The Functional Incorporation of the Mechanosensitive Channel of Large Conductance within a Tethered Lipid Bilayer and the Future Reconstitution of Designer Channels

Material Information

The Functional Incorporation of the Mechanosensitive Channel of Large Conductance within a Tethered Lipid Bilayer and the Future Reconstitution of Designer Channels
Wilson, Danyell
Place of Publication:
[Gainesville, Fla.]
University of Florida
Publication Date:
Physical Description:
1 online resource (159 p.)

Thesis/Dissertation Information

Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Committee Chair:
Duran, Randolph
Committee Members:
Smith, Benjamin W.
Fanucci, Gail E.
Long, Joanna R.
Blount, Paul
Graduation Date:


Subjects / Keywords:
Biochemistry ( jstor )
Cell membranes ( jstor )
Electric current ( jstor )
Electric potential ( jstor )
Electroporation ( jstor )
Ion channels ( jstor )
Ions ( jstor )
Lipid bilayers ( jstor )
Lipids ( jstor )
Proteins ( jstor )
Chemistry -- Dissertations, Academic -- UF
bilayers, channels, dphpc, dphpe, gramicidin, mechanosensitive, tethered, tip
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Chemistry thesis, Ph.D.


When creating a biosensor based on single ion channel activity, the conformational changes within a protein as well as the protein?s ability to reconstitute successfully into non-native lipid environment and remain active must be understood. Therefore, the proximity of key residues in the C-terminal region of the Mechanosensitive Channel of Large Conductance (MscL) was quantified via disulfide bridging of cysteine mutations made to conserved hydrophobic residues in the linker and S3-bundle of this region. The biochemical assays utilizing disulfide bridging provided insight into which residues were dynamic and interactive, enabling them to be studied further via patch clamping. The channel?s C-terminal region during gating was under debate and our results support the theory of the S3-bundle remaining closed during gating. Results from this study also enabled the generation of two new mutant channels that could coordinate heavy metals and the recognition response was a decrease in conductance, and slowed channel kinetics. After the correct configuration of the protein was established, the ideal lipid environment for studying single ion channels in a tethered device for biosensor applications was investigated. The tip-dip electrophysiology method was used to determine an electrically stable lipid environment between different diphytanoyl compositions. Diphytanoyl lipids were vital components of this research due to their increased durability and stability when tethering to solid supports such as the gold surface on the device. Results indicated all of the lipid-ratios suffered from pore formation due to electrical breakdown, both reversible and irreversible. This was the first time that electroporation was reported at such low potentials as 125 mV and 40 mV which may be characteristic of using the tip-dip method with diphtanoyl lipids. The pore formation was random; however distinguishable from single ion channel conductance by configuring I-V curves and evaluating the kinetics of the pores formed in comparison to channel activity. The lipid ratio of 70PC/30PE was chosen as the most stable lipid ratio and was integrated as the synthetic lipid environment for both Gramicidin and MscL on the tBLM device. Results for Gramicidin, indicated that single channel activity within the tBLM were characteristic of the channel. On the other hand, a high applied voltage was required to gate MscL and the conductance response was lower for this channel in comparison to when it is in its natural environment. This lead to the incorporation of a voltage sensitive MscL mutant, K31E, which sensed tension when voltages as low as 85 mV were applied. Preliminary results indicate that this mutant is very active within the tBLM and single channel activity is attainable. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
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.
Thesis (Ph.D.)--University of Florida, 2009.
Adviser: Duran, Randolph.
Statement of Responsibility:
by Danyell Wilson.

Record Information

Source Institution:
Rights Management:
Copyright Wilson, Danyell. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
665095434 ( OCLC )
LD1780 2009 ( lcc )


This item has the following downloads:

Full Text




2 2009 Danyell Wilson


3 This document is dedicated to my grandmother, Ms. Addie Martin.


4 ACKNOWLEDGMENTS First I would like to th ank God for giving me the strength to complete this document. I would also like to thank my committee members, Dr. Randolph Du ran, Dr. Paul Blount, Dr. Gail Fanucci, Dr. Joanna Long, Dr. Benjamin Smith, a nd Dr Tom Lyons, for their advice, guidance, and support. Dr. Blount was a second advisor an d a mentor to me; sincere thanks for opening up the Blounts lab to me. Dr. Fanucci and Dr. Lyons special thanks for opening up the biochemistry labs at UF to me as well. These facilities allowed me perform and complete all of the biochemistry techniques I learned while in Dr. Blounts la b. Dr. Duran, Dr. Smith and Dr. Long, a special thanks for enc ouraging me to continue on to the PhD after my Masters. I would next like to thank Dr. Martin Andersson, Dr. Irene Is cla, Dr. Li-Min Yang, Senior Scientist (Dr.) Robin Wray, and soon to be Dr. Mandy Blackburn as well as Chenyu Zhu, Jeanette Cervantes, Deneyelle Wilson, and Alan Sa lgado. I would also like to thank the Duran group members for always being there. Next, I would love to thank my family a nd friends. My mother Marjorie Wilson who always, always has my back; I am truly grateful to have such a loving charismatic mother. I would love to thank my father, Danny Wilson; a strong, smart, funny, and a blessing to have as a father and friend. I would like to also thank my sisters Sunceray and Ariel Wilson; I am truly blessed to have both of my sister s in life. Sunceray and Ariel we re always just a phone call away and for that I will always be sincerely grat eful. I would like to thank my nephew, Jame Epps, as my inspiration, and the best nephew, ever. I would like to thank all of my friends from childhood, college, graduate school, and church. Without them, I would not have made it through this process. All of thei r prayers, laughter, consoling, partying, editing, and the staying up late nights with me to ensured I became Dr. Wilson. I would also like to acknowledge my dear friend Dr. Charlee Bennett; we did it! We are Ph officially Done!


5 Lastly, I would like to thank th e people who have financiall y supported me and helped me find my passion in science: Dr. Anna Donnelly (SEAGEP), Dr. Jonathan Earl and Ms. Margie Williams (FGLAMP), the Office of Graduate Mi nority Programs (BOE), and both Dr. Blount and Dr. Duran.


6 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........9 LIST OF FIGURES.......................................................................................................................10 LIST OF ABBREVIATIONS........................................................................................................ 13 ABSTRACT...................................................................................................................................14 CHAPTER 1 PROJECT OVERVIEW ......................................................................................................... 16 Introduction................................................................................................................... ..........16 Biological Membrane...................................................................................................... 17 Membrane Proteins and Ion Channels............................................................................. 18 Stretch-Activated Channels.............................................................................................24 E. colis MscL.......................................................................................................... 25 MscL Function and Structure...................................................................................27 S3-COOH-Bundle.................................................................................................... 29 Functionality of the S3-Bundle................................................................................31 Solid-Supported and Tethered Lipid Bilayers.................................................................33 Tethered Lipid Bilayers/Membranes (tBLM)................................................................. 37 DPTL........................................................................................................................39 Surface Binding Region........................................................................................... 39 The Polar Tethered/Hydrophilic Region.................................................................. 40 Diphytanyl Tail Regions.......................................................................................... 40 Diphytanoyl Lipids...................................................................................................41 Summary..................................................................................................................42 2 MATERIALS, METHODS, AND EXPERIMENTAL THEORIES.....................................44 Protein Chemistry and Structural Studies...............................................................................44 Cysteine Mutations in MscL........................................................................................... 44 Biochemistry Assays: Disulfide-trapping........................................................................ 47 Western Blotting.............................................................................................................. 49 Experimental Theory and Pro cedures for Functional Studies................................................ 50 Materials..........................................................................................................................50 Protein Isolation and Purification.................................................................................... 51 MscL Isolation and Purification...............................................................................51 Gramicidin Isolation................................................................................................. 52 Proteoliposome Formation.............................................................................................. 52 DPhPC and DPhPE Vesicle Formatio n and Protein Reconstitution........................ 54


7 Single Channel Measurements: Patc h Clamping Recording T echnique......................... 55 Specialized Patch Clamping Techniques................................................................. 57 Planar Linear Patching: The Tip-Dip Recording..................................................... 57 Tethered Lipid Bilayer............................................................................................. 58 Experimental Setup............................................................................................................. ....59 Tip-Dip Recording Experiment.......................................................................................61 Tethered Devices Experimental Setup............................................................................ 62 The Device...............................................................................................................62 Preparation of the Tethered Bilayer......................................................................... 63 Gigaseal...........................................................................................................................64 Program Analysis for Both Techniques.................................................................................. 65 Analysis of Single Ion Channel Activity................................................................................ 68 Point-Amplitude Histograms...........................................................................................68 Probability Density Function........................................................................................... 70 3 CYSTEINE SUBSTITUTIONS IN THE S3-BUNDLE OF THE MECHANOSENSITIVE CHANNEL OF LARGE CONDUCTANCE................................ 72 Introduction................................................................................................................... ..........72 Materials and Methods...........................................................................................................76 Strains and Cell Growth.................................................................................................. 76 Phenotype studies..................................................................................................... 76 GOF Assays.............................................................................................................. 76 Loss-of-Function Assays................................................................................................. 77 Western Blot Analysis..................................................................................................... 79 Electrophysiology.....................................................................................................80 Results.....................................................................................................................................81 Phenotype Characterization............................................................................................. 81 GOF Studies.............................................................................................................82 LOF Assays..............................................................................................................83 Biochemical Assays......................................................................................................... 84 Discussion...............................................................................................................................90 Conclusion..............................................................................................................................93 4 TESTING THE ELECTRICAL STABLITY OF DIFFERENT LIPID BILAYER COMPOSITIONS OF DPhPC/DPhPE................................................................................... 95 Introduction................................................................................................................... ..........95 Methods and Materials...........................................................................................................97 Results of Lipid Ratios.........................................................................................................100 Pure PC Planar Bilayers................................................................................................ 100 Pure PE Planar Bilayers................................................................................................. 101 50PC/50PE Planar Bilayers........................................................................................... 102 70PC/30PE and 30PC/70PE Planar Bilayers................................................................ 103 Single Channel Results of Gramicidin in 70PC/30PE bilayer............................... 105 Discussion.............................................................................................................................108 The Effect of Hydration State on the Stab ility of Bilayers and Electroporation........... 108


8 Tip-dip and Diphytanoyl Lipid System.................................................................. 111 Conclusion............................................................................................................................112 5 FUNCTIONAL STUDIES OF GRAMICID IN AND MSCL WI THIN A TETHERED LIPID BILAYER MEMBRANE.......................................................................................... 114 Introduction................................................................................................................... ........114 Materials and methods.......................................................................................................... 117 Materials........................................................................................................................117 Proteins....................................................................................................................... ...117 Device............................................................................................................................118 Preparation of the Tethered Bilayer.............................................................................. 118 Protein Reconstitution................................................................................................... 119 Electrophysiology..........................................................................................................120 Results...................................................................................................................................121 Results of Gramicidin Rec onstituted into the tBLM..................................................... 123 Single Channel Analysis of MscL.................................................................................126 Ionic Selectivity...................................................................................................... 127 Lifetime Studies..................................................................................................... 127 Conductance...........................................................................................................127 Discussion.............................................................................................................................132 Conclusions...........................................................................................................................135 6 CONCLUSION..................................................................................................................... 137 APPENDIX A DLS OF THE SIZE DISTRI BUTION OF 70PC/30 PE....................................................... 142 B STABLE BILAYER TRACE AND I-V CURVES.............................................................. 143 C PROTEIN RECTIFICATION WITHIN THE DEVICE...................................................... 145 LIST OF REFERENCES.............................................................................................................146 BIOGRAPHICAL SKETCH.......................................................................................................159


9 LIST OF TABLES Table page 1-1 Different types of ion channels.......................................................................................... 22 2-1 Oligonucleotide primers with cyst eine mutation highlighted in red .................................. 45 5-1 Conductance comparison between the ion ch annels reconstituted in the tBLM and reported conductance. ...................................................................................................... 133


10 LIST OF FIGURES Figure page 1-1 Phospholipids bilayer. And The fluid mosa ic model of the cellular me mbrane. ............. 18 1-2 Equivalent circuit model of the ion channel within the lipid me mbrane. ........................ 20 1-3 Schematic of Gramicidin forming a pore in a synthetic lipid bilayer me mbrane.............. 23 1-4 Schematic representation of E. colis MscL am ino acid sequence and topology of one monomer............................................................................................................................28 1-5 Helical wheel of the S3-bundle.......................................................................................... 30 1-6 Conflicting models of the S3-bundle during gating........................................................... 32 1-7 Solid-Confined or supported membrane............................................................................ 35 1-8 The tethered monolayer component................................................................................... 38 1-9 1,2-Diphytanoyl-sn-Glycer o-3-phosphoethanolamine (DPhPE) and 1,2-Diphytanoylsn-Glycero-3-phospho choline (DPhPC) lipid ................................................................... 41 2-1 Liposome or vesicles fo rmed from phospholipids............................................................. 53 2-2 Schematic of the tip-d ip recording system. ........................................................................ 58 2-3 Experimental Apparatus..................................................................................................... 60 2-4 The tip-dip and tethered stage apparatus........................................................................... 61 2-5 Event detection occurring at level zero and level one....................................................... 67 2-6 The distribution of data points within the hist ogram are fitted by the maximum likelihood with a continuous Gaussian curve.................................................................... 70 3-1 Conflicting models of the S3-bundle during gating........................................................... 74 3-2 Amino acid sequence for MscL and a ribbon design of the linker area and S3 bundle of the C-terminal region with the lo cation of the mutations highlighted. .......................... 75 3-3 Schematic illustration of gain of function assay................................................................77 3-4 Loss of Function schematic representation........................................................................ 78 3-5: Video images obtaine d during the process of ge nerating and patching giant spheroplasts from E. coli ....................................................................................................80


11 3-6 Growth curves for all of the mutants ge nerated in this study as well as for WT MscL and G26H....................................................................................................................... ....82 3-7 LOF study analyzing the percen t survival of the mutants................................................. 83 3-8 Percentage of Dimerization for single mutants A110C, A111C, A112C, a nd A114C ..... 85 3-9 Disulfide crosslinking in single a nd double mu tants identif ied using western blotting...............................................................................................................................86 3-10 Monomers of single mutants and WT as a control exposed to reducing agent DTT......... 86 3-11 Double mutants E119C/V120C, L121C/L122C, L128C/L129C, and AAA exposed to 0.5 M NaCl-LB, water (2nd row), 0.5 M NaCl-LB and 1% H2O2 (3rd row), and water with 1% H2O2.....................................................................................................................87 3-12 Single channel traces of A110H and A112H when exposed to regular patch clamping buffer, and a patch solution of ZnCl2.................................................................................89 3-13 Closed and opened state of MscL with muta tions generated in this study highlighted..... 93 4-1 A reversible hole in th e bilayer formed from elec troporation with an applied potential of -138 mV, and a current transition of more then -200 pA. .............................. 96 4-2 Mini-electroporation of a 100% DPhPC bilayer at 80 mV Pores have a uniform current flow of 10 pA.........................................................................................................97 4-3 SEM images of the borosilicate pipette tips...................................................................... 99 4-4 Electroporation of 100% PC at 120mV with a bilayer lifetime of 56 m inutes................ 101 4-5 Electroporation of 100% PE planar lipid bilayer with poten tials applied fr om 100mV-180 mV............................................................................................................... 102 4-6 Electroporation occurring with a 50PC/50PE bilayer within the 1st 25 sec..................... 102 4-7 Ramp voltages applied to the 70PC/30PE lipid mixture.................................................104 4-8 Episodic trace of the curr ent in pA obtained lipid bilayer compos ed of 70PC/30PE...... 104 4-9 An 18 minute trace of 70PC/30PE lipid mixture............................................................. 105 4-10 Single channel conductance of Gramic idin A is around 67 pS when 60 mV is applied........................................................................................................................ ......106 4-11 Plots of a 1.5 second run at negative 100 mV with and without Gramicidin.................. 106 4-12 Plots IV curves for Gramicidin A for various KCl concentrations.................................. 107


12 5-1 Tethered bilayer membrane array.................................................................................... 116 5-2 AFM images of the monolayer de position and vesicle fusion process............................ 122 5-4 A stable tethered bilayer form from vesicle fusion of 70PC/30PE LUVs....................... 123 5-5 Single channel activity of Gramicidin at an applied voltage of 60 mV with a conductance of 60 pS.. ..................................................................................................... 124 5-6 I-V curves of Gramicidin in the tBLM on the device and within the tip-dip system.. .... 125 5-7 Single channel activity of multiple Gramicid in channels opened within the te thered lipid bilayer.................................................................................................................. ....126 5-8 Single channel activity of WT MscL in the tLBM system : When 300 mV was applied to the membrane MscL opened conducting 266 pS............................................127 5-9 MscL displaying two conducting-states:.........................................................................128 5-10 MscL with the lysine in the 31 position highlighted........................................................ 130 5-10 Single channel recordings of K31E (voltage sensitive mutant of MscL) within the tBLM................................................................................................................................ 130 5-11 Single channel activity of K31E at 143 mV ions was 60 pA providing a conductance of 350 pS..........................................................................................................................131


13 LIST OF ABBREVIATIONS DPhPC 1,2-diphytanoylsn -glycero-3-phosphocholine DPhPE 1,2-diphytanoylsn -3-phosphoethanolamine DPTL 2,3-di-O-phytanoylsn-glucero-1-tetraethylene glycerol-D,L-lipoic acid ester lipid GOF Gain of Function LOF Loss of Function MscL Mechanosensitive Channel of Large Conductance tBLM Tethered Bilayer lipid Membrane


14 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy A PROXIMITY AND FUNCTIONAL STUDY OF THE MECHANOSENSITIVE CHANNEL OF LARGE CONDUCTANCE RECONSTITUTE D IN A TETHERED LIPID BILAYER By Danyell Simone Wilson May 2009 Chair: Randolph Duran Major: Chemistry When creating a biosensor based on single ion channel activity, the conformational changes within a protein as well as the protein s ability to reconstitute successfully into nonnative lipid environment and rema in active must be understood. Therefore, the proximity of key residues in the C-terminal region of the M echanos ensitive C hannel of L arge Conductance (MscL) was quantified via disulfide bridging of cysteine mutations made to conserved hydrophobic residues in the linker and S3-bundle of this regio n. The biochemical assays utilizing disulfide bridging provide d insight into which residues were dynamic and interactive, enabling them to be studied further via patch clamping. The channels C-terminal region during gating was under debate and our results support the theory of the S3-bundle remaining closed during gating. Results from this study also enab led the generation of two new mutant channels that could coordinate heavy metals and the reco gnition response was a decrease in conductance, and slowed channel kinetics. After the correct configuration of the protein was established, the ideal lipid environment for studying single ion channels in a tethered device for biosensor a pplications was investigated. The tip-dip electrophysiology method was used to determine an electrically stable lipid environment between different diphytanoyl co mpositions. Diphytanoyl lipids were vital


15 components of this research due to their increase d durability and stability when tethering to solid supports such as the gold surface on the device. Results indicated all of the lipid-ratios suffered from pore formation due to electr ical breakdown, both reversible a nd irreversible. This was the first time that electroporation was reported at such low potentials as 125 mV and 40 mV which may be characteristic of using th e tip-dip method with diphtanoyl lipids. The pore formation was random; however distinguishable from single ion channel conductance by c onfiguring I-V curves and evaluating the kinetics of th e pores formed in comparison to channel activity. The lipid ratio of 70PC/30PE was chosen as the most stable lipid ratio and was integrated as the synthetic lipid environment for both Gramicidin and MscL on the tBLM device. Results for Gramicidin, indicated that single channel activity within the tBLM were characteristic of the channel. On the other hand, a high applied voltage was required to gate MscL a nd the conductance response was lower for this channel in comparison to when it is in its natural environment. This lead to the incorporation of a voltage sens itive MscL mutant, K31E, which sensed tension when voltages as low as 85 mV were applied. Preliminary results indica te that this mutant is very active within the tBLM and single channel activity is attainable.


16 CHAPTER 1 PROJECT OVERVIEW Introduction Biosensors are small analytical devices that take advantage and m imi cs natures innate ability to sense trace level c oncentrations of analytes ( 1 ). Receptor proteins, for example, are classified as a type of affi nity-based sensor found within many biological sensory systems ( 2). Escherichia colis m echanos ensitive c hannel of l arge conductance, MscL is a receptor protein that plays an important role in maintaining osmo tic homeostasis of these microbes. This enables it to serve as a prototype for mechanosensory tr ansductions, an excellent pharmacological target, and a candidate for biological nanosensors. The goal of this project is to characterize the single channel activity of E. colis MscL within a tethered bilayer lipid membrane (tBLM). The lipid bilayer environment provides both the hydrophobic and hydrophilic requirements of this membrane protein. Tethering the bilayer to a solid support and observing the single channel activity of w ildtype (WT) MscL is the first step in creating a biological nanos ensor. Results from this research can provide insight about the standard characteristics of WT MscL within the tBLM system. Once such standards are configured, future research projects can incorp orate heavy metal, toxin-specific mutants of MscL, or we could generate mutants that alter the single channel activity. For instance, if the large conductance of this cha nnel needs to be decreased to prevent overwhelming the microelectronic device, as pa rt of this study we have mo lecularly engineered a lower conductance MscL. Combining this mutant and the tBLM system can lead to an array of stochastic nanosensors utilizi ng mechanosensitive channels ( 3, 4). Target environments might include biomedical systems such as blood or othe r biological fluids, and/or contaminated water samples (2).


17 In the following sections, a brief introducti on of biological membranes and membrane proteins are presented. Ion cha nnels, and pore forming peptides such as Gramicidin, are also reviewed. The latter will give a historical perspective and an in depth analysis of the advantages and disadvantages of studying bacterial m echanosensitive channels, in particular E. colis MscL. Concluding this chapter will be details as to w hy planar lipid bilayers and tethered bilayer systems provide the best environment for this research. Biological Membrane The plasma membrane for both eukaryotes and prokaryotes serves as a semi-porous barrier encapsulating the cellular constituents ( 5). Phospholipids make up mo st of the lipids present in the plasma membrane ( 5). The structures of these oily substances include a hydrophilic phosphate head group, and a glycerol backbone connecting a hydrophobic tail region. As shown in Figure 1A, these compounds easily self-organiz e such that the hydrophobic tails are oriented inwards forming two integrated layers, known as a bilayer ( 5 ). The hydrophilic heads face the inside of the cell (establishing the intracellu lar face) and the outside (establishing the extracellular face) of the membrane. Since 1970, this system has been described as a fluid mosaic model of lipids and proteins. Isolated and synthetic phospholipid bilayers can closely resemble the cell membrane and are excellent model systems for investigating specific biological proce sses that occur at the cellular level ( 6-9 ). Membrane proteins can be rec onstituted into phosph olipid bilayers, permitting the study of both in vitro functional analysis and prot ein-lipid interactions ( 10, 11). The amphipathic bilayer remains in a fluidic stat e allowing membrane protei ns to diffuse through the double layer to serve as transporters, recep tors and facilitators; enabling the cell to communicate with its environment as shown in Figure 1B.


18 Figure 1-1. A) Phospholipids bilayer. B) The fluid mosaic model of the cellular membrane. Image obtained from Mariana Ruiz Villarreal with permission. Membrane Proteins and Ion Channels Membrane proteins are either attached to or asso ciated with the membrane of a cell ( 12, 13). They exist in two main categories: periphe ral membrane proteins and integral membrane proteins. Peripheral membrane prot eins attach to the lipid bilayer or temporarily to an integral protein by a combination of hydrophobic, electrostrostatic, or other non-covalent interactions ( 13). Integral membrane proteins, on the other hand, are permanently attached to the membrane. They can be further classified according to th eir relationship with the bilayer as integral monotopic proteins or transmembr ane proteins. Integral monot opic proteins are permanently attached to the membrane from either the extracellular or intracellular side ( 13). Transmembrane


19 proteins, the second form of integral membrane proteins, are usually amphipathic and span the entire membrane. One particular type of transmembane proteins are ion channels. Ion channels are found within the membranes of all living cells. They lower the energy barrier for ions to permeate through the cell membrane. ( 14 ). An unequal distribution of ions including Na+, K+, Cland Ca2+ between the intracellular and ex tracellular domain creates an energy barrier (dielectric barrier) termed the membrane potential. This potential which appears across the lipid membranes of all cells can be qu antified in several manners. One estimate is given by the Goldman-Hodgkin-Katz Equation derived from the Nernst equation: e CliKi N i CleKe NClPKPNaaP ClPKPNaaP F RT V ][][][ ][][][ lnrest (1.1) where R is the gas constant, T is temp erature, F is the Faraday constant, Px is the permeability and [X]I and [X]e are the internal and external con centrations of the respective ions (15 ). The equilibrium (resting potential) of a cellular membrane occurs when the intracellular and extracellular ionic concentrations are equal. According to the Gold man-Hodgkin-Katz equation, ion permeability depends on the ions size, mobil ity and concentration. Therefore at different membrane potentials, certain ion channels will translocate specific ions to reach equilibrium. The cellular membrane can be viewed as an RC circuit (refer to Figure 1-2). According to Ohms law the resistance R is equal to the current I, times the voltage V, as seen in equation 1.2: R = IV (1.2) The conductance also according to Ohms law is the inverse of the resistance. Ion channels are like conductors G or gi for specific ions I, across the cells membrane. The electromotive force (Ei) (refer to Figure 1-6) fueling ions to translocate through the channel is the difference of the total potential (potential across the membrane, and the potential of the pore):


20 Ii = gi (E-E0) (1.3) G = 1/R (1.4) The capacitor would represent the membrane separa ting the salt solutions (internal and external ionic solution). Figure 1-2. Equivalent circuit model of the i on channel within the lipid membrane. The ion channel gi, for ion i, the membrane capacitan ce and the electromotive force of the pore are Cm and Ei respectively. This capacitance contributes to defining the numbe r of ions needed to transmit an electrical charge as exemplified in the following equation ( 15): C = Q/E (1.5) C is the capacitance that measures the amount of charge Q, which is required to set up a given potential difference, E across the conductors (the membrane and ev entually ion channels). The change of potential due to current flow Im through the ion channels in the membrane and by small ions that are able to diffuse through micr oscopic opening in the membrane is expressed as: dE/dt = Im/C (1.6) For a given membrane environment, the capacitan ce is relatively constant. When characterizing membrane properties, the capacitance of the lipid bilayer is described by: C=K 0A/d (1. 7)


21 Where K is the dielectric c onstant of the given lipid, 0 is the vacuum permittivity of the system, A is the surface area of the lipid layers and d is the thickness of the bilayer (15). The membrane capacitance adds resistance to the flow of ions at a given time making the previous equation: dE/dt = Im/C = -E/RC = E/ (1.8) Where R is the resistance; we next ta ke into account the initial potential E0, and derive the equation as follows: E = E0 exp (-t/RC) = E0 exp (-t/ ) (1.9) From this we see that the resistance of the membrane and the time constant when ions are transfusing during rest varies from membrane to membrane which is important when considering the results from Chapters 4 and 5. Ion channels are very selective towards specifi c ions; and certain entities may play a major role with the functioning of a channel. For instance, all ion channels ha ve two functional units: 1) a gate that is triggered to open or close via a ligand, a voltage, or changes in the surrounding bilayer 2) a selectivity filter that determines the ions that may pass. Listed in Table 1 are the triggers that are responsible for initiating either a change in conformati on of the ion channel or an opening of a gate/plug within the channel. The opening of an ion channel either allows an influx of specific ions from the extracellular regi on or an outflow of specific ions from the intracellular region. This exchange of ions is known as translocation. The selectivity filters for individual ion cha nnels consist of specific transmembrane amino acids within a certain signature sequence near the N or C terminus and/or a limitation of the translocating pore size ( 16). Dr. MacKinnon, the 2003 Nobel Prize winner in chemistry,


22 discovered conserved signature sequences of am ino acids near the N or C terminus with potassium channels from bacteria to eukaryotic cells ( 16). Table 1-1. Different types of ion channels. Ion Channels Conformational Triggers Voltage-gated channels Open or close, depending on the membrane potential Ligand-gated channels The channel is usually embedded in a larger molecule that has a binding site for a specific ligand on the surface above the membrane. The binding actives opening of the ion channel Cyclic nucleotide-gated channels Internal solutes mediate cellular responses to secondary messengers, along with Ca2+ activate opening of these channels Stretch-activated channels Activated open and closure response from local stretching or compression of the surrounding membrane G-protein-gated channels G-protein attaching to the receptor triggers opening Inward-rectifier K Only a llows for the influx of K+ into the cell only Light-gated channels Dire ctly opened by light Resting channels Constantly open Channel forming Pores Some ion channels are formed from single pep tides that span the membrane, allowing the translocation of ions. These channels do not requ ire a particular stimulus to control the opening and closing of the channel. In nature these channels are usually released from certain organisms as a poisonous defense mechanism or as an an tibody. Some of these single peptides include neutral carriers like Valinonycin and Nonactin, while other transmembrane spanning pores are Nystain, Amphotericin, Tyrocidin and Gramicidin A ( 17, 18). One of the most studied channelforming-pores is Gramicidin. This hydr ophobic peptide antibiotic is produced by Bacillus brevis through an enzymatic synthesis at the onset of sporulation ( 17 ). Five different types of Gramicidin exist: Gram-A, B, C, D, and S. Gramic idin A, B, C, and D all form linear structures while Gramicidin S forms a cyclic peptide. Gram icidin D is a combination of Gramicidin A, B, and C ( 17-19 ). The difference between Gramicidin A, B, and C exist within the amino acid sequence. The trptophan in position 11 for Gram icidin A is replaced by a l-phenylalanine for


23 Gram-B, and l-tyrosine for Gram-C ( 17-20). The duration of this paper will focus on Gramicidin A which will be termed Gramicidin. Gramicidin forms a transmembrane pore when two of the 26 long monomers connect at their N-termini. Six H-bonds at the dimer junction stabilizes th e pore in this conformation as shown in Figure 1-3. Solid-state NMR, 2-dime nsional NMR, circular dichroism and single channel analysis were all used to determine the structure of Gramicidin as a right-handed (RH), single stranded -helical membrane-spanning dimer ( 18). This structure only occurs when the protein is in a bilayer, and/or micellar environment. In organic solvents, Gramicidins structure is either an unfolded (random co il) monomer or several intertwi ned double-stranded (DS) dimer -helixes ( 18 ). The folding of the Gramicidin within the bilayer depends on the length of the acryl chains. Bilayers with a chain length of at least eight carbons tend to provide a sufficient environment for the proper folding of Gramicidin ( 18). The pore diameter is roughly 4 wide and the channel is selectiv e to monovalent cations (K+) and some small anions like Cl( 18). Figure 1-3. Schematic of Gramicidin forming a pore in a synthetic lipid bilayer membrane. As a model system a number of biophysical ch aracteristics have been determined and compared for Gramicidin. Some of them include: effects of the surrounding membrane thickness, the salt concentration in the su rrounding solution, the me mbrane potential and temperature have been assayed for different lipid membrane environments. The short 15 amino acid sequence of alternati ng L-and Dconfigurations, enables it to serve as a model system for understanding basic characteristics of ion channels in synthetic lipid bilayer membranes. In


24 addition, Gramicidin has also served as a m odel for studying ion permeations, both local and long range protein-lipid inte ractions as well as struct ure-function relationships ( 19). In this study, Gramicidin is used to evaluate the tBLM approach as a suitable environment for incorporation of ion channels before in troducing the mechanosensi tive channel into the system. Stretch-Activated Channels Stretch-activated or m echanosensitive (MS) channels participate in many important physiological processes including pain sensation, mi cturition (discharge of urine), hearing, salt and fluid balance, and turgor (pressu re control of the cell) changes ( 21). They react to mechanical stimuli triggering a shift in equilibri um from a closed to open conformation of the protein ( 10). MS channels have been discovered in all the fundamental branches of the phylogenic tree including Bacteria, Archaea, and Eukarya ( 22). In eukaryotes MS channels subsist in the degenerins and epithelial sodium (DEG/ENaC) family, the transient receptor potential (TRP) family, and the mammalian two-pore K+ family. They sense mechanical stimuli in auditory cells as stretch receptors; and they are located in muscle spindles, in vascular endothelium and in neurosensory tissues (23 ). Most of the information obtained concerning the functionality of these channels is inferred from genetic data, both cellular/tissue localization of the proteins or functional manifestations in mutant or knockout animals ( 22). Theses channels usually form multi-component protein complexes, and their actual gating mechanisms are largely unkn own. For these reasons, it is difficult for eukaryote MS channels to serve as model MS channels. Bacterial MS channels have largely served as the model system for mechanosensory transductions for several reasons ( 24, 25 ). 1) Their genetic sequen ce was the first to be cloned and the small gene size allowed for random mutage nesis generating massive libraries of mutant


25 MS channels ( 22, 26-29 ). 2) Conserved sequences were discovered from different bacteria, permitting the identification of conserved regions in 35 MS homologs, providing information relevant to similar regions within these channels ( 22). 3) Bacterial MS channels gate by sensing tension in the membrane independent of other prot eins or artifacts allowi ng them to serve as a simple, easy-to-gate molecular model for functi onal studies. 4) To add to the plethora of information obtained from genetic modeling, both structural and functional evidence of MS channels in eukaryotes and proka ryotes were obtained from studying bacterial mechanosensitive channels. For instance, Escherichia colis MS channel of large conduc tance (MscL) is the most functionally studied MS channel. E. colis MscL To give a historical perspective about the discovery of this protein, there were two incidents that contributed in th e early 1990s: the advancements in the patch clamp technique and the unveiling of E. colis genomic sequence. The patch clamp recording technique at that time measured electrical current from small patches of eukaryotic cellular membranes, allowing the isolation of ion channels ( 30). Modifications to the prep aration of bacterium by Kung et al. which will be discussed in detail in Chapter 3, allowed the patching of these once inaccessible single cell organisms (31). The application of suction to the patch clamp electrodes containing the membrane, led to the discovery of channels that responded to the changes in mechanical characteristics of the bilayer ( 32 ). These channels were novel at the time and identified as mechanosensitive channels. When different variati ons in pressure were applied to the electrode causing changes in tension in the membrane, different conductance states were observed, indicating the existence of multiple mechanosensitive channels. Remaining mindful of the Gram negative classification of E. coli, a series of experiments were conducted to examine the actu al location of these channels within the inner membrane of


26 the cell ( 33 ). This modification took place in 198 9, and it was not until 1993 that Kung et al. discovered the genetic sequence of the m echanos ensitive c hannel of l arge conductance (MscL). The biochemical assay employed to identify the gene responsible for MscL was achieved by a series of membrane fractions being solubi lized using columns that separated the protein according to their biochemical properties ( 34). This was followed by reconstitution of each fraction into azolectin lipids that were assaye d by patch clamping for channel activity. Gel electrophoresis was then used to assay fractions containing a significant number of channels followed by a second separation using an independent column ( 34). This process was performed repeatedly, until results from electrophoresis pr ovided one single protein band indicating the channel was generated from a single gene product. The band was 17kD (34). Consequently, the 37 amino acids that were identified at the Nterminal matched a sequence that was generated unintentionally by a group studying the gene for trkA ( 35). Hamann et al. over sequenced the gene predicting the first 38 residue s of the next gene which was a perfect match for the putative MscL sequence. These key experiments provided the genetic informa tion enabling the cloning and sequencing of the proposed mscL gene encoding a 136 amino acid protein in length with two transmembrane domains by Sukharev et al. ( 34). It was at this point, that the team leading the discovery of E. colis MscL, which included Drs. Sergi Sukharev, Paul Blount, Boris Martinac, Frederick Blattner, and Ching Kung, realized the larger protein band obtained from electroph oresis and a non-denaturing gel filtration column suggested the overall structure had to be a homomultimer. Unsure of the actual structure and mechanisms driving the functional ity of this protein, the authors began an intensive investigation to determine if the gene sequenced was responsib le for the activity of MscL. They generated mscL -null (deletion of the gene encoding for MscL) mutants usi ng insertional disruption and


27 assayed for channel activity by patch clamp; the lack of activity was consistent with MscL playing a role. They also conducted bioc hemical studies that correlated with the electrophysiology studies. The fi nal experiment involved translat ing the gene that encoded the 136 amino acids with a cell-free system, followed by reconstitution of the resulting protein into liposomes and utilizing patch clamping as the electrophysiological assay to observe the single channel activity of MscL ( 36). MscL Function and Structure The proposed function of MscL as a biological emergency release valve for bacteria under hypo-osmotic stress came with the identification of smaller mechanosensitive channels with a redundancy in function ( 26, 32, 37 ). When the bacteria ce ll is placed in hypo-osmotic conditions, MscL senses the tension in the me mbrane causing it to open a conducting pore releasing osmolytes (38-41 ). MscL has a conductance of ~3 nS. When there was no distinct phenotype characteristic for null MscL, MscS an d MscM (mechanosensitive channel of small and mini conductance) open saving the cell by suppressing the phenotype. These two channels also sensed tension in the bilayer and are char acterized according to thei r conductance. The two smaller conducting MS channels are the mechanos ensitive channel of small conductance, MscS (~ 1 nS), and the MS channel of mini conductance, MscM (0.3 nS). This report does not serve as a review for all of the MS channels in E.coli The remainder of this report will focus on MscL from E. coli because it is the most functionally studied and the basis of this research. If the reader is interested in the other MS channels, reference ( 42) provides a very good introduction. Topology studies of the MscL reve aled the location of both the Nand C-termini within the cytoplasm and supported the idea of the protein consisting of two transmembrane domains ( 43). As with most topology studies that lack crystal structures, discrepancies can lead to misinterpretations that in this case suggested the protein was a hexamer instead of a pentamer


28 ( 43). Theses studies did however; lay the foundation for structural studies of this protein. It was not until 1998, when Douglas Rees s group obtained the crystal stru cture of the MS channel from M. tuberculosis using X-ray crystallography data with a resolution of 3.5 that MscL was confirmed to be a pentamer ( 10, 21, 44-48 ). The identification of both the N and C terminal in the cytoplasmic region of the cell was also conf irmed. The identification of two transmembrane domains (TM1 and TM2) joined by a periplasmic loop, as shown in Figure 1-4 was confirmed as well. The TM1 from each peptide forms the 30-40 pore in response to tension in the membrane bilayer, releasing ions and other so lutes from the cytoplasm while the TM2 domain faces the lipid membrane (49). Figure 1-4. Schematic representation of E. colis MscL amino acid sequence and topology of one monomer. Each peptide of the homo pentamer has a S1 region that begins with the N-terminal, two transmembrane regions TM1 (forms the pore) and TM2 which interacts with the lipid bilayer. They each have a perplasmic loop, and S3 region that ends with C-terminal. Images adopted from reference with permission from Dr. Blount ( 43). Mutagenesis studies have indicated the periplasmic loop play s the role of a torsional spring, inhibiting the channel from gati ng under non-osmotic conditions ( 50). The role and conformation during gating of the two cytoplasmi c domains, the N-terminal and the C-terminal,


29 however, is unclear. What is known about the C-te rminal region is that it consists of a linker area starting just after TM2 and the -helical bundle termed S3-bundle. S3-COOH-Bundle A detailed evaluation of the S3-helical bundle obtained from the crystal structure of M. tuberculosis revealed a conform ation in which the hydr ophobic side-chains were exposed outside the S3 helix with the hydrophilic and charged si de-chains packed inside the bundle as shown in Figure 1-5A ( 51). This conformation was originally predicted to be an artifact of the acidic conditions used in the crystallization process, for such a conformation is not predicted to be stable at physiological pH ( 52). In 2001, Sukharev et al. proposed a molecular model of E. colis MscL. Given that the C-terminal structure was in dispute, th is group predicted its structure under physiological conditions and changed the orientation of the helices. In 2003, Sukharev proposed a similar model for the S3-bundle based on an apparent structural homologue, the COMP protein (51). COMP is an extracellu lar protein that mediates the assembly of the cartilage. It has an extr acellular matrix with a five-fold coiled-coil Nterminal stabilized by polar interaction in a bundle ( 51). The C-terminal of E. colis MscL computationally generated from the COMP model has a refined stru cture with a stabilized lefthanded coiled-coil S-3 bundle as shown in Figure 15B. Indeed, a recent reanalysis of the raw X-ray density map has led to a revised model of the M. tuberculosis crystal structure, which is now within the Protein Data Bank ( 20AR) where one of the more significant changes is that the S3 helical bundle m ore closely resembles Sukharev and Guys model. Note, however, that the models described abov e dealt only with the structure of this domain while the channel is in the closed conf ormation. Given that ther e are large structural changes that occur upon channe l gating, an obvious question was whether the S3-bundle


30 remained intact upon channel gating. If so, the li nkers between S3 and TM2 behave as sieves allowing the passage of only smaller intercellular components (see Figure 1-6). Figure 1-5. Helical wheel of the S3-bundle adapted from A) M. tuberculosis and B) the COMP protein. A) The hydrophobic residues, le ucine and isoleuci ne are facing the cytoplasmic region while the hydrophilic arginine is buried inside the bundle. B) The correct orientation at physiological pH of the S3-bundle modeled by Sukharev and Guy, with the hydrophobic residues inside the bundle. Image borrowed from reference (12) with permission from Sukharev. To test this model, disulfide bridging of mutated cysteines within the helical bundle was assayed subsequent to French pressing the cells and following protein purification. These data were compared to patch clamp data. Since di sulfide bridging was obser ved in the biochemical assay, it was assumed the S3-bundle remained intact during gating a nd these results were representative and comparable to the patch clam ping of sheroplast (gia nt bacterial cells). However, biochemical studies us ing disruption buffer, French pressing and/or ambient oxidative conditions may influence oxidatio n either before, after or ev en during gating. Many factors including the degassing of the buffer prior to usage with am bient oxidative conditions and French pressing which destroys the membrane by using extrem ely high pressure may cause changes in the oxidative conditi ons of the protein. Therefore, one can not assume disulfide bridging occurring with these biochemical conditions is comparable to bridging occurring in vivo with sheroplast using the patch clamp. This l eads us to examine the functionality of the S3-


31 bundle for possible clues to its c onformation during gating. There is a possibility that mutating this region can decrease the channel conductance. Functionality of the S3-Bundle In the proposal where MscL S3 domain is mode led af ter COMP, the actual function of this region is believed to limit the size of small i norganic molecules that can pass during osmotic shock (Figure 1.6) ( 51, 53 ). However, in 1997, Cruickshank et al. conducted molecular sieving studies that suggested the pore must be at least 30-40 ( 49). The cytoplasmic proteins released by osmotic shock includes the fairly large th ioredoxin, the elongati on factor Tu, and the molecular chaperone DnaK (54). The mechanism release for these molecules was up for debate until recently. The discrepancy between these two studies was resolved by a subtle difference in experimental approach, indicating these prot eins are actually re leased through MscL ( 54, 55). These studies however support th e possibility of the S3-bundle opening and folding into the bilayer, accommodating the sizes of the proteins being release ( 55). Regardless of which of the models shown in Figure 1-6 are correct, it seems likely that the C-terminus can be manipulated to decrease the conductance of MscL. If the S3-bundle dissembles during gating as shown on the left in Figure 1-6, then locking this area together using cysteine bridging should decrease the conductance. If the other model (on the right) is correct and the S3-bundle remains stable allowing the linke r region to stretch and filter the osmolytes, then creating disulfide bridging in this region and immobilizing the linkers should decrease MscLs conductance. Decreasing the conductance of MscL may enab le it to be reincorporated within the tBLM on a microelect ronic device with other ion cha nnels and its conductance will not over shadow the other lowe r conducting channels. In ad dition, decreasing the conductance would also be a preventive measure in assuring MscLs large conducta nce does not over-whelm the small microelectronic devices due to the large amplification of the signal. Understanding the


32 possible conformational changes in this region may also assist with the reconstitution this protein into the tBLM. If the S3-bundle does remain st able and does not affect the functionality of MscL, then cleaving it may yield greater protein expression and reconstitution. Figure 1-6. Conflicting models of the S3-bundle during gating. The top schematic is the closed state of MscL according the pdb st ructure (OAR). On the right there is a side and top view of the opened state with the S3-bundle se parated and folded into the bilayer. On the left is a side and top view of the ope n state of MscL with the S3-bundle closed during gating and the linker region expanded. Images adopted with permission from references: ( 51, 56) In collaboration with Dr. Paul Blounts laboratory at the University of Texass Southwestern Medical Center we utilized disulfide tr apping to examine the dynamics of the S3bundle and the TM2-S3 linker region. Specifically, we performed disulfide-trapping experiments in vivo to assay the proximity a nd mobility of key residue s in C-terminal region. We chose to mutate amino acids that may be either in close proximity (hydrophobic residues


33 located inside the S3-bundle) or that may interact over time (re sidues in the linker region) to cysteines and assessed the multimerization vi a disulfide bonds using an immune blotting technique. Then, we used the inside-out patch-clam ping method to characterize the function of the mutations that yielded the most intense di sulfide bridging. These mutants were assayed under oxidizing conditions using 0.5% H2O2 to determine if disulfide bridging between amino acids in the C-terminal domain affected the func tionality of the protein. We predicted that a decrease in conductance under thes e conditions would verify bridge formation and a successful alteration of the conformation during gating. By performing th is set of experiments, we determined whether the S3-bundle remains shut a llowing the C-terminal region to behave as a sieve regulating the release of cytoplasmic solu tes. We then took it one step further and attempted to lock the sieve re gion and looked at the overall single channel activity using inside out patch clamping. Finally, we ch aracterized WT MscL within a te thered lipid bilayer as the first step to creating an MscL-ba sed biosensor. Incorporating E. colis MscL into the tethered lipid bilayer, which was done in this study, has added to the ove rall fundamental knowledge of MS channels in different environments. Solid-Supported and Tethered Lipid Bilayers Modifying solid-substrates with lipid bilayers provides mechanical m obility, functional stability and a non-denaturing enviro nment for membrane proteins ( 57). Solid-supports or substrates range from glass, silicon and mica to metals such as platinum and gold ( 11 ). Several technical approaches have been employed to crea te solid-supported lipid bilayers as shown in Figure 1-7. These bio-mimetic supports incl ude solid-supported lipi d bilayer (ssBLM) ( 7, 11, 58, 59), polymer-cushioned lipid bilayer ( 60, 61), hybrid bilayer ( 62, 63), tethered lipid bilayer membrane (tBLM) (64-68 ), freely suspended lipid bilayer ( 69, 70 ), and a supported vesicular layer ( 71, 72 ). It can be quite confusing when such systems are generally referred to as solid-


34 supported lipid bilayers (ssBLM), which is a specific type of solid substrate and bio-mimetic technique. Therefore, for the dur ation of this manuscript, the actua l technique will be referred to as either solid-supported lipid bilayer or ssBLM, and the category or membrane model will be termed solid substrate or solid supported systems. In 1985, Tamm and McConnell conducted one of the first solid supported lipid bilayer studies ( 73). They studied the lateral diffusion of a mixture of lipids sequentially transferred from two monolayers at the ai r-water interface under an oxi dized silicon wafer by extreme pressures ( 73 ). Utilizing fluorescence recovery after pattern phot obleaching, they concluded from lateral diffusion measurements that th e lipids alignment and movement on the solid supports was similar to conventional multilayer systems ( 73). The early experiments by Tamm and McConnell have enabled others to further explore the field by attaching polymers to solid substrates, allowing them to serve as cushions, supporting the bilayer and enabling the recons titution of membrane proteins. Some researchers have also taken advantage of redox reactions that enable thiol-lipids to bind to oxidized metal substrates and form monolayers that can teth er planar bilayers. Today an abundance of both analytical and physical techniques such as electrical impedance spectroscopy (ESI) and atomic force microscopy (AFM), which are used for analyzing globular proteins and re quire a solid surface, can now be employed for characterization of memb rane proteins when reconstituting them into solid-supported bilayers.


35 Figure 1-7. Solid-Confined or supported membra ne. A) Solid-supported lipid bilayer. B) polymer-cushioned lipid bilayer. C) Hybrid bilayer. D) Te thered lipid bilayer. E) Freely suspended lipid bilayer. F-G) supported vesicular layer. Relevant to this work, these pr evious studies have confirmed several key features of solidsupported systems: 1) It is possi ble to fabricate solid substrates that can interact with lipid bilayers. The substrates can either become charged by oxidi zing the surface, or hydrophobic from which a thin aqueous layer between the lipid and the solid surface exist ( 73 ). 2) Van-derWaals, electrostatic, hydration and steric forces control the interactions between the oxidized supports and the planar bilayer. 3) The physioch emical behavior of plan ar bilayers, including a solvent-free environment is still re quired for solid-supported bilayers ( 11 ). 4) Rehydration of the bilayer is required in orde r to ensure the lamellar phase of the lipids. 5) Lastly, bilayers can be formed from either two monolayer transfers or via vesicle fusion ( 63). The goal of this research project, which is to establish a stable, fully flui dic, electrically insulating and robust planar lipid bilayer, is made possible due to the previous rese arch accomplished in this interdisciplinary field.


36 Bilayer requirements for this research incl ude having a conducting solid support, a planar bilayer and a substantial aqueous layer between th e metals surface and the bilayer. Therefore, four out of the seven solid-supported techniqu es were considered for possible membrane systems: the solid-supported BLM, polymer-cus hioned lipid bilayers, hybrid bilayers, freely suspended bilayers and tethered bilayers. So lid-supported lipid bilayers as shown in Figure 17A, are usually formed on a hydrophilic support (glass or oxidized s ilicon wafer) by vesicle fusion or by the sequential transfer of tw o monolayers from the air-water interface ( 68 ). Even though this was the first solid-supported system, ther e were a number of shortcomings that exist within this device, including the small aqueous layer between the bilayer and the solid support. Some researchers hypothesize that this configuration would negativ ely affect the function of the proteins and their lateral mobility, particularly for proteins containing large domains outside of the membrane ( 73). Another concern associated with the ultra thin aqueous layer between the supports surface and the bilayer is the lack of support for total reconstitution of membrane proteins with cytoplasmic regi ons, as in the case for MscL ( 65, 67, 68, 74-77). Polymer-cushioned and hybrid bilayers would both offer long-term stability of the bilayer in this research project. Polymer-substrates form by creating a double tethering of the polymer to the substrates surface and to the bilayer. This interface en hances the robustness of the system ( 61). However, when studying the polymer-cushioned system, Neumann et al. concluded that lateral diffusion within the polymer-cus hioned bilayer is highly restricted ( 61). Lateral diffusion is an important parameter for this study due to the necessity of rec onstituting MscL into the bilayer. When researching the hybrid bilayer system, Anne Plant et al. noted the lack of electrical resistance that occu rs when the synthetic monolayer that forms the hybrid bilayer


37 combines with the natural phosphol ipids upper leaflet. High electrical resi stance enables proper measurement of single ion channel studies( 63 ). Freely suspended lipid bilaye rs as shown in Figure 1-6E resemble the classical planar/black lipid bilayers. The difference be tween the two systems is the formation of the bilayer over an aqueous pore. Pl anar lipid bilayers formation o ccurs over an aperture of 0.1-1 mm. Whereas freely suspended lipid b ilayers are formed over a 25-75 nm hole ( 69). Nonetheless, the membranes resistance within both the freely suspended and planar lipid bilayer is low, which allows for an increase in bac kground noise from leakage through the bilayer ( 58, 69, 78). Tethered lipid bilayers provi de an ideal system for the development of the biosensor studies occurring in this research. They possess a large hydrophilic ionic reservoir for ion mobility and incorporation of the cystoplasmic-terminal region of MscL. Bilayer formation via vesicle fusion in this technique, provides a bila yer with high electrical re sistance and one that is fluidic in nature, allowing lateral di ffusion of different sized proteins ( 61, 65-68). Tethered Lipid Bilayers/Membranes (tBLM) Tethered lipid bilayers are prep ared in a two-step process. First, a monolayer is selfassem bled using a small percentage or lipids which are covalently attached to the substrate. Then lipid vesicles fuse to the monolayer, tethering the bilayer to the system ( 79). Membrane proteins such as ion channels are either incorpor ated into the vesicles before they fuse to the bilayer or proteins can be added to the bila yer once it has self assembled on the monolayer substrate. Cornell et al first developed the tethered bilayer system in an effort to create a biosensor based on ion-channel switches in 1997 ( 57, 80). In 2003, Knoll et al. introduced a stable thiolipid system, 2,3-di-O-phytanoyl -sn-glucero-1-tetraethylene glycerol-D,L-lipoic acid


38 ester lipid (DPTL), that now serves as the monolayer forming component for tethered bilayers developed on gold surfaces (9, 81). This monolayer system consists of a thiol anch or group that covalently attaches to the solid gold surface, a hydrophilic region that forms an a queous reservoir, and an isoprenoid chain tail region that interfaces in the bila yer formed from vesicle fusion as illustrated in Figure 1-8 ( 81). Each component of the thiolipid system plays a ke y role in the overall stability of the tethered bilayer. Figure 1-8. The tethered monolayer component. A) The structur e of 2, 3-di-O-phyl-sn-glucero1-tetraethylene glycerol-D, L-lipoic acid ester lipid (DPTL). B) The length of DPTL attached to a gold surface and the tethered bilayer.


39 DPTL Surface Binding Region The surface binding region of th e tBLM creates the foundation of the system, stabilizing the lipid matrix to the solid support. A gold surf ace is utilized in this research as the solid support. The thiol (R-SH) group from DPTL forms covalent bonds with the oxidized gold surface via a self-assembling monolayer (SAM) process that offers a number of advantages including ( 82 ): Ease of preparation Tunability of surface properties via modification of the molecular structure Enhancement of lateral struct uring on the nanometer scale SAMs spontaneously form, creating complex hierarchical structures from random molecular assemblies ( 61, 82). Thiols (R-SH) on gold surfaces ar e identified as both a traditional and model SAM system where the bond becomes a thiolate (RS-Au) because of sulphurs affinity to first adsorb to the gold ( 82 ). The next step in this chemis orptions process includes the diffusion of DPTL on the golds surface. One of the major benefits of using DPTL is the dithiol (sulphur bond) which covers 5% of the golds surface. Th is allows for maximum insertion of membrane proteins with intracellular components as well as an increase in the aqueous reservoir under the bilayer. Another benefit of having the dithiol bon ds is it ideally decreases the ability of the sulphur to move on the golds surface. The sulphur is known to be mobile on the gold surface, sometimes moving with one or more gold atoms bound to it ( 82). By creating a dithiols, we believe the affinity for movement will be decreased, allowing the development of a stable monolayer. One of the most important requirem ents in developing a thiol-gold SAM, is an extremely clean and ultra smooth surface, which leads to better insulating properties of the tBLM, better self-assembly, and overall heterogeneity of the monolayer ( 9, 65, 82).


40 The Polar Tethered/Hydrophilic Region The tetraethylene-glycol region of DPTL pr ovides a hydrophilic ioni c reservoir between the gold s surface and the membrane. This area is structured similar to that of the cytosol/cytoskeleton found in natural systems (80, 81). It spans roughly 1.4-4.7 nm and ion flow is driven by an externally applied potential betw een the gold surface and a reference electrode in a bulk solution above the bilayer ( 76, 83). An increased length of the ethylene glycol chain affects the stability and viscosity of the ionic re servoir. Adding more ethylene glycols causes the oligimer to adopt motifs slightly distorted in a helical conforma tion, which hinders the formation of a monolayer on the surface ( 9, 68 ). DPTL takes on a 7/2-folded-chain-crystal (FCC) helical conformation (lamellar phase) that is stable and al lows the disulfides to adsorb perfectly to the golds surface ( 79). A decrease in ethylene glycol chains will also decrease the depth of the ionic reservoir and the mobility of ions in this region. Therefore, tetraethylene glycol provides a stable structure suitable for the expe riments taking place in this study. Diphytanyl Tail Regions The isoprenoid chains that ca rry four isoprene units of diphytanyl form the hydrophobic tail of DPTL. This region a ssis ts with forming the lower leaflet of the tethered bilayer ( 66, 79, 81, 84). Serving as the protective membrane surrounding extremeophiles and archaea, 2,3-di-Ophytanyl-sn-glycerol units are known to form stable biomembranes under extreme living conditions ( 81). They have low phase-transition temperatures enabling them to remain in a fluidic structure at room temperature ( 66, 81, 85). Two diphytanyl chains also add to the stability and fluidity of the monolayer compared to a single phytanyl unit, which was used in earlier studies ( 80, 81). As a key component in forming the lower leaflet of the bilayer, the lipids of choice for this research also have a phytanoyl (ester instead of ether connection to the fatty


41 acid chains) tail region. Di phytanyl lipids are utili zed to create the tethered bilayer and also to prevent the development of a hybrid bilayer. Diphytanoyl Lipids 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoethanolam ine (POPE) and 1-palmitoyl-2oleoyl-sn-glycero-3-phosphocholin e (POPC) have served as th e most commonly investigated membrane supporting environment for r econstituted membrane proteins ( 86-89). Diphytanoylbased lipids were chosen for this research because of their enhanced chemical stability throughout a wide temperature ra nge even though they have an ester bond. This bond has a better interaction with the protein then the ether bond pres ent on the DPTL. Diphytanoyl lipids also tether easily to solid surfaces such as gol d substrates which were used for the development of our microelectrode arrays. They function as a more rigid b ilayer in comparison to natural lipids. This may be related to the methyl groups present on the fatty acid chains (refer to Figure 1-8) ( 76 ). Diphytanoyl lipids are not unsaturated like their natural counterparts. The branched chain fatty acid section of the lipid is extremel y disordered, resulting in highly fluidic membrane structures ( 90 ). They are also able to fully mix well with many amphiphiles including cholesterol. Such characteristics enable diphyta noyl lipids to act as a supportive environment for ion channels. O O O O O P O N+ H O O H H H O O O O O P O N+ H O O DPhPE DPhPC O O O O O P O N+ H O O H H H O O O O O P O N+ H O O DPhPE DPhPC Figure 1-9. 1,2-Diphytanoyl-snGlycero-3-phosphoethanolamine (DPhPE) and 1,2-Diphytanoylsn-Glycero-3-phospho choline (DPhPC) lipid


42 The last step in forming the tethered bilayer system occurs via vesicle fusion. Liposomes or vesicles spontaneously form when heterogeneous mixtures of phospholipids are suspended in aqueous solutions ( 9, 68 ). They take on a vesicular structur e whereby the bilayer forms a series of concentric shells encapsulating an aqueous core. Vesicles of lipids DPhPC and DPhPE, 1,2diphytanoylsn -glycero-3-phosphoethanolamine as shown in Figure 1-9 are utilized in this research. Phosphatidylcholine and Phosphatidylet hanolamine head structures make up 64 and 17 percent of the mammalian plasma membranes, respectively ( 12 ). Therefore, vesicle mixtures of these two lipids will be developed and allowed to fuse to the DPTL monolayer creating the tBLM. This fusion process is facilitated by Van der Wals, hydrophobic and hydrophilic interactions whereby the vesicles open and the hydrophobic tail regions ali gn with the tail region of DPTL. Some of the lipids integrated with the DPTL and form the lower leaflet in a manner where the headgroups enter the re servoir and the tails align si de by side with the 2,3-di-Ophtanyl-sn-glycerol units. The remaining lipids form the upper leaflet of the bilayer with the headgroup facing the bulk solution above the bilayer as shown in Figure 1-8B. Vesicle fusion is also a self assembling process were the rate of formation is temperature dependent. Summary In summary, the three main goals of this re search were: 1) Determine the proximity and mobility of like residues in the C-terminal of MscL to understand if this region changes conformation during gating, to decrease the overall conductance of the protein with disulfide bridging. 2) Determine the most stable lipid ratio between DPhPC/DPhP E lipids for use in the tBLM device. This was done by reconstituting a mode l channel, Gramicidin, into a stable planar lipid bilayer and analyzing the single channel activity of this protein using the ti p-dip method. 3) Incorporate both Gramicidin and MscL independ ently into the tBLM a nd measure their single


43 ion channel activity as the first step in creating a stochastic biosensor. The chapter to follow will discuss the isolation and purifica tion of MscL as well as the development of the tethered bilayer system. Chapter 2 will also examine the variati ons of the classical patch clamp technique that was used in this study for analysis of planar linear bilayers. Chapter 3 will examine the results from creating the cysteine mutations while Chapter 4 will examine the most stable lipid ratio using a classical planar lipid bilayer technique Concluding this dissertation will be an assessment of both Gramicidin and MscL, reconstitu ted in the tethered bilayer while utilizing a modified patch clamp technique to an alyze the single channel activity.


44 CHAPTER 2 MATERIALS, METHODS, AND EXPERIMENTAL THEORIES Protein Chemistry and Structural Studies Cysteine Mutations in MscL Mutations, cloning, expression, and biochemi stry assays were done by Robin Wray and me at the University of Texas Southwestern Medi cal Center in Dr. Paul Blounts laboratory. Oligonucleotide primers with cysteine mutati ons at positions A110, A111, A112, A114, E119, V120, L121, L122, L128, and L129 were generated with 15 to 17 base-pair flanking sequences on each side (refer to table 2-1). The Quick-Cha nge kit (Stratagene La Jolla, CA) and a modified MegaPrimer PCR technique were used to exte nd and amplify the target DNA. The QuickChange kit was used for mutants A110C and A111C and the MegaPrimer technique was used for the rest of the mutants. Quick-Change amplifies DNA using PfuTurbo, a cloned version of both PFU DNA Polymerase and ArchaeMaxx-polymerase ( 91). ArchaeMaxx-enhancing factor functions as a dUTPase converting dUTP (caused by dCTP deam ination) to dUMP which improves the total yield of PCR product ( 91 ). Quickchange reagents included: 5 l of Reaction buffer (100 mM KCl, 100mM (NH4)2SO4, 200mM Tris(HCl) (pH 8.8), 20mM MgSO4, 1% Triton X-100, 1 mg/ml BSA), 1 l of dsDNA template (62.5 mg/ l), 1.25 l of oligonucleotide prime 1 (1 g/l), 1 l of dNTP (10 M), 39.50 l of ddH2O, and 1 l of Pfu Turbo DNA polymerase (2.5 U/l). The 1st temperature cycle used to denatu re and unwind the DNA was 95C for 30 seconds. The second temperature cycle included 18 rounds of denaturing the plasmid at 95C for an additional 30 seconds, annealing the oligonucleo tide primers with the cysteine mutation to the plasmid at 55C for one minute, and both extending and in corporating the mutagenic primer with nicked circular DNA strands wa s done at 68C for twelve minutes (2 minutes per kb of the


45 plasmid length). 1 l Dpn I endonuclease was incorporated following the temperature cycling to digest the paternal DNA template. The sample was gently mixed with Dpn I followed by a quick spin. The solution was incubated at 37C for 1 h our and15 minutes in a wate r bath to digest the parental DNA. Table 2-1: Oligonucleotide primers with cysteine mutation highlighted in red Mutation Primers Sequences A110C 5'-CGG AAA AAA GAA GAA CCA TGC GCC GCA CCT GCA CCA AC-3' A110C COMP 3'-GT TGG TGC AGG TGC GGC GCA TGG TTC TTC TTT TTT CCG-5' A111C 5'-GGA AAA AAA GAA GAA CCA GCA TGC GCA CCT GCA CCA AC-3' A111C COMP 3'-GT TGG TGC AGG TGC GCA TGC TGG TTC TTC TTT TTT CC-5' A112C 5'-GAA GAA CCA GCA GCC TGC CCT GCA CCA ACT AAA GAA G-3' A112C COMP 3'-C TTC CTT TAG TTG GTGCGG GCA GGC TGC TGG TTC TTC-5' A114C 5'-G AAC CAG CAG CCG CAR CCT TGC CCA ACT AAA GAA GAA G-3' A114C COMP 3'-C TTC TTC TTT AGT TGG GCA AGG TGC GGC TGC TGG TTC-5' E119C/V120C 5'-C ACC TGC CCA ACT AAA GAA TGC TGC TTA CTG ACA GAA ATT CGT G-3' E119C/V120C COMP 3'-C ACG AAT TTC TGT CAG TAA GCA GCA TTC TTA GTT GGT GCA GGT G-5' L121C/L122C 5'-C ACC AAC TAA AGA AGA AGT TGC TGC ACA GAA ATT CGT GAT TGC-3' L121C/L122C COMP 3'-G CAA ATC ACG AAT TTC TGT GCA GCA TAC TTC TTC TTA GTT GGT G-5' L128C/L129C 5'-C TGA CAG AAA ATT CGT GAT TGC TGC AAA GAG CAG AAT AAC CGC TC-3' L128C/L129C COMP 3'-GA GCG GTT ATT CTG CTC TTT GCA GCA ATC ACG AAT TTC TGT AG-5' A110C/A111C/A112C 5'-GG AAA AAA GAA GAA CCA TGC TGC TGC CCT GCA CCA ACT AAA GAA G-3' A110C/A111C/A112C COMP 3'-C TTC TTT AGT TGG TGC AGG GCA GCA GCA TGG TTC TTC TTT TTT CC-3' Quickchange primer #1 5'-CCA TGA TTA CGC CAA GCG CGC AAT TAA CCC TCA C-3' Quickchange primer #2 5'-GTG AGG GTT AAT TGC GCG CTT GGC GTA ATC ATG G-3' The MegaPrimer procedure was done in a tw o PCR step process, where the product of the first PCR step was used as the primer for the second PCR step. The first primer was the


46 mutation codon flanked by 12 codons at each side. The Megaprimer reagents for the 1st PCR included: 34.5 l of ddH2O, 10 l of 10X turbo Pfu polymerase buffer, 1.5 l of DMSO, 1 l of dNTPs (200 uM), 2.5 l of 5long primer (0.5 M ), 0.29 l of 3 primer with the cysteine mutations (0.5 M), 1 l of DNA template, a nd 0.5 l of Phusion DNA pol ymerase (0.02U/ l). The cycles used with this technique were similar to those of Quickchange with the exception of the extension cycle. The temperature used to extend the PCR product was 72C. The product obtain from this PCR was assayed on an agrose ge l to check for specificity and the correct size. If the proper band was obtained the second PCR step was done for the revers e primer. All of the PCR ingredients remained the same except fo r the 0.5 l of the product from first PCR was added for extension by the 3 long reverse pr imer (0.5 M) for pB10b. Restriction enzymes Xho/Xba were used to cut the plasmid expressi on vector and T4-DNA ligase was used to ligate the mutated DNA product to the pB10b vector. Plasmid constructs were transfor med, via electroporation, into DH10 bacteria to yield DNA for sequencing. DH10 cells and the pB10b construct were mixed and allowed to incubate for 10 minutes on ice before being exposed to a pu lse of 1.8 kV. After transformation, 500 l of SOC (super optimal broth with concentrated magnesium chloride or glucose) media was added to the cells. Next, cells were placed at 37C while shaking at 250 cycles per minute before plating them overnight on Luria-Bertani Broth (LB) plates. Single colonies were selected from each plate and added to 4 ml of liquid LB. Colonies were grown overnight at 37C while shaking at 250 cycles per minute. Mutated DNA was isolated from DH10 cells using QIAprep 8 Turbo Miniprep plasmid DNA isolation and puri fication kit (Qiagen Valencia, CA). This miniprep kit uses alkaline lysis in the presence of high denaturants to isolate DNA. The denaturants assist with the inactivation of exogenous and endogenous nucleases which degrade


47 DNA. Both the lipids and proteins are denature d by strong alkali and deterg ents. The detergents are precipitated out with a neutra lizing salt. The pB10b plasmid c onstructs remain stable in the presence of alkali. Centrifugation was done to separate the plasmid solution from the cellular debris. The solution was then forced though the a pparatus with vacuum in the presence of silica glass beads. The DNA attached to the cationic silica glass was wash ed and eluted with distilled water and stored at -80C. After purification, the pB10b constructs were analyzed by restricti on digestion using Xba and Xho and separated using 1% agarose gel electrophoresis. The ta rget MscL DNA sequence was identified and the mutations in the DNA we re confirmed by automated sequencing at the UTSW McDermott Center for Human Genetic s. After the MscL mutated sequence was confirmed, it was transformed to E.coli strain PB104 (null MscL) ( 26, 27, 43, 92 ). Biochemistry assays were performed in vivo using PB104 strains. Biochemistry Assays: Disulfide-trapping Disulfide-trapping is a classi cal fragme nt scanning method which was used to determine the proximity of selective resi dues in the C-terminal region. Disulfide-trapping uses the oxidation of cysteines, either native or introdu ced, to stabilize the proteins in specific conformations by generating S-S br idges within or between the subunits. This technique has been used by Dr. Blount and coworkers to identi fy key residues in the TM1 and the N-terminal domains of MscL ( 93, 94 ). Disulfide trapping of MscL in vivo can be enhanced by shocking the cells. Shocking the cell occurs when E. coli cells are first grown in high osmolarity media containing high salt concentrations At a certain point when th e cells are healthy and protein expression is high, the cells are s ubsequently diluted into a low os motic environment with little to no salt. This rapid change in environments is termed osmotic downshock, coined by McClure and Britten in their 1962 review ( 95 ). The tugor pressure in the membrane created by the


48 downshock conditions triggers the opening of MscL as a release valve, saving the cell. The key component in this reaction that en hances the intensity of disulfide bridging is the release of the reducing agents usually found in the cytoplasm. With these biochemicals missing, the cysteine residues in close proximity (3.6-6.8 ) are able to crosslink. In these experiments, single colonies we re grown in LB containing 1M NaCl while shaking in an incubator at 37C and rotating at 250 cycles per minute. Culture growth was measured to an optical density (OD)600 at 30 minute intervals. At an early log phase, an OD of 0.2-0.3, protein expression was induced by the addition of 1 mM Isopropyl -D-1thiogalactopyranoside, IPTG (Fisher Scientific, Pittsburgh, PA). IPTG is a reagent that binds and inhibits the Lac 1 repressor, inducing the transcripti on of the protein. At an OD of 0.3-0.4, cultures were diluted 20-fold in either 37C wate r thereby shocking the cells or in 500 mM NaCl LB media which serves as a control for the experi ments. Cultures were then allowed to incubate at 37C while rotating at 250 cycles per minute for 20 minutes. Another control was done to ensure crosslinking, whereby cells were diluted 20 fold in LB containing 1% of oxidizing agent H2O2. Subsequently, cells were pelleted by cen trifugation at 4000 rpm for 20 minutes and resuspended in non-reducing Lammi buffer (62.5 mM Tris pH 6.8, 25% v/v glycerol, 2% w/v sodium dodecyl sulfate (SDS), 0.01% w/v bromophenol blue (Sigma, St. Louis MO)) to a final volume which was normalized based on the OD600 before dilution. In order to study the reversibility of disulfide bri dging, reducing agent, -mercaptoet hanol (5%), was added to the sample buffer for half of the samples. All samples were heated at 70C for 4 minutes to ensure protein denaturation followed by centrifugation for 3 minutes at 12000 rpm. Samples then underwent SDS-Page analysis in a 4-20% gradient Tris-HCl polyacrylamide gel (Bio-Rad


49 Hercules, CA). Briefly, SDS-PAGE (s odium d odecyl s ulfate p olyacrlylamide) g el e lectrophoresis uses polyacrlylamide gel to separa te protein strands using two electrodes. The SDS coats the proteins giving th em a negative charge which ideally enables them to travel vertically to the anode. Smaller proteins travel faster through the pores of the gel while larger proteins move slower. Ladders with known protein molecular weights are used to identify the MW of target proteins. MscLs molecular we ight is around 17 kDa. Therefore when two cysteines form a disulfide bond linking two monomers of MscL to form dimers, the molecular weight, doubles to 34 kD. When three protein subunits crossli nk and form trimers, the MW increases to 57 kD. Four diffe rent subunits crosslinking would be represented by a band 77 kD and be called a tetramer. Western Blotting Once the proteins were fractiona ted, western blot was conduct ed to visualize the protein and to determine if the cysteines were able to crosslink. This immunologi cal technique attaches antibodies to the target protein in a given sam ple for identification and qu antization. First, the proteins fractionated by SDS PAGE are transferred to a membrane sheet. Primary antibodies are next added to the membrane sheet to bind the spec ific proteins. This is followed with enzymetagged antibodies that attach to the primary-antibodies. Binding a s ubstrate to the enzyme allows visualization of the selected protein by eith er colorimetric or fluorimetric detection. MscL was electro-transferred to Immobilon Polyinylidene Fluoride (PVDF) membranes (Millipore Corporation, MA) at 105 mV for 70 minutes. Blots were blocked in 2% milk/TBS for an hour at room temperature. Western Blot an alysis was performed using one of the following: primary-antibodies against the MscL C-termi nus, penta-His-tag anti body or HRP-His-Probe (Thermo Fisher Sic. Pierce Protein Research, Rockford, IL). The 2nd antibody used for MscL was Goat-anti-Rabbit HRP (BioRad). The primary-MscL antibody poorly recognized the double


50 mutants L121C/L122C and L128C/L129C due to the close proximity of the mutation sites to the epitope requiring us to use an tibodies which recognized the His-tag epitope. When protein expression was too low for recognition with the penta-His-antibody, we us ed the HRP-His-Probe (HRP-His-Probe was used for the majority of the study). The HRP-His-probe uses a Nichelating group that binds to the 6His-tag presen t on the protein. The HRP-His-probe reacts with the substrate. This reaction gives off a chemilu minescent signal that was quantified using X-ray sensitive film (Kodak). The film was develope d using the dipping method whereby the film was submerged in each of the following for 90 sec ond intervals: developer, water, and fixer (Developer and Fixer purchased fr om Kodak). The film was then rinsed under running water for 2 minutes and quantification was done by measuring the density of the bands using Scion Image Software (NIH). Experimental Theory and Procedures for Functional Studies Materials Different molar ratios of the archaea phos pholipids, 1, 2-diph ytanoyl-sn-glycero-3phosphocholine (DPhPC) and 1,2-diphytanoyl-sn-glycero-3-phosphoethanolamine (DPhPE), (>99% pure purchased from Avanti Polar Lipids Inc) were used to form the bilayers in both the planar linear and the tethered bilayer studies. 2,3-di-O-phytan oyl-sn-glucero-1-tetraethylene glycerol-D-L-lipoic acid ester lipid, DPTL was used to form the monolayer region of the tethered lipid bilayer. The buffer, Morpholi nopropanesulfonic acid (M OPS), >99.5% ultra grade was purchased from Fluka. Potassium chloride ( certified ACS grade), calcium chloride (certified ACS grade) and potassium hydroxide (certified ACS grade) were purchased from Fisher Scientific and utilized upon receipt.


51 Protein Isolation and Purification Recombinant protein isolation and purification is a process that re quires first the known location of the protein in the cell and a m eans to identify the protein once it is separated from its natural environment. This is achieved by adding a tag to the pr otein and designing the experiments around locating that ta g. Common protein ta gs include Histadines, and Strepadvin. Since MscL is a known membrane protein, the focu s of protein purificati on will be strictly on membrane protein techniques. The first step in this process is to solubilize the cell membrane with either an ionic or nonionic detergent. K nowing the critical micelle concentration (CMC) of the detergent is important for removal of th e detergent soon after the membrane has been disrupted. Another technique th at enhances the disruption of the cellular membrane using mechanical forces is the French Pressing met hod. This techniques works by applying an extreme pressure to the cells that causes them to burst. Once the cell membrane is disrupted, a chromatography technique is required to separate the protein from the cellular debris. These techniques range from affinity chromatography, which binds a certain part of the protein to the beads in the column, to ion exchange chromatography. In the latter method, charged materials within the column attract ionic groups on the prot ein (either cationic or anionic depending on the specificity of the column). As a stronger ion enters the column and reacts with the binding agent, the protein is concentrated an d eluted in a concentrated form. MscL Isolation and Purification In order to isolate and purify MscL, the protein is fi rst generated with a Cterminal 6Histag and expressed in the MscL-null strain PB104 using the pET21a expression vector. The pET vector is able to mass-produce proteins due to its T7-promoter site that binds T7 RNA polymerase and promotes transcription of the targ et protein. Secondly, ce lls were grown to an OD of about 0.6 and induced with 1mM IPTG for 1 hour at 37C. The cells were then


52 centrifuged at 4000 rpm and the pellets were re-sus pended in 40 ml of base buffer (10 mM NaPi and 300 mM pH 8.0) and 100 l of protease inhibito r cocktail (His-tag Protease Inhibitor from Sigma) was added to prevent the denaturing of the proteins. A few cr ystals of DNase and lysozyme was added and allowed to mix for 45 minutes. The samples were then disrupted by French pressing at 16K PSI at 4C and membrane fractions were separated and resuspended in extraction buffer (50 mM Na2HPO4/Nah2PO4, pH 8.0, 300 mM NaCl, 20 mM imidazole plus 2% (v/v) Triton X-100) ( 96). The resulting suspension was incuba ted with 500 l of pre-equilibrated extraction buffer containing Ni-NTA agarose (Qia gen) for 30 min at room temp to bind the proteins. To prepare for the protein purification by affinity chromatography, the NTA-agarose matrix was loaded into a 10 cm column with the protein bound. The column was washed with 10 ml of extraction buffer in the presence of 50m M of imidazole and 1% Triton X-100. Proteins were then eluted in 1.5 ml extraction buffer with 200mM imidazole and 0.2% Triton X-100. They were lastly analyzed by SDS-Page gel for purity and stored at -80C. Gramicidin Isolation Gramicidin A was purchased from Sigma and stored in methanol at -20C. This protein did not require an isolation or purif ication and was used as received. Proteoliposome Formation Purified me mbrane proteins reconstituted into phospholipids vesicles are proteoliposomes. This technique allows for in vitro functional studies of membrane proteins reconstituted into either native lipids, a mixture of di fferent lipids, or synthetic lipids ( 97). This process occurs by first developing the liposomes or vesicles of choice. Liposomes (or vesicles) as shown in Figure 2-1 spontaneously form when heterogeneous mixtures of phospholipids are submerged in aqueous solutions at high temperatures ( 9, 68).


53 They take on a vesicular structure whereby the bilayer forms a series of concentric shells encapsulating an aqueous core. Different pr eparation techniques, including extrusion, sonication, and increasing the pH, can create unila mellar vesicles of certain sizes. Both small unilamellar vesicles (SUVs) and large unilamellar vesicles (LUVs) are formed with the additional steps of either s onication, freeze-thawing, or extr usion through different size membranes ( 98). To create proteoliposomes, the liposome must first be disrupted in order to accommodate the insertion of the purified membrane proteins. Because most membrane proteins once purified are stored in mix solutions of native lipids and de tergent, detergents are used to solubilize the vesicles. The purified proteins are then introduced into the mi xed detergent-lipid solution and allowed to self assemble or reconstitute into the lipids. Removal of the detergent is vital to ensuring the complete reconstitution of the protei n. This is done by either dialysis in the presence of lipids or by adding polystyrene bio beads that absorb the detergents. Figure 2-1 Liposome or vesicles formed from phospholipids.


54 DPhPC and DPhPE Vesicle Formation and Protein Reconstitution DPhPC and DPhPE lipids were mi xed in a 7:3 molar ratio and the solvent was evaporated in a vacuumed rotor vapor for 2 hours. For pr oteoliposomes containing Gramicidin (0.2 mg/ml), 4 l of the protein was added at this step in a ratio of 1:1000 proteins to lipid. Once a thin film of the lipids was observed, indicating the evapor ation of the chloroform, 1 mL MOPS buffer (5 mM MOPS, 25 mM KCl, 0.5 mM CaCl2) was added to rehydrate the lipids. They were next heated while stirring for 2 hours at 50C to form vesicles. The vesicles were cooled to room temperature and sonicated for 5 minutes cr eating a cloudy solution. The homogenizer or extruder along with the 100 nm membrane and f ilters were hydrated for 5 minutes. Vesicles were then extruded 17 times at room temperatur e to ensure homology of the LUVs. Vesicle solutions were then placed at 4C until one of the following occurred: they were prep for an experiment, size verification was done using D ynamic L ight S cattering (DLS), or proteoliposomes with MscL were made. There were two different techniques used to create proteo liposomes of MscL once it was purified. In one of the methods, a 5 ml HiTrap Desalting column (GE healthcare life sciences) packed with cross-linked dextran was used for the removal of salts from the purified protein extracts by gel filtration. This chromatography te chnique separates molecules according to their size as they pass through a gel fi ltration packed column. The column used in this study has a size exclusion limit of MW 5000. The column was first equilibrated with 25 ml or 5 column volumes of degassed buffer (20 mM Tris, 500 mM KCl, 0.1 mM CaCl2, pH 7.4) at a flow rate of 1ml/min. The void volume of the column was 1.5 ml, therefore, 1 ml of the protein fraction was loaded onto the column, at a flow rate of 1 ml/m in, 0.5 ml of buffer is a llowed to elute from the column and discarded, and a total of 2 ml of the desalted sample is collected from the column. MscL was then mixed with the lipid vesicles in a ratio of 1:1000, proteins to lipid.


55 In the other MscL proteoliposome technique th at was used in this study, purified MscL (0.9 mg/ml) was slowly added in a 1:120 protein to lipid mixture after vesicles were extruded and disrupted by the addition of 1% Triton-X 100. The pr otein-lipid-detergent matrix was then heated in the water bath at 60C for 30 minutes While the sample was heating, Biobeads (BioRad) were washed 3X with the MOPs buffer used to create the vesicles. The proteoliposome solution was added to the bio-bead s and incubated with light rota tion for either 4 hours at room temperature or overnight at 4C. The proteoli posome solution was lastly removed from the biobeads and stored at 4C. Single Channel Measurements: Patch Clamping Recording Technique In 1967, Dr. Erwin Neher and Dr. Bert Sakm ann, two cell physiologists, designed the patch clamp technique which m easures ion channe l activity. The most co mmon technique at the time was voltage clamping. Patch clamp offered th e advantage of measuri ng electrical current in the picoamperes (pA) range from single ion channels within the cells membrane ( 99). Information obtained from patch clamp reco rdings includes the conductance, voltage dependence, selectivity, open pr obability, dwell times, resi stance of the membrane, and pharmacological profile of the ion channels ( 99 ). This method can also record channel activity from the interior and exterior of the cell membra ne as well as from ion channels in synthetic membranes. In 1991, both physiologist were awar ded the Nobel Prize in Medicine for their discoveries in the fiel d of electrophysiology ( 99 ). The basis of all patch clamp recordings bega n with the cell attached method. In this method, a glass micropipette tip containing the re ference electrode and electrolyte solution is micromanipulated into the cells surface, isolating a patch of th e membrane. Suction is next applied to create a seal. This seal indicates a significant amount of electrical resistance was obtained between the patch containing the ion channe ls and the tip. This electrical resistance is


56 needed to measure ion channel currents free of b ackground noise. The current flow is reflective of all the ion channels within this patch and meas ured with a specific patc h clamp amplifier. The amplified current transitions are then recorded and analyzed with a data acquisition program. Each channel exhibits a unique amplitude threshold, lifetime profile, and conductance parameters which relates to the single channel ac tivity of the ion channel and can be obtained using patch clamp techniques. One biological process measured using the pa tch clamp is the action potential. Action potential is the opening and closi ng of specific ion channels in an effort to balance the ionic concentration over the interior and exterior of the cell ( 100). This macroscopic process usually occurs with the exchange of Na+ ions from the exterior for K+ ions in the interior (100 ). Once a cell reaches its resting potentia l (70-80mV) the ion channel activity decreases and the cellular membrane is theoretically at eq uilibrium. Whole cell patch clam p configurations are capable of measuring this process. Recording systems that evolved from the cellattached configuration include: whole cell, inside-out, and outside-out recording of excised patches. W hole cell configurations are similar to the attached cell conf iguration; however whole cell measurem ents reflect the electrical current of the entire cell. Inside-out and outside-out methods record activity with the patch detached from the cell when suction is applied. Even th en, the resistance is still high enough to reduce background noise. These configurations allow for measurements of ion channels on the internal and external sides of the memb rane respectively. More recently developed are specialized recording techniques that take advant age of synthetic planar bilayers. Different types of ion channels have been di scovered and characterized by the above patch electrophysiology techniques includi ng voltage-gated ion channels, ligand gated channels, and


57 second messenger-activated channels (78, 99). In vivo studies of MscL using the patch clamp technique requires an increase in E coli cell size to enable patching. It also requires a decrease in the peptidoglycen layer to access the inner membrane. In 1987, Dr. Kung and coworkers adopted a technique to generate giant sphereoplast from the J. Adler group. They were able to harvest patchable bacterial spheroplasts by addi ng chemical reagents: cephalexin, lysozyme, and ethylenediaminetetraacetic acid (EDTA). Cepha lexin inhibits cell division but not growth; leading to long filamentous syncitia termed snakes that collapses to into giant spheres, between 5-10 m; which is acceptable for patching ( 38, 101). Digestion at room temperature with the addition of lysozyme and (EDTA) will hydrolyze the peptidoglycan layer weakening the cell wall causing the E.coli spheroplasts to swell ( 31, 38, 39). MgCl2 and sucrose will be incorporated lastly to stabilize and comp lete the spheroplast for proper patching. Specialized Patch Clamping Techniques Planar Linear Patching: The Tip-Dip Recording Specialized patch clamp tec hniques have been developed to simplify the recording methods in which blockers and other analytes are only used for pharmac ology studies. The tipdip bilayer recording method is a planar lipid bilayer technique th at enables the investigation of ion channels reconst ituted in synthetic lipid environm ents as shown in Figure 2-2 ( 99 ). This configuration also allows one to investigate the electrical stability of the bilayer, which was achieved in this study. In this technique, the ion channel/lipid solu tion is added to the surface of an aqueous solution in order to form a monolayer at the airwater interface. A ground electrode is placed in the aqueous bath solution. The micropipette ti p containing the referen ce electrode, is next micromanipulated onto the surface of the monolay er. Upon retracting the tip from the solution, the hydrophilic lipid head groups a lign the surface of the glass tip forming a monolayer. The tip


58 is then carefully micromanipulated down again just to touch the monolayers surface. At this point a seal and a bilayer are formed from th e two lipid layers mergering. The seal is a measurement of the resistance created between th e two electrodes by the b ilayer. If the ion channel Gramicidin was present, then single ch annel activity would be measured. If an ion channel was not present then the electrical prope rties of the bilayer were studied over time. Figure 2-2. Schematic of the tip-dip recording system. First the pi pette tip is inserted into the monolayer at the air-water interface. When the tip is removed form the surface, the hydrophilic groups align the surface of the gl ass micropipette. The tip is carefully brought back in contact with the monol ayer and the hydrophobic and hydrophilic interactions between the lipid s develop a bilayer and the pr otein or peptides within the bilayer self assemble into an ion channel. Potentials are applied and the electrical current transitions are recorded. (F igure borrowed from previous work) Tethered Lipid Bilayer Solid supported lipid systems as explained in Chapter 1 are fairly new m ethods of incorporating phospholipid bilayers on solid substr ates. One of the most successful techniques and the one currently used in this research is the tethered bilayer lipid membrane system, tBLM. This system is composed of a gold substrate that connects to a probe pad where the ground


59 electrode is placed, a monolayer that is covalently linked to the gold surface and a bilayer that is formed via vesicle fusion. A number of both optical and electrical methods have been employed to study membrane proteins within this environment ( 75, 77, 79, 81, 83, 102-104). In this study, the functional activity of MscL and Gramicidin were analyzed independently for the first time in this system. In this technique, a bilayer is self assemb led via vesicle fusion on top of a monolayer covalently bonded to a gold sensor pad. This b ilayer is either pure lipids or proteoliposomes containing Gramicidin or MscL. The sensor pa d is attached to a probe pad where the ground electrode was placed. In order to maintain a level of hydration for the bilayer, a droplet of buffer or water is added to the sensor pad. A microelectrode filled with the same buffer solution was then micromanipulated into the bilayer soluti on and a 3 mV pulse was passed to measure the electrical resistance (bilaye r covering the gold surface) of the system. If the resistance was in the giga-ohm range then higher voltages were applie d and the change in current was measured for the bilayer or the ion channels. Experimental Setup The experimental setup f or both the tip-dip and the tethered bi layer methods consisted of a patch-clamp amplifier, a micromanipulator, two electrodes, a data recording system, an antivibration table, and a Faraday cage with a dust cove r (refer to Figure 2-3). A noise and dust free environment for the recording systems was esta blished with an anti-vibration or vibration isolation table surrounded by a Faraday cage. To limit surrounding vibrations that may disturb the results, a stack of cinderblocks served as the base of the apparatus. In addition, the apparatus was stationed in the corner of the room where vibrations are at a minimum. A wooden plank with inner-tube air cushions was placed on top of the base to separate a sand filled container. The air cushions suspended the apparatus a little unde r an inch above the wooden plank in the


60 center of the cinderblocks as shown in Figure 2-3. The vibration isolation table was placed on top of the sand box followed by a piezoelectric an ti-vibration table which provided the stage for patch clamp recording. To reduce the amount of electrical noise a copp er mesh Faraday cage with aluminum corner connectors was placed around th e stage. In principle, all outside electrical signals are shielded and unable to enter the cag e due to the copper mesh serving as a conductor and blocking noise. Finally, in order to reduce dust and debris from entering the sample stage a plastic covering was placed over the Faraday cage. Figure 2-3. Experimental Apparatus. The expe rimental setup for both the tip-dip method and the tethered device consists of a patch-clamp amplifier, a micromanipulator, two electrodes, a data recording system, an an ti-vibration table, and a Faraday cage.


61 Figure 2-4. The tip-dip and tethered stage appara tus. A) The patch clamp headstage containing the reference electrode which is also attached to the micro-manipulator. There is also a Teflon-well-plate containing six wells 25 by 25 cm wide where the ground electrode is embedded in the electrolyte solution. B) Th e electronic components for the tethered bilayer experime nt include a microscope that allows visualizing the tungsten ground electrode on the probe pad, a micro-manipulator that lowers the tungsten electrode and a patch clamp headstage containing the reference electrode which is also attached to another micro-manipulator. Once we established a noise and dust free environment the head stage of the patch-clamp amplifier was locked to the base plate and attached to the micr o-manipulator. (refer to Figure 2.4a) All metallic and electrica l objects within the cage were then grounded. The amplifier (Axopatch 200B patch clamp amplifier from Molecular Devices formerly Axon Instruments), data acquisition and digitizer (Digidata 1322A from Molecu lar Devices formerly Axon Instruments), manipulator controls and computer were stationed on a desk next to the cage see Figure 2-3. Tip-Dip Recording Experiment When performi ng a tip-dip recording we first coat a silver wire in a commercial bleach solution for 30 minutes to create a silver/silver chloride Ag/Ag/Cl electrode. This reference electrode was chosen because it is cost efficien t, stable, robust, and compatible with our KCl 250 mM, CaCl2 1 mM, and Morpholinopropanesulfonic acid (MOPS) 5mM filling solutions. Both KCl, and CaCl2 were purchased from Fischer Inc., and MOPS was purchased from Fluka. A


62 Teflon well plate was sanitized, fi rst by rinsing with distilled wa ter from a MilliQ pH 7.4 and with ethanol (ethanol purchased from Fische r). After The buffer so lution (MOPS 5 mM/KCL 250 mM/CaCl2 1mM pH 7.4) was added to th e Teflon well and the ground electrode was inserted as shown in Figure 2-6a. The micropipette tip was prepared by a Fl amming/Brown Micropipette Pulle r (Model 97, Sutter Instruments Co.) w ith a heated filament separating a si ngle borosilicate capillary tube. Borosilicate capillaries purchased from Warner Instruments Inc., were 7.5 cm long and 1.5 mm wide. One capillary created two 1 m2 cross tip diameter because the filament heated the glass vertically in the cente r resulting in two 3.25 cm long micropi pette tips. SEM scanning electron microscope (Hitachi S-4000 FE-SEM) images of th e tip were taken to ensure the tip diameter was 1 m across. Once the electrode and buffer were inserted into the tip, and we initiated the aforementioned tip-dip recording. Tethered Devices Experimental Setup The Device The fabrication of the gold device was done in the Microelectronics Re search Center at the University of Texas Austin in Dr. Ananth Doda balapur laboratory by Dr. Da niel Fi ne. The core element of the tethered lipid bila yer which is based on planar lipid s attached to a solid support is a microelectrode device. The microelectrode a rrays used contained 66 pixels per wafer, with each sensor pad formed by evaporation of 3nm of Ti followed by 500 nm of a 60% Au / 40% Pd alloy on the silicon substrate. A 200 nm gold layer was deposited on top of the alloy and a polyimide resist was photolithographically applied to define the pad si ze as well as to electrically isolate the pads from each other. The device was cleaned with hexane, ace tone, ethanol, water, and treated with UV/ozone for 10 min.


63 Preparation of the Tethered Bilayer The device was submer sed into a DPTL-ethanol so lution, (0.3 mg/ml) and left over night in a saturated ethanol environment a nd the DPTL-ethanol solution was placed in a closed container and stored in a desiccator. The device was then rinsed with ethanol. A chloroform solution (2 mg/ml) containing 70PC/30PE mol% was mixed and the solvents were evaporated under vacuum. The vesicle film was rehydrated with 1 ml of water or buffer as explained earlier. They were heated at 50 oC until a clear solution was obtained ( ca 1 hour). After cooling to room temperature the suspension was sonicated for 5 min, followed by filtration through a 0.45 m filter. The vesicle size was determined by Dynamic Light scattering (DLS) (Precision Detectors). DLS also know as Photon Correla tion Spectroscopy correlates information obtained from small particles scattering li ght from a monochromatic light beam such as a laser. The measurements were performed using a 683 nm laser source and a 90 degrees scattering angle, at 20 C. When needed, additional dilutions were performed until the count rate was between the recommended 200 000 and 400 000 counts per second. C ONTIN analysis was used for the size determinations. These small spherical vesicles move in Brownian motion within solutions causing a Doppler shift in the pres ence of this light beam. A size distribution is correlated from this Doppler shift over a short period of time. Typically, vesicle diameters of 150 50 nm were found. A drop of the vesicle solution was deposited on the gold DPTL treated se nsor pad and left for >5 hours at 4oC or for 2 hours at room temperature with constant rehydration allowing for the ve sicle to fuse, i.e., form the tethered bilayer. Prior to th e electrophysiology examinations a drop (5 l) of buffer solution (5.0 mM MOPS, 250.0 mM KCl and 0.1 mM CaCl2, titrated to pH=7.4 using KOH) was applied to the treated sensor pads. Fabrication of the micr oelectrode tip was done in a similar manner to the tip-dip method stated above. The microelectrode tip was then placed in the headstage and micro-


64 manipulated into the tBLM solution. We used a CCD camera to visualize micro-manipulation of the gold plated tungsten ground electrode on to the surface of the probe pad. Gigaseal Every patch clamp technique including both th e tip-dip m ethod and the tethered bilayer requires high resistances. Th e seal resistance between both th e reference and ground electrode must be at least 1 G in order to observe single ion ch annel activity and ideally create an electrical stable bilayer ( 15, 78). A high resistance enables the complete electrical isolation of the membrane patch, and re duces the current noise ( 78). Experiments were not done if the resistance was not in the giga-ohm range because reduced noise increases resolution, consequently increasing accuracy in detecting tran sitions in current levels. The discrepancy of the current noise is related to Johnson voltage noise Si: Si 2 = 4kTfc/R (2.1) Where k is the Boltzmanns constant, T is temperature (in Kelvins), fc is the low pass filter (Hz) ( 78), and R is the resistance. The Johnson noise for a 1 G resistance at 298 K, with a 1 kHz filter is 0.1 pA, and a 10 G resistance results in current noise of 0.04 pA. Now if the resistance is 100 M under the same conditions as the first seal, the noise becomes 0.4 pA. Therefore the higher the seal resistance, th e less noise observed. This resistance in electrophysiology is termed the gigaseal. A strong gigaseal depends on the pipette materi al used, the size of the tip, and a sample area free of microscopic debris. Both soft (thin wall CEE BEE glass) and hard glass pipettes may be used for patch clamp measurements (78). However, hard glass like borosilicate can establish greater resistance because it has thicke r walls. The diameter of the tip depends on the pulling technique and can range form 0.5 to 5 m ( 78, 105). For instance, in a tip dip recording the tip should roughly be 1 m across in order to establish a proper gigaseal with the planar


65 bilayer ( 78 ). The first dip into the m onolayer establishes the pipett es resistance. When using borosilicate pipettes, the resistance of the first dip should be 6 M ( 99). The second dip reflects the resistance of both the pipettes tip and the bila yer and must be within the giga-ohmic range. There is no limitation within the giga-range because the process of establishing the seal is a hit or miss situation. The resistances readout obtaine d from a gigaseal with th e tethered-bilayer is quite different from previous patch clamp techniques described incl uding the tip dip method. In the tethered-bilayer method the microe lectrode does not actually submerge into the bilayer. The microelectrode is inserted into the bulk buffer so lution above the bilayer. The tethered bilayer is formed on the gold surface of the sensor pad, crea ting barriers between th e two electrodes. The resistance measured is the ability of the bilayer to completely cover the sensor pad. Overall, having a gigaseal eliminates or d ecreases the noise from the systems, other background noises are controlled through the experimental setup, and both of these techniques offer a number of advantages to measuring single channel activ ity including the following: These methods are capable of obtaining high resistance seals The channel activity can be recorded in a timely manner Recording reflects activity of one type of ion channel which reduces the need for blockers The lipid compositions can be exactly defined Program Analysis for Both Techniques Software used during experime nts included Axon Instruments Clampex 8.0 or 9.0 for data collection and Clampfit 9.0 for data analysis. Episodic and gap-free were the two main software protocols involved in characterizing the bilayer and th e single channel activity. The Episodic protocol was used to determine the conductance of the bilayer. During these


66 experiments automated potentials from negative 200 mV to positive 200 mV were applied to the system for several seconds. From these studie s, we determined the stability of the bilayer system. The gap free protocol gave us a chance to observe the ion channels activity over a long period of time with differ ent potentials. With the gap free protocol, we controlled the change in potential manually. Therefore a number of e xperiments were conducted using the gap-free protocol. The recording filter was 5 kHz low pass 8-pole Bessel. Event Detection Data analysis was done using the event detect ion for single channel statistics within the Clampfit program. From the event detection, we obtained statistics on the channels average current during open and closed states, dwell time (average time spent at one level during the duration), and current transitions. We also determined the probability to which the channel was in an open and closed state. In order to conduct th is analysis, signals were first filtered at 1 kHz by an 8-pole low-pass Bessel filter to reduce th e high frequency noise during recording. Once the signal was filtered, the baseline was manually adjusted to zero, for the closed state and event detection was performed. The event detection for single channel analys is goes through a single trace and counts the number of events at different am plitude levels as in Figure 2-5. An event is a continuous section of the trace at one level. In the event viewer, we manually se t the different levels that are detected. Level zero accounts for when the channe l is closed or not conducting and level one, or two correspond to the open states of the channe l. This opening is ca lled a threshold. Each change in amplitude to a different level must me et a maximum threshold in order to be counted as an event. The maximum amplitude Amax/A0, threshold depends on the error function erf, the frequency of the filter recording system and the duration of the event w.


67 Amax/A0 = erf (2.668 fcw) (2.2) Because this value varies, we set a 10% level contribution within the software in which the changes in amplitude contributi ng to a new event have to be within 10% of that levels amplitude. For instance, if the existing level is open at 5 pA and a new event begins at 5.5 pA, the 10% contribution setting will reduce the transition in current to 5.05 pA. (0.9 x 5.0 pA) + (0.1 x 5.5) = 5.05pA (2.3) Rapid deviations of the trace that are not within the range of the current level were ignored. Due to possibility of the channe l being open or closed upon enteri ng the event detection, the first and last events were also excluded from the statis tics. Once a count of all the different events at both the open and closed levels was obtained, single channel ( 15) analysis of the Gramicidin and MscL ion channels was performed. Figure 2-5. Event detection occurring at level zero and level one. Level zero indicates the closed state of the channel or the non-conduc ting level and it is ad justed to fit the baseline. Level one is roughl y at -3.0-3.3 pA. Events are highlighted in red and accounts for the open state of the channel. The duration of this experiment was roughly 12 seconds with an applied potentia l of -100.5 mV. Events ranged from 54.5 ms to 200ms at level one.


68 Analysis of Single Ion Channel Activity Single ion channels exhibit stochastic activity. Theoretically the durations of events and the order in which they o ccur are random variables ( 99, 100 ). The channel has no memory system of opening or closing at a given time, and behaves indepe ndently of specific potentials, and lipid environment. Therefore, parameters re presentative of single channel activity, and the information contained in each event are measured from statistical distributions ( 78, 100). Unlike most ion channels, MscL does not have an ion pref erence. There exist two specific aims when analyzing single ion cha nnel activity of this channel: 1) am plitude studies to understand ion permeation through open channels at different ionic compositi ons, and different membrane potentials, 2) the duration of open and shut tim es or the dwell time for lifetime and kinetic studies ( 100 ). Point-Amplitude Histograms The conductance of biological ion channels is restricted by the channels topology including its ma ximum pore expa nsion, length and structure in the lipid bilayer membrane (15, 106). Because of this restriction, there is a specific range in which channels will conduct. For instance, the conductance of Gramicidin is between 20-90 pS. The conductance for MscL is roughly 3 nS; this pore is very large and non-sele ctive due to the nature of the channel. The large range is also due in pa rt to the calculation of conductance and the method used for recording single channel activity. Conductance can be calculated by stationary fluctuation using voltage clamp measurements on large ensembles of channels or measurements of unitary current analysis using patch clamp ( 15). Variance calculations are used for stationary fl uctuations to calculate N, the number of the independent and identical channels open and c ontributing to macroscopi c current I from single channel conductance i:


69 I 2 = iI-I2/N (2.4) Where I 2 is the variance from each channel ( 15 ). Patch clamp techniques require a gigaseal which allows the isola tion of single ion channels resulti ng in direct measurements of unitary current (50). The conductance is calculated from the slope of current-voltage curves derived from Ohms law. Unitary conductance is calculated from point amplitude histograms at given potentials. All digitized current can be plotted as a hi stogram with a peak for each closed and open level. The area under the peak is proportional to the time spent at that level as shown in Figure 2-6. Point-amplitude histograms for each experi ment at a known potential were plotted from information obtained from the single channel event detection. A statistica l frequency count with a minimum bin width of 0.05 mm was performed and fitted by the maximum likelihood with a continuous Gaussian curve using Origin so ftware as in Figure 2-9. The open Popen /close Pclose probability and the unitary conductance are calculated from this distribution. Both probabilities were estimated as a fraction of the ti me the channel remained in the respective state, divided by the overall time of the recording. Un itary conductance is calc ulated by dividing the average amplitude from the amplitude histograms by the applied potential.


70 Figure 2-6. The distribution of data points within the histogram are fitted by the maximum likelihood with a continuous Gaussian curve. Closed and open events of an ion channel at a potential of -60 mV. The maximum open amplitude is -7 pA and the open probability is 0.53 and the cl osed at level zero is 0.48. Understanding the conductance of biological ion channels indicate that the channels are aqueous pores. In 1959, Mullin combined theories of the past and postulated the modern pore theory in relation to ion channe ls. Past theories assumed ions with their hydration shells are spherical, and that an electrical force lines the walls of a cylindr ical pore (ion channel) determining which would permeate ( 15). Mullins theory examines the ions ability to shed all but the innermost layer of water molecules a nd travel through the pore by solvation obtained from the pores wall ( 15). He also stated that the channel s length and inner diameter served as a selectivity filter allowing a maximum amount of ions at a certain size to translocate the pore ( 15). This theory is accepted today, and the root of the actual m echanistic knowledge is dependent upon ionic flux measurements ( 15 ). Therefore, pore formi ng proteins like Gramicidin and reconstituted ion channels su ch as MscL must reproducibly conduc t in a specific range to be considered as ion channels. Kinetic calcul ations of lifetime are determined by using a probability density function (pdf). Probability Density Function As me ntioned throughout this chap ter, single channel activity exis ts in at least two states: the open and closed state. Considering the simplest model of just two states the transition mechanism from one state to the other is denoted by: Where k21 is the transition from clos ed to the open state, and k12 is the transition from open to the closed state. This unimolecular transiti on reaction involves conformational changes of the 3000 4000 5000 6000 ount N Open= .53 Closed=.48 M2 (V15T) -60mV, 7.6 G 7pA 3000 4000 5000 6000 ount N Open= .53 Closed=.48 M2 (V15T) -60mV, 7.6 G 7pA k12k21closed Open State: 1 2 k12k21closed Open k12k21closed Open State: 1 2


71 ion channel from one state to the other ( 107-109 ). Random thermal movement allows the bonds of the protein to vibrate, bend, a nd stretch in the correct rapid tran sition on the picosecond scale. This unsystematic motion leads to the randomness in the lifetime of the open and closed state ( 78). The probability p that the channel will overcome the energy barrier holding the channel open and fail to close is (1-p)p. The probability that it will succeed in the next attempt r is independent of the first attempt: P(r) = (1-p)r-1p, r = 1, 2, 3, (2.5) The lifetime is therefore a proba bility distribution and is expr essed by a probability density function (pdf) where the area under the curve repres ents the probability that the lifetime is less than the specified time t (15, 78). The exponential density of a time interval with mean of = 1/ and equals: f (t) = i 1e-t/ t>0 (2.6) The area under the curve for any pdf must e qual unity. Therefore an ensemble of ion channels will lead to a correspondi ng number of exponential distributions a, and the total area of the i th component, will also have to equal unity when i is mean ( 15). f (t) = aii -1e-t/ (2.7) Data obtained from pdf are displayed in muti-component histograms where they are usually characterized by three different time values These time values are representative of the different open and closed levels exhibited by the ion channel.


72 CHAPTER 3 CYSTEINE SUBSTITUTIONS IN THE S3 -BUNDLE OF THE MECHANOSENSITIVE CHANNEL OF LARGE CONDUCT ANCE Introduction Mechanosensitiv e (MS) channels in bacteria re gulate the osmotic homeostasis of the cell and are currently the most func tional and structural studied mechanosensitive channels. Incorporating E. colis m echanos ensitive c hannel of l arge conductance, MscL within microelectronic devices will provide a new enviro nment for channel analysis. More importantly, this combination of supported lipid bilaye rs and MscL holds promise for many types of biosensory applications. For inst ance, within a device MscL can serv e as the actual biosensor, or it can serve simply as a release valve. The stim ulus required to gate MscL has been extensively researched and is now reflected by the creation of Designer Mu tated MscL Channels. These channels have been engineered and manipulated to respond to changes in pH, redox reagents, sulfhydryl reagents and toxic heavy metals ( 4, 42, 110, 111) thus producing an analyte sensing property of the protein. Although utilizing MscL presents many advantages for such studies, there are also some likely systematic limitations with incorporating this channel within a microelectronic device. Although the large conducta nce of MscL should yield a large signa l to noise ratio, it may also be overwhelming in some cases. Just a handful of MscL channels opening at a high voltage potential could overwhelm a normal amplifier. In the future, one may wish to combine MscL with other sensors within the same chip. MscL, even with its condensed conductance within a microchip (see Chapter 5) still shows a conductan ce an order of magnitude greater than most high-conducting channels recons tituted into the device ( 64, 77, 112 ). Hence, it would be advantageous to decrease the conductance of the channel.


73 The structure of the C-terminal of the prot ein suggests one possible approach for designing an MscL channel with decreased co nductance. The C-terminal area consists of a linker region that begins directly after the TM2 region at appr oximately residue K96 and ends before the alpha helical S3-bundle. The S3-bundle consists of five identical -helices, one provided from each subunit, and the bundle in one model is c onsidered to began at residue V120 ( 113). The Cterminals function is thought to be that of a si ze exclusion filter. Osmolytes pass through one of the two parts of this region before transversing the pore of the channel. A debate exists to whether the S3-bundle remains stable during gating of the channel, or if it opens and folds into the bilayer as shown in Figure 3-1. Since the crystal structure of M. tuberculosis was resolved in the closed state, at a low pH 3.5 and in the pres ence of detergents, the amino acids were arranged in a conformation that is predic ted to be physiologically probable Molecular models have been designed to correct the structure and placed th e amino acids in the correct orientation ( 51). Disulfide-trapping experiments, with cysteine substitutions to residues located near the end of the S3-bundle, were also done in vitro after French Pressing to analysis the possible conformation of this region during gating (51 ). These experiments also termed cysteine bridging, enable sulfur groups to disulfide bind with neighbori ng cysteines when placed in an oxidative environment, and they provided insi ght concerning the proxim ity of specific amino acids. Surkharev et al. found that cysteine bridging occurr ed under both oxidative and ambient conditions, concluding that this region remains in the cytoplasm during gating. Surkharev et al. then formulated a newer model based off of the COMP (as seen on the right of Figure 3-1) that placed the amino acids in the ideal positions during physiological pH ( 51 ). A recent reevaluation of the M. tuberculosis structure was more consistent with this latter model and uploaded in the pdb-20AR ( 114).


74 Figure 3-1. Conflicting models of the S3-bundle during gating. The top schematic is the closed state of MscL according the pdb st ructure (OAR). On the right there is a side and top view of the opened state with the S3-bundle se parated and folded into the bilayer. On the left is a side and top view of the ope n state of MscL with the S3-bundle closed during gating and the linker region expanded. Images adopted with permission from references: ( 51, 56) In this study, we used cysteine substitutions to restrain movement of portions of this Cterminal region in order to investigate key ar eas that will enable future modulation of conductance. The disulfide bridging resulti ng from the cysteine crosslinking between neighboring subunits will occur if sp ecific residues are consistently in close proximity or if they come into contact with one another (due to mob ility or dynamics of the region). If the S3-bundle does actually disassemble upon gating, then cysteine mutations made to residues E119, V120, L121, L122, L128, and L129 would constrain it s movements and should decrease the


75 conductance by locking the S3-bundle together. On the other hand, if the S3-bundle remains in a stable confirmation during gating, then cystei ne mutations to resi dues A110, A111, A112 and A114 in the linker region between the second tran smembrane domain and S3 should decrease the conductance of the channel as shown Figure 3-2. These above disulfide-trapping experiments were explored in vivo and the results were quantified using the we stern blot technique. Figure 3-2. Amino acid sequence for MscL and a ribbon design of the linker area and S3 bundle of the C-terminal region with the locati on of the mutations hi ghlighted. Molecular model was configured according the MscL crystal structure of M. tuberculosis in the pdb-20AR. Mutants A110, A111, A112, L121 and L122 displayed a high level of crosslinking and were assayed with patch clamping for functiona l analysis under ambient, reduced and/or oxidized conditions. The information obtained from this research has provided insight into the


76 interaction of key residues in the C-terminal region and its dynamics during gating, as well as suggested possible ways of controlling the conductance of the channel. Materials and Methods Strains and Cell Growth Cysteine mutations, A110C, A111C, A 112C, A114C, E119C/V120C, L121C/L122C, L128C/L129C, and A110C/A111C/A112C (AAA), we re m ade using the QuickChange sitedirected mutagenesis kit (Stratagene Corporatio n). This kit provides amplification of doublestranded DNA using the PfuTurbo Polymerase Chain Reaction (PCR) technique. pB10D plasmid constructs were then transformed to DH10 bacteria to yield more DNA for sequencing. Qualitative analysis of the mutation and wildtype genetic sequences were performed using Agarose Gel electrophoresis, and the genetic modifications were verified using automated sequencing. Once the mutated seque nce was confirmed, the constructs were transformed into E.coli strain PB104 ( null MscL), then used for GOF assays and disulfide trapping experiments and MJ455 for LOF studies. Phenotype studies Before performi ng the actual biochemical assa ys and electrophysiol ogy, the affect of the cysteine mutation on the survival of the cell was examined. Two extr eme phenotypes can occur when mutating MscL: Gain of Function (GOF ) or Loss of Function (LOF) mutations. GOF Assays GOF mutants show a slowedor no-grow th phenotype not normally observed. Electrophysiological characterization of these m uta nts have shown that th e MscL channel require less tension for gating. Thus, the interpretati on is that these cha nnels are leaky, and in vivo, open spontaneously making the cell sick, or ev en killing it as shown in Figure 3-3.


77 For these assays overnight cultures of mutants, WT (negative control) and G26H (positive control) were diluted until an early lag phase in LB with 100g/ml ampicillin at 37C while shaking at 250 cycles per minute. The optical density, OD600, was taken every 20 minutes. At an OD of 0.2, the samples were split and Isopropyl-D-thiogalactopyranos ide, IPTG (2mM) was added to induce half of the samples. The OD was measured every 20 mins for the next 5 hours and again the next morning. Growth curves for each mutant were generated and compared to WT and G26H. Figure 3-3. Schematic illustration of gain of function assay. These mutants tend to grow slower because of the leaky channels that compromi se the membranes integrity. Therefore, growing the bacteria in liqui d media and recording the opti cal density in intervals will allow proper screening of GOF. Induc ing the cells, which promotes protein expression essentially reveals the GOF. Loss-of-Function Assays LOF are quite the opposite of GOF; these ch annels only respond to extremely higher me mbrane tension as shown in Figure 3-4. In or der to assay for LOF mutants, expression must take place in bacterial strain s lacking both MscL and MscS (M echanos enitive c hannel of s mall conductance explained in Chapter 1) since MscS has a redundancy of function with MscL ( 114-


78 116). As a result, an inducible plasmid containi ng MscLs gene is transformed into MJ455 cells which lack the genes for both MS channels. To test for LOF mutants, MJF455 cultures we re grown for 8 hours at 37C at a pH 7.0 in Citrate/Phosphate-minimal-media. Cultures were next diluted with 0.5 M NaCl and at an OD600 of 0.25 at 37C they were induced with ITPG fo r 30 minutes. After induction, cells were divided and diluted 1:20 in either water (which shocked the cells) or in LB containing 0.5 NaCl (served as a control) for 20 minutes. Colonies were ne xt diluted into 96-well plates followed by further dilution of 20 l of the cultures into 180 l of me dia. Lastly, 5 l were added to the LB plus ampicillin plates and grown overnight. After 24 hours, the colonies were counted and Colonyforming units (CFU) per volume were calculated. Quantification was achieved by determining the relative survival rate of the different mutants versus cells expressing wild type MscL (negative control) and vectorle ss cells (positive controls) ( 3, 46). Figure 3-4. Loss of Function sc hematic representation. A LOF mutation causes a decreased in survival of bacteria cells due the delaye d responds of MscL sens ing an increase in tension. These mutants are identified upon os motic shock where the mutation inhibits the release of osmolytes causing cell lyses.


79 Western Blot Analysis For a detaile d descriptio n of this procedure see Chapter 2. Briefly, single colonies were grown in LB in the presence of 1 M NaCl while shaking in an incubator at 37C rotating at 250 cycles per minute. Culture growth was measured at 30 min intervals. At an early log phase, an OD of 0.2-0.3, protein expression wa s induced by adding 1 mM IPTG. At an optical density of 0.3-0.4, cultures were diluted in either pre-warmed (37C) destilled water, shocking the cells or in LB with 500 mM NaCl as a control. To in sure crosslinking, anothe r control was prepared using LB containing 1% of oxidizing agent H2O2. To study the revers ibility of disulfide bridging, reducing agent mercaptoethanol was added to the sample buffer for SDS-PAGE analysis. All samples were then heated at 70C for 4 minutes followed by centrifugation for 3 minutes at 12000x g before resolution in a 4-20 % gradient Tris-HCl polyacrylamide gel. Proteins were electro-transfe rred to Immobilon Polyinyliden e Fluoride (PVDF) membranes at 105 mV for 70 minutes for proper vi sualization of the protein. Western Blot analysis was performed using one of the following: primary antibodies against the MscL C-terminus, penta-Hisantib ody or HRP-His-Probe. The secondary antibody for MscL was goat-anti-rabbit HRP Bio-rad). The primary MscL antibody poorly recognized the double mutants L121C/L122C and L128C/L129C due to the close proximity of the mutation sites to the epitope. When protein expression was too low for recognition with the penta-Hisantibody, we used the HRP-His-Probe (which was used for the majority of the study). The HRPHis-probe uses a Ni-chelating group that binds to the 6His-tag pr esent on the protein. The HRPHisprobe reacts with the substrate giving off a chemiluminescent signal that was quantified using X-ray sensitive film (Kodak) and the di pping method. Quantification was completed by measuring the density of the bands using the Scion Image Software.


80 Electrophysiology Electrophysiology studies of Msc L require an increase in E. coli cell si ze to enable patching. It also requires a decrease in the peptidoglycen layer in order to access the inner membrane. In 1987, Dr. Kung and coworkers adopted a technique to generate giant sphereoplast from the J. Adler group. Patcha ble bacterial spheroplasts were obtained by the addition chemical reagents: cephalexin 10 mg/ml, lysozyme 5 mg/ml, ethylenediaminetetraacetic acid (EDTA) 125 mM, and DNAase 5mg/ml. Cephalexin inhibits cell division but not growth: leading to long filamentous syncitia termed snakes that collap se into giant spheres, between 5-10 m, which is acceptable for patching ( 38, 101). Figure 3-5: Video images obt ained during the process of generating and patching giant spheroplasts from E. coli. Image A) rapidly growing E. coli cells. Image B) large filamentous cells also known as snakes generated from treatment with cephalexin. Image C) a giant spheroplast created from collapsed snakes. Image D) the patch is established between the pipette tip and the giant spheroplast and upon suction a gigaohm seal was formed. Image dona ted from Dr. Blount and ref: ( 39) Digestion at room temperature w ith the addition of lysozyme a nd (EDTA) leads to hydroxylation of the peptidoglycan layer, w eakening the cell wall causing the E.coli spheroplasts to swell (31,


81 38, 39). Lastly, MgCl2 and sucrose was incorporated to st abilize and complete the spheroplast shown in Figure 3-5 for proper patching. An inside-out patch clamp configuration that accounts for pressure applied to the membrane was exploited for singl e channel analysis of the muta nts. Negative pressure was applied and monitored using a piezo electric pressure transducer at room temperature. Once a gigaohm resistance was obtained between the micro-pipette tip and the patc h, gentle suction and micro-manipulating of the tip away from the spher oplast disassociated the pa tch. This inside-out patch technique allows exposure of MscLs cyto plasmic region to the bath/buffer solution (200 mM KCl, 90 mM MgCl2, 10 mM CaCl2, and 5 mM HEPES adjusted to a pH of 6). Because the pressure required to gate MscL differs from pa tch to patch (due to the different sizes and geometry of the patched membrane; see ref Moe et al 1995), the pressure required to gate MscL is normalized to that of the MscS. MscL constantly opens at 1.4 times the pressure required to open MscS ( 27). Results To test the proximity and the dynamics of the C-term inal region during gating, we generated cysteine substitutions at a number of positions and assayed thei r ability to disulfide bridge. We individually mutated residue s A110C, A111C, A112C, and A114C; while double mutants were made near the end of the C-terminal at positions E119C/V120C, L121C/L122C, and L128C/L129C as shown in Figure 3-2. We also examined a triple mutant A110C/A111C/A112 (AAA). These residues were chosen because they are highly conserved and are of three different regions of the proposed linker/S3 bundle. Phenotype Characterization Conserved residues can play an important role within the pro tein and their mutation can sometimes indirectly harm the bacterial cell. Therefore, to give a rapid glimpse of possible


82 functional changes in channel ac tivity, we decided to conduct two in vivo phenotype studies to understand the effect of the mutated protein on the survival of the host. GOF Studies Growth curves were generated for all of the mu ta nts used in this study. WT was used as a negative control and G26H (a known GOF mutant) as a positive control to assay for the GOF mutants. Results suggest that none of the muta tions from this study were GOF as shown in the growth curves in Figure 3-6. This may be due to the location of the mutations in the C-terminal which would be in agreement with a deletion study where residues deleted from position 110136 mildly increased the sensitivity to tensi on and sub conducting states. Therefore, the mutations within the C-terminal region did not increase the sensitivity of the protein to the tension of the bilayer. 0.00 0.50 1.00 1.50 2.00 2.50 3.0030 70 1 1 0 150 1 9 0 23 0 270 3 1 0 350 420Time (min)Optical Density Eco MscL (UIND) A110C (UIND) A112C (UIND) A114C(UNID) E119C/120C (UNID) L121/L122 (UNID) L128/L (UNID) AAA (UNID) G26H (UNID) Eco MscL (ND) A110C(ND) A112C (ND) A114C(ND) E119C/120C (ND) L121/L122 (ND) L128/L (ND) AAA (ND) G26H (ND) Figure 3-6. Growth curves for all of the mutant s generated in this study as well as for WT MscL and G26H: UNID-Uninduced and ND-induced (n=3).


83 LOF Assays Single colon ies were counted after a series of dilutions following osmotic downshock and the relative survival rate was calculated for each mutation to determine if they complemented the LOF phenotype. As shown in Figure 3-7 only one mutant in this study has demonstrated the characteristics of a partial LOF phenotype, L121C/ L122C. A full LOF, as determined using an empty plasmid as a positive control, is consistently less than 10% survival. This result is not fully understood, however it does suggest that th ese residues may play some role in channel gating. As shown in Figure 3-4 residues 121 and 122 are located directly at the beginning of the S3-bundle. Bridging these residues together may decrease the conductance of the channel by either locking the S3-bundle to gether, or by decreasing the expansion of the linker region. Either of these actions may incr ease the sensitivity to tension required to open MscL. Nonetheless, actual functio nal studies are required to determine if this partial LOF affects the conductance of the channel. It may ju st affect the single channel kinetics. Figure 3-7. LOF study analyzing the percent survival of the mutants (n=3). G26C was used as a control.


84 Biochemical Assays Disulfidetrapping can utilize th e oxidation of cysteines, ei ther native or mutated, to stabilize a protein in specific conformations by generati ng S-S bridges within or between subunits. Disulfide trapping of MscL in vivo is enhanced by osmotic shock, as shown in Figure 3-8 where the average dimer fo rmation of single mutants A 110C-A114C display a 20 fold increase in intensity upon osmotic shock. Sin ce the cytoplasm contains many reducing pathways using thioredoxin and glutathione/glu taredoxin to maintain reduced thiols, preserving a disulfide bridge in this region is a challenge ( 117, 118 ). In every case measured thus far, it appears that the amount of disulfide bridging increas es subsequent to osmotic shock ( 94). Presumably during the osmotic shock some of the reducing agents including thioredoxin ar e released through MscL ( 54, 55, 94). Therefore, osmotic shock intensifies the dimer formation signal indicating an interaction of residues either by them dynamically passing each other or by them remaining in close enough proximity to bridge during one of the conformational states.


85 Figure 3-8. Percentage of Di merization for single mutants A 110C, A111C, A112C, and A114C (n=12). A) Avg. Dimer formation under os motic shock. B) Avg. Dimer formation under normal conditions. According to our biochemical assays, all cysteines in this study crosslinked at some level. Single mutants formed dimers while double mutants formed dimers, trimers, tetramers, and even some pentamers as shown in Figure 3-9. Monome r bands were observed with all of the mutants indicating not all of the cysteines crosslinked. This was true for both shocked and non-shocked mutants. Reversibility of the disulfide bridging was al so conducted to ensure that the visualized bands were a direct result of cy steine crosslinking. Dimeriza tion was reversed when 5% of reducing agent -mercaptoethanol was added to the sample buffer as shown in Figure 3-10;


86 where only monomer bands were observed. To increase crosslinking or enhance the probability of disulfide bridge formation, the cells were also exposed to an oxidizing agent. Figure 3-9. Disulfide crosslinking in singl e and double mutants iden tified using western blotting. A) Single cysteine substitution forming disulfide bridges under both mock shock and osmotic shock (S) conditions. B) Double cysteine substitutions forming crosslinking between multiple cysteine s. One monomer is about 17 kD. Figure 3-10: Monomers of single mutants and WT as a control exposed to reducing agent DTT. Both Shocked (S) and Nonshocked identified using western blotting


87 Both shocked and nonshock liquid media containing 1% H2O2 reared bands with higher intensity than without the oxidizing agents as shown in Figure 3-11. Dimerization in vivo for all of the double mutants exposed 1% H2O2 yielded higher intensity than under ambient conditions. They both form tetramers and pentamers under bot h ambient and oxidative conditions suggesting their existence with in 3.6-6.8 of thei r analogues in the -helical bundle. While this supports the idea that these residues are in close proximity it tells little of any mo vement or changes that occur in this area upon channel gating. The results from the triple mutant A110C, A111C, and A112C and double mutant L121C/L122C, under both shocked and oxidized c onditions displayed greater crosslinking, making these mutant candidates for the electroph ysiology studies. Therefore, inside-out patching was performed with th ese mutants under ambient, oxi dative, or reduced conditions. Figure 3-11: Double mutants E119C/V120C, L1 21C/L122C, L128C/L129C, and AAA exposed to 0.5 M NaCl-LB (1st row), water (2nd row), 0.5 M NaCl-LB and 1% H2O2 (3rd row), and water with 1% H2O2. Quantified using western blotting. Electrophysiology was preformed in Dr. Blounts laboratory by Drs. Irene Iscla and LiMin Yang. A tetra-mutant of L121C/L122C a nd L128C/L129C was generated and inside-out patch clamped in PB104 cells. During these satu ration experiments, the change in channel


88 activity and conductance were examined in the pr esent of both oxidized a nd reduced conditions. Results from this study indicated there was not a change in the num ber of active channels or their conductance when the tetra-mutant was placed in either oxidative or reduced conditions (results not shown). This suggests that these residues are buried inside th e S3-bundle and that they are unlikely to move upon gating. Electrophysiology of the AAA-MscL included exposing this mutation to buffer solutions containing reducing agent 1 mM dithio threitol (DTT) or oxidizing agent H2O2. In these studies, the single channel ac tivity of this mutant cha nges in consistent ways when exposed to both the reducing and oxidizing agents (res ults not shown). Briefly, oxidizing reagents decreased the conductance of the channel, while reducing agents increased it. However, the triple mutant enables disulfide bridging at multiple positions within the complex and in numerous possible combinations. Therefore, we were unable to determine which sets of cysteinses were crosslinking therein any given f unctional channel. Thus, we next attempted to decrease the conductance of the channel by gene rating Histidine mutations, which can coordinate metal ions. Both A110C and A112C showed a 20 fold increase in dimerization upon osmotic downshock, therefore these two mutants were converted to A110H and A112H using Quickchange site-directed mutation, and sphereoplasts were generated in the same manner as stated above. By mutating them individually to Histidin e, and coordinating heavy metals, we should be able to constrain the movement of this propos ed filtering domain and thus the amount of osmotolytes translocating the pore of the channel, thereby decreasing the channels conductance. Previous studies with wild type MscL have shown no functional changes upon heavy metal treatment ( 110). In contrast, when 1-2 mM of ZnCl2 was added to the bath solution of native


89 bacterial membranes expressing A110H or A112H a decrease in conductance and a change in the single channel kinetics were observed, as shown in Figure 3-12. Figure 3-12. Single channel traces of A110H and A112H when exposed to regular patch clamping buffer, and a patch solution of ZnCl2. Point amplitude histograms of the single channel activity of A110H and A 112H displaying an average current of roughly 80 pA with regular patch clamp buffer and a decrease to 40 pA when exposed to ZnCl2. The current for both mutants was roughly 80 pA before the exposure to ZnCl2 and a holding voltage of -20 mV. Once they were exposed to the heavy metal, the conductance decreased by half from ~4 nS to ~ 2 nS and the channels stayed open longer. Once the ZnCl2 solution was rinsed out, the conductance as well as the activity resumed normal activity. Hence, we have developed an MscL channel in which the conductance can be reversibly decreased by heavy metals.


90 Discussion Despite the fact th at the C-terminal region of MscL was resolved from the crystal structure, its physiological role is unclear Molecular Dynamic stimulations and electron paramagnetic resonance, EPR studies have both supported the notion that part of the C-terminal region does form a -helical bundle at physiological pH (7.0), but the predicted structur e is quite different from that shown in the crystal st ructure. The state of the bundle during the gating process is also controversial, some evidence suggests it remains as a bundle and plays the role of a molecular size filter allowing certain osmolytes to transverse the channels pore (51, 53). On the other hand, each of the previous studies have had their fl aws, and some studies have demonstrated that quite large solutes pass through MscL. This l eads us to question the importance of the Cterminal. Does any conformational change occur in this region? Does it affect the overall function of the channel, and if so, mutations to which area (linker or S3-bundle) will enable a decrease in conductance? Here, in an attempt to better understand the dynamics of the C-terminal region, we probed the proximity of residues A110-A112, A114, E119-L122, L128, and L129 located in the Cterminal region. The results from this study indi cate first, that none of the mutations affect strong phenotypes. This is consistent with the observation that most GOF phenotypes are found along the pore of the channel. For instance, two well known GOF mutation sites are in positions G22 and G26 located are near the constr iction point of the closed channel ( 3). Substitutions with hydrophilic residues to position G22 yields severe GOF mutants, as in the same manner a substitution with Histidine to residue G 26 also yields an ex treme GOF mutant ( 3). The only mutant that did not significantly complement the LOF phenotype was the double mutant L121C/L122C. This mutant, according to Molecular Simulation modeling, spin-labeling, and cysteine substitution, is located in the S3-bundle ( 51, 53). The reasons why this mutant is a


91 LOF mutant are not clear. Howeve r, its location in the upper part of the 3-bundle (refer to Figure 3-4) may affect the stretching of the linker region during opening of the pore if the S3-bundle remains together. If the S3-bundle opens and folds into the membrane as shown in early molecular modeling studies, then the crosslinking that occurs during osmotic downshock at position L121C/L122C may affect the mobility of th e cytoplasmic region. One might anticipate that this will cause MscL to respond to higher tension, creating a loss of function mutant. Mutations at E119C/V120C, which are also expect ed to exist at the op ening of the bundle region but not expected to face each ot her, would not create such a LO F mutant because, as expected, they were not as efficient at forming disulfide bridges and multimers, so presumably had more functional channels. In fact, in most of the experiments, dimers rarely formed with this double mutant. Results from this study secondly supported the concept that the S3-bundle remains in the cytoplasm during gating as shown in Figur e 3-13, right. Mutants L121C/L122C and L128C/L129C were previously generated and th eir ability to form disulfide bridges was observed in a previous study by Sukharev in vitro using biochemical assays and patch-clamping In his study the biochemical a ssays were conducted in ambi ent, reduced, and oxidative conditions, while all of the electrophysiology was only performed in ambient conditions ( 51). He assumed that because crosslinking was observed in both of the oxidative and ambient conditions biochemically, then the bundle remained close during gating and the activity would be the same. In our study, the electrophysiology was taken one step fa rther using saturation experiments. During these experiments the ch ange in channel activ ity and conductance are examined when exposing the cystiene mutations to oxidized and reduced conditions (results not shown). A tetra-mutant of L121C/L122C and L128C/L129C was generated and inside-out patch


92 clamping was formed. Results from this study illustrated there was not a change in the number of active channels or their conductance when the tetra-mutant was placed in either oxidative or reduced conditions, suggesting that these residues are buried inside the S3-bundle and that they are unlikely to move upon gating. With this knowledge, we predicte d that we would be able to generate a lesser-conducting channel by c onstraining the linker between the second transmembrane domain and the S3 bundle. Residues A110, A111, A112, and A114, which are farther up the C-terminal region, were characterized for the first time in this study. Di merization also occurred with these residues. The highest percentage of dimerization occurr ed with A110C and A112C under osmotic shock conditions. These residues are pr edicted to reside just above the S3-bundle in the linker region as shown in Figure 3-13. As they are likely dynami c, cytoplasic, and as we predicted they would come in close enough proximity to crosslink. We also predict that di sulfide bridging would occur due to the dynamics of the region. This supposition was supported by the generation and study of a triple mutant consisting of re sidues A110C/A111C/A112C, (AAA). The disulfide bridging studies of this mutant al so resulted in dimers, trimers, tetramers, and pentamers, and the electrophysiological studies of th is mutant under ambient conditions showed a decrease in the conductance of the channel (resul ts not shown). When exposed to a reducing environment with 5% of ME, the conductance returned back to a normal full-conducting state, implying the disulfide bridging links two or more of the subun its together and changes the characteristics of the protein by decreasing the conductance. These findings lead us to generate additional mutations within this region: site specific single His tidine mutations at residues A110 and A112. We coordinated heavy metals with the his tidines and observed a decrease in conductance by 50% (refer to results in Figure 3-12). Revers ibility was also demonstrated; when the heavy


93 metals were removed from the system, the c onductance normalized. Future research should definitely include adding this molecularly engineered Histidine channel to the tethered lipid bilayer device. If in the future the sensitivity of the channel could be increased to the femtomolar range of detection, this would be the first generation of modulated MscL biosensors for the detection of heavy metals. Figure 3-13. Closed and opened state of Msc L with mutations generated in this study highlighted. Conclusion The aims of this study were 2 fold. First, the objective was to determ ine if the pentam eric -helical bundle within the C-terminal of MscL ch anges conformation from a closed state to an open state during osmotic shock. The second goal wa s to determine what area of the C-terminal region would enable manipulation of single chan nel activity and decrease channel conductance


94 with information obtained from the 1st objective. This was done in vivo using a classical scanning biochemical assay. Results from this study supported the newer molecular model of MscL in the protein data base (pdb), which depicts the hydrophobic residues buried in the bundle and the hydrophilic residues exposed on the outside Our studies strongly suggest that the S3bundle remains closed during gating An even more exciting discovery obtained from this study was the generation of two new mutants that coordinate heavy metals and decrease the conductance of the protein. Moreover, this react ion is reversible when the heavy metals are removed the channel activity returns to normal. Future experiments will include ways of isolat ing and purifying these mutants in order to study them in vitro within liposomes and in a tethered lipid bilayer membrane. Currently, purification of MscL is done using a nickel affinity column that binds to a 6-His-tag placed on the C-terminal of MscL. Possible cleavage of the His-tag could be done once the protein is eluted during purification to ensure the metals are interacting specifically with Histidine mutation and not the His-tag. With an increase in sensitive for the heavy metals, a future experiment could investigate the Histidine mutant as an actual biosensor within the device. Short-term projects will include looking for reve rsible decreases in conduc tance of the Histidines mutants when incorporated in th e tethered lipid bilayer.


95 CHAPTER 4 TESTING THE ELECTRICAL STABLITY OF DIFFERENT LIPID BILAYER COMPOSITIONS OF DPhPC/DPhPE Introduction Planar lipid bilayer systems we re first introduced in 1966 as an electrochemical transducer for detecting antigen-antibody and enzyme-substrate reactions (119). These systems are experimentally engineered over small apertures, between 1 m-1mm in diameter, creating a thin barrier separating two aqueous solutions. Since then, these artificial bilayer lipid membranes have been used as novel biological sensing system s incorporating isolated protein receptors that bind specific analytes. Some recent studies ha ve examined the membrane permeability, while others have exclusively introduced active transporter proteins, allowing for control over ion permeability ( 8, 70, 120). Planar systems offered the advantage of isolation, over traditional studies of ion channels in their native environment. Scientists are able to select th e bilayers composition, control the thickness of the membrane, examine the effects of different lipid architectures, and vary the surrounding solutions of ion channels within planar BLMs ( 6, 8). One disadvantage of using planar systems is the lack of stability associated with establishing the bilayer within either a pore or a hole. For this reason, researchers turned to integrating planar BLMs with solid supports ( 7, 11, 62, 65, 72, 73, 102, 121). Despite this disadvantage and th e frustrating nature of the freely suspended planar system, it is still used to exam ine different lipid compositions, and they provide enough stability to study small i on channels like Gramicidin ( 77, 97, 122, 123). In this study the tip-dip planar bilayer me thod (explained in chapter 2) was used to determine the best lipid ratio be tween DPhPC and DPhPE. This planar linear method establishes the bilayer at the tip of the mi croelectrode and measures the is olated channel activity, once the proper resistance is obtained, using an Ag/Cl electr ode. The lipid ratio that displayed the longest


96 lifetimes, low current leakage, and remained stab le under high applied potentials served as the lipid environment to test Gramicidin as a model channel. An important aspect of forming a stable bi layer is to reduce current leakage and hole formation which relates to random transitions in cu rrent amplitude. If a change in current occurs without the presence of an ion channel, the bilayer is said to suffer from current leakage. Mechanical stress, thermal stress, and electrical stress can all cause conformational changes in the bilayer producing these meta-stable pores Holes formed from electrical stress or electroporation are of particular interest in this research. Figure 4-1. A reversible hole in the bilayer formed from el ectroporation with an applied potential of -138 mV, and a current transition of more then -200 pA. These hydrophilic e-pores are fo rmed from an applied electric field and they maintain lifetimes on the order of several milliseconds. Th is affects the stability of the bilayer and can lead to both reversible and i rreversible breakdown if the pore is of extreme magnitude as shown in the Figure 4-1. In addition to irreversible breakdown, electropora tion can pose a problem to this research. Small pores with smaller conductance can also form in the bilayer when low potentials are applied to the system. This response can be mistaken for single ion channel activity if the


97 current leakage is uniform as shown the Figur e 4-2 when 80 mV are applied to a bilayer composed of 100% DPhPC. Figure 4-2. Mini-electroporati on of a 100% DPhPC bilayer at 80 mV. Pores have a uniform current flow of 10 pA. Other artifacts that contributed to experimental noise include current jumps, background current ramps, and some electrical noise. The sensitivity of the recording system allows it to detect all current changes regardless of the source. When possibl e, quality data from trials with current artifacts are included after background fi ltering, while other runs are disregarded. Therefore, forming a stable bilaye r with a high electrical resistance is vital to the tip-dip method. Two variables that could decrease electroporation, electrical, and backgrou nd noise in the system are the tip diameter and the gigaseals stability. Methods and Materials DPhPC and DPhPE lipids were mi xed in the following molar percentages: 50PC/50PE, 100PC, 100PE, 70PC/30PE and 30PC/70PE. The powde red form of each lipid was dissolved in Chloroform and stored at -20C. This method was explained in detail in Chap ter 2. Briefly, when performing a tip-dip recording, a silver wire was coated in a commer cial bleach solution for 30 minutes to create a silver/silver chloride Ag/Ag/Cl electrode. Ex cess bleach on the surface of the wire was rinsed


98 away with distilled water followed by washes with the buffer solution KCl 250mM, CaCl2 1mM, and Morpholinopropanesulfonic acid, MOPS 5 mM. A Teflon we ll plate was cleaned using distilled water (pH 7.4) and with ethanol. After th e well plate dried, we used disposable pipettes to add the MOPS buffer solution and inserted the ground electrode into one of the wells within the Teflon well plate We then added the lipid mi xture to the surface of the solution. The lipids formed a monolayer on the surface of the buffe r within 10 mins. The micropipette tip was prepared by a Micropipette Puller with a heated f ilament. Borosilicate capillaries were used to create the microelectrode tip and were purchas ed from Warner Instruments Inc., were 7.5 cm long and 1.5 mm wide. One capil lary created two tips because the filament heated the glass vertically in the center resulting in two 3.25 cm long micropipette tips. The tips diameter should be in the range of 0.50-5.00 m ( 78). This diameter range enables the tip to produce an effective gigaseal with the bilayer support. The gigaseal in-turn ensures adequate electrical resistance for the patch, which theoretically prevents current leak age. For these experiments, the Micropipette Puller is set to produce tips with a diameter of 1 m to ensure sensitivity. Over many trials, the following anecdotal observations are made: Freshly pulled tips (within 5 minutes ) effectively formed bilayers Tips covered for days, not exposed to dust in the air, could produce stable bilayers Used tips did not reliably yield gigaseals If the tip touched any solid surface before the experiment, a gigaseal could not be established If air bubbles near the tip resulted from filling the microelectrode, gigaseals are difficult to establish Scanning Electron Microscopy, SEM images of the microelectrodes tip in Figure 4-3 enabled visualization of the obser vations stated above. The first image of the tip indicates the inner diameter is less then 1.00 m across while the second and th ird images show the outer diameter as roughly 1.50 m across.


99 Figure 4-3. SEM images of the borosilicate pipett e tips. A) Freshly pu lled tip with an inner diameter is less then 1.10 m across. B) A tip that was covered for three days the diameter is still less then 1.50 m range. C) A tip after it has been dipped in a bilayer solution of DPhPC prior to SEM analysis. Depending on the patch clamp technique of choi ce, the isolated patch can be between 110,000 m2 and the unitary conductance of ion channels is in the order of several picosiemens (pS). Therefore, the resistance of the bilayer is vital to all patch clamp experiments, especially the tip-dip method. It ensures adequate electrica l isolation of the membra ne patch, a reduction in the current noise recording, and th eoretically prevents current leakage. For single ion channel studies, the tip diameter s hould ideally be between 1-5 m2 which allows the resistance within the Giga-ohm range. This is known as a gigas eal. The experimental setup, the amount of electrical noise in th e environment, the morphology of the microelectrode ti p, and the bilayer support, all affect the establishment of a gigaseal A vibration absorbing base and a copper mesh Faraday cage were utilized to limit electrica l noise, while a plastic cover protected the experimental-setup from dust and debris. The tip diameter as discussed above was kept between

PAGE 100

100 1-1.5 m which helped optimize obt aining the proper resistance. Compared with other planar bilayer methods, the tip-dip method provides a greater probability of obtaining gigaseals ranging from 1.210 G ( 124). We observed gigaseals from 1.20 G to 20 G which are comparable with literature (124). Results of Lipid Ratios The stability of five different lipid ratios of DPhPC and DPhPE (which will be addressed as either PC or PE for the duration of this ch apter) were studied to determine which one would be integrated on the device. The ratios are as followed: 100% PC, 100% PE, 50PC/50PE, 70PC/30PE, and 30PC/70PE. A typical experiment begins with obtaining a gigaseal and applying potentials between 150 mV. Pure PC Planar Bilayers PC planar lipid bilayers were analyz ed first. Gigaseal formation with this lipid composition occurred 56% (n=22) of the time and the bilayer remained stable between potentials 20-80 mV. Pore formation, however occurred at thre shold potentials of 80 mV or greater. The pore formation was heterogeneous as shown in Figure 4-4 where it opened and closed for 1.5 seconds. Lifetimes of the bilayers were usually between 2-10 minutes and once electroporation occurred, it continued at potentials as high as 125 mV and as low as 40 mV. In addition, the baseline was unstable and would drift between 1-3 pA during experiments. Quantitatively short spiky events also appear ed lasting less than 2 seconds throughout the lifetime of the bilayer. We concluded pores produced were independent of th e applied potential, randomly forming large and small pores with conductance of 0.4 nS to 10 pS.

PAGE 101

101 Figure 4-4. Electroporation of 100% PC at 120mV with a bilayer lifetime of 56 minutes. The time spent at120mV was 2mins. The cu rrent changes range between 10-80pA. Pure PE Planar Bilayers The smaller PE head group in 100% PE planar lipid bilayers easily fo rm ed strong gigaseals ranging from 5-10 G 75% of the time (n=15). Threshold potentials were higher than that of 100% PC at 200 mV with electropor ation observed 36% of the time (n =6). After 1-2 minutes of applying potentials, pore formation was consistent and more uniformed than pores formed with 100% PC planar bilayers. The bilayers baseline remained stable havi ng a high recovery rate, with the uniformed pores conducting in the range of 20-40 pS. Th ere was also more electroporation activity at negative potentials for this lip id composition although the buffer is symmetrical. Bilayer lifetimes were as long as 30 minutes to 1 hour be fore electrical breakdow n occurred whereby the seal strength decreased from the giga-ohm to either mega or kilo-ohm range.

PAGE 102

102 Figure 4-5. Electroporation of 100% PE planar lipid bilayer with potentials applied from 100mV-180 mV. 50PC/50PE Planar Bilayers Planar bilayers composed of 50PC/50PE were exceptionally unstable. Form ing a gigaseal with this ratio was rare and the baseline was unstable with very lo w applied potential (3-10 mV). Electroporation that did occur was closely followe d by a disruption of the bilayer. Threshold potentials for electroporation were as low as 40 mV. The bila yers were short lived, with lifetimes less then a minute. Figure 4-6 shows in stability of the bilayer at 40 mV with current fluctuations between 5 pA and zero. Figure 4-6. Electroporation occurring with a 50PC/50P E bilayer within the 1st 25 sec.

PAGE 103

103 Admittedly, the instability with this lipid ra tio is very unclear. We do however; believe both the high hydration state of PC and possibl e raft formation within this system could contribute to the instability of the 50/50 lipid composition. This theory will be examined in detail during the discussion. 70PC/30PE and 30PC/70PE Planar Bilayers Results from both 70PC/30PE and 30PC/70PE showed ma ny similarities. Gigaseals between 2-10 G formed 75% of the time for both ratio s (n=19). Thres hold potentials for electroporation were above 120 mV for both of th e mixtures. When electroporation occurred the activity took on the characteristics of the majority lipid present in the mixture. For instance, epores formed in 70/30 PC/PE were random with current changes ranging from 0 pA-5 pA, and they possessed a drifting baseline. E-pores formed with the 30P C/70PE lipid composition remained uniformed and the current decreased wi th decreasing applied potentials giving the pores very stable unitary conductance. Bilayer lifetimes were also long for both of the lipid mixtures, lasting on the order of several minutes to an hour. The lipid mixture 70PC/30PE was a bit more stable and suitable for our studies in comparison to the lipid composition 30PC/70PE. If the stock solu tion of 30PE/70PC was left at room temperature, a white film would precip itate out of this solu tion and electroporation occurring with this mixture was very characteri stic of single channel activity. Therefore we investigated the 70PC/30PE lipid mixture furthe r. We exposed this lipid system to ramp voltages and the episodic protocol. Ramp voltages as shown in Figure 4-7; automatically applied potentials from zero to 200 mV, then fr om 200 mV to -200 mV w ithin 3 seconds, and we studied Gramicidin in this lipid composition. Results from these experiments indicated th at 70PC/30PE lipid bilayers could survive rapid potential changes applied to the system. Sim ilar to the ramp voltages, the episodic protocol

PAGE 104

104 was used to determine the conduc tance of the bilayer. During these experiments, automated potentials from negative 200mV to positive 200 mV were applied to the system for several seconds as shown in Figure 4-8. Figure 4-7. Ramp voltages applied to the 70PC/30PE lipid mixture. Potentials from zero to 200 mV, then from 200 mV to -200 mV were appl ied the lipid system within 3 seconds. The bilayer followed Ohms law and the curren t did increase when increased potentials were applied; and decreased when lower potentials were applied. Both of the ramp and episodic protocol apply various vo ltages within seconds. Figure 4-8. Episodic trace of th e current in pA obtai ned lipid bilayer co mposed of 70PC/30PE when 0 to both +200 mV and -200 mV were applied simultaneously.

PAGE 105

105 Figure 4-9. An 18 minute trace of 70PC/30PE lipid mixture. A) Raw-data of the current to voltage relationship over time when 50-200 mV was applied for 10 minutes followed by -10 mV to -200 mV and back to zero. B) A trace of the sample obtained at 200 mV C) A portion of the sample at -20 mV. Using the gap-free protocol enabled the study of the stability of the bilayer over long periods of time. Bilayers with this lipid composition were long-lived with electroporation occurring after minutes of the b ilayer existing without any change in currents. As shown in Figure 4-9, this lipid ratio did however experience some reversible bilayer breakdown within a 16 minute trace. Single Channel Results of Gramicidin in 70PC/30PE bilayer Incorporating Gram icidin A as a model channel in the 70PC/30PE lipid composition resulted in single channel activity comparable to literature. As shown in Figure 4-10 the single

PAGE 106

106 channel conductance of Gramicidin is 67 pS whic h is in agreement with literature values of 6070 pS (n=19) ( 19, 123). Figure 4-10: Single channel conductance of Gr amicidin A is around 67 pS when 60 mV is applied. Figure 4-11. Plots of a 1.5 second run at nega tive 100 mV with and without Gramicidin. A) 70PC/30PE pure lipids B) 70PC/30PE with Gramicidin. Plots of the change in current between the opened and closed states (I), against applied potential (V). All points on the IV curve were the average of n 3 different experiments. Standard error bars were plotted; if bars are not r eadily visible, then the SE was approximately the height of the dot. C) 7PC/30PE lipid mi xture. D) 70PC/30PE with Gramicidin A. The single channel kinetics was slow with channel activity occurring for at least one minute continuously. Open dwell times ranged from 1 s10 ms. I-V curves as shown in Figure 4-11 did allow us to distinguish between el ectroporation and singl e channel activity of

PAGE 107

107 Gramicidin. The linear fit and the low error bars proved that current fluctu ations were linked to Gramicidin opening and closing. We also examined the best buffer concentra tion for Gramicidin in the 70/30 lipid ratio. Gramicidin is known to be selective to small mo novalent cations like potassium. Determining an ideal potassium chloride concentration will allow for enough ions to create a measurable current, yet avoid dehydration of the device. We studi ed 3 different KCl c oncentrations: 125 mM, 250 mM, and 350 mM. Figure 4-12. Plots IV curves for Gramicidin A fo r various KCl concentrations. All points on the IV curve were the average of n 3 different experiments. Standard error bars were plotted; if bars are not readily visible, then the SE was approximately the height of the dot. A.) IV curve displaying a conductance of 30 pS when 125 mM KCl is present in the buffer. B.) IV curve displaying a conductance of 57.0 pS with 250 mM of KCl. C.) IV curve displaying a conductance of 49.9 pS when 350 mM of KCl is present in the buffer solution. The concentration of Morpholinop ropanesulfonic acid (MOPs), CaCl2, at pH of 7.4 was kept constant. Results showed that 125 mM gave rise to a lo w conductance. There was little

PAGE 108

108 difference between results from both 250 mM a nd 350 mM, therefore we decided to use 250 mM KCl as shown in Figure 4-12. Discussion When considering the above results, it is important to understand the current theories surrounding the developm ent of electroporation. First, one must understand how the hydration state affects the stability of the bilayer and the behavior of water molecules under an applied electric field. Second, the technique and the li pid composition when analyzing the results must also be considered. In addition on e must also consider the system at hand. This is the first time diphytanoyl lipids have been st udied using the tip-dip method therefore electroporation could possibly be a characteristic of this combined system. The Effect of Hydration State on the St ability of Bilayers and Electroporation The stability of the bilayer is greatly affect ed by the hydration state of the me mbrane. In 2004, Dr. Peter Tieleman used molecular mode ling to investigate th e process in which electroporation occurs and how its dependence on the hydration st ate of the bilayer membrane ( 125). He found that water molecules surround thr ee parts of the lipid bilayer: the carbonyl chains, the phosphate back bone, and the lipi d headgroup. The amount of water molecules surrounding each region is separated into three differe nt water layers adding to the fluidity of the bilayer. Each water layer is depended on the charge and hydrophobicity of each region of the bilayer. The first water layer has the least amount of water mol ecules and is localized around the carbonyls from the fatty acid chains in the hydrophobic part of the membrane ( 125-127 ). Depending on the conformation or structure of the b ilayer (cis or trans) ro ughly five to six water molecules exist in this first layer adding to the hydration of the bilayer. The second water layer exists around the phosphate backbone of th e lipid at the water/lipid interface ( 127). The amount

PAGE 109

109 of water molecules in this region changes, depending on the properties of the headgroup which forms the third water layer of the bilayer ( 127). The size, polarity, charge, and hydrophobicity of the headgroup determine the existence of the 3rd water layer and controls the exposure of the phosphate groups to the bulk water phase surrounding the bilayer ( 127, 128). For example, the choline group of Phosphatidylcholine PC headgroup is extremely hydrophobic due to the thr ee methylene groups bonded to the nitrogen ( 127, 129). The PC headgroup is ther efore pulled inwards towards the membrane. This exposes the phosphate group and its charges to the aque ous phase allowing 18-20 water molecules to bind per lipid (127). The hydrophilic amine group present on Phosphatidylethanolamine (PE) forms a compact layer above the carbonyl hinderi ng mobility of the phosphate backbone and its exposure to water ( 127 ). They link via H-bonding to ad jacent phosphates producing a compact, rigid head group network at the bilayer surface with roughly 4-6 water molecules bonded per lipid (127, 130). The water molecules along with the natural diel ectric potential of th e bilayer creates an overall bilayer dipole potential th at becomes polarized when an external potential is applied ( 127, 131). The two parts contributing to the bilayer dipole potential are the surface potential and the membrane potential. When an external elec tric field is applied to the bilayer in a time dependent manner charged groups and dipoles re-orientate (127). Using molecular dynamic simulations to atomically map this process Tieleman et. al. found that pore formation during electroporation is driven by local electric field gradients at the water/lipid interface when an external potential is applied ( 125). Water molecules move into the field gradient increasing the probabil ity of water defects penetrating into the bilayer ( 125, 128 ). These defects are the mol ecular basis of electroporation

PAGE 110

110 which occurs in the pico-nanosecond time range. Water molecules once inside the bilayer, Hbond with water molecules already present forming what is now termed water fingers that span the bilayer forming water pipes ( 125, 126, 128 ). At this point a deformation of the headgroup occurs whereby it curves into the bilayer to accommodate the water fingers. Both the lipid chains and the headgroups are now exposed to water. Water pipes become fully hydrophilic pores when the headgroups completely align the surface of the pores which results in a change in capacitance or conductance that is observed as a current peak ( 125, 128). When applying these theories to the results obta ined in this study, certa in characteristics of the different lipid mixtures can be better unders tood. For instance, because of its high hydration states, bilayers where PC is the most abundant li pid present could theoreti cally be comprised of an unstable baseline and random pore formation. Gauger et al. compared hydration states of both PC and PE with diphytanoyl-tail, to palmit oyl chains groups using FTIR and Karl-Fischer water titration ( 90). Results suggested that DPhPC has th e highest water uptake in comparison to other phosphocholine (DOPC, POPC, OPPC, DPPC). The ability of this lipid to retain water molecules could explain the in stability and frequent occurre nce of electroporation at low potentials. The study also found that the hydration state of DPhPE was higher in comparison to DOPE, and POPE ( 90). Methyl groups within the diphytan oyl-tail add to the random disordering and water uptake during the lamellar bilayer phase ( 90). This could lead to an increase in water fingers when a high electric field is applied. The uniformity and potential thresholds over 120 mV for bilayers where DPhPE is the most abundant lipid could be explained by consid ering the lower hydration state and the size and hydrophobicity of the headgroup. Because the am ines present on the headgroup H-bond with neighboring phosphates forming a tight compact h eadgroup orientation, more energy i.e. higher

PAGE 111

111 potential thresholds would be require to move more water in the bilayer. In addition, the dipole alignment by H-bonding of the water molecules in th e bilayer is going to be stronger because of the higher energy and the hydrophobic envi ronment of the inner bilayer. Experiments from lipid composition 50/50 PC /PE faced both a high hydration state, and possible raft formation or phase domains. This formation occurs when different lipids do not fully integrate; instead, they form isolated islands leading to th e instability of the bilayer under an applied electric field. They are ideally tightly packed and separated. In biological membranes these domains usually consist of high concentrations of sphingolipids and cholesterol. Computational simulations includ ing Monte Carlo and Coarse-grained Marinkmodels have been employed in modeling these lipid raft or coexisting phases formed by mixed bilayers. Wong et al. used the coarse-gained Marrink-m odel to study the mixtures of dipalmitoyl phosphatidylcholine and ethanolamine DPPC and DPPE and found that DPPE lipids tend to move to the center of the bilayer ( 132). They also found that areas enriched with DPPC bilayers were slightly thicker but co uld not fully explain this phenomenon ( 132). This theory has not fully been proven and is currently under debate. It is however, a possi ble explanation for the unstable bilayers that were formed with 50DPhPC/ 50DPhPE in this research. To date, we have no indication of lipid raft formation or phase separation occurring with the 70PC/30PE lipid mixture. Tip-dip and Diphytanoyl Lipid System The tip-dip me thod may not be the best tec hnique for studying diphytanoyl lipid mixtures due to their large hydration states which may ma ke them more quantifiable to electroporation due to the bilayer forming horizontality at the tip of the pipette. In othe r planar bilayer methods the bilayer is either painted on a small hole sepa rating two buffers filled compartments or formed on a moveable aperture at the air/water interface. Diphytanoyl lipids have been studied using the

PAGE 112

112 painting method where the electroporation threshold potentials were around 250 mV( 6 ). The tipdip method which is both inexpensive, and us er friendly may have lower electroporation thresholds. The details of this phenomenon are not fully understood. However to test the effect of adding a membrane protein, which has been proven to increase the stability of bilayers, Gramicidin A was incorporated into the lipid bilayer. 70/30 DPhPC/DPhPE was chosen as the lipid en vironment because it is one of the most common lipid ratios used in literature, and stab le bilayers were formed with high threshold potentials. The electroporation events were not uniform and they did not follow Ohms law which makes them easy to distinguish from singl e ion channel activity. In addition, since we will be using vesicle fusion to create a bilayer on the device, 70PE/30PC would not form stable vesicles due to the high negative spontaneous cu rvature of the PE group. Spontaneous curvature (SC) relates the cross-se ctional areas of the h eadgroup to the acryl chai n moiety. Negative SC can occur if the lipid has long fatty acid chains and a small headgroup. PE is smaller than PC; therefore it has a higher negative SC value which might have caused it to precipitate out of solution at room temperature. Lastly, 70PC/30P E was chosen because a small amount of PE does increase the stability of the PC bilayer. Conclusion In an attemp t to establish the ideal diphyta noyl lipid environment for studying single ion channels in a tethered device, different lipid ratios were comp ared using the tip-dip method. Diphytanoyl lipids are a vital comp onent of this research due to their increased durability and stability when tethering to solid supports such as the gold surface on the device. In this novel combined system of incorporating diphytanoyl li pids of PC and PE with the tip-dip method, electroporation was unfortunately ob served and characterized. Resu lts indicated all of the lipidratios suffered from e-pore formation due to electrical breakdown. This was the first time that

PAGE 113

113 electroporation was reported at such low pot entials as 125 mV and 40 mV which may be characteristic of using the tip-dip method with diphtanoyl lipids. The phenomenal pore formation was random; however we were able to distinguish it from single ion channel conductance by plotting I-V curves. The lipid ratio of 70PC/30PE was chosen as the most stable lipid ratio and will be used as the synthetic lipid environment to study both Gramicidin and MscL on the device. Even-though the threshold potentials were around 125 mV, studying Gramicidin in this lipid ratio provided positive results with the conductance and dwell times of the channel being comparable to results from literature. This suggests that diphytanoyl li pids in a ratio of 70PC/30PE can form stable bilayers suitable for sing le channel measurements.

PAGE 114

114 CHAPTER 5 FUNCTIONAL STUDIES OF GRAMICIDIN AND MSCL WI THIN A TETHERED LIPID BILAYER MEMBRANE Introduction Mechanosensitiv e (MS) ion channels sense te nsion in the cell membrane causing a change in structural conformation that opens the channels pore. They ar e proposed to be of essential importance in physiological processes such as touch, hearing, balance and osmoregulation (21, 38, 133, 134). Escherichia coils mechanosensitive channel of la rge conductance (MscL) is one of the most studied MS proteins. It has a conductance of ~3 nS with several known intermediate states, and appears to play a crucial role in protecting bacterial cells from osmotic downshock ( 135). MscL is a homopentamer with two subunits that span the membrane twice connected by a perplasmic loop and the Nand C-termini reside in the cytoplasm. The first transmembrane domain forms the constriction point of the cl osed channel; the second transmembrane domain interacts with the lipid matrix. Msc L was first purified and cloned by Kung et al who were also able to retain its functionality in vitro ( 115, 136). E. colis MscL robust properties make it an attractive candidate for nanosensory applications. This protein has been isolated, purified, and reconstituted into several different liposomes, and remained functional (4, 10, 53, 111, 137-141). In vitro studies were done on mutants of this protein to study its conductance, lipid-protein inte ractions, structure, and single channel functionality (46, 51, 53, 130, 139, 142, 143). A number of mutants for E. colis MscL were generated that add sensor y adaptations for pH, heavy metals, light, and redox reactions. For instance, Kloda et al. studied the pH sensitivity of a cluster of residues located within the cytoplasmic region of MscL ( 4). They found that the RKK EE cluster of charged residues inherently affected the tension required to gate MscL. If substitutions where made with the opposite charged residue, this prot on sensor would have an alte red sensitivity to membrane

PAGE 115

115 tension ( 4). Iscla et al. designed metal binding sites using his tidine substitution to the pore lining of MscL, changing the channels sensitivity in the presence of heavy metals ( 110 ). Yoshimura et al. created a cysteine substitution with residue G22 and generated a channel with a greatly enhanced sensitivity to membrane stretching by adding charged sulfhydryl reagents within the channels pore (3). Iscla et al. recently utilized cysteine tra pping of single-cysteine mutations within the N-terminal region; these mutations cr eated a redox-dependent alteration of gating of MscL. Light-sensitive mutants of MscL were achieved with photo-sensitive chemicals tethered to the pore of the protein; Kocer et al. attached synthetic com pounds that underwent lightinduced charge separation to revers ibly open and close MscLs pore ( 111 ). Incorporating any of the above mutants within the tethered lipid e nvironment would have potential as a nanoscaled electrical sensor. As a first st ep, the single channel characteriza tion of the wild type (WT) MscL must be demonstrated and characterized within th e tethered bilayer system before incorporating any of the sensory-designer channels. Here we present results from a technique that utilized tethered bilayer lipid membranes (tBLM) to measure single channel activity of the MscL. The tBLM is part of an engineered microelectronic array chip, illustrated in Figure 5-1, which can ultimately serve as a stabilizing environment for nanosensor-ion channels. The te thers work as stable anchors and serve as spacers to the underlying gold surface which forms an ionic reservoir and allows an ionic flow through the channel. The presence of the tethers in creases the viscosity which results in lowered ion mobility in the reservoir when compared to the bulk solution. Incorporating MscL into the tBLM offe rs many advantages over traditional proteoliposome formation for in vitro functional analysis. tBLM provide mechanical mobility, functional stability, and a non-denaturing environment for membrane protein ( 57). These bio-

PAGE 116

116 mimetic supports enable the functional characteri zation of single ion ch annel activity on a solid substrate. It also enables prot ein-protein and lipidprotein interactions by providing a membrane environment with minimal irreversible bila yer breakdown. Both analytical and physical techniques that require a solid support, such as surface plasmon resonance and Atomic Force Microscopy (AFM), can now be employed fo r characterization w ith the tBLM. Figure 5-1. Tethered bilayer membrane array. A) Optical microscope image of the probe pad and the tungsten electrode. B) Graphical representation of the engineered tethered bilayer membrane array. The lower left co rner shows the gold sensor pad covered by a tethered bilayer lipid membrane (tBLM) incorporating MscL ion channels. C) The molecular structures of the species used in the tBLM. Image designed by Dr. Henk Keizer and borrowed with permission. This approach is new and promises an exciti ng base of fundamental knowledge that can be associated with single ion channel studies. Most importantly, this technique does not require the

PAGE 117

117 tedious patching of blisters for in vitro analysis that is associated with patching reconstructed mechanosensitive channels in proteoliposomes ( 31, 38). Before reconstituting MscL into this system, we studied the single channel activity of Gramicidin as a model channel within the tBLM Gramicidin forms a transmembrane pore when the two 26 long segments connect at their Ntermini. Six H-bonds at the dimer junction stabilizes the pore in this conf ormation. This channel has a simple structure and has been well characterized in its natural environment, planar lipid bilayers, liposomes, tethered-bilayers and polymer cushion-lipid bilayers ( 17, 58, 76, 122 ). To characterize an ion channel in this new sy stem, specific single channel parameters were examined including the conductance, the stability of the bilayer on the microchip, and in the case of MscL, a method of providing en ough tension to the system to cause the channel to open. Materials and methods Materials The ma terials and methods are explained in-det ail in Chapter 2, and are briefly discussed as follows: a 7:3 molar ratio of the arch aea phospholipids, 1,2-diphytanoyl-sn-glycero-3phosphocholine (DPhPC) and 1,2diphytanoyl-sn-glycero-3-phosphoethanolamine (DPhPE), were used to form the bilayer; 2,3-di-O-phytan oyl-sn-glucero-1-tetraethy lene glycerol-D, and L-lipoic acid ester lipid (DPTL) were used as tethers (synthesized according to a literature procedure (81 )). Morpholinopropanesulfonic acid (M OPS), Potassium chloride, calcium chloride, and potassium hydroxide used to make the buffer solutions. Proteins Gramicidin A was comm ercially produced. This protein did not requ ire an isolation or purification; it was stored in methanol at -20C.

PAGE 118

118 MscL containing a Cterminal 6His-tag wa s expressed in the mscL-null strain PB104 using the pB10b expression vector. First, cells were grown to an OD of about 0.6 and induced with 1mM IPTG for 30 min at 37C. Membrane fractions were separated and resuspended in extraction buffer (50 mM Na2HPO4/Nah2PO4, pH 8.0, 300 mM NaCl, 20 mM imidazole plus 2% (v/v) Triton X-100) ( 96). Then the suspension was incuba ted with 500 ul of Ni-NTA agrose (Qiagen) for 30 min at room temperature. The matrix was loaded on a 10 cm column and washed with 10 ml extraction buffer with 50 mM imidazole and 1% Triton X-100. The protein was eluted in 1.5 ml extraction buffer with 200 mM imidazole and 0.2% Triton X-100. Lastly, the protein was analyzed by SD S-Page gel for purity. Device The gold device was fabr icated in the Microelect ronics Research Center at the University of Texas, Austin, in Dr. Ananth Dodabala purs laboratory by Dr Daniel Fine. The microelectrode arrays used contained 66 pixels per wafer. Each sensor pad formed by evaporation of 3 nm of Ti followed by 500 nm of a 60% Au / 40% Pd alloy on the silicon substrate. A 200 nm gold layer was deposited on top of the alloy; a polyimide resist was photolithographically applied to define the pad si ze and electrically isolate the pads from each other. The device was cleaned with a rinse of hexane, acetone, ethanol, and water, and then treated with UV/ozone for 10 minutes. Preparation of the Tethered Bilayer First, the device was submersed into a 0.3 m g/ml DPTL-ethanol solution and left overnight. The wafer was rinsed with ethanol and water. Th e lipid composition of 70PC/30PE (2 mg/ml) was mixed and the chloroform solvent was evaporated under vacuum and the lipids were hydrated in 1 ml of MOPs buffer (5.0 mM MOPS, 250.0 mM KCl and 0.1 mM CaCl2, titrated to pH=7.4 using KOH). The lipids were heated at 50 oC until we obtained a clear solution

PAGE 119

119 (ca 1 hour). After cooling to room temperatur e, the suspension was sonicated for 5 minutes, followed by filtration through a 0.45 m filter. The vesicle size was determined with a Dynamic Light scattering (DLS) detector from Brookhaven Instruments. Vesicle diameters of 150-500 nm were typically found. Finally, a drop of the vesicle solution was deposited on the gold DPTL treated sensor pad and left for >8 hours at 4oC, or at room temperature for 2 hours. This initiated vesicle fusion, i.e. formed the tethered bilaye r. Prior to the electr ophysiology examinations, a drop (5 l) of MOPs buffer solution was a pplied to the treated sensor pads. Protein Reconstitution Since Gramicidin (0.2 mg/m l) was made synthe tically the protein was directly mixed with the lipids to create proteioliposomes. The lipid solution (2 mg/ml in chloroform) containing the 70PC/30PE lipid ratio was mixed with Gramicidin (0.2 mg/ml in methanol) to obtain a final lipid-to-protein ratio of 1000:1. The vesicle preparation and tethered bilayer formation is explained above. MscL was obtained after purification and the product was concentrated to 0.9 mg/ml and gel filtration was used to remove the excess imidazole. MscL was reconstituted into the tethered bilayer with a protocol similar to that of Gramicidin or by the development of proteoliposomes in which detergent was extracted with biobeads. Th e details of this procedure are explained in Chapter 2; briefly, after the lipid vesicle are formed via sonication, 1% Triton X was added to loosen the bilayer. Purified MscL was slowly added in a 1:120 protei n to lipid ratio. The protein-lipid-detergent matrix was then heated in the water bath at 60C for 30 minutes. While the sample was heating, Biobeads (Bio-Rad) were washed 3X with the MOPs buffer used to create the vesicles. The proteoliposome solution was added to the bio-beads and incubated with light rotation for either 4 hours at room temperature or overnight at 4C. The proteoliposome solution was lastly removed from th e bio-beads and stored at 4C.

PAGE 120

120 Electrophysiology Recordings of single channel activity were obtained with two elec trodes. The active electrode was a conventional patch p ipette cont aining an Ag/AgCl electrode and filled with buffer solution. The reference electrode consisted of a gold coated tungst en electrode that was positioned onto the probe pad. Both electrodes were connected by the headstage to an Axon Patch-Clamp amplifier (Axopatch 200B. Molecular Devices Corporation, Union City CA). The signal was passed through a low-pass 5 kHz and di gitized at a sampling rate of 20 kHz using a Digidata 1322A from Molecular Devices Corporat ion. Afterward, a giga-seal was obtained and 2 l of the protein (Gramicidin 0.2 mg/ml or MscL 900 ug/ml) solution was added. Gramicidin channels formed sporadically as monomers within each portion of the bilayer dimerized. In order to gate MscL, we had to create tens ion in the bilayer. As given by the Maxwell stress tensor, an electric field applied over a lipid bilayer induce s a perpendicular stress in the membrane. This stress can be viewed as an increase in the membrane area and can cause changes in the curvature, i.e. flexoelectricit y. This phenomenon has been demonstrated in both lipid bilayers ( 144) and cell membranes ( 145). Because we have no way of monitoring pressure on the device, a voltage was applied to induce te nsion in the membrane. According to Needham and Hochmuth, the membrane stress ( ), as a function of the transmembrane potential (V), may be expressed as: = 0(d(2/de 2)V2 where is the dielectric constant of the surrounding medium (ca 60), 0 is the permittivity of free space constant, d is the bilayer thickness (~ 4 nm), and de is the dielectric thickness (~ 3) ( 146). Proteioliposomes of MscL are known to sense tension around 12 dynes/cm ( 10). This tension corresponds with an electric field of 300 mV according to Needham and Hochmuths equation. This threshold voltage was applied using a patch-clamp amplifier; and used to create enough stress in the membrane. The single ch annel signals were pa ssed through a low-pass 1

PAGE 121

121 kHz filter and digitized at a sampling rate of 50 kHz. Single channel analysis was conducted in the same manner as with traditional patch clamping techniques. Results In this pr esent work, the functional characteri zation of Gramicidin and MscL were studied using a tethered bilayer system. This system is one of seven different methods recently adapted to study membrane proteins on solid supports. Th e first steps in developing the tBLM, were to establish a monolayer on the golds surface, tether the bilayer via vesicle fusion, and test the resistance of the bilayer. The stable thiolipid system, DPTL, served as the monolayer component which tethered the bilayer to the golds surface. Previous experime nts were performed in our laboratory utilizing Atomic Force Microscopy (AFM) to characterize monolayer formation and vesicle fusion. This technique uses a semiconductor tip attached to a cantilever that traces the surface of a material ( 147). The tips deflection along the surface is meas ured by a laser and plotted as a function of tip position. These experiments were conducted in tapping mode with the semiconductors tip tapping across the sample surface. Results from this investigation are shown in Figure 5-2. It was proposed that the monolayer deposition occurred within a 24 hour period wherein DPTL vesicles first rupture a nd then form islands (147). The islands eventua lly swell through solvent diffusion to form the final ~4.7 nm monolayer system ( 147). Unilamellar vesicles of 70DPhPC/30DPhPE, with or without proteins, were added to the monolayer system. The small vesicle size enabled the bilayers to rupture rapidly as they fused with the DPTL monolayer in a six hour time span at room temperature ( 147). Dynamic light scattering, as explained in Chapte r 2, provided insight on the size distribution of the vesicles. The sizes of the LUV ranged between 100200 nm as shown in the appendix.

PAGE 122

122 Figure 5-2. AFM images of the monolayer depos ition and vesicle fusion process. The first image is of the gold substrate. The sec ond image is of the gold substrate with the monolayer self assembled on the surface and spanning roughly 4.7 nm long. The third image has both the tBlM and the monolayer deposited on the golds surface. Together the DPTL and the bilayer is r oughly 12 nm high. DPTL. Images adopted from reference (147) with permission from Henk Kasier and Brain Dorvel. After vesicles fused with the monolayer, we took precautions to avoid electroporation in the membrane since high voltages were applied. A stable giga-seal was always obtained before adding the protein; and no signals were detected before the addition of either one of the ion channels. The bilayer was considerably more stable when tethered to the gold surface in comparison to freely suspending it as in the tip -dip method. Out of 47 bilayers analyzed, 25 exhibited electroporation, while 22 of the bilayers were free of electroporation and were very stable over long periods of time when exposed to high voltages as shown in Figure 5-4. The electroporation threshold potential varied fr om 100-300 mV with the most electroporation observed at 300 mV. These findi ngs are more comparable with the literature (threshold potentials 200-400 mV) than the membrane rupture threshold potentials (120-200 mV) concluded in Chapter 4 where the tip-dip method was utilized with this lipid mol compositions ( 148, 149). The electroporation is seemingly induced by the a pplication of high potentials

PAGE 123

123 applied to a weakly formed bilayer. The pores formed ranged from extremely small to large stable pores that inhibit the bilayers ability to maintain the giga-ohm resistance necessary to observe single channel conductance. The e-pores that did form displayed variable conductance and none of the traces displayed consistence electroporation. Most of the time when it did occur, irreversible breakdown of the bilayer followed. Small current transitions in the range of 0.1-5.0 pA conducting 30-70pS were observed with this sy stem. On the other hand, we also observed bilayers that were extremely stable for over 10 minutes at 300 mV (data in the appendix) and several I-V curves were generated enabling the calculation of the resistance of the bilayers (data in appendix). Figure 5-4. A stable tethered bi layer form from vesicle fusion of 70PC/30PE LUVs. The bilayer was stable for ten minutes while being exposed to voltages ranging from 50-250 mV with a resistance of 2.1 G (calculated using Ohms Law) Results of Gramicidin Reconstituted into the tBLM The simp le helical structure of Gramicidin a llowed the protein to sustain high temperatures and dehydration overnight when reconstituting the pr otein into lipid vesicles of 70PC/30PE. The signals obtained from gramicidin were similar to those found in the literat ure as seen in Figure 55. They also correspond with the signals obtaine d from the tip-dip recordings (Ref. to Figure 5-

PAGE 124

124 5). Experiments and results in Figure 5-5 were published in a collaborati ve effort with Chenyu Zhu and Dr. Martin Andersson ( 77 ). In Figure 5-6, an I-V curve was plotted of the current to voltage relationship for Gramicidin within the tip-dip system and the tBLM device. When negative potentials were applied to the tBLM on the device there were decreases and a small plateau of the current demonstrat ing rectification of the protein (see appendix). Rectification occurs when there is any form of limitation that allows ions to flow through the channel in one direction better than another. Known mechanisms include Mg block and blocking the pore with part of the protein, whether it be a ball and ch ain at one end of the protein or allosteric constraints in the pore. In this case, we hypothesize; it is due to the small ionic reservoir and the slowed mobility, due to high viscos ity, of the ions in the reservoir of the tethered region relative to the large buffer reservoir above the bilayer. Figure 5-5. Single channel activit y of Gramicidin at an appl ied voltage of 60 mV with a conductance of 60 pS. The upper trance was recorded using the tip-dip method, and the lower trace was obtained on the devi ce. In both traces, one open channel corresponding to a current of roughly 4 pA was observed. Two channels were open within the tethered bilayer as indicated by the 2nd current change of roughly 4 pA. When negative voltages ar e applied the positive K+ ions try to rush from the limited ionic reservoir were the tethers and small size of this region alters the conductance observed in the system (see figure and discussion in appendix). Krishna et al. studied the interfacial capacitance

PAGE 125

125 of the reservoir region using the Stern model of ionic distribution, and concl uded that this region has properties similar to a dense hydrated gel with restricted ionic mobility ( 76, 83). Figure 5-6. I-V curves of Gramicidin in the tBLM on the device and within the tip-dip system. Negative potentials were a pplied to the tBLM on the device and current responded in a small plateau. This we configured was due in part to the ionic reservoir and the slowed mobility of the ions in this region.( 77). Gramicidin is known to pass a maximum conductance of 90 pS due to the constricted small size of the pore (18). Larger current transi tions were observed in some of the experiments (41% n=12) as shown in Figure 5-7. These current transitions ranged from 11 pA to 60 pA. The idea of multiple channels being open is not completely ruled out, nor is the less favorable explanation of electroporation occurring. If more than one channel is open simu ltaneously, the conductance and the current would ideally in crease as shown in Figure 5-5 a nd Figure 5-7. The bilayer in Figure 5-6 maintained a giga-seal between 2.9-3.1 G for 28 minutes while facing applied voltages ranging from 20 mV-300 mV. Single cha nnel activity of Gramicidin was observed at 155 mV. We also observed single channel activity at voltages as high as 238 mV (data not shown). This activity was comparable to activity reported in literature where Gramicidins

PAGE 126

126 conductance range is between 20-90 pS and in pr evious experiments conducted on the device by veteran lab members ( 17, 19, 20, 123). We concluded that the 70PC/30PE bilayer was sufficient; and incorporated WT MscL to the tBLM system after 12 experiments of analyzing the activity and optimizing the conditions for pr oteoliposome fusion with Gramicidin. Figure 5-7. Single channel activity of multiple Gramicidin channels opened within the tethered lipid bilayer. Multiple channels open at 155 mV with current transition corresponding between 4-15 pA and the conductance was between 26-90 pS. Single Channel Analysis of MscL Single ion channels exhibit stochastic activity. Theoretically, the durations of events and the orde r in which they o ccur are random variables ( 99, 100 ). Parameters representative of single channel activity, and the information c ontained in each event, are measured from statistical distributions ( 78, 100). Three specific aims exist when analyzing single ion channel activity of MscL reconstituted in this new system: 1) determination of ionic selectivity or preference, 2) the duration of open and shut times for lifetime studies ( 100 ), and 3) amplitude studies to understand ion permeation (conductance) through open channels at different ionic compositions.

PAGE 127

127 Ionic Selectivity MscL is not selective to specific ions because of its large pore size, and n atural response to elute cytoplasmic solutes before cell lyses. Therefore all ions present in the buffer solution within the tBLM permeated the channel. Lifetime Studies Lifetim e distributions of MscL in the tethered system, determined from probability density functions, displayed fast kinetics with the channel open in the ra nge of several ms. These are smaller, and comparable to those found in literature with spheroplast (34, 64, 150). Conductance Analysis of amplitude histogram s assisted in the identification of fully open and sub-states during continuous traces. Sub-st ates are conducting states where the channel is neither fully open nor closed. All digitized current can be plot ted as a histogram with a peak for each shut and open level. Figure 5-8. Single channel ac tivity of WT MscL in the tLBM system: When 300 mV was applied to the membrane MscL opened conduc ting 266 pS. A) The channel is closed at 20 pA and the change in current to the open state is 80 pA B) The unitary conductance of ~300 pS is obtained by divi ding the change in current by the voltage applied ( 64).

PAGE 128

128 The area under the peak is propor tional to the time spent at that level. Unitary conductance is calculated by dividing the mean amplitude calculated from the amplitude histograms by the applied potential. In order to create tension in the bilayer and cau se MscL to open, an electrical field induced membrane stress ( ) strong enough to trigger an opening of the MscL channel was applied. According to Needham and Hochmuth ( 146), the tension as a func tion of the transmembrane potential (V ) can be expressed as: = 0(d(2/de 2)V2 ( 151). This tension corresponds to a transmembrane potential of ~300 mV, which is in very good agreement with our findings. Figures 5-8 and5-9 display single ion channel activity of WT MscL within a 70 DPhPC/30 DPhPE tethered lipid bilayer matrix. The clos ed state of the channel is at 20 pA before normalization, while the open state is around 100 pA ; this gives a unitary conductance of 266 pS when 300 mV is applied according to Ohms law. The conductance of WT MscL obtained from using spheroplast and in vitro using liposome is 3 nS. The lower conductance observed with the tBLM is not completely understood. Figure 5-9: MscL displaying two co nducting-states: A.) Single channe l trace with the lowest substate conducting ~ 3 pS and the large sub-state conducti ng at 255 pS. B.) Point amplitude histogram displaying the multiple levels including a small substrate around 15 pA ( 64). Possible explanations could include, but are not limited to, the diphytanoyl lipid environment, the lack of an external applied pressure, and the presence of the limited ionic

PAGE 129

129 reservoir between the Au surface and the bilayer. In this reservoir the i on mobility is decreased because of the lack of free space th at results from the tethers. The high applied voltage of 300 mV required to gate WT MscL, was a challenge to obtain. Most of the bilayers would rupture and loss thei r giga-seal resistance before 300 mV was applied or the channel did not always respond at 300 mV. Therefore, channel activity was analyzed from a small sample pool. In order to obtain a greate r sample size and decrease the voltage required to gate MscL, Dr. Blount suggested we incorporate a voltage sensitive mutant, K31E. This mutant inhibited the growth of E. coli by creating leaky MscL channels th at responded to lower tension thresholds in the cellular membrane and to negative applied voltages ( 28 ). This was configured from both whole cell physiol ogy and electrophysiology experiments which examined K31Es response to -20 to -100 mV being applied to sphereoplast; K31E s howed a dramatic increase in the probability of opening at the higher voltages ( 28). Blount et. al. concluded that this observation was probably due to th e charge reversal in this po sition (see Figure 5-10), which interfaces with the electric field of the membrane ( 28). Thus, we set out to determine if this mutant would be observed at lower voltages when inserted within the microchip device. The K31E mutant was molecularly engineered and re constituted into the 70PC/30PE lipid mixture in order to create proteoliposomes. Vesicle fusion with K31E proteoliposomes was also researched instead of adding the protein to the tBLM once it formed. This was an attempt to increase the probability of protein reconstitution into the tBLM. Results from K31E reconstituted into the tBLM are shown in Figure 5-10. Single channel activity was observed at lower potentials, as pr edicted. The channel became active when 85 mV were applied while conducting at 178.5 pS whic h is lower then that the conductance of MscL

PAGE 130

130 within the tBLM. However, when 143 mV wa s applied, we observed conductance as high as 350 pS (refer to Figure 5-11). Figure 5-10. MscL with the lysine in the 31 pos ition highlighted. In this study the positively charged lysine was mutated to glutamic acid which has a negative charge. This mutant is voltage sensitive. A) Side view. B) Top view Figure 5-10. Single channel recordings of K31E (voltage sensitive mutant of MscL) within the tBLM. The upper trace displays burst ope n and closing events at 178.5 pS. The lower trace is a point amplitude or all-point s histogram displaying the current levels observed in the above trace. The open current transitions were 20 pA.

PAGE 131

131 Figure 5-11. Single channel activity of K31E at 143 mV. A.) Displa ys burst opening and closing events at 143 mV giving way to curr ent transition as large as 60 pA. B.) A point amplitude or all-points histogram displaying the current levels observed in the above trace. The open current transitions was 60 pA providing a conductance of 350 pS We also observed conductance as low as 50 pS and well above 300 pS as shown in Figure 5-11. MscL is known to have multiple sub-conducting states ( 152, 153). Ghazi et al. performed patch-clamp studies on MscL in both proteolip osomes and shereoplast and found that the channel conductance ranged from 100-2,300 pS depending on the applied negative pressure ( 153). Sub-conductance states may be act ually forming within the tBLM. The single channel kinetics of the K31E muta nt was fast and the number of events was low. This result was comparable to changes in si ngle channel kinetics observ ed with this mutant in blisters and sphereoplast ( 44). Most activity transpired at the beginning of the experiments. Although still preliminary, results obtained from this mutant are extremely promising.

PAGE 132

132 Discussion The tether ed lipid environment is a new concept for integrating ion channels onto solid supports. These ion channels, particularly E. colis MscL, can be modified for nanosensory applications within solid supports In this study, we characterized the single cha nnel activity of Gramicidin, WT MscL, and a voltage sensitive muta nt K31E-MscL within the tBLM. In order to create tension in the tBLM, a high voltage was required and then applied to the system. The resulting conductance of MscL was low in comp arison to the conductance obtained from single channel studies within sphere oplast and protioliposomes ( 10, 26, 36). This result was characteristic of ion channels reconstituted within the tBLM. So far, our lab has reconstituted four different ion channels independently within the tBLM system and observed lower conductance for all of the channels except for Gramicidin ( 64, 77, 112 ). Gramicidins simple structure and small pore(s) may have enabled it to maintain its small conductance between 50-90 pS within the device ( 77 ). We also were able to observe channel activity at different voltages from Gramicidin because channel gating wa s independent of the applied voltage. The observed conductance was lower in tBLM for the other channels than in their natural environment or in liposomes as shown in Table 5-1. The three channels include MscL, Maxi-K, and the synthetic self-assembling protein M2 (a synthetic ion channel based on the channellining domain of the nicotinic ace tylcholine receptor (nACHR) from Torpedo californica ). The Maxi-K channel, which was isolated from mouse brain and muscle, then expressed in Xenopus laevis oocytes, is both Ca2+ activated and vo ltage sensitive (112 ). The conductance is greater than 200 pS. Within the tBLM, th e conductance obtained was 37 pS ( 112 ). The M2 channel consists of three or more single peptides that co me together in the bilaye r and form a pore with a diameter that ranges from 5-15 nm ( 14). The unitary conductance of 15 pS was observed for M2 within the tBLM (112 ). This conductance is three-ti mes lower than the conductance of

PAGE 133

133 templated-M2 under normal patch-clamp conditions. Finally, as reported in this study, the conductance of MscL is also 10 times lower within the tBLM than in other environments. This lower apparent conductance (and the large condu ctance range for all channels) of multiple channels could be caused by the atypical phy tanoyl phospholipids envi ronment, the limited reservoir volume, constriction of the ion channel fr om the isoprenoyl-tail region of the tethers, an interfacial capacitance, or in teractions at the gold surface ( 112 ). A likely trend existing with the results from the four channels st udied in this device is the larger the conductance of the channel, the more attenuation of the conductance observed in the tBLM as shown in Table 5-1. Table 5-1. Conductance comparison between the ion cha nnels reconstituted in the tBLM and reported conductance. While a limited reservoir seems a likely explan ation for the attenuation of the conductance of the channels studied, additiona l artifacts associated with the i ndividual protein could also lead to the lower conductance. For instance, si nce single transmembrane peptides of M2 selfassemble within the bilayer into a trimer, tetr amer, pentamer, or even a hexamer, the lipid environment within the tBLM might favor the fo rmation of a trimer. The conductance of the trimer as examined by Montal et al. using the painting method and results from our research investigating this channel using the tip-dip method, is 15-20 pS (14, 86, 155, 158). The lowered conductance of MscL may in part be due to one of its known sub-conducti ng states. According Ion Channels Reported Conductance (pS) Conductance within the tBLM (pS) References Gramicidin 20-90 50-90 ( 77, 122, 123 ) M2 15-45 13-15 ( 14, 88, 154156 ) Maxi-K 90-300 30-50 ( 97, 156, 157 ) MscL 100-3000 50-300 ( 10, 64, 153 )

PAGE 134

134 to Sukharev et al. MscL has 3 sub-conducting states: S2, S3 and S4, and a fully open state O5 ( 153, 159). The different conductance states range from 1.03 nS to 3.37 nS when the protein is reconstituted in Azolectin liposomes (159). Therefore, when MscL is reconstituted within the tBLM system, the channel may not fully ope n, which may contribute to the lowered conductance. To completely understand the underlying dynamics of the tBLM system and its resulting single channel activity, a more sophisticated equi valent circuit model is required. This model will enable a more accurate account for the actual transmembrane potential within the tBLM system, and the possible control of the conducta nce range. In additi on, it should enable the probing of channel fluctuations in a voltage or current clamp configuration within a microarray system. Another limitation we were able to address was the extreme high voltage requirement used to gate MscL within the tBLM. Many of the bila yers ruptured before withstanding voltages near 300 mV, which seemed to be required to create e nough tension in the bilayer to open MscL. By incorporating the voltage sensi tive K31E-MscL mutant, as expected we were able to observe single channel activity at much lower voltages. However, there were technical differences between the study of the wild type and K31E channels. First, in order to increase expression levels of the usually poorly-expressed K31E mu tant, the C-terminal from residue 110-136 was truncated. While there are no in dications in patch clamp that the truncated channel is more sensitive to voltage, one would lik e to perform the proper control. Thus, a C-terminal truncated MscL was also generated with th e anticipation that its activity would be similar to WT MscL on the device. Early results, however, were inclus ive due to negative result s; the experiments are technically difficult because of the large voltage requirements for opening the channel, which

PAGE 135

135 often disrupts the bilayer. Second, as describe d in Material and Methods, a slightly different procedure was used to incorporate the channel into the tBLM. Here again, it seems unlikely that this change would lead to the increased sensi tivity for the K31E channe l. Hence, while the simplest interpretation of the data is that, as expected, the K31E mutant shows increased voltage sensitivity, additional controls shoul d be performed in the future to rule out the possibility that minor technical alterations played some ro le in the increased sensitivity observed. Conclusions In this study, we examined the functional characteristics of Gramicidin, MscL, and a voltage sensitive conjugate of this channel reconstituted on a tethered membrane bilayer. The channel was gated on an applied voltage, which according to our calcula tions, gives rise to a stress on the lipid bilayer that is similar to pressure induced tensi ons. This high voltage requirement for WT-MscL increased the amount of irreversible breakdown of the bilayers before channel activity could be observed. Therefore, a voltage sensitive mutant K31E was integrated into the tBLM and it gated at much lo wer voltages ranging from 85 mV to 150 mV. Conductance from this channel was comparable to MscL in the device and the fast single channel kinetics were similar to results of this mutant in sphereoplast ( 44 ). Furthermore, these findings show the possibility of using MscL as either a release valve or a designer sensor channel for engineered membrane devices. If MscL was to serve as the actual sensor component within a microchip system, the large conductance of the channel may cause some limitations in the device. Since the conductance of MscL and the voltage sensitive mutant is already 10 times lower within the tBLM then its natural conductance, further extensive studies should examine a range in which these channels conduct in this system. Once this range is determined, and the proper controls are done with both MscL and K31E, then alterations to either channel should be

PAGE 136

136 researched where the conductance can be decr eased or increased. The earlier C-terminal mutation studies in Chapter 3 offer many advantages in decreasing the conductance of the channel without affecting the pore of the channel. Cysteine muta tions to this region can create sensitivity to redox reagents, as well as decrease the conductance by disulf ide-bridging in this area. Histidine mutations to the upper C-terminal region, more specifically A110-A114, can create sensitivity within the channel for heavy metals. The binding events can be interpreted by the decrease in single channel kinetics and c onductance. The beauty of incorporating the Histidines mutants to this region is the reversib ility back to MscLs normal activity once the heavy metals are removed. All in all the ability to incorporate MscL into the tBLM system was the first step, and has opened the door to limitless possibilities of bettering this micro-bioelectronic construct.

PAGE 137

137 CHAPTER 6 CONCLUSION When creating a biosensor based on single ion channel ac tivity, there exist several components at the bio-molecular level that must be fully understood. For instance, conformational changes w ithin the protein must be understood in order to incorporate sensory components to key regions. The proteins ability to reconstitute succe ssfully into non-native lipid environments and remain ac tive should also be investigated. An electrically stable lipid environment must also be studied. In this body of research we have analyzed the proximity and mobility of vital residues in the C-terminal region of the Mechanosensitive Channel of Large Conductance. The structure of this region durin g gating of the channel was under debate and our results support the theory of th is region remaining closed during gating. Results from this study supported the new molecular model of MscL in the protein data base (pdb-2ORA). An even more exciting discovery was obtained from the generation of new mutants. The first mutant, AAA (A110C, A111C, and A112C) is able to decrease th e conductance of MscL when placed in oxidative conditions, and this reac tion is reversible. The second set of mutants are A110H and A112H, which can also decrea se both the conductan ce and single channel activity of MscL in presence of heavy metals. This reaction is also reversible. Future experiments should include ways of isolating an d purifying these mutants in order to study them in vitro within liposomes and a tethered lipid bila yer membrane. Currently, purification of MscL is done using a nickel affinity column that binds to a 6-His-tag placed on the C-terminal of MscL. Possible cleavage of the His-tag could be done once the protein is eluted during purification to ensure the heavy metals interact with the single Histidine mutations rather than the His tag.

PAGE 138

138 The second part of this project was an attemp t to establish the ideal lipid environment for studying single ion channels in a tethered device for biosensor appl ications. Therefore different diphytanoyl lipid ratios were compared using the tip-dip method. Diphyta noyl lipids were vital components of this research due to their increase d durability and stability when tethering to solid supports such as the gold surface on the device. In this novel syst em of incorporating diphytanoyl lipids of PC and PE with the tip dip method; electroporation was unfortunately observed and characterized. Result s indicated all of the lipid-ratios suffered from pore formation due to electrical breakdown, both reversible and irreversible. This was the first time that electroporation was reported at low potentials such as 125 mV and even 40 mV which may be characteristic of using the tip-dip method with diphtanoyl lipids. The lipid ratio of 70PC/30PE was chosen as the most stable lipid ratio and was integrated as the synthetic lipid environment for both Gram icidin and MscL on the tBLM device. Before we decided to move on to the tBLM, we analyzed single channel activity from Gramicidin using the tip-dip method. This was the fi rst time this channel has been st udied in this particular lipid system and analyzed by the tipdip method. This system offered a number of both advantages and disadvantages. For instance, the ease in pr eparation was definitely a plus due to the thermodynamically challenging nature of establishing bilayers (97) This research also added to the fundamental knowledge of Gramicidin chan nel. One of the disadvantages was the appearance of electroporation within the bilayers that could be mistaken for single ion channel activity if using a synthetic ion ch annel or one that is not as char acterized as Gramicidin. The epore formation was random; and we were able to distinguish it from single ion channel conductance by plotting I-V curves and evalua ting the kinetics of the pores formed in comparison to channel activity. Studying Gramic idin in this lipid ratio using this method

PAGE 139

139 provided positive results with th e conductance and dwell times of the channel being comparable to results from literature proving that diphytanoy l lipids in a ratio of 70PC/30PE can actually form stable bilayers suitable for single channel measurements. These results brought us one step closer to understanding the mechan isms of different lipid species, especially those that can be integrated on to solid supports which was comp leted in our next study using first the model channel, Gramicidin, followed by MscL and a voltage sensitive mutant of MscL, K31E which as predicted appears to respond to lower applied voltages. The single channel response of Gramicidin within the tBLM was a bit different then that of MscL within the system. Gramicidins activity wa s comparable with other techniques used to analyze this model pore. MscL on the other hand displayed a 10 fold decrease in conductance on the device. This characteristic has however b een observed with other large channels within the tBLM system. Many reasons including the teth ered ionic reservoir, the lipid environment, and the voltage induced gating coul d have affected the proteins act ivity within this system. The channel was gated upon an applied voltage, which according to our calculations was due to a stress on the lipid bilayer similar to pressure i nduced tension required to open MscL. One of the benefits of inducing gating by an applied potentia l, was the incorporation of a voltage sensitive K31E MscL mutant into the tBLM. This mutant s single channel activity was comparable to MscL within the device were the conductance range was around 300 pS. The kinetics of this channel in the tBLM was also similar to inside-o ut patch clamping assays. More research with this mutant should be conducted to determine a more in depth conductance range and single channel characterization within the tBLM Future changes to the overall configuration of the tBLM syst em should also be considered to qualify control over more parameters on the device. For instance, Wolfgang Knoll et al. has

PAGE 140

140 studied the tBLM system, in part icular the DTPL molecule, for the last ten years in order to understand its affect on the lipid bilayer and the movement of ions in this region ( 61, 65, 66, 68, 79, 81, 104). In 2005, he worked in collaboration with Lizhong He and Renate Naumann at the Max Planck Institute for Polymer Research inco rporating hydrophilic thio ls with the goal of increasing the fluidity of the ionic reservoir within the tBLM system ( 79). With the success of these experiments, Knoll et al. combined polymer supported bilaye r and the tethered bilayer to create the polymer-tethere d lipid bilayer system on modified substrates ( 104 ). This configuration increased the fluidity as well as the distance be tween the bilayer and substrate surface with the overall distance of the polymer-tether being 11.2 nm in diameter instead of the 1.4 nm-3 nm DPTL linker regions. An alternative to measuring the activity of single ion channels w ithin the tBLM should also be considered. Since there is a capacita nce associated with the membrane, the solution interfaces above and below the membrane, and re sistance within the membrane, the electrolyte solution, and the ion channel, a sophisticated equivalent circu it model is required to study the exact ion-channel activity. This model will en able a more accurate account for the actual transmembrane potential within the tBLM system In addition, it should enable the probing of channel fluctuations in a voltage or current cl amp configuration within a micro-array system. Impedance Spectroscopy as been used as the modulating detection system for ion channels within tBLMs ( 62, 65, 69, 74, 76, 79, 80, 102, 160). In this technique an ac potential is applied to an interface between 2 elec trodes and impedance (resistance of alternating current) is analyzed. The results, from this study, nonetheless show the possibility of also using MscL as a release valve for engineered memb rane devices. The large conducta nce of the channel within the

PAGE 141

141 tBLM device is 10 fold lower then reported observa tions in model systems. It is still however, one of the largest conducting channels to date. The fact that it also senses tension in the bilayer can enable this channel to be reconstituted with another biosensor and serve as a release valve whereby: the other channel does th e sensing, and opens allowing ions to translocate into the ionic reservoir between the gold surface and lower leaflet of the bilayer. It is theoretically predicted that the smallest of channels conducting at 50 pS can fill the ionic reservoir within 100 secs. MscL could sense the tension in th e bilayer and open releasing the constituents in the reservoir, thereby increasing the durability of the overall bios ensor device. The earlier C-terminal mutation studies in Chapter 3 offer many advantages in creating designer cha nnels with readable recognition events. Cysteine mutations to this region can create sensitivity to redox reagents, as well as decrease the conductance by disulfide-bridging in this area. Histidine mutations to the upper C-terminal region can create selectivity within the channel for heavy metals. The binding events can be interpreted by the decrease in single channel kinetics. The b eauty of incorporating either one of these mutants to this region is the reversibility back to MscLs normal activity once either the redox reagents or hea vy metals are removed. Therefor e MscL offers two distinct yet important characteristics to future stochastic sensors based on singl e ion channel activity.

PAGE 142

142 APPENDIX 1 DLS OF THE SIZE DISTRIBU TION OF 70PC/30 PE The Figures below show the intensity of the vesicles size distributions directly after extrusion (A ), a week later (B) and tw o weeks after storage at 4C (C).

PAGE 143

143 APPENDIX 2 STABLE BILAYER TRACE AND I-V CURVES This appendix includes supporting data for th e stability of the 70PC/30PE lipid bilayer within the tBLM. The 1st Figure below is a trace of a stab le bilayer formed by the 70PC/30PE lipid composition within the tBLM. The calculated resistance is 4.08 G and the applied voltage is 300 mV. The time scal e of the trace is 30 minutes.

PAGE 144

144 These Figures are I-V curves from three differe nt stable bilayers w ithin the tBLM. The resistance of the I-V curve calculated from Ohms law was 4.5 G for Figure A, 2.1 G for Figure B, and 2.8 for Figure C.

PAGE 145

145 APPENDIX 3 PROTEIN RECTIFICATION WITHIN THE DEVICE This appendix yields to expl ain the flow of potassium ions though Gram icidin and how the dynamics of the tBLM creates protein rectificat ion. When a positive voltage is applied to the Ag/AgCl electrode, Clfrom the buffer solution react with the newly formed Ag+ and the K+ from the bulk solution above the bilayer travel through the open pore of ion channel into the tethered region. When a negative voltage is a pplied to the Ag/AgCl electrode, the Ag/AgCl is reduced to metallic Ag, and Clflows into the solution. This switch in polarity causes the K+ within the limited-tethered ioni c reservoir, between the golds su rface and the bilayer to travel through the pore of the ion channel. The ion mobility within this region is slowed due to the high viscosity and smaller volume of this region ca using, a small rectification with Gramicidin.

PAGE 146

146 LIST OF REFERENCES (1) Subrahm anyam, S., Piletsky, S., and Turner, A. (2002) Application natural receptors in sensors and assays. Anal. Chem. 74, 3942-3951. (2) Bayley, H., and Cremer, P. (2001) Stochastic sensors inspired by biology. Nature 413, 226-230. (3) Batiza, A. F., Yoshimura, K., Schroeder, M., Blount, P., and Kung, C. (1999) Glycine 22 is crucial in mechanical gating of the MscL channel. Biophysical Journal 76, A138-a138. (4) Kloda, A., Ghazi, A., and Martinac, B. ( 2006) C-terminal charged cluster of MscL, RKKEE, functions as a pH sensor. Biophysical Journal 90, 1992-1998. (5) Davidson, M. W. (2004) (1st, Ed.), Florida State University, Tallahassee. (6) Tien, H. T. (1995) Self-assembled lipid bi layers as a smart material for nanotechnology. Materials Science and Engineering C3 7-12. (7) Puu, G., and Gustafson, I. (1997) Planar lipi d bilayers on solid supports from liposomes factors of importance for kinetics and stability. Biochimica et Biophysica Acta (BBA) Biomembranes 1327, 149-161. (8) Winterhalter, M. (2000) Black lipid membranes. Current Opinion in Colloid & Interface Science 5, 250-233. (9) Liu, A. L. (2006) Advances in Planar Lipid Bilayers and Liposomes Vol. 3, 3 ed., Elsevier, London. (10) Moe, P., and Blount, P. (2005) Assessment of potential stimuli for mechano-dependent gating of MscL: effects of pressu re, tension, and lipid headgroups. Biochemistry 44, 12239-12244. (11) Castellana, E., and Cremer P. (2006) Solid supported lip id bilayers: From biophysical studies to sensor design. Surface Science Reports 61, 429-444. (12) Yeagle, P. (2005) Structure of Biological Membranes Vol. 1, 1 ed., CRC. (13) L.Nelson, D., and Cox, M. M. (2005) Lehniger Principles of Biochemistry fouth ed., Sara Tenney, New York. (14) Montal, M. O., Iwamoto, T., Tomich, J., and Montal, M. (1993) Design, synthesis and Functional Characterization of a pentameric Channel Protein that Mimics the Presumed pore structure of the Nicotinic Cholinergic Receptor. Federation of European Biochemical Societies 320 261-266.

PAGE 147

147 (15) Hille, B. (2001) Ion Channels of Excitable Membranes Sinauer Associates, Inc., Sunderland, Massachusetts. (16) MacKinnon, R. (2004) Potassium channels and atomic basis of selective ion conduction (Nobel Lecture). Angewandte Chemie 43, 4265-4277. (17) Urry, D. W. (1971) The Gramicidin A transmembrane channel: A proposed (L,D) eelix. Proc Natl Acad Sci U S A 68, 672-676. (18) Koeppe, R. E. (1996) Engineering the Gramicidin channel. Annu. Rev. Biophys. Biomol. Struct. 25 231-258. (19) Cross, T. A., Arseniev, A., Cornell, B. A., Davis, J. H., Killian, J. A., Koeppe, R. E., Nicholson, L. K., Separovic, F., and Wa llace, B. A. (1999) Gramicidin channel controversy revisited. Nature Structural Biology 6, 610-611. (20) Ring, A. (1992) Influence of ion occupanc y and membrane deformation on gramicidin A channel stability in lipid membranes. 61, 1306-1315. (21) Hamill, O. P., and Martinac, B. (2001) Molecular basis of mecha notransduction in living cells. Physiological Reviews 81, 685-740. (22) Chiang, C.-S. (2005) in Department of Biology pp 164, University of Maryland at College Park, Bethesda (23) Strange, K. (1994) Cellular and Molecular Physiology of Cell Volume Regulation CRC Press. (24) Blount, P., and Moe, P. C. (1999) Bact erial mechanosensitive channels: integrating physiology, structure and function. Trends in Microbiology 7, 420-424. (25) Sukharev, S., and Anishkin, A. (2004) Mechanosensitive channels : what can we learn from `simple' model systems? Trends in Neurosciences 27, 345-351. (26) Blount, P., Sukharev, S. I., Moe, P. C., Na gle, S. K., and Kung, C. (1996) Towards an understanding of the structural and functiona l properties of MscL, a mechanosensitive channel in bacteria. Biology of the Cell 87, 1-8. (27) Blount, P., Sukharev, S. I., Schroeder, M. J., Nagle, S. K., and Kung, C. (1996) Single residue substitutions that change the gating pr operties of a mechanosensitive channel in Escherichia coli. Proc Natl Acad Sci U S A 93 11652-11657. (28) Blount, P., Schroeder, M. J., and Kung, C. (1997) Mutations in a bacterial mechanosensitive channel change the cellular response to osmotic stress. Journal of Biological Chemistry 272 32150-32157.

PAGE 148

148 (29) Sukharev, S. I., Blount, P., Martinac, B., and Kung, C. (1997) Mechanosensitive channels of Escherichia coli: The MscL gene, protein, and activities. Annual Review of Physiology 59, 633-657. (30) Sokabe, M., Sachs, F., and Jing, Z. Q. (1991) Quantitative video microscopy of patch clamped membranes stress, strain, capaci tance, and stretch channel activation. Biophysical Journal 59, 722-728. (31) Delcour, A. H., Martinac, B., Adler, J ., and Kung, C. (1989) Modified reconstitution method used in patch-clamp studies of Escherichia-Coli ion channels. Biophysical Journal 56, 631-636. (32) Sukharev, S. I., Blount, P., Nagle, S., Moe, P., and Kung, C. (1995) MscL, a mechanosensitive channel in Escherichia coli. Journal of General Physiology 106, 9-9. (33) Berrier, C., Coulombe, A., Houssin, C., a nd Ghazi, A. (1989) A patch-clamp study of ion channels of inner and outer membrances and of contact zones of E. coli, fused into giant liposomes. Pressure-activated channels are localizedin the inner membrane. FEBS Letters 259, 27-32. (34) Blount, P., Sukharev, S., Martinac, B., and Kung, C. (1994) Molecular analysis of a gene that encodes a mecanosensitive channel in Escherichia-Coli. Biophysical Journal 66, A170-a170. (35) Hamann, A., Bossemeyer, D., and Baker, E. P. (1987) Physical mapping of the K+ transport trkA gene of Escherichia co li and overproduction of the TrkA protein. J. Bacteriol 169 3138-3145. (36) Sukharev, S. I., Blount, P., Martinac, B., a nd Kung, C. (1994) Molecular-Identification of Large-Conductance Mechanosensitive Cha nnel Protein from Escherichia-Coli. Biophysical Journal 66, A170-a170. (37) Blount, P., Sukharev, S., Sc hroeder, M., Nagle, S., and Kung, C. (1996) Mutations that change gating properties of a mechanosensitive channel in E-coli. Biophysical Journal 70, Wamd7-Wamd7. (38) Martinac, B., Buechner, M., Delcour, A. H., Adler, J., and Kung, C. (1987) Pressuresensitive ion channel in Escherichia-Coli. Proc Natl Acad Sci U S A 84 2297-2301. (39) Blount, P., Sukharev, S. I., Moe, P., Martinac, B., and Kung, C. (1999) Methods in Enzymolgy, Vol. 294, Academic Press, San Diego. (40) Kung, C., and Blount, P. (2004) Channels in microbes: so many holes to fill. Molecular Microbiology 53, 373-380. (41) Kloda, A., Petrov, E., Meyer, G. R., Nguyen, T., Hurst, A. C., Hool, L., and Martinac, B. (2008) Mechanosensitive cha nnel of large conductance. The International Journal of Biochemistry & Cell Biology 40, 164-169.

PAGE 149

149 (42) Corry, B., and Martinac, B. Bacterial mech anosensitive channels: Experiment and theory. Biochimica et Biophysica Acta (BBA) Biomembranes In Press, Corrected Proof (43) Blount, P., Sukharev, S. I., Moe, P. C., Sc hroeder, M. J., Guy, H. R., and Kung, C. (1996) Membrane topology and multimeric structure of a mechanosensitive channel protein of Escherichia coli. Embo Journal 15, 4798-4805. (44) Blount, P., Shroeder, M. J., and Kung, C. (1998) Mutations in MscL, a bacterial mechanosensitive channel, can change the cellular response to osmotic stress. Biophysical Journal 74, A239-a239. (45) Chang, G., Spencer, R. H., Lee, A. T., Ba rclay, M. T., and Rees, D. C. (1998) Structure of the MscL homolog from Mycobacterium t uberculosis: A gated mechanosensitive ion channel. Science 282, 2220-2226. (46) Moe, P. C., Levin, G., and Blount, P. (2000) Correlating a protein structure with function of a bacterial mechanosensitive channel. Journal of Biological Chemistry 275, 3112131127. (47) Poolman, B., Spitzer, J. J ., and Wood, J. M. (2004) Bacterial osmosensing: roles of membrane structure and electrostatics in lip id-protein and protein-protein interactions. Biochimica et Biophysica Acta (BBA) Biomembranes 1666, 88-104. (48) Jeon, J., and Voth, G. A. (2008) Gating of the Mechanosensitive channel protein MscL: the interplay of membrane and protein. Biophysical. Journal. 94, 3497-3511. (49) Cruickshank, C. C., Minchin, R. F., LeDai n, A. C., and Martinac, B. (1997) Estimation of the pore size of the large-conductance mechanos ensitive ion channel of Escherichia coli. Biophysical Journal 73, 1925-1931. (50) Ou, X. R., Blount, P., Hoffman, R. J., a nd Kung, C. (1998) One face of a transmembrane helix is crucial in mecha nosensitive channel gating. Proceedings of the National Academy of Sciences of th e United States of America 95, 11471-11475. (51) Anishkin, A., Gendel, V., Sharifi, N. A., Chiang, C. S., Shirinian, L., Guy, H. R., and Sukharev, S. (2003) On the conformation of the COOH-terminal domain of the large mechanosensitive channel MscL. Journal of General Physiology 121, 227-244. (52) Elmore, D. E., Philipson, K. D., and Dougherty, D. A. (2001) Molecular dynamics simulations of the M-tuberculosis mechanos ensitive channel of large conductance. Biophysical Journal 80, 110A-111A. (53) Perozo, E., Kloda, A., Cortes, D. M., and Martinac, B. (2001) Site -directed spin-labeling analysis of reconstituted MscL in the closed state. Journal of General Physiology 118, 193-205.

PAGE 150

150 (54) Ajouz, B., Berrier, C., Garrigues, A., Be snard, M., and Ghazi, E. (1998) Release of thioredoxin via the mechanosensitive channel MscL during osmotic downshock of Escherichia coli cells. Journal of Biological Chemistry 273, 26670-26674. (55) Ewis, H. E., and Lu, C. D. (2005) Osmotic shock: A mechanosensitive channel blocker can prevent release of cytoplasmi c but not periplasmic proteins. Fems Microbiology Letters 253, 295-301. (56) Kong, Y., Shen, Y., Warth, T. E., and Ma, J. (2002) Conformational pathways in the gating of Escherichia coli mechanosensitive channel. Proc Natl Acad Sci U S A 99, 59996004. (57) BraachMaksvytis, V. L. B., Cornell, B. A., King, L. G., Osman, P. D. J., Pace, R. J., Raguse, B., and Wieczorek, L. (1997) Teth ered lipid bilayer: formation, ionophore incorporation and physical characterisation. Biophysical Journal 72, Th288-Th288. (58) Matsuno, Y., Osono, C., Hirano, A., and Sugawara, M. (2004) Single-Channel Recordings of Gramicidin at agarose-suppor ted bilayer lipid membranes formed by the tip-dip and painting method. Analytical sciences 20 1217-1221. (59) Richter, R. P., Berat, R., and Brisson, A. R. (2006) Formation of solid-supported lipid bilayers: an integrated view. Langmuir 22, 3497-3505. (60) Goennenwein, S., Tanaka, M., Hu, B., Mo roder, L., and Sackmann, E. (2003) Functional incorporation of integr ins into solid supported membranes on ultrathin films of cellulose: impact on adhesion. Biophysical. Journal. 85, 646-655. (61) Nauman, C. A., Prucker, O., Lehmann, T ., Ruho, J., Knoll, W., and Frank, C. W. (2002) The polymer-supported phospholipid bilayer: teth ering as a new approach to substratemembrane stabilization. Biomacromolecules 3, 27-35. (62) Terrettaz, S., Mayer, M., and Vogel, H. (2003) Highly electrically insulating tethered lipid bilayers for probing the function of ion channel proteins. Langmuir 19, 5567-5569. (63) Plant, A. L. (1999) Supported hybrid bilayer membranes as rugged cell membrane mimics. Langmuir 15, 5128-5135. (64) Martin, A., Okeyo, G., Wilson, D., Ke iqer, H., Moe, P., Blount, P., Fine, D., Dodabalapur, A., and Duran, R. (2008) Voltage induced gating of the Mechanosensitive MscL ion channel reconstituted in a tether lipid bilayer membrane. Biosensors and Bioelectronics 23, 919-923. (65) Knoll, W., Frank, C. W., Heibel, C., Naumann, R., Offenhausser, A., Ruhe, J., Schmidt, E. K., Shen, and Sinner, A. (2000) Functional tethered lipid bilayers. Reviews in Molecular Biotechnology 74, 137-158.

PAGE 151

151 (66) Knoll, W., Koper, I., Naumann, R., Sch iller, S., and Duran, R. S. (2004) Functional tethered bimolecular lipid membranes. Abstracts of Papers of the American Chemical Society 227, U853-U853. (67) Vockenroth, I., Atanasova, P., Knoll, W., and Koper, I. (2005 ) Platform: Funtional tethered bilayer membranes as a biosensor Sensors, 2005 IEEE, 603-606. (68) Breffa, C. (2005) in Chemistry pp 150, Max Planck Institu t fur Polymerforschung and Johannes Gutenberg Univer sitat Mainz, Geboren. (69) Romer, W., and Steinem, C. (2004) Impedance analysis a nd single-channel recordings on nano-black lipid membranes based on porous Alumina. Biophys. J. 86, 955-965. (70) Steltenkamp, S., Muller, M. M., Deserno, M ., Hennesthal, C., Steinem, C., and Janshoff, A. (2006) Mechanical properties of pore-spanni ng lipid bilayers probed by Atomic Force Microscopy. Biophys. J. 91, 217-226. (71) Reviakine, I., and Brisson, A. (2000) Fo rmation of supported phospholipid bilayers from unilamellar vesicles investigated by Atomic Force Microscopy. Langmuir 16, 1806-1815. (72) Kim, Y.-H., Rahman, M. M., Zhang, Z.-L ., Misawa, N., Tero, R., and Urisu, T. (2006) Supported lipid bilayer formation by the gian t vesicle fusion induced by vesicle-surface electrostatic attractive interaction. Chemical Physics Letters 420 569-573. (73) Tamm, L. K., and McConnell, H. M. (1985) Supported phospholipid bilayers. 47, 105113. (74) Raguse, B., Braach-Maksvytis, V., Cornell, B. A., King, L. G., Osman, P. D. J., Pace, R. J., and Wieczorek, L. (1998) Tethered lipid bilayer membranes: Formation and ionic reservoir characterization. Langmuir 14, 648-659. (75) Cornell, B. A., Krishna, G., Osman, P. D., Pace, R. D., and Wieczorek, L. (2001) Tethered-bilayer lipid membranes as a support for membrane-active peptides. Biochemical Society Transactions 29 613-617. (76) Krishna, G., Schulte, J., Cornell, B. A., Pace, R., Wieczorek, L., and Osman, P. D. (2001) Tethered bilayer membranes containing ionic reservoirs: The interfacial capacitance. Langmuir 17, 4858-4866. (77) Andersson, M., Keizer, H. M., Zhu, C., Fi ne, D., Dodabalapur, A., and Duran, R. S. (2007) Detection of single ion channel activity on a chip using tethered bilayer membranes. Langmuir 23, 2924-7. (78) Standen, N. B., Gray, P. T. A., and Whitaker, M. J. (1994) Microelectrode techniques, The Plymouth workshop handbook 2nd ed., Company of Biologist, Cambridge, UK.

PAGE 152

152 (79) He, L., Robertson, J. W. F., Li, J., Karche r, I., Schiller, S. M., Knoll, W., and Naumann, R. (2005) Tethered bilayer lipid membranes based on monolayers of Thiolipids mixed with a complementary dilution molecule. 1. incorporation of channel peptides. Langmuir 21, 11666-11672. (80) Cornell, B. A., BraachMaksvytis, V. L. B., King, L. G., Osman, P. D. J., Raguse, B., Wieczorek, L., and Pace, R. J. (1997) A biosensor that uses ion-channel switches. Nature 387, 580-583. (81) Schiller, S. M., Naumann, R., Lovejoy, K., Kunz, H., and Knoll, W. (2003) Archaea Analogue Thiolipids for Tethered bilayer li pid membranes on ultrasmooth gold surfaces. Angewandte Chemie International Edition 42, 208-211. (82) Schreiber, F. (2000) Structure and growth of self-assembling monolayers. Progress in Surface Science 65, 151-257. (83) Krishna, G., Schulte, J., Cornell, B. A., Pace, R. J., and Osman, P. D. (2003) Tethered bilayer membranes containing ionic reservoirs: Selectivity and conductance. Langmuir 19, 2294-2305. (84) Kunze, J., Leitch, J., Schwan, A. L., Faraghe r, R. J., Naumann, R., Sc hiller, S., Knoll, W., Dutcher, J. R., and Lipkowski, J. (2006) New method to measure packing densities of self-assembled Thiolipid monolayers. Langmuir 22, 5509-5519. (85) Mathai, J. C., Sprott, G. D., and Zeidel, M. L. (2001) Molecular mechanisms of water and solute transport across Archaeb acterial lipid membranes. J. Biol. Chem. 276, 2726627271. (86) Montal, M., and Opella, S. J. (2002) The structure of the M2 channel-lining segment from the nicotinic acetylcholine resceptor. Biochimica et Biophysica Acta 1565, 287-293. (87) Montal, M., Montal, M., and Tomich, J. ( 1990) Synporins-synthetic pr oteins that emulate the pore structure of bi ological Ionic channels. Proc. Natl. Acad. Sci. USA 87, 6929-6933. (88) Ferrer-Montiel, A. V., and Montal, M. (1993) A negative charge in the M2 transmembrane segment of the neuronal alpha 7 acetylcholine receptor increases permeability to divalent cations. Federation of European Biochemical Societies 324, 185190. (89) Opella, S. J., Marassi, F. M., Gesell, J. J ., Valente, A. P., Kim, Y., Oblatt-Montal, M., and Montal, M. (1999) Structures of the M2 channel-lining Segments from nicotinic acetylcholine and NMDA receptors by NMR Spectroscopy. Nature Structural Biology 6, 374-379. (90) Gauger, D., Binder, H., Vogel, A., Selle, C., and Pohle, W. (2002) Comparative FTIRspectroscopic studies of the hydration of diphytanoylphosphatidylcholine and ethanolamine. Journal of Molecular Structure 614 211-220.

PAGE 153

153 (91) Hogrefe, H. H., Cline, J., Youngblood, G. L., and Allen, R. M. (2002) Creating randomized amino acid libraries with the QuikChange (R) multi site-directed mutagenesis kit. Biotechniques 33, 1158-+. (92) Blount, P., Sukharev, S. I., Moe, P., and Kung, C. (1997) Mechanosensitive channels of E-coli: A genetic and molecular dissection. Biological Bulletin 192, 126-127. (93) Levin, G., and Blount, P. (2004) Cysteine scanning of MscL transmembrane domains reveals residues critical for mechanosensitive channel gating. Biophysical Journal 86, 2862-2870. (94) Iscla, I., Levin, G., Wray, R., and Blount, P. (2007) Disulfide Trapping the Mechanosensitive Channel MscL into a gating-transition state. Biophysical. Journal. 92 1224-1232. (95) Britten, R. J., and McClure, F. T. (1962) The Amino Acid Pool in Escherichia Coli. Bacteriol Rev 26, 292-335. (96) Blount, P., Sukharev, S. I., Schroeder, M. J., Nagle, S. K., and Kung, C. (1996) Single residue substitutions that change the gating pr operties of a mechanosensitive channel in Escherichia coli. Proc Natl Acad Sci U S A 93, 11652-7. (97) Miller, C., and White, S. H. (1986) Ion Channel Reconstitution Plenum Press, New York. (98) Hauser, H., and Gains, N. (1982) Spontan eous vesiculation of phos pholipids: simple and quick method of forming unilamellar vesicles. Proc Natl Acad Sci U S A 79, 1683-1687. (99) Hamill, O. P., Marty, A., Neher, E., Sa kmann, B., and Sigworth, F. J. (1981) Improved patch-clamp techniques for hi gh-resolution current recording from cells and cell-free membrane patches. Pflugers Archiv-European Journal of Physiology 391 85-100. (100) Byrne, J. H., and Schults, S. G. (1994) An introduction of membrane transport and bioelectricity Raven Press. (101) Ruthe, H. J., and Adler, J. (1985) Fusion of bacterial spheroplasts by electric-fields. Biochimica Et Biophysica Acta 819, 105-113. (102) Terrettaz, S., Ulrich, W. P ., Vogel, H., Hong, Q., Dover, L. G., and Lakey, J. H. (2002) Stable self-assembly of a protein engineering scaffold on gold surfaces. Protein Science 11, 1917-1925. (103) Lei, S. B., Tero, R., Misawa, N., Yama mura, S., Wan, L. J., and Urisu, T. (2006) AFM charaterization of gramicidin-A in teth ered lipid membrane on silicon surface. Chem. Phys. Letters 429, 244-249.

PAGE 154

154 (104) Hwang, L. Y., Gotz, H., Knoll, W., Hawker C. J., and Frank, C. W. (2008) Preparation and characterization of Gl ycoacrylate-based polymer-t ethered lipid bilayers on Benzophenone-modified substrates. Langmuir 24, 14088-14098. (105) Leahmann-Horn, F., and Jurkat-Rott, K. (2003) Nanotechnology for neuronal ion channels. J. Neruo. NeruoSurg. Psychiratry 74 1466-1476. (106) Hombie, F., Cabiaux, V., and Ruysschaert, J.-M. (1998) Channel or channel like Activity associated with pore forming proteins and peptides Molecular Microbiology 27, 12611263. (107) Kessel, A., Shental-Bechor, D., Haliloglu, T., and Ben-Tal, N. ( 2003) Interactions of hydrophobic peptides with lipid bilayers: Monte Carlo Simulations with M2. Biophysical Joural 85, 3431-3444. (108) Miyazawa, A., Fujiyoshi, Y., and Uwin, N. (2003) Structure and gating mechanism of the Acetylcholine receptor pore. Nature 423, 949-955. (109) Saiz, L., and Klein, M. (2005) The Transmembrane domain of the Acetylcholine receptor: Insights from simulations on synthetic peptide models. Biophysical Journal 88 959. (110) Iscla, I., Levin, G., Wr ay, R., Reynolds, R., and Blount, P. (2004) Defining the physical gate of a mechanosensitive channel, Msc L, by engineering metal-binding sites. Biophysical Journal 87, 3172-3180. (111) Kocer, A., Walko, M., Meijberg, W., and Feringa, B. L. (2005) A light-actuated nanovalve derived from a channel protein. Science 309, 755-758. (112) Keizer, H., Dorvel, B., Andersson, M., Fi ne, D., Price, R., Long, J., Dodabalapur, A., Koper, I., Knoll, W., Andeson, P., and Du ran, R. (2007) Functional ion channels in tetherd bilayer membranes-implications for biosensors. ChemBioChem 8, 1246-1250. (113) Blount, P., Iscla, I., Moe, P. C., Li, Y., and Owen, P. H. (2007) MscL: The bacterial Mechanosensitive channel of large conductance, in Current Topics in Membranes pp 201-233, Academic Press. (114) Steinbacher, S., Bass, R., Strop, P., Rees, D. C., and Owen, P. H. (2007) Structures of the Prokaryotic mechanosensitive channels MscL and MscS, in Current Topics in Membranes pp 1-24, Academic Press. (115) Sukharev, S. I., Martinac, B., Arshav sky, V. Y., and Kung, C. (1993) 2 Types of Mechanosensitive channels in the Escherichia-Coli cell-envelope solubilization and functional reconstitution. Biophysical Journal 65, 177-183. (116) Blount, P. (2003) Molecular mechanisms of mechanosensation: Big lessons from small cells. Neuron 37, 731-734.

PAGE 155

155 (117) Stewart, E. J., Aslund, F., and Beckwit h, J. (1998) Disulfide bond formation in the Escherichia coli cytoplasm: an in vivo role reversal for the thioredoxins. Embo Journal 17, 5543-5550. (118) Kadokura, H., Katzen, F., and Beckwith, J. (2003) Protein disulfide bond formation in prokaryotes. Annual Review of Biochemistry 72, 111-135. (119) Liu, A. L. (2006) Advances in Planar Lipid Bilayers and Liposomes Vol. 1, 3 ed., Elsevier, London. (120) Tien, H. T., and Ottova, A. L. (2001) Th e lipid bilayer concept and its experimental realization: from soap bubbles, kitchen sink, to bilayer lipid membranes. Journal of Membrane Science 189, 83-117. (121) Tanaka, M., and Sackmann, E. (2005) Po lymer-supported membranes as models of the cell surface. Nature 437, 656-663. (122) Troiano, G. C., Stebe, K. J., Raphael, R. M., and Tung, L. (1999) The effects of Gramicidin on Electroporation of lipid Bilayers. Biophys. J. 76, 3150-3157. (123) Hirano, A., Wakabayashi, M., Matsuno, Y ., and Sugawara, M. (2002) A single-channel sensor based on Gramicidin controlled by molecolar recognition at bilayer lipid membranes conyaining receptor. Biosensors Bioelectronics 18 973-983. (124) Fukasawa, T., Miyamoto, T., Fujiyama, R., Okada, T., and Sato, T. (2001) Reconstituted ion channels of frog fungif orm Papilla cell membrane. Zoological Science 18, 299-307. (125) Tieleman, D. P. (2004) The molecular basis of electroporation. BioMed Central 10, 1-12. (126) Chen, C., Smye, S. W., and Robinson, M. P. (2006) Membrane electroporation theories: a review. Med. Biol Eng Comput 44, 5-14. (127) Disalvo, E., Lairion, F., Martini, F., Almaleck, H., Diaz, S., and Gordillo, G. (2004) Water in biological membranes at interfaces: Does it play a functional role? The Journal of the Argentine Chemical Society 92 1-22. (128) Tarek, M. (2005) Membrane electroporation: A molecular dynamics simulation. Biophys. J. 88, 4045-4053. (129) Gawrisch, K., Ruston, D., Zimmerberg, J., Parsegian, V. A., Rand, R. P., and Fuller, N. (1992) Membrane dipole potentials, hydrati on forces, and the or dering of water at membrane surfaces. Biophysical. Jounrnal. 61, 1213-1223. (130) Meyer, G. R., Gullingsrud, J., Schulten, K., and Martinac, B. (2006) Molecular dynamics study of MscL interactions with a curved lipid bilayer. Bi ophysical Journal 91, 16301637.

PAGE 156

156 (131) Pasenkiewicz-Gierula, M ., Murzyn, K., Rog, T., and Cazp lewski, C. (2000) Molecular dynamics simulation studies of lipid bilayer systems. Acta Biochimica Polonica 47, 601611. (132) Wong, B. Y., and Faller, R. (2007) Phase behavior and dynamic heterogeneities in lipids: A coarse-grained simulation st udy of DPPC-DPPE mixtures. Biochimica et Biophysica Acta (BBA) Biomembranes 1768, 620-627. (133) Sukharev, S. I., Blount, P., Martinac, B., and Kung, C. (1997) Mechanosensitive channels of Escherichia coli: the MscL gene, protein and activities. Annu. Rev. Physiol 59, 633657. (134) Booth, I. R., and Louis, P. (1999) Managing hypoosmotic stress: aquaporins and mechanosensitive channels in Escherichia coli. Curr. Opin. Microbiol. 2, 166-169. (135) Levina, N., Totemeyer, S., Stokes, N. R., L ouis, P., Jones, M. A., and Booth, I. R. (1999) Protection of Escherichia coli. cells agains t extreme turgor by activation of MscS and MscL mechnosensitive ion cha nnels: identification of genes required for MscS activity. EMBO J. 18, 1730-1737. (136) Sukharev, S. I., Blount, P., Martinac, B., Blattner, F. R., and Kung, C. (1994) A largeconductance mechanosensitive channel in E. Coli. encoded by MscL alone. Nature 368 265-268. (137) Sukharev, S. I., Blount, P., Martinac, B., Blattner, F. R., and Kung, C. (1994) A LargeConductance Mechanosensitive Channel in E. Coli Encoded by Mscl Alone. Nature 368 265-268. (138) Arkin, I. T., Sukharev, S. I., Blount, P ., Kung, C., and Brunger, A. T. (1998) Helicity, membrane incorporation, orientation and th ermal stability of the large conductance mechanosensitive ion channel from E-coli. Biochimica Et Biophysica ActaBiomembranes 1369, 131-140. (139) Blount, P., Ou, X., Hoffma n, R. J., and Kung, C. (1998) Characterization of randomly generated mutants of MscL, a mechanosensi tive channel in bacteria, identifies a functional region of the protein. Biophysical Journal 74, A324-a324. (140) Bartlett, J. L., Levin, G., and Blount, P. (2004) An in vivo assay identifies changes in residue accessibility on mech anosensitive channel gating. Proceedings of the National Academy of Sciences of th e United States of America 101, 10161-10165. (141) Corry, B., Rigby, P., Liu, Z. W., and Martinac, B. (2005) Conformational changes involved in MscL channel gating measured using FRET spectroscopy. Biophysical Journal 89, L49-L51. (142) Okada, K., Moe, P. C., and Blount, P. (2002) Functional design of bacterial m echanosensitive channels Comparisons and contrasts illuminated by random mutagenesis. Journal of Biological Chemistry 277 27682-27688.

PAGE 157

157 (143) Liu, Z. W., and Martinac, B. (2006) MscL adaptation to su stained membrane stretch in liposomes under different amphipaths. Acta Pharmacologica Sinica 27, 196-197. (144) Exerowa, D., Todorov, R., and Nikolov, L. (2004) Amphiphile bilayer films from DPPC: bilayer lipid membranes and Newton black films. Colloids and Surfaces 250 195-201. (145) Zhang, P. C., Keleshian, A. M., and S achs, F. (2001) Voltage-induced membrane movement. Nature 413, 428-432. (146) Needham, D., and Hochmuth, R. M. (1989) Electro-mechanical perm eabilization of lipid vesicles role of membrane tension and compressibility. B iophysical Journal 55, 10011009. (147) Dorvel, B. R., Keizer, H. M., Fine, D., Vuorinen, J., Dodabalapur, A., and Duran, R. S. (2007) Formation of tethered bilayer lipid membranes on gold surfaces: QCM-Z and AFM Study. Langmuir 23, 7344-7355. (148) Baba, T., Toshima, Y., Minamikawa, H ., Hato, M., Suzuki, K., and Kamo, N. (1999) Formation and characterization of planar lipid bilayer membranes from synthetic phytanyl-chained glycolipids. Biochimica et Biophysica Acta (BBA) Biomembranes 1421, 91-102. (149) Baba, T., Minamikawa, H., Hato, M., a nd Handa, T. (2001) Hydration and Molecular Motions in Synthetic Phytanyl-Chain ed Glycolipid Vesicle Membranes. Biophysical Journal 81, 3377-3386. (150) Blount, P., Sukharev, S. I., Moe, P. C., Na gle, S. K., and Kung, C. (1996) Towards an understanding of the structural and functiona l properties of MscL, a mechanosensitive channel in bacteria. Biology of the Cell 87, 1-8. (151) Simon, S. A., and McIntosh, T. J. (1986) Depth of water penetration into bilayers. Methods Enzymol. 127, 511-521. (152) Sukharev, S., Durell, S. R., and Guy, H. R. (2001) Structural m odels of the MscL gating mechanism. Biophysical Journal 81, 917-936. (153) Berrier, C., Besnard, M., Ajouz, B., Coulombe, A., and Ghazi, A. (1996) Multiple mechanosensitive ion channels from Escherichia coli, activated at different thresholds of applied pressure. Journal of Membrane Biology 151 175-187. (154) Raucci, F. J. (2004) in Molecular Biology pp 64, University of Florida, Gainesville (155) Wilson, D. (2006) in Chemistry pp 81, University of Florida, Gainesville. (156) Keizer, H. M., Andersson, M., Chase, C., Laratta, W. P., Proemsey, J. B., Tabb, J., Long, J. R., and Duran, R. S. (2008) Prolonged stoc hastic single ion cha nnel recordings in Slayer protein stabilized lipid bilayer me mbranes. Colloids and Surfaces B: Biointerfaces 65, 178-185.

PAGE 158

158 (157) Latorre, R., Coronado, R., and Vergara, C. (1984) K+ channels gated by voltage and ions. Annual Review of Physiology 46, 485-495. (158) Oiki, S., Danho, W., Madison, V., and Montal M. (1988) M2 dealta, a candidate for the structure lining the ionic channel of the nicotinic cholinergic receptor. Proc. Natl. Acad. Sci. USA 85, 8703-8707. (159) Sukharev, S. I., Sigurdson, W. J., Kung, C., and Sachs, F. (1999) Energetic and spatial parameters for gating of the bacterial la rge conductance mechanosensitive channel, MscL. J. Gen. Physiol. 113 525-540. (160) Yin, P., Burns, C. J., Osman, P. D. J., and Cornell, B. A. (2003) A tethered bilayer sensor containing alamethicin channels and its de tection of amiloride based inhibitors. Biosensors & Bioelectronics 18, 389-397.

PAGE 159

159 BIOGRAPHICAL SKETCH Danyell Wilson was born in Norfolk Virginia in 1981. She attended Booker T. Washington High School where she graduated in the top 15% of her class in 1999. Upon entering Lincoln University, she earned a full sc holarship by com pleting the Lincolns Advance Science and Engineering Reinforcement (LASER ) summer program with a GPA above 3.00 out of 4.00. She participated in several research in ternships throughout her undergraduate career and received the Minority Assess to Research Career s (MARC) National Institute of Health (NIH) fellowship her junior year. In 2003, she graduate d cum laude from Lincoln University with her Bachelor of Science in Chemistry and a minor in Anthropology. Ms. Wilson then received the Alliances for Graduate Education and the Prof essoriate (AGEP) fellowship, and the Grinter fellowship from the University of Floridas Chemis try department where she began this project. In 2006, Ms. Wilson received her Mast ers of Science degree from the University of Florida with a focus in analytical chemistry. She changed research division and joined the biochemistry division and continued her research on single ion channel studies. In 2009 Ms. Wilson, received her PhD in Chemistry from the University of Florida with a focus in Biochemistry. Ms. Wilsons long term career goals are to beco me involved in science education policy.