1 PRODUCTION OF 2,3 BUTANEDIOL IN E SCHERICHIA COLI AND B ACILLUS SUBTILIS By RAFAEL RODRIGUES DE OLIVEIRA A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2014
2 Â© 2014 Rafael Rodrigues de Oliveira
3 To my Mom, Dad, Bruno, Matheus and JÃ³rgia
4 ACKNOWLEDGMENTS I would like to acknowledge my advisor, Dr. Wayne Nicholson, for his mentorship, guidance and friendship. Thank you for cons tructing the scientist that I am now. My understanding of all things science involves jumped orders of magnitude as a result of our discussions. H opefully I will be able to expand and propagate the priceless knowledge I received. I would also like to thank Dr. Patricia Faj ardo Cavazos for sharing cultural experiences that were crucial to my adaptation process of living abroad. I am grateful to my committee members , Drs. Eric Triplet t , Val Ã© rie de Cr Ã© cy Lagard , K. T. Sha n mugam, and Shouguang Jin for providing valuable suggestions and support. I would like to thank S amantha Wat ers, Jennifer Mobberley, Artemis Louyakis, Rebecca L. Mickol , Alexandrea Duscher, Kry stal Kerney , BrÃ© lan Moritz, Deepika Awasthi , Jonathan Martin, Vikarma Brooks, Richard Rosario Passaper ia, Kateryna Zhalnina, Yezhang Ding, Jennie Fagen, Austin Davis Richardson, Fernando Pagliai, Dr s . Adriana Giongo, Ralf MÃ¶ller, Andrew Schuerger , Jamie Foster , Graciela Lorca, C l a udio Gonzalez, James Preston, Tony Romeo, Mun Su Rhee, Christopher Vakuls kas and Julie Maupin Furlow for scientific discussions and friendship. I would also like to thank Janet Gilbert, Christine Gough and Javier Real for excellent bureaucratic and technical assistance. None of this would have been possible without my parents, DÃ©ci o R. de Oliveira and Neusa R. de Oliveira, who have provided constant love and support. I would like to specially thank my wife JÃ³rgia C arbonera de Oliveira for experiencing with me every single mo ment of this incredible journey .
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ ............... 4 LIST OF TABLES ................................ ................................ ................................ ........................... 7 LIST OF FIGURES ................................ ................................ ................................ ......................... 8 LIST OF ABBREVIATIONS ................................ ................................ ................................ .......... 9 ABSTRACT ................................ ................................ ................................ ................................ ... 11 CHAPTE R 1 LITERATURE REVIEW ................................ ................................ ................................ ....... 13 Petroleum dependence ................................ ................................ ................................ ............ 13 2,3 BD properties and application s ................................ ................................ ........................ 15 Organisms producing 2,3 BD ................................ ................................ ................................ . 16 Production of 2,3 BD by Bacillus subtilis WN1038 ................................ .............................. 17 Escherichia coli production of 2,3 BD ................................ ................................ ................... 20 2 THE LYSR TYPE T RANSCRIPTIONAL REGULATOR (LTTR) ALSR INDIRECTLY REGULATES EXPRESSION OF THE BACILLUS SUBTILIS 2,3 BUTANEDIOL (2,3 BD) DEHYDROGENASE GENE BDHA AND 2,3 BD PRODUCTION ................................ ................................ ................................ ....................... 25 Introduction ................................ ................................ ................................ ............................. 25 Materials and Methods ................................ ................................ ................................ ........... 26 Bacterial strains, plasmids, and growth conditions ................................ ......................... 26 Recombinant DNA techniques and plasmid constructions ................................ ............. 27 AlsR protein production and purification ................................ ................................ ........ 27 Enzymatic assays ................................ ................................ ................................ ............. 28 RNA isolation ................................ ................................ ................................ .................. 29 Quantitative reverse transcriptase PCR (qRT PCR) ................................ ....................... 29 Prime r extension mapping ................................ ................................ ............................... 29 Electrophoretic mobility shift assays (EMSAs) ................................ .............................. 30 Statistical analyses ................................ ................................ ................................ ........... 31 Results ................................ ................................ ................................ ................................ ..... 31 Expression of the bdhA lacZ fusion in wild ty pe and alsR::spc strains ......................... 31 Measurement of bdhA transcript levels by qRT PCR ................................ ..................... 32 BDH activity in w.t. vs. alsR::spc strains ................................ ................................ ....... 33 Mapping the bdhA transcription initiation site ................................ ................................ 34 Purifie d AlsR protein does not bind the bdhA promoter region ................................ ...... 34 Discussion ................................ ................................ ................................ ............................... 35
6 3 SYNTHETIC OPERON FOR 2,3 BUTANEDIOL PRODUCTION IN BACILLUS SUBTILIS AND ESCHERICHIA COLI ................................ ................................ .................. 46 Introduction ................................ ................................ ................................ ............................. 46 Materials and Methods ................................ ................................ ................................ ........... 48 Bacterial strains, plasmids, and growth conditions ................................ ......................... 48 Recombinant DNA techniques ................................ ................................ ........................ 48 Plasmid construction ................................ ................................ ................................ ....... 49 Gel electrophoresis ................................ ................................ ................................ .......... 50 Assays ................................ ................................ ................................ .............................. 50 Assay of fermentation products by HPLC ................................ ................................ ....... 50 Statistical analyses ................................ ................................ ................................ ........... 50 Results ................................ ................................ ................................ ................................ ..... 51 Construction of a synthetic alsSDbdhA operon ................................ ............................... 51 Production of 2,3 BD by B. subtilis background strains ................................ ................. 52 Production of 2,3 BD in B. subtilis strains carrying the synthetic alsSDbdhA operon ... 52 Production of 2,3 BD in E. coli ................................ ................................ ....................... 53 Discussion ................................ ................................ ................................ ............................... 54 4 REDOX BALANCED PRODUCTION OF ISOMERICALLY P URE (R,R) 2,3 BUTANEDIOL IN ESCHERICHIA COLI ................................ ................................ ............. 61 Introduction ................................ ................................ ................................ ............................. 61 Materials and Methods ................................ ................................ ................................ ........... 62 Bacterial strains, plasmids, and growth conditions ................................ ......................... 62 Rec ombinant DNA techniques and plasmid constructions ................................ ............. 63 Introduction of T7 RNA polymerase gene into E. coli W1330 and YK29 background s ................................ ................................ ................................ ................. 64 Growth tolerance on diacetyl ................................ ................................ ........................... 65 Detection and quantification of fermentation products by HPLC ................................ ... 65 Statistical analyses ................................ ................................ ................................ ........... 65 Results ................................ ................................ ................................ ................................ ..... 65 Construction of alsS alsD bdhA and alsS bdhA synthetic operons ................................ 65 Redox balanced 2,3 BD fermentation by WN1538 ( alsS bdhA ) ................................ .... 66 Redox unbalanced 2,3 BD fermentation by strain WN1539 ( alsS alsD bdhA ) ............. 67 Fermentation products of WN1534, WN1535, WN1538 and WN1539 ......................... 67 Diacetyl effect on growth ................................ ................................ ................................ 68 Discussion ................................ ................................ ................................ ............................... 68 5 SUMMARY OF RESEARCH ................................ ................................ ................................ 75 LIST OF REFERENCES ................................ ................................ ................................ ............... 77 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ......... 86
7 LIST OF TABLES Table page 2 1 Strains and plasmids used in this study ................................ ................................ .............. 38 2 2 Oligonucleotide primers used in this study ................................ ................................ ........ 39 3 1 Bacterial strains and plasmids used in this study ................................ ............................... 56 4 1 Bacterial strains and plasmids used in this study ................................ ............................... 70 4 2 Fermentation products, glucose consumption and cell growth after 120 h batch fermentation ................................ ................................ ................................ ....................... 7 1
8 LIST OF FIGURES Figure page 1 1 2,3 Butanediol stereoisomers ................................ ................................ ............................. 24 2 1 The 2,3 BD biosynthetic pathway in B. subtilis ................................ ................................ 40 2 2 Beta galactosidase activities expressed from the bdh lacZ transcriptional fusion ............ 41 2 3 Levels of bdhA m RNA measured by qRT PCR ................................ ................................ 42 2 4 BDH activity in 24 h and time course ................................ ................................ ................ 43 2 5 Results of primer extension experiment (inset) in the context of the nucleotide sequence of the bdhA upstream regulatory region ................................ ............................ 44 2 6 Electrophoretic mobility shift assay of purified AlsR protein with the promoter region of bdhA ................................ ................................ ................................ .................... 45 3 1 Construction of synthetic alsSDbdhA operon ................................ ................................ .... 57 3 2 Glucose consumption and production of 2,3 BD and acetoin by B. subtilis background strains ................................ ................................ ................................ ............. 58 3 3 Glucose consumption and production of 2,3 BD and acetoin by B. subtilis strains carrying plasmid pWN1390 ................................ ................................ ............................... 59 3 4 Fed batch flask fermentation profile of E. coli strain WN1390 carrying the synthetic alsSDbdhA operon ................................ ................................ ................................ ............. 60 4 1 2,3 BD synthetic operons and corresponding pathways ................................ .................... 72 4 2 Fermentation products, growth and glucose consumption of uninduced and induced strains WN1538 and WN1539 th roughout 120 h of batch culture ................................ .... 73 4 3 Diacetyl effect on of E. coli W3110 ................................ ................................ .................. 74
9 LIST OF ABBREVIATIONS :: I nsertion mutation 2,3 BD 2,3 butanediol ATP Adenosine triphosphate alsD G ene encoding acetolactate decarboxylase ALDC acetolactate decarboxylase alsR G ene encoding the AlsR transcriptional activator of alsSD AlsR T ranscriptional activator of alsSD alsS G ene encoding acetolactate synthase ALS acetolactate synthase Ap A mpicillin AR Acetoin reductase bdhA G ene encoding butanediol dehydrogenase BDH B utanediol dehydrogenase Da Dalton Em Erythromycin F Forward primer focA G ene encoding b idirectional formate transporter g G ram ( s ) HEPES 4 (2 hydroxyethyl) 1 piperazineethanesulfonic acid h Hour IPTG Isopropyl thio D galactopyranoside J Joule kb kilobase Km Kanamycin
10 L Liter LDH Lactate Dehydrogenase ldh A G ene encoding L Lactate dehydrogenase mRNA Messenger RNA NAD + Nicotinamide adenine dinucleotide , oxidized form NADH Nicotinamide adenine dinucleotide , reduced form NTA Nitrilotriacetic acid OPEC Organization of the Petroleum Exporting Countries p Plasmid PCR Polymerase Chain Reaction PDH Pyruvate dehydrogenase pfl G ene encoding Pyruvate formate lyase PFL Pyruvate formate lyase P spac Hybrid promoter consisting of the 35 region from B. subtilis phage SPO1 and the 10 region from the E. coli lac promoter q RT PCR Q uantitative Reverse Transcriptase P olymerase C hain R eaction R Resistance R Reverse primer Spc Spectinomycin Trp + Tryptophan prototroph trpC2 M utation in gene encoding Indole 3 glycerol phosphate synthase resulting in tryptophan auxotrophy
11 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy PRODUCTION OF 2,3 BUTANEDIOL IN E SCHERICHIA COLI AND B ACILLUS SUBTILIS By Rafael Rodrigues de Oliveira August 2014 Chair: Wayne Nicholson Major: Microbiology and Cell Science The commodity chemical 2,3 butanediol (2,3 BD) has many applications, serving as an: industrial solvent, food additive, and building block for synthetic rubber. The Gram positive bacterium Bacillus subtilis produces 2,3 BD under fermentative conditions . In s ynthesis of 2,3 BD : 1 acetolactate by ALS , 2 acetolactate is decarboxylated to acetoin by ALDC , and 3 ) acetoin is reduced to 2,3 BD by BDH . The alsS and alsD genes form a bicistronic operon, and bdhA is a single unli nked gene. Upstream and divergent from alsSD is the alsR gene that encodes a LysR type transcrip tional regulator. AlsR effect on bdhA expression was assesse d in a wild type strain and a strain carrying an alsR::spc knock out mutation by measuring: (a) expr ession of a transcriptional bdhA lacZ fusion; (b) bdhA mRNA steady state levels by q RT PCR; (c) expression of BDH enzymatic activity ; and d) electrophoretic mo bility shift assay (EMSA) with AlsR and bdhA promoter s . Activation of bdhA expression was lowered, but not abolished, in the alsR::spc mutant. In order to overcome the native regulation of the 2,3 BD pathway and increase 2,3 BD production, a synthetic tricistronic operon was engineered containing the genes for the B. subtilis 2,3 BD pathway ( alsS alsD bdhA ) under control of the IPTG inducible P spac promoter on plasmid pDG148 Stu capable of expression in either B. subtilis or E. coli . When grown on Miller LB medium containing 2 %
12 glucose the plasmid construct showed ~ 3 fold increase in 2,3 BD production relative to the parental strain. The same construct was tested in E. coli, and even higher levels of 2,3 BD were obtained . As a means to bring the pathway into redox balance by adding an extra NAD + regeneration step, the synthetic operon alsS bd hA was constructed including the genes alsS and bdhA but missing the alsD acetolactate spontaneously decarboxylates into diacetyl, which can be reduced into acetoin and 2,3 BD by BDH, thus regenerating 2 NAD + . The alsS bdhA construct was transform ed into E. coli YK29 ( pfl ldh A ), and anaerobic fermentation of glucose produced 0.9 Â± 0.04 g/L 2,3 BD whereas the entire pathway ( alsS alsD bdhA ) produced 17.1Â± 0.88 g/L. We were thus able to restructure the 2,3 BD pathway to produce 2,3 BD in a redox ba lanced manner.
13 CHAPTER 1 LITERATURE REVIEW Petroleum dependence Oil exploitation has allowed an unprecedented growth of humanity, both demographic ally and technological ly (Youngquist, 1999) . However, i t has long been predicted that petroleum is a finite resource (Hubbert, 1956) . As of 2013 there were a total of ~ 1 .7 trillion barrels of prove n oil reserves (OPEC a nd non OPEC). In this very year world oil production was 31.6 billion barrels per day , indicating that at this production rate oil reserves would last only another 53.3 years (BP Statistical Review of World Energy, 2014 ) . In the United States a large major ity (71%) of consumed petroleum goes to combustion as transportation fuels. In addition, 99% of all chemical feedstocks are derived from natural gas and petroleum (McFarlane & Robinson, 2007) . Even though the peak oil debate is extremely biased due to encircling economical interests , and the fact that most of the relevant information are of private nature, some recent reviews have tried to estimate how long petroleum is still going to last. I n t rying to make this estimate , three major groups of factors that determine future prices and availability of oil have bee n proposed (Aguilera, 2014) . The f irst group includes geological factors such as the incidence and nature of the oil/gas resource. The second group takes in to c onsideration per capita income, which will dictate how fast a society utilizes oil and gas reserves. A nd the third group includes discovery of new technologies for extraction that can lower process costs . According to some researchers, the chance of a peak in global oil production around the year 2030 is very serious , and some researchers suggest that we are passing it right now (Sor rell et al. , 2010, Kerr, 2011) . Meanwhile, other sources advocate the peak oil production will not occur before 2035 (Chapman, 2013) .
14 Regardless of the divergences on the precise date for peak oil production, experts agree it will happen sooner or later (Chapman, 2013) . Scarcity will drive oil prices significantly higher an d urban societies will require major rearrangements (Brecha, 2013) . U nemployment and poverty caused by peak oil are likely to bring health consequences such as social and stress related illness es, resulting in decreasing life expectancy (Hanlon & McCartney, 2008) . P ositive aspects may arise concomitantly from oil shortage, such as redu ction of : pollution related respiratory diseases, obesity epidemics, or traffic accidents , all of which can be envisioned as aftermath s of a decrease d availability of motorized vehicles (Roberts, 2008) . It is hard to predict but easy to imagine how strongly affected our lives would be when facing oil exhaustion . The focus of the political discussion should be shifted from when to what can be done to l essen or postpone oil rece ssion . I n this sense, the most immediate measures to be taken are probably i n vestments in new technologies to provide alternatives to fossil fuels for production of energy and chemicals. As a substitute to oil, energy s ources like wind energy, wave power, solar energy, geothermal energy, tidal energy, nuclear power, microbial bio fuel cells, chemical fuel cells, and biomass thermal conversion are all in development by different research groups et al. , 2012, Lumbreras & Ramos, 2012, Rabaey et al. , 2004, Rourke et al. , 2010, Steubing et al. , 2011, Villa et al. , 2013) . These alternative energy sources aim to meet energy demands, but they provide no feedstock for chemical resource production such as polymer building blocks , industrial solvents, pharmaceuticals, fragrance carriers , etc . , which are all largely required by modern societies. Bacterial fermentation products (2,3 butanediol , methanol, e thanol, propanediols, carboxylic acids, furfural, ethyle ne, fatty alcohols , propylene oxide, propylene, acrolein, 3 hydroxypropionaldehyde, acrylic acid, etc . ) serve as chemical feedstock in many applications
15 (Marshall & Ala imo, 2010) . N o single bio based chemical alone will completely alleviate the current demand for fuels and chemicals obtained from oil ; however some molecules might have a more comprehensive role th a n others. In the present study we focused on understanding and improving production of 2,3 butanediol (2,3 BD) , a commodity chemical feedstock with extensive industrial applications. 2,3 BD properties and applications 2,3 BD is a secondary alcohol that can be obtained through fermentation of renewable carbon sources. There are three possible stereoisomers: L (+) (S,S) dextrorotatory 2,3 BD, meso optically inactive 2,3 BD and D ( ) (R,R) levorotatory 2,3 BD (Figure 1 1) (Prescott & Dunn, 1949) . 2,3 BD has a molecular weight of 90.12 poin t close to its high boiling point, the recovery of 2,3 BD through distillation is very expensive (Syu, 2001) . A recently developed procedure utilizing alcohol precipitation and vacuum distillation obtained promising results with lower recovery costs on the process (Jeon et al. , 2013) . The heating value of 2,3 BD (27,198 J g 1 ) is comparable wi th ethanol (29,055 J g 1 ) and methanol (22, 081 J g 1 ), and has be en recommended in the manufact ure of high grade aviation fuel (Qin et al. , 2010, Syu, 2001) . Many valuable molecules can be obtained from 2,3 BD. A simple deh ydrogenation reaction generates acetoin, which is used as food additive to enhance flavor , and in cosmetic products and soaps to obtain desired fragrances (Xiao & Lu, 2014) . A further dehydrogenation leads to diacetyl, used as butter flavor in popcorns and other products. A dehydratation of 2,3 BD generates m ethyl ethyl ketone , an important industrial so lvent. This dehydratation can be performed enzymatically by a diol dehydratase found in Lactobacilus brevis or through reaction with alumina or sulfuric acid (Ji et al. , 2011, Speranza et al. , 1996) . The catalytic dehydration of 2,3 BD, performed through pyrolysis of 2,3 BD diacetate or 2,3 BD dibromide, results in 1,3 -
16 butadiene , a build ing block for synthetic rubber (Winfield, 1945) . Using a de hydroge nase, the carbonyl group of methyl ethyl ketone can be reduced converting the molecule into the solvent and alternative biofuel 2 butanol (Jina et al. , 2011, Speranza et al. , 1997) . 2 Butanol can also be obtained non enzymatically by employing lithium aluminum hydride as a reductant (Leroux & Lucas, 1951) . Without any further transformation, 2,3 BD can be used as an antifreeze agent ; in fact, the D ( ) (R,R) levorotatory form exhibits a lower freezing point th a n the other two isomers (Boutron, 1990) . Organisms producing 2,3 BD The first time 2,3 BD was identified as a microbial fermentation product dates back to 1906 when a Klebsiella strain called at the time Bacillus lactis aerogenes (Escherich) produced large amounts of 2,3 BD during glucose ferment ation (Harden & Walpole, 1906) . Since then, numerous prokaryotes and even the yeast Saccharomyces cerevisiae have been used to prod uce 2,3 BD (Ji et al ., 2011, Kim et al. , 2013) . Currently , the bacterial species that have attained the best performance of 2,3 BD production are Enterobacter aerogenes, Klebsiella pneumoniae , Klebsiella oxytoca and Serratia marcescens , all of which are caus ative agents of human diseases (Bradford et al. , 1997, Hoffmann et al. , 2010, Mahlen, 2011, Sanders & Sanders, 1997) . The productivity [g/(L h)], yield ( g 2,3 BD /g glucose or sucrose for S. marcescens ) and concentration ( g/L ) of 2,3 BD obtained by these strains are as follows: K. pneumoniae rea ches 4.21 g/ (L h), 0.49 g/g and 150 g/L; K. oxytoca reaches 3.22 g/ (L h), 0.5 g/g and 130 g/L; E. aerogenes reaches 5.4 g/ (L h), 0.49 g/g and 110 g/L; S. marcescens reaches 3.4 9 g/(L h), 0.47 g/g and 152 g/L (Ji et al. , 2010, Ma et al. , 2009, Qin et al. , 2006, Qureshi & Cheryan, 1989, Zeng et al. , 1991, Zhang et al. , 2010a, Zhang et al. , 2010b) . The growing need for efficient and carbon neutral chemical production makes photosynthetic micro org anism attractive platforms for 2,3 BD production. Recently the 2,3 BD pathway was introduced into the cyanobacteri um
17 Synechococcus elongates , which used CO 2 as the carbon source to produce 2,3 BD in low amounts ( 2.39 g/L ) , and with low productivity [ / (L h ) ] (Oliver et al. , 2013) . P roduction of 2,3 BD by Bacillus subtilis WN1038 In the present work , the microbial platforms E. coli and B . subtilis strains were used to study the regulation of 2,3 BD pathway as well as to improve its production. B. subtilis is a Gram positive , spore f orming bacterium belonging to the p hylum Firmicutes, class Bacilli, order Bacillales and family Bacil l aceae. It has long being considered a soil organism, however more recently it has be en sug gested to be also part of the normal human gut microbiota (Hong et al. , 2009, Kunst et al. , 1997) . B. subtilis strain WN1038 is a prototroph derived from the auxotroph B. subtilis 168 ( trpC2 ) , by a spontaneous reversion of the trpC2 mutation resulting in a Trp + phenotype (Nicholson, 2008, Spizizen, 1958) . Some of the interesting aspects of using B. subtilis strain WN1038 to produce 2,3 BD are: a) its native capacity f or producing 2,3 BD (Nicholson, 2008, Renna et al. , 1993) ; b) its native ability for secretion of extrace llular polysaccharide degrading enzymes ( amylase, pullulanase, endo 1,4 mannanase, levanase, glucan 1,4 maltohydrolase, pectate 1,4 1,3 1,4 endoglucanase, and endo 1,4 xylanases ) helpful on lowering the costs of utilizing cheap and renewable carbon source s like plant biomass , which is recalcitrant to breakdown by many microorganisms (Antelmann et al. , 2001, Rastogi et al. , 2010) ; c) its feasibility for industrial application due to its Generally Recognized As S afe (GRAS) status (Bradford et al ., 1997, Hoffmann et al ., 2010, Mahlen, 2011, Sanders & Sanders, 1997) ; and d) its status as the premier model organism for G ram positive bacteria. B. subtilis possesses a well characterized an d easily tractable genetic system which is c ompatible with other prom inent G ram positive bacteria such as Paenibacillus sp p. (St John et al ., 2006) with potential for biomass to bio based chemical conversion .
18 Despite the presence of a functional 2,3 BD pathway, the yield (0.044 g/g ) and concentration (0.85 g/L) obtained by B. subtilis WN1038 (described in Chapter 3 , Figure 3 2A ) are low in comparison with the values obtained by the above cited species . In order to meet industrial demands for the economically feasible production of 2,3 BD, th ese production parameters must be improved. For background, a review of 2,3 BD production i n B. subtilis follows. By g lycolysis , one molecule of glucose generates 2 pyruvate s , 2 NADH s , 2 H + s , 2 H 2 O s and 2 ATP s . B. subtilis is a mixed acid alcohol fermenter, and pyruvate can be channeled towards the end products lactate, succinate, ethanol, acetate , acetoin, and 2,3 BD (Nakano et al. , 1997) . No formate accumulation is noted during fermentation (Nakano et al ., 1997) , indicating that pyruvate formate lyase is absent; in support of this observation, the pfl gene is not present in the B. subtilis genome (Kunst et al ., 1997) . Even though B. subtilis does not grow under strict ly anaerobic condition s by glucose fermentation , excess aeration leads to low er levels of 2,3 BD whereas low oxygen availability leads to higher 2,3 BD levels Nakano & Zuber, 1998) . pH also plays an important role on whether the culture will produce more acidic or more neutral compounds , and in general alkaline conditions favor the formation of organic acids, whereas acidic conditions reduc e the formation of organic acid and induce the synthesis of ne utral compounds like 2,3 BD (Garg & Jain, 1995) . Se tting the initial pH of fed batch fermentation at 6.0 gave higher yields of 2,3 BD in B. subtilis th a n with initial pHs of 7.0 or 8.0 (Zhang et al. , 2014) . The genetic organization of the pathway leading from pyruvate to 2,3 BD in B. subtilis has been elucidated (Renna et al ., 1993, Nicholson, 2008) . The pathway for 2,3 BD production begins with the condensation of two pyruvat acetolactate by acetolactate synthase (ALS) encoded by the alsS gene, -
19 acetolactate to acetoin by alsD acetolactate decarboxylase (ALDC) (Voloch et al. , 1985) . Acetoin is then reduced to 2,3 BD by the enzyme 2,3 butanediol dehydrogenase (BDH) encoded by the monocistronic bdhA gene (Nicholson, 2008) . The genes alsS and alsD form a bicistronic operon located at 3,707.8 to 3,710.4 kilobases (kb) on the B. subtilis chromosomal map (Kunst et al ., 1997) , and the monocistronic bdhA is located separately at 677.5 to 678.5 kb. The alsSD operon is positive ly regulated by the LysR like AlsR regulator ; however the bdhA gene is only indirectly regulated by AlsR (FrÃ¤drich et al. , 2012, Oliveira & Nicholson, 2013, Ong et al. , 2011) . Genetic engineering of B. subtilis to improve 2,3 BD production has recently attracted attention of the scientific community. Due to the differential expression of bdhA relative to alsSD (Oliveira & Nicholson, 2013) , the synchronization of both transcriptional units was thought to be a desirable trait. A construct placing the alsSD promoter region upstream from the bdhA coding sequence was cloned in to pAL 10 plasmid and transformed into B. subtilis 168 . In this strategy bdhA gene expression was dependent on the native regulation of P alsSD , thus placing bdhA transcription under direct control of AlsR . Compared with the parental strain , a 4 fold increase in 2,3 BD production was obtained in this recombinant (Biswas et al. , 2012) . Another way of increasing 2,3 BD production would be to increase the expression levels of a lsSD and bdhA by modulating Al sR transcriptional factor expression . Cloning of the alsR gene under regulation of the strong Hpa II promoter resulted in 1.25 and 1.38 fold increase d production of acetoin and 2,3 BD res pectively, whereas cloni n g alsR under P bdhA regulation a slightly higher increase of 1.32 and 1.49 fold in acetoin and 2,3 BD titers respectively was seen compared with the parental strain B. subtilis 168 (Zhang et al. , 2013) . The s trong induction of alsR by Hpa II regulation caused cell s to grow poorly in comparison with when P bdhA was used to regulate
20 expression. Accumulation of 2,3 BD was the highest when alsR was reg ulated by P bd hA but still acetoin accumulated 3 fold more then 2,3 BD (Zhang et al ., 2013) . Other fermentation pathways (e.g., to lactate or acetate), compete for carbon during 2,3 BD production. Elimination of competing pathways has been long used as a way to increase production of a desired fermentation product (Jantama et al. , 2008) . A s eries of gene deletions were constructed in B. subtilis 168 showing that inactivation of the gene encoding phosphotransacetylase ( pta ) increased production of 2,3 BD 2 fold , whereas inactivation of the gene encoding lactate dehydrogenase ( ld h ) lowered 2,3 BD titer by 1.26 fold (Cruz Ramos et al. , 2000) . Despite the discussed advant ages of utilizing B. subtilis as a platform to convert renewable carbon sources into 2,3 BD , there are some constrain t s limiting maximal production . Some possible limiting factors are: i) presence of competing fermentation pathways for pyruvate ; ii) uncoupled expression of alsSD operon and the bdhA gene ; iii) redox unbalanced state of the 2,3 BD p athway ; and iv) lack of anaerobic growth by glucose fermentation . E scherichia coli production of 2,3 BD E. coli is a Gram negative facultative anaerobe belon ging to the phylum Proteobacteria, class Gammaproteobacteria, order Enterobacteriales, and family Enterobacteriaceae. It is a widely prevalent digestive tract inhabitant of mammals, birds, and less prevalently fishes and frogs (Gordon & Cowling, 2003) Although some species pose a threat fo r human health ( e .g. Shiga toxin producing strains) (Kaper et al. , 2004) , i n this work the non pathogenic wild type strain E. coli W3110 ( K 12 ) and the laboratory strain E. coli DH 5 were used as heterologous hosts for 2,3 BD production (Jensen, 1993, Meselson & Yuan, 1968 ) . E. coli W3110 uses mixed acid fermentation to grow anaerobically, producing acetate, ethanol, lactate, formate and succinate from glucose (Kim et al. , 2007) . Some of the interesting aspects of applying E. coli to
21 produce 2,3 BD are: a) its e asy genetic manipulation, with an extensive number of molecular tools ( e.g. p lasmid s, promoters, modified strains); b) its well studied metabolic pathways; c ) the existence of e xcellent E. coli data banks ( e.g. Ecocyc); d ) i ts capacity for anaerobic growth, an important trait fo r industrial scale fermentation; e ) the absence of a native pathway for 2,3 BD synthesis, allowing for introduction of stereospecific 2,3 BD pathway s for production of isomeric ally pure 2,3 BD; and f ) successful track record for industrial production of bio based chemicals. Production of 2,3 BD in E. coli was first performed by externally providing acetoin to a strain expressing the K. pneumoniae bdh gene (Ui et al. , 1996) . Later , a n Sph I fragment of K . pneumoniae containing the entire alsDSB operon (occurring natively in t he order alsD , alsS , and bdh ) was clon ed into a plasmid and transformed into E. coli JM109, yielding 0.27 g/g 2,3 BD from glucose (Ui et al. , 1997) . The first synthetic operon designed for 2,3 BD production in E. coli contained the alsSD operon from B. subtilis combine d with one of four butanediol dehydrogenases , from B. subtilis , Clostridium beijerinkii , Thermoanaerobacter brockii , or K. pneumoniae , obtaining a maximum meso 2,3 BD yield of 0.34 g/g using the K. pneumoniae BDH (Yan et al. , 2009) . Acetolactate is also a precursor of the bran ched chain amino acids leucine, isoleucine, and valine (Goupil Feuillerat et al. , 1997) . Deletion of the gene acetohydroxy acid isomeroreductase ( encoded by the ilvC gene) involved in the leucine/isoleucine synthesis pathway, allowed the mutant E. coli strain carrying a synthetic operon for 2,3 BD production to yield 0.29 g/g of meso 2,3 BD (Nielsen et al. , 2010) . As discussed above , oxygen concentration plays an important role in governing the distribution of fermentati on end products. A derivative of E. coli strain JM109 , expressing from a plasmid a heterologous 2,3 BD pathway, increased
22 the final yield of 2,3 BD from 0.30 to 0.33 g/g simply by lowering the aeration of the cultur e (Li et al. , 2010) . Expressing the entire 2,3 BD pathway divergon alsRDSB (including putative AlsR transcriptional factor) from Enterobacter cloacae in E. coli yielded 0.41 g/g of 2,3 BD (Xu et al. , 2014) . The 2,3 BD pathway starts with the condensation of two pyruvates into the molecule acetolactate . This acid is unstable and upon accumulation is non enzymatically decarboxylated into diacetyl (Zhao et al. , 2009) . The presence of acetolact at e decarboxylase (AL D C ) guarantees the rapid enzymatic decarbox y lation of acetolactate into acetoin, pr eventing significant diacetyl formation (Goupil et al. , 1996) . It has been shown that diacetyl can be reduced to acetoin by the action of a butanediol dehydrogenase, regenerating one NAD + during the process et al ., 2011) . If diacetyl is first reduced to acetoin and then reduced again to 2,3 BD , two NAD + are regenerated , thus placing the pathway in redox balance with glycolysis . P revious studies have utilized engineered E. coli expressing BDH s from K. p neu moniae or Enterobacter cloacae subsp . dissolvens for conversion of externally provided diacetyl into 2,3 BD (Ui et al. , 2004, Li et al. , 2012, Wang et al. , 2013) . In the s e studies , diacetyl and glucose were simultaneously added to cell culture s , which converted diacetyl into 2,3 butanediol . Glucose was added with the purpose of providing reduced cofactors for the NADH dependent reactions of dia cetyl and acetoin reduction by BDH . Although th ese studies illustrate the viability of using a BDH to reduce diacetyl into acetoin, this strategy is cost ly due to the need for diacetyl. In order to be economically feasible the sugar being provided has to b e not just the source of reduced cofactors but also the carbon source for 2,3 BD production. When growing anaerobically by glucose fermentation, E. coli is capable of maintaining its redox balance by producing a mixture of lactate , formate, ethanol and ace tate (Clark, 1989) .
23 The first enzymes in line metabolizing pyruvate to generate these products, are pyruvate formate lyase ( PFL ) , pyruvate dehydrogenase (PDH) and lactate dehydrogenase ( LD H ) . Both enzymes PFL and PDH co nvert pyruvate into acetyl CoA ; howe ver , PDH expression is repre ssed under anaerobic conditions (Kim et al. , 2008) . For this reason, inactivation of the pfl and ldh A genes leads to an anaerobic minus growth phenotype due to redox unbalanced metabolism. Introduction of a 2,3 BD pathway in to a pfl and ld h back ground is highly desirable , because a strain lacking pfl and ld h that expresses the heterologous pathway for 2,3 BD synthesis would not grow anaerobically except by 2,3 BD fermentation. However, the 2,3 BD pathway, as it is found in B. subtilis , is redox unbalanced , using only a single NAD + regeneration step. In this study (Chapter 4) , we constructed an alternative pathway for redox balanced produ ction of 2,3 BD. acetolactate spontaneously decarboxylates into diacetyl. The constructed synthetic operon (described in Chapter 4) harbors only alsS and bdhA genes from B. subtilis , excluding the decarboxylase alsD gene involved in the conversion of acetolactate into acetoin (R enna et al ., 1993, Voloch et al ., 1985) . After transforming this construct into the E. coli strain YK29 ( pfl , ld hA ) 2,3 BD was produced in a redox balanced manner . This is the fir st time 2,3 BD pathway has been engineered to a redox balanced state.
24 Figure 1 1. 2,3 Butan e diol stereoisomers. A) ( S,S) 2,3 BD, B) Meso 2,3 BD and C) (R,R) 2,3 BD
25 CHAPTER 2 THE LYSR TYPE TRANSCRIPTIONAL REGULATOR (LTTR) A LS R INDIRECTLY REGULATES EXPRESSION OF THE BACILLUS SUBTILIS 2,3 BUTANEDIOL (2,3 BD) DEHYDROGENASE GENE BDHA AND 2,3 BD PRODUCTION Introduction Bacillus subtilis is a mixed acid fermenter, capable of converting pyruvate into lactate, acetate, ethanol, acetoin, and 2,3 butanediol (2,3 BD) as fermentative end products (Nakano & Zuber, 1998) . The fermentati on pathway leading from pyruvate to 2,3 BD is of particular recent interest because this compound is an important bio based chemical feedstock that can be converted into industrial solvents and precursors in the manufacture of synthetic rubber and plastics et al ., 2011, Nicholson, 2008) . The genetic organization of the pathway leading from pyruvate to 2,3 BD in B. subtilis has been elucidated (Nicholson, 2008, Renna et al ., 1993) (Fig ure 2 1 ). Two pyruvate molecules are condensed to acetolactate by the enzyme acetolactate synthase (ALS) encoded by the alsS gene, followed by decarboxylation of acetolactate to acetoin by alsD encoded acetolactate decarboxylase (ALDC) (Voloch et al ., 198 5) . Acetoin is then reduced to 2,3 butanediol by the enzyme butanediol dehydrogenase (BDH) encoded by the monocistronic bdhA gene (Nicholson, 2008) . The genes alsS and alsD form a bicistronic operon located at 3707.8 to 3710.4 kilobases (kb) on the B. subtilis chromosomal map (Kunst et al ., 1997) , and the monocistronic bdhA is located separately at 677.5 to 678.5 kb ( Figure 2 1 ). Adjacent to, and divergent from, the alsSD operon at positi on 3710.5 3711.4 kb on the B. subtilis map is located the alsR gene encoding the AlsR protein, which has been shown to positively regulate alsSD transcription in B. subtilis upon binding to the cis regulatory region between alsR and alsS (FrÃ¤drich et al ., 2012, Ong et al ., 2011, Renna et al ., 1993) . By protein se quence analyses, AlsR has been classified as a member of the LysR Type Transcriptional
26 Regulator (LTTR) family, featuring an N terminal DNA binding helix turn helix (HTH) motif and a C terminal co inducer binding domain (Maddocks & Oyston, 2008, Schell, 1993) . Despite considerable conservation both structurally and functionally, LTTR proteins regulate a diverse collection of genes, including those involved in virulence, metabolism, quorum sensing and motility (Maddocks & Oyston, 2008) . Because transcription of the alsSD operon is under A lsR control, we reasoned that bdhA may also be under direct transcriptional control of AlsR. In this communication we show that the influence of AlsR on bdhA expression depends on the physiological state of the cell and that AlsR exerts apparent positively regulation of expression of (i) a bdhA lacZ transcriptional fusion, (ii) bdhA mRNA levels, (iii) levels of BDH enzymatic activity, and (iv) production of 2,3 butandiol. However, we were unable to establish a direct physical role in vitro for AlsR in bdhA promoter binding or bdhA transcription, thus suggesting that AlsR plays an indirect role in bdhA expression. Materials and M ethods Bacterial strains, plasmids, and growth conditions The bacterial strains and plasmids used in this study are listed in Table 2 1 . Miller LB medium (Miller, 1972) was used throughout. Where indicated, glucose was added to LB to a final concentration of 1% or 2% (w/v). When appropriate, antibiotics were added to media at the following final concentrations: ampicillin (Ap), 100 g/ mL ; erythromycin (Em), 5 g/ mL ; spectinomycin (Spc), 100 g/ mL . Aerobic and oxygen reduced growth was performed by cultivating 20 mL of cell culture in a 125 mL sidearm flask on a rotary shaker at a speed of 250 monitored either with a Klett Summerson colorimeter fitted with the No. 66 (red) filter, or in a spectrophotometer set at 660 nm. Under these conditions, 100 Klett units = 1 OD 660 .
27 Recombinant DNA techniques and plasmid constructions Chromosomal DNA was isolated from B. subtil is overnight cultures using a standard protocol (Cutting & Vander Horn, 1990) . The concentration and purity of DNA were determined by UV absorbance at 260 and 280 nm (Sambrook & Russell, 2001) . Competent Escherichia co li DH5 and B. subtilis cells were prepared and transformed by standard techniques (Boylan et al. , 1972, Sambrook & Russell, 2001) . A plasmid for integration of a single copy transcriptional bdhA lacZ fusion into the B. subtilis chromosome was constructed as follows. Primers bdhA +211F and bdhA +812R (Table 2 2 ) were synthesized and used to amplify by PCR a 0.6 kbp fragment internal to the coding sequence of bdhA (Nicholson, 2008) . The PCR product was digested with Hin dIII and Bam HI, gel purified, ligated into Hin dIII Bam HI digested plasmid pMUTIN4 (Vagner et al. , 1998) and introduced into competent E . coli strain DH5 selecting for Ap R , resulting in plasmid pWN1078 (Table 2 1). Plasmid pWN1078 was used to construct congenic wild type and alsR::spc knockout strains carrying the bdhA lacZ fusion as described in Table 2 1. A plasmid for overproduction an d purification of AlsR protein was constructed as follows. The alsR coding sequence was amplified by PCR from wild type B. subtilis strain 168 chromosomal DNA using the primer pair alsR +1F and alsR +880R (Table 2 2 ) . The amplified 0.8 kbp fragment was digested with Nhe I and Xho I and ligated into Nhe I Xho I digested plasmid vector pET 24a (Novagen, EMD Millipore, Billerica, MA USA), which harbors a hexahistidine tag C terminal to the cloning site, resulting in plasmid pWN1471 (Table 2 1 ). AlsR protein production and purification The C terminal histidine tagged AlsR protein was overexpressed in Escherichia coli BL21 (DE3) cells as follows. Cells were grown in 1 L of liquid LB at 37Â°C to an OD 600 of 0.7,
28 then transferred thio D galactopyranoside (IPTG) was added to a final C. Cells were harvested by cen t m L of binding buffer (50 m M HEPES, 500 m M NaCl, 5 m M imidazole, 5% glycerol, pH 7.5), flash frozen in liquid N 2 , and stored at 70Â°C. The thawed cells were lysed in a liquid N 2 refrigerated bead beater in the presence of 1 m M phenylmethylsulfonyl fluoride ( PMSF). The lysate was NTA affinity column (Qiagen). The column was washed with ~30 column volumes of wash buffer (consisting of lysis buffer containing 15 mM imidazole), then Al sR 6xHis protein was removed from the column with elution buffer (lysis buffer containing 500 m M imidazole). Elution buffer was removed from the purified protein by centifugal filtration using an Ultrafree 15 Centrifugal Filter Device 10K NMWL (Millipore Corporation, Bedford, MA), and the retentate was resuspended in reaction buffer (10 m M HEPES, 500 m M NaCl, 2.5% glycerol and pH 7.5). Protein was analyzed by SDS polyacrylamide gel electrophoresis (Sambrook & Russell, 2001) . Protein concentration was determined by fluorometry using the Qubit Protein Assay Kit (Life Technologies). Enzymatic assays galactosidase assays were performed as described previously (Nicholson & Setlow, 1990) and activity was expressed in Miller units (Miller, 1972) . BDH assays were performed as described previou sly (Nicholson, 2008) ; one unit of BDH activity was defined as 1 mol NADH produced per minute per mL per OD 660 of culture using 2,3 BD as substrate . All assays were performed on triplicate cultures.
29 RNA isolation Cells at the indicated times and under the indicated growth conditions were harvested by rapid centrifugation in a microfuge and stored frozen at . Total DNA free RNA was extracted and purified using the RiboPure Bacteria kit (Ambion, Life Technologies, RNA in each sample obtained was measured by UV absorban ce at 260 and 280 nm (Sambrook & Russell, 2001) . Quantitative reverse transcriptase PCR (qRT PCR) To quantify bdhA mRNA levels, total RNA (0.5 g per sample) was subjected to separate qRT PCR reactions using primer pairs bdhA +5F / bdhA +197R or 16S+1075F / 16S+1295R (Table 2 2 ). Reactions were performed using the SuperScript III Platinum SYBR Green One Step qRT PCR kit (Invitrogen) and the manufact urer's recommended protocol. Reactions were performed and monitored in a MiniOpticon real time PCR detection system (Bio Rad , Hercules, CA ). All qRT PCR reactions were performed using biological triplicates and technical quadruplicates. The C t values obtai ned for bdhA mRNA were normalized to C t values obtained for 16S rRNA run in parallel on the same total RNA samples. Results were analyzed using the 2 method (Livak & Schmittgen, 2001) . Primer extension mapping The transcription initiat ion site of bdhA was located by primer extension mapping using the Primer Extension System AMV Reverse Transcriptase kit (Promega, Madison, WI, USA) bdhA +60R (Table 2 1) end with [ 32 P ]ATP (Amersham) using the DNA 5' End Labeling System (Promega). The end labeled primer was hybridized with 10 g of total RNA isolated as described above, and
30 extended with AMV reverse transcriptase using the Primer Extension System (Promega). The exten sion products were separated by electrophoresis through a 6% polyacrylamide/8M urea sequencing gel (Sambrook & Russell, 2001) . Extension products were compared to a DNA sequencing ladder (Sequenase version 2.0 Sequencing kit, USB Corp., Cleveland OH, USA) generated using the same labeled primer and a 501 bp PCR product spanning the entire ydjM bdhA intergenic region, generated using primer pair bdhA 438 / bdhA +60R ( Table 2 2 ). Products wer e visualized and quantified by phosphorimaging . Electrophoretic mobility shift a ssays (EMSAs) Target DNAs were prepared as follows. The alsR alsSD intergenic region, containing the alsSD promoter and AlsR cis binding regions, and the entire ydjM bdhA inter genic region, including all untranslated DNA sequences spanning the bdhA promoter (Fig ure 2 5), were PCR amplified as 155 bp and 437 bp fragments using the appropriate primer pairs (Table 2 2 ) . As a negative binding control, a 146 bp internal fragment of the rpoB coding sequence was also amplified by PCR with its specific primer pair (Table 2 2 ) . All PCR products were purified using the QIAquick PCR Purification Kit (Qiagen). EMSAs were performe d using purified and concentrated AlsR as follows. Different quantities of AlsR protein in reaction buffer were incubated with the PCR products described above in a total volume of 10 L for 20 min at room temperature, mixed with 3 of 50% glycerol, and electrophoresed through a 1% agarose gel cast and run in 1x tris borate buffer pH 8.2 (EDTA was not added to the buffer to avoid possible inhibition of the reaction). Electrophoresis was performed mL and visualized by UV transillumination. Negative images are shown for clarity.
31 Statistical analyses Basic statistical parameters and Analyses of Variance (ANOVA) were performed using commercial statistical softwar e (Kaleidagraph, version 3.6.2; Synergy Software, Reading, PA). Differences with P < 0.05 were considered statistically significant. Results Expression of the bdhA lacZ fusion in wild type and alsR::spc strains In order to study the effect of AlsR on bdhA gene expression in B. subtilis , a bdhA lacZ transcriptional fusion was constructed in plasmid pMUTIN4 and integrated by a single crossover event at the native bdhA locus in the wild type and alsR::spc knockout mutant strains, resulting in strains WN1079 an d WN1152, respectively (Table 2 1 ). Each strain was cultivated in LB medium either in the presence or absence of glucose (1% w/v) and with either high or low aeration. Growth (OD 660 ) and galactosidase activity from bdhA lacZ expression was measured during exponential growth and for 5 hours into stationary phase (Fig ure 2 2 ). When cells were grown in LB without glucose and at high aeration ( Figure 2 2A ), bdhA lacZ fusion expression in wild type strain WN1079 was relatively low during exponential growth, peaked ~1.5 hours after the onset of stationary phase (i.e., t 1.5 ) at ~1,100 Miller U, then declined ( Figure 2 2A ). Kinetics of the bdhA lacZ fusion expression in alsR::spc mutant strain WN1152 grown under the same conditions pa ralleled expression by the wild type strain, but peaked at ~400 Miller U ( Figure 2 2A ). Growth of cells in LB lacking glucose but under reduced aeration ( Figure 2 2B ) resulted in similar kinetics of bdhA lacZ fusion expression to a maximum at ~ t 1.5 , but e xpression was activated to a higher absolute level ( Figure 2 2B ). Again, peak fusion expression by the wild type strain WN1079 (~2,600 Miller U) was higher than the alsR::spc mutant (~1,300 Miller U) ( Figure 2 2B ). Growth of wild type strain WN1079 in LB c ontaining 1% glucose at high aeration resulted in maximal postexponential activation of bdhA lacZ fusion expression, and
32 expression continued to increase throughout stationary phase, reaching ~3,200 Miller U by t 5 ( Figure 2 2C ). In contrast, bdhA lacZ fusi on expression in the alsR::spc mutant strain WN1152 behaved in a marked ly different manner , peaking at ~900 Miller U at ~t 1.5 then declining ( Figure 2 2C ). Finally , wild type cells cultivated in LB containing 1% glucose under reduced aeration ( Figure 2 2D ) showed similar kinetics, but a somewhat lower level, of postexponential activation of bdhA lacZ expression (i.e., ~2,000 Miller U at t 5 ) than the same strain grown with high aeration (compare Figure 2 2C and Figure 2 2D ), whereas expression of bdhA lacZ b y the alsR::spc mutant strain showed similar kinetics and expression level as the same strain grown in LB lacking glucose at reduced aeration (compare Figure 2 2B and Figure 2 2D ). The expression data revealed some observations worthy of note. (i) Activati on of bdhA lacZ expression occurred in the early stationary phase. (ii) Expression of the bdhA lacZ fusion was dramatically enhanced either by glucose alone (compare Figure 2 2A and Figure 2 2C ) or by reduced aeration alone (compare Figure 2 2A and Figure 2 2B ). (iii) Fusion expression appeared to be somewhat decreased when glucose and lowered aeration were both applied ( Figure 2 2D ). (iv) Inactivation of alsR resulted in a reduction, but not a total abolition, of bdhA lacZ fusion expression. Measurement of bdhA transcript levels by qRT PCR Results obtained using the transcriptional bdhA lacZ fusion ( Figure 2 2 ) implied that transcription of bdhA was lowered in the alsR::spc mutant. To test this notion directly, steady state levels of bdhA mRNA were measured using qRT PCR ( Figure 2 3 ). Triplicate cultures of strains WN1038 (w.t.) and WN1192 ( alsR::spc ) were cultivated with aeration in LB either with or without 1% glucose. Total RNA was isolated from cells harvested at t 3 (i.e., 3 hours after the onset of stat ionary phase), when the bdhA lacZ fusion ( Figure 2 2C ) was fully activated, and bdhA mRNA levels were measured by qRT PCR as described in Materials and Methods. The relative level of bdhA transcripts was not significantly different between the wild type an d
33 alsR::spc strains grown in LB in the absence of glucose ( Figure 2 3 ). In contrast, growth in LB + 1% glucose strongly increased (~26 fold) bdhA transcript levels in the wild type strain WN1038, but not significantly in the alsR::spc knockout strain WN119 2 ( Figure 2 3 ). The results indicated that inactivation of alsR reduced the observed postexponential increase in the steady state level of bdhA mRNA. BDH activity in w.t. vs. alsR::spc strains In order to ascertain if expression of the bdhA lacZ fusion ( Fi gure 2 ) and bdh mRNA levels ( Figure 2 3 ) were consistent with BDH activity levels, BDH specific activity was measured in wild type strain WN1038 and in alsR::spc knockout strain WN1192 after 24 h of aerobic growth in LB alone or LB containing 1% glucose ( F igure 2 4A ). In LB without glucose, both strains produced essentially the same l ow amount of BDH activity (27.7 Â± 7.2 and 22.3 Â± 9.3 U, respectively) ( Figure 2 4A ). In LB containing 1% glucose, BDH activity in wild type strain WN1038 was 171 Â± 25.2 U, and activity in the alsR::spc knockout mutant was ~2.3 fold lower (74.7 Â± 5.7 U). While this difference was statistically significant, it was clear that the alsR::spc knockout mutation did not abolish production of BDH ( Figure 2 4A ). Examination of the kinetics of BDH activity during exponential growth and early stationary phase of w.t. strain WN1038 ( Figure 2 4B ) showed essentially the same expression pattern as that observed by the transcriptional bdhA lacZ fusion in WN1038 ( Figure 2 2C ) and was consis tent with the higher steady state level of bdhA mRNA at t 3 ( Figure 2 3 ), suggesting that the increase of BDH activity was a direct consequence of an increase in bdhA mRNA levels. In contrast, BDH activity in the alsR::spc mutant strain WN1192 also exhibite d a low basal level of activity during exponential growth, and was activated to a ~4.5 fold lower level than in the wild type strain by t 3 ( Figure 2 4B ). Again, levels of BDH activity mirrored expression of the bdhA lacZ fusion observed in the alsR::spc mu tant ( Figure 2 2B ) and steady state bdhA mRNA levels measured at t 3 ( Figure 2 3).
34 Mapping the bdhA transcription initiation site The transcription initiation site of the bdhA gene was mapped by primer extension using total RNA isolated from t 3 cultures of either wild type strain WN1038 or alsR::spc mutant strain WN1192, grown in either LB or LB+1% glucose (Fig ure 2 5). In all four cases a full length extension product was detected ( Figure 2 5 inset) corresponding to an adenine nucleotide situat ed 268 nt. upstream from the bdhA initiation methionine codon, and three minor extension products were identified at 153, 146, and 137 nt. preceding the bdhA initiation methionine codon (marked by asterisks in Figure 2 5 ). The longest extension product was presumed to represent the full length transcript; visual inspection of the sequence preceding this site revealed the presence of a region with excellent homology to the consensus 10 region characteristic of B. subtilis promoters recognized by the major s igma A form of RNA polymerase (Helmann, 1995, Jarmer et al. , 2001, Moran et al. , 1982) , but with poor homology to the consensus 35 region ( Figure 2 5 ). Quantification of transcrip t levels by densitometry ( Figure 2 5 inset) revealed that both wild type and alsR::spc mutant cells grown in LB+1% glucose exhibited ~5 fold more bdhA mRNA than cells grown in the absence of glucose. Interestingly, similar amounts of bdhA mRNA were detecte d by LB+1% glucose grown cells of both the wild type and alsR::spc mutant strain ( Figure 2 5 inset). This result was surprising and did not appear consistent with the results of bdhA mRNA quantification by qRT PCR ( Figure 2 3 ) which indicated that the alsR ::spc mutation resulted in a dramatic lowering of bdhA mRNA at t 3 . Purified AlsR protein does not bind the bdhA promoter region Previous work had demonstrated a direct role for AlsR in binding to, and stimulating transcription from, the alsSD promoter (FrÃ¤drich et al ., 2012, Ong et al ., 2011) . To test possible binding of AlsR to the bdhA promoter region, we cloned the wild type alsR gene into an IPTG inducible expression plasmid that would result in overexpression of AlsR protein with a
35 carboxyl terminal histidine tag, purified AlsR protein, and performed binding studies using EMSA ( Figure 2 6 ). In the first experiment ( Figure 2 6A ), a 2.5 fold molar excess of AlsR protein was mixed with three different PCR fragments, spanning: (i) the alsR alsS intergenic region, known to bind AlsR (FrÃ¤drich et al ., 2012, Ong et al ., 2011) and serving as a positive control; (ii) the ydjM bdhA intergenic region; and (iii) a PCR product corresponding to an internal segment of the rpoB gene (Nicholson & Maughan, 2002) , serving as a negative control. In this experiment, AlsR bound to the alsS promoter region, but not to the bdhA promoter region or the rpoB internal coding fragment ( Figure 2 6A ) . However, a small amount of smearing was noted in all complexes formed with a molar excess of AlsR, which may have obscured the bound form, so the experiment was performed again with limiting amounts of AlsR and a molar excess of target DNA ( Figure 2 6B ). This experiment showed AlsR binding to the alsSD promoter region at a protein:DNA ratio as low as 0.07, but no detectable binding of AlsR to the bdhA promoter region at this molar ratio ( Figure 2 6B ). Discussion It has previously been demonstrated that ex pression of the alsSD operon encoding ALS and ALDC is under direct positive transcriptional control of the LTTR family protein AlsR (FrÃ¤drich et al ., 2012, Ong et al ., 2011, Renna et al ., 1993) . The results presented here are also consistent with a positive role for AlsR in bdhA expression, but this role may not be direct. First, expression of a transcriptional bdhA lacZ fusion was lowered, but not abolished, in the alsR::spc mutant strain ( Figures 2 2A, B, C ). However, bdhA lacZ exp ression in cells grown with 1% glucose under low aeration appeared not to depend on a functional alsR gene ( Figure 2 2D ), suggesting that bdhA expression may not be absolutely dependent on AlsR. Second, as measured by qRT PCR, the steady state level of bdh A mRNA was dramatically lower in the alsR::spc mutant than in the wild type strain at t 3 grown in LB with 1% glucose ( Figure 2 3 ); but, as
3 6 measured by primer extension, levels of bdhA mRNA were essentially the same at t 3 in both the wild type and alsR::spc mutant strain grown under the same conditions ( Figure 2 5 ). Third, BDH enzymatic activity was lowered somewhat in the alsR::spc strain measured at either t 3 or at t 24 (Fig ure 2 4 B ), but again BDH activity was not abolished in the alsR::spc mutant. And fourth, we were unable to detect direct binding of purified AlsR protein to the bdhA promoter region, although it quite clearly bound to the alsSD promoter region ( Figure 2 6 ). The above results indicate that regulation of bdhA expression depe nds only indirectly, and only in part, upon the alsR gene product and that its regulation is more complex than previously supposed. In line with this notion, mapping of the bdhA transcription start site revealed that it was separated from the BdhA initiati on methionine codon by a 268 nt 5' UTR ( Figure 2 5 ). In recent years it has become increasingly evident that such 5' UTR regions are associated with attenuators or riboswitches which respond directly to a wide range of small metabolites to exert regulatory control at the level of transcription termination/antitermination, translation, or mRNA stability (Gelfand, 2006, Henkin, 2008, Serganov & Nudler, 2013) . This possibility could explain why the primer extension experiment, which targeted mainly the 5' UTR, detected essentially equivalent amounts of bdhA transcripts initiated by the wild type and alsR::spc strains ( Figure 2 5 inse t), whereas the qRT PCR experiment, which targeted mRNA in the downstream bdhA coding region, (see primer locations in Figure 2 5 ) detected a large difference in bdhA mRNA levels ( Figure 2 3 ). One feature of riboswitches and other attenuators is sequences with potential to form alternate stem loop secondary structures; inspection of the bdhA 5' UTR reveals a number of such regions (data not shown), but their possible significance in bdhA regulation has not yet been tested experimentally. In this light it is interesting to note that primer extension revealed three apparently truncated extension products within the 5' UTR
37 ( Figure 2 5 ). These may have resulted from blockage of reverse transcriptase at regions of secondary structure in the 5' UTR, or alternative ly may simply be due to stalling reverse transcriptase during its passage through monotonous poly T tracts ( Figure 2 5 ). These possibilities will require further experimentation to elucidate. One could envision a number of possible testable explanations fo r the observed effect of the alsR::spc mutation on bdhA expression. For example, it has been demonstrated that mutations inactivating alsR cause severe down regulation of alsSD expression (Renna et al ., 1993) , and thus a dramatic reduction the am ount of ALS and ALDH, hence acetolactate and acetoin, produced. Because ALS and ALDH do not directly participate in re oxidation of reduced NADH during fermentation, expression of bdhA may be directly responsive to intracellular acetolactate and/or acetoin levels in the cell as a measure of the intracellular redox state. Such regulation might not be exerted at the level of transcription initiation, but by a mechanism mediated through the 5' UTR as in other attenuator or riboswitch systems. Improving our und erstanding of the regulatory circuitry governing 2,3 BD production in the model Gram positive platform organism B. subtilis can ultimately lead to its rational manipulation with the goal of maximizing 2,3 BD productivity in a variety of industrially import ant Gram positive microorganisms.
38 Table 2 1 . Strains and plasmids used in this study Strain or plasmid Genotype or relevant characteristics Source (reference) B. subtilis: WN1038 Prototroph (Nicholson, 2008) WN1079 Prototroph, bdhA lacZ ; Em R pWN1078 WN1038 tf; Em R WN1148 (AMBs2) trpC2, pheA1, alsR::spc ; Spc R E. HÃ¤rtig (Cruz Ramos et al ., 2000) WN1152 alsR::spc, bdhA lacZ ; Spc R , Em R WN1148 WN1079 tf; Spc R /Em R WN1192 Prototroph, alsR::spc ; Spc R WN1148 WN1038 tf; Spc R E. coli: deoR , endA1 , gyrA96 , hsdR17 , (r k , m k + ), recA1 , relA1 , supE44 , thi 1 , ( lacZYA argFV169 ), 80 lacZ M15 , F , (Hanahan, 1983) BL21(DE3) F , omp T, hsd SB(r B , m B ), gal, dcm (DE3) Novagen WN1078 DH5 carrying plasmid pWN1078; Ap R This study WN1471 BL21(DE3) carrying plasmid pWN1471 This study Plasmids pMUTIN4 Integrative vector for B. subtilis (Em R ); replicates in E. coli (Ap R ) (Vagner et al ., 1998) p WN1078 pMUTIN4 carrying 0.6 kb Hin dIII/ Bam HI insert internal to bdhA coding region; bdhA lacZ fusion; Ap R in E. coli ; Em R in B. subtilis This study pET24a Protein expression vector, replicates in E. coli (Km R ) Novagen pWN1471 pET24a carrying 0.918 kb Nhe I/ Xho I insert of the alsR 6xHis gene; replicates in E. coli (Km R ) This study Abbreviation: t f , transformation
39 Table 2 2. Oligonucleotide primers used in this study Primer A Sequence Purpose alsR +1F 5 CTTA GCTAGC ATGGAGCTTCGCCATCTTCA Cloning into pET24a; AlsR overproduction alsR +880R ATA CTCGAG TGTACCTGCATCACTCTCTTTAGTTC Cloning in pET24a; AlsR overproduction bdhA +211F GCG AAGCTT GAATTCTCCGGTGAAGTTGCG Cloning in pMUTIN4; bdhA lacZ construction bdhA +812R CCC GGATCC TTTTCCCAAATGCTGACGAT Cloning in pMUTIN4; bdhA lacZ construction bdhA+ 5F AGGCAGCAAGATGGCATAAC qRT PCR bdhA +197R ACAGGTGCCGTTTCATTTGT qRT PCR 16S+1075F CAGCTCGTGTCGTGAGATGT qRT PCR 16S+1295R TGTGGGATTGGCTTAACCTC qRT PCR bdhA 438F ( ydjM bdhA intergenic region) TTCGTCCCCCTGTTTGTTAA PCR template for sequencing; EMSA of bdhA promoter region bdhA 1R ( ydjM bdhA intergenic region) GGATTACCACTCCTATAACTTTTGATG EMSA of bdhA promoter region bdhA +60R TGGCTCTTCGATATGTTCAA TACGGATATCC PCR template for sequencing and primer extension alsR alsS intergenic region F CCCTCACTCCTTATTATGCATTT 3 EMSA of alsS promoter region alsR alsS intergenic region R TTCAATATGCATTCCTTTCCA EMSA of alsS promoter region rpoB internal F CGTCCTGTTATTGCGTCC 3 EMSA of internal rpoB coding sequence rpoB internal R TCCGGCACGCTCACG EMSA of internal rpoB coding sequence A Coordinates in primers are relative to the bdhA translation initiation codon. F, forward; R, reverse. Sequences in italics denote cleavage sites for Bam HI, Hin dIII, Nhe I, and Xho I.
40 Figure 2 1. The 2,3 BD biosynthetic pathway in B. subtilis. Genetic organization of the divergent alsR gene and alsSD operon and the unlinked monocistronic bdhA gene (top line). The pathway from pyruvate to 2,3 BD (bottom line). AlsR is a positive regulator of alsSD transcription (FrÃ¤drich et al ., 2012 , Renna et al ., 1993) , denoted by the plus sign.
41 Figure 2 2. Beta galactosidase activities (squares) expressed from the bdh lacZ transcriptional fusion during growth (circles) of strains WN1079 (w.t., open symbols) and WN1152 ( alsR::spc ; filled symbols). Cultures were grown in LB either with or without glucose ( 1% w/v) and with or without vigorous aeration (denoted by high or low O 2 ), as indicated in each panel A through D . Data shown are averages Â± standard deviations ( n = 3); Beta galactosidase activity error bars not visible are smaller than the plot symbols.
42 Figure 2 3. Levels of bdhA mRNA measured by qRT PCR . WN1038 (w.t.; light gray bars) and WN1192 ( alsR::spc ; dark gray bars) were grown aerobically in LB with or without glucose ( 1% w/v). Data were normalized to both 16S rRNA and strain WN1038 grown in LB without glucose as described in the text. Data shown are averages Â± standard deviations ( n = 3). Lowercase letters denote groups significantly different by ANOVA ( P 0.05).
43 Figure 2 4. Units of BDH activity in 24 h and time course. A) BDH activity in 24 h cultures of strains WN1038 (w.t.; light bars) and WN1192 ( alsR::spc ; dark bars) grown aerobically in LB or in LB + 1% glucose (LB + Glc). Data shown are averages Â± standard deviations ( n = 3). Lowercase letters denote groups sign ificantly different by ANOVA ( P 0.05). B ) Time course of BDH activity (squares) during growt h (circles) of strains WN1038 (w.t.; open symbols) and WN1192 ( alsR::spc ; filled symbols) cultured in LB with both glucose ( 1% w/v) and aeration. Downward arrow denotes t 0 , the transition between exponential and stationary growth phase .
44 Figure 2 5. Results of primer extension experiment (inset) in the context of the nucleotide sequence of the bdhA upstream regulatory region. Primer bdhA + 60R was 5 end labeled with 32 P and hybridized with total RNA extr acted from strain WN1038 (w.t.) or strain WN1192 ( alsR::spc ) grown to t 3 in LB in the presence (+ ) or absence ( ) of 1% glucose. Extension products were mapped relative to a sequencing ladder of the corresponding DNA region using the same primer. The ident ified transcription initiation site is marked +1. The numbers below each full length extension product represent relative intensity of each transcript band, normalized to w.t. grown in LB without glucose. The three bands marked with asterisks denote incomp lete extension products; the locations of which are denoted in the sequence. In the sequence, the transcription initiation site (+1), putative 10, and 35 promoter regions are denoted by bold capital letters, and the putative ribosome binding site (rbs) i s denoted by bold italic capitals. Above each sequence are listed the consensus 35 and 10 regions of sigma A dependent promoters and the consensus ribosome binding sites determined for B. subtilis (Helmann, 1995, Moran et al ., 1982) ; matches are denoted by underlines.
45 Figure 2 6. Electrophoretic mobility shift assay of purified AlsR p rotein with the promoter region of bdhA . DNA fragments used were: alsSD promoter ( alsSD P ; positive control), bdhA ( bdhA P ) , and an internal coding sequence fragment of the rpoB gene ( rpoB I ; negative control). Concentrations of DNA targets and AlsR protein are denoted above the lanes. B , DNA bound to AlsR; F , free DNA. A) EMSA performed with excess AlsR. B) EMSA performed with excess DNA.
46 CHAPTER 3 SYNTHETIC OPERON FOR 2,3 BUTANEDIOL PRODUCTION IN BACILLUS SUBTILIS AND ESCHERICHIA COLI Introduction Although the discovery of new oil and gas reserves and development of more efficient utilization technologies may temporarily stave off an immediate crisis, petroleum depletion over the long term will inexorably lead to steeply increased prices and supply shortages (Aguilera, 2014, Murray & King, 2012) . To alleviate this situation, alternative sources for energy generation are either already established or rapidly comi ng online. Examples include: nuclear, solar, wind, geothermal, tidal or wave energy; microbial fuel cells or hydrogen fuel cells; and biomass thermal conversion. However, petroleum not only provides energy for transport and electricity generation, but is a valuable source of chemical feedstock compounds for the petrochemical industry. T he abovementioned alternative sources do not provide carbon feedstocks for chemical production et al ., 2012, Lumbreras & Ramos, 2012, Rabaey et al ., 2004, Rourke et al ., 2010, Steubing et al ., 2011, Villa et al ., 2013) . Microbial fermentation of renewable carbon sources such as lignocellulose can be employed as a sustainable alternative to petroleum for production of chemical feedstocks. More than thirty years ago it was suggested that a major fraction of the US chemical industry could be supplied by four feedstocks derived from biological fermentation: ethanol, isopropanol, n butanol, and 2,3 butanediol (Palsson et al. , 1981) . Since then, however, only bioethanol production has been actively developed into a mature industry (Otero et al. , 2007) . Thus, exploration of economic routes to the large scale synthesis of other chemical feedstocks is a high priority. The secondary alcohol 2,3 butanediol (2,3 BD) is amenable to a large number of applications such as an industrial solvent, fuel additive, jet fuel, antifreeze, building block for
47 synthetic rubber and bioplastics, foo d additives, and printer ink (Cruz Ramos et al ., 2000, et al ., 2011, Nicholson, 2008, Nakano & Zuber, 1998) . Some members of the family Enterobacteriaceae , such as Klebsiella oxytoca, K. pneumoniae, an d Serratia marcescens, produce high amounts of 2,3 BD. However, because these organisms are nosocomial pathogens, they are inappropriate for synthesis of products for human consumption (Moet et al. , 2007, Podschun & Ullmann, 1998) . As an alternative, the GRAS (Generally Regarded A s Safe) organism Bacillus subtilis produces 2,3 BD and has recently been engineered for efficient cellulolytic breakdown of renewable biomass (Anderson et al. , 2013) . The pathway leading from pyruvate to 2,3 BD in B. subtilis consists of two pyruvate acetolactate synthase (ALS) encoded by the alsS acetolactate to acetoin b y alsD acetolactate decarboxylase (ALDC) (R enna et al ., 1993, Voloch et al ., 1985) . Acetoin is then reduced to 2,3 BD by the enzym e butanediol dehydrogenase (BDH) encoded by the bdhA gene (Nicholson, 2008) . The genes alsS and alsD form a bicistronic operon located at 3,707.8 to 3,710.4 kilobases (kb) on the B. subtilis chromosomal map (Kunst et al ., 1997) . The alsSD operon is under positive transcriptional control by the AlsR protein which is expressed from the divergent alsR gene (Renna et al ., 1993) . In contrast, the monocistronic bdhA gene is located separately from the alsRSD divergon at 677.5 to 678.5 kb and is indirectly regulated by AlsR (Oliveira & Nicholson, 2013) . The poor anaerobic growth capability of B. subtilis , coupled with the complex regulation of its 2,3 BD pathway, lim its its use in industrial scale fermentation processes (Nakano & Zuber, 1998, Nicholson, 2008, Schallmey et al. , 2004) . In contrast, the biotechnology workhorse Escherichia coli lacks a 2,3 BD fermentation pathway but grows well anaerobically, making it
48 an interesting potential heterologous host for industrial scale production of enantio merically pure 2,3 BD (Ji et al ., 2011) . It was therefore considered desirable to simplify genetic manipulation of the 2,3 BD pathway and its high level expression in either E. coli or B. subtilis . I n this study we describe the construction and characterization of a synthetic alsSDbdhA operon under IPTG inducible control and carried on an E. coli B. subtilis shuttle plasmid, with the goal of investigating how to optimize high level production of 2,3 BD production in both platform microorganisms. Materials and M ethods Bacterial strains, plasmids, and growth conditions The bacterial strains and plasmids used in this study are listed in Table 3 1 . Miller LB medium (Miller, 1972) was used throughout. Where indicated, gl ucose was added to LB to a final concentration of 2% (w/v). When appropriate, antibiotics were added to media at the mL mL ; and mL . All cultivations were performed at 37Â°C. Optical density was monitored either with a Klett Summerson colorimeter fitted with the no. 66 (660 nm; red) filter or in a spectrophotometer set at 660 nm. Under these conditions, 100 Klett units = 1 OD 660 . When required, logarithm ic phase cells (0.5 OD 660 D 1 thiogalactopyranoside (IPTG) at the indicated final concentration. Oxygen reduced growth was performed by cultivating 40 mL of cell culture in a 250 mL sidearm flask on a rotary shaker at a speed of 100 rpm. Recombinant DNA techniques Chromosomal DNA was isolated from B. subtilis overnight cultures using a standard protocol (Cutting & Vander Horn, 1990) . DNA concentration was determined by fluorometry using the Qubit system (Invitrogen, Carlsbad, CA, USA) . All PCR reactions were performed
49 using Phusion High Fidelity DNA polymerase with HF buffer (New England Biolabs, Beverly, MA, USA) according to the manufacturer's instructions. Plasmid pDG148 Stu (Josep h et al. , 2001) was digested with the restriction enzyme Stu I . T4 DNA polymerase (New England Biolabs) was used to treat Stu I digested pDG148 Stu and PCR generated fragments (Joseph et al ., 2001) . Competent E. coli and B. subtilis cells were prepared and transformed by standard techniques (Boylan et al ., 1972, Sambrook & Russell, 2001) . Plasmid cons truction A plasmid carrying a synthetic operon containing the B. subtilis alsSD operon fused to bdhA was constructed using the primer pairs described in Figure 3 1A and B. subtilis strain WN1038 genomic DNA as the template. The alsSD operon (with a modified alsS initiation methionine codon from TTG to the canonical ATG sequence ), lacking its natural promoter and rho independent transcriptional terminator was amplified by PCR using Primers 1 and 2 ( Figure 3 1A ). The bdhA gene was amplified using Primers 3 and 4, w hich removed the native bdhA promoter but kept the rho indepen dent transcriptional terminator ( Figure 3 1A ). After individual PCR amplification of alsSD and bdhA with their respective sets of primers, a PCR fusion based approach (Hobert, 2002) was performed to link both fragments into a single synthetic operon, alsSDbdhA. The fusion PCR product s were cloned into pDG148 Stu and introduced into competent cells of E. coli R . Individual Ap R transformants were screened for acetoi n production using a modified Voges Proskauer assay (Nicholson, 2008) ( Figure 3 1B ). The acetoin positive clones were then assayed quantitatively for BDH activity (Nicholson, 2008) ( Figure 3 1C ). Transformant #16 (strain WN1390) was chosen based on its elevated production of both acetoin and BDH activity. Production of high levels of ALS, ALDC and BDH proteins was confirmed by SDS PAGE of cell lysates ( Figure 3 1D ). Plasmid DNA
50 was isolated from strain WN1390 and its proper construction was confir med by sequencing the insert at the University of Florida Interdisciplinary Center for Biotechnology Research (UF ICBR). Plasmid DNA from transformant #16, designated pWN1390, was used in subsequent experiments. Gel electrophoresis DNA and protein were ana lyzed by agarose gel electrophoresis and sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS PAGE), respectively, using standard protocols (Sambrook & Russell, 2001) . Assays Bu tanediol dehydrogenase (BDH) activity was assayed from cell free extracts as described previously (Nicholson, 2008) ; 1 unit produced per minute per milliliter per OD 660 of culture. Acetoin from culture supernatants was assayed by a modified Voges Proskauer test as described in detail previously (Nicholson, 2008) . Assay of fermentation products by HPLC High performance liquid chromatography (HPLC) analyses of fermentation products were performed as describe d in detail previously (Nicholson, 2008, Underwood et al. , 2002) , with the exception that the column was operated at a pressure of 90 bars, a temperature of 60Â°C, and a flow rate of 0.7 mL /min. Chromatograms were processed using Chemstation software (Agilent). Compounds were identified on the basis of their retention times vs. pure standards and the concentration of each fermentation product was determined by peak area integration. Statistical analyses Basic statistical parameters and analyses of variance (ANOV A) were performed using commercial statistical software (KaleidaGraph, version 3.6.2; Synergy Software, Reading, PA). Differences with P
51 Results Construction of a synthetic alsSDbdhA operon With the aim of increasing 2,3 BD production in either E. coli or B. subtilis , an IPTG inducible expression system was constructed in shuttle plasmid pDG148 Stu (Joseph et al ., 2001) capable of replication and expression in either host. Figure 3 1A depicts the scheme for construction of a single synthetic operon encoding the entire 2,3 BD pathway of B. subtilis under control of the IPTG inducible P spac promoter. Notably, construction involved removal of the following natural regulatory regions: (i) the alsSD promoter region; (ii) the alsSD rho independent transcriptional terminator, and; (iii) the bdhA promoter and 5' untranslated region. The bdhA ribosome binding site was placed direc tly downstream from the alsD stop codon, the alsS initiation codon was changed from TTG to the canonical ATG, and the natural bdhA rho independent transcriptional terminator was retained ( Figure 3 1A ). After PCR amplifying the alsSD and bdhA sequences as d ecribed above , both fragments were linked together using fusion PCR (Hobert, 2002) . The re sulting synthetic alsSDbdhA construct was incorporated into digested pDG148 Stu and then transformed into competent E. coli strain DH5 ransformants were picked from LB ampicil l in plates and inoculated overnight in LB + 1% glucose. C ulture supernatants were screened for acetoin production using a rapid Voges Proskauer test ( Figure 3 1B ). Thirteen out of 25 transformants screened were acetoin positive, indicating that ALS and ALDC were being expressed, and these isolates were subsequently screened for BDH activity ( Figure 3 1C ). All 13 acetoin positive clones also expressed BDH, and plasmid DNA from strain WN1390, the clone exhibiting the highest expression, was chosen for further examination.
52 Production of 2,3 BD by B. subtilis background strains The cons truction of wild type B. subtilis reference strain WN1038 was previously described by Nicholson ( 2008 ) and its alsR::spc knockout derivative, strain WN1192 was described in Chapter 2 . Because these two strains were to serve as the B. subtilis host strains for expression of 2,3 BD from the synthetic alsSDbdhA operon, we first characterized their conversion of glucose to acetoin and 2,3 BD in liquid LB medium containing 2% (111 mM) glucose ( Figure 3 2 ). After 48 h, w.t. strain WN1038 had consumed 107.1 Â± 4.4 mM glucose from the medium, whereas WN1192 only consumed 16 Â± 6 mM ( Figure 3 2 ). Strain WN1038 produced a maximum 2,3 BD concentration of 9.5 Â± 1 mM by 24 h, and production remained constant at 48 h; in contrast, strain WN1192 produced no detectable 2,3 BD ( Figure 3 2 ). Wild type strain WN1038 produced 63 Â± 0.7 mM acetoin by 48 h, whereas alsR::spc strain WN1192 produced only 6.5 Â± 0.1 mM acetoin during the same time period ( Figure 3 2 ). Production of 2,3 BD in B. subtilis strains carrying the synthetic als SDbdhA operon The complete 2,3 BD pathway from B. subtilis was constructed as a synthetic tricistronic operon, cloned into the IPTG inducible shuttle plasmid pDG148 Stu, and the resulting plasmid (pWN1390) was introduced by transformation into B. subtilis wild type strain WN1038 and alsR::spc knockout strain WN1192, resulting in strains WN1394 and WN1468, respectively ( Table 3 1 ). Production of 2,3 BD and acetoin from glucose by strains WN1394 and WN1468 was determined in LB medium containing 2% (w/v) glucose ( Figure 3 3 ). Over the course of 48 h, strain WN1394 (w.t.) consumed essentially all the glucose, produced a maximum of 20 Â± 1 mM 2,3 BD at 24 h, and a maximum of 90 Â± 1 mM acetoin at 48 h ( Figure 3 3A ). Induction of expression of the alsSDbdhA syn thetic operon with 1 mM IPTG caused strain WN1394 to produce 27 Â± 1 mM 2,3 BD and 93 Â± 4 mM acetoin by 24 h and 48 h respectively ( Figure 3 3B ). Compared to control strain WN1038 at the same time points ( Figure
53 3 2A ) , IPTG induction of strain WN1394 led to a 2.8 fold and 1.4 fold increase in 2,3 BD and acetoin titer respectively. In contrast, uninduced cells of strain WN1468 ( alsR::spc ) consumed very little glucose over 48 h and yielded only 0.55 Â± 0.02 mM 2,3 BD and 7.1 Â± 0.11 mM acetoin by 48 h ( Figure 3 3C ). This was due to the alsR::spc knockout mutation in its parental strain WN1192 ( Figure 3 2B ). However, IPTG induction of alsSDbdhA expression in strain WN1468 resulted in consumption of nearly all the glucose and yielded 12.5 Â± 1.2 mM 2,3 BD and 64.7 Â± 2.3 mM acetoin by 24 h and 48 h respectively ( Figure 3 3D ) . From these results it appeared that 2,3 BD production in strain WN1468 was derived nearly exclusively from the synthetic alsSDbdhA operon. Production of 2,3 BD in E. coli During construction of t he synthetic alsSDbdhA operon in plasmid pDG148 Stu it was noted that E. coli transformant WN1390 exhibited high levels of extracellular acetoin ( Figure 3 1B ), BDH enzymatic activity ( Figure 3 1C ) and ALS, ALDC, and BDH protein ( Figure 3 1D ). These results suggested that strain WN1390 might also produce high quantities of 2,3 BD. In order to investigate its fermentation profile and to test the maximum amount of 2,3 BD produced, a fed batch experiment was performed in which strain WN1390 was inoculated into LB medium containing 2% (111 mM) glucose either lacking ( Figure 3 4A ) or containing 1 mM IPTG ( Figure 3 4B ). In the uninduced culture, after 24 h nearly all the glucose was consumed and a high amount of 2,3 BD (65 Â± 4 mM) and acetoin (20 Â± 6 mM) was produc ed ( Figure 3 4A ). The 24 h culture was then fed another 111 mM glucose and incubation continued. By the end of 48 h, again most of the glucose was consumed leaving 32 Â± 21 mM glucose, and the 48 h concentrations of 2,3 BD and acetoin had increased to 98 Â± 21 mM and 35 Â± 1 mM, respectively ( Figure 3 4A ). Another 111 mM glucose was added to the 48 h culture and incubation continued.
54 It appeared that 2,3 BD production had ended between 48 h and 72 h, as by 72 h glucose consumption had essentially ceased, and t he titers of 2,3 BD (105.3 Â± 33 mM) and acetoin (48 Â± 15.2 mM) had increased by only modest amounts (data not shown). In stark contrast, induction with IPTG appeared to be exerting an inhibitory effect on the performance of strain WN1390 in fed batch cultu re ( Figure 3 4B ). Only ~28 mM glucose was consumed by 24 h, and glucose consumption had ceased by 48 h ( Figure 3 4B ). Only 4.6 Â± 0.4 mM 2,3 BD was produced by 48 h, and essentially no acetoin was produced ( Figure 3 4B ). Discussion Lignocellulosic biomass i s an abundant and renewable feedstock employable in production of bio based chemicals (Pauly & Keegstra, 2010) . The commonly used laboratory strain of B. subtilis , strain 168, secretes a number of extracellular glucanases, xylanases and arabinases useful for converting plan t biomass into fermentable substrates (Antelmann et al ., 2001) . Strain 168 and its derivatives such as WN1038 can produce 2,3 BD ( Figure 3 2A ); however their productivity of 2,3 BD is lower than other platform organisms like the pathogen Klebsiella pneumoniae (Ji et al ., 2011) ; in addition, B. subtilis grows poorly under anaerobic conditions. On the other hand, the biotechnology workhorse E. coli is capable of growing anaerobically, a required trait for industrial scale fermentation ; however, it presents an inferior capacity for secreting cellulases and other extracellular enzymes (Bokinskya et al. , 2011) , and it completely lacks the 2,3 BD path way. With the intent of producing high levels of 2,3 BD in either B. subtilis or E. coli , we constructed a synthetic alsSDbdhA operon in a bifunctional, IPTG inducible expression plasmid pWN1390, which was introduced into both bacteria . IPTG induction of t he synthetic operon expression in B. subtilis strain WN1394 led to peak production of 27 Â± 1 mM 2,3 BD in 24 h ( Figure 3 3B ), corresponding to a 2.8 fold increased production in comparison to wild type strain
55 WN1038 (9.5 Â± 0.7 mM) ( Figure 3 2A ). Enhanced p roduction of 2,3 BD was shown to be attributable to the synthetic alsSDbdhA operon, as IPTG inducible 2,3 BD production was demonstrated in B. subtilis strain WN1468 ( Figure 3 3C,D ), whose parent strain B. subtilis strain WN1192 carries an alsR::spc knockout mutation abolishing production of 2,3 BD from the endogenous alsSD operon and bdhA gene ( Figure 3 2B ). Placement of the synthetic alsSDbdhA operon into E. coli strain WN1390 led to a dramatically increased titer of 2,3 BD in the culture medium (9 8 Â± 21 mM at 48 h) ( Figure 3 4A ), which corresponded to ~10 fold increase over that seen in the benchmark strain B. subtilis WN1038. It was interesting to note that IPTG induction of strain WN1390 actually resulted in lower conversion of glucose to 2,3 BD and acetoin ( Figure 3 4B ). We attribute this to toxic effects resulting from dramatic overproduction of ALS, ALDC, and BDH in E. coli . This suggestion is supported by the observation that cell growth was impaired in strain WN1390 by IPTG induction ; uninduc ed cells reached ~4 OD 660 units, whereas IPTG induced cells only attained ~2 OD 660 units (data shown). However, high titer production of 2,3 BD in the absence of IPTG induction would potentially be a desirable feature to reduce large scale production costs (Van Dien, 2013) . Given the dramatically higher 2,3 BD titers attained in E. coli relative to B. subtilis , current experiments are aimed toward maximizing 2,3 BD production in E. coli. We were able to increase 2,3 BD production in both B. subtilis and E. coli by introduction and expression of a synthetic tricistronic operon derived from the B. subtilis alsS, alsD, and bdhA genes under control of the IPTG inducible P spac promoter. Both en gineered strains, especially E. coli strain WN1390, showed superior fermentative traits. Further development of both strains will still be required to attain industrial requirements of productivity.
56 Table 3 1 . Bacterial strains and plasmids used in this study Strain or plasmid Genotype or relevant characteristics Source (reference) B. subtilis: WN1038 Prototroph (Nicholson, 2008) WN1192 WN1038 carrying alsR :: spc ; Spc R (Oliveira & Nicholson, 2013) WN1394 WN1038 carrying pWN1390 This study WN1468 WN1 192 carrying pWN1390 This study WN1465 WN1038 carrying pDG148 Stu This study WN1467 WN1192 carrying pDG148 Stu This study E. coli: deoR , endA1 , gyrA96 , hsdR17 , (r k , m k + ), recA1 , relA1 , supE44 , thi lacZYA argFV169 ), 80 , F (Hanahan, 1983) WN1390 alsSDbdhA synthetic operon cloned into pDG148 Stu This study WN1448 Stu empty vector This study Plasmids pDG148 Stu Shuttle vector for protein expression; Ap R ( E. coli); Km R ( B. subtilis ) Bacillus Genetic Stock Center (Joseph et al ., 2001) pWN1390 pDG148 Stu carrying the synthetic operon alsSDbdhA under P spac control This study
57 Figure 3 1. Construction of synthetic alsSDbdhA operon. A) Schematic diagram of the B. subtilis alsSD operon and unlinked bdhA gene. Hooked arrows, transcription initiation sites. Stem loops, rho independent transcription terminators. Numbered arrows, oligonucleotide primers used. Colored portions of primers correspond to highlighted sequences below. B) Liquid overnight cultures of 25 Ap R transformants were screened for acetoin production by Voges Proskauer as say. Negative controls consisted of E. coli are acetoin positive. C) BDH activity of Ap R transformants from panel B) Transformant #16 (highlighted in orange) was selected for further analysis and d esignated strain WN1390 harboring plasmid pWN1390. D) Analysis of proteins on Coomassie Blue stained 10% SDS PAGE. Equivalent amounts of cell free extract , adjusted by OD 660 , were loaded on each lane. ALS, acetolactate synthase; ALDC, acetolactate decarbox ylase; BDH, butanediol dehydrogenase.
58 Figure 3 2. Glucose consumption and production of 2,3 BD and acetoin by B. subtilis background strains. A) Wild type strain WN1038 and B) alsR::spc knockout strain WN1192 were cultivated in LB+2% glucose and assayed at 0 h (black bars), 24 h (white bars), and 48 h (gray) bars. Data are expressed as averages Â± standard deviations ( n = 3). n.d., not detected.
59 Figure 3 3. Glucose consumption and production of 2,3 BD and acetoin by B. subtilis strains carrying plasmid pWN1390. Wild type strain WN1394 (A, B) and alsR::spc knockout strain WN1468 ( C , D) each carrying plasmid pWN1390 (P spac alsSDbdhA ) were cultivated either witho ut induction ( A, C) or after induction with 1 mM IPTG ( B, D) . Samples were taken for assay at 0 h (black bars), 24 h (white bars), and 48 h (gray) bars. Data are expressed as averages Â± standard deviations ( n = 3). n.d., not detected.
60 Figure 3 4. Fed batch flask fermentation profile of E. coli strain WN1390 carrying the synthetic alsSDbdhA operon. Consumption of glucose and production of 2,3 BD and acetoin were assayed in uninduced (A) and IPTG induced (B) cultures grown in LB+2% glucose. Samples were taken for assay at 0 h (black bars), 24 h (white bars), and 48 h (gray) bars. Data are expressed as averages Â± standard deviations ( n = 3). n.d., not detected.
61 CHAPTER 4 REDOX BALANCED PRODUCTION OF ISOMERIC ALLY PURE (R,R) 2,3 BUTANEDIOL IN ESCHERICHIA COLI Introduction H igh p etroleum prices and excessive carbon dioxide emission s have prompted the need of developing processes to produce chemical s and fuels from renewable carbon sources. Through micro bial fermentation the promising commodity chemical 2,3 butanediol (2,3 BD) can be produced, and its further processing can yield important chemicals including: methyl ethyl ketone (an important industrial solvent formed by dehydration of 2,3 BD), 1,3 butad iene ( a precursor for synthetic rubber obtained by 2,3 BD diacetate pyrolysis), 2 butanol (a n industrial solvent and promising biofuel produced by enzymatic dehydration and dehydrogenation of 2,3 BD), diacetyl (butter flavor, obtained from 2,3 BD dehydroge nation) and others (Ji et al ., 2011, S peranza et al ., 1997) . Klebsiella oxytoca, K. pne u moni a e and Serratia marcescens produce high levels of 2,3 BD, but are human health threatening nosocomial pathogens (Jung et al. , 2013) . Bacillus subtilis is a G enerally R ecognized A s S afe (GRAS) organism , but produces low amounts of 2,3 BD when compared with other spe cies and has a very inefficient glucose fermentation pathway, hindering its usage for industrial scale production of 2,3 BD (Ji et al ., 2011, Nakano & Zuber, 1998) . The non pathogenic Escherichia coli strain W1330 grows anaerobically and possesses no 2,3 BD pathway, allowing the production of isomer ic ally pure 2,3 BD molecules in this heterologous host (Yan et al., 2009, Ui et al., 1996) . The best studied 2,3 BD synthesis pathway consists o f : (i) condensation of 2 pyruvates acetolactate by the enzyme acetolactate synthase (ALS) ; (ii) decarboxylation of acetolactate into acetoin by the enzyme acetolactate decarboxylase (ALD C ) ; and (iii) NADH dependent reduction of acetoin into 2, 3 B D , performed by the acetoin reductase activity of the
62 enzyme 2,3 butanediol dehydrogenase (BDH) et a l ., 2011) . A similar but less understood pathway requires no ALD C enzyme, relying on spontaneous decarboxylation of acetolactate into diacetyl (B ryn et al. , 1971) . BDH can convert diacetyl into 2,3 BD through two successive NADH dependent reduction steps regenerating 2 NAD + , thus establishing redox balance when using glucose as a fermentable substrate (Li et al ., 2012, Willey et al. , 2008) . Using the B. subtilis genes encoding the 2,3 BD pathway , we constructed two synthetic operons, one which includ ed all three genes of the pathway ( alsS alsD bdhA ) and an other lacking the alsD gene ( alsS bdhA ) . After introducing the synthetic opero ns in to E. coli strain YK29 pfl and ldh ) , we were able to produce high amounts of (R,R) 2,3 BD and also to prove the viability of a proposed shunt on the 2,3 BD pathway in order to attain the redox balanced 2,3 BD production in E. coli. Materials and Methods Bacterial strains, plasmids, and growth conditions The bacterial strains and plasmids used in this study are listed in Table 4 1. Miller LB medium (Miller, 1972) was used for gene cloning, engineering steps. Fermentation experiments were performed in a medium containing mineral salts , per L: Na 2 HPO 4 , 6.25 g; KH 2 PO 4 , 0.75 g; NaCl, 2 g; (NH 4 ) 2 SO 4 , 1 g; MgSO 4 . 7H 2 O, 0.2 g; FeSO 4 . 7H 2 O, 0.01 g; NaMoO 4 . 2H 2 O, 0.01 g; Na 2 SeO 3 , 0.263 mg (Lee et al. , 1985) , supplemented with 10 g tryptone, 2 g potassium acetate and 50 g glucose. When appropriate, antibiotic s were added to media at the following final concentrations: ampicillin (Ap), 100 mL mL . Wh en required, cells at 0.5 OD 660 of logarithmic phase were induced with 0.04 mM IPTG final concentration. All cultivations were
63 performed by incubating 20 mL of cell culture in a 125 mL sidearm flask on a rotar y shaker at a speed of 100 rpm . When required, anaerobic conditions were established by flushing the medium with N 2 gas, and air locks were used to prevent O 2 from entering the flasks . Optical density was monitored as described before (Chapter 2). Recombinant DNA techniques and plasmid constructions Chromosomal DNA was isolated from B. subtilis overnight cultures using a standard protocol (Cutting & Vander Horn, 1990) . DNA concentration was determined by fluorometry using the Qubit Â® system (Invitrogen, Carlsbad, CA, USA) . All PCR reactions were performed using Phusion High Fidelity DNA polymerase with HF buffer (New England Biolabs, Beverly MA, USA) according to the manufacturer's instructions . Plasmid pET24a (EMD Millipore, Billerica, MA, USA) was digested with the restriction en zymes Nhe I and Xho I . Competent E. coli was prepared and transformed by standard techniques (Boylan et al ., 1972, Sambrook & Russell, 2001) . P lasmids carrying a synthetic operon containing either alsS alsD bdhA or alsS bdhA were constructed as follows. In order to construct alsS als D bdhA synthetic operon , the entire alsSD operon was amplified from the B. subtilis strain WN1038 genome by PCR, us ing the forward ctta gctagc atg ttgacaaaagcaacaaaagaa underlined sequence indicates the Nhe I enzyme restriction site; bold sequence indicates the initiation methionine codon) and the reverse atggattaccactcctataacttt tta ttcagggcttcctt cagttg underlined sequence indicates the bdhA overlapping region for PCR fusion based approach; bold sequence indicates alsD stop codon) were used. In order to PCR amplify the bdhA gene from the genome of strain WN1038, caactgaagg aagccctgaataa aaagttataggagtggtaatcc atg aa underlined sequence indicates the alsD overlapping region for PCR fusion based approach, italic sequence indicates the bdhA RBS region, and bold indicates the bdhA initiation methionine codon) and the
64 ata ctcgag aaggattctggggctgaagt Xho I enzyme restriction site) were designed. To construct the synthetic alsS bdhA operon lacking the alsD gene , the entire alsS co ding sequence was amplified again from the B. subtilis strain WN1038 genome by PCR with the same forward primer used to amplify alsSD and with the atggattaccactcctataacttt cta gagagctttcgttttcatga indicates the bdh A overlapping region for PCR fusion based approach and bold sequence indicates alsS stop codon). To amplify the bdhA gene, the same reverse primer used for alsSDbdhA construction tcatgaaaacgaaagctctctag aaagttataggagtggtaatcc atg aa alsS overlapping region for PCR fusion based approach and bold sequence indicates the bdhA initiation methionine codon). After individual PCR amplification of alsSD, alsS and bdhA with the ir respective sets of primers, a PCR fusion based approach (Hobert, 2002) was performed to link the corresponding fragments into a single synthetic operon, alsS als D bdhA or alsS bdhA. The fusion PCR products were digested with Nhe I and Xho I, cloned into Nhe I Xho I digested plasmid pET24a , and introduced into competent cells of E. coli selecting for Km R (Hobert, 2002, Joseph et al ., 2001, Sambrook & Russell, 2001) . The proper construct s w ere identified by inoculation of kanamycin resistant transformants in to liquid LB+Km containing 1% (w/v) glucose, and screening for acetoin production after 24 h using a modified Voges Proskauer assay (Nicholson, 2008) . The resulting plasmids were designated pWN1526 ( alsS bdhA ) and pWN1528 ( alsS alsD bdhA ), respectively. Introduction of T7 RNA polymerase gene into E. coli W1330 and YK29 backgrounds Using the DE3 lysogenization kit (Novagen) according to the manufacturer's directions, prophage DE3 was used to place the IPTG inducible phage T7 RNA polymerase gene into E.
65 coli strains W3110 (w.t.) and YK29 [ ( ldhA ) ( focA pflB ) ] , resulting in strains WN1533 and WN1531, respectively ( Table 4 1). Growth tolerance on diacetyl In order to determine the inhibitory effect of exogenous diacetyl, wild type E. coli W3110 cells were inoculated at an initial turbidity of 0.05 OD 660 into liquid LB containing various final concentrations of diacetyl ranging from 0 to 10 mM. Cell OD 660 was measured after 24 h. Measurements were performed on triplicate culture s incubated at each diacetyl concentration . Detection and quantification of fermentation products by HPL C High performance liquid chromatography (HPLC) analyses of fermentation products were performed as described in detail previously (Nicholson, 2008, Underwood et al ., 2002) , with the exception that the column was operated at a t emperature of 45Â°C, and a flow rate of 0.4 mL /min. Samples of culture supernatant were run on a Hewlett Packard HP1090 HPLC instrument equipped with an Aminex HPX87 H column (Bio Rad laboratories, Hercules, CA) and UV absorbance and refractive index detectors in series (Agilent Technologies, Santa Clara, CA). Compounds were identified by their retention times relative to pure standards , and the concentration of each fermentation product was determined by peak area integration. St atistical analyses Basic statistical parameters and analyses of variance (ANOVA) were performed using commercial statistical software (KaleidaGraph, version 3.6.2; Synergy Software, Reading, PA). Differences with P nificant. Results Construction of alsS als D bdhA and alsS bdhA synthetic operon s Two synthetic operons , alsS als D bdhA (Figure 4 1A) and alsS bdhA (Figure 4 1B) were constructed and independently cloned into the expression vector pET24a under T7
66 transcriptional promoter regulation. S trains were transformed with empty pET24a vector (control), vector pWN1526 containing the alsS als D bdhA construct and vector pWN1528 containing the alsS bdhA construct. The construct alsS als D bdhA harbors the entire B. subtilis 2,3 BD pathway (Figure 4 1A). Glycolysis generates 2 NADHs per glucose consumed, of which only one is regenerated into NAD + through the B. subtilis 2,3 BD pathway (Figure 4 1A). The elimination of alsD acetolactate) from the synthetic operon (Figure 4 1B), alters the 2,3 BD pathway as follows. Alpha acetolactate is converted into diacetyl by non enzymatic, s pontaneous decarboxylation. BDH reduces diacetyl into acetoin, then into 2,3 BD, in the process regenerating 2 NAD + s and consequently establishing the redox balance of the pathway. Redox balanced 2,3 BD fermentation by WN1538 ( alsS bdhA ) Batch fermentation was conducted under anaerobic conditions for 120 hours in medium containing an initial glucose concentration of 50 g/L. After 120 hours ~half of the glucose had been consumed in both uninduced and induced strain WN1538, producing 0.55 Â± 0.03 g/L and 0.97 Â± 0.0 4 g/L 2,3 BD respectively (Fig ure 4 2 A , B; Table 4 2). The yield obtained with induced cells represents 0.04 g of 2,3 BD per g of glucose, being equivalent to 8% of the theoretical maximum (0.5 g 2,3 BD/g glucose). In 120 h induced cells produced 0.6 Â± 0.05 g/L acetoin, whereas uninduced cells produced 0.2 Â± 0.1 g/L acetoin (Figure 4 2A, B). Induced WN1538 produced 0.3 Â± 0.48 g/L of diacetyl in 120 h, whereas uninduced cells produced 0.02 g/L during the same period of time. Both uninduced and induced WN1538 cells grew better than strains WN1534 and WN1535 carrying the empty vector pET42a in 120 h, however worse than strain WN1539 carrying the complete alsS alsD bdhA synthetic operon (Table 4 2).
67 Redox unbalanced 2,3 BD fermentation by stra in WN1539 ( alsS als D bdhA ) The same conditions as for WN1538 above were used to assess strain WN1539 2,3 BD fermentation performance. Most of the glucose was consumed in 120 h , leaving in the medium only 4.75 g/L for the uninduced condition , and 5.93 g/L f or the induced condition (Figure 4 2 C , D; Table 4 2). Strain WN1539 produced the most 2,3 BD , 17.1 g/L and 14.3 g/L for uninduced and induced cells, respectively. After 120 h, uninduced strain WN1539 yielded 0.4 g of 2,3 BD per gram of glucose consumed, c orresponding to 80% of the theoretical maximum with a productivity of 0.14 g/[L h] . Also the highest cell density ( 3.21 OD 660 ) was reached with uninduced WN1539 in 120 h (Table 4 2). A very small amount of acetoin ( ~0.2 g/L ) was produced in 120 h by strain WN1539 under either uninduced and induced condition s . A trace of diacetyl was detected i n both conditions (~0.075 Â± 0.03 g/L ) . Induction of strain WN1539 with 0.0 4 mM IPTG resulted in an overall lower performance , resulting in lower 2,3 BD production ( 14. 3 Â± 2.08 g/L ) and leaving more glucose ( 6 Â± 7.17 g/L ) in the medium. Fermentation products of WN1534, WN1535, WN1538 and WN1539 E. coli W3110 uses mixed acid fermentation to grow anaerobically, producing acetate, ethanol, lactate, formate and succinate whe n growing in LB + 50 g/L glucose (Kim et al ., 2007) . Lactate was the predominant fermentation product for strain WN1534 in 120 h (7 Â± 0.7 g/L), whe reas WN1535 produced no detectable amount of lactate (Table 4 2). Strains WN1538 and WN1539 , uninduced or induced , accumulated ~0.2 g lactate /L after 120 h (Table 4 2). With the exception of the target fermentation product 2,3 BD, all remaining products accumulated at very low concentrations after 120 h for all strains carrying any of the synthetic operons (Table 4 2) . acetolact ate was not quantified ; however diacetyl accumulated at its highest c oncentration (0.3 g/L) in strain WN1538 lacking the alsD gene.
68 Diacetyl effect on growth Diacetyl is known to react with arginyl residues on proteins (Mathews et al. , 2010, Yankeelov et al. , 1968) . Due to the absence of alsD in strain acetolacta te spontaneously decarboxylates into diacetyl and may acc umulate to toxic levels if not subsequently converted into acetoin and 2,3 BD . In order t o assess the impact of diacetyl on E. coli W3110 growth , cells were inoculated at a initial OD 660 of 0.1 and grown for 24 h in LB containin g a final concentration of diacetyl ra nging from zero to 10 mM (Figure 4 3). D iacetyl concentration s of 5 mM (0.43 g/L) or higher comp letely suppressed cell growth during the first 24 h . Discussion The B. subtilis 2,3 BD pathway is not redox balanced, contain ing only a single NAD + regeneration step (Oliveira & Nicholson, 2013) . This presents a major constraint on the industrial scale fermentation of this secondary alcohol requiring O 2 to oxidize the second NADH generated during glycolysis. Here, we propose and de monstrate a functional alternative redox balanced pathway for 2,3 BD production in E. coli. The engineered 2,3 BD pathway lacks the gene for ALDC ( alsD ), thus introducing an extra NAD + regeneration step (Figure 4 1B). E. coli strain YK29 is not capable of growing anaerobically due to redox unbalanced metabolism. Introducing the synthetic operon alsS bdhA provides a redox balanced pathway for anaerobic growth. However strain WN1539, carrying the redox unbalanced pathway alsS als D bdhA grew better than redox balanced strain WN1538 after 120 h of batch fermentation (Figure 4 2 A, B, C and D; Table 4 2). Examining WN1538 and WN1539 (uninduced and induced) fermentation products, WN1539 yields a significantly higher concentration of 2,3 BD and consumes more glucose after 120 h (Figure 4 2 A, B, C and D). Even though strain WN1538 produced acetoin, diacetyl and 2,3 BD, the concentrations levels of these compounds never reached concentrations
69 as high as for WN1539 (Table 4 2). One possible explanation for this observation is that acetolactate occurs much more slowly than does the ALDC mediated step in the pathway, thus leading to the low 2,3 BD yields seen in strain WN1538 (Figure 4 2A, B). Another aspect to be considered is diacetyl toxicity shown in Figure 4 3, where a concentration higher than 5 mM in the medium proved to be enough to completely hinder E. coli (w.t.) growth (Figure 4 3). Uninduced strain WN1538 accumulated toxic levels of ~ 3.5 mM diacetyl in the medium . Ac cumulation of diacetyl in the fermentation broth of WN1538 suggests that B. subtilis BDH d id not efficiently reduce diacetyl to acetoin. Bacteria that produce high levels of 2,3 BD, such as K. pneumoni a e , possess a BDH enzyme with a K cat 60x higher than B. subtilis BDH used here (Yan et al ., 2009) . Even though K. pneumoniae BDH has a higher Kcat , this enzyme reduces acetoin into meso 2,3 BD , an stereoisomer with inferior antifreeze properties than t he R,R 2,3 BD produced by B. subtilis BDH (Boutron, 1990, Yan et al ., 2009) . A construct carrying a more efficient BDH could prevent diacetyl accumulation and increase the 2,3 BD production levels. The above considerations must cer tainly be addressed by future research.
70 Table 4 1. Bacterial strains and plasmids used in this study Strain or plasmid Genotype or relevant characteristics Source (reference) B. subtilis: WN1038 Prototroph (Nicholson, 2008) E. coli: deoR , endA1 , gyrA96 , hsdR17 , (r k , m k + ), recA1 , relA1 , supE44 , thi lacZYA argFV169 , F , (Hanahan, 1983) W3110 Wild type E. coli K 12 ATCC 27325 YK29 ( ldhA ) ( focA pflB ) Km s (Kim et al ., 2007) WN1526 alsS bdhA synthetic operon cloned into pET24a This study WN1528 alsS als D bdhA synthetic operon cloned into pET24a This study WN1531 YK29 (DE3) DE3 YK29 WN1533 W3110 DE3) DE3 W3110 WN1534 W3110 (DE3) carrying pET24a vector pET24a WN1533 WN1535 YK29 (DE3) carrying pET24a vector pET24a WN1531 WN153 6 W3110 (DE3) carrying alsS bdhA synthetic operon cloned into pET24a pWN1526 WN1533 WN153 7 W3110 (DE3) carrying alsS als D bdhA synthetic operon cloned into pET24a pWN1528 WN1533 WN1538 YK29 (DE3) carrying alsS bdhA synthetic operon cloned into pET24a pWN1526 WN1531 WN1539 YK29 (DE3) carrying alsS als D bdhA synthetic operon cloned into pET24a pWN1528 WN1531 Plasmids pET24a Protein over expression vector, f1 ori ; Km R Novagen pWN1526 pET24a carrying the alsS bdhA synthetic operon ; Km R This study pWN1528 pET24a carrying the alsS als D bdhA synthetic operon ; Km R This study
71 Table 4 2. Fermentation products, glucose consumption and cell growth after 120 h batch fermentation Strain Cell density (OD 660 ) Glucose consumed (g/L) 2,3 Butane diol (g/L) Acetoin (g/L) Diacetyl (g/L) Pyruvate (g/L) Succinate (g/L) Lactate (g/L) Ethanol (g/L) Acetate (g/L) 2,3 BD (g/[L h]) 2,3 BD (g/g) WN1534 1.91 Â± 0.03 14.66 Â± 0.08 ND 0.2 Â± 0.06 ND 0.40 Â± 0.07 0.06 Â± 0.007 7 Â± 0.7 0.15 Â± 0.006 0.66 Â± 0.06 WN1535 1.46 Â± 0.02 15.32 Â± 0.14 ND 0.6 Â± 0.05 ND 0.60 Â± 0.06 0.08 Â± 0.005 ND 0.25 Â± 0.02 1 Â± 0.009 WN1538 uninduced 2.34 Â± 0.11 27.01 Â± 0.08 0.55 Â± 0.03 0.2 Â± 0.1 0.3 Â± 0.48 0.30 Â± 0.06 0.6 Â± 0.016 ND 0.08 Â± 0.07 0.53 Â± 0.17 0.0046 0.02 WN1538 induced 2.54 Â± 0.01 24.26 Â± 1.58 0.97 Â± 0.04 0.6 Â± 0.05 0.02 ND 0.37 Â± 0.053 0.2 Â± 0.3 0.86 Â± 0.07 0.84 Â± 0.04 0.008 0.04 WN1539 uninduced 3.2 Â± 0.01 45.25 Â± 2.54 17.1 Â± 0.88 0.2 Â± 0.06 0.07 Â± 0.02 1.13 Â± 0.01 0.21 Â± 0.06 0.2 Â± 0.07 0.56 Â± 0.05 0.11 Â± 0.006 0.14 0.38 WN1539 induced 3 .0 Â± 0.44 44.07 Â± 7.17 14.3 Â± 2.08 0.2 Â± 0.1 0.08 Â± 0.05 0.20 Â± 0.17 0.48 Â± 0.076 0.2 Â± 0.2 0.92 Â± 0.16 0.12 Â± 0.06 0.12 0.32 Data are expressed as averages Â± standard deviations ( n = 3). ND, not detected.
72 Figure 4 1. 2,3 BD synthetic operons and corresponding pathways . A) Genetic organization of the redox unbalanced alsS als D bdhA synthetic operon and relevant pathway steps from 2 pyruvate s to 2,3 BD , B) g enetic organization of the redox balanced alsS bdhA synthetic operon and relevant pathway steps from pyruvate to 2,3 BD. Arrows are color coded with the constructs, indicating where each enzyme acts in the pathway. Black arrow indicate s the non enzymatic step of the pathway. Structural formula e of each molecule involved in the pathways are represented below each corresponding step.
73 Figure 4 2 . Fermentation products, growth and glucose consumption of uninduced and induced strains WN1538 and WN1539 during 120 h of anaerobic batch culture with 50 g/L glucose . Strain WN1538 ( A , B ) and WN1539 (C, D) were cultivated in LB+2% glucose either uninduced (A, C) or induced with 0.04 mM IPTG (B, D). Symbols are: growth (diamonds); 2,3 BD (circles), acetoin (crossed squares); diacetyl (inverted triangle s ); and glucose (vertical dashes). Data are expressed as averages Â± standard deviations ( n = 3).
74 Figure 4 3 . Diacetyl effect on E. coli W3110 growth after 24 h . Blue bars indicate c ell growth measured 24 h after inoculation in LB + diacetyl ranging from 0 to 10 mM final concentration. Data are expressed as averages Â± standard deviations ( n = 3).
75 CHAPTER 5 SUMMARY OF RESEARCH According to the American Petroleum Institute, oil reserves will be depleted by 2062. Utilization of renewable carbon sources to produce plastics, chemicals, and fuels can alleviate dependence on oil. Because 2,3 BD has multiple applications such as indust rial solvents, food additives, jet fuel, building bl ocks f or synthetic rubber, and others, engineering biocatalysts capable of mass producing this chemical feedstock is desirable en route to reduce d reliance on petroleum . A b etter understanding of the 2,3 BD pathway regulation in B. subtilis is critical to the rational design of strateg ies for improving 2,3 BD production by this bacterium . I t is well established that the LysR like transcriptional factor AlsR directly regulates alsSD operon transcription (FrÃ¤drich et al ., 2012, Renna et al ., 1993) , which corresponds to the first two steps of the 2,3 BD pathway . W e examined the not well understood AlsR influence on B. subtilis bdhA expression, and our results suggest ed that bdhA is indirectly regulated by AlsR (Chapter 2). Organisms producing high amounts of 2,3 BD such as K. pneumoniae and E. cloacae have their 2,3 BD pathway genes ( alsS , alsD , and bdhA ) arranged as a single operon, whereas in B. subtilis the bdhA gene is un linked to the alsSD operon. In an attempt to increase native production of 2,3 BD, an IPTG inducible synthetic ope ron was constructed placing the pathway genes alsS , alsD , and bdhA as part of a single operon . The new genetic organization and regulation increased 2,3 BD production in B. subtilis ~ 2. 8 fold ( Figure 3 3B; Chapter 3). Although E. coli lacks a native 2,3 BD producing pathway it has the great advantage over B. subtilis of being ca pable to grow anaerobicall y , a required feature for economically fea sible industrial scale fermentation . I n order to perform anaerobic fermentation, the pathway to produce a certain compound is required to be in a redox balanced state . Unfortunately , t he native B.
76 subtilis 2,3 BD pathway is not redox balanced, regenerating only a single NAD + per glucose molecule fermented . Therefore in C hapter 4 we describe d the engineering of a redox balanced synthetic p athway for 2,3 BD production . In order to achieve this goal , a n operon carrying only the genes als S and bdhA was con acetolactate dehydrogenase , coded by alsD , leads to acetolactate accumulation, an unstable molecule that spontaneously decarboxylates into diacetyl , whereas in its presence acetolactate is enzymat ically decarboxylated into acetoin. The great advantage of decarboxylati ng acetolactate into diacetyl rather then into acetoin is the fact that diacetyl can suffer two reduction rounds due to the presence of two carbonyl groups . In a n NADH dependent mann er, the enzyme butanediol dehydrogenase is capable of subsequentially reducing diacetyl into acetoin and 2,3 BD, regenerating 2 NAD + s and establishing the redox balance of the pathway. Glucose fermentation by strain WN1538 ( alsS bdhA ) , however, yielded less 2,3 BD than did strain WN1539 ( alsSalsDbdhA ) . Because of the observed accumulation of diacetyl in the culture after 12 0 h by strain WN1538 (Table 4 2 ) , we tested the possible toxicity of diactyl in E. coli (Figure 4 3 ) . Our results showed t hat the level of diacetyl seen in 120 h (3.5 mM) is toxic to the cells. Th us, diacetyl toxicity could be preventing WN1538 from reaching high titer s of 2,3 BD. Evolving strain WN1538 to tolerate higher concentrations of diacetyl might be an importa nt step for future work. This work describes the first rational engineering of the 2,3 BD pathway to obtain redox balanced production of 2,3 BD through glycolysis .
77 LIST OF REFERENCES Aguilera, R.F., (2014) Production costs of global conventional and unconventional petroleum. Energy Policy 64 : 134 140. Anderson, T.D., J.I. Miller, H.P. Fierobe & R.T. Clubb, (2013) Recombinant Bacillus subtilis that grows on untreated plant biomass. Appl Environ Microbiol 79 : 867 876. Antelmann, H., H. Tjalsma, B. Voigt, S. Ohlmeier, S. Bron, J.M. Dijl & M. Hecker, (2001) A proteomic view on genome based signal peptide predictions. Genome Res 11 : 1484 1502. Biswas, R., M. Yamaoka, H. Nakayama, T. Kondo, K . Yoshida, V.S. Bisaria & A. Kondo, (2012) Enhanced production of 2,3 butanediol by engineered Bacillus subtilis . Appl Microbiol Biotechnol 94 : 651 658. Bokinskya, G., P.P. Peralta Yahyaa, A. Georgea, B.M. Holmesa, E.J. Steena, J. Dietricha, T.S. Leea, D. Tullman Erceka, C.A. Voigtg, B.A. Simmonsa & J.D. Keasling, (2011) Synthesis of three advanced biofuels from ionic liquid pretreated switchgrass using engineered Escherichia coli . Proc Natl Acad Sci USA 108 : 19949 19954. Boutron, P., (1990) Levo and dextr o 2,3 butanediol and their racemic mixture: very efficient solutes for vitrification. Criobiology 27 : 55 69. Boylan, R.J., D. Brooks, F.E. Young & N.H. Mendelson, (1972) Regulation of the bacterial cell wall: analysis of a mutant of Bacillus subtilis defe ctive in biosynthesis of teichoic acid. J Bacteriol 110 : 281 290. Bradford, P.A., C. Urban, N. Mariano, S.J. Projan, J.J. Rahal & K. Bush, (1997) Imipenem resistance in Klebsiella pneumoniae is associated with the combination of ACT 1, a plasmid mediated A mpC b lactamase, and the loss of an outer membrane protein. Antimicrob Agents Chemother 41 : 563 569. Brecha, R.J., (2013) Ten reasons to take peak oil seriously. Sustainability 5 : 664 694. Bryn, K., Ã˜. Hetland & F.C. StÃ¸rmer, (1971) The reduction of diacetyl and acetoin in Aerobacter aerogenes evidence for one enzyme catalyzing both reactions. Eur J Biochem 18 : 116 119. butanediol -current state and prospects. Biotechnol Adv 27 : 715 725. Chapman, I., (2013) The end of Peak Oil? Why this topic is still relevant despite recent denials. Energy Policy 64 : 93 101. Clark, D.P., (1989) The fermentation pathways of Escherichia coli . FEMS Microbiol Rev 5 : 223 234.
78 Cruz Ramos, H., T. Hoffmann, M. Marino, H. Nedjari, E. Presecan Siedel, O. Dreesen, P. Glaser & D. Jahn, (2000) Fermentative metabolism of Bacillus subtilis : physiology and regulation of gene expression. J Bacteriol 182 : 3072 3080. Cutting, S.M. & P.B. Vander Horn, (1990) Genetic analysis. In: Molecular Biological Methods for Bacillus. C.R. Harwood & S.M. Cutting (eds). Sussex, UK: John Wiley and Sons, pp. 27 74. Dittmar, M., (2012) Nuclear energy: status and future limitations. Energy 37 : 35 40. F rÃ¤drich, C., A. March, K. Fiege, A. Hartmann, D. Jahn & E. HÃ¤rtig, (2012) The transcription factor AlsR binds and regulates the promoter of the alsSD operon responsible for acetoin formation in Bacillus subtilis . J Bacteriol 194 : 1100 1112. Garg, S.K. & A. Jain, (1995) Fermentative production of 2,3 butanediol: a review. Bioresour Technol 51 : 51:103 109. Gelfand, M.S., (2006) Bacterial cis regulatory RNA structures. Mol Biol 40 : 609 619. f geothermal energy conversion systems: application to the future construction of enhanced geothermal systems in Switzerland. Energy 45 : 908 923. Gordon, D.M. & A. Cowling, (2003) The distribution and genetic structure of Escherichia coli in Australian vertebrates: host and geographic effects. Microbiology 149 : 3575 3586. Goupil, N., G. Cortier, S.D. Ehrlich & P. Renault, (1996) Imbalance of leucine flux in Lactococcus lactis and its use for the isolation of diacetyl overproducing strains. Appl Environ Microbiol 62 : 2636 2640. Goupil Feuillerat, N., M. Cocaign Bousquet, J. J. Godon, S.D. Ehrlich & P. Renault, (1997) acetolactate decarboxylase in Lactococcus lactis subsp . lactis . J Bacteriol 179 : 6285 6293. Gundlach , L., B. Burfeindt, J. Mahrt & W. F., (2012) Dynamics of ultrafast photoinduced heterogeneous electron transfer, implications for recent solar energy conversion scenarios. Chem Phys Lett 545 : 35 39. Hanahan, D., (1983) Studies on transformation of Escheric hia coli with plasmids. J Mol Biol 166 : 557 580. Public Health 122 : 647 652. Harden, A. & G.S. Walpole, (1906) Chemical action of Bacillus lactis aerogenes (Escherich) production of 2:3 butyleneglycol and acetylmethylcarbinol on glucose and mannitol. Proc R Soc Lond B 77 .
79 Helmann, J.D., (1995) Compilation and analysis of Bacillus subtilis A dependent promoter sequences: evidence for extended contact betwe en RNA polymerase and upstream promoter DNA. Nucl Acids Res 23 : 2351 2360. Henkin, T.M., (2008) Riboswitch RNAs: using RNA to sense cellular metabolism. Genes Dev 22 : 3383 3390. Hobert, O., (2002) PCR fusion based approach to create reporter gene construct s for expression analysis in transgenic C. elegans . Biotechniques 32 : 728 730. Hoffmann, K.M., A.C. Deutschmann, M.J. Weitzer, E. Zechner, C. Hogenauer & A.C. Hauer, (2010) Antibiotic associated hemorrhagic colitis caused by cytotoxin producing Klebsiella oxytoca . Pediatrics 125 : 960 963. Hong, H.A., R. Khaneja, N.M.K. Tam, A. Cazzato, S. Tan, M. Urdaci, A. Brisson, A. Gasbarrini, I. Barnes & S.M. Cutting, (2009) Bacillus subtilis isolated from the human gastrointestinal tract. Res Microbiol 160 : 134 143. Hubbert, M.K., (1956) Nuclear energy and the fossil fuels. In: Drilling and Production Practice . American Petroleum Institute, pp. 7 25. Jantama, K., X. Zhang, J.C. Moore, K.T. Shanmugam, S.A. Svoronos & L.O. Ingram, (2008) Eliminating side products and in creasing succinate yields in engineered strains of Escherichia coli C. Biotechnol Bioeng 101 : 881 893. Jarmer, H., T.S. Larsen, A. Krogh, H.H. Saxild, S. Brunak & S. Knudsen, (2001) Sigma A recognition sites in the Bacillus subtilis genome. Microbiology 1 47 : 2417 2424. Jensen, K.F., (1993) The Escherichia coli K 12 "Wild Types" W3110 and MG165 5 h ave an rph frameshift mutation that leads to pyrimidine starvation d ue t o low pyrE expression levels. J Bacteriol 175 : 3401 3407. Jeon, S., D. Kim, H. Song, H.J. L ee, S. Park, D. Seung & Y.K. Chang, (2013) 2,3 Butanediol recovery from fermentation broth by alcohol precipitation and vacuum distillation. J Biosci Bioeng 117 : 464 470. Ji, X.J., H. Huang & P.K. Ouyang, (2011) Microbial 2,3 butanediol production: a state of the art review. Biotechnol Adv 29 : 351 364. Ji, X.J., H. Huang, J.G. Zhu, L.J. Ren, Z.K. Nie, J. Du & S. Li, (2010) Engineering Klebsiella oxytoca butanediol production through insertional inactivation of acetaldehyde dehydrogenase gen e. Appl Microbiol Biotechnol 85 : 1751 1758. Jina, C., M. Yaoc, H. Liuc, C.F. Leed & J. Jib, (2011) Progress in the production and application of n butanol as a biofuel. Renew Sust Energ Rev 15 : 4080 4106.
80 Joseph, P., J.R. Fantino, M.L. Herbaud & F. Denizot , (2001) Rapid orientated cloning in a shuttle vector allowing modulated gene expression in Bacillus subtilis . FEMS Microbiol Lett 205 : 91 97. Jung, S., J. Jang, A. Kim, M. Lim, B. Kim, J. Lee & Y. Kim, (2013) Removal of pathogenic factors from 2,3 butaned iol producing Klebsiella species by inactivating virulence related wabG gene. Appl Microbiol Biotechnol 97 : 1997 2007. Kaper, J.B., J.P. Nataro & H.L.T. Mobley, (2004) Pathogenic Escherichia coli . Nature Rev Microbiol 2 : 123 140. Kerr, R.A., (2011) Peak oil production may already be here. Science 331 : 1510 1511. Kim, S. J., S. O. Seo, Y. S. Jin & J. H. Seo, (2013) Production of 2,3 butanediol by engineered Saccharomyces cerevisiae . Bioresour Technol 146 : 274 281. Kim, Y., L.O. Ingr am & K.T. Shanmugam, (2007) Construction of an Escherichia coli K 12 mutant for homoethanologenic fermentation of glucose or xylose without foreign genes. Appl Environ Microbiol 73 : 1766 1771. Kim, Y., L.O. Ingram & K.T. Shanmugam, (2008) Dihydrolipoamide dehydrogenase mutation alters the NADH sensitivity of pyruvate dehydrogenase complex of Escherichia coli K 12. J Bacteriol 190 : 3851 3858. Kunst, F., N. Ogasawara, I. Moszer, A.M. Albertini, G. Alloni, V. Azevedo, M.G. Bertero, P. Bessieres, A. Bolotin, S. Borchert, R. Borriss, L. Boursier, A. Brans, M. Braun, S.C. Brignell, S. Bron, S. Brouillet, C.V. Bruschi, B. Caldwell, V. Capuano, N.M. Carter, S.K. Choi, J.J. Codani, I.F. Connerton, A. Danchin & et al., (1997) The complete genome sequence of the gram p ositive bacterium Bacillus subtilis . Nature 390 : 249 256. Lee, J.H., P. Patel, P. Sankar & K.T. Shanmugam, (1985) Isolation and characterization of mutant strains of Escherichia coli altered in H 2 metabolism. J Bacteriol 162 : 344 352. Leroux, P.J. & H.J. L ucas, (1951) L( ) 2 butanol from D( ) 2,3 butanediol. J Am Chem Soc 73 : 41 42. Li, L., Y. Wang, L. Zhang, C. Ma, A. Wang, F. Tao & P. Xu, (2012) Biocatalytic production of (2S,3S) 2,3 butanediol from diacetyl using whole cells of engineered Escherichia col i . Bioresour Technol 115 : 111 116. Li, Z. J., J. Jian, X. X. Wei, X. W. Shen & G. Q. Chen, (2010) Microbial production of meso 2,3 butanediol by metabolically engineered Escherichia coli under low oxygen condition. Appl Microbiol Biotechnol 87 : 2001 2009. Livak, K.J. & T.D. Schmittgen, (2001) Analysis of relative gene expression data using real time quantitative PCR and the 2( Delta Delta C(T)) Method. Methods 25 : 402 408.
81 Lumbreras, S. & A. Ramos, (2012) Offshore wind farm electrical design: a review. Wind Energy 10 : 1002 1498. Ma, C.Q., A.L. Wang, J.Y. Qin, L.X. Li, X.L. Ai, T.Y. Jiang, H. Tang & P. Xu, (2009) Enhanced 2,3 butanediol production by Klebsiella pneumoniae SDM. Appl Microbiol Biotechnol 82 : 49 57. Maddocks, S.E. & P.C. Oyston, (2008) Structure and function of the LysR type transcriptional regulator (LTTR) family proteins. Microbiology 154 : 3609 3623. Mahlen, S.D., (2011) Serratia infections: from military experiments to current practice. Clin Microbiol Rev 24 : 755 791. Marshall, A. & P.J. Alaimo, (2010) Useful products from complex starting materials: common chemicals from biomass feedstocks. Chem Eur J 16 : 4970 4980. Mathews, J.M., S.L. Watson, R.W. Snyder, J.P. Burgess & D.L. Morgan, (2010) Reaction of the butter flavorant diacetyl (2,3 butanedione) with N acetylarginine: a model for epitope formation with pulmonary proteins in the etiology of obliterative bronchiolitis. J Agric Food Chem 58 : 12761 12768. McFarlane, J. & S.M. Robinson, (2007) Survey of alternativ e feedstocks for commodity chemical manufacturing. In: (Oak Ridge National Laboratory, TN). O.R.N. Laborator (ed). Oak Ridge, Tennessee: U.S. Department of Energy, pp. 1 28. Meselson, M. & R. Yuan, (1968) DNA restriction enzyme from E. coli . Nature 217 : 11 10 1114. Miller, J.H., (1972) Experiments in Molecular Genetics . Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York. Moet, G.J., R.N. Jones, D.J. Biedenbach, M.G. Stilwell & T.R. Fritsche, (2007) Contemporary causes of skin and soft tissue i nfections in North America, Latin America, and Europe: report from the SENTRY Antimicrobial Surveillance Program (1998 2004). Ann Clin Microbiol Antimicrob 57 : 7 13. Moran, C.P., N. Lang, S.F. LeGrice, G. Lee, M. Stephens, A.L. Sonenshein, J. Pero & R. Los ick, (1982) Nucleotide sequences that signal the initiation of transcription and translation in Bacillus subtilis. Mol Gen Genet 186 : 339 346. Nature 481 : 433 435. Nakano, M.M., Y.P. Dailly, P. Zuber & D.P. Clark, (1997) Characterization of anaerobic fermentative growth of Bacillus subtilis genes required for growth. J Bacteriol 179 : 6749 6755. Nakano, M.M. & P. Zuber, (1998) Anaerobic growth of a "strict aero be" (Bacillu s subtilis). Annu Rev Microbiol 52 : 165 190.
82 Nicholson, W.L., (2008) The Bacillus subtilis ydjL ( bdhA ) gene encodes acetoin reductase/2,3 butanediol dehydrogenase. Appl Environ Microbiol 74 : 6832 6838. N icholson, W.L. & H. Maughan, (2002) The spectrum of spontaneous rifampin resistance mutations in the rpoB gene of Bacillus subtilis 168 spores differs from that of vegetative cells and resembles that of Mycobacterium tuberculosis. J Bacteriol 184 : 4936 494 0. Nicholson, W.L. & P. Setlow, (1990) Sporulation, germination, and outgrowth. In: Molecular biological methods for Bacillus. C.R. Harwood & S.M. Cutting (eds). New York: J. Wiley & Sons, pp. 391 450. Nielsen, D.R., S. H. Yoon, C.J. Yuan & K.L.J. Prather, (2010) Metabolic engineering of acetoin and meso 2,3 butanediol biosynthesis in E. coli . Biotechnol J 5 : 274 284. Oliveira, R.R. & W.L. Nicholson, (2013) The LysR type transcriptional regulator (LTTR) AlsR indirectly regulates expression of the Bacillus s ubtilis bdhA gene encoding 2,3 butanediol dehydrogenase. Appl Microbiol Biotechnol 97 : 7307 7316. Oliver, J.W.K., I.M.P. Machadoa, H. Yonedaa & S. Atsumia, (2013) Cyanobacterial conversion of carbon dioxide to 2,3 butanediol. Proc Natl Acad Sci 110 : 1249 1 254. Ong, E.B.B., T.S.T. Muhammad, M.R. Samian & N. Najimudin, (2011) Purification of the Bacillus subtilis regulatory protein AlsR and the localization of its binding site. Asia Pac J Mol Biol Biotechnol 19 : 29 37. Otero, J.M., G. Panagiotou & L. Olsson, (2007) Fueling industrial biotechnology growth with bioethanol. Adv Biochem Eng Biotechnol 108 : 1 40. Palsson, B.O., S. Fathi Afshar, D.F. Rudd & E.N. Lightfoot, (1981) Biomass as a source of chemical feedstocks: an economic evaluation. Science 213 : 513 517. Pauly, M. & K. Keegstra, (2010) Plant cell wall polymers as precursors for biofuels. Curr Opin Plant Biol 13 : 305 312. Podschun, R. & U. Ullmann, (1998) Klebsiella spp. as nosocomial pathogens: epidemiology, taxonomy, typing methods, and pathogeni city factors. Clin Microbiol Rev 11 : 589 603. Prescott, S.C. & C.G. Dunn, (1949) The production and properties of 2,3 butanediol. Industrial Microbiology 22 : 487 523. Qin, J., W. Gao, Q. Li, Y. Li, F. Zheng, C. Liu & G. Gu, (2010) Improvement of thermostab ility of beta 1,3 1,4 glucanase from Bacillus amyloliquefaciens BS5582 through in vitro evolution. Chinese J. Biotechnol. 26 : 1293 1301. Qin, J.Y., Z.J. Xiao, C.Q. Ma, N.Z. Xie, P.H. Liu & P. Xu, (2006) Production of 2,3 butanediol by Klebsiella pneumoniae using glucose and ammonium phosphate. Chinese J Chem Eng 14 : 132 136.
83 Qureshi, N. & M. Cheryan, (1989) Production of 2,3 butanediol by Klebsiella oxytoca . Appl Microbiol Biotechnol 30 : 440 443. Rabaey, K., N. Boon, S.D. Siciliano, M. Verhaege & W. Verstraete, (2004) Biofuel cells select for microbial consortia that self mediate electron transfer. Environ Microbiol 70 : 5373 5382. Rastogi, G., A. Bhalla, A. Adhikari, K.M. Bischoff, S.R. Hughes, L.P. Christopher & R.K. Sani, (2010) Characterizatio n of thermostable cellulases produced by Bacillus and Geobacillus strains. Bioresour Technol 101 : 8798 8806. Renna, M.C., N. Najimudin, L.R. Winik & S.A. Zahler, (1993) Regulation of the Bacillus subtilis alsS, alsD , and alsR genes involved in post expone ntial phase production of acetoin. J Bacteriol 175 : 3863 3875. Roberts, I., (2008) The economics of tackling climate change. BMJ 336 : 165 166. Rourke, F.O., F. Boyle & A. Reynolds, (2010) Tidal energy update 2009. Appl Energy 87 : 398 409. Sambrook, J. & D.W. Russell, (2001) Molecular cloning: a laboratory manual . Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Sanders, W.E. & C.C. Sanders, (1997) Enterobacter spp.: pathogens poised to flourish at the turn of the century. Clin Microbiol Rev 1 0 : 220 241. Schallmey, M., A. Singh & O.P. Ward, (2004) Developments in the use of Bacillus species for industrial production. Can J Microbiol 50 : 1 17. Schell, M.A., (1993) Molecular biology of the LysR family of transcriptional regulators. Annu Rev Micro biol 47 : 597 626. Serganov, A. & E. Nudler, (2013) A decade of riboswitches. Cell 152 : 17 24. Sorrell, S., J. Speirs, R. Bentley, A. Brandt & R. Miller, (2010) Global oil depletion: a review of the evidence. Energy Policy 38 : 5290 5295. Speranza, G., S. Co rti, G. Fontana & P. Manitto, (1997) Conversion of meso 2,3 butanediol into 2 butanol by Lactobacilli . Stereochemical and enzymatic aspects. J Agric Food Chem 45 : 3476 3480. Speranza, G., P. Manitto, G. Fontana, D. Monti & A. Galli, (1996) Evidence for ena ntiomorphic enantiotopic group discrimination in diol dehydratase catalyzed dehydration of meso 2,3 butanediol. Tetrahedron Letters 37 : 4247 4250. Spizizen, J., (1958) Transformation of biochemically deficient strains of Bacillus subtilis by deoxyribonuc leate. Proc. Natl. Acad. Sci. USA 44 : 1072 1078.
84 St John, F.J., J.D. Rice & J.F. Preston, (2006) Paenibacillus sp. strain JDR 2 and XynA1: a novel system for methylglucuronoxylan utilization. Appl Environ Microbiol 72 : 1496 1506. Steubing, B., R. Zah & C. Ludwig, (2011) Life cycle assessment of SNG from wood for heating, electricity, and transportation. Biomass Bioenergy 35 : 2950 2960. Syu, M. J., (2001) Biological production of 2,3 butanediol. Appl Microbiol Biotechnol 55 : 10 18. Ui, S., M. Odagiri, A. Mimura, H. Kanai, T. Kobayashi & T. Kudo, (1996) Preparation of a chiral acetoinic compound using transgenic Escherichia coli expressing the 2,3 butanediol dehydrogenase gene. J Biosci Bioeng 81 : 386 389. Ui, S., Y. Okajima, A. Mimur a, H. Kanai & T. Kudo, (1997) Molecular generation of an Escherichia coli strain producing only the meso isomer of 2,3 butanediol. J Ferment Bioeng 84 : 185 189. Ui, S., Y. Takusagawa, T. Sato, T. Ohtsuki, A. Mimura, M. Ohkuma & T. Kudo, (2004) Production o f L 2,3 butanediol by a new pathway constructed in Escherichia coli . Lett Appl Microbiol 39 : 533 537. Underwood, S.A., S. Zhou, T.B. Causey, L.P. Yomano, K.T. Shanmugam & L.O. Ingram, (2002) Genetic changes to optimize carbon partitioning between ethanol a nd biosynthesis in ethanologenic Escherichia coli . Appl Environ Microbiol 68 : 6263 6272. Vagner, V., E. Dervyn & S.D. Ehrlich, (1998) A vector for systematic gene inactivation in Bacillus subtilis. Microbiology 144 : 3097 3104. Van Dien, S., (2013) From the first drop to the first truckload: commercialization of microbial processes for renewable chemicals. Curr Opin Biotechnol 24 : 1061 1068. Villa, D.C., S. Angioni, E. Quartarone, P.P. Righetti & P. Mustarelli, (2013) New sulfonated PBIs for PEMFC applicatio n. Fuel Cells 13 : 98 103. Voloch, M., N.B. Jansen, M.R. Ladisch, G.T. Tsao, R. Narayan & V.W. Rodwell, (1985) 2,3 Butanediol. p. 933 947. Wang, Y., L. Li, C. Ma, C. Gao, F. Tao & P. Xu, (2013) Engineering of cofactor regeneration enhances (2S,3S) 2,3 butan ediol production from diacetyl. Scientific Report 3 : 2643. Willey, J., L. Sherwood & C. Woolverton, (2008) Prescott's Principles of Microbiology, p. 960. McGraw Hill Science/Engineering/Math. Winfield, M.E., (1945) The catalytic dehydration of 2,3 butanedi ol to 1,3 butadiene. J Counc Sci Ind Res 18 : 412 413.
85 Xiao, Z. & J.R. Lu, (2014) Strategies for enhancing fermentative production of acetoin: a review. Biotech Adv 32 : 492 503. Xu, Y., H. Chu, C. Gao, F. Tao, Z. Zhou, K. Li, L. Li, C. Ma & P. Xu, (2014) Systematic metabolic engineering of Escherichia coli for high yield production of fuel bio chemical 2,3 butanediol. Metab Eng 23 : 22 33. Yan, Y., C. Lee & J.C. Liao, (2009) Enantioselective synthesis of pure (R,R) 2,3 butanediol in Escherichia coli with stereospecific secondary alcohol dehydrogenases. Org Biomol Chem 7 : 3914 3917. Yankeelov, J.A., Jr., C.D. Mitchell & T.H. Crawford, (1968) A simple trimerization of 2,3 butanedione yielding a selective reagent for the modification o f arginine in proteins. J Am Chem Soc 90 : 1664 1666. Youngquist, W., (1999) The post petroleum paradigm -and population. Popul Environ 20 : 297 315. Zeng, A.P., H. Biebl & W.D. Deckwer, (1991) Production of 2,3 butanediol in a membrane bioreactor with cell recycle. Appl Microbiol Biotechnol 34 : 463 468. Zhang, L., J. Sun, Y. Hao, J. Zhu, J. Chu, D. Wei & Y. Shen, (2010a) Microbial production of 2,3 butanediol by a surfactant (serrawettin) Serratia marcescens H30. J Ind Microbiol Biotechnol 37 : 857 862. Zhang, L., Y. Yang, J. Sun, Y. Shen, D. Wei, J. Zhu & J. Chu, (2010b) Microbial production of 2,3 butanediol by a mutagenized strain of Serratia marcescens H30. Bioresour Technol 101 : 1961 1967. Zhang, X., T. Bao, Z. Rao, T. Yang, Z. Xu, S. Y ang & H. Li, (2014) Two stage pH control strategy based on the pH preference of acetoin reductase regulates acetoin and 2,3 butanediol distribution in Bacillus subtilis . PLoS One 9 : e91187. Zhang, X., R. Zhang, T. Bao, T. Yang, M. Xu, H. Li, Z. Xu & Z. Rao , (2013) Moderate expression of the transcriptional regulator AlsR enhances acetoin production by Bacillus subtilis . J Ind Microbiol Biotechnol 40 . Zhao, L., Y. Bao, J. Wang, B. Liu & L. An, (2009) Optimization and mechanism of diacetyl accumulation by En terobacter aerogenes mutant UV 3. World J Microbiol Biotechnol 25 : 57 64.
86 BIOGRAPHICAL SKETCH Rafael Rodrigues de Oliveira was born in Rio Grande Rio Grande do Sul, Brazil, in 1982. He completed his B.Sc. degree in Biological Sciences with a minor in Molecular, Cellular and Functional Biology from the Federal University of Rio Grande do Sul (UFRGS) , Porto Alegre/RS, Brazil, in 2007. Rafael completed his M.Sc. degree in Genet ics and Molecular Biology in 2010, where his thesis focused on studying a family of soybean cell wall proteins (HyPRPs) responsive to biotic and abiotic stresses. After studying plant cell walls, a promising renewable carbon source for fermentation process , he decided to examine how synthetic biology could be used for engineering of metabolic pathways for producing bio based chemicals from renewable feed stock . As his model system, Rafael has been concentrating on the pathway for 2,3 butanediol synthesis in B. subtilis and E. coli. His Ph.D. was conducted under the direction of Dr . Wayne Nicholson at the University of Florida/ Space Life Sciences Laboratory at Kennedy Space Center NASA . As a Ph.D. student, he has presented his research at the 97 th Southeaster n Branch of the American Society of Microbiology Annual Meeting (2011), and at the 114 th General Meeting of the American Society for Microbiology (2014).
Incorporationofextracellularfattyacidsbyafattyacid kinase-dependentpathwayin Staphylococcusaureus JoshuaB.Parsons,MatthewW.Frank, PamelaJackson,ChitraSubramanianand CharlesO.Rock* DepartmentofInfectiousDiseases , St.JudeChildren's ResearchHospital , 262DannyThomasPlace, Memphis,TN38105,USA . Summary Acyl-CoAandacyl-acylcarrierprotein(ACP)synthetasesactivateexogenousfattyacidsforincorporationintophospholipidsinGram-negativebacteria. However,Gram-positivebacteriautilizeanacyltransferasepathwayforthebiogenesisofphosphatidic acidthatbeginswiththeacylationof sn -glycerol-3phosphatebyPlsYusinganacyl-phosphate(acylPO 4 )intermediate.PlsXgeneratesacyl-PO 4 fromthe acyl-ACPend-productsoffattyacidsynthesis.The plsX geneof Staphylococcusaureus wasinactivated andtheresultingstrainwasbothafattyacidauxotrophandrequired denovo fattyacidsynthesisfor growth.Exogenousfattyacidswereonlyincorporatedintothe1-positionandendogenousacylgroups werechanneledintothe2-positionofthephospholipidsinstrainPDJ39( ! plsX ).Extracellularfattyacids werenotelongated.Removaloftheexogenousfatty acidsupplementledtotherapidaccumulationof intracellularacyl-ACPandtheabruptcessationof fattyacidsynthesis.Extractsfromthe ! plsX strain exhibitedanATP-dependentfattyacidkinaseactivity, andtheacyl-PO 4 wasconvertedtoacyl-ACPwhen puriÃžedPlsXisadded.Thesedatarevealtheexistenceofanovelfattyacidkinasepathwayforthe incorporationofexogenousfattyacidsinto S.aureus phospholipids. Introduction Phosphatidicacid(PtdOH)isauniversalintermediatein thebiosynthesisofmembranephospholipidsineubacteria (YaoandRock,2013).Inthe Escherichiacoli model, PtdOHformationisinitiatedbytheacylationof sn -glycerol3-phosphate(G3P)byPlsB,aG3Pacyltransferase.PlsB of E.coli hasbeenextensivelycharacterizedandutilizes eitheracyl-acylcarrierprotein(ACP)oracyl-CoAthioesterstoacylatethe1-positionofG3P(Lightner etal ., 1980;Green etal .,1981;Rock etal .,1981a).PlsBis responsiblefortheselectionoffattyacidsincorporatedinto the1-positionofmembranephospholipidsandisakey regulatorypointinthepathway(Rock etal .,1981b;Heath etal .,1994;CronanJrandRock,1996;YaoandRock, 2013).However,mostbacteria,includingimportanthuman Gram-positivepathogenssuchas Streptococcuspneumoniae and Staphylococcusaureus ,lacka plsB geneand insteadusethePlsX/PlsYpathwayfortheacylationofG3P (Lu etal .,2006)(Fig.1).Inthesesystems,thesolublePlsX isanacyl-ACP:PO 4 transacylasethatconvertstheacylACPend-productsof denovo fattyacidsynthesistotheir acyl-PO 4 derivatives.Theseactivatedfattyacidsarethen usedbytheintegralmembraneproteinPlsYtoacylate G3P(Lu etal .,2006;2007).Asecondacyltransferase, PlsC(Coleman,1992),isuniversallyexpressedinbacteria andcompletesthesynthesisofPtdOHbytransferringa fattyacidtothe2-positionofacyl-G3P.The E.coli PlsC useseitheracyl-ACPoracyl-CoAasacyldonors,butthe Gram-positivePlsCsuseonlyacyl-ACPassubstrate(Lu etal .,2006;YaoandRock,2013). Exogenousfattyacidscanaccesstheseacyltransferase systemsaftertheiruptakebythecellandactivation.In E.coli ,thepathwayinvolvestheconversionofthefatty acidtoanacyl-CoAthioesterbyFadD,anacyl-CoAsynthetase(YaoandRock,2013). E.coli acyltransferasesuse acyl-CoAsassubstrates,butthisorganismcannotconvert fattyacidsoracyl-CoAstoacyl-ACP.Thus,thereisno elongationofexogenousfattyacidsbyFASIIorincorporationintolipopolysaccharideviatheacyl-ACP-speciÃžc acyltransferases.However,somebacteriapossessan acyl-ACPsynthetasethatligatesfattyacidstoACP.The expressionofthisenzymeallowsexogenousfattyacidsto notonlybeusedbytheacyltransferases,butalsotoenter thefattyacidbiosyntheticpathwayandbeelongated (Jiang etal .,2006;2010).Ourrecentworkwith Staphylococcusaureus showsthatexogenousfattyacidsareboth usedbytheacyltransferasesandelongatedbyFASII showingthattheyareconvertedtoacyl-ACP(Parsons etal .,2011).Experimentswithcrudeextractsshowedthe Accepted17February,2014.*Forcorrespondence.E-mailcharles .firstname.lastname@example.org;Tel.( + 1)9015953491;Fax( + 1)9015953099. MolecularMicrobiology (2014) 92 (2),234Ã245 ! doi:10.1111/mmi.12556 Firstpublishedonline11March2014 Â©2014JohnWiley&SonsLtd
formationofacyl-ACPfromalabelledfattyacid,ATPand Mg 2 + ,butthegeneencodingtheputativeacyl-ACPsynthetasewasnotidentiÃžed. Thegoalofthisstudyistodetermineifexogenousfatty acidsaretakenupin S.aureus viaanacyl-ACPoracylPO 4 dependentpathway(Fig.1).Thekeytodistinguishingthesetwopossibilitiesisthephenotypeofa plsX deletionmutant.Iffattyacidsareactivatedtoacyl-ACPs, thena ! plsX strainwouldbenon-viablebecausethey wouldbeunabletosynthesizeacyl-PO 4 fromeitherexogenousfattyacidsorFASII.However,ifafattyacidkinase pathwayisoperating,a ! plsX strainwouldbeableto produceacyl-PO 4 fromexogenousfattyacidsforPlsYand obtainacyl-ACPforPlsCfrom denovo fattyacidsynthesis.PlsXisconsideredanessentialgenebasedonthe growthphenotypeofa Bacillussubtilis strainconditionallydeÃžcientin plsX expression(Paoletti etal .,2007). However,themediausedintheseexperimentswasnot supplementedwithfattyacids.WeÃžnd Staphylococcus aureus ! plsX strainsarebothfattyacidauxotrophsand requiredenovofattyacidssynthesisforgrowth.ExogenousfattyacidsarenotelongatedinPlsX-null S.aureus , andcellextractsexhibitfattyacidkinaseactivity.Thus,the incorporationofhostfattyacidsintothephospholipids ofGram-positivepathogensinvolvesanovelfattyacid kinasetoactivateexogenousfattyacidsforincorporation intophospholipids(Fig.1). Results Aspreviouslyreported(Parsons etal .,2011),wefound 100 " Moleateinhibited[ 14 C]acetateincorporationinto S.aureus strainRN4220(Fig.2A).Oleatewasusedin mostofourexperimentsastheprototypicalexogenous fattyacidbecauseitsincorporationandmetabolismwas easilyseparatedfromtheFASII-producedsaturated branched-chainfattyacids,andoleateisanabundanthost fattyacid.ThereductioninFASIIactivitywasaccompanied byincorporationoftheexogenousoleate(18:1 ! 9),andits elongationproduct20:1 ! 11,intophospholipidsreplacing thenormalbranched-chainsaturatedfattyacids. S.aureus lacksthecapacityfor # -oxidationandtherewereno [ 14 C]acetate-labelledfattyacidsdetectedinthemediumin theseexperiments.Theregulatoryeffectofexogenous fattyacidwasnotauniquepropertyofoleate(Fig.2A). Lauricacid,amixtureof anteiso 15:0/17:0,or17:0also reducedtheamountof[ 14 C]acetatelabelling.Neitherexogenousshort-chainfattyacids(octanoicacid)norvery-long chainfattyacids(19:0and21:0)reducedtherateof acetateincorporation,suggestingthemosthighlyincorporatedfattyacidswereacylchainsthatwerenormallyfound inphospholipids.Thesedatasuggestedthat S.aureus wasabletosparesomeenergyexpendedinfattyacid synthesiswhenexogenousfattyacidsareavailable. In E.coli ,acyl-CoAsynthetase(FadD)playsakeyrolein theincorporationofexogenousfattyacidsbygenerating theacyl-CoAsubstratesforthePlsB/PlsCacyltransferases (DiRusso etal .,1999).Therewasnoapparentrolefor acyl-CoAinphospholipidsynthesisinbacteriathatemploy theacyl-PO 4 /acyl-ACP-dependentPlsX/PlsY/PlsCacyltransferasesystem(Lu etal .,2006).Wedidnotdetect acyl-CoAsynthetaseactivityin S.aureus extracts(not shown).Thereweretwoopenreadingframesinthe S.aureus genomeannotatedaspotentialacyl-CoAsynthetases,SA0226andSA0533,whichhave23%and28% identitywithFadDrespectively.Thesetwogeneswere inactivatedandtheresultingstrainsdidnothaveadeÃžciencyinexogenousfattyacidincorporationintophospholipidsasassessedby[ 14 C]oleatelabellingcomparedwith theparentstrainRN4220(notshown).Oleoyl-CoAwasnot Fig.1. Fattyacidmetabolismin S.aureus .Phosphatidicacid(PtdOH)issynthesizedbythestepwiseacylationof sn -glycerol-3-phosphateby PlsYthattransfersafattyacidtothe1-positionfromacyl-phosphate(acyl-PO 4 )followedbyPlsCthatacylatesthe2-positionusingacyl-ACP. Acyl-ACPisproducedbythetypeIIfattyacidbiosyntheticpathway(FASII),andPlsXcatalysestheinter-conversionofacyl-ACPand acyl-PO 4 .Exogenousfattyacidsenterthecellandmaybeactivatedbyeitherafattyacidkinaseoracyl-ACPsynthetasepathway.Our experimentsshowstrainPDJ39( ! plsX )requiresbothexogenousfattyacidsand denovo fattyacidsynthesisforgrowthandhasfattyacid kinaseactivityincellextracts.Thesedataruleoutthepresenceofanacyl-ACPsynthetase(Aas,red)andshowthepresenceofafattyacid kinase(Fak,green)pathwayfortheincorporationofexogenousfattyacidsintophospholipids.Theacyl-PO 4 mayeitherbeusedbyPlsYor convertedbyPlsXtoacyl-ACPthatcaneitherbeelongatedorutilizedbyPlsC.Thegene(s)encodingFakareunknown. Fattyacidmetabolismin S.aureus 235 Â©2014JohnWiley&SonsLtd, MolecularMicrobiology , 92 ,234Ã245
Fig.2. Long-chainacyl-CoAdoesnothavearolein S.aureus fattyaciduptake. A.Inhibitionof[ 14 C]acetateincorporationintofattyacidinthepresenceofexogenousfattyacids.Wild-typestrainRN4220wasgrowntoan A 600 = 0.5.EachfattyacidwasaddedtoLBmediumcontaining0.1%Brij-58toaÃžnalconcentrationof100 " M,and30minlater[ 14 C]acetate wasadded.Attheendofa30minlabellingperiod,thecellswereharvested,extractedandtheamountofradioactivityinthelipidfraction determined.Thedatawerenormalizedtotheradiolabelintheuntreated,wild-typestrainsetto100%(140000dpmper7 $ 10 8 cells).Fatty acidsareabbreviatedasnumberofcarbonatoms:numberofdoublebonds.TheÃ”a'designationreferstoan anteiso branched-chainfatty acid. B.Expressionofacyl-CoAsynthetaseimpairs S.aureus growth.The E.colifadD genewasclonedintotheregulatedexpressionplasmid pG164(emptyvectorcontrol)togeneratepFadD,andtransformedintowild-typestrainSA178R1.Strainsweregrowninthepresenceof1mM IPTGtoinduce fadD expression,and18:1wasaddedataconcentrationof200 " Minearlylogphaseasindicatedbythearrow. C.MassspectrometryanalysisoftheCoAthioesterpoolin18:1 ! 9-treatedcells.StrainSA178R1harbouredeitherpG164,theemptycontrol expressionplasmid,orpFadD,apG164derivativeexpressing E.coli acyl-CoAsynthetase(FadD).Bothstrainsweregrowninthepresenceof 1mMIPTGeitherinthepresenceorabsenceof0.2mMoleate.AtanA 600 of1.0,thecellswereharvestedandprocessedformass spectrometryasdescribed(Parsons etal .,2011).Propionyl-CoAwasaddedasaninternalstandardandthesignalsforCoA,acetyl-CoAand oleoyl-CoAwerenormalizedtothepropionyl-CoAsignal. 236 J.B.Parsons etal . ! Â©2014JohnWiley&SonsLtd, MolecularMicrobiology , 92 ,234Ã245
detectedin S.aureus stainRN4220growninthepresence ofexogenousoleateusingESI-MS/MS(notshown). Tovalidatethesenegativeresults,weinsertedthe E.colifadD geneintoplasmidpG164,whichtogetherwith strainSA178R1,constitutedatightlyregulated S.aureus expressionsystem(D'Elia etal .,2006).StrainSA178R1 wasderivedfromstrainRN4220andhastheT7polymeraseandtheLacI q repressorgenesintegratedintothe chromosomeunderthecontrolofthe Pspac promoter/ operatorallowingIPTG-dependentexpressionfromthe pG164plasmid.Theadditionof200 " Moleicaciddidnot affectthegrowthofstrainSA178R1/pG164eitherinthe presenceorabsenceofinducer(Fig.2B).When fadD expressionwasinducedbyIPTGinstrainSA178R1/ pJP106( fadD ),thepresenceofoleicacidcausedasigniÃžcantdecreaseingrowthrate,andtherewasaslight effectongrowthintheabsenceofinduction(Fig.2B). Thesedatasuggestedthattheformationofoleoyl-CoAin thestrainexpressingFadDhadadeleteriouseffect.We havenotexploredthebasisforthegrowthinhibitionby FadDexpression,butitisknownthatacyl-CoAsare potentinhibitorsoftheacyl-PO 4 -dependentPlsYacyltransferase(Lu etal .,2006),andtheymayhaveinhibitory effectsonotherenzymes.Oleoyl-CoAwasnotdetected inIPTG-treatedstrainSA178R1/pG164exposedtooleic acid(Fig.2C),consistentwithourpreliminaryexperimentswithstrainRN4220.Incontrast,therewasasigniÃžcantaccumulationofoleoyl-CoAdetectedinextracts fromstrainSA178R1/pJP106expressingFadD.These dataruledoutthepresenceofanunknownproteinthat convertsexogenousfattyacidstoacyl-CoAin S.aureus , andwereconsistentwiththeconclusionthatlong-chain acyl-CoAhasnorolein S.aureus fattyacidmetabolism. StrainslackingplsXwerefattyacidauxotrophs TheÃžrstexperimentsuggestingthat plsX knockoutstrains werefattyacidauxotrophscamefromtheanalysisofthe growthof B.subtilis strainLP39inthepresenceand absenceoffattyacids.Thisstrainhasachromosomal plsX knockoutcoveredbyanintegratedxylose-inducible plsX gene(Paoletti etal .,2007).Whendeprivedofthe xyloseinducer,growtharrestoccurredwhenPlsXlevels weredepletedbelowthethresholdrequiredforgrowth (Paoletti etal .,2007).Wefoundthatthesupplementation ofstrainLP39withamixtureofbranched-chainfattyacids resultedinthecontinuedgrowthofthePlsX-depleted strain(Fig.3A).Thesedatasuggestedthatexogenous fattyacidswereabletoovercomeablockadeatthePlsX step,whichappearedtoruleoutanacyl-ACPsynthetase mechanismforfattyacidincorporationintophospholipid. The B.subtilis geneticsystemwascomplexanddidnot allowforthestudyofstrainsthatwerecompletelydevoid ofPlsX.Therefore,weconstructed S.aureus strain PDJ39( ! plsX )thatinactivatedthe plsX genebytheinsertionofagroupIIintronataminoacid122ofthePlsX sequenceintotheparentstrainSA178R1(Fig.3B, inset ). Becauseweanticipatedthe ! plsX straintobeafatty acidauxotroph,theselectionstepswereperformedinthe presenceofafattyacidsupplementconsistingofa mixtureof anteiso 15:0/17:0fattyacids(Parsons etal ., 2011).StainPDJ39( ! plsX )wasindeedastrictfattyacid auxotroph(Fig.3B).Complementationofthefattyaciddependentgrowthphenotypeof ! plsX strainwithplasmid pPlsXexpressingthe plsX geneshowedthatthatthe deÃžciencyinPlsXwasbothnecessaryandsufficientfor thegrowthphenotype(Fig.3C).AFN-1252speciÃžcally inhibitstheenoyl-ACPreductase(FabI)stepintheelongationcycleofFASII(Kaplan etal .,2012).Althoughstrain PDJ39( ! plsX )wasafattyacidauxotroph(Fig.3B),the strainremainedassensitivetoAFN-1252asthewild-type parentstrainSA178R1(Fig.3D).StrainPDJ38( ! accD ) cannotproducemalonyl-CoA,doesnotcarryoutFASII, and ! accD strainswerecharacterizedasafattyacidand lipoateauxotrophs(Parsons etal .,2011;2013).Because FASIIwasnotoperationalinstrainPDJ38,AFN-1252 hadnoeffectonitsgrowth.Thus,strainPDJ39( ! plsX ) requiredbothendogenousfattyacidsynthesisfromFASII andanexogenousfattyacidsupplementforgrowth. Fattyacidcompositionandphospholipidstructurein strainPDJ39( ! plsX) Ourpreviousstudyonexogenousfattyacidmetabolismin S.aureus showedthatexogenous18:1 ! 9waselongated to20:1 ! 11byFASII,andboth18:1and20:1wereincorporatedintothe1-positionofPtdGro(Parsons etal .,2011). Thesedatashowedthatexogenousfattyacidswereconvertedtoacyl-ACPforelongationbyFASIIandtoacyl-PO 4 forutilizationbyPlsY.Phosphatidylglycerol(PtdGro)was themostabundantphospholipidin S.aureus ,andthe predominantPtdGromolecularspeciesinwild-typecells consistedof17:0inthe1-positionand15:0inthe2-position (Fig.4A).ThemoststrikingaspectPtdGrostructurein S.aureus wasthat15:0wasalmostexclusivelyfoundin the2-positionofPtdGro(Parsons etal .,2011),illustrating thehighselectivityofPlsCforthe15-carbonacyl-ACP.In wild-typestrainSA178R1,thereweretwomajorPtdGro molecularspeciesdetectedincellsgrownwith18:1 ! 9 correspondingto18:1/15:0and20:1/15:0(Fig.4B). However,instrainPDJ39( ! plsX )onlyasingleprominent PtdGromolecularspecieswasdetectedconsistingof18:1 pairedwith15:0(Fig.4C).About50%ofthe18:1was elongatedbyFASIIbeforeincorporationinPtdGrointhe wild-typestrain,buttherewasnoelongationof18:1 ! 9to 20:1 ! 11inPlsX-nullcells.Oneaspectoftheseexperimentswasthat18:1wasnormallychanneledtothe 1-positioninwild-typecellsduetothesubstratespeciÃžcity Fattyacidmetabolismin S.aureus 237 Â©2014JohnWiley&SonsLtd, MolecularMicrobiology , 92 ,234Ã245
ofPlsC(Parsons etal .,2011);however,incorporationof 15:0intothe1-positioncanoccurasevidencedfromthe existenceofthe30-carbonPtdGromolecularspeciesin strainSA178R1(Fig.4A).Tocorroboratetheconclusion thatexogenousfattyacidswerefunneledtothe1-position in ! plsXs train,itwasgrownwithan anteiso 15:0supplementandthePtdGrocompositiondeterminedbymass spectrometry.Onlyasinglemolecularspeciesconsisting 15:0/15:0PtdGrowasdetectedinstrainPDJ39( ! plsX ) conÃžrmingexogenousfattyacidswerechanneledintothe 1-positionofthephospholipidsandwerenotelongated (Fig.4D).Thesedataruleouttheexistenceofanacyl-ACP Fig.3. GrowthphenotypesofPlsX-nullstrains. A. B.subtilis strainLP39wasconstructedtoallowconditionalexpressionofPlsXfromaxylose-dependentpromoter(Paoletti etal .,2007).In theabsenceofxylose,growtharrestoccurs,whichwasattenuatedbytheadditionofexogenousfattyacids[FA;a1mMmixtureof anteiso branched-chain15:0/17:0in10mgml % 1 fattyacidfreebovineserumalbuminpluslipoate(0.1 " gml % 1 )]. B. S.aureus strainPDJ39( ! plsX ),aderivativeofthewild-typestrainSA178R1,wasafattyacidauxotroph.ThestrainsweregrowninLB mediumplus10mgml % 1 fattyacidfreebovineserumalbumineitherwithorwithouta1mMoleatesupplement.Theinsetshowsamultiplex PCRgenotypingresultforthedisrupted plsX geneinthe ! plsX straincomparedwiththewild-typegeneinstrainSA178R1.Thewild-type allelegivesa509bpbandandthemutant plsX allelegivesa305bpproduct. C.ThestrainPDJ39( ! plsX )growthphenotypewascomplementedbythepresenceofpPlsX,aplasmidderivedfrompG164thatexpresses the plsX gene.ThecontrolplasmidwastheemptypG164,andbothstrainsweregrownwithoutfattyacids. D.MinimuminhibitoryconcentrationforAFN-1252instrainsSA178R1(wild-type),PDJ39( ! plsX )andPDJ38( ! accD ).AFN-1252isapotent inhibitorof S.aureus enoyl-ACPreductaseandeffectivelyblocks denovo fattyacidsynthesis(Kaplan etal .,2012).Allstrainsweregrownon LBmediumwith S.aureus branched-chainfattyacids(1mMofamixtureof anteiso 15:0/17:0in10mgml % 1 fattyacidfreebovineserum albumin).ControlstrainPDJ38( ! accD )wasafattyacidauxotrophthatcannotmakeanyendogenousfattyacids,dependsonexogenousfatty acidsforacyl-PO 4 andacyl-ACPformation,andwascompletelyrefractorytoAFN-1252growthinhibition(Parsons etal .,2011;2013). 238 J.B.Parsons etal . ! Â©2014JohnWiley&SonsLtd, MolecularMicrobiology , 92 ,234Ã245
synthetase-dependentpathwayforfattyacidmetabolism andsuggestedtheexistenceofafattyacidkinasethat convertedfattyacidstoacyl-PO 4 ,whichweresubsequentlyeitherconvertedtoacyl-ACPbyPlsXforelongation byFASIIorusedforacylationofthe1-positionofG3Pby PlsY. Thechannelingofexogenousfattyacidstothe1positionandendogenouslysynthesizedfattyacidstothe 2-positionwascorroboratedwithmetaboliclabelling experimentsfollowedbythedigestionofthelabelled PtdGrowithphospholipaseA 2 todeterminethepositional distrubtionofthelabel(Fig.5).StrainsSA178R1and PDJ39( ! plsX )werelabelledwith[ 14 C]acetateandthe [ 14 C]PtdGroisolatedbythin-layerchromatography.As expected,theacetatelabelwasdistributedbetweenthe1and2-positionsinthewild-typestrainSA178R1(Fig.5A). However,[ 14 C]acetatewasonlyincorporatedintothe 2-positionofthePtdGroisolatedfromthe ! plsX strain (Fig.5B).ThesedatashowthatFASIIonlyprovidedacyl chainsforthe2-positionacyltransferase(PlsC)inthe Fig.4. Phosphatidylglycerol(PtdGro)molecularspeciessynthesizedfromexogenousfattyacidsinwild-typestrainSA178R1andits ! plsX derivative,strainPDJ39( ! plsX ). A.PtdGromolecularspeciessynthesizedbywild-type S.aureus strainSA178R1growninLBplus10mgml % 1 fattyacidfreebovineserum mediumwithoutafattyacidsupplement.Thespeciesarelabelledwiththe1-positionacylchainoverthe2-positionacylchain.Themost abundantPtdGromolecularspeciesconsistedof17:0inthe1-positionand15:0inthe2-position.TheseidentiÃžcationswerebasedonamore detailedanalysisofPtdGrostructure(Parsons etal .,2011),whichshowedthat15-carbonfattyacidswerealmostexclusivelylocalizedtothe 2-positionof S.aureus PtdGro. B.Wild-typestrainSA178R1bothincorporated18:1 ! 9intophospholipidandelongateditto20:1 ! 11.Thisresultedintwoprominentmolecular speciescontaininganunsaturatedfattyacid(either18:1or20:1)derivedfromthemediumanda15-carbonbranched-chainfattyacidfrom de novo biosynthesis. C.StrainPDJ39( ! plsX )didnotelongate18:1 ! 9to20:1 ! 11resultinginasinglemolecularspeciesofPtdGroconsistingof18:1anda 15-carbonbranched-chainfattyacid. D.PtdGromolecularspeciesderivedfromstrainPDJ39( ! plsX )growninthepresenceof anteiso 15:0(0.25mMin10mgml % 1 bovineserum medium).OnlyasinglePtdGromolecularspecieswasdetectedconsistingoftwo15:0fattyacids. Fattyacidmetabolismin S.aureus 239 Â©2014JohnWiley&SonsLtd, MolecularMicrobiology , 92 ,234Ã245
PlsX-nullstrain.Takentogether,thedatainthissection showedthat ! plsX cellsrequiredexogenousfattyacidsto acylatethe1-positionofG3PandFASIItoacylatethe 2-position. AnalysisofFASIIactivityinPlsX-nullstrains WenextexaminedFASIIactivityinstrainPDJ39( ! plsX ) followingtheremovaloftheexogenousfattyacidsupplement.Theremovaloftherequiredfattyacidsupplement fromthe ! plsX straininearlylogarithmicgrowthledtothe cessationofgrowth(Fig.6A).Initially,therewaslittledifferencebetweenthegrowthofthesupplementedand non-supplementedstrains,butwithinanhourtherewasno increaseintheopticaldensityoftheculturewithoutfatty acid.Exogenousfattyacidswererequiredfor denovo fatty acidsynthesisinstrainPDJ39( ! plsX ).The ! plsX strain wasgrowntoearlylogphaseinthepresenceoffattyacids, thecellsharvested,thefattyacidsupplementremovedand thecellsresuspendedinmediumeitherwithorwithout fattyacidasillustratedinFig.6A.Thecultureswerethen labelledwith[ 14 C]acetateandtheincorporationoflabelinto lipidsmonitoredfor20min(Fig.6B).Theincorporationof acetateintolipidswasreduced > 98%intheabsenceof exogenousfattyacidsshowingthatendogenousFASII wastightlycoupledtotheacquisitionofextracellularfatty acidsinthePlsX-nullstrain.Samplesfromthe ! plsX strain growninthepresenceoffattyacidand30minafterthe removaloffattyacidsfromthemediumwereisolatedand theACPpoolcompositionwasdeterminedbyconformationallysensitivegelelectrophoresisfollowedbyvisualizationoftheACPspeciesbyimmunoblottingwithanti-ACP antibodies(Fig.6C).Normally,theACPpoolin S.aureus consistedprimarilyofnon-esteriÃžedACP(Parsons etal ., 2011),andonlynon-esteriÃžedACPwasdetectedinthe ! plsX samplesgrownwithfattyacids.Followingthe removalofthefattyacidsupplement,long-chainacyl-ACP accumulatedinthe ! plsX strain(Fig.6C).Thesedata wereconsistentwiththemodelthatacyl-ACPutilization dependedontheformationofacyl-G3PbythePlsYreaction.However,notalloftheACPwasconvertedtoacylACPshowingthatFASIIceasedbeforealloftheACPwas depleted.Thiswasincontrasttotheconversionofallthe ACPtoshort-chainacyl-ACPintermediatesin S.aureus treatedwiththeFabIinhibitorAFN-1252(Parsons etal ., 2011).Theaccumulationoflong-chainacyl-ACPcorrelatedwiththecessationofFASII,suggestingthatacyl-ACP mayplayaregulatoryrolein S.aureus similartothe acyl-ACPregulationofFabHandacetyl-CoAcarboxylase in E.coli (HeathandRock,1996;DavisandCronanJr, 2001). Metabolismoffattyacidsincellextracts Previously,wereportedthatextractsfrom S.aureus formedacyl-ACPthatwasidentiÃžedbygelelectrophoresis fromfattyacids,ATPandMg 2 + (Parsons etal .,2011). Thesedataindicatedanacyl-ACPsynthetasemaybe present;however,theanalysisofPlsX-nullcellextracts Fig.5. Positionaldistributionoflabelledprecursorsincorporated intoPtdGro.StrainPDJ39( ! plsX )wasgrowninthepresenceof 0.5mMoleate,10mgml % 1 BSA.AtanA 600 = 0.05,5 " Ciml % 1 of [ 14 C]acetatewasaddedandthecellsgrownfor7h.Strain SA178R1wasgrownandlabelledinthesamemannerusingmedia withoutafattyacidsupplement.Thelipidswereextracted, [ 14 C]PtdGrowasisolatedbythin-layerchromatography,and digestedwithphospholipaseA 2 .Thesampleswerethenseparated bythin-layerchromatography,andthedistributionsoflabelbetween the1-and2-positionswasindicatedbytheratiooflabelled lysoPtdGro(1-position)toFA(2-position)usingtheBioscan ImagingDetector. A.Positionaldistributionof[ 14 C]acetate-labelledPtdGroderived fromwild-typestrainSA178R1.Theincompletedigestionshowsthe locationofthesubstrate(PtdGro)andthetwoproducts,1-acylglycerophosphoglycerol(LysoPtdGro)andfattyacid(FA). B.Positionaldistributionof[ 14 C]acetate-labelledPtdGroderived fromstrainPDJ39( ! plsX ). 240 J.B.Parsons etal . ! Â©2014JohnWiley&SonsLtd, MolecularMicrobiology , 92 ,234Ã245
revealedthatacyl-ACPformationrequiredthepresenceof PlsX(Fig.7A).Thelocationofacyl-ACPandacyl-PO 4 in thethin-layerchromatographyanalysiswasdetermined bypreparing[ 14 C]oleoyl-ACPandpartiallyconvertingit to[ 14 C]acyl-PO 4 usingpuriÃžedPlsX.Extractsfromstrain PDJ39( ! plsX )didnotproduce[ 14 C]acyl-ACPfrom [ 14 C]oleate,ACP,ATPandMg 2 + ,butratherthefattyacid wasconvertedtoaproductthatco-migratedwiththe acyl-PO 4 standard.TheadditionofpuriÃžedPlsXtothe ( ! plsX )extractresultedintheformationof[ 14 C]acyl-ACP, and[ 14 C]acyl-PO 4 wasnotdetected.Thus,acyl-ACPwas notformedbythedirectacylationofACPbyasynthetase asdescribedinothersystems(Jiang etal .,2006;2010), butratherthroughtheinitialformationofacyl-PO 4 followed byitstransacylationtoACPviaPlsX.Theseresultsshowed thepresenceofafattyacidkinaseactivityinthe S.aureus cellextracts.Also,theexperimentsshowedthatacyl-PO 4 wasrapidlyconvertedtoacyl-ACPbyPlsXaccountingfor primarilydetectingacyl-ACPinsimilarreactionsperformed previouslywithwild-typecellextracts(Parsons etal ., 2011).AÃžlterdiscassaywasdevelopedtomeasurefatty acidkinaseactivity,whichwaslinearwiththeprotein concentrationinthePlsX-nullextract(Fig.7B).Fattyacid kinaseactivityrequiredATPplusMg 2 + .InadditionthereactionwassigniÃžcantlystimulatedbypresentingthefattyacid substrateinTritonX-100micelles(Fig.7C),asaremany otherlipid-utilizingenzymes(Carman etal .,1995).There isasingleacetatekinase( ackA )homologuein S.aureus . S.aureus strainKB8000( ! ackA )(Sadykov etal .,2013) wasnotdefectivein[ 14 C]oleateuptake(notshown),ruling outadualfunctionforacetatekinase.Thesedatadetected thepresenceofafattyacidkinaseactivityin S.aureus that wasresponsiblefortheformationofacyl-PO 4 fromexogenousfattyacidsinPlsX-nullcells. Discussion Thisworkrevealstheexistenceofafattyacidkinase pathwayfortheincorporationofexogenousfattyacidsinto thephospholipidsof S.aureus (Fig.1).Thekeyintermediateinphospholipidsynthesis,PtdOH,isformedbythe Fig.6. Couplingoffattyacidandphospholipidsynthesisinstrain PDJ39( ! plsX ). A.GrowthofstrainPDJ39( ! plsX )haltsabruptlyfollowingremoval oftherequiredfattyacidsupplement.Thestrainwasgrowntoan A 600 of0.5inLBmediumcontaining1mMoleateplus10mgml % 1 bovineserumalbumin.Theculturewasthensplitandtheoleate supplementremovedbycentrifugationandthecellsre-suspended inmediaeitherwithorwithoutthefattyacidsupplementand growthmonitored. B.Fattyacidsynthesisrequiresexogenousfattyacidsinstrain PDJ39( ! plsX ).The ! plsX strainwasgrowntoanA 600 of0.5,and theoleatesupplementremoved.Thecellswereresuspendedin mediaeitherwithorwithouttheoleatesupplementand30min later,5 " Ciml % 1 of[ 14 C]acetatewasaddedtobothcultures.Atthe indicatedtimes,thecellswereharvestedandtheamountoflabel incorporatedintothelipidfractionwasdetermined. C.ArepresentativeexampleoftheACPpoolcompositionunder differentgrowthconditionsdeterminedbyconformationallysensitive gelelectrophoresisandimmunoblottingwithanti-ACPantibody. Lane1,strainPDJ39( ! plsX )followingtheremovalofthe exogenous18:1 ! 9supplementfor30min;Lane2,strainPDJ39 grownwiththeoleatesupplement;andLane3, anteiso 17:0-ACP standard. Fattyacidmetabolismin S.aureus 241 Â©2014JohnWiley&SonsLtd, MolecularMicrobiology , 92 ,234Ã245
acylationofthe1-positionbyPlsYfollowedbytheacylACP-dependentacylationofthe2-positionbyPlsC.Exogenousfattyacidsaccesstheacyltransferasesystemby Ãžrsttranslocatingacrossthebacterialcellmembrane.Protonatedfattyacidsareknowntorapidlyandspontaneously ÃŸipacrossaphospholipidbilayer(Garlid etal .,1996).This propertyprovidesaplausiblemechanismforfattyacid entryintocellsbecausethereisnoevidenceforaproteinmediatedcytoplasmicmembranetransporterinanybacteria.Oncethefattyacidsareaccessibletothecytosol, theyareconvertedtoacyl-PO 4 byanATP/Mg 2 + -dependent fattyacidkinase(Fak).Theacyl-PO 4 maythenbeusedfor theacylationofthe1-positionofG3PbyPlsY,orconverted toacyl-ACPbyPlsX.Theacyl-ACPmayeitherbeelongatedbyFASIIorutilizedtoacylatethe2-positionvia PlsC. S.aureus PlsChasahighselectivityfor15-carbon branched-chainfattyacidswhichmeansthatexogenous fattyacidswithchainlengthsgreaterthan15carbonsare almostexclusivelyfoundinthe1-positionofPtdGro.The geneticinactivationofPlsXpreventstheinter-conversion ofacyl-PO 4 andacyl-ACP.Thus,PlsX-nullstrainsarenot abletoelongateextracellularfattyacidsandrequirebotha fattyacidsupplementtosupplyacyl-PO 4 viafattyacid kinase(Fak)toPlsYand denovo FASIItosupplythe acyl-ACPtoPlsC.Thenextimportantstepinthisareaof researchwillbetheidentiÃžcationofthegene(s)encoding thefattyacidkinasecatalyticactivity. Thefattyacidkinasepathwayrepresentsanewparadigmfortheassimilationofextracellularfattyacidsin bacteria,andstandsincontrasttothepathwayforexogenousfattyacidmetabolisminGram-negativebacteria. E.coli hasbeentheclassicalmodelforextracellularfatty acidmetabolism,andotherGram-negativebacteria possessasimilarsetofgenesrequiredfortheactivation offattyacidsderivedfromtheenvironmenttoacyl-CoAs byacyl-CoAsynthetase(FadD).Also, E.coli usesexogenousfattyacidsasacarbonsourcevia # -oxidation (BlackandDiRusso,1994).ManyGram-negativebacteria usefattyacidsasacarbonsource,andtheligationof exogenousfattyacidstoCoA,ratherthanACP,prevents # -oxidationintermediatesfrombecomingintermingled withFASIIintermediates.Ourexperimentsruleoutarole foracyl-CoAin S.aureus fattyacidmetabolism(Fig.2). ThisÃžndingisconsistentwiththeimportantdifferencesin theacyldonorspeciÃžcitiesbetweenthe E.coli PlsB/PlsC acyltransferasesystemandthePlsX/PlsY/PlsCpathway. Fig.7. Acyl-PO 4 andacyl-ACPsynthesisincell-freeextracts. A.Thin-layerchromatographicanalysisofthereactionproducts formedfromtheincubationofcelllysatespreparedfromstrain PDJ39( ! plsX )containingATP,Mg 2 + ,ACP,TritonX-100and [ 14 C]oleateasdetailedinthemethods.Thereactionmixtureswere separatedonSilicaGelGlayersdevelopedwith chloroform:methanol:aceticacid(90/10/10)andthedistributionof radioactivitydeterminedwithaBioscanImagingdetector.Thetop traceshowstheconversionof[ 14 C]oleoyl-ACPto[ 14 C]acyl-PO 4 by puriÃžedPlsX.Themiddletraceisthereactionproductsformedby cell-freeextractsofstrainPDJ39( ! plsX )spikedwithpuriÃžedPlsX alongwithATP,Mg 2 + and[ 14 C]oleate.Thebottomtraceisthesame ! plsX extractintheabsenceofaddedPlsX. B.AÃžlterdiscassayforfattyacidkinasewasperformedusingthe cellextractsfromstrainPDJ39( ! plsX )containingATP,Mg 2 + ,Triton X-100and[ 14 C]oleateasdetailedinthemethods. C.MaximalfattyacidkinaseactivityrequiredATP,Mg 2 + andTriton X-100.Theenzymesourcewasthecytosolfromthe ! plsX strain (300 " gprotein). 242 J.B.Parsons etal . ! Â©2014JohnWiley&SonsLtd, MolecularMicrobiology , 92 ,234Ã245
E.coli PlsBandPlsCutilizeeitheracyl-ACPoracyl-CoA, whereas S.aureus PlsYusesonlyacyl-PO 4 andPlsC usesonlyacyl-ACP.Somebacteriapossessanacyl-ACP synthetasethatbelongstothesameAMP-bindingprotein superfamilyasacyl-CoAsynthetase(Jiang etal .,2006); however,ourexperimentswithPlsX-nullstrainsruleout acyl-ACPsynthetaseinvolvementinphospholipidsynthesisfromextracellularfattyacidsin S.aureus .Howwidespreadthefattyacidkinasepathwayforexogenousfatty acidmetabolismisremainstobedetermined,butitseems likelythatitwouldbepresentinthoseGram-positive pathogensthatutilizethePlsX/PlsY/PlsCpathwayfor phospholipidsynthesis,andareabletoscavengefatty acidsfromtheenvironment. Experimentalprocedures Strains,plasmidsandmaterials S.aureus strainRN4220wasobtainedfromRichardNovick (Kreiswirth etal .,1983). S.aureus strainSA178R1and shuttlevectorpG164wereobtainedfromMerck(D'Elia etal ., 2006). B.subtilis strainLP39thatexpresses plsX onlyinthe presenceofxylosewasdescribedpreviously(Paoletti etal ., 2007).StrainPDJ39( ! plsX )wasconstructedfromstrain SA178R1bytheinsertionofagroupIIintron366bpintothe plsX geneusingtheprimerdesignsoftwareandplasmid systemprovidedintheTargetronGeneKnockoutKit(SigmaAldrich)(Zhong etal .,2003).Genotypingwasperformed usingamultiplexPCRreactioncontainingaprimerspeciÃžc fortheintron(5 & -CGAAATTAGAAACTTGCGTTCAGTAAAC andtwo plsX gene-speciÃžcprimers280F,5 & -CAGCAGG TAATACTGGTGCTTTAATGTCAGand789R,5 & -ATCTTT CTTCA ATATTGCACCTGC.Thewild-type plsX genegavea 509bpproductandtheknockoutallelegavea305bp product.StrainsPDJ23( ! SA0226)andPDJ25( ! SA0533) wereconstructedbyinsertionofagroupIIintronatbp306 andbp93oftheSA0226andSA0533genesrespectivelyin strainRN4220,andveriÃžedbyPCRusingprimersoutsidethe introninsertionsite.StrainPDJ38( ! accD )wasgenerated usingmethodsandplasmidusedtogeneratethesame knockoutinRN4220(Parsons etal .,2011).The E.colifadD genewasclonedintotheBamHI/EcoRIsitesofpG164witha C-terminalFLAGtagusingprimers5 & -GGATCCATGAAGAA GGTTTGGCTTAACCGTTATCCCGCGGACand5 & -TTATTTA TCATCATCATCTTTATAATCTATGGCTTTATTGTCCACTTTGCCGCGCGCTTC. S.aureusplsX genewasampliÃžed usingprimers5 & -ATCGCATATGAAAAAAATCGCAGTAGAT GCCATGGand5 & -CGATAAGCTTTTAGTGGTGGTGGTGG TGGTGTTCTCCTGAAAATTCACGCGCAGTCandligated intotheNdeIandHindIIIsitesofpET28atogeneratepJY010. Inordertoligate plsX intopG164, plsX genewasampliÃžed usingprimers5 & -AAATGTCATATGATAAGCGAGGATAAAA TTATGGand5 & -TTGCTGGGATCCTCATTTGATTCACCTA CAGTCTCTTTCandligatedintoTOPOPCR2.1cloning vector.AninternalNdeIsitewasidentiÃžedandremovedby site-directedmutagenesisusingprimers5 & -GCTAAAGGT AATAGTTTAACGAAAAAATCTTATGAGTTATTAAATCATGA TCATTCATTand5 & -AATGAATGATCATGATTTAATAACTC ATAAGATTTTTTCGTTAAACTATTACCTTTAGC.The plsX genewassubsequentlyligatedintotheNdeIandBamHIsites ofpPJ131andrestrictedwithNheIandBlpItoallowligation intopG164.GrowthmediausedwaseitherLuriaBroth(LB), LBsupplementedwith10mgml % 1 bovineserumalbuminor LBsupplementedwith0.1%Brij58plus500 " Mexogenous fattyacid.[114 C]Aceticacid(55mCimmol % 1 )and[114 C]oleic acid(55mCimmol % 1 )werepurchasedfromAmericanRadiolabelledChemicalsandPerkin-Elmerrespectively.AffinitypuriÃžedanti-rabbitACPantibodywasdescribedpreviously (JackowskiandRock,1983).Fattyacidfreebovineserum albuminandBrij-58werefromSigma-Aldrich.Fattyacids werefromLarodanFineChemicalsorSigma-Aldrich. [114 C ]Acetateincorporationinfattyacid-treatedcells S.aureus strainRN4220wasculturedinLBmediumcontaining0.1%Brij-58untilA 600 = 0.5whentheculturewasdivided into10mlaliquots.Indicatedamountsoffattyacidinavolume of10 " lDMSOwereaddedtotheculturesandthenwere incubatedat37Â¡Cwithvigorousshaking.After30min,50 " Ci of[114 C]acetatewasaddedtoeachculturewhichwere returnedtothe37Â¡Cincubatorfor30min.[114 C]Acetate incorporationinDMSO-treatedcellsrepresented100%incorporation.Thecellswereharvestedbycentrifugation,washed twotimeswithPBS.Thecellpelletwassuspendedin100 " l waterandlipidsextractedwith360 " lchloroform:methanol:hydrochloricacid(1:2:0.02).Phaseswereseparated afteradditionof120 " lchloroformand120 " l2Mpotassium chloride.RadiolabelledlipidswerequantiÃžedbyscintillation counting. Minimalinhibitoryconcentration Theminimalinhibitoryconcentrations(MICs)forAFN-1252 againststrainsSA178R1,PDJ38andPDJ39weredeterminedusingabrothmicrodilutionmethodinLBmedium supplementedwith10mgml % 1 bovineserumalbuminand 1mMmixtureof15:0/17:0 anteiso branched-chainfatty acids. S.aureus culturesweregrowntoA 600 = 1.0anddiluted 30000-foldinmedium.A10 " laliquotofdilutedcellswas addedtoeachwellofaU-bottomed,96-wellplatecontaining 100 " lofmediumwithindicatedAFN-1252concentration.A fattyacidconcentrationof1mMwasusedbecausethisconcentrationsupportednormalgrowthof S.aureus fattyacid auxotrophs( ! accD )(Parsons etal .,2013).Theplatewas incubatedat37Â¡Cfor20handreadusingaFusionplate readerat600nm. Lipodomics PtdGromolecularspecieswereobtainedfromlipidextracts from50mlculturesofstrainsSA178R1andPDJ39( ! plsX ) growninLBplus10mgml % 1 bovineserumalbuminwithor without1mM18:1or0.2mM anteiso 15:0supplements. Lipidextractionandmassspectrometrywasperformedas describedpreviously(Parsons etal .,2011).Fattyacidmethyl esterspreparedandquantiÃžedusingaHewlett-Packard5890 gaschromatographasdescribedpreviously(Zhang etal ., 2002). Fattyacidmetabolismin S.aureus 243 Â©2014JohnWiley&SonsLtd, MolecularMicrobiology , 92 ,234Ã245
Thepositionaldistributionoflabelledprecursorswasdeterminedbylabellingstrainsin10mlculturesgrowninLB mediumplus10mgml % 1 bovineserumalbumincontaining either5 " Ciml % 1 [114 C]acetateor0.5 " Ciml % 1 [114 C]oleic acidandgrowntoA 600 = 2.0.Cellswerepelleted,washed 2 $ 10mlLBcontaining10mgml % 1 bovineserumalbumin and2 $ 10mlH 2 O.Thecellpelletwassuspendedin100 " l waterandlipidsextractedwith360 " lchloroform:methanol:hydrochloricacid(1:2:0.02).Phaseswereseparated afteradditionof120 " lchloroformand120 " l2Mpotassium chloride.[ 14 C]PtdGrowasisolatedbypreparativethin-layer chromatographyonSilicaGelHlayerswithchloroform:methanol:aceticacid(55/20/5,v/v/v).The[ 14 C]PtdGro wasdigestedwith Najanaja snakevenomphospholipaseA 2 (Parsons etal .,2011).After3h,thelipidswereextracted andappliedtoaSilicaGelHlayerdevelopedinchloroform:methanol:aceticacid(55/20/5,v/v/v)andtheplate imagedwithaBioscanImagingdetectortodeterminethe extentofdigestionandthedistributionoflabelbetweenthe1and2-positions. ACPimmunoblotting A40mlcultureof S.aureus wasgrowntoanA 600 = 0.7,the cellswereharvestedbycentrifugationandwashedtwice withLBmediumcontaining10mgml % 1 bovineserum albumin.Thecellpelletwassplitbetweenmediumeitherwith orwithoutoleateandgrownforanadditional30min.Cells from20mlcultureswerecentrifugedat4000 g ,for10min andpelletsresuspendedin110 " lphosphate-bufferedsaline containing1mgml % 1 lysostaphin,0.1mgml % 1 DNaseIand proteaseinhibitorcocktailbeforeincubationonicefor3h. Theextractswerecentrifugedat20000 g for15minandthe supernatantremovedandproteinquantiÃžedusingBradford assay(Bradford,1976).Celllysates(30 " g)and100ngof 17:0-ACPwereloadedontoa2.5Murea,15%acrylamide conformationallysensitivegel.Theseparatedproteinswere transferredtoapolyvinylidenediÃŸuoridemembranebyelectroblotting.Theprimaryanti-ACPantibodywasusedat1:500 dilutionandsecondaryanti-rabbitIgGconjugatedwithalkalinephosphataseata1:5000dilution.TheblotwasdevelopedusingtheECFsubstrateandtheÃŸuorescentsignal recordedusingaTyphoonPhosphoImager9500. PlsXpuriÃžcation PlasmidpJY010expressingHis-taggedPlsXwastransformedintoBL21(DE3)Tunercellsandusedtoinoculatea1l cultureofLBwhichwasgrownat37Â¡Cwithvigorousshaking. TheculturewasgrownuntilA 600 = 0.7whenproteinexpressionwasinducedbytheadditionof1mMIPTG.After3h,the cellswereharvestedbycentrifugationandresuspendedin bindingbuffer(20mMTris-HCl,pH8.0,0.5MNaCl).The cellswerelysedusingamicroÃŸuidizeranddebrisremovedby centrifugationat36000 g for15min.Thesupernatantwas appliedtoaNi-NTAcolumnandwashedwithincreasing concentrationsofimidazole.PuriÃžedPlsXwaselutedin 400mMimidazoleandwassubsequentlydialysedovernight against20mMTris-HCl(pH7.5),5mMEDTA,100mMNaCl and1mMDTT(Lu etal .,2006). Fattyacidkinasebiochemicalassay Extractsformeasuringfattyacidkinaseactivityweregeneratedfrom10mlculturesofstrainPDJ39( ! plsX )grownin LB/BSAwith1mMoleicaciduntilA 600 = 1.0.Cellswerecentrifugedandresuspendedin110 " lphosphate-buffersaline containing1mgml % 1 lysostaphin,0.1mgml % 1 DNaseIand proteaseinhibitorcocktailbeforeincubationoniceforthree hours.Theextractswerecentrifugedat20000 g for15min andthesupernatantremoved.Theradiochemicalassaycontained0.1MTris-HCl,pH7.5,20mMMgCl 2 ,10mMATP,1% TritonX-100,0.1 " Ci[114 C]oleicacidandindicatedamount ofcellextractinavolumeof60 " l.Assayswereincubatedat 37Â¡Cfor20minbeforeanalysis.FortheÃžlterdiscassay, 50 " lofreactionmixturewasspottedontoaDE81Ãžlterdisc andsubsequentlywashed3timesfor30minwith1%acetic acidinethanol.Thediscsweredriedandanalysedbyscintillationcounting.Thin-layerchromatographywasperformed using20 " lofreactionmixturespottedontoaSilicaGelG layer,developedinchloroform:methanol:aceticacid(90/ 10/10)andanalysedwithaBioscanImagingdetector.The countingefficiencyoftheBioscanwithourSilicaGelGplates was2%.[114 C]18:1-ACPwasgeneratedusing[114 C]oleic acid, S.aureus ACPandthe Vibrioharveyi acyl-ACPsynthetaseenzyme(Fice etal .,1993).[114 C]oleoyl-phosphate standardwasgeneratedusing[114 C]18:1-ACPandPlsXas describedpreviously(Lu etal .,2006). Acknowledgements WethankNachumKaplanforhisgiftofAFN-1252and KennethBaylesforstrainKB8000.ThisresearchwassupportedbyNationalInstitutesofHealthGrantGM034496 (C.O.R.),CancerCenterSupportGrantCA21765andthe AmericanLebaneseSyrianAssociatedCharities.Thecontent issolelytheresponsibilityoftheauthorsanddoesnotnecessarilyrepresenttheofficialviewsoftheNationalInstituteof GeneralMedicalSciencesortheNationalInstitutesofHealth. References Black,P.N.,andDiRusso,C.C.(1994)Molecularandbiochemicalanalysesoffattyacidtransport,metabolism,and generegulationin Escherichiacoli . BiochimBiophysActa 1210: 123Ã145. Bradford,M.M.(1976)Arapidandsensitivemethodforquantitationofmicrogramquantitiesofproteinutilizingtheprincipleofprotein-dyebinding. AnalBiochem 72: 248Ã254. Carman,G.M.,Deems,R.A.,andDennis,E.A.(1995)Lipid signalingenzymesandsurfacedilutionkinetics. JBiol Chem 270: 18711Ã18714. Coleman,J.(1992)Characterizationofthe Escherichiacoli genefor1-acylsn -glycerol-3-phosphateacyltransferase ( plsC ). MolGenGenet 232: 295Ã303. Cronan,J.E.,Jr,andRock,C.O.(1996)Biosynthesisof membranelipids.In Escherichiacoli and Salmonellatyphimurium: CellularandMolecularBiology.Neidhardt,F.C., Curtis,R.,Gross,C.A.,Ingraham,J.L.,Lin,E.C.C.,and Low,K.B.(eds).Washington,DC:AmericanSocietyfor Microbiology,pp.612Ã636. 244 J.B.Parsons etal . ! Â©2014JohnWiley&SonsLtd, MolecularMicrobiology , 92 ,234Ã245
Davis,M.S.,andCronan,J.E.,Jr(2001)Inhibitionof Escherichiacoli acetylcoenzymeAcarboxylasebyacyl-acyl carrierprotein. JBacteriol 183: 1499Ã1503. D'Elia,M.A.,Pereira,M.P.,Chung,Y.S.,Zhao,W.,Chau,A., Kenney,T.J., etal .(2006)Lesionsinteichoicacidbiosynthesisin Staphylococcusaureus leadtoalethalgainof functionintheotherwisedispensablepathway. JBacteriol 188: 4183Ã4189. DiRusso,C.C.,Black,P.N.,andWeimar,J.D.(1999)Molecularinroadsintotheregulationandmetabolismoffatty acids,lessonsfrombacteria. ProgLipidRes 38: 129Ã197. Fice,D.,Shen,Z.,andByers,D.M.(1993)PuriÃžcationand characterizationoffattyacyl-acylcarrierproteinsynthetase from Vibrioharveyi . JBacteriol 175: 1865Ã1870. Garlid,K.D.,Orosz,D.E.,Modriansky,M.,Vassanelli,S.,and Jezek,P.(1996)Onthemechanismoffattyacid-induced protontransportbymitochondrialuncouplingprotein. JBiol Chem 271: 2615Ã2620. Green,P.R.,Merrill,A.H.,Jr,andBell,R.M.(1981)Membranephospholipidsynthesisin Escherichiacoli :puriÃžcation,reconstitution,andcharacterizationof sn -glycerol-3phosphateacyltransferase. JBiolChem 256: 11151Ã 11159. Heath,R.J.,andRock,C.O.(1996)Inhibitionof # -ketoacylacylcarrierproteinsynthaseIII(FabH)byacyl-acylcarrier proteinin Escherichiacoli . JBiolChem 271: 10996Ã11000. Heath,R.J.,Jackowski,S.,andRock,C.O.(1994)Guanosinetetraphosphateinhibitionoffattyacidandphospholipidsynthesisin Escherichiacoli isrelievedby overexpressionofglycerol-3-phosphateacyltransferase ( plsB ). JBiolChem 269: 26584Ã26590. Jackowski,S.,andRock,C.O.(1983)Ratioofactiveto inactiveformsofacylcarrierproteinin Escherichiacoli . J BiolChem 258: 15186Ã15191. Jiang,Y.,Chan,C.H.,andCronan,J.E.(2006)Thesoluble acyl-acylcarrierproteinsynthetaseof Vibrioharveyi B392 isamemberofthemediumchainacyl-CoAsynthetase family. Biochemistry 45: 10008Ã10019. Jiang,Y.,Morgan-Kiss,R.M.,Campbell,J.W.,Chan,C.H., andCronan,J.E.(2010)Expressionof Vibrioharveyi acylACPsynthetaseallowsefficiententryofexogenousfatty acidsintothe Escherichiacoli fattyacidandlipidAsyntheticpathways. Biochemistry 49: 718Ã726. Kaplan,N.,Albert,M.,Awrey,D.,Bardouniotis,E.,Berman, J.,Clarke,T., etal .(2012)Modeofaction, invitro activity, and invivo efficacyofAFN-1252,aselectiveantistaphylococcalFabIinhibitor. AntimicrobAgentsChemother 56: 5865Ã5874. Kreiswirth,B.N.,Lofdahl,S.,Betley,M.J.,O'Reilly,M., Schlievert,P.M.,Bergdoll,M.S.,andNovick,R.P.(1983) Thetoxicshocksyndromeexotoxinstructuralgeneis notdetectablytransmittedbyaprophage. Nature 305: 709Ã712. Lightner,V.A.,Larson,T.J.,Tailleur,P.,Kantor,G.D.,Raetz, C.R.H.,Bell,R.M.,andModrich,P.(1980)Membrane phospholipidsynthesisin Escherichiacoli :cloningofa structuralgene( plsB )ofthe sn -glycerol-3-phosphateacyltransferase. JBiolChem 255: 9413Ã9420. Lu,Y.-J.,Zhang,Y.-M.,Grimes,K.D.,Qi,J.,Lee,R.E.,and Rock,C.O.(2006)Acyl-phosphatesinitiatemembrane phospholipidsynthesisingram-positivepathogens. Mol Cell 23: 765Ã772. Lu,Y.-J.,Zhang,F.,Grimes,K.D.,Lee,R.E.,andRock,C.O. (2007)TopologyandactivesiteofPlsY:thebacterial acylphosphate:glycerol-3-phosphateacyltransferase. J BiolChem 282: 11339Ã11346. Paoletti,L.,Lu,Y.-J.,Schujman,G.E.,deMendoza,D.,and Rock,C.O.(2007)Couplingoffattyacidandphospholipid synthesisin Bacillussubtilis . JBacteriol 189: 5816Ã 5824. Parsons,J.B.,Frank,M.W.,Subramanian,C.,Saenkham,P., andRock,C.O.(2011)Metabolicbasisforthedifferential susceptibilityofGram-positivepathogenstofattyacidsynthesisinhibitors. ProcNatlAcadSciUSA 108: 15378Ã 15383. Parsons,J.B.,Frank,M.W.,Rosch,J.W.,andRock,C.O. (2013) Staphylococcusaureus fattyacidauxotrophsdonot proliferateinmice. AntimicrobAgentsChemother 57: 5729Ã5732. Rock,C.O.,Goelz,S.E.,andCronan,J.E.,Jr(1981a)Phospholipidsynthesisin Escherichiacoli .Characteristicsof fattyacidtransferfromacyl-acylcarrierproteinto sn -glycerol-3-phosphate. JBiolChem 256: 736Ã742. Rock,C.O.,Goelz,S.E.,andCronan,J.E.,Jr(1981b)Phospholipidsynthesisin Escherichiacoli .Characteristicsof fattyacidtransferfromacyl-acylcarrierproteinto sn -glycerol-3-phosphate. JBiolChem 256: 736Ã742. Sadykov,M.R.,Thomas,V.C.,Marshall,D.D.,Wenstrom, C.J.,Moormeier,D.E.,Widhelm,T.J., etal .(2013)InactivationofthePta-AckApathwaycausescelldeathin Staphylococcusaureus . JBacteriol 195: 3035Ã3044. Yao,J.,andRock,C.O.(2013)Phosphatidicacidsynthesisin bacteria. BiochimBiophysActa 1831: 495Ã502. Zhang,Y.-M.,Marrakchi,H.,andRock,C.O.(2002)The FabR(YijC)transcriptionfactorregulatesunsaturatedfatty acidbiosynthesisin Escherichiacoli . JBiolChem 277: 15558Ã15565. Zhong,J.,Karberg,M.,andLambowitz,A.M.(2003)Targeted andrandombacterialgenedisruptionusingagroupII intron(targetron)vectorcontainingaretrotranspositionactivatedselectablemarker. NucleicAcidsRes 31: 1656Ã 1664. Fattyacidmetabolismin S.aureus 245 Â©2014JohnWiley&SonsLtd, MolecularMicrobiology , 92 ,234Ã245