 under both sulfur and non-sulfur conditions

Figure 3.5: Phenotype of recombinant Saci strains on sulfur Parent strain MW001 grows and produces some sulfate on sulfur. RK06 grows poorly on sulfur, but specific sulfate production is higher than MW. RK34 grows well on sulfur and has the highest specific sulfate production.

110

Figure 3.6: Transcriptional differences within strains under sulfur and non-sulfur conditions The boxes indicate 2-fold and 4-fold differential regulation, distance along the horizontal axis reflects differences in parent strain with and without sulfur, while RK34 is on the vertical axis. The prevalence of genes near the diagonal y=x indicates similar responses to sulfur in both strains. Colored ovals highlight groups of genes that appear to be co-regulated.

111

Figure 3.7: Transcriptional differences between strains MW001 and RK34 The boxes indicate 2-fold and 4-fold differences, distance along the horizontal axis reflects differences between MW001 and RK34 when both are grown without sulfur, while the vertical axis is differences when both are grown on sulfur. Saci_1149 is expressed at much lower levels in RK34 than MW001, whether the cells are grown on sulfur or not, whereas the gene cluster Saci_0763-69 appears to be more highly expressed in MW001 than RK34 when both are grown without sulfur, but this trend reverses after sulfur is added.

112

Figure 3.8: qPCR results for key genes in modified strains Expression is relative to the reference gene Saci_0817, coding for a component of the DNA polymerase. No bars indicate cases where amplification never occurred, or where the Cq was later than for control wells.

Figure 3.9: Transcript levels of 3HP/4HB genes in Saci vs. Msed as a percentile of all genes in array Average LSMEAN percentiles for genes from the 3HP/4HB cycle, with min and max observed percentiles shown by the error bars

113

Supplemental Table 3.2: Primers used in this study Name Use Sequence (5’->3’) qPCR primers AJL.157q qPCR detection of TGAGAGATTAAGGTCCGTAGTCAG AJL.158q Msed_0406 – E10 TTCATCGTCAAGAAGCCTTATGTC AJL.160q qPCR detection of GACGGATTTCACAGTTATGGTAGG AJL.161q ST1127 – SOR AAGTATCATCCCTCCTAGGAATCC AJL.196q qPCR reference gene: AGCAGATTCAGATTGAAGGTGATG AJL.197q Saci_0817 TTCAGAGAATGCAGAAGCCTTAAG JAC_Q4F qPCR reference gene: GGATAAGCCATGGGAGTCTTAC JAC_Q4R Saci 16S GCTACAAACGGGAAATAGCCTATA BMZ03q qPCR detection of GGCATATGGTAATGTTGTAGGACC BMZ04q ST1856 (TQOb) GTGCCATGTTCCATTAGATATTTCC BMZ11q qPCR detection of CCTCATGTGGTACGATTTAGATACC BMZ12q Saci_1149 – E10 CCAGCCTTTATAGTAGCTAAGAACG BMZ19q qPCR detection of Abr TGGTAAAACTGCAGATAAAACTTGC BMZ20q accC (E1a) GAACGTAGCTACCAGATTCTATTCC BMZ21q qPCR detection of Abr GCATTTGGAGAACATACAGTAATTCC BMZ22q SOR GGCTCTCTAATAACTGATGGAATCC BMZ23q qPCR reference gene AGGCTGAAACTTAAAGGAATTGGC BMZ24q Abr 16S ACTTAACCGGACATTTCACAACAC A519F Sequencing of the CAGCMGCCGCGGTAA 1492R(l) archaeal 16S gene GGTTACCTTGTTACGACTT Construction of RK06 cloning construct AJL.120 pUC19 backbone TCTAGAGTCGACCTGCAGG AJL.121 GGTACCGAGCTCGAATTC AJL.128 Saci_1149 5’ flanking GAATTCGAGCTCGGTACCCCACTCTAATTTCTCTCA AJL.129 region CCAAGTACTAGAACTGCTCAAAATTTAATATTGAGCTTAAT AA AJL136 Saci_1149 3’ flanking CCTGAATGAATATCATTTATAATAAATCCTTTTTTAAATTTT region TCTCGCGTAACAAGAACC AJL.137 CTGCAGGTCGACTCTAGAGAGCTAATTTTGGAGG AJL.142 ST1127 (SOR) gene GAACGAGTACGTCTTCTGAGGGTGTATGTGTATGCCGAAAC CATA AJL.135 ATTTAAAAAAGGATTTATTATAAATGATATTCATTCA AJL.180 Saci_2355 (SlaA) AATCTTTAAATAAATTATTAAGCTCAATATTAAATGTCTAA promoter AGGGTGTTCTT AJL.140 ACTTTTCACTTTCAAGG AJL.141 Msed_0406 (E10) gene GATGTTTTTCCTTGAAAGTGAAAAGTATGGTTACCGTTCAA AJL.133 TCAGAAGACGTACTCG AJL.173 Sso pyrBEF selection TTTGAGCAGTTCTAGTACT AJL.174 cassette CTGCAGGTCGACTCTAGAGACCGGCTATTTTT AJL.183 Screening/sequencing GGTATTATGGAAAGGAGTGTCGTC AJL.184 Saci_1149 site (RK06 & CCCCTCTTAGCAACTTCAACTAAG 34)

114

Supplemental table 3.2 (continued) Construction of RK34 cloning construct BMZ_100 Backbone, pyrBEF, and TAAATCCTTTTTTAAATTTTTCTCGCGTAACAAGAACC BMZ_101 3’ flanking from RK06 GGTACCGAGCTCGAATTCACTG plasmid BMZ_102 End of SOR gene of ggccagtgaattcgagctcggtaccGACGGATTTCACAGTTATG BMZ_103 RK06 (new 5’ flanking aaacagtggagaaattatcactgtGAGAAAAATTTAAAAAAGGATTTATT region) ATAAATG BMZ_104 Saci_0155 (gdh) ttaaatttttctcACAGTGATAATTTCTCCACTGT BMZ_105 promoter ttcattgaagtttatctttgacatATTTTATCCAAGTAATTCTTCTCTC BMZ_106 ST1855-56 (TQO) genes gagaagaattacttggataaaatATGTCAAAGATAAACTTCAATGAAG BMZ_107 gttcttgttacgcgagaaaaatttaaaaaaggatttaTCAACTACTTATCGTGAAG TAAG

115

3.9. References

Albers S-V, Siebers B. 2014. The Family Sulfolobaceae. In: Rosenberg Eea, editor. The Prokaryotes. Berlin: Springer-Verlag. p 323-346. Albers SV, Jonuscheit M, Dinkelaker S, Urich T, Kletzin A, Tampe R, Driessen AJ, Schleper C. 2006. Production of recombinant and tagged proteins in the hyperthermophilic archaeon Sulfolobus solfataricus. Appl Environ Microbiol 72(1):102-11. Angelini S, Deitermann S, Koch HG. 2005. FtsY, the bacterial signal-recognition particle receptor, interacts functionally and physically with the SecYEG translocon. EMBO Reports 6(5):476-481. Auernik KS, Kelly RM. 2008. identification of components of electron transport chains in the extremely thermoacidophilic crenarchaeon Metallosphaera sedula through iron and sulfur compound oxidation transcriptomes. Appl Environ Microbiol 74(24):7723-7732. Auernik KS, Kelly RM. 2010. Physiological versatility of the extremely thermoacidophilic archaeon Metallosphaera sedula supported by transcriptomic analysis of heterotrophic, autotrophic, and mixotrophic growth. Appl Environ Microbiol 76(3):931-935. Bathe S, Norris PR. 2007. Ferrous iron- and sulfur-induced genes in Sulfolobus metallicus. Appl Environ Microbiol 73(8):2491-2497. Berg IA, Kockelkorn D, Buckel W, Fuchs G. 2007. A 3-hydroxypropionate/4-hydroxybutyrate autotrophic carbon dioxide assimilation pathway in archaea. Science 318(5857):1782- 1786. Berkner S, Wlodkowski A, Albers S-V, Lipps G. 2010. Inducible and constitutive promoters for genetic systems in Sulfolobus acidocaldarius. Extremophiles 14(3):249-259. Brito Ja, Sousa FL, Stelter M, Bandeiras TM, Vonrhein C, Teixeira M, Pereira MM, Archer M. 2009. Structural and functional insights into sulfide:quinone oxidoreductase. Biochemistry 48(24):5613-5622. Brock TD, Brock KM, Belly RT, Weiss RL. 1972. Sulfolobus: A new genus of sulfur-oxidizing bacteria living at low pH and high temperature. Arch Mikrobiol 84(1):54-68. Estelmann S, Hügler M, Eisenreich W, Werner K, Berg IA, Ramos-Vera WH, Say RF, Kockelkorn D, Gad, apos and others. 2011. Labeling and enzyme studies of the central carbon metabolism in Metallosphaera sedula. J Bacteriol 193(5):1191-1200. Fuchs T, Huber H, Burggraf S, Stetter KO. 1996. 16S rDNA-based phylogeny of the archaeal order sulfolobales and reclassification of Desulfurolobus ambivalens as Acidianus ambivalens comb. nov. Syst Appl Microbiol 19(1):56-60. Govantes F, Orjalo AV, Gunsalus RP. 2000. Interplay between three global regulatory proteins mediates oxygen regulation of the Escherichia coli cytochrome d oxidase (cydAB) operon. Mol Microbiol 38(5):1061-1073. Grogan DW. 1989. Phenotypic characterization of the archaebacterial genus Sulfolobus - Comparison of 5 wild-type strains. J Bacteriol 171(12):6710-6719. Hawkins AB, Adams MWW, Kelly RM. 2014. Conversion of 4-hydroxybutyrate to acetyl-coa and its anapleurosis in the Metallosphaera sedula 3-hydroxypropionate/4- hydroxybutyrate carbon fixation pathway. Appl Environ Microbiol 80(8):2536-45 Hawkins AS, Han Y, Bennett RK, Adams MWW, Kelly RM. 2013. Role of 4-hydroxybutyrate- CoA synthetase in the CO2 fixation cycle in thermoacidophilic archaea. J Biol Chem 288(6):4012-4022.

116

Huber G, Spinnler C, Gambacorta A, Stetter KO. 1989. Metallosphaera sedula gen, and sp. nov. represents a new genus of aerobic, metal-mobilizing, thermoacidophilic archaebacteria. Syst Appl Microbiol 12(1):38-47. Jiang CY, Liu LJ, Guo X, You XY, Liu SJ, Poetsch A. 2014. Resolution of carbon metabolism and sulfur-oxidation pathways of Metallosphaera cuprina Ar-4 via comparative proteomics. J Proteom 109:276-289. Kletzin A. 1989. Coupled enzymatic production of sulfite, thiosulfate, and hydrogen sulfide from sulfur: purification and properties of a sulfur oxygenase reductase from the facultatively anaerobic archaebacterium Desulfurolobus ambivalens. J Bacteriol 171(3):1638-1643. Kletzin A. 2006. Metabolism of Inorganic Sulfur Compounds in Archaea. In: Garrett RA, Klenk H-P, editors: Blackwell Publishing Ltd. p 261-274. Leyn SA, Rodionova IA, Li X, Rodionov DA. 2015. Novel transcriptional regulons for autotrophic cycle genes in Crenarchaeota. J Bacteriol 197(14):2383-2391. Liao H, McKenzie T, Hageman R. 1986. Isolation of a thermostable enzyme variant by cloning and selection in a thermophile. Proc Natl Acad Sci U S A 83(3):576-80. Loder AJ, Han Y, Hawkins AB, Lian H, Lipscomb GL, Schut GJ, Keller MW, Adams MWW, Kelly RM. 2016. Reaction kinetic analysis of the 3-hydroxypropionate/4-hydroxybutyrate CO2 fixation cycle in extremely thermoacidophilic archaea. Metab Eng 38(October):446- 463. Mao D, Grogan D. 2012. Genomic evidence of rapid, global-scale gene flow in a Sulfolobus species. ISME Journal 6(8):1613-1616. Morales M, Arancibia J, Lemus M, Silva J, Gentina JC, Aroca G. 2011. Bio-oxidation of H2S by Sulfolobus metallicus. Biotechnol Lett 33(11):2141-2145. Müller FH, Bandeiras TM, Urich T, Teixeira M, Gomes CM, Kletzin A. 2004. Coupling of the pathway of sulphur oxidation to dioxygen reduction: characterization of a novel membrane-bound thiosulphate:quinone oxidoreductase. Mol Microbiol 53(4):1147-1160. Nixon A, Norris PR. 1992. Autotrophic growth and inorganic sulphur compound oxidation by Sulfolobus sp. in chemostat culture. Arch Microbiol 157(2):155-160. Protze J, Müller F, Lauber K, Naß B, Mentele R, Lottspeich F, Kletzin A. 2011. An Extracellular Tetrathionate Hydrolase from the Thermoacidophilic Archaeon Acidianus Ambivalens with an Activity Optimum at pH 1. Front Microbiol 2: article 68 Purschke WG, Schmidt CL, Petersen A, Schäfer G. 1997. The terminal quinol oxidase of the hyperthermophilic archaeon Acidianus ambivalens exhibits a novel subunit structure and gene organization. J Bacteriol 179(4):1344-1353. Rocha DJP, Santos CS, Pacheco LGC. 2015. Bacterial reference genes for gene expression studies by RT-qPCR: survey and analysis. Antonie van Leeuwenhoek 108(3):685-693. Shimizu T, Shen J, Fang M, Zhang Y, Hori K, Trinidad JC, Bauer CE, Giedroc DP, Masuda S. 2017. Sulfide-responsive transcriptional repressor SqrR functions as a master regulator of sulfide-dependent photosynthesis. Proc Natl Acad Sci U S A 114(9):2355-2360. Veith A, Urich T, Seyfarth K, Protze J, Frazão C, Kletzin A. 2011. Substrate pathways and mechanisms of inhibition in the sulfur oxygenase reductase of Acidianus Ambivalens. Front Microbiol 2: article 37 Wagner M, van Wolferen M, Wagner A, Lassak K, Meyer BH, Reimann J, Albers S-V. 2012. Versatile genetic tool box for the crenarchaeote Sulfolobus acidocaldarius. Front Microbiol 3:214-214.

117

Wheaton G, Counts J, Mukherjee A, Kruh J, Kelly R. 2015. The confluence of heavy metal biooxidation and heavy metal resistance: Implications for bioleaching by extreme thermoacidophiles. Minerals 5(3):397-451. Wolf J, Stark H, Fafenrot K, Albersmeier A, Pham TK, Muller KB, Meyer BH, Hoffmann L, Shen L, Albaum SP, et al. 2016 A systems biology approach reveals major metabolic changes in the thermoacidophilic archaeon Sulfolobus solfataricus in response to the carbon source L-fucose versus D-glucose. Mol Microbiol 102(5):882-908. Wood AP, Kelly DP, Norris PR. 1987. Autotrophic growth of four Sulfolobus strains on tetrathionate and the effect of organic nutrients. Arch Microbiol 146(4):382-389. Yang H, Lipscomb GL, Keese AM, Schut GJ, Thomm M, Adams MWW, Wang BC, Scott RA. 2010. SurR regulates hydrogen production in Pyrococcus furiosus by a sulfur-dependent redox switch. Mol Microbiol 77(5):1111-1122. Zillig W, Yeats S, Holz I, Böck A, Rettenberger M, Gropp F, Simon G. 1986. Desulfurolobus ambivalens, gen. nov., sp. nov., an autotrophic archaebacterium facultatively oxidizing or reducing sulfur. Syst Appl Microbiol 8(3):197-203. Zimmermann P, Laska S, Kletzin A. 1999. Two modes of sulfite oxidation in the extremely thermophilic and acidophilic archaeon Acidianus ambivalens. Arch Microbiol 172(2):76- 82.

118

CHAPTER 4

Synthetic enzymatic pathway for thermophilic acetone production incorporating an

unusually thermostable enzyme from mesophilic Clostridium acetobutylicum

Benjamin M. Zeldes, Christopher Straub, Robert M. Kelly

1Department of Chemical and Biomolecular Engineering, North Carolina State University,

Raleigh, NC 27695-7905

To be submitted to Biotechnology and Bioengineering

119

4.1. Abstract

One of the potential advantages of an extremely thermophilic metabolic engineering host is facilitated purification of volatile chemicals from the vapor phase of an actively growing culture. This process would reduce purification costs and alleviate toxicity to the cells by continuously removing product, a process we are calling “bio-reactive distillation”. While synthetic pathways for use in extreme thermophiles can be inspired by existing mesophilic versions, there need to be thermophilic homologs of the constituent enzymes, which may or may not be known or available. The pathway for production of acetone from acetyl-CoA in the mesophilic bacterium Clostridium acetobutylicum requires three enzymes: thiolase (Thl), acetate:acetoacetyl-CoA CoA transferase (CtfAB), and acetoacetate decarboxylase (Adc). A previously reported thermophilic acetone production pathways functioned only as high as 55°C, because the Adc from the moderate thermophile Bacillus amyloliquefaciens lacked sufficient thermostability. Here, we report a synthetic enzymatic pathway for acetone production that functions up to at least 70°C in vitro, made possible by the unusual thermostability of Adc from the mesophile C. acetobutylicum and the anticipated thermostability of heterodimeric CoA transferases from extreme thermophiles Thermosipho melanesiensis and Caldanaerobacter subterraneous. The CoA transferase beta subunits, when purified separately, survived only a few minutes at 70°C. In contrast, the intact alpha-beta complex had a half-life of hours at 70°C. The three enzymes were shown to produce acetone in vitro at temperatures of at least 70°C, paving the way for metabolic engineering of an extreme thermophile for acetone production to demonstrate the concept of bioreactive distillation.

120

4.2. Introduction

Acetone is widely used industrially as a solvent and polymer precursor. An acetone shortage during World War I led to one of the first instances of large-scale industrial fermentation, leveraging the ability of Clostridium acetobutylicum to convert organic acids into acetone, butanol, and ethanol (Sauer, 2016). However, routes to produce acetone from petroleum eventually eclipsed fermentative processes, and most recent research on C. acetobutylicum has focused on production of bio-butanol as a drop-in replacement for gasoline, in which context acetone is often seen as an undesired byproduct to be minimized (Jang et al. 2012). However, acetone remains a valuable product in its own right (Luo et al. 2016). As a commodity chemical, acetone is valued less than butanol but more than ethanol. Acetone is a feedstock in the production of bisphenol A and methyl methacrylate-based polymers, used as a fuel additive, and of course as a solvent (direct solvent use represents ~30% of total demand) (Wu et al. 2007).

Acetone is also considerably less toxic to cells than butanol, which can have severe toxic affects even at low concentrations due to its fluidizing effects on the cell membrane (Peabody and Kao

2016). Additionally, recombinant expression of enzymes for acetone production in native acetate producers has improved growth (Bermejo et al. 1998; Shaw et al. 2015), since acetone appears to be less toxic than acetate, especially at low pH.

Acetone’s high volatility makes it a potentially strategic metabolic engineering product for an extremely thermophilic host (Topt >70°C), where continuous recovery of product from the bioreactor, termed “bio-reactive distillation”, could be possible even at atmospheric pressure and relatively moderate titers (Figure 4.1). However, no microorganism growing at such elevated temperatures is known to produce acetone. In fact, production of solvents is rare among the fermentative extreme thermophiles, which tend to produce organic acids and hydrogen gas

121

instead. The highest reported temperatures for native production of acetone, ethanol, and butanol are 43°C (Weimer 1984), 72°C (Svetlitchnyi et al. 2013), and 58°C (Freier-Schroder et al. 1989), respectively, while metabolically engineered hosts expressing recombinant enzymes have allowed production as high as 55°C (Shaw et al. 2015), 75°C (Basen et al. 2014), and 60°C

(Keller et al. 2015), respectively, as summarized in Table 4.1.

The lack of native producers means that generating acetone in an extremely thermophilic host will require establishing a synthetic production pathway using enzymes from multiple thermophilic organisms, as has been reported previously for n-butanol (Keller et al. 2015).

Acetone production pathways at elevated temperatures rely on the native, mesophilic three- enzyme pathway in C. acetobutylicum as a template. First, Thiolase (Thl) extends carbon chains by condensing two acetyl-CoA molecules into an acetoacetyl-CoA. Acetoacetyl-CoA:acetate

CoA transferase (Ctf) then transfers the CoA subunit from acetoacetyl-CoA to acetate, generating acetoacetate and recovering an acetyl-CoA. Finally, acetoacetate decarboxylase (Adc) decarboxylates acetoacetate to acetone and CO2. Of course, thermostable, thermoactive versions of these three enzymes needed to be identified.

The most straightforward way of identifying thermostable enzymes is by searching for homologs expressed in thermophiles with optimum growth temperatures around the desired working temperature (in this case, at least 70oC). This was the approach taken previously (Shaw

o et al. 2015), utilizing Thl from Thermoanaerobacterium thermosaccharolyticum (Topt 60 C), Ctf

o from Thermosipho melanesiensis (Topt 70 C), and Adc from Bacillus amyloliquefaciens (Topt

50oC). While this collection of enzymes successfully produced acetone at 55oC, the Thl and Adc were from only moderately thermophilic organisms, and were unlikely to work for our process.

122

Thermophilic homologs can be found for many mesophilic enzymes, but in the case of highly specialized metabolisms, or reactions that are thermodynamically less favorable at higher temperatures, no thermophilic candidates may be available. However, there are also cases where enzymes isolated from mesophilic organisms show unusual thermostability. In fact, several highly thermostable industrial enzymes have been derived from mesophilic hosts, such as the α- amylase from Bacillus licheniformis (Saito 1973). Another example can be found in the acetone pathway; acetoacetate decarboxylase from C. acetobutylicum was reported to be active at 70°C when first characterized in partially purified forms (Davies 1943). Subsequent work determined the enzyme retained 50% activity after 30 minutes at 80°C, and activity actually increased following an hour incubation at 70°C (Autor and Fridovich 1970).

C. acetobutylicum grows optimally at 35°C, so why it would have an enzyme activated at

70°C is not clear. One hint comes from the observation that C. acetobutylicum spores are activated by brief heat shock at temperatures as high as 80°C, and that the sporulation process is regulated in concert with solvent formation (Jones and Papoutsakis 2014). The role of heat in the activation of both suggests that Adc may play an important role in the transition from inactive spore to actively growing cell.

A number of determinants for protein thermostability have been proposed, although many are not well supported by data (Kumar et al. 2000). Proposed amino acid differences in thermophiles include a reduction in the thermolabile amino acids asparagine, glutamine, and cysteine, and increases in rigidity and compactness due to hydrophobic amino acids, as well as proline and alanine (Zhou et al. 2008). Predictive features, such as tighter packing, more salt bridges, or increased alpha helices, require structural data for the protein of interest or a close homolog in order to be useful, and online tools are available that predict protein stability with

123

varying degrees of accuracy (Khan and Vihinen 2010). SCooP, a structure based bioinformatics tool for predicting protein stability (Pucci et al. 2017), predicts a relatively high melting point for all the Adc structures available from PDB (Table 4.2), suggesting that thermostability may be a unifying feature of this whole class of proteins, rather than something unique to Cace Adc. One caveat is that SCooP is designed to work on monomeric proteins, so accuracy may be low with the dodecameric Adc.

In addition to searching for either thermophilic or mesophilic enzymes that already exhibit thermostability, an alternative approach would be to engineer thermostability into currently thermolabile proteins. Here, the two competing approaches can be broadly classified as either random mutagenesis or rational design. Random mutagenesis can be highly effective if a simple method for subsequent selection is available, as in the case where a kanamycin resistance gene had its thermostability increased by 20oC (Hoseki et al. 1999). Rational design requires an understanding of the factors that dictate thermal tolerance, which as discussed above is still lacking. However, one rational design strategy that has worked for several protein types is introducing new cysteine-cysteine disulfide bonds (Liu et al. 2014), although this may also hinder flexibility important for catalytic function.

The question addressed here is whether a synthetic biochemical pathway for acetone production can be designed and demonstrated to function at 70°C or higher. If available, this would pave the way for developing metabolic engineering strains that form acetone at temperatures high enough to facilitate its recovery and purification.

124

4.3. Methods

Identification of thermophilic gene candidates

Thermophilic homologs to C. acetobutylicum acetone pathway enzymes Thl

(AAK80816.1), CtfAB (NP_149326 & NP_149327), and Adc (NP_149328) were identified by

BLASTp (NCBI) searches limited to microbial groups known to be made up primarily of thermophilic organisms. Results were narrowed down to the most promising candidates by focusing on organisms with optimum growth temperatures above 70°C.

Protein expression

Genes encoding candidate enzymes were cloned from genomic DNA (Tmel CtfA, CtfB,

Csub Thl, CtfA, CtfB, Case Adc) or synthesized (Vdis and Sulfolobus sp. Adc) (Integrated DNA technologies, Skokie, IL), and inserted into PCR amplified plasmid backbones pET-46, pRSF, or pCDF (EMD Millipore, Billerica, MA) with flanking regions appropriate for Gibson assembly, which was performed using a 2x master mix (New England Biolabs, Ipswich, MA). All constructs included the N-terminal 6-histidine tag except for the Adc enzymes. Cell lines used for cloning were chemically competent E. coli cells 5-α (New England Biolabs) for plasmid screening and amplification, and Rosetta 2 (EMD Millipore) for protein expression. Protein expression was carried in shake flasks containing one liter of ZYM-5052 autoinduction medium

(Studier 2005) and appropriate antibiotics at 37°C for 20-24 hours. Cells were harvested by centrifugation at 10,000g for 10 minutes.

Protein purification

Cells were re-suspended in 5mL/mg pellet weight of IMAC buffer A (300mM NaCl,

50mM sodium phosphate, 1mM MgCl2, 20mM imidazole, 10% glycerol, pH8 for Csub Thl,

125

increased to 20% glycerol and 100mM sodium sulfate added for all four Ctf subunits), or lysis buffer (50mM Potassium Phosphate, pH 5.9) for Adc, and lysed in a French pressure cell. Lysed cells were heat-treated at 65°C for 10 minutes to denature E. coli proteins, and centrifuged at

24,000g for 20 minutes to generate soluble heat-treated cell extract.

No further purification was performed for the Adc candidates. The heat-treated cell extract was buffer exchanged into storage buffer (50 mM Tris-HCl, 100 mM sodium chloride,

50% glycerol, pH 7.5) in Vivaspin 20 10,000 Da MWCO filters (Sartorius, Goettingen,

Germany).

Histidine-tagged proteins were purified by immobilized metal affinity chromatography using 5 mL HisTrap HP columns. Binding was in IMAC buffer A (described above), followed by elution in a gradient up to 500mM Imidazole. Fractions containing the elution peak were pooled, concentrated, and buffer exchanged as described above. Storage buffers consisted of: 50 mM Tris-HCl, 100 mM sodium chloride, 1 mM DTT, 50% glycerol, pH 7.5, for Csub Thl;

50mM MOPS, 500mM ammonium sulfate, 50% glycerol for all four Ctf subunits. Purified enzymes were stored at -20°C.

Enzyme assays

Individual enzymes were assayed for activity on a Lambda 25 spectrophotomer with

PTP-1 Peltier heaters (Perkin Elmer, Waltham, MA) using 100uL Quartz Cuvettes (Starna Cells,

Atascadero, CA). All assays were performed at 70°C unless otherwise noted. Thiolase assay mixture consisted of 100mM MOPS pH 7.9, 0.3 mM NADH, and appropriately diluted enzyme.

The reaction was started by adding 0.5 mM acetyl-CoA substrate and consumption of NADH was monitored by absorbance at 340 nm. Acetate:acetoacetyl-CoA CoA transferase was monitored by tracking consumption of the Mg-enolate form of acetoacetyl-CoA by decrease in

126

absorbance at 310 nm (Cary et al. 1990). The assay mixture consisted of 100mM Tris, 150 mM potassium acetate, 20 mM MgCl2 and 5% glycerol at pH 7.5. Acetoacetyl-CoA was added, and absorbance was monitored for 30 seconds to establish the non-enzymatic rate of acetoacetyl-CoA hydrolysis, then an appropriately diluted mix of the Ctf subunits was added (subunit B was always in slight excess). For inhibition studies substrate concentrations ranged from 15 to 720 mM acetate and 0.03 to 0.4 mM acetoacetyl-CoA. Acetoacetate decarboxylase assay mix consisted of 50 mM potassium phosphate, 300mM lithium acetoacetate, pH 5.9. The assay was started by adding appropriately diluted enzyme, and consumption of acetoacetate was monitored by absorbance at 290 nm. Thermal inactivation studies involved enzymes diluted to appropriate concentration in assay buffer and incubating in a thermocycler at 70°C for a range of times, then assaying for residual activity.

The three enzymes were assayed together as an in vitro pathway in an mixture containing

100mM Tris, 10mM MgCl2, 150mM potassium acetate, 5mM acetyl-CoA, pH 7.5, with enzymes added to activities of 5 U/mL for Thl and Adc, 15 U/mL for Ctf. Controls consisted of reaction mixture with each enzyme missing individually, and a no-enzyme control. The resulting mixtures were incubated at 70°C in a thermocycler, and acetone was detected by GC.

Other methods

Acetone was detected on a GC-2014 gas chromatograph (Shimadzu, Kyoto, Japan) equipped with a ZB-WAXplus 30 m long, 0.53-mm ID capillary column (Phenomenex,

Torrance, CA) and flame ionization detector. The GC oven temperature was initially held at

35°C for 3 min, increased to 150°C at 20°C/min, and held for 6 min. The injector was held at

220°C and FID detector at 300°C. Nitrogen was used as the carrier gas at a column flow of 30 cm/s. Samples of 0.1 µL were injected with a 1:10 split ratio using an AOC-20i autosampler.

127

Protein concentrations were determined by the Bradford method with BSA standards.

SDS-PAGE was done using BioRad TGX 4 to 12% gels with standard Tris-glycine buffer.

Samples were heated at 95°C for 15 minutes to ensure denaturation of the thermostable enzymes, and gels were stained with GelCode blue. Blue native PAGE used the Novex NativePAGE kit, with a 4-16% Bis-Tris gel. Imaging relied on running the gel with Dark Blue Cathode Buffer, then de-staining overnight.

4.4. Results

Thermophilic acetone pathway candidates

Blast searches for homologs to the C. acetobutylicum acetone pathway enzymes indicated several promising candidates. The thiolase from Caldanaerobacter subterraneus subsp. tengcongensis (formerly Thermoanaeobacter tencongensis) is 68% identical at the amino acid level, and in fact was already characterized and used as part of a synthetic pathway for n-butanol production, which starts with the same condensation reaction of two acetyl-CoA molecules

(Loder et al. 2015). C. subterraneous also topped the CoA transferase prospects, with 65% identity for both subunits, followed closely by Thermosipho melanesiensis with 53 and 64% identity for the alpha and beta subunits, respectively. The T. melanesiensis enzyme hit was also not surprising, having been identified in the search for a more moderately thermophilic acetone pathway functional at 55°C (Shaw et al. 2015). However, while that report confirmed activity by acetone formation at 55°C and room temperature activity assay of unpurified cell extracts, there was no characterization of the enzyme or its thermostability. The same study utilized an acetoacetate decarboxylase from Bacillus amyloliquefaciens with good sequence identity (65%) to C. acetobutylicum, but this organism’s optimum growth temperature of 50°C was too low for

128

the extremely thermophilic pathway. Unfortunately, the acetoacetate decarboxylase is a highly specialized enzyme for solvent production, and the only extremely thermophilic candidates identified, from Vulcanisaeta distributa and a metagenomics sequence annotated as ‘from a

Sulfolobus species’, had rather poor sequence homology (< 30% amino acid ID). One surprising finding in the first report characterizing the activity of the C. acetobutylicum acetoacetate decarboxylase (Davies 1943) indicated that the enzyme retained activity up to 70°C, suggesting that the native enzyme could be used even in an extremely thermophilic process. The results of the search for extremely thermophilic acetone enzyme homologs are summarized in Table 4.3, and alignments to C. acetobutylicum enzymes are in Figure 4.2.

Expression of acetone pathway enzymes

The thiolase from C. subterraneus had been purified and characterized previously (Loder et al., 2015). The gels in Figure 4.4 confirm that the IMAC purified preparation used here is of high purity and forms the expected homotetramer.

Therefore, we turned our attention to the next enzyme in the pathway, Ctf. The CtfAB complex has been purified from E. coli (Sramek and Frerman 1975b) and C. acetobutylicum

(Wiesenborn et al. 1989), and in both cases is described as a fastidious enzyme requiring at least

20 wt% glycerol and 500mM ammonium sulfate to remain stable. Both reports purified the alpha and beta subunits together as a complex, as did a subsequent study expressing the Cace enzyme in E.coli (Cary et al. 1990). The salting out purification described by these earlier reports was utilized here for mixtures of the T. melanesiensis and C. subterraneus subunits expressed individually in E. coli. This resulted in partially purified cell extracts with Ctf activity at 70oC

(confirmation the C. subterraneous enzyme functions as a thermophilic Ctf), but purified enzyme was needed for subsequent assays.

129

Histidine-tagged fusion proteins allow for easy purification (as was used for Thl), but the tag can interfere with enzyme activity if the tagged end (N or C terminus) is near the enzyme active site or dimerization interfaces. Fortunately, a crystal structure of the E. coli CtfAB complex is available (PDB ID: 5DBN), and the active site for all “Family I” CoA transferases consists of well-conserved sequences in both subunits near a nucleophilic glutamate on the beta subunit (Heider 2001). An analysis of the crystal structure indicates that both N and C termini for all subunits in the E. coli complex are on the outer surface, away from both the tetramer interfaces and the active site (Figure 4.3). Given the structural similarities between family-I transferases, even those from vastly different lineages (Coros et al. 2004), our thermophilic candidates seemed likely to share this general structure. Therefore, N-terminal histidine-tagged alpha and beta subunits of both Ctf complexes were expressed in E. coli for purification. An initial lysis in standard IMAC buffer (lacking glycerol and sulfate) resulted in cell extracts with a strong band for both Tmel and Csub alpha subunits on SDS-PAGE, but bands for the beta subunits were only visible in the insoluble cell fraction. Subsequent lysis in IMAC A buffer reformulated for Ctf purification was able to recover both subunits in soluble form, but yields of the beta subunits were low, since most was still lost to the insoluble fraction (particularly for

Csub). Purification using IMAC buffers containing glycerol and sulfate resulted in highly purified alpha subunits, as well as purified Tmel beta, but the Csub beta subunit remained only partially purified (78%), as determined by SDS-PAGE densitometry (Figure 4.4).

In contrast to previous efforts, individually purified Cft alpha and beta subunits from separate expressions were obtained for biochemical characterization; the E. coli alpha subunit had been previously purified alone for structural but not biochemical analysis (Korolev et al.

2002). The results shed light on the difficulty of recovering active Ctf in previous studies. It

130

appears that the cause of Ctf’s low stability is specifically attributable to the beta subunit, since alpha subunits from each thermophile purified easily even without ammonium sulfate and glycerol as stabilizers. It is also possible that the beta subunit could only be recovered alone in this case due to the greater stability inherent in thermophilic proteins. Blue Native PAGE analysis confirms that the Tmel Ctf complex exists as the expected heterotetramer, although bands for the hetero-dimer and hetero-octamer are also visible. The Ctf subunits were also loaded individually at equal levels, but both beta subunits are only faintly visible as broad smears

–likely reflecting denaturation or aggregation in the gel, and further evidence of their low stability. No oligomers are visible for the Csub complex, although this could be a result of the low purity of the beta subunit.

The acetoacetate decarboxylase candidate from Sulfolobus sp. was expressed in both His- tagged and untagged form, neither of which showed any acetoacetate decarboxylase activity. The

Vulcanisaeta distributa enzyme was expressed in E. coli but could not be solubilized under any conditions tested. In contrast, the Adc of C. acetobutylicum was functionally expressed in both his-tagged and untagged forms, but the tagged enzyme exhibited significantly reduced activity.

The active enzyme complex is a homododecamer (Ho et al., 2009), and it is likely that the tag interferes with subunit assembly. Therefore, untagged Adc was expressed and the cell extract was heat-treated, resulting in a surprisingly pure protein (84% according to densitometry), which

Native PAGE confirmed formed the expected 12-subunit complex (Figure 4.4).

Characterization of acetone pathway enzymes

The C. subterraneus Thl was characterized in detail previously (Loder et al. 2015). The functional enzyme complex is a homo-tetramer displaying remarkable thermostability, with a half-life at 70oC of over 200 min. The enzyme also functions well in the desired direction

131

(formation of acetoacetyl-CoA), with vmax = 74 U/mg and a strong affinity for the acetyl-CoA substrate KM = 271 µM. This is desirable because thiolases catalyzing the reverse reaction

(cleavage of acetoacetyl-CoA to two acetyl-CoAs) are prevalent in the carbon fixation cycle of the extremely thermoacidophilic Sulfolobales (Berg et al. 2007).

Purification of the individual alpha and beta subunits of Ctfs from Tmel and Csub allowed us to assay the individual subunits for activity: neither subunit from either species was capable of catalyzing the CoA transfer reaction alone. While Ctf complexes from E. coli and C. acetobutylicum have been purified and characterized for kinetics and substrate preferences, subunits from Ctf enzymes have not been purified separately for analysis. There is one instance of the co-purified E. coli subunits being separated by subsequent urea denaturation, which claimed the pure beta subunit had ~2% the activity of the intact complex (Frerman and

Duncombe 1979). However, as the experimenters started with the CtfAB complex, the possibility of a small amount of residual contamination in the individual subunit fractions cannot be ruled out. The authors describe the alpha subunit as involved in “structural support or maturation”, which agrees with the findings here that it dramatically increases the stability of the beta subunit, but seem to have ruled out the possibility of alpha playing a catalytic role simply because it does not contain the nucleophilic glutamate involved in catalysis. More recent data suggests the alpha subunit plays an important role in binding the CoA group (Korolev et al.

2002), in addition to the structural support it provides the beta subunit. The presence of highly conserved residues neighboring the active site in both subunits also confirms that both are important for catalysis (Figure 4.3)

While we observed no activity from either the alpha or beta subunits alone, activity was observed using hybrid assemblies (Csub CtfA with Tmel CtfB and vice-versa), but required

132

substantial over-loading of both subunits. Interestingly, the individual subunits also displayed dramatically different thermostabilities; incubation of the beta subunit of either species at 70oC for a few minutes was enough to eliminate activity, while the alpha subunit showed no reduction in activity after an hour (Figure 4.5). The combined subunits of Tmel exhibited a stability intermediate between the two pure subunits, with a half-life at 70oC of 95 minutes. Thermal degradation of Tmel CtfAB complex followed a two-step inactivation, where activity was rapidly reduced to approximately 60% in the first 15 minutes of high temperature incubation, but then declined much more slowly, so that nearly 20% of activity was present even after 12 hours

(Figure 4.5). The low purity of Csub CtfB complicated activity assays, since residual E. coli proteins quickly denatured at assay temperature and contributed to substantial background absorbance. Low yield also meant there was not enough Csub enzyme for all the assays required for a full characterization. However, based on limited data the Csub enzyme complex had thermostability comparable to Tmel, with residual activity evident even after 12 hours at 70 oC.

Family-I CoA transferases are known to exhibit Ping-Pong enzyme kinetics (also occasionally called “double displacement” or “substituted-enzyme” kinetics). In this case, one substrate binds the enzyme, is modified, and then dissociates as the first product, leaving the enzyme in a modified intermediate state, followed by binding of the second substrate, which is modified and dissociates as product 2, recovering the original enzyme. In the case of Ctf, the net reaction is:

Acetoacetyl-CoA + Acetate → Acetoacetate + Acetyl-CoA which can be broken up into the two components:

Acetoacetyl-CoA + Enzyme → Enzyme-CoA + Acetoacetate

Acetate + Enzyme-CoA → Enzyme + Acetyl-CoA

133

The two substrates and sequential nature of the reaction leads to unusual kinetics, where substrate inhibition appears at relatively low concentrations, but can be overcome by increasing the concentration of the other substrate (Wiesenborn et al. 1989). Assays of Tmel Ctf exhibit this unusual substituted-enzyme substrate inhibition, as evidenced by the Hanes plot

([Substrate]/velocity vs. [Substrate]) with the characteristic features: a series of assays with one substrate held constant form straight lines without a common intersection point, but individual lines intersect to the right of the y-axis (Figure 4.6) (Cornish-Bowden 1995). This result confirms that catalysis in the thermophilic Tmel Ctf proceeds through the same ping-pong mechanism observed in mesophilic versions of the enzyme.

While the complexity of Ping-Pong kinetics makes it difficult to report typical enzyme kinetic parameters, the specific activities observed for the purified Tmel Ctf complex were similar to those reported for the E. coli enzyme. With substrate concentrations of 25mM acetate and 0.2mM acetoacetyl-CoA we observed 80 µmol/min/mg protein (or U/mg), while under comparable conditions Ecoli Ctf has specific activity around 150 U/mg (Sramek and Frerman

1975a). The value for purified C. acetobutylicum Ctf under similar assay conditions was 29.1

U/mg, compared to 0.36 U/mg in raw cell extracts (Wiesenborn et al. 1989). A specific activity of 3.57 U/mg was reported in cell extracts of recombinant T. saccharolyticum over-expressing

Tmel CtfAB, although in this case the thermophilic enzyme was being assayed at room temperature, and using a very different method (Shaw et al. 2015).

C. acetobutylicum Adc has been thoroughly characterized. The exact nature of its active site and catalytic mechanism have been of interest to biochemists (Highbarger et al. 1996; Ho et al. 2009), and historically its surprising thermostability was a topic of study (Autor and Fridovich

1970; Neece and Fridovich 1967). Our results confirm that incubation at 70oC leads to an

134

increase in activity (Figure 4.7). Most early kinetic analysis made use of an arbitrary activity scale based on absorbance units, but two more recent reports give kcat = 165/sec (Ho et al. 2009) and 1560/sec (Highbarger et al. 1996) with acetoacetate. Given that these values differ by roughly a factor of 10, and calculation of kcat depends on molar concentration of enzyme, it seems likely that the larger value was calculated using the molecular weight of the dodecameric holozyme (330 kDa), while the smaller used the subunit weight of 27.5 kDa. If this is the case, converting the kcat values to specific activity gives 360 and 280 U/mg, respectively. Our Cace

Adc exhibited activities in excess of 1000 U/mg, which is probably attributable to heat-activation from incubation at 70 oC, since neither reference above mentions heat treatment.

Full in vitro pathway

Mixtures of all three enzymes were capable of converting substrates acetyl-CoA and acetic acid to acetone at 70oC, while omitting any one enzyme eliminated acetone production

(Figure 4.8). Samples are run alongside a 2.5 mM acetone standard in assay buffer, which seems to distort the acetone peak, since it displays a broad right shoulder in both the standard and in vitro reaction. The relative size of the peaks suggests incomplete conversion, since stoichiometrically 5mM acetyl-CoA would result in 2.5 mM acetone. The peak at 4.1 minutes is likely acetoacetic acid, since it is most prominent in the no-Adc control, but a smaller peak is also visible as an intermediate in the full reaction.

4.5. Conclusions

Acetone is a promising candidate for production and bio-reactive distillation of a bio- based chemical produced in a thermophile. The pathway requires three enzymes with good stability up to at least 70oC (Figure 4.9). The previously characterized C. subterraneous thiolase

135

meets this requirement, as does the surprisingly thermostable C. acetobutylicum acetoacetate decarboxylase, which is fortunate since thermophilic homologs showed no activity. The ease with which untagged Cace Adc was separated from contaminating E. coli proteins and brought to greater than 80% purity by simple heat treatment serves as a reminder of why recombinant thermophilic proteins are so popular for research and crystallography. This also suggests that other unusually thermostable proteins from mesophiles could be identified simply by heat- treating mesophilic cell extracts to remove the most labile proteins, then screening the supernatant for residual activity, or running the soluble fraction on a protein gel.

The missing link for acetone production was an extremely thermophilic acetate:acetoacetyl-CoA CoA transferase, two of which have been reported here. Either T. melanesiensis or C. subterraneus could serve as the basis for a thermophilic acetone production, since both complexes display good activity and thermostability. In addition to reporting the first purified thermostable Ctfs, this work sheds light on the properties of CoA transferases in general.

The function of the hybrid Tmel/Csub Ctf complexes is likely due to the strong structural and sequence similarities among CoA transferases, particularly around the active site, and warrants further investigation to see if this cross-functionality is evident among homologs from other species as well. Meanwhile, the dramatic difference in thermostability between the Ctf alpha and beta subunits, and poor stability of the beta subunits in general, helps to explain why this class of enzymes has been so challenging to purify in the past. The dramatically improved stability of the full Ctf complex compared to the beta subunit highlights the role that the alpha subunit plays in stabilizing its partner, and serves as example of how important quaternary structure interactions can be for protein stability (the surprisingly thermostable dodecameric Cace Adc is another good example).

136

The relative thermal stabilities of the enzymes in the synthetic pathway should be taken into account when designing cloning constructs for recombinant expression in thermophilic hosts. While the discovery that Tmel CoA transferase has a half-life of over 1 hour at 70oC means that the proposed extremely thermophilic acetone pathway can function, it remains the least thermostable of the three enzymes, inactivating more rapidly than the Csub Thl.

Unexpectedly, the mesophilic Cace Adc is the most stable, and the dramatic increase in activity at the beginning of high temperature incubation meant that even after 12 hours activity was still greater than prior to heat-treatment. Given that Adc also appears to be the most active enzyme, it could be expressed at the end of an operon, or separately under the control of a weaker promoter.

Thl and Ctf have comparable stabilities and activities, so roughly equivalent expression would be appropriate. Given the instability of Ctf beta, and the fact that all Ctf genes appear to exist in a ctfAB operon (often with overlapping start/stop codons), joint expression of the two subunits seems to be essential, especially at high temperatures.

As shown in Figure 4.9, only the reaction catalyzed by Adc is strongly favored thermodynamically. The equilibrium of the thiolase reaction favors acetyl-CoA, meaning that intracellular acetoacetyl-CoA concentrations will be low. One way to drive the reaction forward at the point of Ctf is to increase the concentration of acetate, which is a known factor in driving the switch to solventogenesis in C. acetobutylicum (Wiesenborn et al. 1989). Using a host that is a good natural acetate producer and can handle relatively high concentrations of acetate would facilitate acetone production. Fortunately, there are a number of fermentative extreme thermophiles that produce acetate as a major fermentation product and for which genetic tools are available (Loder et al. 2017).

137

4.6. Acknowledgements

This work was supported with grants to RMK from the US National Science Foundation

(CBET 1264052) and US Dept. of Agriculture (2018-67021-27716). BMZ acknowledges support from an NIH Biotechnology Traineeship (2T32GM008776). CTS acknowledges support from a

US Department of Education (P200A140020) GAANN Fellowship

138

Table 4.1: Highest temperatures for solvent production natively and in metabolically engineered hosts Native Engineered Acetone 43°C (Weimer 1984) 55°C(Shaw et al. 2015)

Bacillus macerans Thermoanaerobacterium saccharolyticum Ethanol 72°C (Svetlitchnyi et al. 2013) 75°C (Basen et al. 2014)

Caldicellulosiruptor sp. Pyrococcus furiosus Butanol 58°C (Freier-Schroder et al. 1989) 60°C (Keller et al. 2015)

Clostridium thermosaccharolyticum Pyrococcus furiosus

Table 4.2: SCooP melting temperatures of Adc structures Predicted melting temperatures for all Adc structures in the PDB.

Organism Growth Topt Adc PDB ID SCooP Tm C. acetobutyicum 35°C 3bh2 63.5°C Chromobacterium violaceum 32°C 3bgt 53.1°C Legionella pneumophila 36°C 3c8w 66.1°C Methanoculleus marisnigri 25°C 3cmb 55.8°C

Table 4.3: Enzyme candidates for thermophilic acetone pathway Source Organism Topt Accession % amino acid Evidence number identity Thiolase (Thl) – C. acetobutylicum: AAK80816.1 C. subterraneus 70°C WP_011024972 68% (99% Characterized, 70°C subsp. coverage) (Loder et al. 2015) tengcongensis Acetate acetoacetyl-CoA transferase (CtfAB) – C. acetobutylicum: NP_149326 & 27 C. subterraneus 70°C WP_009610465 65% (97% cov) Homology (Shaw et subsp. al. 2015) tengcongensis WP_011025123 65% (100% cov) T. melanesiensis 70°C WP_012057350 53% (96% cov) Activity, 55°C (Shaw et al. 2015) WP_012057349 64% (94% cov) Acetoacetate decarboxylase (Adc) – C. acetobutylicum: NP_149328 C. acetobutylicum 35°C NP_149328 100 (100% cov) Activity, 70°C (Davies 1943) V. distributa 85°C WP_013335399 29% (92% cov) Homology Sulfolobus contig 75°C WP_009989587 24% (98% cov) Homology

139

Figure 4.1: Vapor-liquid equilibria for acetone and ethanol The gray box highlights the estimated range for biologically feasible product titers, indicating that bio-reactive distillation with acetone as a product could be accomplished at temperatures 15°C lower than with ethanol. (data from VLE-calc.com)

140

Figure 4.2: Sequence alignments of acetone enzymes Amino acid alignments (of proteins listed in Table 4.3), dark gray highlights = identical, light gray = similar. The lower homology among the Adc enzymes compared to the others is immediately apparent. Alignments performed in Geneious 8.1.9 (http://www.geneious.com).

141

Figure 4.3: Structure of E. coli Ctf and multispecies alignment of active site sequences The crystal structure of E. coli Ctf (PDB ID: 5DBN) heterotetramer, with each peptide’s N and C termini highlighted red and blue, and active site residues (Prosite entries PS01273 and PS01274) in green. Sequence alignments of the active site residues of five Family I CoA-transferases are shown. Alignments in Geneious

142

Figure 4.4: SDS-PAGE and Blue-native PAGE of acetone enzymes SDS-PAGE densitometry indicates that all enzymes have the expected Mr and are at least 80% pure (except for Csub CtfB). Blue Native-PAGE also confirms that most enzymes exhibit the expected multimolecular arrangement, with the exception of Csub Ctf (possibly a result of the low purity of the beta subunit). Neither Ctf-beta subunit is visible on native PAGE, even though all lanes are loaded to equal mass. The smear in both lanes suggests the subunits may have denatured.

Figure 4.5: Thermal inactivation of Tmel Ctf Tmel Ctf complex exhibits the common two-step thermal inactivation, as evidenced by the weak fit to a first-order inactivation model. The individual subunits also have vastly different thermal stabilities.

143

Constant Acetoacetyl-CoA 0.4 mM 0.2 mM 0.1 mM 0.05 mM 0.03 mM 180

130

80 [acetate]/v 30

-100-20 100 300 500 700

[acetate]

Figure 4.6: Hanes plot of Tmel Ctf kinetics Tmel Ctf exhibits typical ping-pong kinetics substrate inhibition, visible on a Hanes plot as lines intersecting to the right of the Y-axis.

Figure 4.7: Thermal stability of Cace Adc The first hour of high-temperature incubation leads to an increase in Adc activity, after which activity slowly declines. Even after 12 hours the activity remains greater than the unheated reference.

144

Figure 4.8: In vitro function of full acetone pathway In vitro function of the three enzymes (Rxn) is confirmed by production of acetone. Omitting any one enzyme is enough to prevent acetone production, although a peak consistent with the acetoacetate intermediate is visible in the no-Adc control.

Figure 4.9: Three enzyme pathway to acetone production The enzymes catalyzing the three steps of the acetone pathway (along with equilibrium constants for the reactions in the direction shown). High acetic acid concentrations drive the reaction forward by causing the Ctf enzyme to favor acetoacetate production. The pathway can be used with a native acetic acid producer, where it simultaneously detoxifies acetic acid while generating acetone. Equilibrium constants from eQuilibrator (Flamholz et al. 2012).

145

4.7. References

Autor AP, Fridovich I. 1970. The thermal inactivation of acetoacetate decarboxylase. J Biol Chem 245(20):5223-5227. Basen M, Schut GJ, Nguyen DM, Lipscomb GL, Benn RA, Prybol CJ, Vaccaro BJ, Poole FL, Kelly RM, Adams MWW. 2014. Single gene insertion drives bioalcohol production by a thermophilic archaeon. Proc Natl Acad Sci U S A 111(49):17618-17623. Berg IA, Kockelkorn D, Buckel W, Fuchs G. 2007. A 3-hydroxypropionate/4-hydroxybutyrate autotrophic carbon dioxide assimilation pathway in archaea. Science 318(5857):1782- 1786. Bermejo LL, Welker NE, Papoutsakis ET. 1998. Expression of Clostridium acetobutylicum ATCC 824 genes in Escherichia coli for acetone production and acetate detoxification. Appl Environ Microbiol 64(3):1079-1085. Cary JW, Petersen DJ, Papoutsakis ET, Bennett GN. 1990. Cloning and expression of Clostridium acetobutylicum ATCC 824 acetoacetyl-Coenzyme A : acetate / butyrate : Coenzyme A- transferase in Escherichia coli. Appl Environ Microbiol 56(6):1576-1583. Cornish-Bowden A. 1995. Fundamentals of Enzyme Kinetics. London: Portland Press. 343 p. Coros AM, Swenson L, Wolodko WT, Fraser ME. 2004. Structure of the CoA transferase from pig heart to 1.7 Å resolution. Acta Cryst. D 60(10):1717-1725. Davies R. 1943. Studies on the acetone-butanol fermentation. Biochem J 37(2):230-238. Flamholz A, Noor E, Bar-Even A, Milo R. 2012. EQuilibrator - The biochemical thermodynamics calculator. Nucleic Acids Res 40(D1):770-775. Freier-Schroder D, Wiegel J, Gottschalk G. 1989. Butanol formation by Clostridium thermosaccharolyticum at neutral pH. Biotechnol Lett 11(11):831-836. Frerman FE, Duncombe GR. 1979. Studies on the subunits of Escherichia coli coenzyme a transferase: Reconstitution of an active enzyme. Biochimica et Biophysica Acta 580(2):289-297. Heider J. 2001. A New Family of CoA-transferases. FEBS Letters 509:345-349. Highbarger LA, Gerlt JA, Kenyon GL. 1996. Mechanism of the reaction catalyzed by acetoacetate decarboxylase. Importance of lysine 116 in determining the pKa of active- site lysine 115. Biochemistry 35(1):41-46. Ho MC, Ménétret JF, Tsuruta H, Allen KN. 2009. The origin of the electrostatic perturbation in acetoacetate decarboxylase. Nature 459(7245):393-397. Hoseki J, Yano T, Koyama Y, Kuramitsu S, Kagamiyama H. 1999. Directed evolution of thermostable kanamycin-resistance gene: a convenient selection marker for Thermus thermophilus. J Biochem 126(5):951-6. Jang Y-S, Lee JY, Lee J, Park JH, Im JA, Eom M-H, Lee J, Lee S-H, Song H, Cho J-H and others. 2012. Enhanced Butanol Production Obtained by Reinforcing the Direct Butanol- Forming Route in Clostridium acetobutylicum. mBio 3(5):1-9. Jones MAAH, Papoutsakis ET. 2014. σK of Clostridium acetobutylicum is the first known sporulation-specific sigma factor with two developmentally separated roles, one early and one late in sporulation. J Bacteriol 196(2):287-299. Keller MW, Lipscomb GL, Loder AJ, Schut GJ, Kelly RM, Adams MW. 2015. A hybrid synthetic pathway for butanol production by a hyperthermophilic microbe. Metab Eng 27:101-106. Khan S, Vihinen M. 2010. Performance of protein stability predictors. Human Mutation

146

31(6):675-684. Korolev S, Koroleva O, Petterson K, Gu M, Collart F, Dementieva I, Joachimiak A. 2002. Autotracing of Escherichia coli acetate CoA-transferase α-subunit structure using 3.4 Å MAD and 1.9 Å native data. Acta Cryst. D 58(12):2116-2121. Kumar S, Tsai C-J, Nussinov R. 2000. Factors enhancing protein thermostability. Protein Eng. Des. Sel. 13(3):179-191. Liu L, Deng Z, Yang H, Li J, Shin HD, Chen RR, Du G, Chen J. 2014. In silico rational design and systems engineering of disulfide bridges in the catalytic domain of an alkaline α- amylase from Alkalimonas amylolytica to improve thermostability. Appl Environ Microbiol 80(3):798-807. Loder AJ, Zeldes BM, Conway JM, Counts JA, Christopher T. 2017. Extreme Thermophiles as Metabolic Engineering Platforms: Strategies and Current Perspective. In: Wittmann C, Liao JC, editors: Wiley-VCH. p 507-580. Loder AJ, Zeldes BM, Garrison GD, Lipscomb GL, Adams MWW, Kelly RM. 2015. Alcohol selectivity in a synthetic thermophilic n-butanol pathway is driven by biocatalytic and thermostability characteristics of constituent enzymes. Appl Environ Microbiol 81(20):7187-7200 Luo H, Ge L, Zhang J, Ding J, Chen R, Shi Z. 2016. Enhancing acetone biosynthesis and acetone–butanol–ethanol fermentation performance by co-culturing Clostridium acetobutylicum/Saccharomyces cerevisiae integrated with exogenous acetate addition. Bioresour Technol 200:111-120. Neece MS, Fridovich I. 1967. Acetoacetic decarboxylase: Activation by heat. J Biol Chem 242(12):2939-2944. Peabody GL, Kao KC. 2016. Recent progress in biobutanol tolerance in microbial systems with an emphasis on Clostridium. FEMS Microbiol Lett 363(5):1-6. Pucci F, Kwasigroch JM, Rooman M. 2017. SCooP: an accurate and fast predictor of protein stability curves as a function of temperature. Bioinformatics 33(May):3415-3422. Saito N. 1973. A thermophilic extracellular α-amylase from Bacillus licheniformis. Arch Biochem Biophys 155(2):290-298. Shaw AJ, Miller BB, Rogers SR, Kenealy WR, Meola A, Bhandiwad A, Sillers WR, Shikhare I, Hogsett Da, Herring CD. 2015. Anaerobic detoxification of acetic acid in a thermophilic ethanologen. Biotechnol Biofuels 8(1):1-12. Sramek SJ, Frerman FE. 1975a. Escherichia coli coenzyme A-transferase: Kinetics, catalytic pathway and structure. Arch Biochem Biophys 171(1):27-35. Sramek SJ, Frerman FE. 1975b. Purification and properties of Escherichia coli coenzyme A- transferase. Arch Biochem and Biophys 171(1):14-26. Studier FW. 2005. Protein production by auto-induction in high density shaking cultures. Protein Expr Purif 41(1):207-234. Svetlitchnyi V, Kensch O, Falkenhan D, Korseska S, Lippert N, Prinz M, Sassi J, Schickor A, Curvers S. 2013. Single-step ethanol production from lignocellulose using novel extremely thermophilic bacteria. Biotechnol Biofuel 6(1):31. Weimer PJ. 1984. Control of Product Formation During Glucose Fermentation by Bacillus macerans. Microbiology 130(1):103-111. Wiesenborn DP, Rudolph FB, Papoutsakis ET. 1989. Coenzyme A Transferase from Clostridium acetobutylicum ATCC 824 and Its Role in the Uptake of Acids. Appl Environ Microbiol 55(2):323-329.

147

Wu M, Wang J, Liu J, Huo H. 2007. Life-Cycle Assessment of Corn-Based Butanol as a Potential Transportation Fuel. Report nr ANL/ESD/07-10. 59 pp Zhou XX, Wang YB, Pan YJ, Li WF. 2008. Differences in amino acids composition and coupling patterns between mesophilic and thermophilic proteins. Amino Acids 34(1):25- 33.

148