<%BANNER%>

Manipulations of the Peripheral Stalk in the F1FO ATP Synthase

Permanent Link: http://ufdc.ufl.edu/UFE0042384/00001

Material Information

Title: Manipulations of the Peripheral Stalk in the F1FO ATP Synthase
Physical Description: 1 online resource (218 p.)
Language: english
Creator: Welch, Amanda
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: Biochemistry and Molecular Biology (IDP) -- Dissertations, Academic -- UF
Genre: Medical Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: MANIPULATIONS OF THE PERIPHERAL STALK IN THE F1FO ATP SYNTHASE The F1FO adenosine triphosphate (ATP) synthase is expressed in almost every organism. The enzyme is primarily responsible for the production of the high-energy compound, ATP. The enzyme is located in the cytoplasmic membrane of bacteria, the thylakoid membrane of chloroplasts, and the inner mitochondrial membrane of eukaryotes. The mechanism and architecture of the enzyme is highly conserved from bacteria to mammalian organisms. The enzyme is composed of two sectors, F1 and FO. The F1 sector houses the catalytic sites responsible for generating ATP from ADP and Pi. Catalytic sites are located at three alpha/beta interfaces and exist in three different conformational states. Sequential passage of each catalytic site though the different conformations is triggered by rotation of the central, rotor stalk. The torque to move the rotor stalk is generated by the passage of protons down an electrochemical gradient through the FO sector. The function of the peripheral stalk is to hold the F1 sector catalytic subunits in place against the movement of the rotor stalk. This work examined structure function relationships within the peripheral stalks of both bacterial and mitochondrial enzymes. Previous work had shown that the bacterial peripheral stalk had high inherent plasticity. These observations were extended by examining the positional constraints on a highly conserved arginine (bR36) located in the Escherichia coli peripheral stalk. Additional studies were conducted to facilitate biophysical experimentation on the peripheral stalk. During the course of this work, developments in other laboratories suggested that the eukaryotic peripheral stalk might not share the plasticity observed in the bacterial stalk. Therefore, a genetic approach was adopted to investigate the plasticity of the S. cerevisiae peripheral stalk. The S. cerevisiae stalk could be extended in a manner similar to the bacterial stalk. The result indicated considerable plasticity within the eukaryotic peripheral stalk. The current structural and functional information is incorporated into a new peripheral stalk model.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Amanda Welch.
Thesis: Thesis (Ph.D.)--University of Florida, 2010.
Local: Adviser: Cain, Brian D.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0042384:00001

Permanent Link: http://ufdc.ufl.edu/UFE0042384/00001

Material Information

Title: Manipulations of the Peripheral Stalk in the F1FO ATP Synthase
Physical Description: 1 online resource (218 p.)
Language: english
Creator: Welch, Amanda
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: Biochemistry and Molecular Biology (IDP) -- Dissertations, Academic -- UF
Genre: Medical Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: MANIPULATIONS OF THE PERIPHERAL STALK IN THE F1FO ATP SYNTHASE The F1FO adenosine triphosphate (ATP) synthase is expressed in almost every organism. The enzyme is primarily responsible for the production of the high-energy compound, ATP. The enzyme is located in the cytoplasmic membrane of bacteria, the thylakoid membrane of chloroplasts, and the inner mitochondrial membrane of eukaryotes. The mechanism and architecture of the enzyme is highly conserved from bacteria to mammalian organisms. The enzyme is composed of two sectors, F1 and FO. The F1 sector houses the catalytic sites responsible for generating ATP from ADP and Pi. Catalytic sites are located at three alpha/beta interfaces and exist in three different conformational states. Sequential passage of each catalytic site though the different conformations is triggered by rotation of the central, rotor stalk. The torque to move the rotor stalk is generated by the passage of protons down an electrochemical gradient through the FO sector. The function of the peripheral stalk is to hold the F1 sector catalytic subunits in place against the movement of the rotor stalk. This work examined structure function relationships within the peripheral stalks of both bacterial and mitochondrial enzymes. Previous work had shown that the bacterial peripheral stalk had high inherent plasticity. These observations were extended by examining the positional constraints on a highly conserved arginine (bR36) located in the Escherichia coli peripheral stalk. Additional studies were conducted to facilitate biophysical experimentation on the peripheral stalk. During the course of this work, developments in other laboratories suggested that the eukaryotic peripheral stalk might not share the plasticity observed in the bacterial stalk. Therefore, a genetic approach was adopted to investigate the plasticity of the S. cerevisiae peripheral stalk. The S. cerevisiae stalk could be extended in a manner similar to the bacterial stalk. The result indicated considerable plasticity within the eukaryotic peripheral stalk. The current structural and functional information is incorporated into a new peripheral stalk model.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Amanda Welch.
Thesis: Thesis (Ph.D.)--University of Florida, 2010.
Local: Adviser: Cain, Brian D.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0042384:00001


This item has the following downloads:


Full Text

PAGE 1

1 MANIPULATIONS OF THE PERIPHERAL STALK IN THE F1FO ATP SYNTHASE By AMANDA KUHNS WELCH 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 2010

PAGE 2

2 2010 Amanda Kuhns Welch

PAGE 3

3 To my husband, parents, and dog

PAGE 4

4 ACKNOWLEDGMENTS I would like to acknowledge and thank my advisor, Dr. Brian Cain, for all his help, support, and proofreading. H e has provided me with an excellent laboratory and graduate school experience, given me advice, and tolerated my eccentricities. I would like to thank the members of my committee Drs. Linda Bloom, Michael Kilberg, and Wolfgang Streit. They have provided me with help and valuable insight through out the dissertation process. I appreciate all the help that my undergraduate research assistants Rachel Grimes, Caleb Bostwick, and Randy Quesadahave provided over the past four years. I would also like to acknowledge Dr. Shane Claggett for his advice and always making me laugh. I am grateful to Mollie Jacobs for her assistance with the experiments in this dissertation, constant support, and provision of magic water. My thanks go to my parents, John and Jana Kuhns, for always encouraging and believing in me. I appreciate the support from all my friends, especially Jennifer Lavoie and Rici Hutchinson. Finally, I would like to thank my husband, Stephen Welch, for all his support and not becoming frightened when th e dissertation process made me crazy.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................. 4 LIST OF TABLES ............................................................................................................ 8 LIST OF FIGURES .......................................................................................................... 9 ABSTRACT ................................................................................................................... 12 CHAPTER 1 BACKGROUND AND SIGNIFICANCE ................................................................... 14 Introduction ............................................................................................................. 14 Coupling Adenosine Triphosphate Generation to Cellular Respiration ................... 15 Cellular Respiration and the E lectron Transport Chain ........................................... 17 The Search for the Phosyphorylated Intermediate .................................................. 19 The Chemiosmotic Hypothesis ............................................................................... 21 The Binding Change Mechanism ............................................................................ 23 Subunit Composition of the F1FO ATP Synthase .................................................... 27 Cat alytic and Rotational Domains .................................................................... 28 The and subunits. ................................................................................ 29 The rotor stalk ............................................................................................ 31 Proton Translocation Domain ........................................................................... 35 The a subunit ............................................................................................. 35 The c subunits ............................................................................................ 38 The Peripheral Stalk ......................................................................................... 39 Mutagenic Analysis of the Bacterial and Eukaryotic Peripheral Stalks ................... 41 Mutagenesis of the Bacterial Peripheral Stalk .................................................. 41 Membrane domain of the bacterial b subunit ............................................. 42 Tether domain of the bacterial b subunit .................................................... 43 Dimerization domain of the bacterial b subunit .......................................... 45 The F1 b subunit ............................. 47 Chimeras of the bacterial b subunit ............................................................ 48 Right handed versus left handed coiledcoil debate .................................. 49 ............................................................................... 53 Mutagenic Analysis of the Mitochondrial Peripheral Stalk ................................ 55 Transmembrane domain I .......................................................................... 56 Intermembrane space loop ........................................................................ 57 Transmembrane domain II ......................................................................... 58 The h and d subunit interaction domain ..................................................... 58 The F1 and Oligomycin Sensitivity Conferring Protein binding domain ...... 59 Mutations in the mitochondrial d and h/ F6 subunits .................................... 61 Mutations in the mitochondrial OSCP subunit ............................................ 63

PAGE 6

6 Structures of the Peripheral Stalk ........................................................................... 64 Peripheral Stalk Models .......................................................................................... 66 2 INTRAGENIC SUPPRESSORS OF THE ARGININE 36 MUTATIONS IN THE BACTERIAL PERIPHERAL STALK ........................................................................ 81 Introduction ............................................................................................................. 81 Materials and Methods ............................................................................................ 84 Materials ........................................................................................................... 84 Mutagenesis Strains and Media ....................................................................... 85 Preparative Procedures .................................................................................... 86 Assay of F1FO ATP Synthase Activity .............................................................. 86 Immunoblot Analysis ........................................................................................ 87 Results .................................................................................................................... 88 Construction and Growth Characteristics of Mutants ....................................... 88 Assembly of Intact F1FO ATP Synthase ............................................................ 89 Coupled F1FO Activity ....................................................................................... 90 Discussion .............................................................................................................. 91 3 CHEMICAL MODIFICATION OF CYSTEINES IN THE BACTERIAL PERIPHERAL STALK ........................................................................................... 102 Introduction ........................................................................................................... 102 Materials and Methods .......................................................................................... 105 Materials and General Molecular Methods ..................................................... 105 Isol ation of the F1 Sector ................................................................................ 106 Labeling of the F1 Sector ................................................................................ 109 Reconstitution of F1FO ATP Synthase ............................................................ 110 Results .................................................................................................................. 110 Generation of F1FO ATP Synthases with Mutant b subunits ........................... 110 Construction of mutant b and bN2C,C21S,+11 genes .................... 111 Coupled activity of enzymes containing b and bN2C,C21S,+11 subunits ................................................................................................ 112 Two Plasmid Expression System for Introduction of a Cysteine into One b Subunit ........................................................................................................ 113 Construction of plasmids for the twoplasmid expression system. ........... 114 Coupled activity of enzymes from the twoplasmid expression system. ... 115 Chemical modific ......... 115 Attempted reconstitution of the chemically modified F1FO ........................ 117 Chemical mod ification of intact F1FO ATP synthase (cysless operon, C64) ...................................................................................................... 118 Discussion ............................................................................................................ 118 4 MANIPULATIONS OF THE PERIPHERAL STALK OF THE EUKARYOTIC F1FO ATP SYNTHASE .................................................................................................. 134 Introduction ........................................................................................................... 134

PAGE 7

7 Materials and Methods .......................................................................................... 136 Materials and General Molecular Biology Methods ........................................ 136 Transformations ............................................................................................. 136 Colony PCR .................................................................................................... 138 Construction of ATP4syn Expression Vectors .................................................. 139 Isolation of Genomic and Plasmid DNA from S. cerevisiae ............................ 140 Sequencing of Genomic or Plasmid DNA ....................................................... 141 Preparative Procedures .................................................................................. 142 Assay of F1FO ATP synthase activity .............................................................. 144 Immunoblot analysis ....................................................................................... 144 Results .................................................................................................................. 145 Construction of ATP4syn Expression Vectors .................................................. 145 ATP4 Strain ................................................ 147 Construction of Mutant Strains ....................................................................... 149 Growth Properties of Mutant Strains .............................................................. 152 Expression of Recombinant b Subunits .......................................................... 152 F1FO ATPase Activity ...................................................................................... 154 Discu ssion ............................................................................................................ 155 5 CONCLUSIONS AND FUTURE DIRECTIONS .................................................... 174 Conclusions .......................................................................................................... 174 Intragenic Suppressors of the Arginine 36 Mutations in the Bacterial Peripheral Stalk ........................................................................................... 176 Chemical Modification of Cysteines in the Bacterial Peripheral Stalk ............. 177 Manipulation of the Eukaryotic Peripheral Stalk in the F1FO ATP Synthase ... 179 Future Directions .................................................................................................. 183 Fluorescent Techniques in the Eukaryotic F1FO ATP Synthase ..................... 184 Further Validation of the Partial Peripheral Stalk Structure ............................ 187 Summary .............................................................................................................. 190 APPENDIX: YEAST MEDIA AND IDIOSYNCRASIES ................................................ 198 LIST OF REFERENCES ............................................................................................. 200 BIOGRAPHICAL SKETCH .......................................................................................... 218

PAGE 8

8 LIST OF TABLES Table page 1 1 Subunit composition of bacterial and mitochondrial F1FO ATP synthases. ......... 69 1 2 The subunit composition of the Escherichia coli F1FO ATP synthase. ................ 69 1 3 S. cerevisiae F1FO ATP synthase subunit composition. ...................................... 70 1 4 B. taurus F1FO ATP synthase subunit composition. ............................................ 70 2 1 Aerobic growth properties of mutants with double amino acid substitutions. ...... 95 3 1 Primers to generate b subunit genes for cysteine introduction ......................... 121 3 2 Plasmids generated for cysteine introduction ................................................... 121 4 1 Plasmids generated for ATP4syn expression .................................................... 159 4 2 Primers used in S. cerevisiae mutagenesis experiments ................................. 160 4 4 S. cerevisiae F1FO ATP synthase activity ......................................................... 162 5 1 Primer sequences and substitutions for bL123V126 ............................................ 192

PAGE 9

9 LIST OF FIGURES Figure page 1 1 Mitochondrial electron transport chain ................................................................ 71 1 2 Binding change mechanism ............................................................................... 72 1 3 Binding change equation .................................................................................... 72 1 4 Prokaryotic and eukaryotic F1FO ATP synthases ............................................... 73 1 5 Ro tor stalk interactions with F1 in S. cerevisiae ................................................. 74 1 6 Topology of the a subunit .................................................................................. 75 1 7 Mechanism of proton translocation ..................................................................... 76 1 8 Mechanism of proton translocation in the c ring ................................................. 77 1 9 Kane Dickson peripheral stalk structure ............................................................. 78 1 10 Partial B. taurus peripheral stalk structure docked with the B. taurus F1c10 structure inside the S. cerevisiae F1FO ATP synthas e cryoelectron microscopy image ............................................................................................... 79 1 11 Structure of the B. taurus peripheral stalk with the F1 sector .............................. 80 2 1 Sense strand sequences of recombinant uncF(b) genes ................................... 96 2 2 Immunoblot analysis of uncF(b) gene mutant membranes ................................. 97 2 3 ATP driven energization of membrane vesicles prepared from isoleucine uncF(b) gene mutants ........................................................................................ 98 2 4 ATP driven energization of membrane vesicles prepared from isoleucine uncF(b) gene mutants ........................................................................................ 99 2 5 ATP driven energization of membrane vesicles prepared from glutamic acid uncF(b) gene mutants ...................................................................................... 100 2 6 Model of mutated region in pTAM46 ( bV5) and pAW1 ( bR36E, E39R, V5) ................ 101 3 1 Construction of pAW18 ( b) .............................................................. 122 3 2 Construction of pAW19 ( bN2C, C20S, +11) .............................................................. 123 3 3 Coupled F1FO ( b bN2C, C20S, +11) ......................................... 124

PAGE 10

10 3 4 Construction of pAW12 ( b) .......................................................... 125 3 5 Construction of pAW32 ( b) ................................................... 126 3 6 b) .......................................................... 127 3 7 Construction of pAW15 (Cysteineless unc C64) ................ 128 3 8 Coupled F1FO Activity in membrane vesicles from strains ........................................ 129 3 9 Initial labeling reaction of F1 with AlexaFluor 488 .......................................................................................... 130 3 10 Labeled fractions of F1 AlexaFluor 488 ................................................................................................. 131 3 11 Attempted reconstitution of Alexa Fluor 488 labeled with FO ............................ 132 3 12 Labeled intact F1FO ATP synthase ................................................................... 133 4 1 Scheme for constructing E. coliS. cerevisiae shuttle vector ............................ 163 4 3 Scheme for sequencing genomic and plasmid DNA from S. cerevisiae ........... 165 4 4 Design of Synthetic ATP4 gene ........................................................................ 166 4 5 Scheme for plasmid shuffling ............................................................................ 167 4 6 Agarose gel electrophoresis of S. cerevisiae strain AW2 ................................. 168 4 7 YPD plate showing dissected tetrads ............................................................... 169 4 8 Sequence and agarose gel electrophoresis of strain AW3 ............................... 170 4 9 Growth of wild type and ATP4syn strain on YPG ............................................... 171 4 10 Amino acid sequence alignment of B. taurus and S. cerevisiae b subunits ...... 171 4 11 Double strand sequences of recombinant ATP4(b) genes ............................... 172 4 12 Growth of ATP4synmutant strains on YPGD ...................................................... 173 4 13 Immunoblot analysis of atp4(b) gene mutant mitochondria .............................. 173 5 1 Hinge Stiff model .............................................................................................. 193 5 2 Amino acids and distances used to determine angles of manipulated b subunits ............................................................................................................ 194

PAGE 11

11 5 3 Model of the shortened and lengthened peripheral stalk complex in B. taurus 195 5 4 Proposed FRET r esidues ................................................................................. 196 5 5 Close view of b d subunit interaction ................................................................ 197

PAGE 12

12 Abstract of Dissertation Presented to the Graduate School of the University of of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy MANIPULATIONS OF THE PERIPHERAL STALK IN THE F1FO ATP SYNTHASE By Amanda Kuhns Welch December 2010 Chair: Brian Cain Major: Medical Sciences The F1FO adenosine triphosphate (ATP) synthase is expressed in al most every organism. The enzyme is primarily responsible for the production of the highenergy compound, ATP. The enzyme is located in the cytoplasmic membrane of bacteria, the thylakoid membrane of chloroplasts, and the inner mitochondrial membrane of eukaryotes. The mechanism and architecture of the enzyme is highly conserved from bacteria to mammalian organisms. The enzyme is composed of two sectors, F1 and FO. The F1 sector houses the catalytic sites responsible for generating ATP from ADP and Pi. Catalytic sites are located at three / interfaces and exist in three different conformational states. Sequential passage of each catalytic site though the different conformations is triggered by rotation of the central rotor stalk. The torque to m ove the rotor stalk is generated by the passage of protons down an electrochemical gradient through the FO sector. The function of the peripheral stalk is to hold the F1 sector catalytic subunits in place against the movement of the rotor stalk This work examined structure function relationships within the peripheral stalks of both bacterial and mitochondrial enzymes Previous work had shown that the bacterial

PAGE 13

13 peripheral stalk had high inherent plasticity. These observations were extended by examining the positional constraints on a highly conserved arginine ( bR36) located in the Escherichia coli peripheral stalk. Additional studies were conducted to facilitate biophysical experimentation on the peripheral stalk. During the course of this work, developm ents in other laboratories suggested that the eukaryotic peripheral stalk might not share the plasticity observed in the bacterial stalk. Therefore, a genetic approach was adopted to investigate the plasticity of the S. cerevisiae peripheral stalk. The S. cerevisiae stalk could be extended in a manner similar to the bacterial stalk. The result indicated considerable plasticity within the eukaryotic peripheral stalk. The current structural and functional information is incorporated into a new peripheral stal k model.

PAGE 14

14 CHAPTER 1 BACKGROUND AND SIGNI FICANCE Introduction The F1FO Adenosine Triphosphate Synthase (ATP synthase) is the primary enzyme responsible for the generation of ATP in most organisms, including humans. ATP is the primary energy source in the c ell. The breaking of the high energy bonds in the molecule are used directly or indirectly to drive most of the energy requiring reactions in the cell. Examples include, but are not limited to, muscle contraction, biosynthesis of macromolecules, and nervous activity. Cells maintain ATP levels in the low millimolar range to perform necessary reactions for cell survival and growth. If ATP levels fall, cell growth is impaired and eventually the cell will undergo apoptosis. Logically, the enzyme that generat es ATP is extremely important and nearly ubiquitously expressed. The importance of this enzyme makes its mechanism the focus of extensive biochemical, genetic, and biophysical investigation. The mechanism of the ATP synthase was highly controversial. The level of controversy involving this enzyme has made for a very interesting history of the field of bioenergetics. There have been tremendous arguments over the mechanism of the enzyme. At one point in time the bioenergetics session of conferences were known for their entertainment value, involving arguments and sharpwitted jabs (Prebble, 2002). There were missteps in determining the mechanism of the enzyme, such as the intensive search for a phosphorylated intermediate that led many fine scientists astr ay (Schatz, 1996). ATP synthase is the terminal complex in the oxidative phosphorylation pathway. A F1FO ATP synthase is located in the cytoplasmic membrane of bacteria, the thylakoid

PAGE 15

15 membrane of chloroplasts, and the inner mitochondrial membrane of eukar yotes. For the sake of simplicity, most of the following discussion of the electron transport chain is about mammalian cells/mitochondria (Figure 11). However, there are some differences in the electron transport chain between bacteria, chloroplasts, and mitochondria. The electron transport chain in Escherichia coli is similar to mitochondria, but it lacks a Complex III and has two kinds of cytochrome c. In chloroplasts, there are three complexes (photosynthetic complex I, photosynthetic complex II, and cy tochrome b6f ). These complexes are connected by two mobile electron carriers called plastoquinone and plastocyanin (Voet and Voet, 2004). In mitochondria, the electron transport chain is composed of four complexes along with two additional electron carrie rs, quinone and cytochrome c. The chain transports protons to the intermembrane space of the mitochondria. As a result, the net transfer of protons the electron transport chain is responsible for generating a chemiosmotic gradient across the membrane. The ATP synthase is able to able to use the proton gradient and synthesize ATP from ADP and Pi. Coupling Adenosine Triphosphate Generation to Cellular Respiration The coupling of cellular respiration to the generation of ATP was demonstrated by Wladimir Engelhardt in the early 1930s (Engelhardt, 1982). It was well known at that time that ATP was formed from anoxidative breakdown of glucose through either glycolysis and/or fermentation. Additionally, the process of cellular respiration was being discerned by David Keilin amongst others. However, the connection between ATP synthesis and cellular respiration was not yet known. Engelhardt was interested in cellular respiration. He stated in his article Life and Science that he was always tinkering with thi ngs. He enjoyed the visible nature of

PAGE 16

16 cellular respiration. However, unlike Keilin and others, he used an unusual model for his studies. Most people used either S. cerevisiae or liver and muscle tissue because they were available in abundance. Engelhar dt, however, was at an institution, Kazan University in Russia, where nucleated avian blood cells were plentiful. These cells underwent unusually intense cellular respiration. Avian erythrocytes have intense respiration is due to an unusually large nucleus and the cells have a high concentration of ATP. This is unlike human erythrocytes that do not contain a nucleus. He found that as soon as cellular respiration was interrupted by cyanide poisoning or removal of oxygen by repeated evacuation and flushing with pure nitrogen, rapid dephosphorylation of ATP is observed, accompanied by a corresponding increase of inorganic phosphate (Engelhardt, 1982). This gave Engelhardt two options for interpretation of his data: (i) respiration stopped the dephosphory lation of ATP; or (ii) hydrolysis of ATP continued in the presence of respiration, but respiration compensated for this by making more ATP. The latter was more likely because there was no known process in which hydrolysis was stopped by the presence oxyge n alone. This indirect evidence was later confirmed by a direct demonstration that respiration in some manner compensated for the continual dephosphorylation of ATP, possibly by the generation of more ATP. Engelhardt introduced nucleated avian cells into anaerobic conditions. During this time, inorganic phosphate was produced, presumably from the hydrolysis of ATP. When the same cells were reintroduced to aerobic conditions, the inorganic phosphate underwent esterification, that is ADP was phosphorylated to yield ATP (Engelhardt, 1982).

PAGE 17

17 Cellular Respiration and the Electron Transport Chain The knowledge that ATP was a product of cellular respiration created interest in the field of bioenergetics (Friedkin and Lehninger, 1948). David Keilin was princi pally responsible for determining the order of the electron transport chain (Green and Tzagoloff, 1966). The mechanism that Keilin found was almost as we understand it today. However, the topological organization of the complexes would not be known until m uch later. By the early 1920s was already known that biological tissues were sensitive to cyanide (Keilin, 1925). Thunberg had shown that many organic substrates, such as succinic acid, could be oxidized by enzymes that are specific for a particular substr ate. These enzymes would later be called dehydrogenases. Wieland and Thunberg went on to postulate that the fundamental process is the activation of oxygen by an ironcontaining respiratory enzyme (Slater, 2003). In 1925, Keilin published a groundbreaki ng paper in Proceedings of the Royal Society entitled On cytochrome, a respiratory pigment, common to animals, S. cerevisiae, and higher plants (Keilin, 1925; Slater, 2003). This paper characterized the pigmented compounds of the cell. Keilin first show ed that the substance previously referred to as either myoor histo hematin was not limited to muscle cells or hemoglobin as previously thought, but was found even in unicellular organisms. The pigmented compounds characterized by Keilin are now known to be the ironsulfur and heme complexes that shuttle electrons in the electron transport chain. Keilin proposed the term cytochrome because there was evidence that the cytochrome was not a simple structure, but a complex one. Indeed, it was later shown t hat there was more than one form of cytochrome (Keilin and Hartree, 1955; Green et al. 1959; Green and

PAGE 18

18 Hatefi, 1961). The cytochromes that Keilin studied had very similar absorbance spectra across species. Cytochromes also exhibited a difference in the absorbance spectrum from the oxidized and reduced states. Further research showed that the cytochromes were oxidized in the resting state and became partially reduced in the active state. From this evidence, Keilin concluded that cytochromes were common to mammals, prokaryotes, plants, and fungi (Keilin, 1925). By 1939, Keilin was able to describe the order of cellular respiration as: dehydrogenase cytochrome b cytochrome c cytochrome a cytochrome a3 oxygen (Green et al. 1959; Slater, 2003) I n 1946, Hogeboom et al separated fractions from rat liver (Hogeboom et al. 1946) The research found that between 70 74% of the cytochrome oxidase activity was in the mitochondrial fraction. This observation led to the discovery that cellular respiration was located in the mitochondrial membrane. Then, in 1955, cytochrome c1 was added to the electron transport chain (Keilin and Hartree, 1955). Further work was done looking at the exact number of electrons transported. Eventually these experiments led t o the discovery of the complexes associated with the electron transport chain. It appeared that the known inhibitors of the electron transport chain, like cyanide, acted on different complexes. Using inhibitors, F.L. Crane in David Greens laboratory disc overed ubiquinone, which was later found to be both a hydrogen and electron carrier (Crane et al. 1957; Slater, 2003). In the 1960s, copper was found in cytochrome c oxidation (Beinert and Palmer, 1964) and ironsulfur centers in Complexes I and III (Beinert and Sands, 1960; Rieske et al. 1964). Today, it is known that the electron transport chain in eukaryotes is composed of four complexes: Nicotinamide adenine dinucleotide:Coenzyme Q oxidoreductase (also

PAGE 19

19 known as: NADH dehydrogenase or Complex I), Succinate:Coenzyme Q oxidoreductase (also known as: Succinate Dehydrogenase or Complex II), Coenzyme Q:Cytochrome c oxidoreductase (also known as: cytochrome bc1 complex or Complex III), and Cytochrome c oxidase (also known as: COX or Complex IV) (Voet and V oet, 2004). Complex I removes two electrons from NADH and reduces ubiquinone to ubiquinol. In the process of doing this, Complex I transports four protons from the matrix across the mitochondrial intermembrane (Figure 11). Complex II reduces ubiquinone to ubiquinol, as well, through electrons from FAD. Ubiquinol diffuses through the membrane to Complex III, which oxidizes the ubiquinol back to ubiquinone. Complex III uses the electrons from two ubiquinols to reduce cytochrome c located in the intermemb rane space. Additionally, Complex III releases four protons, from two ubiquinols, into the intermembrane space. Cytochrome c transports electrons to Complex IV. These electrons are used to reduce molecular oxygen to form water. At the same time, four pr otons are translocated from the matrix to the intermembrane space. The translocation of protons, in conjunction with the transport of electrons, sets up the electrochemical gradient that the F1FO ATP synthase uses to generate ATP from ADP and Pi. The Sear ch for the Phosphorylated Intermediate Now that it was known that ATP was coupled to the electron transport chain, the question became how was ATP synthesized. Many enzymes, such as kinases, were known to catalyze phosphorylation reactions. Generally, these kinases would take a phosphate group bound to either the kinase itself or another compound, and pass it to the substrate for phosphorylation of the substrate molecule. It was thought that ATP would be synthesized according to a similar mechanism and the phosphorylated

PAGE 20

20 intermediate hypothesis was born. It presumed that one of the complexes involved in oxidative phosphorylation would serve as a phosphorylated intermediate. The ATP synthase would then remove the phosphate group from that intermediate and use the phosphate group to phosphorylate ADP and generate ATP. The phosphorylated intermediate was thought to be some molecule in the electron transport chain because the electron transport chain had previously been shown to be linked to ATP synthesis (Engelhardt, 1982) This was seemingly a very logical, biochemical hypothesis. After a remarkable effort, it was felt in the bioenergetics field that perhaps the phosphorylated intermediate would never be found. In the words of Paul Boyer, Our basic knowledge of the chemistry involved does not appear adequate for the task, and the problem is likely to be with us for some time (Boyer, 2002). The problem of the phosphorylated intermediate was stated rather nicely in a paper by E.C. Slater in 1953, some 20 years after the discovery that the generation of ATP is coupled to the electron transport chain. The proposed chemical reaction was as follows: (1) AH2+ B + C A~C + BH2 (2) A~C + ADP +H3PO4 A + C + ATP (3) A~C + H2O A + C Where AH2 and B are close to each other in the respiratory chain and C is an additional factor required for the interaction, perhaps the elusive ATP synthase. The highenergy bond between A and C (A~C) is what gave the energy for ATP synthesis. In this paper, Slater proposed t hat the most likely candidate for the highenergy

PAGE 21

21 intermediate was the glceraldehydephosphate3 phosphate dehydrogenase, because phosphorylation that occurred there (Slater, 1953). The Chemiosmotic Hypothesis There were several problems with the search for the phosphorylated intermediate. The largest was that the intermediate was proving to be rather elusive. In addition, the phosphorylated intermediate did not explain why the coupling of phosphorylation took place at the membrane. There were problems w ith stoichiometry. Cellular respiration changed with stress and exertion, but the phosphate:oxygen ratio did not vary. A very puzzling concern was that hydrolysis of ATP outside the mitochondria, caused changes within the mitochondria and uncoupling occurred with a variety of chemically distinct inhibitors by apparently differing mechanisms. Finally, there was swelling and shrinking of the mitochondria associated with oxidative phosphorylation (Slater, 1953; Mitchell, 1961). These questions drove Peter M itchell to look for alternatives to the phosphorylated intermediate. Taking inspiration from other systems, such as ion uptake in plant roots and acid secretion in the stomach that showed a connection between ion movement and metabolism, Mitchell proposed that a similar mechanism was involved in oxidative phosphorylation (Prebble, 2002). Mitchells model, which was subsequently called the chemiosmotic theory, proposed that the electron transport chain was coupled to oxidative phosphorylation via an ATP synt hesis mechanism (Mitchell, 1961). That is the energy stored in the proton gradient generated by the electron transport chain could be used to drive the formation of ATP (Mitchell, 1961, 1975, 1979). This was a very controversial idea. First, there was l ittle experimental evidence supporting the chemiosmotic theory. Indeed, Mitchells 1961 paper showed no new

PAGE 22

22 experimental evidence for the theory. Second, many scientists had spent the majority of their careers searching for the phosphorylated intermediat e. The concept of the phosphorylated intermediate was very entrenched in the field at this time. This resulted in many heated debates. It was reported that one bioenergeticist got so angry during a discussion that he was seen hopping around on one foot. Other reports stated that scientists attended the the oxidative phosphorylation sessions of the Federation meetings because they knew a good punchup was on the cards (Prebble, 2002). What followed were many years of researchers working hard to test Mitchells theory. Meanwhile, work was proceeding on the ATP synthesis system. Maynard Pullman, a researcher in Effraim Rackers group, was interested in what factor synthesized ATP. It was observed that ATP was hydrolyzed to ADP and Pi. Pullman hypothesized that the same factor that hydrolyzed ATP might be responsible for its synthesis by the reverse mechanism (Engelhardt, 1982). To test this the proteins that coupled ATP synthesis to the electron transport chain were solubilized using a mechanical shaker with glass beads. This was reportedly a rather frightening machine that, when in use, shook the entire building (Schatz, 1996). Following recovery by an ultracentrifugation step, these submitochondrial particles were capable of respiration, but failed to synthesize ATP. However, if the supernatant was added back, ATP synthesis was restored (Pullman et al. 1960; Zalkin et al. 1965). Further work supported the hypothesis that the same factor was responsible for both ATP synthesis and hydrolysis. Fractionation of the supernatant in the above experiment allowed for identification of the fraction that restored ATP synthesis. The factor that restored coupling of ATP synthesis to respiration was called coupling factor one or simply F1

PAGE 23

23 (Pullman et al. 1960). Later it was found that there was a membrane bound sector that was sensitive oligomycin and that sector was given the designation of FO for oligomycin sensitive coupling factor (Kagawa and Racker, 1966a). A key experiment in support of the chemiosmot ic theory was one performed by Effraim Racker and Walther Stoeckenius (Racker and Stoeckenius, 1974) The researchers reconstituted F1FO ATP synthase and bacteriorhodopsin into liposomes. If the chemiosmotic theory was valid, then the protons pumped out by bacteriorhodopsin would flow back in through the reconstituted F1FO complex and generate ATP. This was indeed the case and showed that the F1FO ATP synthase could take advantage of a proton gradient to synthesize ATP. Through this and other work by various researchers eventually Mitchells hypothesis became more accepted. In 1974, E.C. Slater, formerly a major proponent of the highenergy phosphorylated intermediate theory, declared the search over and recognized that the chemiosmotic theory was a l ikely explanation (Prebble, 2002). Finally, Mitchell was recognized with the Noble Prize for his chemiosmotic theory in 1978. The Binding Change Mechanism Now, that the F1FO ATP synthase was identified there was the question of how the enzyme worked. In 1972, Paul Boyer was sitting in an afternoon seminar that he did not quite understand and began thinking about the enzyme (Boyer, 2002) Boyer had spent a lot of time searching for the phosphorylated intermediate. During the pursuit, Boyers group used many different systems including both bacteria and plants. The wide variety of model systems aided in his formulation of the Binding Change Mechanism (Figure 12) (Boyer, 2002). A major experimental approach in the

PAGE 24

24 lab was using isotope exchange rates, especially O18 and P32, to study how enzymes worked. A phosphorylated histidine residue in a protein was found that looked to be a very promising possibility for the phosphorylated intermediate. However, it turned out to be the phosphorylated intermediate for substrate level phosphorylation. Substrate level phosphorylation is the phosphorylation of ADP that occurs in conjunction with glycolysis. Further efforts to identify the phosphorylated intermediate were unsuccessful. Boyer still rejected the idea of the chemiosmotic coupling of ATP synthesis, favoring an energized state captured in conformational changes of proteins (Boyer, 2002). This led to the day in 1972 when Boyer let his mind wander in a seminar. It there struck him that maybe the energy was not used to bind Pi to ADP, but to release a tightly bound ATP from the active site of the enzyme (Boyer, 2002). This would explain one of the major problems mentioned by Mitchell in 1961. Specifically, that uncoupling agents all seemed to act via different m echanisms, and that few of those mechanisms seemed to be involved in an attack on the Pi oxygen. That is the Pi HOH oxygen exchange rate remains unchanged in the presence of uncouplers because that exchange is not directly involved in the coupling of the system (Figure 13). Boyer correctly hypothesized that the reversible binding of ATP at a catalytic site would protect those Pi oxygens. Richard Cross, a post doctoral fellow in Boyers lab, found further evidence that Pi oxygens were insensitive to uncoupling agents (Cross and Boyer, 1975). Cross and Boyer submitted a manuscript to The Journal of Biological Chemistry outlining this conformational change mechanism, but the journal rejected this paper due to lack of evidence (Boyer, 2002). Boyer then submi tted the paper to Proceedings of the

PAGE 25

25 National Academy of Science using the sponsored track in order to get it published (Boyer et al. 1973; Cross and Boyer, 1975). The binding change mechanism postulated that there were at least two, later found to be thr ee, sites of catalysis. These sites would exist in different conformational states. The Open state would allow ADP and Pi to enter the site (Figure 12, yellow). Following a conformational change, the site would be converted to the Loose state (Figure 1 2, blue). This would bring ADP and Pi into the correct orientation for forming the highenergy bond, although that bond would not be formed until the next conformational state. The next conformational change forms the Tight site and allows the ADP a nd Pi already in the correct orientation to bind to form ATP. The newly formed ATP is held tightly in the site (Figure 12, green). The catalytic site would then undergo another conformational change to return to the Open site, the newly formed ATP woul d be released, and the process could repeat itself (Boyer et al. 1973; Cross and Boyer, 1975). The rapid reversal of this process (i.e. the hydrolysis of ATP) in the presence of uncouplers would allow the Pi HOH oxygen change to continue, but would inhibit the exchanges involving ATP (i.e. Pi ATP and the ATP HOH exchange) (Figure 13) (Boyer et al. 1973). Boyers proposal of the binding change mechanism was a tacit acknowledgement that the electrochemical gradient could provide the driving energy for conformational change associated with ATP synthesis. Confirmation of the binding change mechanism was shown by the crystal structure obtained by John Walkers group in 1994 (Abrahams et al. 1994), which showed all three catalytic sites in different conformational states Physical rotation of within the F1ATPase during ATP hydrolysis was demonstrated by

PAGE 26

26 Noji et al providing the final confirmation of the essential features of the mechanism of the F1FO ATP synthase (Noji et al. 1997). As predicted by t he work of Boyer and many others the F1FO ATP synthase was found to have three catalytic sites (Boyer et al. 1973; Schuster et al. 1975; Recktenwald and Hess, 1977; Slater et al. 1979). The catalytic sites showed an unusually high degree of cooperativi ty. It was eventually shown that participation of alternating sites was mandatory for synthesis of ATP. Studies into unisite catalysis, which is only possible under very low, nonphysiological ATP conditions (<1 nM), showed that ATP binds with a very high affinity to the catalytic site, 1012 M1. However, the turnover rate was so slow at 104 s1 that unisite catalysis was unlikely to account for biological ATP synthesis. It was later shown that multisite catalysis exhibited a much faster turnover rate (a 106 fold increase) due to catalytic cooperativity (Cross et al. 1982; Grubmeyer et al. 1982). Consistent with the expected properties of oxidative phosphorylation there has been some lingering debate about whether trisite or bisite catalysis occurs wit hin the cell. Boyer originally envisioned the binding change mechanism as a bisite catalytic mechanism with alternating sites of catalysis (Kayalar et al. 1977). Studies involving the attachment of a gold beadactin filament to the gamma subunit can be interpreted as support for bisite catalysis (Noji and Yoshida, 2001; Yasuda et al. 2001). Adachi et al showed bisite catalysis by performing single molecule experiments with F1 using a Cy3ATP at a concentration of 150 nM (Adachi et al. 2007). However, under saturating ATP concentrations trisite catalysis was observed by using fluorescent probes to detect site occupancy (Weber et al. 1993, 1996; Lbau et al. 1998; Senior et al. 2000). It is likely

PAGE 27

27 that bisite and trisite catalysis is dependent upon the concentration of ATP. The experiments showing trisite catalysis were performed under conditions more similar to that found in the cell. Thus, it appears likely that trisite catalysis is the physiologically relevant mechanism. Subunit Composition of the F1FO ATP Synthase Determination of the subunits of the F1FO ATP synthase in bacterial, chloroplast, and mitochondrial enzymes was a major undertaking. The earliest work on the Escherichia coli enzyme was done using genetic complementation by Frank Gibson and colleagues (Downie et al. 1979). Mutants that could not grow on a nonfermentable carbon source were screened for the loss of F1FO ATP synthase. Those genes were confirmed by subunit overexpression assays (Foster et al. 1980). The final order of the unc operon genes that encodes all eight structural subunits of the F1FO ATP synthase was determined by Simoni and coworkers (Gunsalus et al. 1982). Shortly, thereafter Walker and colleagues reported the complete sequence of the unc operon (Walker et al. 1984). By this time, it was already clear that the mitochondrial enzyme had a number of additional subunits (MacLennan and Tzagoloff, 1968; Tzagoloff et al. 1968; Kanner et al. 1976). The subunit composition of the F1FO ATP synthases from prokaryotes to eukaryotes is largely similar and the mechanism of the enzyme is the same (Table 11). The E. coli F1FO ATP synthase, located in the cytoplasmic membrane, has a molecular mass of approximately 450,000 Da. The enzyme is composed of 22 subunits in a stoichiometry of 33 c10ab2. The eukaryotic enzyme, located in the inner mitochondrial membrane, has a slightly larger molecular mass of approximately 500,000

PAGE 28

28 Da. Accordingly, the mitochondrial enzyme is composed of 24 subunits and has a stoichiometry of 33 (OSCP) c10abdh. The mitochondrial enzyme has the following associated proteins: IF1 (inhibitory factor 1), f, e, and g that are not essential for ATP synthesis. The Saccharomyces cerevisiae enzyme has two other associated proteins, i and k. The additional proteins associated with the mitochondrial F1FO ATP synthase are involved in dimerization of the enzyme in the mitochondria. The dimerization and possible oligomerization of the mitochondrial enzyme may enhance energy transduction and aid in stabilization of the enzyme in the inner mitochondrial membrane (Arnold et al. 1998). It also appears that mitochondrial cristae morphology is dependent not only upon the dimerization of the enzyme, but organization of those dimers (Arselin et al. 2004; Fronzes et al. 2006; Strauss et al. 2008). Catalytic and Rotational Domains Early electron micrographs clearly showed F1 attached through the membrane via a stalk like structure, the overall structure was reminiscent of a knob on a stick (Gogol et al. 1987). The knobs coul d be removed through ultracentrifugation and were found, when in the supernatant, to have ATPase activity (Kagawa and Racker, 1966b). The same sort of ATPase activity was found in a variety of organisms and the subunit composition of F1 was determined. The F1 sector was found to contain three copies of the and subunits, with single copies of other subunits in all species examined (Boyer, 2002). The F1 sector had a stoichiometry of 33 E. coli as opposed to 33 et al. E. coli is

PAGE 29

29 E. coli subunit (Figure 14) (Walker et al. 1987). About the same time the binding change mechanism emerged, Catterall et al proposed that the bulk of the catalytically active F1 ATPase had a stoichiometry of 33 (Catterall et al. 1973). This proved universally true of all F1ATPases. The 2.8 structure o f the F1 sector of the Bos taurus F1FO ATP synthase in the presence of physiological Mg2+ confirmed the 33 organization of the F1 sector (Abrahams et al. 1994). The structure showed the and subunits were in an alternating hexamer with the subuni t extending through the middle of the hexamer. The catalytic sites were located at three / interfaces and each catalytic site was in a different conformation. Three additional noncatalytic nucleotide binding site were found at the other / interfaces. Importantly the subunit participated in unique proteinprotein interactions with each / pair. If rotated each / pair would sequentially pass through differing conformations as predicted by the binding change mechanism. The and subunits The and subunits display considerable homology between the E. coli, B. taurus and Saccharomyces cerevisiae F1FO ATP synthases. The subunit is larger than the subunit and has higher molecular mass (Table 12). The E. coli subunit is highly conserved with 72% sequence identity to both the Homo sapien and B. taurus subunit, according to a BLAST search (Altschul et al. 1997, 2005). The high degree of homology is likely due to the similarity in catalytic mechanism amongst the enzymes. The subunit of the E. coli enzyme is not as highly conserved with 57% sequence identity to both Bos taurus and H. sapein subunit, according to a BLAST search (Altschul et al. 1997, 2005). According to crystal structures of the S. cerevisiae and B.

PAGE 30

30 taurus F1 sector, the nonconserved amino acids generally lie on the surface F1 and do not participate in the catalytic mechanism (Abrahams et al. 1994; Gibbons et al. 2000; Mueller et al. 2004; Puri et al. 2005). In S. cerevisiae strains that are null for the or gene, the cells cannot grow on a nonfermentable carbon source. Still, less than one percent of these strains went or 0 (Lai Zhang et al. 1999). The terms or 0 refer to either cells that have damaged or missing mitochondrial DNA, which can occur w hen F1FO ATP synthase subunit genes are defective or deleted. Interestingly, the B. taurus subunits can complement the null or gene strains (Lai Zhang and Mueller, 2000). This not only speaks to the extent of structural homology, but also indicates that the S. cerevisiae cells can express and import the B. taurus and subunits into the mitochondria. An additional experiment investigated the ability of the B. taurus subunits to complement a S. cerevisiae strain. The B. taurus subunits initially could not complement the quintuple knockout S. cerevisiae strain However, the daughter cells of the S. cerevisiae strain expressing the B. taurus subunits grew on a nonfermentable carbon source indicating complementat ion. Further analysis of the daughter cells showed that mutations occurred in genes outside of the F1 sector and not in the B. taurus genes. This result showed that the B. taurus F1 subunits could be incorporated into active complexes with the S. cerevis iae FO sector (Puri et al. 2005). In the 2.8 B. taurus crystal structure the F1 sector is 80 high and 100 across with a dimple at the top of the structure (Abrahams et al. 1994). The remarkable feature of this highresolution crystal structure is that it showed all three catalytic sites in different conformations (Abrahams et al. 1994). The S. cerevisiae 3.43

PAGE 31

31 F1c10 enzyme showed most of the features of the B. taurus F1 and added much of the central stalk (Dautant et al. 2010). In both F1 sect ors, the catalytic sites are largely housed in the subunits and the noncatalytic sites are mostly in the subunits. Abrahams et al named the three subunits, based on conformation: TP, DP, E. The TP and DP sites contained the ATP and ADP molecules, respectively, and the E site was lacking a nucleotide (Abrahams et al. 1994). In the S. cerevisiae F1c10 structure, an unusual feature was that the conformation of the TPTP pair was similar to that of the DPDP pair found in the B. taurus F1 ( Gibbons et al. 2000; Dautant et al. 2010). Additionally, the DP/ DP pair in the S. cerevisiae structure was more open than those found in the B. taurus structure (Dautant et al. 2010). The authors hypothesize that this structure is of the ADP inhibit ed form of the enzyme and is probably occurs during the resting state when the enzyme is waiting to be activated by the proton motive force (Dautant et al. 2010). The rotor stalk The rotor stalk in E. coli F1FO ATP synthase is composed of the and proteins. In the mitochondrial enzyme, the rotor stalk consists of the subunits, with the E. coli subunit (Table 11, 12, 13) The sequence similarity between the E. coli and B. taurus subunits is 29 % (Altschul et al. 1997, 2005). The mature S. cerevisiae subunit is slightly longer than the B. taurus subunit at 278 amino acids, compared to 273 amino acids in the B. taurus subunit, and has a slightly higher molecular mass (Table 13) (Consortium, 2010). The subunit of the E. coli mitochondrial enzyme with sequence similarity of approximately 26 %. The E. coli subunit is an inhibitor of the enzyme with

PAGE 32

32 a Ki of 10 nM (Sternweis and Smith, 1980). Truncation of the subunit showed that the carboxyl terminus of the protein contained the necessary amino acids for the inhibition of the enzyme (Kuki et al. 1988). It is presumed that the S. cerevisiae the same function as the subunit in the E. coli enzyme. The subunit in the mitochondrial enzyme has no apparent equivalent subunit in the E. coli enzyme. The rotor stalk is necessary for assembly and function of the F1FO ATP synthase. S cerevisiae with a null mutation for either the fermentable carbon source and 100 % of the cells went either or 0 (Lai Zhang et al. 1999). The S. cerevisiae null mutant was still capable of slow growth on a nonfermentable carbon source. The B. taurus subunits and complemented null mutations in S. cerevisiae for the or subunits, respectively. However, the B. taurus S. cerevisiae (L ai Zhang and Mueller, 2000). Interestingly, in the S. cerevisiae ) the B. taurus B. taurus F1 subunits, complemented the deletion strain (Puri et al. 2005). The high resolution crystal structures of the B. taurus and S. cerevisiae F1 sector showed that the conserved amino acids lie at the interaction sites between the rotor stalk and the 33 hexamer (Dautant et al. 2010). Since the location of the conserved sites were mainly i n sites of subunit subunit interaction, this is likely what allowed the B. taurus subunits ( ) to complement the S. cerevisiae ) (Abrahams et al. 1994; Gibbons et al. 2000; Mueller et al. 2004; Puri et al. 2005). The 2.8 crystal structure of the B. taurus F1 sector showed the subunit as 90 long extended through the center of the 33 hexamer (Abrahams et al. 1994). The

PAGE 33

33 central stalk is similarly shown in the 3.43 S. cerevisiae F1c10 crystal structure (Figure 1 5) (Dautant et al. 2010). The S. cerevisiae structure revealed the subunit to lie along the narrowest channel in the 33 ring. This channel is lined with prolines at P290, P291, and P276, which interact with the conserved G273 (Figure 15A). Below this region, the amino and carboxyl termini of the subunit form a coiledcoil. It shou ld be noted that the long helix of the carboxyl terminus of the S. cerevisiae subunit has a much higher curvature than the B. taurus subunit. At the base of the subunit, the globular domain interacted with the cctions with both the and subunits forming a foot like structure atop the cring. The amino sandwich, with which the second helix of the carboxyl terminal domain interacts (Figure 15B). The subuni t interacts subunit through its amino terminal helix (Dautant et al. 2010). Work by McCarty et al showed that the modification of a thiol group in the E. coli subunit by N ethylmaleimide inhibited the enzyme (McCarty et al. 1972) Additional work by Duncan et al showed that in the E. coli F1FO ATP synthase, covalently crosslinking a single subunit to the subunit also inhibited the enzyme, and that breaking the crosslink reversed the inhibition (Aggeler et al. 1995; Duncan et al. 1995). By far, the most impressive experiment was performed by Noji et al In a novel experiment, the 33 ring was fixed to a Ni NTA coated coverslip via histidine tags on the subunits. Then, a fluorescently labeled actin filament was attached to the streptavidin conjugated subunit. In the presence, and only in the presence, of 2 mM ATP, the actin filament rotated in a counter clockwise direction (Noji et al. 1997; Noji and Yoshida, 2001). Work by Walkers group showed that the subunit inter action with

PAGE 34

34 each catalytic site dictates its conformation (Leslie et al. 1999). By 2000, it was well established that the binding change mechanism was correct and that rotation of the subunits in E. coli or the catalytic sites. The question was no longer if the central stalk rotated, but how did the central stalk rotate. With three catalytic sites and three conformational states, it would make sense that each catalytic site would completely transition through each state per 360 rotation of the subunit. This would mean that for every 120 rotation of the subunit, each catalytic site should change into a different conformation. In 2001, Yas uda et al showed that this was generally correct (Yasuda et al. 1998, 2001). However, it turns out that each 120 step is not a smooth transition, but occurs in two smaller steps. First, there is an ATP waiting dwell, followed by an 80 rotation. Nex t, there is another dwell called the catalytic dwell followed by a rotation of 40 completing the 120 movement of the subunit (Nishizaka et al. 2004; Sielaff et al. 2008). Biochemical experiments extended and confirmed the rotational studies. For example, in the E. coli enzyme the C87 was covalently crosslinked to either D380C or E381C. When the subunit was crosslinked to the subunit the enzyme retained the hydrolysis, but failed to release the Pi. This indicated that the ratelimiting ste p for rotation occurs after hydrolysis, but before Pi release (Baylis Scanlon et al. 2008). The binding of ATP to the catalytic site increases with forward rotation of the rotor stalk (Iko et al. 2009). The central stalk adopts a different position duri ng the ATP waiting state during catalysis than in the MgADP inhibited state (Hirono Hara et al. 2005). Recently, the torsional stiffness of the internal portion of the rotor stalk (i.e. the

PAGE 35

35 subunit embedded in the 33 ring) and the external portion of the rotor stalk in the 33 complex (i.e. the portion of outside of the ring) was measured. The internal portion had a torsional tension of 223 pNnm/radian and the external portion was measured to be 73 pNnm/radian. The difference in torsional tensio n shows the elasticity of the rotor stalk and may account for the smooth transference of torque (Okuno et al. 2010) (Waechter et al submitted for publication). Proton Translocation Domain Proton translocation occurs through the FO sector. The sector has a stoichiometry of ac10 b2 in E. coli and ac10b dh (or F6) in eukaryotes. The proton translocation ability of FO is dependent upon the cring, a subunit, and only the membrane domain of the b subunit (Greie et al. 2004). The a subunit is thought to hous e two proton half channels (Vik and Antonio, 1994; Junge et al. 1997). A proton from the mitochondrial intermembrane space enters through one half channel and protonates a single c subunit. Nearly simultaneously, a different single c subunit is deprotonated and that proton leaves through the second half channel in the a subunit exiting the FO sector on the opposite side of the membrane. These protonation and deprotonation events generate torque, which drives rotation of the central stalk. The main role of the peripheral stalk, is to hold the and subunits of F1 in place against the rotation of the central stalk. The a subunit Maximum conductance through the FO proton channel is 10 fS or approximately equivalent to 6,500 protons per second at an elec tric driving force of 100 mV (Feniouk et al. 2004). However, torque is not generated by the protons entering into the a

PAGE 36

36 subunit. The net movement of protons through the a subunit drives rotation of the ring of c subunits. Proton translocation appears t o be obligately coupled to not only ATP synthesis, but ATP hydrolysis, as well. S everal site directed mutagenesis studies showed that the proton channel is largely housed in the membraneembedded FO a subunit (Cain, 2000). The lack of proton conductance through FO due to a lack of a functional a subunit suppresses the ATP hydrolysis activity of the F1 sector (Gardner and Cain, 1999; Ono et al. 2004). As of the writing of this dissertation there is no highresolution crystal structure for the a subunit of t he F1FO ATP synthase. The majority of the information for the topology of the a subunit comes from mutagenic analysis and a few biochemical studies (Figure 16). The E. coli a subunit is thought to be composed of five transmembrane domains (Long et al. 1998). These transmembrane domains were predicted through hydropathy plots, mutagenesis, and chemical reagent accessibility studies. For example, individual cysteine residues were introduced into the a subunit, which contains no native cysteines, at many positions. It should be noted that in the wild type protein ends with aH271 and the cysteine at a277 was engineered for these particular experiments. Introduction of those cysteine substitutions had no apparent effect on enzyme function. The enzyme was r eacted with the polar labeling agent 4 acetamido4 maleimidylstilbene 2,2 disulfonic acid (AMDA) on the F1 side of the membrane. This polar reagent should only react with the thiol groups on cysteines that are accessible to the cytoplasm. Following the tr eatment with the AMDA, the same enzymes were reacted with the less polar 3(N maleimidylpropionyl) biocytin (MPB) that could react with the cysteine residues on the opposite side of the membrane from F1.

PAGE 37

37 The a subunits were analyzed to see if the AMDA prevented the thiols on the introduced cysteines from reacting with the MPB. The cysteines at positions aS69C, aG172C, aK176C, and aC277 could be blocked from reacting with MPB by AMDA, suggesting these residues were in the cytoplasm. In contrast, the cystei nes at positions aP8C, aR24C, aS27C, and aE131C could only be labeled with MPB, indicating that these residues were in the periplasm (Long et al. 1998). Through this and other studies, the two cytoplasmic loops were identified. The first cytoplasmic loop, L12 ( a6599), apparently interacts with the b subunit (Wada et al. 1999). A cysteine introduced into position 74 ( aK74C) was crosslinked to the b subunit using a photoactivatable reagent (Long et al. 2002). There was no evidence that the second cytoplasmic loop, L34 ( a167202) was near the peripheral stalk subunits. Crosslinking experiments suggested that transmembrane helices 2, 3, 4, and 5 formed a four helix bundle with transmembrane helix 4, which contains the key arginine at position 210 (discus sed further below), at the ca interface (Figure 16). Other key amino acids involved in the proton channel (discussed below) on transmembrane helices 2, 4, and 5 were found to face the interior of the four helix bundle (Schwem and Fillingame, 2006). Spin labeling further supported this interpretation (Dmitriev et al. 2007). Spin labels were placed along transmembrane helices 4 and 5, and NMR was used to determine the location of these residues relative to one another. The results suggested that the transmembrane helices 4 and 5 would fold back on themselves. Mutations affecting aR210 located on helix four of the E. coli a subunit resulted in the loss of proton translocation. Several groups have characterized different substitutions for aR210 that reproduced the observation (Lightowlers et al. 1987; Cain

PAGE 38

38 and Simoni, 1989; Vik and Antonio, 1994). An experiment demonstrated that this arginine is not position dependent and can be moved to position 252 ( aQ252R), which is on helix five and faces the aR210 on helix four with retention of partial activity (Hatch et al. 1995). The glutamine at aQ252 can be substituted by a lysine and at least some partial activity was detectable (Ishmukhametov et al. 2007). The c subunits The number of c subunits varies from eight to fifteen depending on species (Junge et al. 2009; Matthies et al. 2009; Vollmar et al. 2009; Watt et al. 2010). The c subunits are arranged in a ring. Based on the highresolution crystal structure of the c ring from Spirulina platensis each c subunit is composed of two helical segments that form a hairpinlike structure (Pogoryelov et al. 2009). The overall structure of the ring of c subunits is reminiscent of an hourglass. There is an essential, conserved negatively charged amino acid in the middle of the hourglass. In the case of S. platensis, it is cE62 and in E. coli it is cD61. This negatively charged amino acid is thought to be the site of protonation and deprotonation in the c ring. The majority of the work done on proton transloc ation through the FO sector has been performed on the prokaryotic enzyme. For the sake of clarity, the following discussion will focus on the prokaryotic enzyme. The mechanism of proton translocation is based on the concept that the ring of c subunits carr ies out Brownian rotational fluctuations relative to the a subunit (Junge et al. 1997). In the ATP synthesis direction, a proton from periplasmic side of the membrane enters through a half channel in the a subunit (Figure 17). The entering of this proton interrupts an electrostatic interaction between a positively charged amino acid in the a subunit ( aR210 in E. coli) and negatively

PAGE 39

39 charged amino acid in a single c subunit ( cD61 in E. coli or cE62 in S. platensis ) (Vik and Antonio, 1994; Junge et al. 1997; Pogoryelov et al. 2009). Structural studies in S. platensis suggested that protonation of the c subunit causes the c subunit to assume the proton locked conformation (Figure 18) (Pogoryelov et al. 2009). Essentially simultaneously, a different c subunit that is protonated enters the more hydrophilic, water accessible environment of the second proton half channel. This triggers the c subunit to assume an outward orientation and favoring release of the proton on the cytoplasmic side of the membrane. The release of the proton allows the negatively charged amino acid on the c subunit to again enter into a salt bridge with the positively charged amino acid on the a subunit (Pogoryelov et al. 2009). In E. coli, at least, it appears that there are other important, but not necessarily essential, amino acids in the proton channel (Figure 16). Most of these positions are highly conserved. The newly released proton from the c subunit interacts with a negatively charged amino acid in the a subunit ( aE219 E. coli ). At the cytoplasmic surface, there is a histidine at position 245 that appears to facilitate the protonation and deprotonation of the glutamic acid at position 219 (Hartzog and Cain, 1994; Vik and Antonio, 1994). A role for the cytoplasmic loops L12 and L34 has been proposed for further proton translocation. The residues located on the two loops, L12 and L34, are sensitive to inhibition by Ag+(Moore et al. 2008) This indicates that that the loops may be involved in proton translocation and perhaps, due to their location, involved in proton release into the cytoplasm The Peripheral Stalk The function of the peripheral stalk is thought to hold the F1 33 catalytic ring against the rotation of the central, rotor stalk. The E. coli peripheral stalk is composed of

PAGE 40

40 two identical b subunits that extend from the bottom of the membrane in FO, along one side of the enzyme, to the top of F1 where at least one b subunit. The mitochondrial peripheral consists of b dh(OSCP), wit h the h subunit being equivalent to the F6 subunit. The mitochondrial b subunit has no appreciable sequence similarity to the E. coli b subunit, even though the two are denoted by the same name. In fact, the b subunit of the E. coli enzyme has no appreciable sequence similarity with any of the mitochondrial peripheral stalk subunits (Altschul et al. 1997, 2005). The mitochondrial b subunit transverses the membrane twice, then extends along one side of F1 to bind to the OSCP. The OSCP of the mitochondrial enzyme is equivalent to the E. coli enzyme and has roughly 25% sequence similarity (Altschul et al. 1997, 2005). The h subunit interacts with the OSCP, b subunit, and d subunit (Rubinstein et al. 2005). It appears that the d subunit only int eracts with the h and b subunits and does not extend down to make contact with other FO subunits (Bueler and Rubinstein, 2008). As described above the highresolution structure of the F1ATP synthase was determined in 1994 (Abrahams et al. 1994). Many more structures of the F1ATP synthase followed. It was presumed that a structure of the entire complex was soon to follow. While there have been a plethora of structures involving the F1 sector, over fifteen years later there still not a structure of the entire F1FO ATP synthase of any organism (Kagawa et al. 2004; Mueller et al. 2004; Bowler et al. 2007). Therefore, much of what was known about the peripheral stalk in both the bacterial and mitochondrial enzyme was obtained through mutagenesis studies

PAGE 41

41 Mutagenic Analysis of the Bacterial and Eukaryotic Peripheral Stalks In 1998, both negative stained electron microscopy and cryoelectron microscopy images clearly showed two stalks in the E. coli F1FO ATP synthase complex. The central, broad stalk had previously been shown to be comprised of the and subunits. The thin peripheral stalk was presumed to be composed of the b2 (Wilkens and Capaldi, 1998a, 1998b). It was already thought that the bacterial b subunit was equivalent to the ATP4 subunit in S. cerevisiae and the b subunit in B. taurus (Walker et al. 1987; Velours et al. 2001). However, it became clear that the eukaryotic peripheral stalk was more complicated. The eukaryotic peripheral stalk is composed of the b d, F6 (or h in S. cerevisiae ) subunits, and the Oligomycin Sensitivity Conferring Protein (OSCP) (Walker et al. 1987). Mutagenesis of the Bacterial Peripheral Stalk Chemical labeling experiments from Hoppe et al showed that the amino terminus of the b subunit was embedded in the membrane and the carboxyl terminus extended above the membrane to contact the F1 sector (Hoppe et al. 1983). In 1984, Schneider and Altendorf found that the b subunit of the F1FO ATP synthase was critical for proton translocation and binding of F1 to FO (Schneider and Altendorf, 1984). Circular dichroism experiments in this region demonstrated that in liposomes the secondary structure of the b subunit consisted of 80 % helix and 14 % turns (Greie et al. 2000b). Porter et al reported the first singleamino acid substitution mutations that altered F1FO ATP synthase function (Porter et al. 1985). The researchers determined that the bG131D in the carboxyl terminus was critical for both proton translocation and F1 binding. At the amino terminus bG9D was critical for proton translocation, but not F1

PAGE 42

42 binding. Taken together these results suggested that the b subunits spanned the entire F1FO complex and played a role in coupling ATP synthesis to proton translocation. The b subunit is s ubdivided into four functional domains: (i) the membrane spanning domain, (ii) the tether domain, (iii) the dimerization domain, and (iv) the F1 (Dunn et al. 2000b). Membrane domain of the bacterial b subunit The membrane domain cons ists of the first 24 amino acids located at the amino termini of the b subunits that compose the transmembrane helix (Dunn et al. 2000b). The tolerance of the F1FO ATP synthase to mutations in the membrane domain of the b subunit was probed (Hardy et al. 2003). Mutations affecting residues towards the cytoplasmic side of the membrane had little effect on enzyme function. However, substitutions on the periplasmic side of the membrane showed defects in enzyme assembly. The exception to this was the bN2A,T6A,Q10A triple mutant that assembled into an oxidative phosphorylation defective F1FO ATP synthase. It is possible that the mutations towards the periplasmic side of the membrane interfered with either b b subunit or b a subunit interactions. In fact, the b subunit is proximal to at least on c subunit. A bN2C subunit was crosslinked to a c subunit at its carboxyl terminus (Jones et al. 2000). For each cysteine in the c subunit a maximum of 50% of the b subunit was crosslinked. This was likely because only one b subunit is in close proximity to the c ring. ATP driven proton pumping was stopped by a disulfide bridge between bN2C and cV78C. The importance of the observation was not clear because the disulfide bond did not inhibit ATP hydrolysis, indicating that the enzyme was uncoupled. A stable ab2 subcomplex was isolated without any crosslinks between the subunits (Greie et al. 2000a). This ab2 subcomplex could be reconstituted with the cring and F1. The

PAGE 43

43 reconstituted complex had wildtype ATP hydroly sis and proton translocation abilities (Stalz et al. 2003). Possibly, the mutations affecting sites towards the cytoplasmic side of the membrane did not yield phenotypes because the b subunits were not in contact with other proteins in the complex. A NM R structure and crosslinking data using a polypeptide modeling the b subunit membrane domain was interpreted as suggesting that the b subunits were in close contact with one another on the periplasmic side, but separated as the b subunits approached the cy toplasmic side (Dmitriev et al. 1999). A later NMR structure of soluble portions of the b subunit introduced by Steigmiller et al also suggested that the b subunits might continue to flare apart as they exit the membrane in the tether domain (Steigmiller et al. 2005). Thus, it is possible that the on the periplasmic side of the membrane the b subunits have close interactions, but not on the cytoplasmic side. Tether domain of the bacterial b subunit The tether domain of the b subunit is located from the top of the membrane bilayer to the bottom of F1 and consists of roughly b25 60 (Dunn et al. 2000b). There is an evolutionarily conserved arginine at position 36 in the bacterial F1FO ATP synthase (Caviston et al. 1998). When this arginine was substitut ed with an isoleucine, serine, or cysteine, the bacteria were unable to grow on minimal media supplemented with succinate and enzyme function was impaired to varying degrees. The most severe mutation was the bR36I substitution that exhibited no growth on a nonfermentable carbon source and ATP driven proton pumping activity similar to the negative control. It was found that the bR36E substitution was a true uncoupling mutation (Caviston et al. 1998). Both F1ATPase and FO proton translocation activities w ere essentially normal,

PAGE 44

44 but coupling between the two sectors was disrupted. Later work showed that the bR36 mutations could be overcome by an extragenic suppressor. Incorporation of two different b subunits, one with the bR36I or bR36E substitution and one wild type b subunit, formed functional, intact F1FO ATP synthases (Grabar and Cain, 2004). When the bR36C subunit was modified using a photoactivatable agent, irradiation resulted in crosslink formation with the first cytoplasmic loop, L12, of the a subunit (McLachlin et al. 2000). Although the data was never published, a review article indicated that bA32C could also be crosslinked to the a subunit (Greie et al. 2000a). Together with the genetic data, the evidence indicated that in one subunit the bR36 is likely to interact with the first cytoplasmic loop, L12, of the a subunit. The b subunit of the E. coli enzyme can tolerate large changes in length in its tether domain. Sorgen et al found that up to 11 amino acids ( bL54Q64) could be deleted from the tether domain of the b subunit and functional, intact F1FO ATP synthases could still be formed (Sorgen et al. 1998b). A follow up experiment found that the same region of the b subunit accepted as many as 14 amino acids ( bL54K67) and intact, functi onal enzyme was still formed (Sorgen et al. 1999). In either case, long insertions and deletions yielded fewer assembled F1FO complexes, but those formed appeared to be fully active F1FO ATP synthases. The difference between the largest total insertions a nd deletions represented a change in length of 36 in a region of the b subunit that is estimated to be only 45 long. To put this in perspective, that is an 80% change in length for this region of the b subunit. Additionally, the E. coli F1FO ATP synt hase can incorporate b subunits of unequal length into the enzyme complex. More specifically, the enzyme can incorporate one b subunit with a deletion of seven amino acids and one

PAGE 45

45 with an insertion of seven amino acids both in the tether domain ( bL54S60) into functional, intact F1FO ATP synthases (Grabar and Cain, 2003). This is a difference of approximately 18 between the two subunits within a peripheral stalk. Dimerization domain of the bacterial b subunit The demarcation between the tether and dimerization domains is not strictly defined. The dimerization domain consists of roughly residues b54 122 of the b subunit. A soluble portion of the b subunit between b54146 and was sufficient to form a dimer. This portion of the subunit was determined to be in an helix according to circular dichroism measurements and was capable of binding F1 (Dunn, 1992). One face of the helix is populated with hydrophobic residues, especially alanine (Priya et al. 2009). Further deletion work was done to show that the b53122 was sufficient to form dimers (Revington et al. 1999). Crosslinking experiments demonstrated that disulfide bonds were more readily formed between residues in the dimerization region than in any other domain, indicating that the b subunits were i n close proximity in this region (McLachlin and Dunn, 1997). In a fragment of the b subunit, b3082, it was possible to form crosslink between b subunit dimers at position 61, 68, and 72, but not 70 (Priya et al. 2009). These results corroborated the ext ensive crosslinking studies performed by Dunn and coworkers (McLachlin and Dunn, 1997; Wood and Dunn, 2007) Research using the entire soluble portion of the b subunit ( bsol) showed that bsol, A79L had a dimerization defect with only about 8% existing as a dimer (Sorgen et al. 1998a). Similarly, it was found that a substitution of a proline for bA79, bR82, bR83, or bQ85 caused impairment in enzyme function associated with assembly defects, with the largest defect found in the bA79P substitution (McCormick and Cain, 1991). Further

PAGE 46

46 substitutions involving bA79 found that activity of the enzyme was lost in the bA79K substitution, as well (McCormick et al. 1993). In this dimerization domain, the b subunit comes into close proximity with subunits in the F1 sector. A photoactivatable reagent anchored at bA92C was found to crosslink to both the and subunits. Similar experiments resulted in crosslink formation between bI109C and bE110C and the subunit (McLachlin et al. 2000). Moreover, attempts to inse rt or delete amino acids in the dimerization domain of the b subunit, bA121V132, abolished enzyme function (Bhatt et al. 2005). This is in stark that had little effec t. Interestingly, the defective b subunits were suppressed by pairing it with a fully functional b subunit (Bhatt et al. 2005). Apparently, only one normal length b subunit was required to form a functional peripheral stalk. The dimerization domain has been the primary focus of structural work and secondary structure predictions. A crystal structure of the monomeric fragment encompassing the entire dimerization domain, b62122, showed that segment to be in a 90 long helix (Del Rizzo et al. 2002). Pri ya et al showed an NMR structure of b22156 as an helix as well (Priya et al. 2008). However, this structure was problematic because the length of the b subunit structure was 16 nm long and the enzyme is thought to be slightly less than 15 nm long based on electron microscopy of intact F1FO ATP synthase ( Gogol et al. 1987; Lcken et al. 1990). This cast doubts about whether or not the structure was physiologically relevant. Later another NMR structure of b3082 also by Priya et al was shown to be 4. 8 nm in length (Priya et al. 2009). This would

PAGE 47

47 make the entire b subunit approximately 14.4 nm and its length is more reasonable in terms of the measured size of the F1FO complex from electron microscopy studies. The F1 rial b subunit Amino acids 123156 of the b subunit compose the F1 binding domain. A variety of mutagenic studies show that disruption of this sequence impairs binding of the b inal of the E. coli b subunit were deleted, the enzyme displayed no coupled activity. This was due to both impaired assembly and lost FO proton translocation (Takeyama et al. 1988; Bhatt et al. 2005). Crosslinking studies later showed that this deletion disrupted the binding of the b et al. crosslinked to the b subunit or a single subunit without a loss of enzymatic activity (Ogilvie et al. 1997; McLachlin and Dunn, 2000). This was particularly important because it clearly established that the peripheral stalk and the F1 33 subunits were stator elements that did not move relative to one another during catalytic activity. The site of interaction was apparently at the carboxyl termini of both the b subunits an d this made the case for b2 Chandler, 1998; McLachlin et al. 1998). To that end several cysteine mutations were generated, specifically bS84C, bA144C, bL156C. The bS84C and bA144C mutants covalently linked within F1FO complexes without the loss of enzyme activity. The bL156C subunit formed a disulfide bond with C90 at the top of F1, but activity was lost for reasons that are not clear (Rodgers and Capaldi, 1998). The crosslinking data showed that the b subunits must extend to the top of F1 to interact with the (Rodgers et al. 1997; Dunn et al. 2000a).

PAGE 48

48 Further evidence showed that binding of b2 to F1 altered the catalytic site (Kersten et al. 2000). Spin labeled Y331C had apparently normal catalytic activity. However, when a truncated form of the b subunit, b24 156, bound F1 the catalytic site underwent significant conformational changes by increasing the population of open sites. Single molecule fluorescence between F1 and FO reveal a Kd of 2.7 nM for F1 binding FO, but a Kd of 9 10 nM for F1 b inding either the ab2 subcomplex and b2 dimer (Krebstakies et al. 2005). Together with electron microscopy studies and genomic deletion assays, confirmed that both the peripheral stalk and the central stalk subunits were necessary for F1 to bind to FO (S chneider and Altendorf, 1984; Gogol et al. 1989) Chimeras of the bacterial b subunit Large portions of the b subunit tether and dimerization domains can be replaced with homologous portions of either the b or b subunits from Thermosynechococcus elongatus The b subunit that had the segment from amino acids bE39I86 replaced with homologous regions from the b or b subunit from T. elongatus grew better than and had higher levels of F1FO ATP synthase than the smaller replacements, bE39D53 or bL54I86. Expression of the b chimeric subunits yielded more activity than the b chimeric subunits (Claggett et al. 2007). More interesting experiments involved expression of both chimeric subunits together (Claggett et al. 2009). The chimeras of the E. coli b subunit with T. elongatus b and b subunits showed preferential crosslinking, during short incubation times, between TbA83C and TbA90C. This preferential crosslinking indicated that in the enzyme complex the b subunits favor a staggered conformation. Im portantly, trapping the peripheral stalk in the staggered conformation had no detectable effect on F1FO ATPase activity. This indicated that the staggered conformation is a fully

PAGE 49

49 functional arrangement. Under less stringent conditions, crosslinking between TbA90C and TbA90C occurred. Therefore, a dynamic equilibrium between staggered and parallel conformations may exist. Nevertheless, the favored conformation is the staggered one. Only one Tb b subunit participates with the key b Structural changes of the b2 1. ESR was used to look at the b2 dimer upon binding to th e F1 sector (Zaida et al. 2009). At room temperature, ESR showed a possible pivot point at position 62 due to increased motion around the amino acid. Frozen ESR spectra showed the formation of a bubble between b3364 that was too big for coiledcoil pac king. Right handed versus left handed coiledcoil debate Dunn and coworkers investigated a soluble form of the b subunit that included the region critical to dimerization, b53122. Sedimentation results from that study were most consistent with a coiledcoil packing of the b2 dimer (Revington et al. 1999). Later thermal denaturation indicated that the b2 dimer, at least from residues 53122, were in a noncanonical parallel coiledcoil (Revington et al. 2002). A controversy developed in the field over whether this is a right handed or left handed parallel coiledcoil. This question has been probed by crosslinking of model polypeptides and bsol, ESR experiments, and molecular modeling. However, none of these experiments has been conducted on the b2 di mer in the context of the entire enzyme. Investigations into the left handed coiledcoil configuration of the b2 dimer have been led primary by the laboratory of Dr. Pia Vogel and Dr. John Wise. The investigators have performed numerous ESR experiments on site specific spin labeled

PAGE 50

50 bsol to determine interspin distances. These interspin distances were then used in molecular modeling studies to examine the overall structure of the b2 dimer. The b subunits contain heptad repeats that are characteristic of left handed coiled coils. In these structures, the periodicity of the helix is reduced from 3.6 amino acids per turn to 3.5 amino acids per turn. Therefore, every seventh amino acid occupies the same face of the helix. The amino acids that make up thes e heptad repeats are denoted a g Amino acids a and d form the hydrophobic faces of interaction and b,c,e,f, and g from the hydrophilic outer surface (Lupas, 1997). However, in the view of the Wise and Vogel group, the heptad repeats had two stutters in t he E. coli b subunits between b31116 (Wise and Vogel, 2009). The two stutters consist of the deletions of three heptad repeat positions. Molecular modeling aided with interspin distance measurements from ESR showed that the E. coli b2 dimer could accomm odate these stutters by local deformations (Hornung et al. 2008; Wise and Vogel, 2009). T he heterodimeric, bb subunits, from the genetically engineered cyanobacteria, Synechocystis sp. PCC 6803 was spin labeled. The heterodimeric nature of the b subunits of this enzyme allowed for differentiation of the signal from each of the subunits. The results suggested that the two subunits had a 100 residue long helix and the carboxyl terminal of the bb complex was disordered. Through interspin distances, heptad repeats, and molecular modeling studies Volkov et al predicted that the bb subunits formed a conventional left handed coiledcoil (Volkov et al. 2008). Analysis of other bacterial, cyanobacterial, and plant b and b subunits found little sequence similarity, but there were apparent heptad repeats. Spin labeling of the E. coli bsol fragment correlated with molecular modeling left handed coiled coils (Hornung et al.

PAGE 51

51 2008). Hypothetical molecular models of right handed and left handed coiledcoils showed that the left handed coiled coil model is the most likely structural conformation (Wise and Vogel, 2008). The crystal structure of the b62122 monomer could be superimposed on top of the molecular models (Del Rizzo et al. 2002; Wise and Vogel, 2009). This demonstrated that the crystallized b62122 monomer could pack as a left handed coiledcoil. It was suggested that the stutters caused deformations that might be required for formation of functional F1FO ATP synthases (Wise and Vogel, 2009). Investigations into the right handed coiledcoil model have been led primarily by the laboratory of Dr. Stanley Dunn. The dimerization domain of the b subunit, b62122, was crystallized by Del Rizzo et al (Del Rizzo et al. 2002). It crystallized as a monomer in an helix that was 90 in length. The b62122 fragment was expressed in solution and analyzed by small angle x ray scattering (SAXS). The SAXS data showed that the b62122 dimer was found to be in an elongated helical coiled coil. Furthermore, the authors found evidence of a conserved undecad repeat that is indicative of a right handed coiledcoil (Lupas, 1997; Del Rizzo et al. 2002). The undecad repeat is made up of eleven amino acids labeled abcdefghijk There is a hydrophobic core, for coi led coil interactions, formed by amino acids at the a d e and h positions (Lupas, 1997). Dunn and coworkers interpreted the crystal structure, the SAXS data, and the conserved undecad repeat as the b subunit dimer being in a right handed coiledcoil (D el Rizzo et al. 2002). Additional work by Del Rizzo et al showed that after a 24 hour incubation crosslinks preferentially formed between cysteines introduced at position a of one b subunit and position h of an adjacent b subunit. Thermal denaturation of the crosslinked heterodimers and homodimers showed that the heterodimers were more

PAGE 52

52 stable. (Del Rizzo et al. 2006). The authors suggested this was further support for the right handed coiledcoil, with the caveat that proteins in the right handed coil ed coil conformation are rare in nature. Wood and Dunn engineered b53156 with cysteines in a variety of positions (Wood and Dunn, 2007). The researchers found that after a 72hour dialysis there was more efficient disulfide bond formation between heter odimers than homodimers. That is, the more efficient disulfide bond formation was between cysteines that were only proximal to one another in an offset configuration. The offset configuration was characteristic of right handed coiledcoils. Further evidence of this offset configuration came from b53156 dimers that were engineered to either be either aminoterminally or carboxyl terminally shifted. In these dimers, only the deletion of the last four amino acids from the aminoterminally shifted monomer weakened the b chimeric b subunit studies in the Cain laboratory were consistent with that interpretation (Claggett et al. 2009). Bi et al found that parts of the b subunit could be replaced by parts from homologous proteins that also had undecad repeats (Bi et al. 2008). Amino acids 55110 of the b subunits were replaced with sequences from both Bacillus subtilis and Thermotoga maritime and fully functional enzyme complexes formed. Replacements of b5595 from a more distantly related bacterium, the E subunit of the Chlamydia pneumoniae V type ATPase, and a completely unrelated protein, Ag84 from Mycobacterium tuberculosis also yielded functional enzyme. However, replacements with left handed coiled coil seque nces from the leucine zipper of the S. cerevisiae GCN4 inactivated the enzyme. However, it should be noted that the loss of activity in

PAGE 53

53 the GCN4/ E. coli b subunit chimeras the GCN4 used in the chimeras contains perfect heptad repeats. Wise and Vogel responded to this by suggesting that the two stutters found in the b subunit heptad repeats, which were not found in the GCN4 sequence, were necessary for enzyme function (Bi et al. 2008; Wise and Vogel, 2009). A recent, important highresolution structure has emerged of the A1AO ATP synthase from the distantly related organism, Thermus thermophilus (Lee et al. 2010) Its peripheral stalk, subunit E, appears to be in a right handed coiled coil. As of the writing of this dissertation, this is only one dimeric right handed coiledcoil found in nature (Lee et al. 2010). The right handed coiledcoil could provide a means for compensating for the torque generated by rotation of the rotor stalk. In the view of Dunn and coworkers, the right handed coiled coil of t he b subunits would tighten, supercoiling the subunits to compensate for the torque produced by the rotor stalk (Del Rizzo et al. 2006). two domains: an amino terminal six helix bundle and a carboxyl terminal domain. The two domains could be crosslinked via two native cysteines at positions C64 and C140 (Wilkens et al. 1997). It appeared that the binding of the b2 dimer to F1 was not strictly dependen in vitro (Motz et al. 2004). The soluble fragment of the b subunit, b22156, had single cysteine substitutions introduced evenly throughout and each of those cysteines was labeled with spin probe. The spectra of the ( b22146)2 di mer depleted F1 The ESR spectra indicated that the binding of the b2 1 was not ordered. Additionally, it appeared that from b80 to the carboxyl terminus, with some exceptions,

PAGE 54

54 the b subunit packs rather tightly to the F1 sector (Motz et al. 2004). This likely occurs by proteinprotein contacts with an / pair along a noncatalytic site interface. 1FO complex in vivo Very few mutations affecting the subunit caused binding defects with the F1 sector. For example, Stack and Cain generated an alanine scan mutation that altered Q23endQ29endQ74endQ150N, caused decreased levels of ATPase activity of the enzyme (Jounouchi et al. Q176end V174end, showed no detectable change in ATPase levels. Weber et al used tr yptophan 1 (Weber et al. A14D A14L, of the ten generated had large 1, and these did not prevent assembly of an intact, functional F1FO ATP synthase. Importantly, inclusion b34156 in an F1increased affinity for F1 ( The Kd for binding b34156 to F1 large Kd may re present some redundancy in the binding. Perhaps that is the reason b 1 (Weber et al. 2003a). Further tryptophan fluor escence experiments showed that upon binding to F1W28 fluorescence was enhanced by 50% and the Kd was 1.4 nM (Weber et al. 2002). The Kd was equivalent to 50.2 kJ/mol, which is the energy required to resist rotation of the central stalk.

PAGE 55

55 Only the f irst 22 amino acids of the d of 130 nM (Weber et al. 2003b). The 122 fragment was able to bind subunits that contained a variety of substitutions without any significant change in Kd. However, binding of 122 was impaired when the subunits contained substitutions l ocated in helices 1 or 5. Senior et al found that, while 122 was sufficient to bind in order for to bind the full length subunit, a complex containing the subunit with the other F1 su bunits ( 23) was required (Senior et al. 2006). The authors proposed a model where the 22 amino acids of the amino terminal of the subunit are sequestered until binding to the subunit. Upon binding the subunit, the 22 amino acids of the amino ter minal are exposed to allow the binding of stoichiometry (i.e. the reason why the stoichiometry is 3 33) because steric 33 ring (Senior et al. 2006). Mutagenic Analysis of the Mitochondrial Peripheral Stalk Velours et al isolated the FO sector of the S. cerevisiae F1FO ATP synthase and analyzed its subunit composition (Velours et al. 1987). The b subunit was christened the ATP4 subunit because it was the fourth protein in descending order of molecular mass. The other subunits of the S. cerevisiae peripheral stalk were found to be the d subunit, h (or F6 in B. taurus ) subunit, and OSCP (analo E. coli) (Collinson et al. 1994). Sequencing of the first 10 amino terminal amino acids of the b subunit demonstrated that the subunit underwent proteolytic processing. The b subunit gene in S. cerevisiae encodes 244 amino acids, but the mature protein contains only 209 amino

PAGE 56

56 acids, and has a molecular weight of 23,250 Da protein (Velours et al. 1988, 1989). Despite little structural similarity and no sequence identity, the b subunit of the S. cerevisiae enzyme is functionally homologous to the b subunit of the E. coli enzyme (Velours et al. 1988). There is a much higher degree of sequence similarity (44% sequence conservation) of the S. cerevisiae b subunit to the b subunit found in the B. taurus enzyme (Walker et al. 1987; V elours et al. 1988). Importantly, disruption of the ATP4 gene, and subsequent loss of expression, caused 70% of the cells to become (Paul et al. 1989). While the mitochondrial b subunit has not been subdivided into functional domains like the bacteri al b subunit, I propose a five functional domain scheme here: (i) transmembrane domain I, (ii) intermembrane space loop, (iii) transmembrane domain II, (iv) h and d subunit interaction domain and (v) the F1 and OSCP binding domain. Transmembrane domain I Transmembrane domain I is defined as the first 46 amino acids, S. cerevisiae b146, of the amino terminus. The absence of transmembrane domain I in the S. cerevisiae b subunit decreased the amount of g subunit present and there was no oligomerization or dimerization of F1FO ATP synthase complexes in the mitochondrial innermembrane. A surprising finding was that upon the deletion of transmembrane domain I, the mitochondria showed strange morphology. The cristae became more onionlike and no longer ruffled. However, these changes in morphology caused no apparent change in oxidative phosphorylation showing the F1FO ATP synthase was active in vivo A similar result was seen in the S. cerevisiae (Soubannier et al. 2002). Additional experiments showed crosslink formation between

PAGE 57

57 S. cerevisiae bK7C and bK14C subunits and the g subunit, using dithiobis(succinimidyl propionate) with a spacer arm of 12 (Soubannier et al. 1999). These experiments demonstrated that the transmembrane domain I was important for levels of subunit g and ATP synthase dimerization in the mitochondria. Intermembrane space loop The intermembrane space loop contains S. cerevisiae b4755 and is proximal to the a subunit. Cysteine residues were introduced into the apparent intermembrane space loop at S. cerevisiae bN47C, bL49C, bD54C, and bE59C. The introduced cysteines were labeled in intact mitochondria by N (7 (dimethylamino) 4 methyl 3 coumarinyl)maleimide (DACM). To confirm the location of the presumed b subunit hydrophilic loop, the thiol groups on the introduced cysteines were reacted with water soluble maleimide, 4acetamido4 maleimidylstilbene 2,2 disulfonic acid (AMDA). AMDA labeled bD54C and aC23. This reaction with AMDA prevented labeling of the thiol groups on bD54C and aC23 with another soluble maleimide, 3( N maleimidylpropionyl)biocytin (MPB). This led to the conc lusion that the bD54C was in the intermembrane space. A disulfide bond was observed between bD54C and aC23. This indicated that the b and a subunits were in close proximity to one another in the mitochondrial FO sector (Spannagel et al. 1998b). Spontaneous crosslinks were formed between bD54C or bE55C on a neighboring F1FO complex (Spannagel et al. 1998a). Deletion of the intermembrane space loop of the b subunit, S. cerevisiae b 55, caused defects in dimerization and oligomerization (Weimann et al. 2008). There were changes in mitochondrial morphology reminiscent of the deletion of the first transmembrane domain of the b subunit (Soubannier et al. 2002). The bN25C no longer crosslinked to gC75, but a new crosslink was formed between aC23 and bN25C. This was

PAGE 58

58 surprising because it reflected an apparently major change in the organization of FO. That is the amino terminus is accessible to the innermembrane space, instead of to the matrix. A new crosslink was also formed between eC28 and aC23. The authors suggested that the intermembrane space loop is necessary to position the amino terminus of the b subunit, the e subunit, and g subunit in FO. Transmembrane domain II Few subs titutions seem to affect transmembrane domain II, S. cerevisiae b56 75 (Paul et al. 1992; RazakaJolly et al. 1994). The chromosomal ATP4 gene in a haploid S. cerevisiae strain was replaced with five different ATP4 genes that contained mutations in trans membrane domain II. Out of the five mutant strains generated, only one had an affect upon the activity of the ATP synthase. The double mutation of bL68R, V69E exhibited no growth at 37C and slow growth at 30C with a lower yield. This strain also showed decreased sensitivity to oligomycin without changing sensitivity to DCCD (RazakaJolly et al. 1994). This observation indicated that the transmembrane domain II might have an effect on the conformation of the oligomycin binding site. The h and d subunit interaction domain The h and d subunit interaction domain, S. cerevisiae b76168, is the primary site of interaction between the b subunit, and the d and h subunits. A cursory study of S. cerevisiae b subunit interactions with the crosslinking agent, dithiobis(succinimidyl propionate) showed that the b subunit is within 12 of the h and d subunits in the S. cerevisiae F1FO ATP synthase (Soubannier et al. 1999). Upon closer examination, a crosslinking agent with a 9 spacer arm, pazidophenacyl bromide, showed that the carboxyl terminus of the b subunit came within 9 of subunits h and d. This agent generated crosslinks between the S. cerevisiae bK98C subunit and the h subunit.

PAGE 59

59 Additional crosslinks were formed between the d subunit and bK141C, bK143C, bK151C subunits (Soubannier et al. 1999). The F6 subunit in B. taurus (analogous to the h subunit in S. cerevisiae ) was crosslinked with both the B. taurus b subunit and d subunits (Joshi and Burrows, 1990). Subcomplex formation in reconstitution exper iments with B. taurus b d, and F6 subunits showed binary complex formation between a carboxyl terminal fragment that contained the d and h subunit interaction domain of the B. taurus b subunit and the d and F6 subunits (analogous to the S. cerevisiae h subunit) (Collinson et al. 1994). This demonstrated that the h and d subunits interacted with the b subunit proximal to its carboxyl terminus. The F1 and Oligomycin Sensitivity Conferring Protein binding domain The F1 and OSCP binding domain contains roughly b169209 of the S. cerevisiae b subunit. Collinson et al investigated the binding order and subcomplex formation in B. taurus of F1 with OSCP, b d, and F6 (Collinson et al. 1994). The researchers found that the only binary complex formed with B. tau rus F1 was F1OSCP. The OSCP could also form a binary complex with the b subunit. Various sized fragments of the b subunit were reconstituted with F1 and the OSCP. There were only substoichiometric amounts of the b subunit fragments in a complex with F1OSCP. However, when F6 was added stoichiometric amounts of the b79214 and b121 214 were found in a complex with F1OSCPF6. Once the d subunit was added, the complexes F1OSCPF6b d and F1OSCPF6b79214d were formed. Additional evidence showed that the amino terminus of the b subunit transversed the membrane twice. Quantitative sequencing confirmed that there was only one copy of the OSCP, F6, b, and d subunits per enzyme complex (Collinson et al. 1996). Crosslinking studies showed more specifical ly that the OSCP was proximal to both the and subunits (Joshi and Burrows, 1990). These

PAGE 60

60 crosslinking results demonstrated that the b subunit interacted with all subunits that compose the S. cerevisiae and mammalian peripheral stalk. A mutagenesis experiment by Paul et al showed that the removal of the last 10 amino acids from the carboxyl terminus of the S. cerevisiae b subunit caused the loss of ATP4 ( b subunit) and ATP6 ( a subunit) expression (Paul et al. 1992). The result was comparable to the removal of the carboxyl terminal amino acids of the bacterial b subunit. However, removal of the last eight amino acids revealed a temperature sensitive mutant that had wild type like growth at 30 C, but not 37 C. The substitution of a variety of polar residues in the membrane extrinsic portion of the b subunit yielded no change in growth or coupling. Further crosslinking experiments showed that the carboxyl terminus of the S. cerevisiae b subunit came within 9 of the OSCP and The S. cerevisiae bK174C subunit crosslinked to both and the OSCP. The S. cerevisiae OSCP had an additional crosslink with bK209C (Soubannier et al. 1999). Lysines toward the carboxyl terminus of the b subunit were substituted with cysteines. The cysteines were reacted with p azidophenacyl bromide, a crosslinking agent with a spacer arm of 9 The S. cerevisiae bK209C and bK174C subunits were crosslinked to the OSCP. The bK174C subunit was also crosslinked to the subunit (Velours et al. 2000). This showed that the b subunit interacts with both the subunit and OSCP in the F1 and OSCP binding domain. FRET studies were performed between the OSCP and the b subunit to consider their relative positions in the mitochondrial F1FO ATP synthase. These were performed by generati ng fusion proteins to the carboxyl termini of the S. cerevisiae OSCP and b subunit. This generated the fusion proteins, OSCP BFP11 and b YEGFP3. FRET was

PAGE 61

61 seen between the two proteins, and there was no change in FRET during ATP hydrolysis. The authors c oncluded from these results that the OSCP and b subunits did not move relative to one another (Gavin et al. 2003). However, the possibility that the two fluorescent proteins were too big to detect the relative movements of the OSCP and b subunits could not be ruled out. Mutations in the mitochondrial d and h/F6 subunits It is generally thought that the d and h (F6 in B. taurus ) subunits functionally substitute for the hydrophilic portion of one bacterial b subunit. The d subunit in the S. cerevisiae F1FO ATP synthase was found to have 22% sequence identity and 44% sequence similarity to the d subunit in the B. taurus enzyme. There are no equivalent subunits in bacterial F1FO ATP synthases. The d subunit in S. cerevisiae is 173 amino acids and the one in B. taurus is 160 amino acids (Walker et al. 1987; Norais et al. 1991; Walker et al. 1991). Deletion of the d subunit gene ( ATP7 ) in S. cerevisiae resulted in the lack of growth on a nonfermentable carbon source and lack of expression of the ATP6 gene (Norais et al. 1991). The disruption of the ATP4 gene decreased the amount of d subunit to 8% of wildtype levels (Norais et al. 1994). The F6 subunit in the B. taurus F1FO ATP synthase had no apparent homolog in S. cerevisiae until 1996 when the novel subunit h was found in S. cerevisiae The h subunit, expressed from the ATP14 gene, was 112 amino acids as synthesized and was processed to 92 amino acids. This is considerably different from the B. taurus F6 that is only 76 amino acids in the mature for m. The S. cerevisiae h subunit was discovered when a null mutant was unable to grow on a nonfermentable carbon source and displayed oligomycin insensitive ATPase activity (Arselin et al. 1996). Loss of the ATP14 gene also correlated with the lack of ATP6 protein. The h subunit may play a

PAGE 62

62 role in dimerization of the F1FO ATP synthase in the mitochondria (Fronzes et al. 2006). In accordance with this role of dimerization, atp14 null S. cerevisiae display the altered mitochondrial cristae morphology (Goyon et al. 2008). T he h subunit has only about 14.5% sequence identity with the B. taurus F6 subunit. This is not a high sequence identity, but it was found that the B. taurus F6 subunit complemented an atp14 null gene mutant (Velours et al. 2001). Inter estingly, S. cerevisiae with the B. taurus F6 subunit was fully capable of oxidative phosphorylation at 28C. An NMR structure showed that the B. taurus F6 subunit was composed of two helices loosely packed together by a hydrophobic core, connec ted by an unstructured linker (Carbajo et al. 2004). As the B. taurus F6 subunit was able to complement the lack of an h subunit in S. cerevisiae it was presumed that the F6 and h subunits had similar structures. Crosslinking studies performed between s ubunit h and the b subunit in S. cerevisiae indicated that the two proteins interact in an area proximal to the carboxyl terminus of the b subunit (Soubannier et al. 1999). Similarly, reconstitution studies with the B. taurus F6 and b subunits showed that the two interact towards the carboxyl terminus of the b subunit (Collinson et al. 1994). The F6 subunit was crosslinked to a single subunit, the subunit, the b subunit, and the d subunit. Trypsin accessibility studies suggested that F6 did not interact with the membrane embedded portion of FO (Joshi and Burrows, 1990). Furthermore, chemical crosslinks were detected between subunit h, d, and b subunits (Fronzes et al. 2003). The variety of contacts was in line with the NMR structure of the B. taurus F6 subunit. This led the authors to conclude that

PAGE 63

63 the h subunit in S. cerevisiae occupied the same space as the F6 subunit in B. taurus (Soubannier et al. 1999; Velours et al. 2001). Mutations in the mitochondrial OSCP subunit Localization of the domains of the OSCP began with Joshi et al (Joshi et al. 1997). The deletion of the last ten amino ac ids from the carboxyl terminus, OSCPK181 L190, of the B. taurus OSCP caused defects in both assembly and coupling of B. taurus F1FO ATP synthase. However, the last three amino acids, OSCPE188L190, could be deleted without any effect on the enzyme activity (Joshi et al. 1992). Like E. coli and S. cerevisiae, the polar amino acids found in the B. taurus OSCPK181 R187 were found to be important to coupling (Haz ard and Senior, 1994; Mao and Mueller, 1997). If the amino acids found in the B. taurus OSCPK181R184 were substituted with glutamine or alanine, there was no change in the helical content as seen with circular dichroism or coupling. However, if those amino acids were substituted with prolines, then coupling was lost and helical content decreased. The defective OSCP subunits could still interact with F1 (Joshi et al. 1997). The observation indicated that the fivehelix bundle retained its structure a llowing proteinprotein interaction with the subunit amino terminus. More recently, an NMR study showed that the adjacent, hydrophobic helices 1 and 5 of the OSCP interact with the hydrophobic amino terminal 25 amino acids of the subunit (Carbajo et al 2005, 2007). Cryoelectron microscopy localized the OSCP to the top of F1, where it extended 10 nm along the surface of F1 towards FO to meet with the peripheral stalk (Rubinstein and Walker, 2002). In summary, the properties of the nd mitochondrial OSCP subunit establish that both serve the same functional roles as the F1 sector peripheral stalk docking site in F1FO ATP synthases.

PAGE 64

64 Structures of the Peripheral Stalk At the outset of my dissertation research the only structures of the peripheral stalk subunits, were partial structures of isolated subunits or fragments of subunits (Del Rizzo et al. 2002; Carbajo et al. 2005). Highresolution NMR structures were available 6 subunit, (Wilkens et al. 1997; Carbajo et al. 2004). However, intact F1FO complex structures were limited to the lowresolution cryoel ectron microscopy studies (Gogol et al. 1987; Lcken et al. 1990; Rubinstein et al. 2003). In late 2006, a subcomplex of the B. taurus peripheral stalk was crystallized and the structure determined at 2.8 resolution (Figure 19) (Dickson et al. 2006) The structure contained amino acids 79183 of the b subunit, 3123 of the d subunit, and 570 of the F6 subunit. The first 78 amino acids, including both transmembrane domains and the carboxyl terminus ( b184214) were deleted from the b subunit polypept ide. Of the three proteins in the structure, the F6 subunit is the most complete. Out of the 76 amino acids that compose the F6 subunit, only six are not in the structure. About one quarter of the d subunit proximal to the carboxyl terminus was not resolv ed, as d124160 was not in the structure (Dickson et al. 2006). Importantly, several very highly conserved amino acids were absent from the d subunit structure. The d subunit interacts with the b subunit through one parallel and two anti parallel coiledcoil interactions in the middle of the b subunit (approximately b99162). As expected, the F6 subunit interacts more towards the carboxyl terminus of the b subunit. The NMR structure of the F6 subunit is very similar to the F6 in the x ray crystallography structure (Carbajo et al. 2004; Dickson et al. 2006). The subunit consists of two helices coupled by linker region. However, both helices appear to be extended and

PAGE 65

65 they do not interact with each other via their hydrophobic patches as originally proposed (Carbajo et al. 2004). Instead, it appears that the hydrophobic patches on each F6 subunit make contact with the b subunit (Dickson et al. 2006). The cryoelectron microscopy images showed the h subunit of S. cerevisiae which is analogous to the F6 subunit of B. taurus to be located alongside the peripheral stalk rather than at the extreme carboxyl terminus of the b subunit (Rubinstein et al. 2005). The high resolution structure is compatible with the cryoelectron micrograph and showed that the F6 subunit extends quite some distance along the b subunit alongside F1 (Dickson et al. 2006). A cryoelectron microscopy structure of the entire B. taurus F1FO ATP synthase with a resolution of 32 had been solved earlier (Rubinstein et al. 2003). The highresolution structure of the partial B. taurus peripheral stalk, the NMR structure of the OSCP amino terminus, and the S. cerevisiae F1c10 highresolution structure was docked Inside this low resolution structure of the F1FO complex (Stock et al. 1999; Carbajo et al. 2005; Dickson et al. 2006). In the docked structure, the peripheral stalk was almost perpendicular to the membrane and positioned the OSCP not directly on top of the F1 sector, but slightly to the side (Dickson et al. 2006). In 2008, a new 24 cryoelectron microscopy structure of the S. cerevisiae F1FO ATP synthase was solved. When the B. taurus partial peripheral stalk structure and the S. cerevisiae F1c10 structure were docked again inside this improved cryoelectron microscopy structure, the peripheral stalk had to curve to fit around the F1 sector (Figure 110) (Lau et al. 2008). Importantly, in both docked structures it appeared that the peripheral stalk narrowed to the width of a single helix towards the membrane.

PAGE 66

66 In 2009, W alkers group published a new highresolution structure of the B. taurus F1 sector bound to a partial peripheral stalk (Figure 111) (Rees et al. 2009). In this structure, the and subunits are very similar to the ground state structure (Bowler et al. 2007; Rees et al. 2009). The carboxyl terminal amino acids of the b subunit, b122207, were resolved. Only amino acids F6(5 25) and F6(3557) were resolved and this was the region that interacts with both the b and d subunits. The d subunit was seen i n three fragments, d3040, d6574, and d8591. In this structure, the curve of the peripheral stalk was much more pronounced. The structure also highlighted the possibility of a hinge region at bP146 (Rees et al. 2009). The more pronounced curve and the possibility of the hinge region supported a new model for how the peripheral stalk acts in the context of the entire F1FO ATP synthase. Peripheral Stalk Models All peripheral stalk models must account for fixed points of contact between the peripheral s talk and subunits located in both the F1 and FO sectors. Mutagenic and structural studies of the bacterial and eukaryotic F1FO ATP synthases have led to the formation of two competing views of how the peripheral stalk functions in the context of the enzyme. The most recent structure of the F1peripheral stalk complex allows for the formulation of a third model. The original two models were the rigid rod model and flexible rope model. The rigid rod model proposed that the peripheral stalk acts as a ri gid rod to hold the 33 hexamer in place against the rotation of the central stalk. The flexible rope model stated that the peripheral stalk is a highly flexible structure, possibly with the capacity to store the elastic energy from the rotation of the c entral stalk. The third developing model, proposed in this dissertation, is the hinge stiff model.

PAGE 67

67 The rigid rod model was developed because the main role of the peripheral stalk is to hold the 33 ring in place against the rotation of the central stalk. This model stated that the best way to accomplish this was for a stiff rod to hold the ring in place. This rod would be stiff enough to resist any torque from the rotation of the central stalk. The peripheral stalk would basically act like a clamp t o hold F1 in place. Initially this idea was supported by the 2006 highresolution structure of the B. taurus peripheral docked into the cryoelectron micrograph image of the F1c10 complex. In this position, the b subunit was perpendicular to the membrane with little or no possibility for flexibility while maintaining the fixed points of contact with F1 and FO (Dickson et al. 2006). The flexible rope model arose from mutagenesis studies of the E. coli peripheral stalk. The E. coli peripheral stalk could tolerate large changes in length. Up to two turns of an helix could be deleted and inserted into the peripheral stalk and the F1FO ATP synthase was essentially wildtype (Sorgen et al. 1998b, 1999). Even longer insertions and deletions yielded functional F1FO ATP synthases. Additionally, the peripheral stalk was able to incorporate b subunits of two different lengths into the same enzyme without abolishing activity of the enzyme (Grabar and Cain, 2003). The ability of the b subunits to adapt to large changes in length suggested substantial plasticity in the tether domain. However, the amino and carboxyl termini of the b subunits were not as tolerant of substitutions because of the need to maintain proteinprotein interactions with F1 and FO (Takeyama et al. 1988; Sorgen et al. 1998a; Hardy et al. 2003; Bhatt et al ., 2005). The hinge stiff model is newly formed based on work in this dissertation and the latest B. taurus F1peripheral stalk structure. This model takes into account that the E.

PAGE 68

68 coli peripheral stalk, while not rigid, is not as pliable as the rotor stalk (Waechter et al. submitted for publication). However, it also accommodates the high tolerance for a large degree of mutation in the tether domain. The hingestiff model, along with i ts implications, will be discussed further in Chapter 5.

PAGE 69

69 Table 11. Subunit composition of bacterial and mitochondrial F1FO ATP synthases. Escherichia coli Saccharomyces cerevisiae a Bos taurus Catalytic domain Rotor Stalk H + Translocation c c (ATP9) c Domain a a (ATP6) a Peripheral Stalk OSCP OSCP b b (ATP4) b d (ATP7) d h (ATP14) h Associated Proteins e e g g i k aGene names are in parenthesis. Note: some literatur e refers to the subunits as a product of their gene name. Table 12. The subunit composition of the Escherichia coli F1FO ATP synthase. Subunit Length a Molecular Mass b 513 (N/A) 55,222 459 (460) 50,325 287 (N/A) 31,577 138 (139) 15,068 a 271 (N/A) 30,303 c 79 (N/A) 8,256 177 (N/A) 19,322 b 156 (N/A) 17,264 aLength is expressed in amino acids. The precursor length is in parenthesis. bMolecular mass is in Daltons (Da) of the precursor protein.

PAGE 70

70 Table 13. S. cerevisiae F1FO ATP synthase subunit composition. Subunit Leng th a Molecular Mass b 512 (545) 55,394 478 (511) 51,256 278 (311) 30,649 61 (62) 6,612 138 (160) 14,555 c (ATP9) 76 (76) 5,850 a (ATP6) 249 (259) 27,943 OSCP 195 (212) 20,874 b (ATP4) 209 (244) 23,253 d (ATP7) 173 (174) 19,722 h (ATP14) 92 (124) 10,408 aLength is expressed in amino acids. The precursor length is in parenthesis. bMolecular mass is in Daltons (Da) of the mature protein (Velours and Arselin, 2000). Table 14. B. taurus F1FO ATP synthase subunit composition. Subu nit Length a Molecular Mass b 480 (528) 55,263 478 (511) 51,563 273 (298) 30,256 50 (51) 5,652 146 (168) 15,065 c 75 (136) 7,608 a 226 (226) 24,788 OSCP 190 (213) 20,930 b 214 (256) 24,669 d 160 (161) 18,562 F 6 76 (108) 8,958 aLength is expressed in amino acids. The precursor length is in parenthesis. bMolecular mass is in Daltons (D a) of the precursor protein.

PAGE 71

71 Figure 11. Mitochondrial electron transport chain. Shown are the complexes of the mitochondrial transport chain located in the inner mitochondrial membrane. Electron transport is shown by the black arrows. Proton trans location is shown by the green arrows. Ubiquinone and cytochrome c are shown in the orange and purple circles, respectively.

PAGE 72

72 Figure 12. Binding change mechanism. The binding change mechanism as proposed by Paul Boyer. The Open, Loose, and Tight sites are denoted by the letters O, L, and T. In ATP synthesis the mechanism goes from left to right and the opposite direction in ATP hydrolysis. Modeled after the Binding Change Mechanism proposed by Paul Boyer (Boyer, 2002). Figure 13. Binding change equation. The equation for the binding change mechanism shows that the oxygens that belong to the Pi are protected from inhibitors because uncoupling does not affect the Pi equilibrium. The circle encompasses the reactions that occur within the bi nding site. Based on Boyers Binding Change Equation (Boyer et al. 1973).

PAGE 73

73 Figure 14. Prokaryotic and eukaryotic F1FO ATP synthases. Panel A, shows the bacterial ATP synthase. Panel B, depicts the eukaryotic enzyme. Homologous subunits retain the same color.

PAGE 74

74 Figure 15. Rotor stalk interactions with F1 in S. cerevisiae Panel A shows the residue interactions of the subunit (cyan) with the and subunits (magenta and lavender). Residues P290, P291, P276, are colored in red and residue G273 is colored blue. Panel B shows the subunit, colored in cyan, interactions with the subunit, brown. Structures were rendered in MacPyMol from PDB number 2WPD.

PAGE 75

75 Figure 16. Topology of the a subunit. The a subunit is proposed to pri marily consist of five transmembrane helices. Areas boxed or circled in red denote the location of amino acids mentioned in the text.

PAGE 76

76 Figure 17. Mechanism of proton translocation. The a subunit from E. coli is shown on the left as blue helices. The important aR210 is shown as sticks and is colored orange. The other orange residues in the a subunit are mentioned in the text. On the right is a c ring from S. plantesis The important cE62 is colored in yellow. The two black cylinders represent the two proposed proton half channels. The a subunit was rendered in MacPyMol from PDB number 1C17. The ring of c subunits was rendered in MacPyMol from PDB number 2WIE with the aid of PISA. The composite figure was made in Adobe Photoshop.

PAGE 77

77 Figure 18. Mechanism of proton translocation in the c ring. The a subunit from E. coli is shown on the left as blue helices. The fragment of the c ring from S. plantesis is shown in green on the right. The negatively charged amino acid, blue, on the c subunit is in the inward locked conformation (1). Once the c subunit enters the more hydrophilic channel it can form a possible salt bridge with the positively charged amino acid on the a subunit, orange, as seen in step 2. Once the proton is transferred to the pr oton channel, the c subunit can again enter the inward, locked conformation as seen in 3. The a subunit was rendered in MacPyMol from PDB number 1C17. The ring of c subunits was rendered in MacPyMol from PDB number 2WIE with the aid of PISA. The composit e figure was made in Adobe Photoshop.

PAGE 78

78 Figure 19. Kane Dickson peripheral stalk structure. Residues b79183, d2123, and F6(5 70) of the B. taurus peripheral stalk are shown in blue, pea green, and orange, respectively. Importantly, as shown, the per ipheral stalk narrows to the width of a single helix as the stalk enters the membrane. Structure rendered in MacPyMol from PDB number 2CLY (Dickson et al. 2006).

PAGE 79

79 Figure 110. Partial B. taurus peripheral stalk structure docked with the B. taurus F1c10 structure inside the S. cerevisiae F1FO ATP synthase cryoelectron microscopy image. The b d, and F6 subunits from the Dickson structure are shown in magenta, orange, and green, respectively. The OSCP from NMR studies is shown in blue. Reprinted wit h permission from Dr. John Rubinstein (Lau et al. 2008)

PAGE 80

80 Figure 111. Structure of the B. taurus peripheral stalk with the F1 sector. In this highresolution structure, b122207, d3040, 65748591, and F6(5 25,3557) (blue, pea green, orange, respec tively) are shown with OSCP1146,169189 (yellow) and the 33 subunits (deep purple, light blue, cyan, pink, and sand, respectively). It is important to note that large portions of the d and b subunits are not resolved in this structure. The structur e was rendered in MacPyMol from PDB number 2WSS.

PAGE 81

81 CHAPTER 2 INTRAGENIC SUPPRESSO RS OF THE ARGININE 3 6 MUTATIONS IN THE BACTERIAL PERIPHERAL STALK Introduction The F1FO ATP synthase is the principle enzyme for ATP generation in most organisms. The overall s tructure of the enzyme is highly conserved in all organisms. The F1 sector houses the catalytic sites for the production of ATP from ADP and Pi (Stock et al. 2000; Boyer, 2002; Senior et al. 2002). Highresolution structural analysis of the B. taurus F1 sector has shown that each of the catalytic sites exists in different conformations based on interactions with the rotor (Leslie et al. 1999). In bacteria, F1 consists of nine subunits present in 33 and subunits are arra stalk that extends up through the middle of the 33 hexamer. Rotational energy derived from proton translocation through the FO sector is translated to chemical energy in t he F1 sector (Boyer, 1993). In E. coli, FO consists of three subunits present in the ab2c10 stoichiometry. The a subunit houses two proton half channels. The protons flow through the periplasmic half channel and bind a single c subunit. Almost simultaneously, a proton is released from a different single c subunit into the cytoplasmic half channel. Protonation and/or deprotonation of individual c subunits drive the rotational movement of the c subunits. The function of the peripheral stalk, b21 against the movement of the rotor. A highresolution structure of the B. taurus peripheral stalk suggests that it bends around F1 (Dickson et al. 2006). The b subunit has been divided into three functional domains. The F1 binding domain, bQ123L156, interacts with an pair and with

PAGE 82

82 b subunit (McLachlin et al. 1998; Revington et al. 1999; McLachlin and Dunn, 2000; McLachlin et al. 2000). One of the two b et al. 1997), providing direct support for the suggestion that these subunits act as a stator to hold the F1 sector in place. The dimerization domain falls in a highly helical polar segment, b53122 (Del Rizzo et al. 2002). The b subunits are thought to interact via hydrophobic residues located in this largely polar region (Motz et al. 2004). The membranespanning d omain consists of the amino terminal 24 amino acids that are largely hydrophobic and responsible for anchoring the b subunit in the membrane. In between the membranespanning domain and the dimerization domain lays the tether domain. The tether domain, bV 25S60, roughly corresponds to the segment of the peripheral stalk from the top of the membrane to the bottom of F1. The membranespanning and tether domains of the b subunits interact with the FO sector a and c subunits. The transmembrane segments of each b subunit, together with subunits a and c, is the minimum subcomplex required for proton translocation (Greie et al. 2004). Additionally, the b subunits copurified with a His tagged a subunit, showing that an ab2 subcomplex could be purified from the intact enzyme (Stalz et al. 2003). Several crosslinking experiments have shown that the a and c subunits interact with the b subunit. The bN2C subunit was crosslinked to cysteine residues introduced at positions c74, c75, and c78 at the carboxyl terminus of the c subunit (Jones et al. 2000). Crosslinking of the two subunits inhibited ATP driven proton pumping, and this was interpreted as the inhibition of the rotation of the c ring (Jones et al. 2000). A site located in the first cytoplasmic loop of the a subunit, aK74C, has been crosslinked to the b

PAGE 83

83 subunit using a photoactive crosslinking agent with little effect on proton translocation (Long et al. 2002). In a similar experiment, the bR36C was found to crosslink to the a subunit (McLachlin et al. 2000). In the absence of a highresolution structure of the entire b subunits, mutational studies have provided the most useful information on the b subunit dimer (Cain, 2000). In the membranespanning domain, replacement of several amino acids as a group resulted in a loss of function of the enzyme (Hardy et al. 2003). Mutagenesis of the tether region suggested that this domain was quite adaptable. Functional F1FO ATP synthase was obtained using b subunits with insertions of up to 14 amino acids or deletions up to 11 amino acids (Sorgen et al. 1998b, 1999). The tether domain also contains the evolutionarily conserved barg36 residue. Several mutants were constructed to replace bR36. The most interesting were the bR36I and bR36E substitutions. B oth resulted in a defective F1FO ATP synthase and the latter was a true uncoupling mutation (Caviston et al. 1998). The phenotypes of both the bR36I and bR36E mutations could be suppressed by coexpression with a b subunit containing bR36, forming heterodimers (Grabar and Cain, 2004). Additionally, a variety of other substitutions were made that allowed limited assembly and F1FO ATP synthase activity. An exception to this was the bR36K mutation that resulted in a phenotype similar to that of wildtype. This suggested that while the bR36 was not essential for function, it did exert an important influence on the enzyme. In the dimerization domain, mutations appear to result in a loss of dimerization and failure to efficiently assemble the F1FO complex (Dunn, 1992; McLachlin and Dunn, 1997; Sorgen et al. 1998a). The F1binding domain was found to be particularly

PAGE 84

84 sensitive to mutation. For example, substitution or deletion of the hydrophobic residues in the socalled VAILAVA sequence, bV124A130, resulted i n major functional defects in the enzyme (Porter et al. 1985; Takeyama et al. 1988; Rodgers et al. 1997; Dunn et al. 2000a; Bhatt et al. 2005). Again, the phenotypes of these mutations could be overcome by coexpression with a wild type b subunit (Grabar and Cain, 2003; Bhatt et al. 2005). The present work further investigates mutations affecting the highly conserved bR36 residue. A series of double mutants were constructed in order to move the arginine residue vertically along the b subunit to consi der positional effects for the basic group. Second site substitutions moved the arginine, replacing either a naturally occurring glutamate or isoleucine. The bR36I mutation could not be suppressed with an arginine located either above or below bR36. However, we present biological and biochemical evidence of a strong intragenic suppressor to the bR36E mutation that restored near normal F1FO ATP synthase activity. Materials and Methods Materials Molecular biology enzymes and oligonucleotides were obtained from New England Biolabs and Invitrogen, respectively. Reagents were acquired from SigmaAldrich and Fisher Scientific. Western blotting reagents and high performance chemiluminescence film were purchased from Amersham Pharmacia Biotech. The monoclonal antibodies against the V5 epitope tag were obtained from Invitrogen. The anti mouse, horseradish peroxidase linked whole antibody (from sheep), was a product of GE Healthcare, UK Limited.

PAGE 85

85 Mutagenesis Strains and Media The wild type b subunit plasmid, pTAM46 ( bV5), and the parental mutant plasmids, pTAM54 ( bR36E, V5) and pTAM53 ( bR36I, V5), that were used to generate the double mutants have been previously described (Grabar and Cain, 2004). Site directed mutagenesis was either performed by a Stratagene QuikCh ange Kit or by ligationmediated mutagenesis between the Mfe I site and PpuM I site (Figure 21). Plasmids were purified using the Qiagen Mini Prep and Maxi Prep kits (Qiagen, Inc.). Restriction endonuclease digestions, ligations, and transformations were conducted according to the recommendations of the manufacturers. Double stranded DNA sequence determinations were performed by either the automated sequencing at the core facility at the University of Florida Interdisciplinary Center for Biotechnology Research or at the Center for Mammalian Genetics. The sequence of the mutagenized region of each plasmid constructed, along with any restriction sites inserted for screening purposes, is shown in Figure 21. All recombinant plasmids and the control plasmids b ), pTAM46 ( bV5), pTAM54 ( bR36E, V5), and pTAM53 ( bR36I, V5) (Grabar and Cain, 2004) were expressed in E. coli b ) so that the only b subunits expressed were that of the plasmid gene (McCormick and Cain, 1991). For membrane preparati ons, strains were grown in Luria broth supplemented with 0.2% (w/v) glucose (LBG). For plate growth, isopropyl 1 thio b D to media as necessary. All cultures were incubated at 37 1FO ATP synthase activity was determined by streaking all strains onto plates containing a Minimal A media supplemented with 0.2% (w/v) succinate and incubated at either 37 or 25

PAGE 86

86 Preparative Procedures Inverted membrane vesicles wer e prepared from 500 mL cultures of LBG (Ap) ( 50mM Tris 4), and then suspended in TM Buffer with 10 ss at 14,000 p.s.i. Unbroken cells were pelleted via two successive centrifugations at 9,000 rpm for 6 minutes each. Then membranes were collected using ultracentrifugation at 48,000 rpm for 1.5 hours. Membranes were washed in TM buffer and pelleted usi ng ultracentrifugation at 48,000 rpm for 1 hour. Membranes were suspended in 1 mL TM buffer, centrifuged for 10 minutes at 13,000 rpm, and stored at 4 using the modified Lowry assay for membrane proteins by Markwell, et al. (Markwell et al. 1978). Growth on a plate with Minim al A media supplemented by 0.2% (w/v) succinate was used as an in vivo measure of enzyme viability. Assay of F1FO ATP Synthase Activity Membrane energization was detected by fluorescence quenching of 9amino6 chloro2 methoxyacridine (ACMA) (Aris et al 1985). This assay was performed using a Photon Technology International Fluorimeter and the associated Felix32 software or a Perkin Elmer PS3 spectrofluorimeter as seen in Fig. 3A. A time based assay was performed for 720 seconds. The excitation and emission wavelengths were set at 410 nm and 490 nm, respectively. In a cuvette, 500 g of protein was diluted in 3 mL MOPS buffer (50mM MOPS, 10mM MgCl of 0.15 M ATP was added. The quenching of fluorescence was observed for the remaining 630 seconds.

PAGE 87

87 Membraneassociated ATP hydrolysis activity was assayed using the acid molybdate method (Cain and Simoni, 1989). A high pH buffer (50 mM Tris HCl, 1mM Mg Cl2, pH 9.1) was used to relieve the effects of FO on F1. The reaction consisted of 4 HCl, pH 7.5) at iced stop buffer (9% HCl (v/v), 0.6% ammonium molybdate (v/v), 1.7% SDS (w/v)). To amino2 naphthol 4 sulfonic acid) was added and incubated at 25 each sample. Immunoblot Analysis glycine SDS BioRad Ready Gel. Following electrophoresis, the proteins were transferred onto a nitrocellulose membrane using and BioRad electroblot apparatus (100 mV, 1 ho ur, 4 membrane was washed three times for with TBS ( 1mM Tris HCl, 15 mM NaCl pH 7.2 ). The proteins were then stained with Fast Green stain to confirm transfer to the membrane. The membrane was rinsed with Fast Green Destain and washed three ti mes with T TBS ( 1mM Tris HCl, 15 mM NaCl pH 7.2 with 0.1 % Tween80 (v/v)). The membranes were incubated overnight at 4 fat milk in TBS. The membrane was washed three times with TTBS. A 1:4000 dilution of the anti V5 ant ibody in 2 % BSA in T TBS was incubated at room temperature for one hour. The membrane was washed three times in TTBS and then incubated at room temperature in a 1:20,000 dilution of the anti mouse, horseradish peroxidase linked whole antibody (from sheep). Following this incubation, the membrane was washed five times with T TBS. The antibody was detected using chemiluminescence (ECL Plus,

PAGE 88

88 Amersham). The signals were visualized using high performance chemiluminescence film using a Kodak X Omat. Signal strength was assessed using UnScan It gel digitizing software (Silk Scientific, Inc.). Results Construction and Growth Characteristics of Mutants Previous studies have shown that amino acid substitutions, bR36I and bR36E, were sufficient to inactivate F1FO ATP synthase (Caviston et al. 1998; Grabar and Cain, 2004). Three double mutants were constructed with an isoleucine or glutamate at bR36 and the arginine residue replacing a naturally occurring glutamate or isoleucine. This segment of the b subunit is thought to be predominantly helical (Del Rizzo et al. 2002), so the positions amino acid replacements were designed to occur at one turn of the helix in either direction of the bR36 amino acid (Figure 21). In addition, a single mutant was made, pAW5 ( bE39R, V5), to consider whether two arginines in close proximity affected enzyme function. Interestingly, two spontaneous mutations arose when generating the bR36I,E39R subunit after selection on rich media supplemented with ampicillin. The spontaneous recombinant subunits each had a mutation at position 32, giving pAW10 ( bA32G, R36I, E39R, V5) or pAW8 ( bA32E, R36I, E39R, V5). All plasmids were transformed into the KM2 ( b ) E. coli cell line, which contains a chromosomal deletion for the uncF(b) gene. This eliminated any endogenous b subunit expression. The V5epitope tag was engineered at the carboxyl terminus of the wild type and all mutant b subunits to facilitate detection of recombinant subunits. Previous reports have demonstrated that the V5 epitope tag has little effect upon enzyme assembly or activity (Grabar and Cain, 2004).

PAGE 89

89 First, the recombinant plasmids were tested for their ability to complement the deletion strain for growth on minimal A media supplemented with succinate. Since, F1FO ATP synthase is required for oxidative phosphorylation, growth on succinate as a carbon source served as an initial indication of the assembly of mutant b subunits into a functional enzyme. As expected, the single isoleucine mutant, KM2/pTAM53 ( bR36I, V5), failed to grow on succinate (Caviston et al. 1998; Grabar and Cain, 2004). Attempts to suppress the phenotype by using plasmids pAW4 ( bR36, I40R, V5) and pAW6 ( bI33R, R36I, V5) were unsuccessful (Table 21). The presence of isoleucine at bR36 was apparently sufficient to prevent biologically significant levels of F1FO ATP synthase activity. However, a more interesting result was obtained for the bR36E mutant (Caviston et al. 1998; Grabar and Cain, 2004). The strain KM2/pTAM54 ( bR36E, V5) was not able to grow on succinate. The first indication of suppression of a bR36 defect was that the mutant strain KM2/pAW1 ( bR36E, E39R, V5) grew slowly on succinate at 37 1). At 25 bR36E, E39R, V5) grew just as well as the wildtype control, KM2/pTAM46 ( bV5) (Table I). The results indicated that the bR36E, E39R, V5 subunit was functionally incorporated into an F1FO ATP synthase complex. Interestingly, addition of an arginine in pAW5 ( bE39R) yielded a plasmid that also supported growth on a succinate minimal media. However, it should be noted that strain KM2/pAW5 ( bE39R) did not grow as well as the wild type control, KM2/pTAM46 ( bV5), at eith er 371). Assembly of Intact F1FO ATP Synthase Western blots were performed on membrane samples from the mutant strains using anti V5 mouse monoclonal antibody to the mutant b subunits (Figure 2 2). When IPTG was included in the growth medium to achieve maximum production, the recombinant b subunits were found at levels comparable to control KM2/pTAM46 ( b )

PAGE 90

90 membranes. The exceptions were membranes from strains KM2/pAW5 ( bE39R) and KM2/pTAM53 ( bR36I, V5) that were expressed at levels of approximately one and half times greater than that of control KM2/pTAM46 ( bV5). The increased levels of these subunits were reproducible, and we have no explanation for the enhanced stability. However, the important issue is that all recombinant b subunits were present, indicating that expression and incorporation into the membrane were not limiting for F1FO ATP synthase assembly. In the absence of FO, F1 has very limited affinity for the membrane. Consequently, the level of F1ATPase activity was used as an indication of the amount of intact F1FO ATP synthase present in the sample. Assaying ATP hydrolysis at high pH relieves influence of FO on F1, allowing for the measurement of maximal F1ATPase activity (Cain and Simoni, 1989). All of the mutant strains yielded reduced levels of intact F1FO complexes (Table I). The range for strains expressing b subunits with the bR36I substitution was between about 3055% of positive control values for KM2/pTAM46 ( bV5). The absolute numbers were somewhat higher than the Cain laboratory has reported in the past, reflecting improvements in procedures for washing and suspension of membrane vesicles. The levels bR36E subunits also yielded reduced levels of ATPase activity, at approximately 4075% of the positive control. H owever, all b subunit mutant membranes had ATP hydrolysis values substantially in excess of the negative control b ) membranes, suggesting assembly of intact F1FO complexes containing the recombinant b subunits. Coupled F1FO Activity F1FO media ted ATP driven proton pumping was determined as an indication of coupled activity in membrane vesicles. Two sets of fluorescent traces are depicted for

PAGE 91

91 each sample. Earlier experiments showed reduced ATP driven membrane acidification for either KM2/pTAM53 ( bR36I, V5) or KM2/pTAM54 ( bR36E, V5) (Caviston et al. 1998; Grabar and Cain, 2004). These data were produced using an old Perkin Elmer PS 3 spectrofluorimeter (Figure 23A). However, the new PTI Q4SE fluorimeter, with extended sensitivity and muchimp roved signal to noise ratio, detected low levels of activity for both membrane preparations (Figure 23B, Figure 25). There was no evidence that either KM2/pAW4 ( bR36I, I40R, V5) or KM2/pAW6 (bI33R, R36I, V5) membranes had higher proton pumping activity than pTAM53 ( bR36I, V5) (Figure 23A). However, very strong proton pumping activity was observed in KM2/pAW1 ( bR36E, E39R, V5) (Figure 2 4). At this point, it became apparent that an arginine at position 39 could suppress the bR36E mutation. A question t hen arose as to whether this bE39R substitution could be sufficient to suppress the bR36I mutation. Therefore, another recombinant b subunit was generated: bR36I, E39R, V5. Although this mutant had adequate assembly, approximately 55% of that of wildtype ( bV5), it did not exhibit any ATP driven proton pumping. Additionally, the two recombinant subunits generated via spontaneous mutation did not show any proton pumping activity despite showing sufficient assembly (Figure 24). The growth study, ATP hydrol ysis assay, and coupled F1FO activity results indicate that the bE39R mutation was an effective intragenic suppressor for the bR36E mutation. Discussion The bR36 residue within the b subunit is a highly conserved amino acid located in the tether domain where the subunit emerges from the membrane bilayer. Amino acid substitutions for the conserved bR36 have been shown to impair F1F0 ATP synthase by differing mechanisms. The bR36I substitution inhibited proton translocation in an intact enzyme complex, and the bR36E uncoupled ion movement from catalytic activity

PAGE 92

92 (Caviston et al. 1998). Although extragenic suppressors have been engineered between the b subunit genes in the F1F0 complex (Grabar and Cain, 2004), the present work focuses on intragenic suppres sion of mutations altering bR36. No highresolution structural information is available for the tether domain of the b subunit, but secondary structure predictions suggested that the area of the subunit containing bR36 was likely to be et al. 2002). Therefore, second site mutations were constructed to move the important arginine up or down one helical turn within the b subunit. The positions of arginine and isoleucine were swapped in the bI33R, R36I, V5 and bR36I, I40R, V5 subunits, and a similar inversion was made for the glutamate and arginine in the bR36E,E39R, V5 protein. Only expression of the latter recombinant subunit yielded an active F1F0 ATP synthase. The bR36E, E39R, V5 subunit was readily incorporated in the F1F0 complex and yielded an essentially wild type phenotype in vivo The bR36E, V5 subunit uncoupled the enzyme (Caviston et al. 1998), and the double substitution was sufficient to effectively restore coupling. ATP driven proton translocation i n bR36E, E39R, V5 was greatly enhanced in comparison to bR36E, V5 substitution (Figure 25). The evidence argues that the presence of the arginine is important for coupling, but it cannot be viewed as essential since several other mutants exhibited at least partial function (Caviston et al. 1998). More specifically, a basic amino acid is needed for efficient coupling since the bR36K was almost fully functional (Caviston et al. 1998). A subunit with two tether domain arginines, bE39R, V5 subunit, was al so incorporated into an active F1F0 ATP synthase (Table 21, Figure 25).

PAGE 93

93 However, it seems very unlikely that bR36 substitutions alter interactions involved in the b b dimer. Dimer interactions are governed by hydrophobic interactions at multiple posit ions along much of the length of the b subunit, but an arginine contributed by each of the two b subunits would be expected to repel each other rather than strengthen the proteinprotein interaction. Although there was a reference to unpublished results p urporting disulfide bridge formation between bQ37C subunits in a review article some years ago (Greie et al. 2000a), no clear evidence has been reported of b b subunit interactions in the tether domain between the sites occupied by bQ10 and bA59 (McLachli n and Dunn, 1997; Dmitriev et al. 1999; Dunn et al. 2000b; Greie et al. 2000a). Involvement of bR36 in an interaction with the a subunit offers a more reasonable interpretation. Chemical crosslinking has been reported between photoactivatable agents bound to bR36C and the a subunit (Long et al. 2002) and between aK74C in the first cytoplasmic loop of the a subunit and the tether domain of subunit b (McLachlin et al. 2000). Although an interaction of the bR36 with the a subunit appears likely, the exa ct role of bR36 cannot be positively ascertained in the absence of more information about subunit interactions within FO. However, the coupling defect observed in the original bR36E substitution suggested either an opening of the a subunit proton channel or a loss of stator integrity. Normal function was recovered by reintroduction of an arginine into the tether domain in the bR36E, E39R, V5 subunit. Therefore, it is the presence of the arginine at a site low in the tether domain that is important. The length of the arginine side chain apparently facilitates positioning the guanidino group to perform its function (Figure 26).

PAGE 94

94 The bulky hydrophobic group in the bR36I subunit apparently resulted in an irreversible defect in the F1FO complex. There was a significant assembly defect, and proton translocation was lost in the fully assembled intact F1FO. Clearly, introduction of an arginine at a nearby secondsite was not sufficient to restore measurable enzyme activity (Table 21, Figure 23). Subunits containing the bR36I substitution do not form a productive interaction within the FO sector regardless of the presence of an additional arginine positioned in close proximity. Presumably, the bR36I subunits result in an inappropriate conformation voiding b subunit function, but the substitution does not further supported by the spontaneous generation of mutants in response to the bR36E,E39R mutation. However, the bR36I mutation was readily suppressed by an extragenic suppressor (Grabar and Cain, 2004), so it again seems likely that it is an important intersubunit interaction between a single b subunit and the a subunit that has been disturbed. The extraordinary degree of conservation of bR36 among bacterial species has been documented (Caviston et al. 1998; Poetsch et al. 2007). However, the importance of a basic amino acid located in the b subunit near the membrane surface may extend to higher organisms, as well. A lysine can be found at comparable position in mitochondrial b subunits from animals (Kim et al. 2004) and subunit IV of plants (Poetsch et al. 2007). This information suggests that a basic amino acid contributes to efficient F1FO ATP synthase coupling, not only in E. coli as evidenced by the bR36E, E39R double mutation, but in higher eukaryotes as well.

PAGE 95

95 Table 21. Aerobic growth properties of mutants with double amino acid substitutions. a E. coli strains were grown aerobically on Minimal A media supplemented with succinate. Colony size was scored after a 72 hour incubation at either 37 degrees as: +++, 0.5 mm; no growth. b ATPase activity was measured as described under Experimental Procedures. Units 4 released per mg of protein per minute standard deviation. Units were calculated from the slope of the line based on six time points with incubations of 12 minutes. All values were performed in triplicate. c Mutant activity less the activity of KM2/pBR322 divided by the activity of KM2/pTAM46 less the activity of KM2/pBR32 2 and converted to a percentage. Strains Description Growth a Specific Activityb % Wild Type Activityc 37 25 KM2/pTAM46 b V5 Ap r +++ +++ 2.442 0.71 100 % KM2/pBR322 b Ap r 0.296 0.21 0 % KM2/pTAM53 b R36I, V5 Ap r 0.918 0.05 29 % KM2/pTAM54 b R36E, V5 Ap r 1.223 0.13 43 % KM2/pAW1 b R36E, E39R, V5 Ap r ++ +++ 1.764 0.12 68 % KM2/pAW4 b R36I, I40R, V5 Ap r 1.182 0.10 41 % KM2/pAW5 b E39R, V5 Ap r + + 1.904 0.48 75 % KM2/pAW6 b I33R, R36I, V5 Ap r 1.467 0.19 55 %

PAGE 96

96 Figure 21. Sense strand sequences of recombinant uncF(b) genes. Top, the wildtype nucleotide sequence and deduced amino acid sequences of pTAM46 ( bV5) corresponding to amino acids 3254 of the b subunit. Subsequent lines describe novel constructs. The plasmid name and mutation are noted on the left, mutated nucleotides resulting in a change in the amino acid sequence are bolded, and restriction sites used for screening or mutagenesis purposes are underlined and labeled.

PAGE 97

97 Figure 22. Immunoblot analysis of uncF(b) gene mutant membranes. Membrane proteins were separated on a 15% Tris HCl BioRad Ready Gel (See Experimental Procedures) and the proteins were transferred to nitrocellulose membranes. The pr esence of b subunit was detected using an antibody against the V5 epitope tag located on the carboxyl terminus of the b subunits. Lanes, membranes prepared from strains: 1) KM2/pBR322, 2) Blank, 3) KM2/pTAM36, 4) KM2/pTAM54, 5) KM2/pAW1, 6) KM2/pAW5, 7) K M2/pTAM53, 8) KM2/pAW4, 9) KM2/pAW6. Amino acid replaces in each subunit are specified above the lanes. Signal strengths on this and other blots were determined using a much shorter exposure and the UnScan It gel digitizing software (Silk Scientific, In c.).

PAGE 98

98 Figure 23. ATP driven energization of membrane vesicles prepared from isoleucine uncF(b) gene mutants. Membranes from IPTG induced strains were prepared by differential centrifugation (see Experimental Procedures). Proton pumping was assayed using ACMA and 500 Panel A, traces obtained using a Perkin Elmer PS3 spectrofluorimeter. Panel B, traces obtained using a PTI Q4SE fluorimeter. Traces: 1) KM2/pBR322 b ), 2) KM2/pAW4 ( bR36I, I40R, V5), 3) KM2/pAW6 (bI33R, R36I, V5), 4) KM2/pTAM53 ( bR36I, V5), 5) KM2/pTAM46 ( bV5).

PAGE 99

99 Figure 24. ATP driven energization of membrane vesicles prepared from isoleucine uncF(b) gene mutants. Membranes from IPTG induced strains were prepared by differential centrifugation (see Experimental Procedures). Pr oton pumping was assayed using ACMA and 500 Traces were obtained using the PTI Q4SE fluorimeter. Traces: 1) b ), 2) KM2/pAW8 ( bA32E, R36I, E39R, V5), 3) KM2/pAW10 ( bA32G,R36I,E39R, V5), 4) KM2/pTAM46 ( bV5).

PAGE 100

100 Figure 25. ATP driven energization of membrane ves icles prepared from glutamic acid uncF(b) gene mutants. Membranes from IPTG induced strains were prepared by differential centrifugation (see Experimental Procedures). Traces were obtained using the PTI Q4SE fluorimeter. Traces: 1) b ), 2) KM2/pTAM54 ( bR36E, V5), 3) KM2/pAW1 ( bR36E, E39R, V5), 4) KM2/pAW5 ( bE39R, V5), 5)KM2/pTAM46 ( bV5).

PAGE 101

101 Figure 26. Model of mutated region in pTAM46 ( bV5) and pAW1 ( bR36E, E39R, V5). An helical conformation was assumed based on secondary structure predictions. The model on the left shows residues 29 to 40 of the wildtype b subunit, with bR36 shown in black and bE39 in gray. The model on the right shows the same region of pAW1 ( bR36E,E39R, V5).

PAGE 102

102 CHAPTER 3 CHEMICAL MODIFICATION OF CYSTEINES IN THE BACTERIAL PERIPHERAL STALK Introduction The primary role of the peripheral stalk in the F1FO ATP synthase is to hold the F1 33 hexamer in place against the rotation of the central st alk during ATP synthesis (Chapter 1). The peripheral stalk has a stoichiometry of b2 E. coli F1FO ATP synthase. The two identical b subunits are thought to be in parallel, extended helix conformations. The subunits extend from the amino terminus on the periplasmic side of the cytoplasmic membrane up along one side of the enzyme to the top of F1, where at least one b et al. 2009). The majority of the proteinprotein contacts between b2 and with other subunits in the F1FO ATP synthase occur at either end of the peripheral stalk in the membrane spanning and F1 binding domains. In the membrane spanning domain the b2 dimer is thought to primarily interact with the a subunit. At the F1 binding domain, the interactions appear to be with a single pair and the et al. 2005). Within the tether domain, the region from the top of the membrane to the bottom of F1, there is very little interaction with other subunits. An important exception to this is the likely interaction at the bottom of the tether domain between bR36 and the large cytoplasmic loop, L12, of the a subunit (see Chapter 2). The peripheral stalk exhibits a high degree of tolerance to single amino acid substitutions altering many of its residues. Previous work has shown that the peripheral stalk can tolerate individual amino acid substitutions at many positions in all four

PAGE 103

103 functional domains (Sorgen et al. 1998b, 1999; Hardy et al. 2003; Bhatt et al. 2005). Claggett et al showed that approximately onethird of the E. coli b subunit encompassing the tether domain and a part of the dimerization domain can be replaced with a similar segment from the b or b subunit of Thermosynechococcus elongatus a thermophillic cyanobacterium (Claggett et al. 2007) These chimeric b subunits as sembled into functional F1FO ATP synthases. Epitope tags engineered at both the amino and carboxyl termini of the b subunit did not significantly affect enzymatic activity (Grabar and Cain, 2003). Importantly, Grabar and Cain devised a twoplasmid expressi on system for incorporating two different b subunits into the same enzyme (Grabar and Cain, 2003). Using the twoplasmid expression system, b subunits of unequal lengths were integrated into the same F1FO ATP synthase. Specifically, these complexes had one b subunit with an insertion of seven amino acids and one with a deletion of seven amino acids. Again, these enzymes proved intact and functional. The collective set of data showed that the E. coli F1FO ATP synthase can function with large changes in content and in length of its peripheral stalk. The evidence was indicative of considerable plasticity in the bacterial stalk structure. The plasticity exhibited by the peripheral stalk makes it a candidate for elastic power transmission between the F1 and FO sectors. The catalytic activity of the F1 sector is coupled with proton translocation through the FO sector by rotation of the rotor stalk ( ) and the ring of c subunits ( c815). In the F1 sector, each 120 step of the rotor stalk ( ) causes sequential conformational changes of the / catalytic sites. The rotation in the FO sector occurs with the protonation and deprotonation of individual c subunits, which causes the c ring to rotate. The step size varies based on the number of c

PAGE 104

104 subunits in the ring and is calculated to be between approximately 24 in a c15 ring to 45 in a c8 ring. The enzyme is thought to compensate for the asymmetry in step sizes between the rotor stalk and c ring by an elastic power transmission between the F1 and FO sectors (Sielaf f et al. 2008). Whether the elastic power transmission occurs through the rotor and/or peripheral stalk is unknown. Experiments have been performed to determine the stiffness of the subunit when associated with the 33 hexamer (Okuno et al. 2010). I n these experiments the hexamer was fixed to a Ni NTA modified coverslip through His tags on the and subunits, and a magnetic bead was attached to the external end of the subunit. The change of the angle of the magnetic bead was visualized while the enzyme underwent ATP hydrolysis. By tracking the change in angle of the bead, the torsional stiffness of the subunit was calculated. The portion of the subunit embedded in the 33 hexamer was calculated to have a torsional stiffness, which is a meas ure of resistance torsional force, of 223 pNnm/rad. This is in comparison the torsional stiffness of the subunit external to the hexamer measured at 73 pNnm/rad. It should be noted that these torsional stiffness measurements were done with only the F1 se ctor and do not account for any possible constraints on the subunit by the FO sector. There has been some suggestion that the peripheral stalk is relatively stiff within the entire enzyme (Waechter et al. submitted for publication). In these studies the entire F1FO complex was fixed to a Ni NTA modified coverslip by His tags on the and subunits and a magnetic bead was attached to a bN2C subunit. The amino terminus of the b subunit was readily accessible because the enzyme was not associated with the membrane. The stiffness of the b subunits was calculated by applying a magnetic force and forcing

PAGE 105

105 rotation of the enzyme. The results indicated that the b subunits were relatively stiff. However, it is unclear whether the experiment considers the constraints on the b subunits due to its interaction with the F1 and FO sectors. Additionally, the experiment does not exclude the possibility of local deformities in the b subunits. It should also be noted that the measurements of the stiffness of the and b subunits are not directly comparable because the subunit studies are done in a portion of the enzyme, while the b subunit studies are done with the entire enzyme. A direct measurement of plasticity of the peripheral stalk within a F1FO complex situated in a membrane environment is necessary to determine if the peripheral stalk can participate in a transmission of elastic power. The intent of the present work was to develop tools to allow biophysical analysis of the peripheral stalk in the E. coli F1FO ATP syn thase. In this chapter, I show the design and construction of a variety of mutant enzymes that contained only one cysteine in the F1 and FO sectors. The functionality and ability to chemically modify some of these mutants are demonstrated. Materials and M ethods Materials and General Molecular Methods All chemicals were obtained from Sigma Aldrich Co. or Fisher Scientific, Inc. unless otherwise noted. Restriction enzymes and T4 polynucleotide kinase were obtained from New England Biolabs, Inc. T4 DNA ligas e was bought from Invitrogen. QuickChange mutagenesis kit was obtained from Agilent Technologies. Restriction enzyme digests, T4 polynucleotide kinase reactions, ligations, and QuickChange mutagenesis protocols were performed according to manufacturers instructions.

PAGE 106

106 Isolation of the F1 Sector Wild type and recombinant F1FO ATP synthase were expressed from plasmids in E. coli unc operon. For isolation of the F1 sector, strains were grown in Luria broth supplemented with 0.2 % (w/v) glucose (LBG) and grown at 37 C. Ampicillin (10 g/mL) and chloramphenicol (25 ng/mL) were added to the media as needed. A total of 16 L of cells were grown for each enzyme purification. These cells were grown in 800 mL aliquots, inoculated with 10 mL of an overnight E. coli culture, without antibiotic for 4 hours at 37 C. Cells were harvested by centrifugation at 10,000g for 6 minutes at 4 C. From this point on cells and buffers were kept on ice or at 4 C. Harvested cells were suspended in STEM buffer (250 mM sucrose, 100 mM TES, 20 mM Mg(CH3COO)24H2O, 0.25 mM EGTA, 40 mM EACA, pH 7.0) and transferred to a preweighed, prechilled 250 mL centrifuge tube. Centrifuging at 10,000g for 6 minutes was sufficient to pellet the cells. The pellet was weighed after discarding the supernatant to determine cell mass. Using STEM buffer again, the cells were suspended at 2 mL/g wet cells for overnight storage. Cells were never stored longer than 12 hours. The cells were lysed in the presence of 1.25 mL 1 mg/mL DNase I. At this point, the suspension was divided into 4 5 fractions, in order to accommodate the capacity of the French Press cell. Each fraction was passed through the French Press twice to ensure maximum yield of lysed cells. Next, the lysates were centrifuged at 25,000g for 20 minutes. The pellet was discarded and the supernatant was transferred to a clean 15 mL ultracentrifuge tube with a screw cap. This provided a safeguard against spilling the sample. The supernatant was centrifuged at 144,000g for 2.5 hours. Following centrifugation, the supernatant was discarded and the pellets suspended in TES50

PAGE 107

107 buffer (50 mM TES, 15 % glycerol, 5 g 6aminohexanoic acid, 1.0 g paminobenzamidine (PAB), pH to 7.0 with NaOH). It should be noted that to avoid damaging the enzyme, suspensions were performed as gently as possible with the aid of a paintbrush. The suspension was returned to the same ultracentrifuge tube and centrifuged, again, at 144,000g for 2.5 hours. To remove the peripheral membrane proteins, the pellet was suspended in TES5 plus PAB (5 mM TES, 15% glycerol, 0.5 mM DTT, 0.5 mM EDTA, 5 g 6aminohexanoic acid, 1 g PAB, pH to 7.0 with NaOH). This low ionic strength buffer was used to strip off the peripheral membrane proteins, but F1 remained associated with the membrane because the PAB stabilizes its interaction with FO. The cell fre e suspension was centrifuged at 144,000g for 2.5 hours. The supernatant was discarded, the pellet was suspended again in TES5 plus PAB, and the centrifugation step was repeated. F1 was removed from the membrane by suspending in TES5 without PAB (5 mM TES, 15% glycerol, 0.5 mM DTT, 0.5 mM EDTA, 5 g 6aminohexanoic acid, pH to 7.0 with NaOH) and centrifuging at 144,000g for 2.5 hours. The supernatant was saved. This suspension and centrifugation step was repeated three times. At the end, the supernatant fr om each centrifugation step was pooled into one fraction and the total volume was determined. The supernatant could be saved at 0 C for 12 days if necessary, but longer storage was avoided. The protein was precipitated using PEG 6000. The following solutions were added ice cold on a per 100 mL basis supernatant and in this order: 5 mL 1 M TES pH 7.0, 0.2 mL 0.5 M DTT, 5 mL 1 M MgCl2. The mixture was stirred for 1 minute. If it was cloudy, the mixture was centrifuged at 8,000g for 60 minutes and the volume was

PAGE 108

108 determined. Next, 25 mL 50% PEG 6000 was added per 100 mL supernatant and stirred for 10 minutes. The mixture was centrifuged at 8,000g for 60 minutes. The supernatant was discarded and the inside of the tube was dried using a kimwipe. The pellet was suspended in 80 mL Futai buffer (50 mM Tris, 10% glycerol, 2 mM MgCl2, 1 mM DTT, 5 g 6aminohexanoic acid, pH to 7.4 with HCl, and 520 mg Na2ATP added after the pH was altered) with freshly prepared 1 mM PMSF. This volume of Futai buffer and PMSF was not dependent upon the volume of the supernatant so it was not scaled down. The suspension was centrifuged at 31000g for 15 minutes. The supernatant was transferred to a clean tube and the pH was adjusted to 7.4 using 0.1 M NaOH. Finally, the mixture was stored overnight at 20 C. All chromatography steps were carried out in the cold room. The Whatman DE52 resin was prepared according to the manufacturer's instruction for a total column volume of 45 mL and the pH was adjusted to 7.4. The column was equilibr ated by running 250 mL of Futai buffer (pH 7.4) through the column until the pH of the effluent was 7.4. The supernatant was applied to the column and the flow was adjusted to 1 mL/min. First, the column was washed with 50 mL Futai buffer (pH 7.4). Next, i t was washed with 100 mL Futai buffer (pH 7.4) containing 40 mM Na2SO4. Finally, the F1 was eluted off the column with 200 mL Futai Buffer (pH 7.4) containing 80 mM Na2SO4. The eluent was collected until the column had run dry. A 1 mL HighTrapQ column (Am ersham) was used to concentrate the F1 from the eluent without further damaging the protein. The column was set up using two pumps and run according to the manufacturer's instructions. The starting buffer was Futai

PAGE 109

109 buffer (pH 7.4) containing 80 mM Na2SO4 a nd the elution buffer was 1 M NaCl in the Futai buffer with Na2SO4. To further purify F1, a size exclusion column (Sephacryl S 300) was used. The Sephacryl S 300 was poured into the 1.5 cm x 45 cm column using a funnel. The column was packed at 1 mL/min using the running buffer, Futai Buffer (pH 7.0) containing 40 mM Na2SO4. Following sample loading, the column was run at 0.5 mL/min and 5 mL fractions were collected overnight. In the morning, fractions were analyzed for ATP hydrolysis activity as described under Materials and Methods in Chapter 2. Fractions containing ATP hydrolysis activity were pooled and concentrated using the HiTrapQ column. The concentration was performed according to the method described above, except that the starting buffer was Fu tai Buffer (pH 7.0) containing 40 mM Na2SO4. Labeling of the F1 Sector The concentrated protein was dialyzed 1:1000 overnight against Futai buffer (pH 7.0) containing 40 mM Na2SO4 in the absence of DTT. Following dialysis, the protein concentration was determined using a modified Lowry method (Markwell et al. 1978). The protein was either diluted or concentrated to 50 M. A 10 mM stock of the fluorescent dye Alexa Fluor 488 maleimide was prepared in DMSO. It was important to prepare this solution in a amber microcentrifuge tube and cover the tube in aluminum foil to prevent photobleaching. 100 L of the 10 mM fluorescent dye was added to 1 mL 50 M F1. The reaction was covered in foil and placed on the Nutator in the cold room to gently mix the solution ov ernight. The next day, the reaction was stopped by adding approximately 50 L of betamercaptoethanol. To separate the labeled protein from excess label, the reaction was applied to a Sephadex G 25 column in the cold room.

PAGE 110

110 The column size was 1.5 cm x 20 c m and the running buffer was Futai buffer (pH 7.0) containing 40 mM Na2SO3. Reconstitution of F1FO ATP Synthase Reconstitution was attempted with labeled F1 and unlabelled FO. The unlabeled FO was prepared following the procedure described previously in Isolation of the F1 sector from the beginning to the end of the TES5 without PAB steps, but with a few differences. Only a 500 mL culture was used and the pellet, not the supernatant, was saved after the TES5 without PAB steps. After the TES5 without PAB steps were complete, the pellet was suspended in reconstitution buffer (20 mM Tricine/NaOH, pH8.5, 20 mM succinate, 2.5 mM MgCl2, 0.6 mM KOH) to a final protein concentration of 40 nM. Labeled F1 was added in a 1:1 molar ratio to the FO. At that time the NaCl concentration was adjusted to 50 mM and the mixture was heated in a water bath to 37 C. Next, the mixture was placed on ice for 90 minutes and then centrifuged in an ultracentrifuge for 2.5 hours at 158,000g. The supernatant was placed in a clean conical tube and placed at 4 C. The pellet was suspended in modified reconstitution buffer (20 mM Tricine/NaOH, pH8.5, 20 mM succinate, 2.5 mM MgCl2, 0.6 mM KCl, 4% Glycerol) and stored at 4 C. Results Generation of F1FO ATP Synthases with Mutant b subunits Ongoing biophysical studies of the F1FO ATP synthase peripheral stalk require the ability to chemically modify the b is to position chemically reactive cysteines within the subunits by sitedirected

PAGE 111

111 mut agenesis. It was important that the cysteine mutation used in biophysical experiments not alter F1FO ATP synthase activity. Construction of mutant b and bN2C,C21S,+11 genes enerous gift from Dr. Nakamoto). This plasmid contained the entire unc operon with all of the cysteine codons replaced with alanine codons. Enzymes expressed from this plasmid showed no apparent difference in activity when compared to the wildtype enzyme (Kuo et al. pBR322. The cysteine less (cysless) unc operon was inserted between the Hin DIII site and the Nde I site, eliminating the tetracycline resistance gene. In addition to aid in purif ication, a FLAG epitope tag was engineered at the amino terminus of the subunit. The insertion of eleven codons within the b subunit gene of the cys less unc operon was generated by digesting pTAM3 ( bC20S,+11) (Table 32)(Grabar, 2004) and pSD306 ( bN2C) (Gift from Dr. Stanley Dunn) with Ppu MI and Bss HII. Ligating the two together yielded pBDC307 (cys less unc operon, bN2C,C20S,+11). However, once this E. coli unc operon), a very low amount of recombinant F1FO was obtained. Plasmids pSD306 and pBDC307 were used as a source of F1FO ATP synthase for a biophysical study (Waechter et al submitted), but the low yield of enzyme was problematic. Upon further inspection, it was found that the plasmid lacked a proper promoter for the unc operon. Therefore, a decision was made to move the bN2C,C20S,+11 less unc operon) plasmid. This was accomplished by digesting pBDC307 (promoter less, cys less unc operon, bN2C,C20S,+11 less unc operon) with Bss HII and Bsr GI and ligating

PAGE 112

112 the two together. This yielded plasmid, pAW18 (cys less unc operon, bN2C,C20S,+11) (Figure 31) (Table 32). To generate the unc operon with the bN2C,C20S, subunit, pTAM4 ( b) (Grabar, 2004) (Table 32) was mutated via QuickChange Mutagenesis (Agilent) using primers AW562 and AW572 (Table 31). This PCR reaction mutated an asparagine codon to a cysteine codon at position two, yielding plasmid pAW13 ( bN2C,C20S, ) (Table 3 2). This plasmid was unsuitable for our purposes because it did not contain the entire cysless unc operon. To place the b subunit gene into the plas containing the cys less unc bN2C,C20S, ) were digested with Bss HII and Bsr GI. The two resulting fragments were ligated together to yield pAW19 (cys less unc operon, bN2C,C20S, ) (Figure 32) (Table 32). Coupled activity of enzymes containing b and bN2C,C21S,+11 subunits ATP driven proton pumping mediated by F1FO is considered an indication of coupled enzyme activity. The approach was used to determine if the cysteine substitutions compromis ed coupled F1FO ATP synthase activity. In order to determine coupled activity of F1FO ATP synthases expressed from plasmids pAW18 ( bN2C, C20S, +11) and pAW19 ( b unc genes so the only enzyme present was expressed from the plasmids. It appeared that both pAW18 ( bN2C, C20S, +11) and pAW19 ( b) provided abundant levels of coupled F1FO ATP synthase to drive proton pumping assays (Figure 33). However, the two recombinant enzymes with alterations in the b subunit exhibited less activity than compared to the wild type enzyme. Sorgen et al showed that enzymes containing b+11 and b subunits had reduced amounts of coupled activity in the membrane fractions, b ecause there were fewer intact F1FO ATP

PAGE 113

113 synthases (Sorgen et al. 1998b, 1999). It is likely that this is true of enzymes containing these recombinant b subunits, as well. However, these vectors were significantly better than earlier constructs. Both plasm ids pAW18 ( bN2C, C20S, +11) and pAW19 ( bN2C,C20S, ) have been requested by labs currently preparing further biophysical studies. TwoPlasmid Expression System for Introduction of a Cysteine into One b Subunit A central problem with studies based on fluorescent labeling of the b subunit was that both b subunits would have the cysteine. Therefore, it was necessary to construct mutants with a single cysteine amino acid in one of the b The key part of these experiments was that the cysteine residue to be modified must be loc ated in only one of the b subunits. This was challenging because the two identical b subunits are expressed from a single gene within the unc operon. To overcome this obstacle, a twoplasmid expression system was devised in the Cain laboratory (Grabar and Cain, 2003). The two plasmids had different origins of replication to allow both plasmids to be replicated in the same cell. For screening purposes, each plasmid had different antibiotic resistance genes. One plasmid contained the cys less unc operon and t he ampicillin resistance gene. The other plasmid, which had the chloramphenicol resistance gene, contained a b subunit gene with a number of modifications. Modifications in this b subunit gene included a histidine tag at its carboxyl terminus (denoted as + His), a truncation of four amino acids from the carboxyl terminus (denoted expression of these two plasmids resulted in two populations of F1FO ATP synthases: (i) an enzyme with two full length b subunits and (ii) an enzyme with one full length b subunit and one truncated b subunit with a histidine tag and a cysteine residue. No F1FO complexes were present with two truncated b

PAGE 114

114 subunits because those enzymes will not assemble properly (Takeyama et al. 1988; Grabar and Cain, 2004; Bhatt et al. 2005). The existence of a cysteine in only one b subunit allowed for selective labeling with a fluorescent maleimide. Construction of plasmids for the twoplasmid expression system. To build the plasmid that would be used for the truncated b subunit, pSBC109 ( b+His) (Claggett, 2008) was selected as the starting plasmid. This plasmid had a histidine tag (+His) on the amino terminus of the b subunit. The cysteine codon at position 20 was mutated to a serine codon (Table 31). This yi elded plasmid pAW11 ( bC20S, +His, Cmr). The last four codons from the b subunit of this plasmid were used in a QuickChange protocol (Agilent) (Table 31). This yielded plasmid pAW12 ( b +His, Cmr) (Figure 34). This plasmid was used to generate the bS84C mutation in plasmid pAW32 ( b, Cmr) (Figure 35). An additional b subunit mutant was necessary. This was a nontruncated b subunit with a deletion of eleven amino acids in the tether domain. This mutant b less unc operon) so that it could be coexpressed with plasmid pAW32 ( b, Cmr (cysless unc operon) was digested with Ppu MI, the sticky ends were filled in using the large Klenow fragment from DNA polymerase I, and then blunt end ligated. This yielded plasmid pAW38 (cys less unc operon, b) (Figure 36). e are two naturally plasmid (cys less unc operon). The cysteine codon at position 64 was restored by less unc C140S) with Bss HII and Sph I

PAGE 115

115 and ligating the two fragments together to yield plasmid pAW15 (cys less unc operon, C64, C140S) (Figure 37). Coupled activity of enzymes from the twoplasmid expression system. To determine if the cysteine substitutions or the double expression system resulted in a loss of coupled F1FO ATP synthase activity combinations of plasmids were activity, then the substitution would not be useful for chemical modification with a fluorescent probe. Earlier work had already shown that the bC20S, +His subunit did not interfere with coupled F1FO activity (Claggett et al. 2009). The F1FO ATP synthas e from the strain expressing pAW12 ( b) andpAW15 ( b C64) (Trace 1, Figure 38) has comparable F1FO mediated proton pumping activity to the wildless unc operon) (Trace 2, Figure 38). This indicated that the enzyme from two plasmids had the same activity as the cysteineless enzyme. Fluorescent C64 in the F1 sector of the enzyme was not trivial. There was a variety of challenges for this experi ment, such as separating the F1 sector from the FO sector. However, the largest problem was that F1 is a large complex and the labeling reaction required a relatively high concentration of protein. Generating enough protein for the labeling reaction was labor intensive. It required large quantities of bacteria to be grown and large quantities of the F1 sector had to be purified without degradation. Nevertheless, the initial attempt at chemical modification of the C64 using the Alexa Fluor 488 maleimide (AF488) was successful. The AF488 was mixed with F1 and

PAGE 116

116 the reaction was stopped with 100 L betamercaptoethanol. To remove excess AF488, the labeled F1 fraction was dialyzed overnight at 1:1000 against 40 mM Na2S O4 Futai buffer (pH 7.0) in the absence of DTT. The fluorescent spectra of unlabeled and AF488modified F1, both preand post dialysis were recorded. The fluorescent signals from both the preand post dialysis samples (green and blue traces, respectiv ely, Figure 39) were similar and suggested that the majority of the AF488 in the reaction had labeled the F1 sector. The fluorescent signal of the post dialysis AF488modified F1 sector (blue trace, Figure 39) was above the background, unlabeled F1 (red trace, Figure 39) indicating that AF488 had successfully modified the F1 sector. However, the volume of the sample during dialysis increased resulting in a decreased concentration of the AF488modified F1 sector. The decrease in concentration of the modi fied F1 sector was unsuitable for use in future experiments. Therefore, in future labeling experiments a Sephadex G 25 column was used to remove any excess AF488 present. After the first proof of concept purification and modification experiment, the amount of F1 purified was increased. A larger amount of F1 was desired for use in reconstitution experiments. The newly purified F1 sector was labeled as in the proof of concept experiment. Once the labeling reaction was stopped with betamercaptoethanol, the entire labeling reaction was applied to a Sephadex G 25 column to separate any excess label. The eluent from the column was collected in 1 mL fractions. Each fraction was analyzed for fluorescent signal and ATP hydrolysis activity. ATP hydrolysis activity and fluorescent signal was confined to two consecutive fractions (blue and green traces, Figure 310). Additionally, ATP hydrolysis rates correlated with

PAGE 117

117 the amount of fluorescence in the sample, indicating that the AF488 label did not interfere with F1 catalytic activity. Attempted reconstitution of the chemically modified F1FO Reconstitution of the F1FO complex was attempted for the AF488modified F1ATPase from strain 1100 pAW12 ( b) and pAW15 ( b C64). After the reconstitution attempt, no F1FO ATP driven proton pumping observed in the sample (Figure 311C). The lack of F1FO ATP driven proton pumping indicated that the reconstitution was unsuccessful. However, it was still possible that reconstitution did occur, but the membrane vesicle was leaky and prevented energization of the membrane. If the membrane vesicles were intact, addition of NADH to the sample would acidify the membrane. This would appear as an approximately 85% decrease in fluorescence within 30 seconds of addition of NADH. Only a slight decrease in fluorescence was observed upon addition of NADH, suggesting that the membrane vesicles were not intact (Figure 311D). This was probably due to F1less FO in the membrane. Another possible explanation was that the enzyme did reconstitute properly, but the AF488 interfered with F1FO ATP driven proton pumping activity If the AF488 did interfere with coupled activity, but not rec onstitution, then all AF488modified F1 should be associated with the membrane vesicles. The membrane portion of the sample was separated by centrifuging the mixture for 1.5 hours at 158,000g in an ultracentrifuge. The supernatant was saved in a clean micr ocentrifuge tube and the pellet containing the membrane portion was suspended in modified reconstitution buffer. The supernatant and the pellet were each analyzed for the presence of a fluorescent signal (Figure 311A and 311B). Only the

PAGE 118

118 supernatant showed any fluorescence, indicating that the modified F1 failed to bind to the membranebound FO sector. Chemical modification of intact F1FO ATP synthase (cysC64) To determine if the fluorescent label interfered with F1FO ATP synthase activity, the intact enzyme was labeled. The plasmids pAW12 ( b) and pAW15 ( b C64 E. col i cells. The protocol for preparation of F1FO ATP synthases in inverted membrane vesicles was similar to that described in the Isolation of the F1 sector except that the protocol was terminated following the TES5 plus PAB steps. Chemical modification of the F1FO complex with AF488 was conducted as described previously. Following the modification of the complex, fractions were analyzed for the presence of the AF488modified F1 and F1FO ATP driven proton pumping activity (Figure 312). Interestingly, the label did not affect coupled enzyme function as F1FO ATP driven proton pumping was observed in abundance (Figure 312B). Although reconstitution failed, the intact and coupled chemically modified F1FO complexes may yet prove useful for future experimentation. Discussion Work in this chapter showed the generation of a variety of b mutants, the functionality of the recombinant proteins in F1FO ATP synthase, and 1 and intact F1FO ATP synthase. The F1FO ATP synthase assembled properly including subunit s with cysteines introduced into cysteineless enzyme. Additionally, this chapter showed that C64 did not affect catalytic activity of both F1ATPase and F1FO ATP synthase.

PAGE 119

119 This work showed that the F1FO ATP s ynthase accepted the introduction of a reactive thiol could be chemically modified was answered. This was shown when the AlexaFluor 488 maleimide readily labeled the cy C64 was the only cysteine present, it was known with certainty that the only the twoplasmid system provided a means for future labeling of one b subunit. Modification of one b subunit could be used in anisotropy or, in conjunction with a experiments. The line of research was terminated for sever al reasons. First, the reconstitution experiments did not work. The possibility exists that the fluorescent label interfered with reconstitution by altering the bb is absolutely essential for this interac tion. The fluorescent label may have prevented this critical protein protein contact. The argument against this is that it appears that labeling of the intact F1FO complex did not interfere with coupled enzymatic activity. Another possibility was that the F1 subunit is known to dissociate from F1 1 under reconstitution condit ions. However, recovery of fluorescent label in the same fraction with the ATP hydrolysis activity following column chromatography suggests that stability of intact F1 was maintained through the chemical modification step.

PAGE 120

120 More importantly, work ended her e because a crystal structure from B. taurus of part of the peripheral stalk by Kane Dickson et al was published (Dickson et al. 2006). This highresolution structure highlighted the major differences in structure between the prokaryotic and eukaryotic enzymes. Not only was the subunit composition different in the eukaryotic peripheral stalk, but it also appeared that interactions between those subunits were different than in the prokaryotic stalk. For example, the b subunits of the prokaryotic stalk were thought to be in a parallel coiledcoil conformation (Bi et al. 2008; Wise and Vogel, 2008). However, the b and d subunits in the eukaryotic stalk were in one parallel and two anti parallel coiled coils in the B. taurus structure (Dickson et al. 2006). The major differences also called into question the validity of using the prokaryotic enzyme as a model for the eukaryotic enzyme for studies of the peripheral stalk. Therefore, work stopped here on the E. coli enzyme and it was decided that the eukaryoti c F1FO ATP synthase was the more appropriate model for any future experiments. Nevertheless, the technology developed in the bacterial experiments may be useful for designing protocols for study of the eukaryotic stalk. However, it became essential to det ermine if the two peripheral stalks have common physical properties. For example, is the plasticity seen in the prokaryotic stalk applicable to the eukaryotic peripheral stalk? There were indications that this might not be the case. The Kane Dickson struct ure was initially interpreted as indicative of a rigid structure rising perpendicular to the plane of the membrane. This question is further explored in the work in Chapter 4.

PAGE 121

121 Table 31. Primers to generate b subunit genes for cysteine introduction Prime r Name Primer Sequence TB12 5 ctgttcgttctgttctccatgaagtacgtttggccgcc 3 TB13 5 ggcggccaaacgtacttcatggagaacagaacgaacag 3 TB16 5 gcgaacaaacgccgctgccagattctcgacgaagc 3 TB17 5 gcttcgtcgagaatctggcagcggcgtttgttcg 3 TB40 5 catcgtggataagctttaaggaggga ggggctg 3 TB41 5 cagcccctccctccttaaagcttatccacgatg 3 AW56 2 5 ctaaatagaggcattgtgacatgtgtcttaacgcaac 3 AW57 2 5 gttgcgttaagacacatgtcacaatgcctctatttag 3 Table 32. Plasmids generated for cysteine introduction Plasmid Name Description Notes pAW 11 b C20S +His Cm r pJLG1 based pAW12 b Cm r pJLG1 based pAW13 b Cm r pTAM3 based pAW15 C64S pAW18 b N2C, C20S, +11 Ap r pAW19 b Ap r pAW32 b Cm r pJLG1 based pAW38 b Ap r

PAGE 122

122 Figure 31. Construction of pAW18 ( b). Panel A, Complete plasmid map that shows the entire unc is colored in green. Panel B, Stu I restri ction site was inserted for screening in generation of pAW18 ( b). Lanes: 1 1 kb ladder, 2 pAW18 Stu I, 3 pAW18 Stu I, 4 pAW18 Stu I.

PAGE 123

123 Figure 32. Construction of pAW19 ( bN2C, C20S, +11). Panel A, Complete plasmid map that shows the entire u nc is colored in green. Panel B, AflIII Restriction site inserted for screening in generation of pAW18 ( b). Lanes: 1, 1 kb ladder, 2 uncut, 3 AflIII, 4 pAW19 AflIII.

PAGE 124

124 Figur e 33. Coupled F1FO ( b bN2C, C20S, +11). Membranes were prepared as described in Chapter 2 under Materials and Methods. Traces: 1 b ), 2 re bN2C,C20S, +11), 3 b), 4

PAGE 125

1 25 Figure 34. Construction of pAW12 ( b). Panel A, Complete plasmid map that shows the b subunit with the Hin DIII site from the primers TB40and TB40. Panel B, Lanes: 1 1 kb ladder, 2 pAW12 uncut, 3 pAW12 Hin DIII, 4 pAW122 uncut, 5 pAW122 Hin DIII.

PAGE 126

126 Figure 35. Construction of pAW32 ( b). Panel A, Complete plasmid map that shows the b subunit with the Eco RI rest riction site, inserted for screening purposes. Panel B, Lanes: 1 1 kb ladder, 2 pAW32 uncut, 3 pAW32 Eco RI.

PAGE 127

127 Figure 3b). Panel A, Complete plasmid map that shows the entire unc operon. The section replaced in the pB plasmid is colored in green. Panel B, Agarose gel electrophoresis showing the restriction site screen for the loss of a Pvu II site. Lanes : 1 1 kb ladder, 2 Pvu II, 3 A Pvu II, 4 B Pvu II, 5 C Pvu II, 6 D Pvu II.

PAGE 128

128 Figure 37. Construction of pAW15 (Cysteineless unc C64). Panel A, Complete plasmid map with restriction sites used for construction shown. The DNA fragment inserted to form the plasmid is shown in green. Panel B, Agarose gel electrophoresis showing the restriction s ite screen for the insertion of the Eco RI and abolishment of the Cla I sites. Top. Lanes : 1 1 kb ladder, 2 3 Cla I, 4 Eco RI, 5 pAW15 Cla I, 6 pAW15 Eco RI.

PAGE 129

129 Figure 38. Coupled F1FO Activity in membrane vesicles from strains 1 Membranes were prepared as described in Chapter 2 under Materials and Methods. Traces: 1 bC20S, C64), 2 blue, less unc operon).

PAGE 130

130 Figure 39. Init ial labeling reaction of F1 with AlexaFluor 488. The F1 sector was purified and labeled as under Materials and Methods. Traces: Red background fluorescence of buffer, Blue predialysis labeled F1, Green post dialysis labeled F1.

PAGE 131

131 Figure 310. Labeled fractions of F1 AlexaFluor 488. Traces: Red background fluorescence of buffer, Blue peak labeled F1, Green fraction after peak.

PAGE 132

132 Figure 311. Attempted reconstitutio n of Alexa Fluor 488 labeled with FO. Reconstitution was attempted as described previously under Materials and Methods. Panel A, Fluorescent scan of supernatant from reconstitution attempt; Panel B, Fluorescent scan of membranes from reconstitution attempt; Panel C, ATP driven F1FO proton pumping.; Panel D, NADH acidification of membrane vesicles.

PAGE 133

133 Figure 312. Labeled intact F1FO ATP synthase. Membranes were prepared in a manner similar to that described under Materials and Methods. Panel A, Fluor escent scan of the intact enzyme labeled with AlexaFluor 488. Panel B, ATP driven F1FO proton pumping of intact enzyme labeled with AlexaFluor 488.

PAGE 134

134 CHAPTER 4 MANIPULATIONS OF THE PERIPHERAL STALK OF THE EUKARYOTIC F1FO ATP SYNTHASE Introduction Although the prokaryotic and eukaryotic peripheral stalks apparently serve the same core function, the subunit structures are very different. In bacteria, the peripheral stalk of the F1FO ATP synthase is composed of two identical b subunits that are thought to be in an extended helix. The b subunits extend from the bottom of FO, along one noncatalytic / interface to the top of F1, where at least one subunit interacts with the coil conformation from the surface of the membrane to the top of F1. It remains a matter of controversy whether or not the two b subunits are in a left or right handed coiled coil (Bi et al. 2008; Volkov et al. 2008). Hornung et al performed EPR spin experiments on t he soluble portion of the b subunit ( bsol), and subsequently carried out sophisticated molecular modeling based on those experiments to investigate this question (Hornung et al. 2008). According to the molecular modeling analysis, the two b subunits appeared likely to adopt a left handed coiled coil. However, mutagenesis and circular dichroism experiments performed on the b subunits by Bi et al suggested the opposite interpretation of a right handed coiledcoil (Bi et al. 2008). A very recent developm ent lends support the right handed coiledcoil argument. Lee et al have crystallized and solved the structure of the peripheral stalk of a related enzyme, the A1AO ATP synthase (Lee et al. 2010). A1AO ATPases have two peripheral stalks per enzyme comple x, so the structure may not be directly applicable. Nevertheless, the A1AO peripheral stalk in the highresolution structure was in a right handed coiled coil.

PAGE 135

135 In contrast, the eukaryotic peripheral stalk is composed of four different subunits: OSCP, h (or F6 in Bos taurus ), d, and b ( Figure 14 ) in Saccharomyces cerevisiae. The eukaryotic b subunit differs structurally from the bacterial b subunit in that it has two transmembrane domains at the amino terminus rather than one. After traversing the memb rane twice, the subunit extends in a extended helix along one side of F1. The b subunit participates in proteinprotein interactions with subunit d from approximately lysine 98 to glutamic acid 161 (all amino acid numbers are from the S. cerevisiae enzym e unless otherwise noted). Continuing on, the b subunit interacts with the h subunit from approximately threonine 130serine 182. Finally, it reaches the top of F1 interacts with the h subunit as well (Rees et al. 2009). The major differences in subunit composition of the eukaryotic and prokaryotic peripheral stalks raised a questi on about whether the two stalks have similar physical and functional characteristics. Previous work done by Sorgen et al and Grabar and Cain showed that the peripheral stalk of the E. coli enzyme displayed a high degree of plasticity(Sorgen et al. 1998b, 1999; Grabar and Cain, 2003). Removal of 11 amino acids and insertion of 14 amino acids in the b subunits resulted in assembled, fully functional F1FO ATP synthases. However, the number of intact enzymes was reduced. An active enzyme could even contain b subunits of two different lengths with a size difference of up to 14 amino acids. The E. coli F1FO ATP synthase was able to compensate for size differences in the b subunits and form functional, intact F1FO ATP synthases. These

PAGE 136

136 data seemed incompatible wi th previous concepts that the peripheral stalk of the E. coli enzyme was a rigid rod with fixed points of contact in F1 and FO. In the present work, the length of the eukaryotic peripheral stalk was manipulated as a test of its plasticity. A synthetic AT P4 gene was designed to be suitable for sitedirected mutagenesis and a novel expression system was developed within S. cerevisiae deficient for the ATP4 gene. Although only relatively small deletions were compatible with assembly of a functional S. cerev isiae F1FO complex, long insertions comparable to the bacterial insertions resulted in assembly of an active F1FO ATP synthase. The implications of these observations for plasticity of the eukaryotic stalk are considered. Materials and Methods Materials and General Molecular Biology Methods Molecular biology enzymes and oligonucleotides were obtained from New England Biolabs and Invitrogen, respectively. Reagents, chemicals, and media were acquired from SigmaAldrich, Fisher Scientific, and MP Biomedicals The Rodeo ECL Western Blotting Reagent Pack was obtained from Affymetrix. The rabbit antiserum against the S. cerevisiae b subunit was a generous gift from Dr. Rosemary Stuart (Marquette University). Restriction digests, ligations, and polynucleotide k inase treatments were done according to the manufacturers instructions. Transformations Transformations of S. cerevisiae were performed essentially as described by Geitz et al (Gietz and Woods, 2002). A 10 mL overnight culture was inoculated using complete synthetic media without the amino acid uracil with 2% (w/v) glucose. The following morning the culture was diluted to an OD600 of 0.5 in 10 mL of YPD.

PAGE 137

137 After growing the S. cerevisiae for 3 5 hours at 30 C at 300 rpm, the cultures were centrifuged at 3000g for 5 minutes. The pellet was suspended in 5 mL sterile ddH2O and centrifuged again at 3000g for 5 min. Next, the pellet was suspended in 200 L 100 mM LiCOOH. It was then transferred to 1.5 mL microcentrifuge tube and centrifuged using a table t op centrifuge at 18000g for 5 seconds. The cells were suspended in 100 L of 100 mM LiCOOH and 50 L of the suspension were transferred into two separate, sterile microcentrifuge tubes. One tube was for the DNA transformation; the other tube was for the negative control transformation (i.e. using sterile ddH2O instead of DNA). Again, the tubes were centrifuged at 18000g for 5 seconds. The supernatant was removed and the following were added the in this order: 240 L 50% (w/v) PEG 3500, 36 L 1 M LiCOOH, 50 L 2 mg/mL singlestranded herring testes DNA (Clontech), 34 L DNA or sterile ddH2O. This transformation mixture was mixed using a vortex for approximately 1.5 minutes. This was at times difficult because the 50% PEG was very viscous. In those cases the mixture was incubated at room temperature for about five minutes. In either event, it took about 1.53 minutes to suspend properly. The mixture was incubated at 30 C for at least 30 minutes. After that incubation period, the tube was mixed using a vortex briefly and incubated at 42 C for 45 minutes. Next, the transformation mixture was centrifuged for 5 seconds at 18000g and the supernatant was removed. If the selective marker was an antibiotic resistance gene, like the KanMx6 cassette, then the pellet was suspended in 1 mL YPD and grown, shaking, at 30 C for 35 hours. If the selectable marker was an auxotrophic marker then, this step was skipped and the pellet was suspended in 1 mL sterile ddH2O. 10 L and 100 L of the resultant suspensions were plated on to appropriate selective

PAGE 138

138 media and placed in a 30 C incubator for at least 2 days. Resulting colonies were streaked and patched. Colony PCR Colony PCR is a technique used for isolating genomic and plasmid DNA directly from a S. cerevisiae colony that is suitable for PCR. The DNA tended to be of lower quality and, as such, was used only for screening purposes. On those occasions where the goal was to amplify a portion of genomic DNA for cloning purposes, or other uses where high quality D NA was important, then the genomic DNA isolation protocol was used (see below). Using a micropipette tip, a small amount of the colony was transferred to 30 L 0.2 % SDS in a microcentrifuge tube and mixed using a vortex on high for 30 seconds. The tube was then transferred to a 95 C heating block and incubated for 4 minutes. The tube was placed in a centrifuge and spun for 1 minute at 18000g. The supernatant was transferred to a clean tube. The supernatant contained the crude DNA and could be stored at 20 C for up to one month. For the PCR step, Platinum Taq polymerase (Invitrogen) was used because it is a high fidelity polymerase that accepts higher salt conditions found in crude DNA extracts. The reaction mixture was prepared on ice. The follow ing components were combined: 5 L 10X Platinum Taq PCR Buffer, 1.5 L 50 mM MgCl2, 1 L 10 mM dNTPs, 2 L 50 ng/L Primer AW89, 2 L 50 ng/L Primer AW78S, 2 L 25 % TritonX 100, 37.2 L sterile ddH2O, 0.3 L Platinum Taq polymerase, and the last component added was1 L crude DNA extract. There were several important steps in this procedure. First, the crude DNA extract had to be added last. Second, only 1 L of the DNA extract could be added. If the PCR failed, then 0.5 L of the crude DNA was used

PAGE 139

139 in a second attempt. Finally, to make the 25 % TritonX, the water was heated to 65 C and the TritonX 100 was added slowly to the hot water. The PCR cycling conditions were as follows: (1) 95 C for 1 min, (2) 95 C for 30 seconds, (3) 54 C for 1 min, (4) 72 C for 1 min/ kb product, (5) repeated steps 2 4 for 35 cycles total, (6) 72 C for 6 min, (7) held at 4 C. For an agarose gel assay, 5 L of the PCR product was used. For sequencing the PCR product, first the PCR product was purified using a QiaQuick PCR Clean Up kit (Qiagen) according to the manufacturers instructions. However, the DNA was eluted in the optional 30 L of buffer EB. Between 6 10 L of the purified PCR product was sufficient for DNA sequencing by the core facility. Construction of ATP4syn Expression Vectors The ATP4syn shuttle vectors were constructed by a multiplestep protocol. The ATP4syn gene was constructed by Genscript to my specifications. It was originally received as an insert in the plasmid pUC57 and named pAW16 (Apr) (Table 41). The plasmid pAW16 (Apr) was digested with Sac II to remove an extra seven codons located at nucleotide number 2279 and ligated, yielding plasmid pAW17 (Apr). To facilitate future mutagenesis experiments the Kas I site located in the pAW17 (Apr) plasmid backbone was removed by digesting the plasmid with Nde I and Nru I. The sticky ends from those restriction sites were filled using the Klenow fragment of DNA polymerase I and then blunt end ligated. The resulting plasmid was named pAW20 (Apr) and was used as the initial target vector for all further mutagenesis experiments. To move the ATP4syn gene from pAW20 (Apr) the plasmid was digested with Eco RI and Not I. This allowed the intact ATP4syn gene to be ligated into plasmid pRS313 (Apr, HIS3 ) that had previously been digested with Eco RI and Not I, yielding

PAGE 140

140 plasmid pAW26 (Apr, HIS3 ) (Figure 41A). PCR primers, AW Prom Sal and AW Prom Nco, were used to amplify the 5 regulatory region from genomic DNA of the BY4743 S. cerevisiae strain (Table 42). Included in the primers were flanking SalI and Nco I sites used to insert the PCR product upstream of the ATP4syn gene to yield plasmid pAW31 (Apr, HIS3 ) (Figure 41B). Similarly, the downstream region of the ATP4 gene was amplified using primers, AW Term Not and AW Term Sac, containing Not I and Sac I sites, respectively (Table 42). That PCR product containing the downstream region and the plasmid pRS313 (Apr, HIS3 ) were digested with Not I and Sac I and the two were ligated together to generate pAW28 ( Apr, HIS3 ) (Figure 41C). Agarose gels showing the PCR products from amplification of the 476 bp 5 regulatory region and the 807 bp region downstream of the ATP4 gene are shown in Figures 42A and 42B. Next, pAW26 (Apr, HIS3 ) and pAW28 (Apr, HIS3 ) were digested with Sal I and Not I. The resulting fragments containing the 5 regulatory region and ATP4syn, and the other fragment containing the 3 region in pRS313 (Apr, HIS3 ) were ligated together to give pAW33 ( ATP4syn Apr, HIS3 ) (Figure 41D). At the e nd of all construction, the final 5 regionATP4syn3 region construct was sequenced. To obtain an E. coliS. cerevisiae shuttle expression vector that contained the auxotrophic marker URA3 a similar process to the above was followed for plasmid pRS416 ( ATP4syn, Apr, HIS3 ) (Figure 41E). Isolation of Genomic and Plasmid DNA from S. cerevisiae A tube holding 10 mL of an appropriate medium was inoculated from a S. cerevisiae patch. The culture was grown at 30 C overnight, then diluted to an OD600 of 0.5 and allowed to grow for 4 hours at 30 C. The culture was then centrifuged at 3000g for 5 min in a refrigerated centrifuge. The cell pellet was suspended in 5 mL

PAGE 141

141 sterile ddH2O and centrifuged at 3000g for 5 minutes. The pellet was suspended in 0.5 mL s terile ddH2O, transferred to a clean microcentrifuge tube, and centrifuged at 18000g for 30 seconds. Next, the following solutions were added in this order: 200 L breaking buffer (2 % (v/v) Triton X 100, 1 % (v/v) SDS, 100 mM NaCl, 10 mM Tris HCl (pH 8.0), 1 mM EDTA (pH 8.0), stored for less than one year), 0.3 g acid washed glass beads (Sigma Catalog # G877210G), and 200 L 25:24:1 phenol: chloroform: isoamyl alcohol. The mixture was mixed using a vortex at maximum speed for 3 minutes. Following the m ixing step, 200 L TE (10 mM Tris (pH 8.0), 1 mM EDTA (pH 8.0)) was added and the sample was mixed briefly using a vortex. The mixture was centrifuged at 3000g for 5 minutes. The top aqueous layer was removed and placed in a clean microcentrifuge tube. To the tube with the aqueous layer, 1 mL of 100 % ethanol and 10 L 7.5 M NH4OAc was added. The mixture was centrifuged at 18000g for 5 minutes. The supernatant was aspirated off and the pellet was suspended in 100 L TE. Two L of DNA was used for all PCRs. Sequencing of Genomic or Plasmid DNA First, it is important to note that it was difficult to separate plasmid DNA from genomic DNA isolated from a S. cerevisiae cell. In order to sequence either DNA, it was necessary to perform two PCRs. However, i f the plasmid DNA was isolated from a bacteria cell, only the second PCR step was necessary. The first reaction required a primer specific to the plasmid DNA or genomic DNA of interest (Primer AW45 in Figure 4 3) and another primer either up or downstream (Primer AW Term Sac in Figure 43). This reaction was carried out as discussed under Colony PCR except that 2 L of DNA was used. Following the PCR, the reaction was purified using the QiaQuick PCR Clean up Kit (Qiagen). Next, a sequencing PCR was performed using a sequencing

PAGE 142

142 primer that was compatible with the resulting fragment (Primer AW78S in Figure 43). The reaction conditions for the sequencing PCR were as follows: 1 L DMSO, 1 L 50 ng/L Sequencing primer, 8 L BigDye, 10 L DNA. The cycl ing conditions were: (1) 96 C for 1 minute, (2) 96 C 30 seconds, (3) 50 C 15 seconds, (4) 60 C 4 minutes, (5) repeated steps 24 for a total of 35 cycles, (6) held at 4 C. Sequence determination was done by automated sequencing in the Center for Epigenetics core facility at the University of Florida. Preparative Procedures For mitochondrial isolation, the strains were grown in rich media with 2% galactose as a carbon source to allow for growth of the strains. Biological F1FO ATP synthase activity w as determined by plating dilutions on plates containing rich media with 2% glycerol and 0.1 % glucose and incubating at 30 C for 3 days. Preparation of mitochondria was done essentially as described by Meisinger et al (Meisinger et al. 2006). An overnight of each strain with 5 mL CSM His with glucose was inoculated. From that 5 mL overnight, 50 mL YPGal was inoculated and incubated with shaking at 30 C for 12 days. After cells have reached stationary phase 500 mL of YPGal was inoculated with between 520 mL of the 50 mL culture and grown for 20 hours. The cells were harvested by centrifugation at 3000g for 5 minutes, washed once with sterile doubledistilled H2O (ddH2O) and weighed. The cells were suspended in the DTT buffer that had been prewarmed t o 30 C at 2 mL buffer/g wet cells. The suspension was incubated with shaking slowly (80 rpm) for 20 minutes at 30 C. The cells were centrifuged at 3000g for 5 minutes and suspended in zymolyase buffer (ZB) (1.2 M sorbitol, 20 mM potassium phosphate, pH 7.4) at 7 mL/g wet cells. The

PAGE 143

143 suspension was centrifuged again at 3000g for 5 minutes. Again the cells were suspended in ZB at 7 mL/g wet cells with zymolyase 20 T (Fisher Scientific) added at 3 mg/g wet cells and incubated at 30 C with slow shaking for 3045 minutes. After digestion, the cells were centrifuged at 3000g for 5 minutes. The pellet was carefully suspended in ZB at 7 mL/g wet cells and centrifuged again at 3000g for 5 minutes. The pellet was carefully suspended in homogenization buffer (0. 6 M sorbitol, 10 mM Tris HCl, pH7.4, 1 mM ethylenediaminetetraacetic acid (EDTA), and 1 mM phenylmethylsulfonyl fluoride (PMSF) added fresh) that had been precooled to 4 C at 6.5 mL/g wet cells. From this point forward everything was kept ice cold and all buffers, equipment, and tubes were kept on ice. The spheroplasts were lysed with 15 strokes of a glass Teflon homogenizer. Afterwards the homogenate was diluted twofold with homogenization buffer. Then, the diluted homogenate was centrifuged at 1500g for five minutes to pellet cell debris and nuclei. The supernatant was transferred to a clean centrifuge tube and the supernatant was centrifuged at 4000g for 5 minutes. The supernatant was transferred to a second clean centrifuge tube and centrifuged for 15 minutes at 12000g. This pellet contained the mitochondrial fraction. The pellet was gently suspended in 250 L of 4 C SEM Buffer (250 mM sucrose, 1 mM EDTA, 10 mM MOPS KOH, pH 7.4) by using either a 1000 L tip or a 200 L tip with the last few millimeters at the end cut off. The mitochondria were stored at 4 C until ATPase assays or a western blot could be performed. The longest period for storage at 4 C was one week. Remaining mitochondria were flash frozen and stored at 80 C. Protein concentration was determined using the bicinchoninic acid method (Smith et al 1985).

PAGE 144

144 Assay of F1FO ATP synthase activity Assays of ATP hydrolysis were performed essentially as described by Tzagoloff and Ackerman (2006). The assay was run in duplicate (i.e. two reactions per sample) plus one blank. First, the mitochondrial protein samples were diluted to 5 mg/mL. Next, each reaction was set up using 0.5 mL 2x ATPase assay buffer (0.1 M Tris H2SO4, pH 8.5, 8 mM MgSO4) and 0.38 mL ddH2O. Finally, 20 L mitochondria protein (5 mg/mL) was added to each reaction tube. To the second set of reactions 5 L of oligomycin (2 mg/mL in ethanol) was added. The tubes were incubated at 37 C for one minute. The reactions were started by adding 100 L of 0.1 M ATP, pH 7.2 (adjust pH with NaOH) and each tube was mixed using a vortex briefly. The reactions were incubated at 37 C for 12 minutes. To stop the reaction 0.2 mL of 50% trichloroacetic acid was added and mixed using a vortex. Clean tubes were set up to develop the reacti on, one for each reaction, 5 mL of 0.5% trichloroacetic acid solution were added. Next, 0.2 mL of the stopped assay solution was pipeted into each tube. Then, 0.5 mL of 5% ammonium molybdate solution was added and each tube was mixed using a vortex brief ly. Finally, 0.15 mL ANS reagent (15% (w/v) NaHSO3, 6% (w/v) Na2SO3, and 0.25% (w/v) amino napthol sulfonic acid) was added to each tube and mixed using a vortex briefly. The reactions were incubated at room temperature for 10 minutes and absorbance was at 660 nm against the blank. The absorbance reading is equal to the micromoles of inorganic phosphate (Pi) released per minute. The specific activity was calculated by dividing the absorbance reading by the milligrams of protein added to the assay. Immunoblot analysis Mitochondrial protein samples (30 g protein /well) were loaded onto a 12% Tris glycine SDS BioRad TGX Gel. Following electrophoresis, the mitochondrial proteins

PAGE 145

145 were transferred to a PVDF membrane (BioRad) by electrobloting (110 V, 20 mi nutes, 4 C). Detection was performed using the Rodeo ECL Western Kit (Affymetrix). The membrane was washed twice with TBS (1 mM Tris HCl, 15 mM NaCl, pH 7.2). The membrane was then blocked for one hour shaking, at room temperature with TBS S (1 mM Tris H Cl, 15 mM NaCl, pH7.2, 0.05% Saddle Soap, Affymetrix) with 2% Rodeo Blocker (Affymetrix). The rabbit antiserum against the b subunit was added to the blocking buffer at a concentration of 1:1000 and incubated, shaking, at room temperature for one hour. N ext, the membrane was washed three times with blocking buffer. The secondary antibody, anti rabbit IgG, was added to the blocking buffer at a 1:30,000 concentration and incubated with shaking at room temperature for one hour. Following this incubation, t he membrane was washed three times with TBS S. The antibody was detected using chemiluminescence (Affymetrix). The signals were visualized using high performance chemiluminescence film (GE Life Sciences) and a Kodak X Omat. Signal strength was assessed using UnScan It gel digitizing software (Silk Scientific, Inc). Results Construction of ATP4syn Expression Vectors The native ATP4 gene is a difficult target for sitedirected mutagenesis because it is both AT rich and has very few naturally occurring restriction sites. To facilitate the mutagenesis experiments in this report and future experiments, a synthetic gene was designed. The primary considerations for synthetic ATP4 gene design were: (i) to position flanking restriction sequence sites for moving the synthetic gene into the expression vector; and (ii) to maximize the number of unique restriction sequence sites spaced over the entire coding sequence of the gene (Figure 44). Eco RI and Not I

PAGE 146

146 restriction sites were chosen to flank the 5 and 3 ends of the sequence, respectively (Figure 44A). The restrictions sites within the coding sequence were introduced through silent mutations (Figure 44B). However, these silent mutations altered codon usage. An attempt was made to balance codon usage by mutating other nucleotides to increase the number of S. cerevisiae high preference codons designed into the synthetic ATP4 gene ( ATP4syn)1 (Figure 44B). The flanking restriction sites were used to facilitate the movement of the ATP4 gene from pAW20 (Apr) (see Materials and Methods) to the novel E. coli S. cerevisiae shuttle vector (described below). To construct an ATP4syn expression system for generating recombinant b subunits, it was necessary to build an E. coliS. cerevisiae shuttle vector. On the surf ace, many commercially available E. coli S. cerevisiae shuttle vectors appear have to a promoter and terminator that might have served the needs of the planned experiments. However, use of any of those vectors would have required either very extensive S cerevisiae strain construction (i.e. for use of the Met promoter) or would have required addition of carbon sources to culture media that complicated the determination of the mutant phenotypes (see below). Therefore, we elected to build a new expression vector with the 5 regulatory region and the 3 region of the native ATP4 The reasons for this were twofold. The native promoter is thought to be a strong, constitutive promoter and previous work done by other groups showed that the promoter yielded appropriate levels of b subunit (Velours et al. 1989; Gavin et al. 2003). The parental plasmids, pRS313 and pRS416, were chosen to serve as platforms for expression shuttle vector construction for two reasons: (i) they are stable 1 Italicized ATP4syn denotes the gene. Nonitalicized ATP4syn denotes either the protein or strain, which is defined in the text.

PAGE 147

147 due to the centromeric ori gin of replication; and (ii) these plasmids offered two different auxotrophic markers. An option for choosing an auxotrophic marker was needed for Construction of Mutant Strains (s ee below). The ATP4syn shuttle vectors were constructed by a multiplestep protocol (see Materials and Methods). The ATP4syn gene was moved from the plasmid pAW20 (Apr) into either plasmid pRS313 (Apr, HIS3 ) or pRS416 (Apr, URA3 ). The 5 regulatory re gion and 3 region were amplified from genomic DNA (Figures 42). The PCR products were inserted upstream and downstream, respectively, of the ATP4syn gene to yield either pAW33 (Apr, HIS3 ) or pAW36 (Apr, URA3 ) (Figures 41D and 41E). At the end of all c onstruction, the final plasmid containing the 5 regulatory region, ATP4syn gene, and 3 region was sequenced. ATP4 Strain The deletion of the ATP4 gene from the genome causes S. cerevisiae cells to go either or 0. That is their mitochondrial DNA either becomes damaged, -, or entirely missing, 0 (Velours et al. 1988). This is possibly due to the lack of F1FO ATP synthases present in the mitochondria. It has previously been shown that the enzyme was important to formati on of mitochondrial cristae. To overcome this problem, a scheme was devised to maintain a copy of either the native ATP4 gene or the ATP4syn gene in the cell at all times during all steps involved in chromosomal ATP4 gene deletion strain construction and m utant ATP4syn plasmid transformation (Figure 45). Parental S. cerevisiae strain BY4743 was a diploid strain homozygous for the wild type ATP4 gene. A KanMx6 cassette, which contains the kanamycin resistance gene and its own

PAGE 148

148 promoter and terminator sequence, was used to replace the ATP4 gene on one chromosomal allele. The selectable marker was chosen because there is little or no background of nontransformed cells on G418 selection medium. Primers AW83 and AW84 were used to amplify the KanMx6 cassette a h obtained from ATCC (catalog no. 4012750) (Table 42). This resulted in an 1780 bp PCR product containing the KanMx6 cassette flanked by 150 bp of homology both upstream and downstream of the ATP4 gene. The linear PCR product was transformed directly into the strain BY4743 (Figure 45). S. cerevisiae has an excellent homologous recombination system and will readily incorporate linear pieces of DNA that contain regions of homology into their genome (Sikorski and Boeke, 1991). Incorporation of the KanMx6 cassette into one ATP4 allele of the cells was detected by screening for growth on solid YPD media supplemented with 250 g/mL (active) G418 (YPD+G418) and YPG. Transformants capable of growth on YPD+G418 media were selected and genomic DNA was prepared. The genomic DNA was amplified and sequenced for both the native ATP4 gene and the KanMx6 cassette (Figure 46). The resulting heterozygous strain was named AW2 (MATa /MAT hisuraatp4::KanMx6/+). Next, plasmid pAW36 ( ATP4syn, Apr URA3 ) was transformed into strain AW2. Sporulation was induced by a shift to solid media containing potassium acetate and reduced amount of glucose and nutrients. The resulting tetrads were dissected by use of a micromanipulator (Figure 47). Haploi d colonies were screened for growth on three types of media: (i) solid synthetic media without uracil, (ii) solid YPD+G418 media, and (iii) solid rich media supplemented with 3% glycerol (YPG) to test for functional ATP

PAGE 149

149 synthase activity. Those colonies t hat grew on all three media were genotyped for a chromosomal KanMx6 cassette disruption of the chromosomal copy of the ATP4 gene and the presence of the plasmidborne ATP4syn gene (Figure 48). The strain was subsequently tested for mating type and was found to be Mat a and was named AW3 (Mata hisuraatp4::KanMx6, pAW36 ( ATP4syn, Apr, URA3 )). To determine if the ATP4syn gene was capable of complementation of the chromosomal ATP4 gene disruption, the strain was patched on to solid YPG medium to test in vivo function. The ATP4syn strain grew robustly on YPG medium, and the level of growth was comparable to that of the wildtype parent strain BY4743 (Figure 49). This showed that the plasmid synthetic ATP4syn gene allowed phenotypically normal F1FO ATP synt hase activity in vivo Construction of Mutant Strains Mutations were designed to test the plasticity of the eukaryotic peripheral stalk by lengthening and shortening the b subunit of S. cerevisiae Previous work in our lab has shown that the E. coli peripheral stalk can tolerate deletions of up to 11 amino acids and insertions of up to 14 amino acids within the tether domain of a functional F1FO ATP synthase (Sorgen et al. 1998b, 1999). However, it was not known whether the plasticity detected in the prokaryotic stalk was a property shared with the eukaryotic enzyme. Therefore, the ATP4syn genes were constructed with up to 14 codons deleted and inserted in ATP4syn gene of S. cerevisiae Specifically, deletions and insertions of four, seven, eleven, and fourteen codons were designed. The number of codons inserted or deleted was based on the addition or removal of turns of an helix. The site for the manipulations was selected by inspection of the highresolution crystal structure

PAGE 150

150 of the B. taurus p eripheral stalk (Dickson et al. 2006) and the cryoelectron microscopy image of the S. cerevisiae F1FO complex (Rubinstein et al. 2003). In both structures, the peripheral stalk narrowed to the width of a single helix (Figures 19 and 110). This narr owing of the peripheral stalk coincides with the same region in the E. coli enzyme where the insertion and deletion work was done by Sorgen et al (Sorgen et al 1998b,1999) Next, the target in the S. cerevisiae peripheral stalk was chosen by comparing deduced amino acid sequence of the B. taurus b subunit to that of the S. cerevisiae. The two genes share 44% sequence similarity (Figure 410). This comparison showed that the narrow region in the B. taurus peripheral stalk structure was analogous to amino acids 78 94 in the S. cerevisiae b subunit. However, there were two short segments of highly of conserved amino acids: 7983 and 100108. Care was taken to avoid changing these regions as much as possible. Therefore, the segment of the ATP4syn gene encoding amino acids 81 and 94 was selected as the target for mutagenesis. The majority of the manipulations were constructed in the sequence encoding amino acids 8393. All initial mutagenesis protocols for the insertion and deletion constructs were done us ing plasmid pAW20 (Apr, ATP4syn). Mutagenic oligonucleotides were designed to either insert or delete 4, 7, 11, or 14 codons (Figure 411). In the case of the insertions, the coding sequence for the previous 4, 7, 11, or 14 amino acids were repeated with alternative codons for the amino acids wherever possible. Each mutant ATP4syn gene was constructed in essentially the same manner by ligationmediated site directed mutagenesis. First, the pAW20 plasmid was digested with the appropriate restriction enzy mes for the mutagenic oligonucleotide set. For example, to generate an

PAGE 151

151 ATP4syn gene with an insertion of four codons, the plasmid pAW20 was digested with Bss HII and Xho I. Next, the digested pAW20 plasmid was ligated to the doublestranded oligonucleotide specifying the insertion of codons lysine 87 to aspartic acid 90, as shown in Figure 411. In the case of the insertion of four codons this generated plasmid pAW40 (Apr) (Table 41). To move the ATP4syn+4 gene pAW40 (Apr) was digested with Sac II and Not I and the fragment was ligated into the E. coliS. cerevisiae shuttle vector pAW33 ( ATP4syn, Apr, HIS3 ). The resulting plasmid pAW42 ( ATP4syn+4 Apr, HIS3 ) (Table 41), was screened by restriction site digestion for the insertion of an Aat II restriction si te and the nucleotide sequence determined using primer AW45 (Table 41). Other E. coliS. cerevisiae shuttle vectors generated in this manner are shown in Table 4 3. Mutant ATP4syn expression plasmids were transformed individually into the S. cerevisiae strain AW3 (hisuraatp4:: KanMx6, pAW36 (Apr, URA3 ATP4syn)) to yield cells that contained both the pAW36 plasmid (Apr, URA3 ATP4syn) and a mutant plasmid (Apr, HIS3 mutant ATP4syn) (Figure 45). The pAW36 plasmid was selected against by replica plating the S. cerevisiae cells on solid YPD+5 Fluoroorotic Acid (5FOA) media. The URA3 gene encodes the enzyme orotine5'monophosphate decarboxylase. That enzyme converts 5FOA to 5 fluorouracil. 5fluorouracil inhibits thymidylate synthase, an enzyme that contributes to the biosynthesis of thymidine, and is, therefore, toxic to the cell. After selection on solid YPD+5FOA media, surviving cells are patched on to solid CSM His media. If there was any doubt that the pAW36 plasmid had been removed from the cell, the patches were replica plated onto solid CSM URA medium. Lack of growth on the medium provided assurance that the plasmid had been

PAGE 152

152 removed. Genomic DNA from selected patches grown on solid CSM His medium were isolated and sequenced for the presence of the desired mutant ATP4syn gene. Growth Properties of Mutant Strains Recombinant ATP4syn genes were tested for the ability to complement the disruption of the endogenous ATP4 gene in vivo by growth on YPG or YPG with 0.1 % glucose (YPGD) media. Use of glycerol as a carbon source requires a functional, intact F1FO ATP synthase for growth. Inclusion of a limited amount of glucose in YPGD media allowed slow for growth of petite mutants lacking a functional enzyme. The insertion and deletion strains showed drastically different growth phenotypes. The deletion strains, with the exception of the ATP4syn growth on glycerol and little growth on YPGD medium. This suggested that the majority of the deletion strains exhibited the petite phenotype on YPGD (Figure 412A). Of all the deletion strains, the only one that indicated complementat ion was the ATP4syn strain that showed growth on YPG and YPGD comparable to that of the control ATP4syn strain. In contrast to the majority of the deletion strains, all of the insertion strains grew on both YPG and YPGD media. With the exception of the ATP4syn+14 strain, the insertion strains showed robust growth on both types of media (Figure 412B). Only the largest insertion in the ATP4syn+14 strain showed visibly less growth on both YPG and YPGD. The colonies were very small and the growth yield was poor indicating that the ATP4syn+14 strain F1FO ATP synthase was just barely capable of biologically significant function. Expression of Recombinant b Subunits Western blots were performed on mitochondria isolated from the mutant strains using anti b subunit antiserum to determine steady state levels of the recombinant b

PAGE 153

153 subunits (Figure 413). The presence of the b subunit band indicated that the protein was imported to the mitochondria properly. To ensure that the anti b antiserum was specific a negative control from strain on the western blot (Lanes 1 in Figures 413A and 413B). The positive control from the ATP4syn strain showed a strong signal indicating that the synthetic gene product was expressed and imported into the mitochondria (Lanes 2 in Figures 413A and 413B). Expression levels of the ATP4syn+4 and ATP4syn b subunits were comparable to wild type and reflected the growth phenotype of those strains. However, the similarity of the expression levels between the inser tion and deletion strains ended there. The gene products of the ATP4syn+7 and ATP4syn+11 strains were expressed at levels approximately 60% of that of wild type. In contrast, the ATP4synsyn b subunits were detected at much lower levels, 34% and 16% of wild type respectively (Figure 143B). The most dramatic difference between the insertion and deletion strains was between the strains containing F1FO complexes with ATP4synsyn+14 subunits. The band from ATP4syn b subunit strain was detectable on the original film, but only after a very long exposure time, but the ATP4syn+14 gene product was readily detectable at levels reaching 49% of wild type. It is interesting to note that while the ATP4synsyn+11 proteins were readily detectable, the former does not grow on media containing glycerol as a carbon source and the latter grew very well on YPG media (Table 43). Additionally, the expression level of the ATP4syn+14 protein 49% of wildtype, representing only a 10% decrease in expression levels from the ATP4syn+7 and ATP4syn+11 proteins, but there was a significant decrease in the growth on YPG and YPGD media. Therefore,

PAGE 154

154 expression level of the b subunit did not necessarily directly correlate to growth phenotypes of the mu tant strains. F1FO ATPase Activity In mitochondriaenriched samples, the greatest single source of ATP hydrolysis activity comes from the F1FO ATP synthase. Traditionally, the amount of Pi release that is specifically assigned to the enzyme is the level l ost by addition of the inhibitor, oligomycin. Oligomycin interacts with the FO proton channel at a site thought to be located at the interface between the a subunit and the ring of c subunits to inhibit the enzyme (Nagley et al. 1986). In the absence of oligomycin, the ATP4syn strain mitochondria had significantly syn (5.3 versus 3.03 mole Pi min1mg protein1) (Table 44). With the exception of the ATP4syn mitochondria from all strains exhibited inorganic phosphate release levels reflective of their growth phenotype. That is, mitochondria from the largest deletion strains that did not grow on YPG showed ATP hydrolysis levels similar to the negative control. In contrast, the insertion strains and the smallest deletion strain that grew on YPG had ATP hydrolysis levels approaching the ATP4syn strain. The exception was mitochondria from the ATP4syn levels. Although the ATP4syn in vivo the complex was apparently unstable during preparation of the mitochondria. In the presence of oligomycin, the ATP4syn strain showed a 57% decrease in ATP hydrolysis activity, as expected based on previous literature (Gavin et al. 2003; Weimann et al. syn strain, showed negligible oligomycin sensitivity (2%). In line w ith the results from in the absence of oligomycin,

PAGE 155

155 most of the deletion strains showed no appreciable oligomycin sensitivity (Table 44). In contrast, ATP4syn ATP hydrolysis. For the insertion F1FO complexes, this was a significant departure from their previous pattern essentially of wildtype characteristics. The similarity of the oligomycin sensitivity result for the ATP4syn ATP4syn+7 strains (35% and 47%, respectively) was interesting. The two strains exhibit opposite growth phenotypes. The ATP4syn+4, and ATP4syn+11 strains showed negligible oligomycin sensitivity, but had similar growth phenotypes to the ATP4syn strain. This led us to suggest that in at least the ATP4syn+4, and ATP 4syn+11 strains the oligomycin may be prevented from binding to the a subunit in the appropriate way to inhibit an otherwise fully active F1FO ATP synthase. Discussion In this chapter, I report the construction, expression and characterization of a series of mutations insertion and deletion of amino acids in the b subunit of S. cerevisiae F1FO ATP synthase. Insertions of up to eleven amino acids yielded a wild type phenotype for growth on glycerol as a carbon source. The mitochondriaenriched fractions had readily detectable levels of the insertion subunits and abundant ATP hydrolysis activity. This activity showed surprisingly little oligomycin sensitivity when considered in the context of growth phenotype. Additionally, expression of the ATP4syn+14 b subunit complemented the ATP4 deletion strain for at least limited growth and measurable F1FO ATPase activity was apparent in the mitochondria. In contrast, most of the deletion mutations resulted in loss of F1F0 ATP synthase function. Only the smallest deletion subunit ATPsyn

PAGE 156

156 However, the F1FO complex with this shortened subunit had no apparent activity in vitro suggesting enzyme instability during preparation mitochondrial enriched fractions. Previous work done in the Cain laboratory showed that the E. coli F1FO ATP synthase could tolerate large changes in length of the b subunits (Sorgen et al. 1998b, 1999). Those experiments were interpreted as evidence that considerable plasticity was an inherent property of the bacterial peripheral stalk. The experiments presented here suggest that the eukaryotic peripheral stalk has similar plasticity, at least with respect to the insertions. Normal growth characteristics on nonfermentable carbon sources indicated that active eukaryotic F1FO ATP synthases capable of accommodating b subunits ranging from four amino acids shorter to eleven amino acids longer than the normal subunit. This represents a change in length of the b subunit of approximately 18 To put this in perspect ive, the total distance from the top of the membrane to the bottom of F1 is approximately 40 so a range of 18 represents a change of 45% in the length of the peripheral stalk segment between the F1 and FO sectors in an intact F1FO complex. An addition al turn of an helix in the ATP4syn+14 subunit resulted in substantially less enzyme assembly. The evidence from the deletion experiments is less clear. The eukaryotic enzyme lost all function when deletions were extended beyond only 4 amino acids, while the bacterial enzyme accommodated much longer deletions of up to 11 amino acids (Sorgen et al. 1998b). In bacteria, the largest deletions resulted in fewer assembled F1FO ATP synthases. These enzymes were apparently fully active and stable during membrane preparation. A defect in assembly of the F1FO complex likely accounted for the loss of function found in the S. cerevisiae The lack of properly

PAGE 157

157 assembled enzyme is associated with deformation or loss of the mitochondria leading to an overall defici ency in energy metabolism (Dautant et al. 2010). The longer deletion strain cells failed to generate enough ATP to survive on a nonfermentable carbon source. Although there was abundant ATP synthesis activity apparent in vivo, even the ATP4syn b subunit F1FO complex proved unstable during mitochondrial preparation. The evidence would seem to suggest that the wild type S. cerevisiae b subunit is very near its minimum functional length. Considering the sequence similarity between S. cerevisia e and B. taurus b subunits, especially within the segment crystallized by Dickson et al it appears likely that eukaryotic peripheral stalks evolved to favor the shortest practicable peripheral stalks (Dickson et al. 2006). One of the most surprising fi ndings was the lack of oligomycin sensitivity associated with the majority of the mutant ATP4syn F1FO ATP synthases. With the exception of the ATP4syn+7 enzyme, all of the mutant strains that contained functional F1FO ATP synthases were essentially insensi tive to oligomycin. The result suggests that a subtle conformational change occurred F1FO as a result of incorporation of the ATP4syn+4, ATP4syn+11, and ATP4syn+14 subunits. The shift appeared to be propagated within the FO sector to the oligomycin bindi ng site within the proton channel. Importantly the shift did not cause any apparent loss in channel function, only inhibitor binding seemed to be affected. There are two important implications of the result. First, this calls into question the validity of the widespread use of oligomycin sensitivity as a measure the amount of ATP hydrolysis attributable to F1FO ATP synthase in mitochondrial preparations. Certainly, the approach is not valid for investigation of mutations affecting the b subunit and by extension other peripheral stalk subunits. This

PAGE 158

158 conclusion is further supported because the h subunit can be functionally replaced by the B. taurus F6 subunit and inhibition by oligomycin was abolished (Velours et al. 2001). Second, propagation of a conf ormation shift in the b subunit implies that changes in peripheral stalk length may not be fully accounted for by structural changes within the stalk itself. This idea will be more fully explored in Chapter 5.

PAGE 159

159 Table 41. Plasmids generated for ATP4syn e xpression Plasmid Name Description of Plasmid Restriction site(s) for screening Selection Marker Notes pAW17 ATP4 syn Xho I Ap pAW20 pAW17 Kas I in ATP4 syn only Kas I Ap pAW17 based pAW21 ATP4 syn Ap pAW20 based pAW24 ATP4 syn +7 Aat II Ap pAW20 based pAW26 ATP4 syn in pRS313 Xho I Ap, HIS3 pAW20 based pAW28 pRS313 Term Ap, HIS3 Terminator PCR pAW31 Prom+ ATP4 syn Xho I Ap Correct promoter pAW26 based pAW33 Prom+ ATP4 syn +Term Xho I Ap, HIS3 pRS313 based pAW34 Prom+ ATP4 syn +7 +Term Aat II Ap, HIS3 pRS313 based pAW35 Prom+ ATP4 syn +Term Xho I Ap, HIS3 pRS313 based pAW36 Prom+ ATP4 syn +Term Xho I Ap, URA3 pRS416 based pAW37 Term+pRS416 Ap, URA3 Terminator PCR pAW39 ATP4 syn Ap pAW20 based pAW40 ATP 4 syn + 4 Aat II Ap pAW20 based pAW41 ATP4 syn in pAW33 Ap, HIS3 pAW33 based pAW42 ATP4 syn +4 in pAW33 Aat II Ap, HIS3 pAW43 ATP4 syn +14 Aat II Ap pAW20 based pAW44 ATP4 syn +11 in pAW33 Aat II Ap, HIS3 pAW45 ATP4 syn +14 in pAW33 Aat II Ap, HIS3 pCJB1 ATP4 sy n Ap pAW20 pCJB2 ATP4 syn in pAW33 Ap, HIS3 pCJB3 ATP4 syn Ap pAW20 pCJB4 ATP4 syn in pAW33 Ap, HIS3 pRQ1 ATP4 syn +11 Aat II Ap pAW20 based

PAGE 160

160 Table 42. Primers used in S. cerevisiae mutagenesis experiments Primer Designation Sequence a A W Prom Sal CCCCCCC GTCGAC GTGTTGTGACGCAACTGC AW Prom Nco CCCCCCCC CCATGG AAGGACAACGAACACCTTGGC AW Term Not CCC GCGGCCGC TCACAACAGTAACTGCG AW Term Sac GGGG GAGCTC GCCAATGGTTCTATCCAAAAGGG AW45 CAAGGTGTTCGTTGTCCTTCGATGGTGATT AW78 S CTTGTCGCAGTTACTGTTGTGATTACTTC AW83 CAAGAAGAGATATATAACCTGAGCATCC AW84 CAATCACGACGCTTTTTCTCTTCAC AW89 GAGGATACACCCTTAAGAATGTG aThe restriction sites for SalI, Nco I, Not I, Sac I are underlined, bold, underline and italicized, and bold and italicized respectively.

PAGE 161

161 Table 43. Growth phenotype and genotype of ATP4syn strains generated. Designation ATP4 Gene Genotype Growth Characteristics a YPGD YPG AW3 ATP4 syn atp4::KanMx6, pAW36 ( URA3 + ATP4 ) +++ +++ AW4 ATP4 syn atp4::KanMx6, pAW33 ( HIS3+, ATP4 ) +++ +++ AW8 ATP4 syn atp4::KanMx6, pAW41 ( HIS3+, ) +++ +++ AW6 ATP4syn atp4::KanMx6, pAW35 ( HIS3+, ) AW9 ATP4 syn atp4::KanMx6, p CJB4 ( HIS3+, ) AW10 ATP4 syn atp4::KanMx6, pCJB2 ( HIS3+, ) AW11 ATP4 syn +4 atp4::KanMx6, pAW42 ( HIS3+, atp4+4 ) +++ +++ AW5 ATP4 syn +7 atp4::KanMx6, p AW34 ( HIS3+, atp4+7 ) +++ +++ AW12 ATP4syn+11 atp4::KanMx6, pAW44 ( HIS3+, atp4+11 ) +++ +++ AW14 ATP4 syn +14 atp4::KanMx6, pAW45 ( HIS3+, atp4+14 ) ++ ++ AW7 atp4::KanMx6, pRS313 ( HIS3 + ) aSymbols: +++ large, healthy colonies, ++ medium size colonies, very small colonies, no growth

PAGE 162

162 Table 44. S. cerevisiae F1FO ATP synthase activity b subunit ATPase Activity mole P i min 1 mg protein 1 % oligomycin sensitive ol igomycin + oligomycin ATP4 syn 5.30 0.89 2.30 0.61 57 % ATP4 syn b 3.03 1.07 2.98 0.76 2 % ATP4 syn 4 3.89 0.55 3.49 0.87 10 % ATP4 syn 7 3.27 0.78 2.12 0.40 35 % ATP4 syn 11 1.46 0.46 1.20 0.40 18 % ATP4 syn 14 2.27 0.80 2.08 0.83 8 % ATP4 syn +4 5.82 0.58 5.11 0.65 12 % ATP4 syn +7 5.72 0.45 3.03 0.34 47 % ATP4 syn +11 5.82 0.58 5.11 0.65 10 % ATP4 syn +14 4.56 0.64 4.25 0.73 7 %

PAGE 163

163 Figure 41. Scheme for constructing E. coliS. cerevisiae shuttle vector. Panel A, The construction of pAW26 from pAW20 and pRS313). Panel B, The green PCR product from the amplification of the 5 regulatory region is combined with pAW26 and generated pAW31 (5 region+ ATP4syn in pRS313). Panel C, The plasmid pRS313 and the amplified 3 region, shown in red, were ligated to yield pAW28 (3 region in pRS313). Panel D, The plasmids pAW28 and pAW31 are ligated together to generate, the ATP4syn expression vector, pAW33. Panel E, The ATP4syn expression vector pAW36 (Apr, URA3 ) was generated in a similar process to plasmid pAW33.

PAGE 164

164 Figure 42. Agarose gel electrophoresis of PCR products containing the 5 and 3 regulatory regions. Panel A, Lanes: (1) Low mass ladder, (2) 476 bp PCR product containing the 5 regulatory region. Panel B, Lanes: (1) Low mass ladder, (2) 807 bp PCR product containing the 3 regulatory region.

PAGE 165

165 Figure 43. Scheme for sequencing genomic and plasmid DNA from S. cerevisiae The sequencing was done in two steps. (Top) A small piece of DNA was amplified using two primers, with AW45 being specific for the ATP4 or ATP4syn gene. (Bottom) The sequencing reaction was performed on the amplified piece of DNA.

PAGE 166

166 Figure 44. Design of synthetic ATP4 gene. The ATP4syn gene was designed to facilitate mutagenic experiments. Panel A, Restriction map of the ATP4syn gene. The distance of 100 bp is marked by a bar. B, Complete sequence of ATP4syn gene. In yellow are nucleotides changed to insert enzyme restriction sites through silent mutations. Nucleotides changed to alter codon preference are colored in cyan. Underlined are restriction sites. The name for the enzyme for each restriction site are below the underlined sequence.

PAGE 167

167 Figure 45. Scheme for plasmid shuffling. Shown in the upper left is a simplified schematic of a S. cerevisiae cell. In the nucleus are two double helices r epresenting the two alleles of the ATP4 gene. Outside of the nucleus, in the cytoplasm, one mitochondrion is shown. Plasmids pAW36 and pAW33 are in red and blue, respectively.

PAGE 168

168 Figure 46. Agarose gel electrophoresis of S. cerevisiae strain AW2. Two pri mers were used to amplify the region surrounding the ATP4 gene. This generated the 2 kb PCR fragment containing the KanMx6 cassette and the 1 kb fragment containing the ATP4 gene. Lanes: (1) 1 kb ladder, (2) homozygous ATP4 (3) homozygous KanMx6, (4) hom ozygous ATP4 (5) heterozygous ATP4/ KanMx6, (6) heterozygous ATP4/ KanMx6, (7) homozygous ATP4

PAGE 169

169 Figure 47. YPD plate showing dissected tetrads. Tetrads were dissected on to thin, flat YPD plates. Arrow points to colony that gave rise to haploid strai n AW3 (ura-, his-, atp4:: KanMx6).

PAGE 170

170 Figure 48. Sequence and agarose gel electrophoresis of strain AW3. To determine if AW3 was indeed the correct strain, the DNA from the cell was both sequenced and studied by restriction enzyme digestion. Panel A, Top. Sequencing result from ATP4syn gene. This was determined by aligning the sequencing result with the nucleotide sequence of the ATP4syn gene. Bottom. Sequencing result of the KanMx6 cassette. Panel B, PCR products were digested with Aat II, which was only found in the ATP4syn gene. The fragments indicating presence of the atp4:: KanMx6 allele were: 3 kb, 1.8 kb, 600 bp. Lanes: 1, 1 kb Ladder; 2, Nondigest PCR Product; 3, Aat II digested PCR product; 4, Non digest PCR Product; 5, Aat II digested PCR product.

PAGE 171

171 Figure 49. Growth of wild type and ATP4syn strain on YPG. BY4743 (left) and AW3 (right) strains were each patched on YPG media and grown for 3 days at 30 C. Glycerol is a nonfermentable carbon source and, in order to grow, S. cerevisiae requires a function F1FO ATP synthase. Figure 410. Amino acid sequence alignment of B. taurus and S. cerevisiae b subunits. The protein sequences were obtained from the Pubmed protein database and were aligned using protein BLAST. The method used was the com positional adjustment method. The bit score is included because it can be used to compare different alignments. The expected value is the number of other sequences that would have the equivalent bit scores in the database. Identities are the number of exact matches, while positives is based on the both exact matches and similarity. Gap score is based on the gaps inserted into a sequence to compensate for insertions or deletions in one sequence relative to another.

PAGE 172

172 Figure 411. Doublestrand sequences of recombinant ATP4(b) genes. (A) Top, wild type synthetic sequence and deduced amino acid sequence corresponding to amino acids 7397. Below are novel deletion constructs with the deduced amino acid sequences. (B) Top, wild type synthetic sequence and deduced amino acid sequence corresponding to amino acids 8297. Below are novel insertion constructs with the deduced amino acid sequences. Mutations are noted on the left Restriction sites used for mutagenesis and screening purposes are underlined and labeled.

PAGE 173

173 Figure 412. Growth of ATP4synmutant strains on YPGD. Strains were serial diluted from 106 cells/mL to 100 cells/mL or 107 cells/mL to 103 cells/mL shown in Panel A and Panel B, respectively. Plates were placed in an incubator at 30 C and grown for 3 days. Five microliters of each dilution was spotted onto rich media with 3% glycerol and 0.1% glucose. Panel A, Deletions in b subunit. Panel B, Insertions in b subunit. Figure 413. Immunoblot analysis of atp4(b) gene mutant mitochondria. Mitochondria proteins were separated on a SDS gel and proteins were transferred to a PVDF membrane (See Materials and Methods). The presence of a b subunit was detected using anti serum against the b subunit in S. cerevisiae Panel A, Lanes syn, (x) empty, (2) ATP4syn, (3) ATP4syn ATP4synsynsyn Lanes syn, (x) empty, (2) ATP4syn, (3) ATP4syn+4, (4) ATP4syn+7, (5) ATP4syn+11, (6) ATP4syn+14. Signal strengths on these and other blots were calculated using shorter exposure times and UnScanIt gel digitizing software (Silk Scientific, Inc.)

PAGE 174

174 CHAPTER 5 CONCLUSIONS AND FUTURE DIRECTIONS Conclusions The work in this dissertation demonstrates that the inherent plasticity in the peripheral stalk of the E. coli F1FO ATP synthase is also found in the S. cerevisiae enzyme. In 2006, the first highresolution structure of a peripheral stalk complex from either a bacterial or a eukaryotic enzyme was reported (Dickson et al. 2006). This structure feature d a peripheral stalk that had only a slight bend, allowing it to wrap around F1. When the structure debuted, it was remarked that the idea of the peripheral stalk being flexible was incompatible with the structure. Part of the argument involved the abilit y of the peripheral stalk complex to form well ordered crystals. The idea being that if the peripheral stalk could be crystallized as a complex, it must be rigid and well ordered. The other feature of the model showed the peripheral stalk subcomplex docked with the rest of the enzyme such that only an upright, perpendicular arrangement could be adopted. Complicating the interpretation is that the eukaryotic peripheral stalk subcomplex did not readily crystallize. No peripheral stalk has ever been crystall ized within an F1FO ATP synthase. The bacterial stalk has never been successfully crystallized and a structure solved at high resolution. despite attempts by at least four groups over a period exceeding a decade. Moreover, the proposal of a rigid peripheral stalk was not compatible with previous work that showed the length of the bacterial b subunit, could be readily manipulated (Sorgen et al. 1998b, 1999; Grabar and Cain, 2003). Assuming that fixed points of contact were maintained in F1 and FO, there was no way to accommodate the types of insertions and deletions engineered in the bacterial enzyme

PAGE 175

175 into the initial B. taurus F1FO model. However, later models that docked the B. taurus peripheral stalk within the F1FO complex positioned the stalk so that it curved around the F1 sector of the enzyme (Lau et al., 2008; Rees et al. 2009). It appeared that in these models the enzyme could accommodate changes in length of the eukaryotic peripheral stalk. It must be noted that bacterial and eukaryotic enzymes are under different functional constraints. The E. coli F1FO ATP synthase, like many other bacterial F1FO ATP synthases, has to be able to act as both an ATP driven proton pump and as an ATP synthase. In contrast, the eukaryotic enzymes only work as ATP synthases. It is possible that the peripheral stalk evolved specifically to serve this function. Thus, the possibility existed that the bacterial and mitochondrial peripheral stalks might be fundamentally different. Support for this idea is that the eukaryoti c peripheral stalk is made up three very different proteins that interact with each other via both parallel and anti parallel coiled coils. In the prokaryotic and chloroplast enzymes, the peripheral stalk consists of either two identical or two very simil ar proteins that interact with each other through an extended parallel coiledcoil. Moreover, the archeabacterial A1AO peripheral stalk has identical subunit right handed coiledcoil structures (Lee et al. 2010). Therefore, it became very important to examine both peripheral stalks in the context of the entire enzyme to determine if the two stalks shared similar properties. The work in Chapter 2 extended the Cain laboratory investigation of the E. coli peripheral stalk. Specifically, studies in this chapter examined the positional effects of an important arginine at position 36 in the tether domain of the E. coli b subunits. Chapter 3 provided tools for current and future biophysical experiments in the E. coli

PAGE 176

176 peripheral stalk (Waechter et al submitted f or publication). It also provided a framework for generating biophysical tools for study of the eukaryotic enzyme as well. Chapter 4 bridged work done on the bacterial enzyme to the eukaryotic enzyme. It showed that the peripheral stalk in the S. cerevi siae mitochondrial F1FO ATP synthase had the same inherent plasticity observed in the bacterial stalk. Finally, a new model was proposed to address the current information on how the peripheral stalk functions in the context of the entire enzyme. Intragenic Suppressors of the Arginine 36 Mutations in the Bacterial Peripheral Stalk Research reported in Chapter 2 studied an arginine residue that is conserved in all prokaryotic b subunits. The amino acid, bR36 in E. coli was moved up one turn of the helix in the b subunit with minimal effect on F1FO ATP synthase activity. Previous work had shown that substituting this arginine with an isoleucine, glutamic acid, serine, or cysteine caused the enzyme to lose activity (Caviston et al. 1998). In the case of the glutamic acid, it was a true uncoupling mutation. Additionally, it was found that a bR36 was only required on one b subunit in the peripheral stalk (Grabar and Cain, 2004). The other b subunit could have either a bR36I or bR36E substitution and the enzyme remained active. Although this extragenic suppression of bR36I was successful, my work demonstrated that the bR36I substitution could not be overcome by an intragenic suppressor. The bR36I substitution apparently caused an irrevocable change in the F1FO complex. The most likely defect resulting from bR36I was in the b a subunit interactions. More interestingly, the defect caused by the bR36E substitution could be overcome by an intragenic suppressor. The bR36E, E39R subunit showed apparent wildt ype like growth on minimal media supplemented with succinate and ATP driven protonpumping

PAGE 177

177 activity. This is in contrast to the uncoupling phenotype resulting from the bR36E substitution. The bR36E substitution showed no growth on a nonfermentable carbon source and very little ATP driven proton pumping in isolated membrane vesicles (Table 2 1, Figure 25). These results showed that an arginine at or around b36 is important, but not essential, to coupling the catalytic activity of the F1 sector to the proton translocation of the FO sector. It is possible that this important arginine, bR36, interacts with the a subunit and, more specifically with the first cytoplasmic loop, L12, of the a subunit. Previous work in other laboratories has showed that an amino acid in the first cytoplasmic loop of the a subunit, aK74C, can be chemically crosslinked to the b subunit (Long et al. 2002). The reverse is also true. The bR36C subunit could be crosslinked to the L12 loop of the a subunit through use of a photoactivat ible crosslinking reagent (McLachlin et al. 2000) There has been some additional studies that show the L12 loop of the a subunit may be involved in proton translocation. More specifically the loop might be involved in the release of the proton into the c ytoplasm (Moore et al. 2008). This would account for experiments showing that proton translocation is not only dependent upon the a and c subunits, but the b subunit as well (Greie et al. 2004). Through interaction of the bR36 with the L12 loop, the b subunit might play an indirect structural role in proton translocation. Chemical Modification of Cysteines in the Bacterial Peripheral Stalk The work in Chapter 3 described the development of tools for biophysical experiments. It is notable that many amino acid substitutions can be made in the peripheral stalk subunits in the F1FO ATP synthase with little or no effect on activity. Some of the most important remaining questions focus on the conformation of the

PAGE 178

178 peripheral stalk in the resting versus the act ive state of the enzyme. Most of the ongoing biophysical experiments on the peripheral stalk rely on introducing individual cysteines into the peripheral stalk and subsequently modifying the thiol groups with fluorescent and spin labels. Novel cysteines w ere introduced into a cysteineless F1FO ATP synthase with out affecting enzyme activity, and some of these are likely to be employed in ongoing biophysical experiments (Figures 33 and 38). C64) was labeled with Alexa Fluor 488, a fluorescent maleimide (Figure 310). The Alexa Fluor 488C64 subunit did not inhibit the ATPase activity of the F1 sector. Moreover, the intact F1FO ATP synthase could be la C64 with out affecting coupled ATP driven proton pumping (Figure 3OSCP, it should be possible to modify the S. cerevisiae mitochondrial enzyme at a co mparable site. The reconstitution experiments involving the Alexa Fluor 488modified F1 sector with unlabeled FO in membrane vesicles were unsuccessful (Figure 311). The lack of reconstitution might be due to the Alexa Fluor 488 interfering with the nec essary b interactions. Another possibility was that the subunit disassociated from the F1 sector free F11 on to FO cannot occur. Clearly, the Alexa Fluor 488 modification did not destabilize the F1FO complex, as the label did not affect coupled enzyme activity. Ultimately, this line of work was terminated because the B. taurus partial peripheral stalk structure provided an opportunity to study the

PAGE 179

179 peripheral stalk in the context of the highresolution structure. This is still not possible for any bacterial or chloroplast peripheral stalk. Manipulation of the Eukaryotic Peripheral Stalk in the F1FO ATP Synthase Chapter 4 sought to investigate whether the properties of the eukaryotic and prokaryotic peripheral stalks were similar. Previous work in the E. coli F1FO ATP synthase showed that the enzyme could tolerate large deletions and insertions in both b subunits and up to a fourteen amino acid size difference between the two b subunits (Sorgen et al. 1998b, 1999; Grabar and Cain, 2003). The evidence favored an interpretation that the peripheral stalk of the E. coli enzyme had a high degree of plasticity. The highresolution structure of the B. taurus peripheral stalk subcomplex varied significantly from the bacterial peripheral stalk (Dickson et al. 2006). The proposed docking of the B. taurus peripheral stalk structure in the context of the F1FO ATP synthase suggested that the B. taurus peripheral stalk might intersect the membrane perpendicularly and was rigid within the F1FO complex. There was little apparent space for the changes in length acceptable in the prokaryotic stalk. In order to address the plasticity of the eukaryotic peripheral stalk, insertion and deletion mutations were constructed. The target area in the S. cerevisiae b subunit was the area where the subunit emerges from the membrane, which is comparable to the segment manipulated in the E. coli enzyme. This required significant construction of both the expression strain and expression vector. The S. cerevisiae F1FO ATP synthase tolerated a deletion of four amino acids and insertions of up to fourteen amino acids in the b subunit. Although the eukaryotic peripheral stalk showed a somewhat reduced degree of plasticity compared to the bacterial peripheral stalk, it displayed a much higher degree of plasticity than predicted by the rigid stalk model. The enzymes that

PAGE 180

180 contained a deletion of four amino acids or insertions of four, eleven, or fourteen amino acids in the b subunit exhibited a surprising lack of oligomycin sensitivity. Normally enzymes from strains that grow on a nonfermentable carbon source display oligomycin sensitivity. These results suggest that a subtle conformational shift occurred in the peripheral stalk that propagated to the FO sector to inhibit oligomycin binding. I propose that this conformational change could be accommodated by a new model of peripheral stalk function called the hinge stiff model. There are three proposed models for how the peripheral stalk functions: (i) the rigid rod model, (ii) the flexible rope model, and (iii) a new hingestiff model. It had been proposed that the peripheral stalk of the eukaryotic F1FO ATP sy nthase was a rigid rod and had no plasticity. This was based on the ability of the membrane extrinsic portion of the peripheral stalk to be crystallized and erroneous docking of the structure within a cryoelectron microscopy structure of the F1FO complex (Dickson et al. 2006). This is in contrast to the flexible rope model, where the peripheral stalk was thought to be entirely flexible in the tether domain, which had largely been based on mutagenesis of the bacterial peripheral stalk. Additional suppor t for the flexible rope model was that no bacterial peripheral stalk has ever been crystallized. Recently, another model has emerged that I am calling the hingestiff model where there are hinge regions in the peripheral stalk. The two hinge regions woul d allow for a relatively high degree of plasticity in the h/d subunit interaction domain during assembly. Once the enzyme was fully assembled, the peripheral stalk could assume a relatively stiff conformation. Any changes in length of the b subunit would be accommodated for by adjusting the angle at two different hinge regions located within the subunit (Figure 51).

PAGE 181

181 There was evidence for two hinge regions in the b subunit. One appeared to be located in the area where the b subunit emerges from the membrane and the other towards the carboxyl end of the d/h subunit interaction domain. The evidence for these hinge regions came from both the recent highresolution structure of the F1peripheral stalk complex in B. taurus as well as mutagenesis and electron spin resonance work done in E. coli (Rees et al. 2009; Dmitriev et al. 1999; Hardy et al. 2003; Steigmiller et al. 2005; Zaida et al. 2009). A NMR structure of b134 from the E. coli subunit revealed an helix from b422, a bend around b2326, and the helix resuming at bP27 (Dmitriev et al. 1999). The bend around b2326, located where the b subunit emerges from the membrane, could serve as a hinge region. Later mutagenesis work in E. coli in the b1824 region failed to detect any indication of im portant b b interactions and electron spin resonance experiments on bsol showed no evidence of b b interaction in the tether domain (Hardy et al. 2003; Steigmiller et al ., 2005; Zaida et al ., 2009). The negative data from these experiments suggested the possibility that the two b subunits flare apart near the cytoplasmic surface of the membrane. If that were the case, then the peripheral stalk would be expected to be highly flexible in that localized area. This further supported the location of hinge regi on in that segment. The new highresolution structure of the B. taurus F1peripheral stalk complex showed the possibility of a hinge around B. taurus b146 (Figure 51) (Rees et al. 2009). The two hinges, one located near the surface in the membrane and t he other in the h/d subunit interaction domain, might act as pivot points for the b subunit. Changing the angle of these hinges would alter the wrapping of the peripheral stalk around the F1 sector allowing incorporation of a shortened or a lengthened b su bunit into the F1FO complex (Figure 51). In the context

PAGE 182

182 of the entire F1FO complex, the peripheral stalk would be locked into its final conformation. Overall, the peripheral stalk might indeed be more rigid than the rotor stalk within the enzyme (Okuno et al ., 2010; Waechter et al. submitted for publication). However, it is possible that there are local segments of higher plasticity, especially around the hinge regions, within the peripheral stalk. Previous work has shown that the average curved helix has a radius of curvature of 60 or 14.3 (Barlow and Thornton, 1988). It should be noted that in general an helix that is curved by more than 20 is considered kinked. The proposed hinge regions would have a radius of curvature less than 60 in order to accommodate the changes in length shown here. Protein structural plasticity with regards to insertions and deletions within the T4 lysozyme has been investigated using crystallography (Vetter et al. 1996). Within various regions of the T4 lysozyme, 32 insertions and five deletions were constructed in the nine helices of the protein. These recombinant proteins were express, purified, and where possible studied by x ray crystallography. The changes in structure that occurred between the mutant and wil d type protein structures were observed. In the cases where the insertion was flanked by fixed points of contact with other domains of the protein, the inserted amino acids were looped out. This looping out was offset by as much as 25. This is simila r to what was proposed by the hingestiff model, where the insertions or deletions was flanked by fixed points of contact with other subunits in the F1FO ATP synthase. The angles of the shortened or lengthened b subunits between the two hinge regions was calculated by using the available highresolution structures for the B. taurus F1peripheral stalk complex and the S. cerevisiae F1c10 complex (Rees et al. 2009;

PAGE 183

183 Dautant et al. 2010). The distances from the hinge at B. taurus b146 to the adjacent subunit and from the subunit to the top of the FO sector were calculated in MacPyMol (Figure 52). The length of the shortened or lengthened b subunit was calculated using MacPyMol to measure the length of the repeated or deleted region (Figure 53A). The resulting angles were graphed and overlaid onto the hingestiff model (Figure 53B). This showed that only slight adjustments to the angle of the peripheral stalk between the two proposed hinge regions were necessary to accommodate changes in the b subunit length. The hingestiff model is the only current model that accounts for the measured physical properties of the peripheral stalk and large changes in length of the b subunit. Future Directions The work in this dissertation focuses on the nature of the peripheral stalk in both the bacterial and mitochondrial F1FO ATP synthase. However, the vast majority of what is known about both peripheral stalks is the interactions of the peripheral stalk with the F1 sector. This is because of extensive genetic, biochemical, and structural work on the F1 sector. For example, a search of the protein database (www.pdb.org) yields 48 structures involving the subunits of the F1 sector. Walker and coworkers contributed t he majority of those 48 structures that are from the mammalian F1 sector. Only two of those structures contain subunits from a bacterial F1 sector. Out of the 48 structures, only one shows the peripheral stalk with the F1 sector (Rees et al. 2009). The structures provide additional insight when designing experiments to probe the probable interactions between the peripheral stalk and the F1 sector. To date there is no highresolution structure of an intact FO and only a model polypeptide NMR structure of the

PAGE 184

184 transmembrane domain of the E. coli b subunit (Dmitriev et al. 1999). Progress on structures of FO is limited to the ring of c subunits and not necessarily relevant to the peripheral stalk (Meier et al ., 2005; Pogoryelov et al ., 2009) Therefore, much of what has been derived about the peripheral stalk contacts in FO has been through using mutagenic analysis. In order to completely probe these interactions, a structure of either just the FO sector or, more favorably, the entire F1FO complex must be solved. However, it must be stressed that any s tructure for the overall complex must be validated using other techniques to show that the structure is relevant in the active state of the enzyme. Fluorescent Techniques in the Eukaryotic F1FO ATP Synthase While the work in this dissertation has revealed that the eukaryotic peripheral stalk shares a degree of plasticity with the bacterial stalk, it has also shown that overall this plasticity is significantly reduced in the eukaryotic enzyme. However, local areas of high plasticity within the eukaryotic peripheral stalk may exist. The highresolution structure of the B. taurus F1peripheral stalk complex can be used as a guide for selecting useful sites to investigate the existence of regions of higher plasticity in the eukaryotic peripheral stalk. Hopefull y, enzyme preparation and chemical modification techniques described in Chapter 3 can be adapted to accommodate the study of the S. cerevisiae mitochondrial F1FO ATP synthase. The thiol groups on cysteine residues could be used for labeling in the S. ce revisiae enzyme like in the bacterial enzyme. The entire S. cerevisiae enzyme has eight cysteine codons. It appears that the cysteines that are present in the enzyme complex are either in integral membrane proteins and have a low probability of being modi fied, in the cases of the g, a, and e subunits, or buried at subunit subunit

PAGE 185

185 interfaces, in the subunits (Dautant et al. 2010). In terms of the S. cerevisiae peripheral stalk subunits, the only cysteine is in the OSCP. The exact location of t his cysteine residue in the OSCP is unknown, but the analogous residue in the B. taurus OSCPN97, is buried at the F1OSCP interface (Rees et al. 2009). Therefore, given the homology between the two enzymes, the S. cerevisiae cysteine, OSCPC100 may be loc ated within the F1OSCP interface as well. One might expect low chemical reactivity at this site within the OSCP, and this could be easily tested by attempting to label the intact enzyme with an Alexa Fluor maleimide. Conditions for the determination of t he efficiency of modification of the OSCP could use the technology E. coli enzyme (Chapter 3). Alternatively, a cysteineless OSCP gene ( ATP5 ) could be constructed and used in a gene replacement experiment (Chapter 4). I ntroduced cysteines could then be labeled and used for a variety of experiments, such as chemical modification, crosslinking, fluorescence anisotropy, FRET, or spin resonance experiments. Fluorescence anisotropy experiments targeting the peripheral stal k subunits should be run in both the resting and active state of the F1FO ATP synthase (i.e. in the absence and presence of ATP). Fluorescence anisotropy examines the local environment of the fluorescent probe. That is, the more constraints the environment places on movement of the fluorescent probe, the higher the anisotropy. If the peripheral stalk structure became more ordered under the active state of the enzyme, then fluorescence polarization would increase. The two states could then be compared to determine the changes, if any, in the local environment of the b subunit of the S. cerevisiae peripheral stalk associated with catalytic activity.

PAGE 186

186 FRET experiments present a different challenge in that one would have to selectively label the b subunit and the OSCP. In some ways this might be more easily accomplished in the S. cerevisiae F1FO complex than in the E. coli enzyme. The b subunits in the peripheral stalk of eukaryotes are present in only single copies in the enzyme. In submitochondrial particles from S. cerevisiae the F1 sector can be removed from the FO sector, each sector could be labeled with a different fluorescent maleimide on introduced cysteines. Again labeling conditions optimized in Chapter 3 might be directly applicable. Then the F1 a nd FO sectors could be reconstituted (Collinson et al. 1994; Ackerman and Tzagoloff, 2007). It is this difficulty that was not overcome in experiments on the E. coli F1FO complex in Chapter 3. However, there is no reason to believe that given several di fferent labeled cysteines in the S. cerevisiae F1 that this cannot be accomplished. Gavin et al showed that there were no changes in FRET between the OSCP and b subunit during the resting or active state of the enzyme (Gavin et al. 2003). However, the authors used GFP and BFP as fluorescent probes. These fluorescent proteins are very large when compared to the Alexa Fluor maleimides (approximately 27 kDa for GFP and 0.7 1.0 kDa for the Alexa Fluor series). It was unlikely that GFP based fluorescence woul d have the sensitivity to detect stretching and contracting in the peripheral stalk associated with function of the enzyme. If one were to adopt the Alexa Fluor labeled cysteine approach, then construction of OSCPK174C and bK121C in the S. cerevisiae enzym e would be excellent choices (Figure 54). These positions are based on the structure of the F1peripheral stalk and are a distance of 74 in the B. taurus structure. This should yield approximately the same distance in the S. cerevisiae enzyme and allow for the usage of the Alexa Fluor 546

PAGE 187

187 Alexa Fluor 647 pair, which has an RO value of 74 This pair would give a much more sensitive view of the relative movements of the OSCP and b subunit in the active and resting state of the enzyme. Further Validat ion of the Partial Peripheral Stalk Structure The B. taurus partial peripheral stalk structure is a major advancement in the study of the peripheral stalk of the eukaryotic F1FO ATP synthase. However, it is unknown whether this structure is in fact physiologically relevant or valid in the context of the entire enzyme. Since the structure was first solved in 2006, there have been many changes proposed with respect to its location in the structure in the enzyme (Dickson et al. 2006; Lau et al. 2008; Rees et al. 2009). Originally, the structure was shown to intersect the membrane perpendicularly (Dickson et al. 2006). Later publications curved the peripheral stalk around the F1 sector (Lau et al. 2008). The most recent work showed the possibility of a hinge in the B. taurus peripheral stalk at b146 (Rees et al. 2009). Therefore, it is increasingly necessary to validate the structure by functional analysis in an intact F1FO ATP synthase. One possible way to do this is through mutagenic analysis directed by available structures. Currently, I am directing work in the Cain laboratory to evaluate the interactions between the b and d subunits. For example, there are four amino acids in the b subunit, S. cerevisiae bV123L126, that are completely conserved between the B. taurus S. cerevisiae, and other mitochondrial enzymes, including H. sapien (Figure 55). These four amino acids also appear to be in contact with the d subunit in the B. taurus partial peripheral stalk structure (Dickson et al. 2006). Cur rently three of those amino acids, S. cerevisiae bF124 V126, are being substituted with alanines to determine if those amino acids are involved in an essential contact with the S. cerevisiae d subunit. If these

PAGE 188

188 amino acids are in contact, then substituting them with alanines would likely interfere with subunit subunit interaction. This will be manifest in F1FO ATP synthase deficiency probably in the form of an assembly defect. Single amino acid substitutions in each position are also currently under constr uction to determine the effects of replacing each individual amino acids (Table 51). It is important to note that a lack of effect for any of these substitutions would only show that those amino acids are not essential and could not entirely rule out subunit subunit interaction taking place in that region. Additional studies would be necessary such as disulfide bridge formation between cysteines introduced at positions in close proximity to look at the issue. It would also be interesting to investigate t he potential hinge region at amino acid b146 in the B. taurus (Rees et al. 2009). The comparable position in S. cerevisiae is b145. It is of interest to note that the amino acid at position b146 in the B. taurus is a positively charged arginine, and at t he comparable position b145 in the S. cerevisiae is a negatively charged glutamic acid. However, this position is close to a more conserved region at bL146 in S. cerevisiae ( bV147 in B. taurus ) and a patch of similar residues ( S. cerevisiae b148149, b151, b154155). It is possible that this region provides the flexibility needed during assembly for the hingestiff model. A glycine scan mutagenesis to soften or an alanine scan to induce a more rigid helical conformation could be useful for determining t he effect of the bend. The same concepts could be used to investigate the hb subunit and OSCP b interactions within the complex. For example, there are three conserved amino acids in the b subunit that in the B. taurus F1peripheral stalk structure appear to interact with the OSCP. These three amino acids are bV177, bR179, and bV180 in the S. cerevisiae

PAGE 189

189 They appear to be interacting with an helical segment of the OSCP. In the segment OSCPK169 I189 in B. taurus or OSCPK174S194 in S. cerevisiae enzym e. In that helical segment there are six amino acids between the B. taurus and S. cerevisiae OSCP: S. cerevisiae OSCPT182,K183,I184,Q185,K186,L187. One could investigate the interactions of these subunits by performing an alanine scan by a series of specific single amino acid substitutions and introduction of cysteines to facilitate disulfide bridge crosslinking. This would require the generation of a S. cerevisiae strain containing a disruption in both the OSCP and b subunit genes. Although technically challenging, the same plasmid shuffling technique as outlined in Chapter 4 could be used to perform these experiments. Investigating the hb subunit interactions using this approach would be even more difficult because of the low homology between the h and F6 subunits. There is the possibility of a different hinge region located within the OSCP (Rees et al. 2009). The B. taurus F1peripheral stalk structure shows a flexible linker region between the amino and carboxyl domains of the OSCP, approximately OSCPL131Q146 in B. taurus that corresponds to OSCPF134 K151 in S. cerevisiae Performing mutagenesis and crosslinking experiments in this region to fix the domains in the OSCP in a specific conformation would provide a loss of activity if the flexibility of this region were required for enzyme function. Some of the above experiments require showing crosslink formation between the peripheral stalk subunits. Currently the lab has anti sera to the S. cerevisiae b subunit. However, antibodies to the S. cer evisiae OSCP, d, or h subunits are not commercially available. One option would be to generate the antibodies for the S. cerevisiae subunits. Another possibly easier option would be to construct epitope tags on the

PAGE 190

190 subunits. It was previously shown that a 25 kDa protein (BFP) on the carboxyl terminus of the OSCP did not unduly affect enzyme activity. Therefore, it is reasonable to think that a smaller histidine, c myc or V5 epitope tag would have little effect. The expression of an epitopetagged OSCP could be done by either plasmid shuffling or direct integration of the fusion protein into the chromosome. Summary In summary, the major steps towards understanding the peripheral stalk function in the F1FO ATP synthase would be to crystallize the entire peripheral stalk by itself or in the context of the entire complex, directly measure the plasticity of the peripheral stalk, and to determine a unified view of peripheral stalk assembly and function. Currently the highresolution structures available are only of partial peripheral stalk complexes (Dickson et al. 2006; Rees et al. 2009). These complexes most notably lack transmembrane domain I, intermembrane space loop domain, and transmembrane domain II of the b subunit and conserved portions of the d s ubunit. These regions of the peripheral stalk are important to understanding the function of the stalk. Completing the peripheral stalk structure would allow further examination of subunit interactions within the stalk, especially the interactions of the highly conserved d subunit residues missing from the current structures. A complete structure of the entire F1FO complex would allow for more targeted mutagenic analysis studies on the interactions of the peripheral stalk with other subunits in the enzyme. An entire enzyme structure might illuminate how the b subunit affects proton translocation and even how the peripheral stalk participates in the dimerization and oligomerization of ATP synthases in mitochondria. Additionally, the plasticity of the peripheral stalk needs to be measured directly. Sitedirected mutagenic analysis of the peripheral stalk is very helpful, as it demonstrates the importance and

PAGE 191

191 possible function of certain amino acids. The plasticity of the E. coli peripheral stalk is currently being examined by the laboratory of Dr. Wolfgang Junge (Waechter et al submitted for publication) using force applied on a magnetic bead attached to the E. coli b subunit in an intact complex. However, this method does not examine the local changes in plasticity that is implied in the hingestiff model. Possible techniques for measuring these local changes like fluorescence anisotropy or FRET have been proposed in a previous section in this chapter (see Fluorescent Techniques in the Eukaryotic F1FO AT P Synthase). Direct measurement of localized changes in plasticity of the peripheral stalk is necessary in order to prove the hingestiff model. Currently, the hingestiff model is the only model that accommodates for all the data on the peripheral stalk of both the prokaryotic and eukaryotic enzymes. This model presents a much needed unified view of both assembly and function of the peripheral stalk in the context of the entire F1FO ATP synthase.

PAGE 192

192 Table 51. Primer sequences and substitutions for bL123V126 Primer Name Primer Sequence a Substitution or Note AW100 TGGCGGCCGCTTCCAAGGAAACTGT Alanine Scan AW101 CGACAGTTTCCTTGGAAGCGGCCGCCAGTAC Antisense to AW100 AW102 SCTTTGATGTTTCCAAGGAAACTGT b L123P b L123R AW103 CGACAGTTTCCTTGGAAACATCAAAGSGTAC Antisen se to AW102 AW104 TGYRTGATGTTTCCAAGGAAACTGT b F124H, b F124R, b F124Y, b F124C AW105 CGACAGTTTCCTTGGAAACATCAYRCAGTAC Antisense to AW104 AW106 TGTTTSRAGTTTCCAAGGAAACTGT b D125Q, b D125R, b D125E, b D125G AW107 CGACAGTTTCCTTGGAAACTYSAAACAGTAC Antisense to AW106 AW10 8 TGTTTGATSRTTCCAAGGAAACTGT b V127H, b V127R, b V127D, b V127G AW109 CGACAGTTTCCTTGGAAYSATCAAACAGTAC Antisense to AW108 aPrimer sequences are written in the 5 to 3 direction

PAGE 193

193 Figure 51. HingeStiff model. Shown is the B. taurus F1 sector with the subunits colored in light blue, deep purple, cyan, pink, and sand, respectively. The OSCP is shown in yellow. The peripheral stalk subunits b d, and F6 are shown in blue, pea green, and orange, respectively. The two presumed hinge points are ci rcled, with B. taurus b146 colored in yellow. The region between to two hinge points is bracketed.

PAGE 194

194 Figure 52. Amino acids and distances used to determine angles of manipulated b subunits. For both structures the coloring is as follows: The subunits are colored light blue; are in deep purple; OSCP is in yellow; is in teal; is in dirty violet; is in sand. The peripheral stalk is colored orange, split pea, and blue for F6, d, and b respectively. The ring of c subunits are shown in green. Panel A, The F1 sector and partial peripheral stalk structure of the B. taurus enzyme rendered using coordinates from the PDB, 2WSS. Shown in red on the right, within the oval, is amino acid 146 in the b subunit, which was chosen as it is in the proposed hinge region {Rees 2009}. On the left side of the oval, shown in red, as well, is amino acid 131 of the subunit. Panel B, The F1c10 structure from S. cerevisiae was developed using pymol and PDB structure 2WPD{Dautant 2010}. In the subunit colored in red, within the oval, is residue 130, which is analogous to residue 131 in the B. taurus enzyme. Also within the oval and colored in red on the ring of c subunits is residue 42. The distance between the two residues was measured using Pymol and the distance, 89.8 is shown.

PAGE 195

195 Figure 53. Model of the shortened and lengthened peripheral stalk complex in B. taurus Panel A, The model is based on the 2CLY protein database coordinates. Shown are the b d and F6 subunits in blue, pea green, and orange, respec tively. The model on the far left shows the wild type complex. The vertical bars indicate the areas of the insertions and deletion, with the the deletion of four amino acids; the region of the deletion is indicated by the horizontal line. The yellow of the +7 and +11 model indicates the duplicated and inserted region. Panel B, Shows the angles required in order for the functional insertion and deletion b subunits to reach fro m the top of FO (residue 79) to the proposed hinge region, residue 146. The figure was rendered in MacPyMol from 2CLY and 2WSS. The composite image was made in Adobe Photoshop.

PAGE 196

196 Figure 54. Proposed FRET residues. Shown is the B. taurus F1 sector with t he subunits colored in light blue, deep purple, cyan, pink, and sand, respectively. The OSCP is shown in yellow. The peripheral stalk subunits b d, and F6 are shown in blue, pea green, and orange, respectively. Colored in red are the res idues proposed for mutagenesis and labeling in FRET experiments.

PAGE 197

197 Figure 55. Close view of b d subunit interaction. The B. taurus b and d subunits are shown in blue and pea green, respectively. The conserved B. taurus residues b124127 (analogous to b123126 in S. cerevisiae) are colored yellow. Note the close interactions of the conserved b subunit residues to the d subunit.

PAGE 198

198 APPENDIX YEAST MEDIA AND IDIOSYNCRASIES Media All media for plates are made in 2 L flasks with 500 mL of liquid media. To sterilize the media, cover the top of the flask with foil and autoclave the flask in a nalgene bin for 20 minutes. For liquid media, omit the agar. All media protocols can be scaled down or up as necessary. It is preferable to use Difco brand agar, yeast extract, and peptone. Rich S. cerevisiae medium (YPD): Yeast Extract 5 g Bacto Peptone 10 g Dextrose 10 g Agar 8 g ddH2O 500 mL For rich medium (YP) substitute 15 mL glycerol for the Dextrose and autoclave as above. For YPD+G418 medium add 2.5 mL 50 mg/mL active G418 (filter sterilized) to autoclaved YPD medium after the medium has cooled to 65 C Synthetic Dropout Medium (CSM): Yeast Nitrogen Base 3.35 g Amino Acid Drop Out Mix 0.385 g Dextrose 5 g Agar 8 g ddH2O 500 mL Sporulation Medium: P otassium Acetate 10 g Yeast Extract 1 g Dextrose 0.25 g Amino Acid Drop Out Mix 0.1 g Agar 8 g ddH 2 O 500 mL Strains sporulated after 35 days.

PAGE 199

199 5 Fluoroorotic Acid Medium: This medium required preparation in two parts. Part One: Yeast Nitrogen Base 6.7 g Drop Out Mix without Uracil 0.385 g Dextrose 20 g Agar 16 g ddH 2 O 500mL Autoclave the mixture for 20 minutes and cool until the mixture reaches approximately 65 C. Part Two: 5 Fluoroorotic Acid Powder 1 g Uracil (2.4 mg/mL) 5 mL ddH 2 O 500 mL Heat the mixture on low with stirring for about an hour until the 5FOA is fully dissolved. Combine the two mixtures together and pour plates promptly. Yeast Idiosyncrasies When spreading plates after transformations, use the minimal number of strokes to spread the cells. If the cells are spread until the plate is dry, fewer transformants grow. G418 does not select well when used with Complete Synthetic Medium (CSM). For selection purposes, use only in conjunction with YPD medium Plates for tetrad dissection should be poured no thicker than half the height of the Petri dish and care should be taken so the medium is level. Velvets for replica plating should be wrapped in foil, upside down, and autoclaved. If no clean velvets are available, filter paper can be used. However, filter paper should not be used for replica plates that will be photographed. Cells replica plated on YPG plates can take 710 days to grow. To generate permanents: Inoculate 10 mL of appropriate CSM and allow to grow overnight. In the morning measure the OD600 of the culture and dilute to an OD600 of 0.15. Grow the culture for 3 to 5 hours. Then, combine 750 L of the culture with 300 L of 50% Filter sterilized Glycerol. Store at 80 C. The website www.yeastgenome.org is a valuable resource for yeast gene information, homology, and pathways.

PAGE 200

200 LIST OF REFERENCES Abrahams JP, Leslie AG, Lutter R, Walker JE (1994) Structure at 2.8 A resolution of F1ATPase from bovine heart mitochondria. Nature, 370: 621 628 Ackerman S, Tzagoloff A (2007) Methods to Determine the Status of Mitochondrial ATP Synthase Assembly. Methods Mol Bio, 372: 363 378 Adachi K, Oiwa K, Nishizaka T, Furuike S, Noji H, Itoh H, Yoshida M, Kinosita K (2007) Coupl ing of rotation and catalysis in F(1) ATPase revealed by singlemolecule imaging and manipulation. Cell, 130: 309 321 Aggeler R, Haughton MA, Capaldi RA (1995) Disulfide bond formation between the COOH terminal domain of the beta subunits and the gamma and epsilon subunits of the Escherichia coli F1ATPase. Structural implications and functional consequences. J Biol Chem 270: 9185 9191 Altschul SF, Madden TL, Schffer AA, Zhang J, Zhang Z, Miller W, Lipman DJ (1997) Gapped BLAST and PSI BLAST: a new generation of protein database search programs. Nucleic Acids Res 25: 3389 3402 Altschul SF, Wootton JC, Gertz EM, Agarwala R, Morgulis A, Schffer AA, Yu YK (2005) Protein database searches using compositionally adjusted substitution matrices. FEBS J 272: 5101 5109 Aris JP, Klionsky DJ, Simoni RD (1985) The Fo subunits of the Escherichia coli F1FoATP synthase are sufficient to form a functional proton pore. J Biol Chem 260: 11207 11215 Arnold I, Pfeiffer K, Neupert W, Stuart RA, Schgger H (1998) Yeast mitoc hondrial F1F0ATP synthase exists as a dimer: identification of three dimer specific subunits. EMBO J 17: 7170 7178 Arselin G, Vaillier J, Graves PV, Velours J (1996) ATP synthase of yeast mitochondria. Isolation of the subunit h and disruption of the ATP 14 gene. J Biol Chem 271: 20284 20290 Arselin G, Vaillier J, Salin B, Schaeffer J, Giraud M F, Dautant A, Brthes D, Velours J (2004) The modulation in subunits e and g amounts of yeast ATP synthase modifies mitochondrial cristae morphology. J Biol Chem 279: 40392 40399 Barlow DJ, Thornton JM (1988) Helix geometry in proteins. J Mol Biol, 201: 601 619 Baylis Scanlon JA, AlShawi MK, Nakamoto RK (2008) A rotor stator cross link in the F1 ATPase blocks the rate limiting step of rotational catalysis. J Biol Chem 283: 26228 26240

PAGE 201

201 Beinert H Sands RH (1960) Studies on Succinic and DPNH Dehydrogenase Preparations by Paramagnetic Resonance (EPR) Spectroscopy. Biochem Biophys Res Commun 3: 41 46 Beinert H, Palmer G (1964) OxidationReduction of the Copper Component of Cytochrom Oxidase. Kinetic Studies with a Rapid Freezing Technique. J Biol Chem 239: 1221 1227 Bhatt D, Cole S, Grabar TB, Claggett SB, Cain BD (2005) Manipulating the length of the b subunit F1 binding domain in F1F0 ATP synthase from Escherichia coli. J Bioenerg Biomembr 37: 67 74 Bi Y, Watts JC, Bamford PK, Briere LA, Dunn SD (2008) Probing the functional tolerance of the b subunit of Escherichia coli ATP synthase for sequence manipulation through a chimera approach. Biochim Biophys Acta, 1777: 583 591 Bowler MW, Montgomery MG, Leslie AG, Walker JE (2007) Ground state structure of F1ATPase from bovine heart mitochondria at 1.9 A resolution. J Biol Chem 282: 14238 14242 Boyer PD (1993) The binding change mechanism for ATP synthase--some probabil ities and possibilities. Biochim Biophys Acta 1140: 215 250 Boyer PD (2002) A research journey with ATP synthase. J Biol Chem 277: 39045 39061 Boyer PD, Cross RL, Momsen W (1973) A new concept for energy coupling in oxidative phosphorylation based on a m olecular explanation of the oxygen exchange reactions. Proc Natl Acad Sci USA, 70: 2837 2839 Bueler SA, Rubinstein JL (2008) Location of subunit d in the peripheral stalk of the ATP synthase from Saccharomyces cerevisiae. Biochemistry 47: 11804 11810 Cain BD (2000) Mutagenic analysis of the F0 stator subunits. J Bioenerg Biomembr 32: 365 371 Cain BD, Simoni R (1989) Proton translocation by the F1F0ATPase of Escherichia coli. Mutagenic analysis of the a subunit. J Biol Chem 264: 3292 3300 Carbajo RJ, Silv ester JA, Runswick MJ, Walker JE, Neuhaus D (2004) Solution structure of subunit F(6) from the peripheral stalk region of ATP synthase from bovine heart mitochondria. J Mol Biol, 342: 593 603 Carbajo RJ, Kellas FA, Runswick MJ, Montgomery MG, Walker JE, Neuhaus D (2005) Structure of the F1binding domain of the stator of bovine F1FoATPase and how it binds an alphasubunit. J Mol Biol, 351: 824 838

PAGE 202

202 Carbajo RJ, Kellas FA, Yang J C, Runswick MJ, Montgomery MG, Walker JE, Neuhaus D (2007) How the N terminal domain of the OSCP subunit of bovine F1FoATP synthase interacts with the N terminal region of an alpha subunit. J Mol Biol, 368: 310 318 Catterall WA, Coty WA, Pedersen PL (1973) Adenosine triphosphatase from rat liver mitochondria. 3. Subunit composition. J Biol Chem 248: 7427 7431 Caviston T, Ketchum C, Sorgen P, Nakamoto R, Cain BD (1998) Identification of an uncoupling mutation affecting the b subunit of F1F0 ATP synthase in Escherichia coli. FEBS Lett 429: 201 206 Claggett SB (2008) Construction and D isulfide Crosslinking of Chimeric b Subunits in the Peripheral Stalk of F1FO ATP Synthase from Escherichia coli. Biochemistry and Molecular Biology Ph.D.: 219 Claggett SB, Grabar TB, Dunn SD, Cain BD (2007) Functional incorporation of chimeric b subunits into F1Fo ATP synthase. J Bacteriol 189: 5463 5471 Claggett SB, Plancher MO, Dunn SD, Cain BD (2009) The b subunits in the peripheral stalk of F1Fo ATP synthase preferentially adopt an offset relationship. J Biol Chem 284: 16531 16540 Collinson IR, van R aaij MJ, Runswick MJ, Fearnley IM, Skehel JM, Orriss GL, Miroux B, Walker JE (1994) ATP synthase from bovine heart mitochondria. In vitro assembly of a stalk complex in the presence of F1ATPase and in its absence. J Mol Biol, 242: 408 421 Collinson IR, Sk ehel JM, Fearnley IM, Runswick MJ, Walker JE (1996) The F1F0ATPase complex from bovine heart mitochondria: the molar ratio of the subunits in the stalk region linking the F1 and F0 domains. Biochemistry 35: 12640 12646 Consortium U (2010) The Universal P rotein Resource (UniProt) in 2010. Nucleic Acids Res 38: D142 148 Crane FL, Hatefi Y, Lester RL, Widmer C (1957) Isolation of a quinone from beef heart mitochondria. Biochim Biophys Acta, 25: 220 221 Cross RL, Boyer PD (1975) The rapid labeling of adenosi ne triphosphate by 32P labeled inorganic phosphate and the exchange of phosphate oxygens as related to conformational coupling in oxidative phosphorylation. Biochemistry 14: 392 398 Cross RL, Grubmeyer C, Penefsky HS (1982) Mechanism of ATP hydrolysis by beef heart mitochondrial ATPase. Rate enhancements resulting from cooperative interactions between multiple catalytic sites. J Biol Chem 257: 12101 12105 Dautant A, Velours J, Giraud M F (2010) Crystal structure of the Mg.ADP inhibited state of the yeast F1c10 ATP synthase. J Biol Chem 285: 29502 29510

PAGE 203

203 Del Rizzo P, Bi Y, Dunn SD, Shilton B (2002) The "second stalk" of Escherichia coli ATP synthase: structure of the isolated dimerization domain. Biochemistry 41: 6875 6884 Del Rizzo PA, Bi Y, Dunn SD (2006 ) ATP synthase b subunit dimerization domain: a right handed coiled coil with offset helices. J Mol Biol, 364: 735 746 Dickson V, Silvester J, Fearnley I, Leslie A, Walker JE (2006) On the structure of the stator of the mitochondrial ATP synthase. Embo J 25: 2911 2918 Dmitriev OY, Freedman KH, Hermolin J, Fillingame RH (2007) Interaction of transmembrane helices in ATP synthase subunit a in solution as revealed by spin label difference NMR. Biochim Biophys Acta, 1777: 227 237 Dmitriev OY, Jones P, Jiang W, Fillingame RH (1999) Structure of the membrane domain of subunit b of the Escherichia coli F0F1 ATP synthase. J Biol Chem 274: 15598 15604 Downie JA, Gibson F, Cox GB (1979) Membrane adenosine triphosphatases of prokaryotic cells. Annu Rev Biochem 48: 1 03 131 Duncan TM, Bulygin VV, Zhou Y, Hutcheon ML, Cross RL (1995) Rotation of subunits during catalysis by Escherichia coli F1ATPase. Proc Natl Acad Sci USA, 92: 10964 10968 Dunn SD, Chandler J (1998) Characterization of a b2delta complex from Escherichi a coli ATP synthase. J Biol Chem 273: 8646 8651 Dunn SD (1992) The polar domain of the b subunit of Escherichia coli F1F0 ATPase forms an elongated dimer that interacts with the F1 sector. J Biol Chem 267: 7630 7636 Dunn SD, McLachlin D, Revington M (2000a) The second stalk of Escherichia coli ATP synthase. Biochim Biophys Acta, 1458: 356 363 Dunn SD, Revington M, Cipriano DJ, Shilton B (2000b) The b subunit of Escherichia coli ATP synthase. J Bioenerg Biomembr 32: 347 355 Engelhardt WA (1982) Life and S cience. Annu Rev Biochem 51: 1 19 Feniouk BA, Kozlova MA, Knorre DA, Cherepanov DA, Mulkidjanian AY, Junge W (2004) The of ATP synthase: ohmic conductance (10 fS), and absence of voltage gating. Biophys J 86: 4094 4109 Foster DL, Mosher ME, Futai M, Fill ingame RH (1980) Subunits of the H+ ATPase of Escherichia coli. Overproduction of an eight subunit F1F0ATPase following induction of a lambdatransducing phage carrying the unc operon. J Biol Chem 255: 12037 12041

PAGE 204

204 Friedkin M, Lehninger AL (1948) Phosphor ylation coupled to electron transport between dihydrodiphosphopyridine nucleotide and oxygen. J Biol Chem 174: 757 Fronzes R, Chaignepain S, Bathany K, Giraud MF, Arselin G, Schmitter JM, Dautant A, Velours J, Brthes D (2003) Topological and functional s tudy of subunit h of the F1Fo ATP synthase complex in yeast Saccharomyces cerevisiae. Biochemistry 42: 12038 12049 Fronzes R, Weimann T, Vaillier J, Velours J, Brthes D (2006) The peripheral stalk participates in the yeast ATP synthase dimerization independently of e and g subunits. Biochemistry 45: 6715 6723 Gardner JL, Cain BD (1999) Amino acid substitutions in the a subunit affect the epsilon subunit of F1F0 ATP synthase from Escherichia coli. Arch Biochem Biophys 361: 302 308 Gavin PD, Devenish RJ, Prescott M (2003) FRET reveals changes in the F1stator stalk interaction during activity of F1F0ATP synthase. Biochim Biophys Acta 1607: 167 179 Gibbons C, Montgomery MG, Leslie AG, Walker JE (2000) The structure of the central stalk in bovine F(1) ATPa se at 2.4 A resolution. Nat Struct Biol, 7: 1055 1061 Gietz RD, Woods RA (2002) Transformation of yeast by lithium acetate/singlestranded carrier DNA/polyethylene glycol method. Meth Enzymol 350: 87 96 Gogol EP, Lcken U, Capaldi RA (1987) The stalk connecting the F1 and F0 domains of ATP synthase visualized by electron microscopy of unstained specimens. FEBS Lett 219: 274 278 Gogol EP, Aggeler R, Sagermann M, Capaldi RA (1989) Cryoelectron microscopy of Escherichia coli F1 adenosinetriphosphatase decorated with monoclonal antibodies to individual subunits of the complex. Biochemistry 28: 4717 4724 Goyon V, Fronzes R, Salin B, di Rago JP, Velours J, Brethes D (2008) Yeast cells depleted in Atp14p fail to assemble Atp6p within the ATP synthase and exhibit altered mitochondrial cristae morphology. J Biol Chem 283: 9749 9758 Grabar TB (2004) Contributions of the Individual b Subunits to the Function of the Peripheral Stalk of F1FO ATP Synthase. Biochemistry and Molecular Biology Ph.D.: 203 204 Grabar TB, C ain BD (2003) Integration of b subunits of unequal lengths into F1F0ATP synthase. J Biol Chem 278: 34751 34756 Grabar TB, Cain BD (2004) Genetic complementation between mutant b subunits in F1F0 ATP synthase. J Biol Chem 279: 31205 31211

PAGE 205

205 Green DE, Ziegl er DM, Doeg KA (1959) Sequence of components in the succinic chain of the mitochondrial electron transport system. Arch Biochem Biophys 85: 280 282 Green DE, Hatefi Y (1961) The mitochondrion and biochemical machines. Science, 133: 13 19 Green DE, Tzagoloff A (1966) The mitochondrial electron transfer chain. Arch Biochem Biophys 116: 293 304 Greie JC, Deckers Hebestreit G, Altendorf K (2000a) Subunit organization of the stator part of the F0 complex from Escherichia coli ATP synthase. J Bioenerg Biomembr 32: 357 364 Greie JC, Deckers Hebestreit G, Altendorf K (2000b) Secondary structure composition of reconstituted subunit b of the Escherichia coli ATP synthase. Eur J Biochem 267: 3040 3048 Greie JC, Heitkamp T, Altendorf K (2004) The transmembrane domai n of subunit b of the Escherichia coli F1F(O) ATP synthase is sufficient for H(+) translocating activity together with subunits a and c. Eur J Biochem 271: 3036 3042 Grubmeyer C, Cross RL, Penefsky HS (1982) Mechanism of ATP hydrolysis by beef heart mitoc hondrial ATPase. Rate constants for elementary steps in catalysis at a single site. J Biol Chem 257: 12092 12100 Gunsalus RP, Brusilow WS, Simoni RD (1982) Gene order and genepolypeptide relationships of the protontranslocating ATPase operon (unc) of Es cherichia coli. Proc Natl Acad Sci USA, 79: 320 324 Hardy A, Grabar TB, Bhatt D, Cain BD (2003) Mutagenesis studies of the F1F0 ATP synthase b subunit membrane domain. J Bioenerg Biomembr 35: 389 397 Hartzog PE, Cain BD (1994) Secondsite suppressor mutat ions at glycine 218 and histidine 245 in the alpha subunit of F1F0 ATP synthase in Escherichia coli. J Biol Chem 269: 32313 32317 Hatch LP, Cox GB, Howitt SM (1995) The essential arginine residue at position 210 in the alpha subunit of the Escherichia col i ATP synthase can be transferred to position 252 with partial retention of activity. J Biol Chem 270: 29407 29412 Hazard AL, Senior AE (1994) Mutagenesis of subunit delta from Escherichia coli F1F0ATP synthase. J Biol Chem 269: 418 426 Hirono Hara Y, I shizuka K, Kinosita K, Yoshida M, Noji H (2005) Activation of pausing F1 motor by external force. Proc Natl Acad Sci USA, 102: 4288 4293

PAGE 206

206 Hogeboom GH, Claude A, HotchKiss RD (1946) The distribution of cytochrome oxidase and succinoxidase in the cytoplasm of the mammalian liver cell. J Biol Chem 165: 615 629 Hoppe J, Montecucco C, Friedl P (1983) Labeling of subunit b of the ATP synthase from Escherichia coli with a photoreactive phospholipid analogue. J Biol Chem 258: 2882 2885 Hornung T, Volkov OA, Zaida TM, Delannoy S, Wise JG, Vogel PD (2008) Structure of the cytosolic part of the subunit bdimer of Escherichia coli F0F1ATP synthase. Biophys J 94: 5053 5064 Iko Y, Tabata KV, Sakakihara S, Nakashima T, Noji H (2009) Acceleration of the ATP binding rate of F1ATPase by forcible forward rotation. FEBS Lett 583: 3187 3191 Ishmukhametov RR, Pond JB, Al Huqail A, Galkin MA, Vik SB (2007) ATP synthesis without R210 of subunit a in the Escherichia coli ATP synthase. Biochim Biophys Acta 1777: 32 38 Jones P, Hermolin J, Jiang W, Fillingame RH (2000) Insights into the rotary catalytic mechanism of F0F1 ATP synthase from the cross linking of subunits b and c in the Escherichia coli enzyme. J Biol Chem 275: 31340 31346 Joshi S, Cao GJ, Nath C, Shah J (1997) Olig omycin sensitivity conferring protein (OSCP) of bovine heart mitochondrial ATP synthase: highaffinity OSCPFo interactions require a local alphahelix at the C terminal end of the subunit. Biochemistry 36: 10936 10943 Joshi S, Burrows R (1990) ATP synthase complex from bovine heart mitochondria. Subunit arrangement as revealed by nearest neighbor analysis and susceptibility to trypsin. J Biol Chem 265: 14518 14525 Joshi S, Javed AA, Gibbs LC (1992) Oligomycin sensitivity conferring protein (OSCP) of mito chondrial ATP synthase. The carboxyl terminal region of OSCP is essential for the reconstitution of oligomycinsensitive H(+) ATPase. J Biol Chem 267: 12860 12867 Jounouchi M, Takeyama M, Chaiprasert P, Noumi T, Moriyama Y, Maeda M, Futai M (1992) Escheri chia coli H(+) ATPase: role of the delta subunit in binding Fl to the Fo sector. Arch Biochem Biophys 292: 376 381 Junge W, Lill H, Engelbrecht S (1997) ATP synthase: an electrochemical transducer with rotatory mechanics. Trends Biochem Sci 22: 420 423 Junge W, Sielaff H, Engelbrecht S (2009) Torque generation and elastic power transmission in the rotary F(O)F(1) ATPase. Nature, 459: 364 370

PAGE 207

207 Kagawa R, Montgomery MG, Braig K, Leslie AG, Walker JE (2004) The structure of bovine F1ATPase inhibited by ADP and beryllium fluoride. EMBO J 23: 2734 2744 Kagawa Y, Racker E (1966a) Partial resolution of the enzymes catalyzing oxidative phosphorylation. IX. Reconstruction of oligomycinsensitive adenosine triphosphatase. J Biol Chem 241: 2467 2474 Kagawa Y, Racker E (1966b) Partial resolution of the enzymes catalyzing oxidative phosphorylation. X. Correlation of morphology and function in submitochondrial particles. J Biol Chem 241: 2475 2482 Kanner BI, Serrano R, Kandrach MA, Racker E (1976) Preparation and characterization of homogeneous coupling factor 6 from bovine heart mitochondria. Biochem Biophys Res Commun 69: 1050 1056 Kayalar C, Rosing J, Boyer PD (1977) An alternating site sequence for oxidative phosphorylation suggested by measurement of substrate binding patterns and exchange reaction inhibitions. J Biol Chem 252: 2486 2491 Keilin D (1925) cytochrome, a respiratory pigment, common to animals, yeast, and higher plants. Proceedings of the Royal Society of London Series B 98: 312 339 Keilin D, Hartre e EF (1955) Relationship between certain components of the cytochrome system. Nature, 176: 200 206 Kersten MV, Dunn SD, Wise JG, Vogel PD (2000) Sitedirected spinlabeling of the catalytic sites yields insight into structural changes within the F0F1ATP s ynthase of Escherichia coli. Biochemistry 39: 3856 3860 Kim M, Kim I, Kim S, Kim S, Lee Y, Kang H, Kweon H, Hwang U, Kim M, Lee J, Yoon Y (2004) CDNA cloning and expression of ATP synthase subunit b from the fire bellied frog Bombina orientalis (Anura: Di scoglossidae). DNA Seq 15: 196 201 Krebstakies T, Zimmermann B, Graber P, Altendorf K, Borsch M, Greie JC (2005) Both rotor and stator subunits are necessary for efficient binding of F1 to F0 in functionally assembled Escherichia coli ATP synthase. J Biol Chem 280: 33338 33345 Kuki M, Noumi T, Maeda M, Amemura A, Futai M (1988) Functional domains of epsilon subunit of Escherichia coli H+ ATPase (F0F1). J Biol Chem 263: 17437 17442 Kuo PH, Ketchum CJ, Nakamoto RK (1998) Stability and functionality of cyst eineless F(0)F1 ATP synthase from Escherichia coli. FEBS Lett 426: 217 220 Lai Zhang J, Mueller DM (2000) Complementation of deletion mutants in the genes encoding the F1ATPase by expression of the corresponding bovine subunits in yeast S. cerevisiae. E ur J Biochem 267: 2409 2418

PAGE 208

208 Lai Zhang J, Xiao Y, Mueller DM (1999) Epistatic interactions of deletion mutants in the genes encoding the F1ATPase in yeast Saccharomyces cerevisiae. EMBO J 18: 58 64 Lau WC, Baker LA, Rubinstein JL (2008) CryoEM Structure of the Yeast ATP Synthase. J Mol Biol, 382: 1256 1264 Lee LK, Stewart AG, Donohoe M, Bernal RA, Stock D (2010) The structure of the peripheral stalk of Thermus thermophilus H(+) ATPase/synthase. Nat Struct Mol Biol, 17: 373 378 Leslie A, Abrahams J, Braig K, Lutter R, Menz R, Orriss G, van Raaij M, Walker JE (1999) The structure of bovine mitochondrial F1ATPase: an example of rotary catalysis. Biochem Soc Trans 27: 37 42 Lightowlers RN, Howitt SM, Hatch L, Gibson F, Cox GB (1987) The proton pore in the E scherichia coli F0F1 ATPase: a requirement for arginine at position 210 of the asubunit. Biochimica et biophysica acta, 894: 399 406 Lbau S, Weber J, Senior AE (1998) Catalytic site nucleotide binding and hydrolysis in F1F0ATP synthase. Biochemistry 37: 10846 10853 Long J, DeLeonRangel J, Vik S (2002) Characterization of the first cytoplasmic loop of subunit a of the Escherichia coli ATP synthase by surface labeling, cross linking, and mutagenesis. J Biol Chem 277: 27288 27293 Long JC, Wang S, Vik SB (1998) Membrane topology of subunit a of the F1F0 ATP synthase as determined by labeling of unique cysteine residues. J Biol Chem 273: 16235 16240 Lcken U, Gogol EP, Capaldi RA (1990) Structure of the ATP synthase complex (ECF1F0) of Escherichia coli fro m cryoelectron microscopy. Biochemistry 29: 5339 5343 Lupas A (1997) Predicting coiledcoil regions in proteins. Curr Opin Struct Biol, 7: 388 393 MacLennan DH, Tzagoloff A (1968) Studies on the mitochondrial adenosine triphosphatase system. IV. Purificat ion and characterization of the oligomycin sensitivity conferring protein. Biochemistry 7: 1603 1610 Mao Y, Mueller DM (1997) Structural interactions of the oligomycin sensitivity conferring protein in the yeast ATP synthase. Arch Biochem Biophys 337: 8 16 Markwell M, Haas S, Bieber L, Tolbert N (1978) A modification of the Lowry procedure to simplify protein determination in membrane and lipoprotein samples. Anal Biochem 87: 206 210

PAGE 209

209 Matthies D, Preiss L, Klyszejko AL, Muller DJ, Cook GM, Vonck J, Meier T (2009) The c13 ring from a thermoalkaliphilic ATP synthase reveals an extended diameter due to a special structural region. J Mol Biol, 388: 611 618 McCarty RE, Pittman PR, Tsuchiya Y (1972) Light dependent inhibition of photophosphorylation by N ethylmaleimide. J Biol Chem 247: 3048 3051 McCormick K, Cain BD (1991) Targeted mutagenesis of the b subunit of F1F0 ATP synthase in Escherichia coli: Glu77 through Gln85. J Bacteriol 173: 7240 7248 McCormick K, DeckersHebestreit G, Altendorf K, Cain BD (1993) Characterization of mutations in the b subunit of F1F0 ATP synthase in Escherichia coli. J Biol Chem 268: 24683 24691 McLachlin D, Bestard J, Dunn SD (1998) The b and delta subunits of the Escherichia coli ATP synthase interact via residues in their C terminal regions. J Biol Chem 273: 15162 15168 McLachlin D, Coveny A, Clark S, Dunn SD (2000) Sitedirected cross linking of b to the alpha, beta, and a subunits of the Escherichia coli ATP synthase. J Biol Chem 275: 17571 17577 McLachlin D, Dunn SD (1997) Dimerization interactions of the b subunit of the Escherichia coli F1F0 ATPase. J Biol Chem 272: 21233 21239 McLachlin D, Dunn SD (2000) Disulfide linkage of the b and delta subunits does not affect the function of the Escherichia coli ATP synthase. Biochemistry 39: 3486 3490 Meier T, Polzer P, Diederichs K, Welte W, Dimroth P (2005) Structure of the rotor ring of F Type Na+ ATPase from Ilyobacter tartaricus. Science 308: 659 662 Meisinger C, Pfanner N, Truscott KN (2006) Isolation of yeast mitochondr ia. Methods Mol Biol, 313: 33 39 Mitchell P (1975) The protonmotive Q cycle: a general formulation. FEBS Lett 59: 137 139 Mitchell P (1979) Keilin's respiratory chain concept and its chemiosmotic consequences. Science 206: 1148 1159 Mitchell P (1961) Cou pling of phosphorylation to electron and hydrogen transfer by a chemi osmotic type of mechanism. Nature, 191: 144 148 Moore KJ, Angevine CM, Vincent OD, Schwem BE, Fillingame RH (2008) The cytoplasmic loops of subunit a of Escherichia coli ATP synthase may participate in the proton translocating mechanism. J Biol Chem 283: 13044 13052

PAGE 210

210 Motz C, Hornung T, Kersten M, McLachlin D, Dunn SD, Wise J, Vogel P (2004) The subunit b dimer of the FOF1ATP synthase: interaction with F1ATPase as deduced by sitespecifi c spin labeling. J Biol Chem 279: 49074 49081 Mueller DM, Puri N, Kabaleeswaran V, Terry C, Leslie AG, Walker JE (2004) Crystallization and preliminary crystallographic studies of the mitochondrial F1ATPase from the yeast Saccharomyces cerevisiae. Acta Crystallogr D Biol Crystallogr 60: 1441 1444 Nagley P, Hall RM, Ooi BG (1986) Amino acid substitutions in mitochondrial ATPase subunit 9 of Saccharomyces cerevisiae leading to oligomycin or venturicidin resistance. FEBS Lett 195: 159 163 Nishizaka T, Oiwa K, Noji H, Kimura S, Muneyuki E, Yoshida M, Kinosita K (2004) Chemomechanical coupling in F1ATPase revealed by simultaneous observation of nucleotide kinetics and rotation. Nat Struct Mol Biol, 11: 142 148 Noji H, Yasuda R, Yoshida M, Kinosita K (1997) D irect observation of the rotation of F1ATPase. Nature, 386: 299 302 Noji H, Yoshida M (2001) The rotary machine in the cell, ATP synthase. J Biol Chem 276: 1665 1668 Norais N, Prom D, Velours J (1991) ATP synthase of yeast mitochondria. Characterization of subunit d and sequence analysis of the structural gene ATP7. J Biol Chem 266: 16541 16549 Norais N, Vaillier J, Velours J (1994) Over expression of the yeast ATP synthase subunit D in Escherichia coli: use of polyclonal antibodies directed against rec ombinant subunit D. Biochem Biophys Res Commun 200: 877 883 Ogilvie I, Aggeler R, Capaldi R (1997) Cross linking of the delta subunit to one of the three alpha subunits has no effect on functioning, as expected if delta is a part of the stator that links the F1 and F0 parts of the Escherichia coli ATP synthase. J Biol Chem 272: 16652 16656 Okuno D, Iino R, Noji H (2010) Stiffness of gamma subunit of F(1) ATPase. European biophysics journal : EBJ epub ahead of print Ono S, Sone N, Yoshida M, Suzuki T (2004) ATP synthase that lacks F0asubunit: isolation, properties, and indication of F0b2subunits as an anchor rail of a rotating cring. J Biol Chem 279: 33409 33412 Paul M, Guerin B, Velours J (1992) The C terminal region of subunit 4 (subunit b) is essent ial for assembly of the F0 portion of yeast mitochondrial ATP synthase. Eur J Biochem 205: 163 172

PAGE 211

211 Paul MF, Velours J, Arselin de Chateaubodeau G, Aigle M, Guerin B (1989) The role of subunit 4, a nuclear encoded protein of the F0 sector of yeast mitochondrial ATP synthase, in the assembly of the whole complex. Eur J Biochem 185: 163 171 Poetsch A, Berzborn R, Heberle J, Link T, Dencher NA, Seelert H (2007) Biophysics and bioinformatics reveal structural differences of the two peripheral stalk subunits in chloroplast ATP synthase. J Biochem (Tokyo) 141: 411 420 Pogoryelov D, Yildiz O, FaraldoGmez J, Meier T (2009) Highresolution structure of the rotor ring of a protondependent ATP synthase. Nat Struct Mol Biol, 16: 1068 1073 Porter A, Kumamoto C, Alda pe K, Simoni R (1985) Role of the b subunit of the Escherichia coli protontranslocating ATPase. A mutagenic analysis. J Biol Chem 260: 8182 8187 Prebble J (2002) Peter Mitchell and the ox phos wars. Trends Biochem Sci 27: 209 212 Priya R, Biukovic G, Ga yen S, Vivekanandan S, Gruber G (2009) NMR solution structure of the b3082 domain of subunit b of Escherichia coli F1FO ATP synthase. J Bacteriol 191: 7538 7344 Priya R, Tadwal VS, Roessle MW, Gayen S, Hunke C, Peng WC, Torres J, Gruber G (2008) Low res olution structure of subunit b (b (22156)) of Escherichia coli F (1)F (O) ATP synthase in solution and the bdelta assembly. J Bioenerg Biomembr 40: 245 255 Pullman ME, Penefsky HS, Datta A, Racker E (1960) Partial resolution of the enzymes catalyzing ox idative phosphorylation. I. Purification and properties of soluble dinitrophenol stimulated adenosine triphosphatase. J Biol Chem 235: 3322 3329 Puri N, LaiZhang J, Meier S, Mueller DM (2005) Expression of bovine F1ATPase with functional complementation in yeast Saccharomyces cerevisiae. J Biol Chem 280: 22418 22424 Racker E, Stoeckenius W (1974) Reconstitution of purple membrane vesicles catalyzing light driven proton uptake and adenosine triphosphate formation. J Biol Chem 249: 662 663 RazakaJolly D Rigoulet M, Gurin B, Velours J (1994) Mutation in the hydrophobic domain of ATP synthase subunit 4 (subunit b) of yeast mitochondria disturbs coupling between proton translocation and catalysis. Biochemistry 33: 9684 9691 Recktenwald E, Hess B (1977) A llosteric influence of anions on mitochondrial ATPase of yeast. FEBS Lett 76: 25 28

PAGE 212

212 Rees D, Leslie A, Walker JE (2009) The structure of the membrane extrinsic region of bovine ATP synthase. Proc Natl Acad Sci USA, 106: 21597 21601 Revington M, McLachlin D Shaw G, Dunn SD (1999) The dimerization domain of the b subunit of the Escherichia coli F(1)F(0) ATPase. J Biol Chem 274: 31094 31101 Revington M, Dunn SD, Shaw GS (2002) Folding and stability of the b subunit of the F(1)F(0) ATP synthase. Protein Sci, 11: 1227 1238 Rieske JS, Hansen RE, Zaugg WS (1964) Studies on the Electron Transfer System. 58. Properties of a New OxidationReduction Component of the Respiratory Chain as Studied by Electron Paramagnetic Resonance Spectroscopy. J Biol Chem 239: 3017 3 022 Rodgers A, Capaldi R (1998) The second stalk composed of the band deltasubunits connects F0 to F1 via an alphasubunit in the Escherichia coli ATP synthase. J Biol Chem 273: 29406 29410 Rodgers A, Wilkens S, Aggeler R, Morris M, Howitt S, Capaldi R (1997) The subunit deltasubunit b domain of the Escherichia coli F1F0 ATPase. The B subunits interact with F1 as a dimer and through the delta subunit. J Biol Chem 272: 31058 31064 Rubinstein J, Walker J (2002) ATP synthase from Saccharomyces cerevisiae: location of the OSCP subunit in the peripheral stalk region. J Mol Biol, 321: 613 619 Rubinstein JL, Walker JE, Henderson R (2003) Structure of the mitochondrial ATP synthase by electron cryomicroscopy. EMBO J 22: 6182 6192 Rubinstein JL, Dickson VK, Runswick MJ, Walker JE (2005) ATP synthase from Saccharomyces cerevisiae: location of subunit h in the peripheral stalk region. J Mol Biol, 345: 513 520 Schatz G (1996) Efraim Racker: June 28, 1913 September 9, 1991. Biogr Mem Natl Acad Sci, 70: 321 346 Sc hneider E, Altendorf K (1984) Subunit b of the membrane moiety (F0) of ATP synthase (F1F0) from Escherichia coli is indispensable for H+ translocation and binding of the water soluble F1 moiety. Proc Natl Acad Sci U S A, 81: 7279 7283 Schuster SM, Ebel RE, Lardy HA (1975) Kinetic studies on rat liver and beef heart mitochondrial ATPase. Evidence for nucleotide binding at separate regulatory and catalytic sites. J Biol Chem 250: 7848 7853 Schwem BE, Fillingame RH (2006) Cross linking between helices within subunit a of Escherichia coli ATP synthase defines the transmembrane packing of a four helix bundle. J Biol Chem 281: 37861 37867

PAGE 213

213 Senior A, Nadanaciva S, Weber J (2002) The molecular mechanism of ATP synthesis by F1F0ATP synthase. Biochim Biophys Acta 1 553: 188 211 Senior A, Weber J, Nadanaciva S (2000) The catalytic transition state in ATP synthase. J Bioenerg Biomembr 32: 523 529 Senior AE, Muharemagic A, WilkeMounts S (2006) Assembly of the stator in Escherichia coli ATP synthase. Complexation of al pha subunit with other F1 subunits is prerequisite for delta subunit binding to the N terminal region of alpha. Biochemistry 45: 15893 15902 Sielaff H, Rennekamp H, Engelbrecht S, Junge W (2008) Functional halt positions of rotary FOF1 ATPase correlated w ith crystal structures. Biophysical journal 95: 4979 4987 Sikorski RS, Boeke JD (1991) In vitro mutagenesis and plasmid shuffling: from cloned gene to mutant yeast. Methods Enzymol 194: 302 318 Slater EC (1953) Mechanism of phosphorylation in the respiratory chain. Nature, 172: 975 978 Slater EC (2003) Keilin, cytochrome, and the respiratory chain. J Biol Chem 278: 16455 16461 Slater EC, Kemp A, van der Kraan I, Muller JL, Roveri OA, Verschoor GJ, Wagenvoord RJ, Wielders JP (1979) The ATP and ADP binding sites in mitochondrial coupling factor F1 and their possible role in oxidative phosphorylation. FEBS Lett 103: 7 11 Sorgen P, Bubb M, Cain BD (1999) Lengthening the second stalk of F(1)F(0) ATP synthase in Escherichia coli. J Biol Chem 274: 36261 36266 Sorgen P, Bubb M, McCormick K, Edison A, Cain BD (1998a) Formation of the b subunit dimer is necessary for interaction with F1ATPase. Biochemistry 37: 923 932 Sorgen P, Caviston T, Perry R, Cain BD (1998b) Deletions in the second stalk of F1F0ATP syntha se in Escherichia coli. J Biol Chem 273: 27873 27878 Soubannier V, Rusconi F, Vaillier J, Arselin G, Chaignepain S, Graves PV, Schmitter JM, Zhang JL, Mueller D, Velours J (1999) The second stalk of the yeast ATP synthase complex: identification of subuni ts showing cross links with known positions of subunit 4 (subunit b). Biochemistry 38: 15017 15024 Soubannier V, Vaillier J, Paumard P, Coulary B, Schaeffer J, Velours J (2002) In the absence of the first membranespanning segment of subunit 4(b), the yeast ATP synthase is functional but does not dimerize or oligomerize. J Biol Chem 277: 10739 10745

PAGE 214

214 Spannagel C, Vaillier J, Arselin G, Graves PV, Grandier Vazeille X, Velours J (1998a) Evidence of a subunit 4 (subunit b) dimer in favor of the proximity of A TP synthase complexes in yeast inner mitochondrial membrane. Biochimica et biophysica acta, 1414: 260 264 Spannagel C, Vaillier J, Chaignepain S, Velours J (1998b) Topography of the yeast ATP synthase F0 sector by using cysteine substitution mutants. Cross linkings between subunits 4, 6, and f. Biochemistry 37: 615 621 Stack AE, Cain BD (1994) Mutations in the delta subunit influence the assembly of F1F0 ATP synthase in Escherichia coli. J Bacteriol 176: 540 542 Stalz W, Greie JC, Deckers Hebestreit G, Al tendorf K (2003) Direct interaction of subunits a and b of the F0 complex of Escherichia coli ATP synthase by forming an ab2 subcomplex. J Biol Chem 278: 27068 27071 Steigmiller S, Borsch M, Graber P, Huber M (2005) Distances between the bsubunits in the tether domain of F(0)F(1) ATP synthase from E. coli. Biochim Biophys Acta, 1708: 143 153 Sternweis PC, Smith JB (1980) Characterization of the inhibitory (epsilon) subunit of the protontranslocating adenosine triphosphatase from Escherichia coli. Biochem istry 19: 526 531 Stock D, Gibbons C, Arechaga I, Leslie AG, Walker JE (2000) The rotary mechanism of ATP synthase. Curr Opin Struct Biol, 10: 672 679 Stock D, Leslie AG, Walker JE (1999) Molecular architecture of the rotary motor in ATP synthase. Science 286: 1700 1705 Strauss M, Hofhaus G, Schroder RR, Kuhlbrandt W (2008) Dimer ribbons of ATP synthase shape the inner mitochondrial membrane. EMBO J 27: 1154 1160 Takeyama M, Noumi T, Maeda M, Futai M (1988) Fo portion of Escherichia coli H+ ATPase. Carboxyl terminal region of the b subunit is essential for assembly of functional Fo. J Biol Chem 263: 16106 16112 Tzagoloff A, MacLennan DH, Byington KH (1968) Studies on the mitochondrial adenosine triphosphatase system. 3. Isolation from the oligomycinsens itive adenosine triphosphatase complex of the factors which bind F1 and determine oligomycin sensitivity of bound F1. Biochemistry 7: 1596 1602 Velours J, Arselin de Chateaubodeau G, Galante M, Gurin B (1987) Subunit 4 of ATP synthase (F0F1) from yeast mitochondria. Purification, aminoacid composition and partial N terminal sequence. Eur J Biochem 164: 579 584 Velours J, Durrens P, Aigle M, Gurin B (1988) ATP4, the structural gene for yeast F0F1 ATPase subunit 4. Eur J Biochem 170: 637 642

PAGE 215

215 Velours J Paumard P, Soubannier V, Spannagel C, Vaillier J, Arselin G, Graves PV (2000) Organisation of the yeast ATP synthase F(0):a study based on cysteine mutants, thiol modification and cross linking reagents. Biochimica et biophysica acta, 1458: 443 456 Velou rs J, Arselin G, Paul MF, Galante M, Durrens P, Aigle M, Guerin B (1989) The yeast ATP synthase subunit 4: structure and function. Biochimie 71: 903 915 Velours J, Vaillier J, Paumard P, Soubannier V, Lai Zhang J, Mueller DM (2001) Bovine coupling factor 6, with just 14.5% shared identity, replaces subunit h in the yeast ATP synthase. J Biol Chem 276: 8602 8607 Velours J, Arselin G (2000) The Saccharomyces cerevisiae ATP synthase. J Bioenerg Biomembr 32: 383 390 Vetter IR, Baase WA, Heinz DW, Xiong JP, Snow S, Matthews BW (1996) Protein structural plasticity exemplified by insertion and deletion mutants in T4 lysozyme. Protein Sci, 5: 2399 2415 Vik SB, Antonio BJ (1994) A mechanism of proton translocation by F1F0 ATP synthases suggested by double mutants of the a subunit. J Biol Chem 269: 30364 30369 Voet D, Voet J (2004) Electron Transport and Oxidative Phosphorylation. In Biochemistry Harris D, Fitzgerald P (eds.), pp 802 820. John Wiley & Sons, Inc.: Hoboken Volkov OA, Zaida TM, Voeller P, Lill H, Wis e JG, Vogel PD (2008) Denovo modeling and ESR validation of a cyanobacterial F(o)F(1) ATP synthase subunit bb' left handed coiled coil. Biochim Biophys Acta 787: 183 190 Vollmar M, Schlieper D, Winn M, Buchner C, Groth G (2009) Structure of the C14rotor ring of the proton translocating chloroplast ATP synthase. J Biol Chem 284: 18228 18235 Wada T, Long JC, Zhang D, Vik SB (1999) A novel labeling approach supports the fivetransmembrane model of subunit a of the Escherichia coli ATP synthase. J Biol Chem 274: 17353 17357 Waechter A, Yumin X, Dunn S, Cain BD, Wintermann F, Engelbrecht S, Junge W (2010) Elastic Power Transmission in EFOF1ATPase. Submitted for publication Walker JE, Lutter R, Dupuis A, Runswick MJ (1991) Identification of the subunits of F 1F0ATPase from bovine heart mitochondria. Biochemistry 30: 5369 5378 Walker JE, Runswick MJ, Poulter L (1987) ATP synthase from bovine mitochondria. The characterization and sequence analysis of two membraneassociated subunits and of the corresponding cDNAs. J Mol Biol, 197: 89 100

PAGE 216

216 Walker JE, Saraste M, Gay NJ (1984) The unc operon. Nucleotide sequence, regulation and structure of ATP synthase. Biochim Biophys Acta 768: 164 200 Watt IN, Montgomery MG, Runswick MJ, Leslie AG, Walker JE (2010) Bioenerget ic cost of making an adenosine triphosphate molecule in animal mitochondria. Proc Natl Acad Sci USA, 107: 16823 16827 Weber J, Bowman C, Senior AE (1996) Specific tryptophan substitution in catalytic sites of Escherichia coli F1ATPase allows differentiati on between bound substrate ATP and product ADP in steady state catalysis. J Biol Chem 271: 18711 18718 Weber J, WilkeMounts S, Lee RS, Grell E, Senior AE (1993) Specific placement of tryptophan in the catalytic sites of Escherichia coli F1ATPase provides a direct probe of nucleotide binding: maximal ATP hydrolysis occurs with three sites occupied. J Biol Chem 268: 20126 20133 Weber J, WilkeMounts S, Nadanaciva S, Senior A (2004) Quantitative determination of direct binding of b subunit to F1 in Escheri chia coli F1F0ATP synthase. J Biol Chem 279: 11253 11258 Weber J, WilkeMounts S, Senior AE (2002) Quantitative determination of binding affinity of delta subunit in Escherichia coli F1ATPase: effects of mutation, Mg2+, and pH on Kd. J Biol Chem 277: 1 8390 18396 Weber J, WilkeMounts S, Senior AE (2003a) Identification of the F1binding surface on the deltasubunit of ATP synthase. J Biol Chem 278: 13409 13416 Weber J, Muharemagic A, WilkeMounts S, Senior AE (2003b) F1F0ATP synthase. Binding of delta subunit to a 22residue peptide mimicking the N terminal region of alpha subunit. J Biol Chem 278: 13623 13626 Weimann T, Vaillier J, Salin B, Velours J (2008) The intermembrane space loop of subunit b (4) is a major determinant of the stability of yeast oligomeric ATP synthases. Biochemistry 47: 3556 3563 Wilkens S, Capaldi RA (1998a) Electron microscopic evidence of two stalks linking the F1 and F0 parts of the Escherichia coli ATP synthase. Biochimica et biophysica acta, 1365: 93 97 Wilkens S, Capaldi RA (1998b) ATP synthase's second stalk comes into focus. Nature, 393: 29 Wilkens S, Dunn SD, Chandler J, Dahlquist FW, Capaldi RA (1997) Solution structure of the N terminal domain of the delta subunit of the E. coli ATPsynthase. Nat Struct Biol, 4: 198 2 01 Wise JG, Vogel PD (2008) Subunit b dimer of the Escherichia coli ATP synthase can form lefthanded coiled coils. Biophys J 94: 5040 5052

PAGE 217

217 Wise JG, Vogel PD (2009) Accommodating discontinuities in dimeric left handed coiled coils in ATP synthase external stalks. Biophys J 96: 2823 2831 Wood KS, Dunn SD (2007) Role of the asymmetry of the homodimeric b2 stator stalk in the interaction with the F1 sector of Escherichia coli ATP synthase. J Biol Chem 282: 31920 31927 Yasuda R, Noji H, Yoshida M, Kinosita K Itoh H (2001) Resolution of distinct rotational substeps by submillisecond kinetic analysis of F1ATPase. Nature, 410: 898 904 Yasuda R, Noji H, Kinosita K, Yoshida M (1998) F1ATPase is a highly efficient molecular motor that rotates with discrete 120 degree steps. Cell, 93: 1117 1124 Zaida TM, Hornung T, Volkov OA, Hoffman AD, Pandey SJ, Wise JG, Vogel PD (2009) Conformational changes in the Escherichia coli ATP synthase bdimer upon binding to F(1) ATPase. J Bioenerg Biomembr 40: 551 559 Zalkin H, Pul lman ME, Racker E (1965) Partial resolution of the enzymes catalyzing oxidative phosphorylation. IV. Formation of a complex between coupling factor 1 and adenosine diphosphate and its relation to the 14C adenosine diphosphateadenosine triphosphate exchang e reaction. J Biol Chem 240: 4011 4016

PAGE 218

218 BIOGRAPHICAL SKETCH Amanda (ne Kuhns) Welch was born in 1982 in Kissimmee, Florida to John and Jana Kuhns. She grew up in Kissimmee, FL and graduated in 2000 from Osceola High School, which happened to be the same high school that her father had attended. After high school graduation, she entered the University of Florida as a chemistry major. She successfully completed her Bachelor of Science in 2004 and subsequently entered the management program in the War ren College of Business at the University of Florida. She received her Master of Science with a major in Management in 2005. That same year she married her husband, Stephen Welch, and entered the Interdisciplinary Program in Biomedical Sciences in the College of Medicine at the University of Florida.