1 INORGANIC BINDING PEPTIDES DESIGNED BY PHAGE DISPLAY TECHNIQUES FOR BIOTECHNOLOGY APPLICATIONS By CHIH WEI L IAO A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQ UIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2010
2 2010 ChihWei Liao
3 T o my family, friends, and lab menbers
4 ACKNOWLEDGMENTS First, I would like to thank Dr. Laurie Gower, my advisor, for providing me the chance to start the new research field, molecular recognition in organic inorganic hybrid materials, in her group. During my study in the United States she inspired me to have the creative thinking and taught me how to be a scientist. I also want to express m y gratitude to my committee members, Dr. Ronald Baney, Dr. Hassan El Shall, Dr. Kevin Powers, and Dr. Joanna Long. Thank them to induce me how to figure out the strate gies for solving the problems in my research. In addition, I am so glad to to work with nice and talented teammates, Taili Thula, Sang Soo Jee, Douglas Rodriguez, Mark Bewernitze, Yawen Yeh, and Myong Hwa Lee. During my study, we shared research experiences with each other. They always kindly gave me support when I needed their help in my rese arch. I also would like to thank Seonhoo Kim in Dr. Nortons group for fabricating the indium oxide devices for us. With their technical support, we could have the enough research resource for the development of reversibly electro activated biosensors. I appreciate Dr. Paul Gulig and Julio Martin in t he department of microbiology for advis ing me in the molecular biology techniques and provide the fluorescence microscope. In addition, I must ment ion our collaborators in Univer s i ty of Washin g ton, Dr. Mehm et Sarikaya and Dr. Candan Tamerler. In the beginning of my research, they gave a chance to learn phage display techniques in their group. Be c a us e of their help, we can develop phage display techniques in some practical applications. Finally, I need to sa y many thanks to my fa mily, and my friends from the b ottom of my heart. For my family and friends, they always accompany with me mentally, even if they are far away from me. When I feel frustrating or stressful, they always supply spiritual energy to chee r
5 me up. For my love, Yi Ling Lin, she is the best supporter. She took care of my emotion and every thing in the living. Thus, I can focus on finishing this dissertation.
6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ...............................................................................................................4 LIST OF TABLES ...........................................................................................................................8 LIST OF FIGURES .........................................................................................................................9 LIST OF ABBREVIATIONS ........................................................................................................13 ABSTRACT ...................................................................................................................................15 CHAPTER 1 INTRODUCTION ..................................................................................................................17 1.1 Overview ...........................................................................................................................17 1.2 Molecular R ecognition of B iomacromolecules ................................................................19 1.3 Challenges of Biopolymers in Biomimetics .....................................................................21 1.4 Small Peptides Based on Combinatorial Techniques .......................................................21 1.5 Newly Designed Binding Peptides through Bioinformatics .............................................25 1.6 Applications of Inorganic Binding Peptides .....................................................................28 2 I NORGANIC BINDING PEPTIDES BASED ON PHAGE DISPLAY FOR B IOSENSOR APPLICATIONS .............................................................................................43 2.1 Motivation .........................................................................................................................43 2.2 Background and Significance ...........................................................................................44 2.2.1 D efinition of B iosensors .........................................................................................45 2.2.2 Immobilization Methods of Biosensing Materials .................................................46 2.2.3 A Challenge in Biosensor Development ................................................................48 2.3 Materials and Methods .....................................................................................................50 2.3.1 Materials .................................................................................................................50 18.104.22.168 Dodecapeptide p hage display peptide l ibrary (Ph.D.12) ............................50 22.214.171.124 E. coli. ER2738 host strain ...........................................................................50 126.96.36.199 Baterial culture medium ...............................................................................51 188.8.131.52 Stock solution ...............................................................................................52 184.108.40.206 Buffer solutions ............................................................................................52 220.127.116.11 Elution buffers ..............................................................................................52 2.3.2 Methods ..................................................................................................................53 18.104.22.168 Fabrication of Indium Zinc Oxide (IZO) as a target material ......................53 22.214.171.124 Phage displa y protocol .................................................................................55 126.96.36.199 Blue white screening ....................................................................................59 188.8.131.52 The determination of e xpressed 12mer p eptides from M13 phage DNA ...63 184.108.40.206 Immunofluorescence (IF) m icroscopy e xperiment ......................................64
7 220.127.116.11 Enzyme linked immunosorbent assay (ELISA) ...........................................65 18.104.22.168 Calculation of surface coverge of a phage clone on substrate sheets ...........66 2.4 Results and Discussions ....................................................................................................66 3 PARTICLE SEPARATION USING INORGANIC BIND ING PEPTIDES DESIGNE D BY PHAGE DISPLAY TEC HNIQUES ................................................................................86 3.1 Motivation .........................................................................................................................86 3.2 Background and Significance ...........................................................................................87 3.2.1 Liberation ...............................................................................................................87 3.2.2 Concentration .........................................................................................................88 3.2.3 Challenges of Current Mining Processing Technologies .......................................90 3.2.4 Role of Inorganic Binding Peptides in Separation of Minerals ..............................92 3.3 Materials and Methods .....................................................................................................93 3.3.1 Materials .................................................................................................................93 3.3.2 Methods ..................................................................................................................93 3.4 Results and Discussions ....................................................................................................95 4 M13 PHAGE AMPHIPHILE AS COLLECTORS FOR THE SPECIFIC SEPARATION OF MINERAL PARTICLES ................................................................................................112 4.1 Motivation .......................................................................................................................112 4.2 Background and Significance .........................................................................................112 4.3 Materials and Methods ...................................................................................................115 4.3.1 Materials ...............................................................................................................115 4.3.2 Methods ................................................................................................................115 4.4 Results and Discussions ..................................................................................................118 5 CONCLUSIONS ..................................................................................................................133 LIST OF REFERENCES .............................................................................................................136 BIOGRAPHICAL SKETCH .......................................................................................................146
8 LIST OF TABLES Table page 11 Consensus peptide sequences that have binding affinity to the various inorganic materials .............................................................................................................................32 21 P hage titers in the I ZO system with low pH elution buffer ...............................................72 22 P hage titers in the I ZO system with high salt elution solution ...........................................72 31 Phage titers for the phages selected from dolomite particles ...........................................101 32 Charge of Histidine in peptide sequence QTLPLPLTIAHP and zeta potential of dolomite surfaces in 3.3% PC buffer at pH 4.0, pH 7.4, and pH10.0 ..............................101 33 Charge of Aspartic acid in peptide sequence GFASDPSSSPWT and Zeta potential of dolomite surfaces in 3.3% PC buffer at pH 4.0, pH 7.4, and pH10.0 ..............................101 34 Phage titers for phage selected from Forur Corner francolite pebbles .............................102 35 Ph age titers for phage selected from South Fort Meade francolite pebbles .....................102 36 C hemical properties of peptide sequences selected from FC and SFM francolite pebbles ..............................................................................................................................103 37 The comparison of peptide sequences selected from francolite and dolomite powders ..103 41 Summary for the surface coverage of selecte d phage clones on francolite and dolomite ...........................................................................................................................125 42 Contact angle on francolite and dolomite in 3.3% PC buffer solution at pH7.4 ..............125 43 Contact angle on francolite and dolomite in distilled water at pH7.4 ..............................125 44 ICP AES analysis for the pure minerals ...........................................................................126 45 ICP AES analysis of mixed minerals (Francolite:Dolomite=1:1) at pH7.4 .....................126 46 ICP AES analysis of mixed minerals (Francolite:Dolomite=3:1) at pH7.4 .....................126 47 ICP AES analysis of mixed minerals (Francolite:Dolomite=1:3) at pH7.4 .....................127
9 LIST OF FIGURES Figure page 11 TEM images for n ano m agnetosomes particles within the bacterium Aquaspirillum Magnetotacticum ................................................................................................................33 12 Images of t he spicules with start shaped tips on the body of the spong Rosella racovitzea ..........................................................................................................................33 13 Images for the n acre structure of mother of pearl mollusks. .............................................34 14 Schematic for t he generatio n of random peptide librar ies. ................................................34 15 The configuration of a filamentous M13 phage ................................................................35 16 S chematic of inserted peptide str uctures within pIII proteins ..........................................35 17 Schematic of the p hage display (PD) technique ...............................................................36 18 Schematic of the blue white screen ...................................................................................36 19 M echanism of the blue white screen .................................................................................37 110 M etaboli sm of lactose Galactosidase ................................................37 111 M galactosidase .........................................................38 112 Image of a phage clone with the specific binding affinity to platinum. ............................38 113 S chematic of the bioinformatics approach .........................................................................39 114 The scoring matrix for quartz binding peptides .................................................................39 115 The morphology of gold nanocrystal s ...............................................................................40 116 S chematic of the formation of silver nanoparticles in silver nitrate solution ....................40 117 TEM analysis of silver nanoparticles. ................................................................................41 118 Schematic for the spatial control of an array of silver nanocrystals using silver binding peptides ................................................................................................................41 119 Images of gold quantum dots ............................................................................................42 120 Images of smectic ordered self supporting Auvirus films ................................................42 21 Schematic of the electro releasing mechanism using electro activated peptides. .............73
10 22 S chematic of the market size and potent ial application s of biosensors in the worldwide market .............................................................................................................73 24 Schematic for the deposition of IZO on a sapphire sheet ..................................................74 25 Design of the device for releasing the phage clones ..........................................................74 26 Serial dilution of phage samples ........................................................................................75 27 N terminal sequence of random 12mer peptide gIII fusion for M13 phage DNA in Ph.D.12 phage library .......................................................................................................75 28 Schematic of immunofluorescence analysis .....................................................................76 29 Schematic of enzyme linked immunosorbent assay .........................................................76 210 Measurement of the surface coverage of a phage clone on the IZO surface .....................77 211 12mer amino acid sequence s of selected phages from InZnO with low pH elution ........77 212 Images of ELISA plate containg phage clones eluted from IZO with low pH elution buffer. .................................................................................................................................78 213 UVVisible absorbance of the enzymatic substrate solution .............................................79 214 Images of the immuofluorescence analysis for t he negative control experiment ..............80 215 IF analysis for t he phage clone TKNMLSLPVGPG .........................................................80 216 IF analysis of t he phage clone MNRPSPPLPLWV ...........................................................81 217 Photography of t he etching InZnO with using the low pH elution buffer .........................81 218 12mer amino acid sequences of selected phages from InZnO with high salt elution ......82 219 I mage of ELISA plate containing 6 phage clones from the third biopanning with high salt elution buffer after 10 minute development time ........................................................82 220 UVVisible absorbance of the enzymatic substrate solution from ELISA plate in Figure 2 19 .........................................................................................................................83 221 IF imag es for t he phage clone SHAPDSTWFALF from the low pH elution ....................83 222 Releasing test for the mixture of phage clones. .................................................................84 223 Electro releasing test for the phage cloneNMTMSFPTYPIA. .........................................85 31 Schematic of a peptide amphiphile as a flotation agent ...................................................103
11 32 A curve representing the relationship between recovery and grade of concentrate ........104 33 S chematic for the flotation process ..................................................................................104 35 T ypes of collectors used in the flotation concentration ...................................................105 36 Summary of expressed 12 mer dolomite binding peptide sequences. The letters correspond to the commonly used one let ter code for the 20 amino acids. .....................106 37 Images of the phage clone ADYFTARPGPIT on dolomite particles at the objective magnification 20X and 40X .............................................................................................106 38 Classification of phages with expressed 12mer peptide sequences into three categories based on the surface coverage on dolomite ....................................................107 39 The binding affinity of t hree representatives of phages on dolomite particles ................107 310 Zeta potential for francolite/dolomite powders in the PC buffer .....................................108 311 Images of the phage clone QTLPLPLTIAHP on dolomite surfaces. ..............................108 312 Images of phage clone GFASDPSSSPWT on dolomite surfaces. ..................................109 313 Four Corner francolite binding peptide sequences ..........................................................109 314 South Fort Meade francolite binding peptide binding sequences ....................................110 315 The surface coverage of phage clones with expressed francolite binding peptides on francolite and dolomite respectively ................................................................................110 316 Images of the phage clone with ex pressed peptide WSITTYHDRAIV on francolite particles and dolomite particles. ......................................................................................111 41 S chematic of froth flotation using M13 phage with expressed francolite binding peptides as collectors for separation of francolite particles from dolomite particles ......127 42 The set up for the bench top flotation. ............................................................................128 43 The meth od of the contact angle measurements. .............................................................128 44 Images of francolite particles modified with phage clones selected from francolite ......129 45 Images of dolomite particles modified with phage clones selected from dolomite .........129 46 The recovery rate of pure francolite and dolomite particles versus dose of fatty acid as a collector in distilled water at pH 7.4 .........................................................................130 47 The recovery of pure francolite and dolomite particles versus dose of phage clone WSITTYHDRAIV as a collector in 3.3% PC buffer at pH 7.4 .......................................130
12 48 The recovery of pure francolite and dolomite particles versus dose of phage clone TNSNWTPFWPLP as a collector in 3.3%PC buffer at pH 7.4 .......................................131 49 The recovery of pure francolite and dolomite particles versus dose of phage clone SSMTHQHARVDT as a collector in 3.3%PC buffer at pH 7.4 ......................................131 410 Comparison of recovery rate of francolite showing all three representative phage clones as collectors respectively at pH 7.4 ......................................................................132 411 The hydrophobicity of francolite through phage clone WSITTYHDRAIV in 3.3%PC buffer solution at pH7.4. ..................................................................................................132
13 LIST OF ABBREVIATION S AA Amino acids ALP A lkaline P hosphatase BG D galactosidase BioLBL B iomimetic L ayer by L ayer BLAST Basic Local Alignment Search Tool BSA B ovine S erum A lbumin B/W Screen Blue White Screen CSD C ell S urface D isplay DI D eionized ELISA E nzyme L iked I mmunosorbent A ssay FITC F luorescein I sothiocyanat e GC G as C hromatography HPLC H igh P erformance L iquid C hromatography HRP H orseradish P eroxidae I CBR The Interdisciplinary Center for Biotechnology Research ICP AES Inductively Coupled Plasma Atomic Emission Spectrometer IF I mmunofluor scence IgG I mmunoglobulins IPTG I sopropyl D 1thiogalactopyranoside I TO Indium T in O xide IZO Indium Znic Oxide MHDASAM 16mercaptohexadecanoic A cid S elf A ssembly M onolayer MIMIC M icromoulding in C apillaries MM M agnetosaome M embrane
14 MTB M agnetotactic B acterium NCBI National Center for Biotechnology Information PD P hage D isplay PDMS P olydimethyl S iloxane PEG Poly (ethylene glycol) POM P olarized O ptical M icroscopy PZT P iezoelectric T ransducer QCM Q uartz C rystal M icrobalance QCM D Q uartz C rystal M icrobalanceDissipation RI R ef ractive I ndex SPR Surface P lasmon R esonance TIRF T otal R eflection F luorescence TMB 3, 3, 5, 5 tetramethylbenzidine
15 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the R equirements for the Degree of Doctor of Philosophy INORGANIC BINDING PEPTIDES DESIGNED BY PHAGE DISPLAY TECHNIQUES FOR BIOTECHNOLOGY APPLICATIONS By ChihWei Liao December 2010 Chair: Laurie B. Gower Major: Materials Science and Engineering Biomacromol ecules play an important role in the control of hard tissue structure and function via specific molecular recognition interaction s between proteins of the matrix and inorganic species of the biomineral phase During the construc tion of the tissue biomacro molecules are usually folded into a certain comformation, analogous to a lock for fitting with other proteins or smaller molecules as a key Currently the rational design of molecular recognition in biomacro molecules is still hard to accomplish becau se the protein conformation is too complex to precisely predict based on the existing conformational information of proteins found in biological systems In the past two decades, the combinatorial approach ( e.g. phage display techniques) has been used to s elect short binding peptides with molecular recognition to a n inorganic target material without a prior knowledge of the amino acid sequence required for the specific binding The technique has been referred to as biopanning because bacteriophage s are us ed to screen for peptides that exhibit strong binding to a target material of interest. In this study, two diverse applications were chosen to demonstrate the utility of the biopanning approach. In one project, phage display techniques were used to pan for Indium Zinc Oxide ( InZnO ) binding peptides to serve as linkers between tran s ducer devices and biosensing
16 element s for demonstrat ion of the feasibility of reversib ly electro activated biosensors. The amorphous InZnO with its homogeneous surface, led to three consensus peptide sequences AGFPNSTHSSNL, SHAPDSTWFALF, and TNSSSQFVVAIP. In addition, it was demonstrated that some selected phage clones of the InZnO binding peptides were able to be released from the InZnO surface after applying a voltage of 1400 mV on an electroactivated releasing device. In the second project phage displ a y tec hni qu es were used to se lect phage clones that bind specifically to francolite mineral in order to achieve separation of francolite particles from dolomitic partic les within Florida phosphate ore A phage clone with a 12mer francolite binding peptide of WSITTYHDRAIV was able to concentrate the content of francolite from 25% to 42% in a bench top flotation process of mixed minerals The first system demonstrates an advanced technology application of the biopanning approach for the development of novel biosensors, while the second system demonstrates application of the biotec hnology approach to a commodity industry
17 CHAPTER 1 INTRODUCTION 1.1 Overview In the pa st two decades, phage display techniques have brought a new dimension to the design of advanced materials Beyond this field, however, there still exist s abundant pot ential applications of organic inorganic hybrid systems based on phage display techniques. This dissertation seeks to explore the diversity of these phage display techniques by focusing on two main projects, the first of which is centered on creating advanced materials for biosensor applications, whereas the second project seeks to evaluate the feasibility of these techniques for industrial production of commodities. In the advanced materials project applications of phage display techniques were explored for the development of electroactive peptides which could be us ed for self cleaning biosen sors; the commodities project investigates the preferential separation of mineral particles based on surface modification of inorganic materials using inorg anic binding peptides In the first application of phage display techniques, we considered that biosensing activities of biosensors can suffer because the receptors become clogged by analyte, whereby loss of detection usually causes underestimation in the measurement of concentration of analyte. Thus, self cleaning sensing components may be considered one of approach to provide biosensors with an effectively continuous detection. Thus, we proposed the concept of reversibly electro activated peptide linkers to reversibly immobilize bioreceptors in close proximity to the surface of a transducer device for a self cleaning sensing component With this idea in mind we hypothesized that reversibly electro activated peptides designed by phage display techniques could not only display specific binding affinity to an inorganic material of interest, but could also be released after applying a voltage to an inorganic surface. First we ultilized
18 immunofluorescence analysis and enzyme linked immunosorbent assay to evaluate the binding affinity and specificity of these peptides to inorganic materials. Then, we design e d an electro releasing device to test if these inorganic binding peptides could be released from the surface of the target material coated on the electro releasing device. With respect to the specific separation of mineral particles using inorganic bindin g peptide s selected by phage display techniques, we consider ed that current commercial surfactants used in the phosphate mining industry are not able to remove dolomite from francolite in the froth flotation process. Recently, microorganisms have been found that are able to serve as flotation agents to float valuable minerals specifically, but low recovery rate of minerals were achieved at neutral pH, and a high dose compared to conventional chemical surfactants was required Thus, we chose francolite pebbl es as a target material for biopanning with M13 phage clones in order to find francolite binding peptides that could be used as flotation agents to achieve specific separation of francolite from phosphate ore containing dolomite contaminations It was hypo thesized that the phage display system could be based on a neutral working environment of phage display biopannings and the small bindin g area of M13 phages on mineral particles might allow for a low concentration to be used. The feasibility of M13 phage amphiphiles as flotation agents in the commercial process of mineral recovery was evaluated where the some representatives of phage amphiphiles were compared to standard commodity surfactants by way of recovery abilities. In this project, immunofluorescen ce analysis was used to choose clones with specific binding affinity to francolite relative to dolomite We hypothesized that the phage body, which is relative ly hydrophobic, could enhance the adhesion of francolite particles to air bubbles for floating fr ancolite particles from phosphate ore containing dolomite particles Therefore, we evaluated the hydrophobicity of phage coat proteins
19 using captive air bubble techniques to correlate with the effect of phage coat proteins on the recovery rate of minerals. Inductively coupled plasma atomic emission spectroscopy (ICP AES) was adopted to evaluate the purity of francolite in floated minerals after a mixture of francolite and dolomite was floated by a phage clone with an expressed francolite binding peptide. 1.2 Molecular R ecognition of B iomacromolecules In biological systems biomacromolecules are usually involved in organization of molecules to hierarchical structures via their molecular specificity  to make up a myriad of different functions of soft tissues and hard tissues. For example, magnetite nanoparticles cubooctahedral single crystal of iron oxide, are aligned within Aquaspirillum Magnetotacticum  a kind of magnetotactic bacterium (MTB) (See Figure 1 1). Those magnetosomes are usually aligned in a chain within the MTB to form a permanent magnetic dipole. Thus, MTB swim along the magnetic field. For the formation mechanism of magnetosomes, the magnetosome sp ecific proteins within magnetos ome membrane (M M) are though to control the accumulation of iron, redox, and the nucleation of iron oxide  Antarctic sponge, Rosella racovitzea is another interesting example. The living environment of Rosella racovitzea is usually around 200 m under the ocean. However, green alga can sur vive in this environment s lack of light by establishing symbiotic relation s with Rosella racovitzea bodies  The poss ible explanation is that silica based spicules with star shaped tips serve as light collector s Those tips guide the light from optical silica based fibers to the outer wall of the sponge. The structure of silica based optical fibers is a layer by layer si lica shell su rounding the central core as shown in Figure 12  Some major protein components from the central core may behavior like enzymes that catalyze ions from seawater to form silica 
20 Furthermore, nacre, the structure of mother of pearl, is found at the interior of mollusk shell s Nacre is also a famous repr esentative of the hybrid ization of organic and inorganic materials in nature. The apparent characteristic of nacre with high fracture toughness and strength is highly organized aragonite tablet s separated by thin organic layer s comprised of proteins and polysaccharides [6, 7] (See Figure 13  ) Those natural biological structures ha ve inspired scientists and engineers to design human made materials for practical applications utilizing the tools of molecular recognition that biological systems exemplify Basically, the concept of molecular recognition is defined a s the relationship between lock and key host to guest molecules  Then, guest molecules key must effectively fit into the hosts lock In biochemistry, enzymes behave as the locks which fit with the desired substra tes as the keys. For biomacromolecules, locks usually mean the crevices within the protein surface created by a certain conformation of the macro molecule, or the hallow sites within the molecular aggregates. However, keys are usually small molecules captur ed partially or completely by those crevice or hollow sites. In the other words they are complementarily in molecular structure. As to the flexibility of host/guest molecules, their conformations are not necessar ily rigid. In most cases their conformatio ns are highly affected by the environment such as the change of pH, ion conductivity, and temperature. In addition, those conformations may continuously change to adjust the optimal steric arrangement during the occurrence of the host gu e st binding event. The driving force of interaction between host and guest molecules is often entropic ally driven  as discussed below The chemical interaction between host and guest molecules consist s of three components  : steric, polar and hydrophobic fit. Steric fit means the interaction range between host and guest mol ecules must be within their van der W aals radii. Furthermore, their empty space
21 approaches a minimum. Electrostatic fit represents the maximum of polar interaction s such as ionic bonding, hydrogen bonding, and aromatic ring cation bonding Finally, hydrophobic fit is the trend of association between nonpolar groups in the biophase an aqueous medium with dissolved ions and small molecules. This association tendency is defined by the density difference between host molecules and water. Thus, an association reaction becomes spontaneous because net dispersion attraction forces between host and guest atoms is larger that those between both of them separately and water while there is an entr opic gain from the rel a ese of hyd r a tion waters. 1.3 C hallenge s of Biopolymers in Biomimetics Biological tissues are constructed with genetic control under mild physiological condition s and aqueous environments. During the construction of those tissues, bi omacromolecules ( proteins, polysaccharides, polynucleotides, or lip id s) control those construction processes via their molecular recognition. In biomimetics, biomacromolecules are used to mimic the structure and biological function of proteins within the tissues. Those biomacromolecules are usually extracted from tissues follo wed by the process of purification. After the purification process, the composition of extraction may contain several biomacromolecules [11, 12] In addition, the current knowledge leading to the prediction of prote in structure and surface binding chemistry is not sufficient to perform the rational design of proteins  1.4 Small Peptides Based on Combinatorial Techniques At the start of this century, a combinatorial approach developed by the microbiology community was adopted, and that has brought a whole new dimension to the field of biomimetic engineering. The combinatorial biological techniques such as phage display (PD) [2, 1454] and cell surface display (CSD)  had bee n adopted to select the peptide sequences that preferentially bind to inorganic or organic surface s Bacteriophages are virus particles that infect
22 bateria duri ng their replication cycle. In combinatorial biological techniques, libraries need to be constructed to express random peptides on the phage or bacterial For the construction of libraries, a set of random foreign oligonucleotides are inserted into the phage genomes or bacterial plasmids, which are responsible for expressing proteins on the surface of virus or bacterial cell. After those foreign oligonu celotides fuse with the phage genomes or bacterial plasmid respectively, they express random foreign pept ides within those surface proteins (See Figure 14  ) In this study, we focused on applications of phage display techniques. Using this approach, phage display libraries are used to pan for peptides that bind to the desired target materials. In the other word, those biopanning processes are basically an affinity selection technique which selects peptides that show binding affinity to a target. In the initial phage display techniques antibodies we re usually the target materials. I n our studies, target materials were solid phase inorganic mat erials. P hage display techniques usually consist of four steps. F irst, phage libraries are constructed by inserting foreign desired gene segments into a region of the bacteriophage genome the entirety of an organism's hereditary information. Thus, random peptides can display on the surface of a bacteriophage Secondly, a phage library is incubted with a target material for capturing the phage on the surface of a target material Subse quently unbound phages are wash ed away. Then, bound phages are eluted. Expressed peptides from the DNA sequencing results of bound pha ges usually display strong and specific binding affinity to a target material. S ome commercial phage display libraries su ch as Ph .D .12TM ( New England Biolab, Inc ) consist of 2.7 x 109 electroporated sequences The type of phages used in phage display includes Ff filame ntous phages (M13, fd, and fl) and phages with the capsid shape ( Lambda and T7 ) The amplification of Ff phage family is via a
23 nonlytic propagation mechanism. That means that all components of the phage coat are exported through the inner membrane of the bacteria prior to the assembly of mature phages. In this nonlytic mechanism, the density of coat protein s of filame ntous phages is lower than phages with the capsid confugration because only proteins that can withstand export are displayed. However, the size of Ff phages is not relat ed to their DNA. Thus, the insertion of foreign DNA within the genome does not affect the size of phages significantly For Lambda and T7, their assembly is through a lytic mechanism in which the construction of the capsid occurs in the cytoplasm of cells. Thus, it is easy to display a high density of coat protein. Most phage display libraries use filamentous phage strains due to their flexible and robust properties in display, even if the Ff family has the disadvantage of a low density of coat protein. Figure 15 show s the configuration of filamentous M13 phages which are flexible length and 6 nm in diameter  Filamentous phages are composed of the major coat protein (pVIII) and m inor coat proteins (pIII, pVI, pVII, and pIX). The phage body is composed of 2700 copies of pVIII protein (50 a mino a cids (AA) in length ) with helical arrangement. Each end of a phage body is capped with two kinds of minor coa t proteins: One end consist s of pVII (5 copies, 33AA) and pIX (5 copies, 32AA). The other end includes pIII (5 copies, 406AA) and pVI (5 copies, 112AA). Phage libraries can display random peptides within pVIII using insertion of foreign olignucle otides into the gVIII genome. In genera l, the size of the expressed peptides is limited in this system. Furthermore, low and high affinity binders often can not be discriminated due to avidity effect which means the combined strength of multiple bond interactions However, this system is an ideal candidate if low affinity binders are mainly involved in studies. Currently, many phage libraries are generated to express random peptides at the amino terminus of pIII.
24 Alth ough pIII proteins only has 5 copies to cause small avidity effect, the significant change of pIII proteins such as the fusion of heterologous proteins can prevent the viral assembly and block infectivity. In this research, the adopted phage library is New England BiolabPhD.12TM which expresses random peptides in the terminus of pIII proteins. For the structure of inserted petptides in pIII proteins, there are two types current ly available as commercial filamentous phage libraries: N terminal constrained p eptide inse r t and N terminal linear peptide insert (See Figure 16  ) The phage libraries that we used express N terminal linear peptide insert in protein pIII. In gener al, 12mer peptides dont display obvious conformations However, 12mer peptides expressed in pIII proteins composed of 406 amino acids may be folded into a periodic arrangement after binding on an inorganic surface. In phage display selection (biopanni ng), which is illustrated in the schematic in Figure 1 7, a large variant of phage mutants, 109 random peptide sequences, are exposed to a desired target. Some phages with weak b onding to the target are removed with extensive washes. Then, more strongly bound phages can be eluted with a low pH solution. The eluted phages are amplified by infecting E. coli E2738 host bacterial strain, isolated, titered, and reexposed to fresh target to enrich the population of strong binders. The whole procedure is called a biopanning. Three to five biopannings are performed to evolve to the phages with strong binding affinity to targets ( arrows in Figure 17) DNA of those selected phages is sequenced to determine the peptide binding sequences that provide molecular recog nition to the target inorganic surface. During the isolation of phage clones, a blue white (B/W) screen is used for the purpose of phage isolation. The B/W screen ing method i s a molecular biology technique for the detection of successful ligation in vector based gene cloning which is based on the secretion of the enzyme galactosidase. In the PhD. 12TM M13 phage library, E. Coli ER2738 host strain does not
25 contain lac Z gene within its DNA. Thus, it cannot secrete galactosidase until lac Z within M13 pha ges fuse with its DNA to become bacteriophages (See Figure 18 (a) ). The phage clones appear as blue plaques on the LB agar plate. An example is shown in Figure 18 (b) The molecular mechanism of the B/W screen is related to the lac operon in E. coli str ain as a host cell combined with a complementary subunit vector of the phage Figure 1 9  shows galactosidase can be secreted in the presence of inducer when the recombinant DNA of E. coli host strain with the target vector still keep s the existence of lacZ gene. Otherwise, the secretion of galactosidase will be blocked if lacZ gene is disrupted by foreign DNA or repressors interact with operator. galactosidase is the intracellular enzyme that can cleave the disaccharide lactose into glucose and galactose (See Figure 110  ). However, galactosidase is colorless. How do we detect the appearance of galactosidase? In general, the indicator, Xgal, is usually chosen to detect the existence of galactosidase. Xgal is a colorless modified galactose sugar that can be catalyzed by galactosidase to produce an insoluble blue product (5bromo 4chloroindole) (See Figure 111 ). At the same time, isopropyl D 1thiogalactopyranoside (IPTG), which functions as the inducer of the lac operon, interacts with repressors to avoid t hem blocking the secretion of galactosidase. 1.5 New ly Designed Binding Peptides through Bioinformatics After phage display selection, a consensus amino acid sequence of selected peptides which show strong binding affinity to the material of in terest usu ally can be obtained. Figure 112 show s that platinum binding peptides on phage clones labeled with a f luorescein isothiocyanate ( FITC ) fluorescence probe display preferential binding affinity to the platinum area on a patterned substrate, rather than the quartz area  Beside s using phage displ ay to pan for binding peptides from the biological and organic targets such as antibodies, cell s and polymer s
26 inorganic binding peptides also ha ve been selected for some practical applications. Table 11 includes some consensus amino acid sequences of i norganic binding peptides  After post selection of phage clones, a consensus amino acid sequence of a peptide insert just guarantees it exhibit s strong binding affinity to a target of interest, rather than specific binding affinity. Thus, while some of the consensus peptide sequences display preferential binding affinity to the target of interest, other consensus peptide sequences can have binding affinity to several materials beside a target. Is there any strategy to pursue the peptides with specific binding abilities with using the biopanning approach ? Basically, a high degree of specificity is possible in biopanning through different approaches [24, 42, 46] One approach named anti selection is to take a set of clones that were determi ned to have affinity for the targeted surface, but then test the affinity of those clones with other surfaces that may be present in a given device ( e.g. using immunofluorescence), and hope to find a clone that exhibits preferential affinity for the material of interest. Another interesting approach was demonstrated by Fang et al  who used subtractive bacteriophage biopanning to identify 12 mer peptides that bind selectively, a s well as induce the precipitation of t itania but not silica. Their approach consisted of two steps of biopanning: i) removal of phage particles containing silicabinding peptides from the phage library (the subtractive step), and then ii) isolation of phage particles bearing peptides that bind strongly to titania Interestingly, the subtractive biopanning process yielded several acidic peptides enriched in hydroxyl bearing residues, while prior reports of phage display biopanning with silica and t itania t argets led to the isolation of polycationic peptides enriched in basic residues. Thus, selectivity may need to be determined according to the demands of the system (or selectively may not even be required if a device only uses the one material that was tar geted).
27 As a third option, the bioinformatics approach being developed by Sarikayas group also has the potential to design selective peptides de novo  In this method (See Figure 113  ) a set of experimentally selected peptides are categorized for their binding affinities (such as quantify ing % coverage of bound phage by immunofluoresence microscopy) and scoring matrices are defined. These include similarities within strong binding sequences and the differences between the strong and weak binders. Experimentally, this entails categorizing a relatively large number of clones as strong, medium, and weak binders. They have demonstrated the capabilities of the approach by classifying experimentally characterized quartz binding peptides and computationally designing new sequences with specific affinities, and found that binding of the computationally designed peptides correlated with their predictions with high accuracy  as described below. The following example shows the approach of bioinformatics to design a new amino acid sequence of inor ganic binding peptides. In this example, experimental ly characterized quartz binding peptides (Quartz I), BLOSUM 62, and PAM 250 were classified into strong, medium, and weak binding sets as the base respectively for developing the bioinformatic approach. Note BLOSUM 62, and PAM 250 include quartz binding peptides sequences from natural proteins. Subsequently, the data base including 1,000,000 random peptide sequences were used to compare with the strong binder set in BLOSUM 62, PAM 250, and QUARTZ I to d educe new ly design ed quartz binding peptide sequences in the scoring matrix in F igure 114  : The scoring matrix is the indication of similarity between two compared peptide sequences. Finally, those predicted peptides must be confirmed by experimental v alidation such as immunofluorscence (IF) analysis to confirm their binding affinity match es with the experiment al results. Basically, the quantity and quality of initial data determine the precision of the scoring
28 matrix. It is feasible to enhance the pr ecision of the scoring matrix by exp a nding the quantity of experimental peptides This is done by a dding predicted peptides corresponding to the experimental validat ion of Quartz I to establish Quartz II for the second generation of scoring matrix which c an then design more peptides with specific binding affinity to targets. 1.6 Applications of Inorganic Binding Peptides Inorganic binding peptides selected from phage display ha ve been demonstrated for a variety of applications such as morphol o gy modifier, template nucleat or quantum dots, nanoparticle s ynthesis and molecular linkers via their molecular recognition for two inorganic materials at opposite ends of dual peptide linkers A dditives can regulate the morphology of a crystal because t he crystallographic faces with the adhesion of additives are stabilized by lowing surface energy. Thus, a distinctive application of inorganic binding peptides is as morphology modifiers. As in the example shown in Figure 115(a) [18, 61] gold can alter its crystal habit that is expressed when gold binding peptides interact with specific crystallographic faces. After gold binding peptides adhere onto the (111) face s of gold particles, the morphology of gold was changed into flat triangular or pseudo hexagonal shapes Under the equilibrium conditions without any additive, the shape of the gold particles is cubooctahedral as shown in Figure 115(b)  Furthermore, inorganic binding peptides also can be applied in the fab rication of nanoparticles. Naik et al. proposed a mechanism for silver nanoparticle formation using silver binding peptide to create a siliver reduction layer as depic ted in Figure 116  Silver binding peptides interact with the clusters of silver metal atoms in aqueous silver nitrate solution. Silver ions tend to underg o a reduction reaction to silver metal atoms, which deposit onto the s i l ver clusters. This phenomenon is mainly as result of the reduc ing environment which is produced around the clusters via the interaction of silver binding peptides with silver clusters. Finally,
29 silver atoms from the reduction reaction aggregate around the clusters to form nanoparticles t hat precipitate from solution. The TEM images of silver nanoparticles created with the silver binding peptide base process from Naik et al. are shown in Figure 117  The size of the silver particles is 60 50 nm. The electron diffraction pattern indicates a face centered cubic lattice structure which matched with that of silver. Th e silver binding peptides were further utilized by patterning the silver nanoparticles In a micromoulding in capillaries (MIMIC) technique  the elastomer polydimethyl siloxane (PDMS) with patterned microfluid ic channels was placed onto a glass substrate as a stamp (See Figure 118 a  ) A s ilver binding peptides solution was guided to flow through those microfluid ic channels as templates for s i l ver deposition in a spatially ordered array. Next, silver binding peptides on the surface of microfluid ic channels were incubated with silver nitrite solution to nucleate silver particles as shown in Figure 118b  When the silver nanocrystal array was illuminated with a mercury lamp, fluorescence due to light scattering of silver nanoparticles was observed (See Figure 1 18c  ). In the micro electr onics industr y topdown photolithography methods are used for the control of x y positioning. However, a b iomimetic layer by layer (BioLBL) approach, which involve s the usage of inorganic binding peptides can enable one to control the stacking of molecules in the z coordinate. By combining BioLBL with photolithography, it is possible to fabricate 3D nanoscale structures in an x y z controlled manner  Furthermore, inorganic binding peptides also can be applied in the assembly and immobilization of inorganic nanoparticles in 2D and 3D geometries. In general, quantum dots are produced with vacuum techniques such as molecular beam epitaxy as shown in Figure 120 a  A desirable alternative wou ld be not only to synthesize inorganic nanodots under mild conditions, but also to
30 immobilize/self assemble the nanodots A suitable method is to use inorganic binding peptides with specific recognition for inorganics for nanoparticle assembly. The advanta ge of this approach is inorganic binding peptides can genetically or synthetically be fused to other functional biomolecular units or ligands to form he terofunctional molecular ent ities. Figure 1 20bc  shows the assembly of nanogold particles on a plate polystyrene surface coated with inorganic binding peptides, which resemble the distribution of quantum dots obtained by high vacuum deposition techniques. Liquid crystalline materials have been the key component in optical and electronic devices that contain liquid crystal displ ay s The molecular str ucture of liquid crystalline materials often comprises two components : rod like segment and flexible segment. Becher and coworkers ha ve showed that rod shape virus es display a distinct liquid crystalline phase [64, 65] Filamentous M13 phage also fit with the requirements of liquid crystals: The bodies of M13 phages have a rod like structure. pIII minor proteins of M13 phages behave as flexible chains. Furthermore, peptide insert s in pIII of M13 phage can bind and nucleate desired inorganic materials at the nanometer scale. Thus, an ordered nanocrystal line thin film can be fabricated using the liquid crystalline phase displayed by the nanocrystal functionalized M13 virus [66, 67] The dry Auphage thin film was prepared in a diluted Auphage solution (~ 6 mg ml1) as shown in Figure 121 (a)  T his viral nanocrystal hybrid thin film was transparent, corresponding to the optical property of nanocrystalline materials. Upon characterization by polarized optical microscopy (POM), the texture of the striped pattern in F igur e 121 (b)  matches with the s me c tic phas e of liquid crystals  In scanning electron microscopy (SEM) analysis, the morphology of the Au virus thin film also corresponded to a long range order ed s me c tic phase: the spacing of the zigzag periodic band (9.34+ the distinct feature
31 of a chiral s me c tic C structure  Individual complex es of Au M13 phage can be observed by TEM from particles extracted from a much diluted s me c tic suspension solution (See Figurer 121 (d))  The TEM image in F igure 1 21 (d) shows that a 10 nm gold particle is bound to the pIII end of one phage. Thus, the recognition of inorganic binding peptides based on phage display techniques can achieve molecular construction for a variety of applications.
32 Table 1 1. Consensus peptide sequences that have binding affinity to the various inorganic materials. Table taken from Nat Mater 2003 Sep;2(9):577585 By Sarikaya M, Tamerler C, Jen AKY, Schulten K, Baneyx F  Permission was granted by Nature Publishing Group.
33 Figure 1 1. TEM images for n ano m agnetosomes p articles within the bacterium Aquaspirillum Magnetotacticum Image taken from Acta Biomater 2007 May;3(3):289 299 by Tamerler C, Sarikaya M.  Permission was granted by Elsevier. Figure 12. Images of t he spicules with start shaped tips on the body of the spong Rosella racovitzea Image taken from Acta Biomater 2007 May;3(3):289299 by Tamerler C, Sarikaya M.  Permission was granted by Elsevier.
34 Figure 13. Images for the n acre structure of mother of pearl mollusks. Image taken from Acta Biomater 2007 May;3(3):289299 by Tamerler C, Sarikaya M.  P ermission was granted by Elsevier. Figure 14. Schematic for t he generation of random peptide librar ies Schematic taken from Acta Biomater 2007 May;3(3):289299 by Tamerler C, Sarikaya M.  P ermission was granted by Elsevier.
35 pIII: 5 c opies, 406aa pVI: 5 copies, 112aa pVIII: 2700 copies, 50aa pVII: 5 copies, 33aa pIX: 5 copies, 32aa Figure 15. The configuration of a fi lamentous M13 phage Image taken from Plant MolBiol 2002 Dec;50(6):837 854 by Willats WGT.  P ermission was granted by Oxford University Press Figure 16. S chematic of inserted peptide structures within pIII proteins. (a) Minor proteins pIII (orange) and major proteins pVIII (green) of filamentous M13 phage are highl ighted (b) The single stran d DNA of M13 phages includ es genome gIII and genome gVIII (c) Two types of structure s of inserted peptides at the terminus of protein pIII. Image taken from J Mater Chem 2003;13(10):24142421 by Flynn CE, Mao CB, Hayhurst A, Williams JL, Georgiou G, Iverson B, et al.  Permission was granted by the Royal Society of Chemistry.
36 Figure 17. S chematic of the p hage display (PD) technique (a) lacZgIII M13 ori M13 circular single strain DNAE. coli ER2738 host strain: F lacIq(lacZ)M15 proA+B+ zzf:: Tn10(TetR)/fhuA2 supE thi( lac proAB ) ( hsdMSmcr B)5 ( rk mk McrBC)+ Recombinant of E. coli gene with M13 phage gene Bacteriophage lacZ genegalactosidase lacZgIII M13 ori M13 circular single strain DNA lacZgIII M13 ori M13 circular single strain DNAE. coli ER2738 host strain: F lacIq(lacZ)M15 proA+B+ zzf:: Tn10(TetR)/fhuA2 supE thi( lac proAB ) ( hsdMSmcr B)5 ( rk mk McrBC)+ Recombinant of E. coli gene with M13 phage gene Bacteriophage lacZ genegalactosidase (b) Figure 18. Schematic of the blue white screen. (a) S chematic of bacteriophage with lacZ gene. (b) An example of the B/W screen: t he blue spots on LB agar plate are bacterial colonies that cont a in transfected M13 phage cl ones.
37 mRNA lacl promoter operator lacZ lacY lacA b. Inducer Depression inducer transcription Galactosidase Permease Transacetylase Repressor monomer mRNA lacl promoter operator lacZ lacY lacA b. Inducer Depression inducer transcription Galactosidase Permease Transacetylase Repressor monomer Figure 19. M echanism of the blue white screen O O O H O H C H2O H O H O H C H2O H O H O O O H O H O H C H2O H O H O H O H O H C H2O H O H O H + L a c t o s e G a l a c t o s e G l u c o s e galactosidase O O O H O H C H2O H O H O H C H2O H O H O O O H O H O H C H2O H O H O H O H O H C H2O H O H O H + L a c t o s e G a l a c t o s e G l u c o s e galactosidase Figure 110. M etaboli sm of lactose G alactosidase
38 O O O H O O H O H N B r C l H H N H O C l B r H O O H O H O H O H O H+ N B r C l O N B r C l O H H Xgal (5 bromo 4chloro 3indolyl D galactopyranoside ) Galactosidase H2O 5bromo 4chloro 3hydroxyindole galactose Oxidation Insoluble blue 5,5 dibromo 4,4 dichloroindigo O O O H O O H O H N B r C l H H N H O C l B r H O O H O H O H O H O H+ N B r C l O N B r C l O H H Xgal (5 bromo 4chloro 3indolyl D galactopyranoside ) Galactosidase H2O 5bromo 4chloro 3hydroxyindole galactose Oxidation Insoluble blue 5,5 dibromo 4,4 dichloroindigo Figure 111. M galactosidase Figure 112. Image of a phage clone with the s pecific binding affinity to platinum Image taken from Acta Biomater 2007 May;3(3):289299 by Tamerler C, Sarikaya M.  Permission was granted by Elsevier.
39 Figure 113. S chematic of the bioinformatics approach Schematic taken from Acta Biomater 2007 May;3(3):289299 by Tamerler C, Sarikaya M.  P ermission was granted by Elsevier Figurer 1 14. The scoring matrix for quartz binding peptides. The amino acids are colored according to their chemical properties (hydrophobic, acidic, basic and polar). Image taken from Bioinformatics 2007 Nov;23(21):28162822 by Oren EE, Tamerler C, Sahin D, Hnilova M, Seker UOS, Sarikaya M, et al.  Permission was granted by Oxford University Press.
40 A B Figure 115. The morphology of gold nanocrystal s. (a) gold binding peptides on (111) faces ; (b) no additive Image taken from J Mol Biol 2000 Jun;299(3):725735 by Brown S, Sarikaya M, Johnson E.  P ermission was granted by Elsevier. Figure 116. S chematic of the formation of silver nanoparticles in silver nitrate solution. Schematic taken from Nat Mater 2002 Nov;1(3):169172 by Naik RR, Stringer SJ, Agarwal G, Jones SE, Stone MO.  P ermission was granted by Elsevier.
41 Figure 117. TEM analysis of silver nanoparticles. (a) A TEM image of silver nanoparticles created using silver binding peptides: The size of the particle s is a r ound 60150 nm. (b) High magnification of one silver na noparticle. Insert: The array of crystal spots in the electron diffraction pattern indicates the structure of the nanoparticles corresponds to a face centered cub ic arrangement of silver metal. Images taken from Nat Mater 2002 Nov;1(3):169172 by Naik RR, Stringer SJ, Agarwal G, Jones SE, Stone MO.  P ermission was granted by Elsevier. Figure 118. Schem a tic for the spatia l contr ol of an array of silver nanocrystal s using silver binding peptides. (a) A patterned elastomer (polydimethyl siloxane, PDMS) mould was used to create microfluidic channels that serve to guide the AG4 siliver binding peptide solution on the glass substrate by capillary action. The peptides adsorb on the glass surface in a pattern defined by the network. (b) Optical b right field image showing the linear arrays of silver obtained after incubation of the AG4 patterned glass substrate with 0.1 mM silver nitrate for 48 h at room temperature. (c) Autofluorescence of the biomimetically synthesized silver particles when excited with a mercury lamp. Images taken from Nat Mater 2002 Nov;1(3):169172 by Naik RR, Stringer SJ, Agarwal G, Jones SE, Stone MO.  P ermission was granted by Elsevier.
42 (a) (b) (c) Figure 119. Images of gold quantum dots (a)Th e atomic force microscope image show s quantum dots (GaInAs) assembled on GaAs subst rate through vacuum techniques (b) Assembly of gold nanoparticles through inorganic binding peptides (c) Schematic illustration of the process used to create nanopaticle array in b, showing dual peptide linker that can bind to both the glutaradehyde and gold. PS: polystyrene substrate, GA: glutaradehyde crosslinking agent GEP 1: inorganic binding peptides with specificity to gold. Image taken from Nat Mater 2003 Sep;2(9):577585 by Sarikaya M, Tamerler C, Jen AKY, Schulten K, Baneyx F.  Permission was granted by Nature Publishi ng Group. (a) (b) (c) (d) Figure 120. Images of s me c tic ordered self supporting Auvirus films (a) Photography of a dry Au phage thin film (b) P olarized O ptical M icroscopy image of the Au phage film (Scal e M image of Au phage film (Scal e ) TEM image of an individual complex of Auphage. (Inser t: A fast fourier transform image and lattice fringe image of a gold nanoparticle at the pIII end of the phage.) Images taken from Adv Mater 2003 May;15(9):689692 by Lee SW, Lee SK, Belcher AM.  Permission was granted by Wiley VCH.
43 CHAPTER 2 I NORGANIC BINDING PEPTIDES BASED ON PHAGE DISPLAY FOR B IOSENSOR APPL ICATIONS 2.1 Motivation In this study, we appl ied the biopanning process to search for peptides that do not necessarily have a strong binding affinity for a surface, but instead have a reversible binding affinity. If the reversibility could be precisely re gulated, then one could design smart surfaces for a whole host of applications. One possibility which is explored here is to pan for electroactivated peptides which can be electrically triggered to either adsorb or desorb (or both) from an electronic mat erials surface as an electric field is applied. An exciting application that result s from electroactivated peptides is the possibility of self cleaning device surfaces by electrically triggered release of a coating. Biofouling contributes to a significan t number of infections created with implanted biomaterials and devices, and is a significant problem with biosensor arrays as well. One could conceivably design a microsystem that could reduce or eliminate a biofilm. This could also help to overcome the hurdle in continuous sensing devices, because a system could be designed that replenishes spent receptors that have become clogged upon binding to the analyte or impurities (Figure 21). In this design, t he surface of a transducer device could be patterned f or multicomponents as well. One could envision a system where the device is electronically stimulated to release the spent receptors, which is followed by a flow through system tha t contains a new batch of fresh receptors, which then bind and assemble on the sur face once it is returned to the de activated (or activated binding) configuration. This could then provide for long term continuous biosensing if such a s replen ishment is done on an automated basis. This would be particularly valuable for fieldwork applications, ranging from biosensors for soldiers at the front, to point of care diagnostics in personal healthc are, to testing of animal/plant food products in the agricultural arena.
44 2.2 Background and Significan ce Healthcare is an imp ortant issue worl dwide. The dea th rate of diseases in the countries with few medical sources is higher than developed nations. The clinics in those countries have a need for drugs to treat illness. However, the problem is that diagnostic equipments are not widespread due t o unaffordable prices for them. Diseases often are not identified, misdiagnos ed or therapy is delayed Thus, it is necessary to develop easy, economic, immediate diagnostic tool s for the patients welfare. Biosensors can be strong candidates for overcoming unbalanced distribution of medical sources between cities and countries based on several reasons as follow: Inexpensive cost of production is affordable for general clinics. All components of biosensors can be integrated in a small chip. In addition, bios ensors have low energy consumption. This makes it possible to carry biosensor s to poor areas. Furthermore, fast detection time also reduce s the waiting time to allow doctors to initiate the effective therapy for patients immediately. In the past 40 years, biosensors ha ve been developed as analytical tools for the detection of biological and chemical agents in diagnostic, environmental field monitoring, agriculture production, the pharmaceutical manufacturing and food processing. The worldwide market of bios ensors in 2003 was about $7.3 billion (Figure 22)  In 2007, the market grew to 10.8 billion with 10.4% growth rate even under a weak global economy. The trend of increasing biosensor development is mainly in the health care industry. For example, 6% of the populations in Europe and America are diabetic. Thus, availability of glucose biosensors with simple, rapid, accurate detecti on is in urgent demand for the diagnos is of diabetes  The research areas in biosensor development are wide and interdisciplinary including biological, biochemistry, physical chemistry, electrochemistry, electrical engineering, and software engineering.
45 2.2.1 D efinition of B iosensors What are biosensors? Biosensors are defined as analytical devices which can specifically, rapidly, and continuously convert biological/chemical response into an el ectric signal. The term biosensor is usually referred to as the sensor device which quantif ies the c oncentration of biological substances when a biological system can not be used directly. In general, a biosensor device consist s of a biosensing component chemically or physically immobilized onto the transducer element which converts a biological, chemical, or biochemical signal to a quantifiable and processable electrical signal  As shown in F igure 23, biosensing probes could be enzymes  antibodies  organelles, nucleic acids  cells, tissues, microbes [75, 8288] and more. The transducer in a biosensor device may be electrochemical (such as amper ometric  potentiometric [92, 93] conductive  impedance  voltammetric  and etc.), optical (including optical fiber, surface plasmon resonance [95, 96] absorption, chemiluminescence, bioluminescence  fluorescence, and etc.), piezoelectric (quartz crystal microbalance [30, 63, 75, 98, 99] and surface acoustic wave  magnetic  calorimetric, and others [104, 105] ) When the detectable signal from an analytebiosensing probe binding event is transferred to a transducer, this type of biosensing material can be directly linked to the transducer as a label free biosensor. Otherwise, labeling is required t o achieve the delivering of a signal through the transducer. Basically, the function of labels is to amplify biological signals. Labels could be fluorescence, chemiluminescence, bioluminescence, enzymes, metal particles, and nanoparticles. Finally, a computer system is required for data processing, network connection, wireless communication, and the usage of a data base. What characteristics are necessary for an effective biosensor? A successful biosensor must possess at least some of the following benefic ial features:
46 Specificity and stability: The biosensing probe must be highly specific to analyses, be stable under normal storage conditions and show good stability over a large number of assays. Less sampling: It should a llow the minimum pre treatment of samples for loading onto the biosensor device. Binding event with less inferference: When biosens ing materials interact with analytes, the binding event must not be altered by pH, temperature, and other ph ysical parameters. In addition, if the binding eve nt will involve co zymes. The best way is that cozymes are co immobilize d with the enzyme. Effective signal: Electrical noise should not interfere with the re al signal from a biosensor. The signal must be precise, accurate, reproducible and linear over th e effective measurement range. Biocompability: The probe must be tiny and biocompatible, with no toxi n releas e if the biosensor is to be used for monitoring situations in biological s ystems. In addition, biosensors must have antifouling properties because biosensing materials such as enzymes may decompose under an autoclave environment. Cost effectiveness and easy operation: The ideal biosensor should be cheap, small, portable and easy to operat e 2.2.2 Immobilization Methods of Biosensing Materials One critical aspect in the fabrication of biosensors is how to link the biological components to the transducer device, and in a way that allows transduction of the signal, such as the occurrence of a receptor binding event. Thus, it is necessary to immobilize the biosening molecules close enough to the transducer surface to provide sensitivity, operation stability, and response at a satisf actory level during the detection of analytes. The immobilization of biosens ing components on transducer elements can be di vided into chemical and physical pathways [84, 85] In chemical immobilization methods, the interaction between biosen s ing materials and transducer surfaces involve the formation of covalent binding or crosslinking For covalent immobilization, biosening materials s uch as enzymes, antibodies, oligonucleotides, and carbohydrates can be bound onto the transducer surface by chemical bonding through amino, carboxy l, sulfhydryl, or aromatic side groups of biosening molec ules. In
47 addition, viable cells also can be covalently immobilized onto the transducer surface by the chemical reaction between functional groups on the cell walls of microorganisms and the transducer surface. During the formation of covalent bonding, micr oorganisms inevitably contact harsh chemicals. This may damage the cell membranes and biological activities. Another chemical immobilization method is cross linking which involve s the formation of a network structure via the linkages between functional gr oups on biosens ing molecules and multifunctional agents such as glutaraldehyde, hexamethylene diisocyanate, and 1,5 dinitro 2,4difluorobenzene  Cross linking is widely adopted in the immobilization of microorg anisms onto transducer surfaces if the detection of analytes just relies on intracellular enzymes, rather than cell viability [84, 85] The advantage of this method is that it can reduce the wear of biosensing prob es However, high mechanical strength of a transducer surface can not be obtained via crosslinking immobilization. If a viable cell is required during the detection event the formation of covalent bonds between microorganisms and transducers should be avoided. The suitable strategy to keep the viability of cells is physical immobilization including adsorption and entrapment. Physical adsorption is a simple method with less disruption to biosensing materials. However, the interaction between biosens ing materials and the transducer surface is weak caus ing a short lifetime (several days) in this kind of immobilization method. Biosens ing materials including enzymes, antibodies, and microbes can be adhered onto a transducer matrixe via ionic, hydrophilic hydr ophilic, hydrophobic hydrophobic, and hydrogen bond interactions. Although biosens ing materials can survive in those mild environments, poor long term stability is the drawback in adsorptive immobilization.
48 With respect to the entrapment immobilization of biosens ing materials, biosen s ing materials are confined to the proximity of the transducer surface by dialysis membranes. A dialysis membrane provides sev eral functions such as structur al framework, selective ion permeability, and mediation of the electron transfer process. Another way is to mix bioche mical/polymer gels such as ag a r os, alginate, or polyurethane with microorganisms, and then coat those mixtures onto the transducer surface. The main disadvantages in entrapment methods are low sensitivity and detection limit due to the diffusion problem of entrapment materials. 2.2.3 A Challenge in Biosensor Development Biological components immobilized onto a transducer device through physical interaction including adsorption and entrapment usually have the diffusion problems which cause low stability of the biosensors. Thus, functionalizing the inorganic interface with a covalent linker is usually adopted to achieve high stability of these biosensors. However, one of the major problems in the biosensors with covalent linkers is the loss of activity of molecular probes after surface immobilization because the molecular probes are gradually clogged by analyte. Thus, one challenge with a biosensor device is the ability to have continuous detection, which means that the biosen s i ng component needs to release its analyte and be restored back to its original active state. One way to achieve this might be to use a compound with relatively weak binding characteristics so that sufficient flow will pull off the compound after the detection event. An alternative approach is proposed in this study, in which inorganic binding peptides designed by phage display techniques could be tailored as linker s between the organic inorganic interface, such as for the attachment of bios en s ing components (e.g DNA aptomers, antibodies and peptide epitopes) to inorganic transducer element s and be released by an electrical trigger from the transduction element in a biosensor device. An additional advantage of this system is
49 that peptides designed by phage display techniques are short (e.g random expressed peptides in phage libraries kit from New England Biolab, Inc. are usually 7mers or 12 mers) which bring the analyte closer to the transducer surface for improved sensitivity as compared to avidinbiotin linkages which are macromolecular proteins There are some literature reports that support the premise of using an electric field to release a bound organic, where electrodesorption has been of interest for anti biofouling surfaces. T ang et al  found that electrical stimulation could remove a triblock copolymer of co(propylene sulfide block ethylene glycol) from indium tin oxide (ITO) surfaces. These copolymers have been studied as protein resistant coatings that originally were found to chemisorb onto gold surfaces. This paper shows they also adsorb onto the transparent conducting surface of ITO, which was postulated to occur through direct sulfide indium or tin interactions. They applied an ascending anodic electrical stimulus to the surface of the modified samples, and found that copolymer was steadily removed, with complete loss of a polymeric monolayer at a potential of 2000 mV. Yeh et al  have also demonstrated electric field desorption, and in this case the protein bovine serum albumin (BSA) from a lead zirconate titanate substrate (a kind of piezoelectric t ransducer (PZT)) coated with either fired silver or titanium. They compared DC ver sus sinusoidal AC signals, and found the vibration mode of the piezoelectric aided in the removal, where 58% protein could be removed from the silver coated PZT, while 39% could be removed from the titanium coated PZT. Through modeling, they believe that t he applied electric potential was the major contributor in reducing the adhesive force between protein and surface, where the desorbed protein was then taken away by acoustic streaming shear stress. They describe a mechanism which considers application of the voltage to lead to an accumulation of
50 charges on the surface due to the capacitance character of PZT, where at a certain voltage, the charge polarity will be the same as that of the adsorbed proteins, leading to repulsive forces between surface and pro teins. In addition, because the acoustic streaming velocity is dependent on vibration amplitude, and the shear stress is linear to the streaming velocity, they proposed that the PZT plates with larger vibration amplitude would have greater BSA desorption, which was observed. These reports demonstrate that the electrodesorption principle is feasible, and given the large size of the synthetic copolymer or protein, both of which may have many binding sites, it seems reasonable to speculate that smaller peptides should be desorbed more quickly and completely, and at lower potential if the proper binding chemistry is selected. Therefore, my goal was to use phage display techniques to pan for short inorganic binding peptides that could be desorbed by an elect ric field for achieving reversibly electroactivated biosensors. 2.3 Materials and Methods 2.3.1 Materials 22.214.171.124 Dodecapeptide p hage d isplay p eptide l ibrary (Ph.D.12) New England Biolabs Ph.D. 12 phage display kit is based on M13 phage vector modified f or pentavalent display of peptides as N terminal fusions to the minor coat protein pIII. The phage display Ph. D.13 pfu/ml (1.29 x 109 12mers peptide sequences supplied in TBS with 50% glycerol). There is a short linker se quence between the displayed peptide and pIII: Gly Gly Gly Ser. 126.96.36.199 E. coli. ER2738 host strain F proA+B+ lacIq (lacZ)M15 zzf::Tn10 (TetR)/fhuA2 glnV thi (lac proAB) (hsdMSmcrB)5 (rk mk McrBC). Host strain supplied as 50% glycerol culture was i ncluded in New England Biolabs Ph.D.12 phage library kit.
51 188.8.131.52 Baterial culture medium Lauria Bertani ( LB) Lennox m edium : T wenty gram LB Lennox (Fisher Brand) (trypton: yeast extraction: NaCl = 2:1:1) were dissolved in 1 liter distilled water and adjus ted pH to 7.07.5. The LB medium was sterilized for 15minutes at 1.5 atm ( 121) in an autoclave (NAPCO, model 800DSE). LB agar m edium : F orty gram LB agar (Fisher brand) w as dissolved in 1 liter distilled water until the solution become transparent by heating. The pH was adjusted to 7.07.5 and the solution was sterilized for 15 mi nutes at 1.5 atm (121) in an autoclave. Top agar m edium : T wenty gram LB (Fisher brand) and 15g agar were dissolved 1 liter distilled water by heating until the solution became transparent. The t op agar medium was sterilized for 15 minutes at 1.5 atm (121) in an autoclave. 10x M inimal s alts (MS) : T hree gram Na2HPO47H2O (Fisher brand), 1.5g KH2PO4 (Fisher brand), 0.25g NaCl (Fisher brand), and 0.5g NH4Cl (Fisher brand) were dissolved in 50 ml distilled water. M9 Solid m edium p late : Ten ml MS and 1.5g agar (Fisher brand) were dissolved in 89 ml distilled water by heating until the solution became transparent and sterilized for 15 minutes at 1.5 atm (121) in an autoclave. Then, sterilize 1M MgSO46 H2O (Sigma Aldrich), 1M CaCl2 (Fisher brand), and 40(w/v) % glucose (Fisher brand) were filtered sterile syringe filter. Then, 0.2 ml sterilized 1M MgSO46 H2O, 0.01 ml sterilized 1M CaCl2, and 0.5 ml sterilized 40 % (w/v) glucose were added into the agar/MS mixture. Finally, pour liquid M 9 medium was poured onto 90 mm petri dish until solidification.
52 184.108.40.206 Stock solution Tetracycline s tock : Twenty mg/ml tetracycline hydrochloride (Sigma Aldrich) was dissolved in 70% ethanol and stored at 20. Xgal/IPTG s tock : Five gram IPTG (ultrapure g rade, dioxane free, Molecul A) and 4 g Xgal (Molecul A) were dissolved in 100 ml DMF (Sigma Aldrich) and stored at 20. Glycerol s tock s olution : Eighty ml glycerol (Sigma Aldrich) and 20ml distilled water were mixed to make 80% (w/v) and sterilized for 15 minutes at 1.5 atm (121) in an autoclave. 220.127.116.11 Buffer solution s PEG NaCl : Twenty gram Poly (ethylene glycol) (PEG) 8000 (Sigma) and 14.61g NaCl (Fisher) were dissolved in distilled water up to 100ml and sterilized for 15 minutes under 1.5 atm at 121C Detergent s tock s olution : Two ml (w/v) Tween 20 ( Enzyme grade, Fisher) and 2 ml (w/v) Tween 80 (for molecular biolgy, Sigma Aldrich) were added into 6 ml distilled water to get 20% (w/v) Tween 20/ 20% (w/v) Tween 80. PC b uffer : Three gram KH2PO4 ( Fisher) 1.90 g Na2CO3 ( Fisher), and 3.50 g NaCl (Fisher) were dissolved in 400ml distilled water to get 55 mM KH2PO4, 45 mM Na2CO3, and 200 mM NaCl A n a ppropriate amount of detergent was added from the detergent stock into the PC buffer according to the desire d detergent concentration (eg. 0.02%, 0.1%, 0.5%). PC buffers were sterilized by using a 18.104.22.168 Elution buffer s Low pH e lution b uffer s olution : Three g ram glycine (Sigma Aldrich) and 400 mg BSA (Sigma Aldrich) were dissolved in distilled water up to 200 ml to get 0.2 M glycine and 1 mg/ml
53 BSA, and the pH was adjusted to 2.2 with a combination of 10 M HCl and 1M HCl. The solution was sterilized by using a High s alt e lution b u ffer s olution : Five M MgCl2.6H2O was prepared and sterilized using a 2.3.2 M ethods 22.214.171.124 Fabrication of I ndium Zinc Oxide (IZO) as a t arget m aterial In this study, Indium Z i nc Oxide (IZO) was chosen for the initi al substrate as a target material for pann ing with a M13 phage library to screen for IZO binding peptides. Amorphous IZO thin film, n type semiconductors with high electron mobility (~ 10cm2V1S1), can be deposited onto suitable substrates such as Sapphir e, SiO2, or Si, using sputtering technique s near room temperature. IZO has the useful properties of chemical stability and thermal stability. Thus, IZO is a good candidate for the transducer element in a biosensor. Furthermore, consensus amino acid sequences that meam some amno acid residues frequently occur in some positions of a peptide seuqence are dependent of chemical composition, pH of the environment, crystal structure, morphology, surface roughness, and the size of grains or particles. Amorphous IZO is anticipated to provide a more homogeneous surface than a crystalline one because it should not have grain boundaries and crystal defects, where consensus sequence is not always achieved. This property can ha ve a high possibility of lead ing to a well defined consensus sequence in peptides. IZO thin films and devices for the phage releasing test s were provided by Dr. Nortons group in the Department of Materials and Engineering. The detail ed fabrication procedure of IZO thin films and device s is as follow : D eposition of IZO t hin f ilm : Indium zinc oxide (IZO) thin film s w ere deposited on 5mm square sapphire wafers in argon plasma with using rfmagnetron sputtering in Nanoscale Research Facilities (NRF) University of F lorida. Before deposition, the sputt er chamber was
54 pumped down to less than 5x106Torr. During deposition of the IZO thin films sapphire substrates were washed with trichloroethylene, acetone, and methanol in ultrasonic bath for 5 minutes succesively for removing contamination from the sapphire surface, and then were dried by a nitrogen gas stream. In addition, a commercial 3 inch diameter IZO pellet was used as a target for producing IZO vapor by applying 200 W power under 5 mTorr working pressure (Figure 2 4) D evices coated with IZO for electro releasing experiments : In the fabrication of the IZO device s SiO2 / Si wafers with 5 mm x 10 mm size were used as substrates Substrates were cleaned in an ultrasonic bath for 5 minutes each in trichloroethylene, acetone, and methanol, and were bl own dry by nitrogen gas to remove organic or inorganic contamination. Substrates were heated in an oven at 105oC for 10 minutes to remove moisture on the substrate surface. S1813 positive photoresist (PR) (Shipley) was coated onto the substrate by a spinne r ( Headway ) at 5000 rpm for 50 seconds Subsequently, PR coated substrates were baked at 100oC for 90 seconds on a hot plate. Then, PR covered substrates were exposed to light with wavelength in 365 nm for 15 seconds by Karl Suss MA 6 Contact aligner syste m with hard contact mode under exposure intensity in 8 m W/cm2. To develop photoresist after exposure, MF 300 developer was used. We developed for 50 seconds and washed for 2 minutes in deionized (DI) water. After the developing process, we used the sputte ring system to make metal electrodes. Ti and Au thin films were deposited by KJL CMS 18 Multi source sputtering system. Ti was deposited with 10 nm and Au was deposited with 80 nm thickness. Final metal layers were Ti / Au / Ti. After metal deposition, a c irc ular pattern was made with the PR (S1813) to connect gold wires later, which would then be removed after the InZnO (IZO) film deposition (Figure 25 )
55 126.96.36.199 Phage display protocol Phage display was performed by incubating the phage display library (Ph.D. 12, New England Biolabs ) with the desired target material, washing away unbound phage, and then eluting the strongly bound phages. The eluted phages were amplified by infecting the host bacterial strain ER2738, and purified. These steps are called biop anning rounds, which enrich the pool in favor of binding sequences. This sequence of steps was repeated to enrich the phage clones with the binding affinity to the target material until consensus peptide sequences are obtained. However, not every material leads to consensus peptide sequences. In this case, the cycle of biopanning is stop until peptide sequences with preferentially strong binding affinity to a specific target material. After each round, single colonies were selected for DNA sequencing, and t hen characterized to evaluate the specific binding affinity for the target with Enzyme Linked Immunosorbent Assay (ELISA) or ImmunoFluorescence Microscopy (IF), for qualitative comparison. All chemicals and labware were autoclaved, and the reactions perfo rmed in an Airclean 600 PCR workstation laminar flow hood. Phage display protocol and the other methods applied during the phage display procedure are discussed in detail as follow s Binding s tep : 12 (New England Bi olabs, MA) was exposed to an amorphous Indium Zinc Oxide (IZO) thin film coated on the top and bottom of a sapphire (0001) plate in PC buffer. The IZO phage solution was rotated by an agitator (Labqueake, Barnestend Thermolyne) for 30 min in order to obta in sufficient time for phage IZO interaction. Washing s tep : After a 30 min rotation, several washing cycles were performed in order to remove the nonspecific phage from IZO surface. Phages that strongly bind to the target substrate are retained, while the nonbinder ones are washed away. These washing cycles were repeated for ten to thirteen times for each biopanning round. The detergent concentration was
56 increased gradually up to 0.5%. Applied washing procedure is described below First, one IZO sheet was put into a 1.5 ml microfuge tube and o ne ml PC was added into a microfuge tube. Then, ten Ph.D. 12) was added into a 1.5ml microfuge tube (Fisher brand), and the m icrofuge tube containing IZO and phage peptide library was put the on an agitator 30 min for phage IZO interaction. After 30 min agitation, the f irst supernatant was put into another microfuge tube and the IZO sheet was washed with 1 ml PC containing 0.1 % detergent. The washing steps were repeated 6 times. The supernatant was put in a fresh microfuge tube at the end of the each washing step Next a m icrofuge tube containing the IZO sheet with bound phages in 1ml PC containing 0.1% detergent was put the on an agitator 30 min for removing unbound phage again. The supernatant was transferred into a fresh microfuge tube. 1ml PC containing 0.1 % dete rgent was added into the microfuge tube including the IZO sheet and the m icrofuge tube containing the IZO sheet was put on an agitator 30 min for removing unbound phages further. The supernatant was transferred into a fresh microfuge tube and t he IZO sheet was washed with 1 ml PC containing 0.1 % detergent. The microfuge tube containing IZO with bound phages was left on an agitator overnight. The supernatant was discarded and transferred into microfuge tube. Subsequently, t he IZO sheet with bound phages wa s washed with 1 ml PC which contains 0.1% detergent. The microfuge tube containing IZO with bound phages was put on an agitator 30 min. The supernatant was discarded and transferred into microfuge tube. IZO was washed with 1 ml PC containing 0.1% detergent The microfuge tube containing the IZO sheet with bound phage s was put on an agitator 30 min. The supernatant was put into microfuge tube. Elution s tep : Strongly bound phages were recovered from the IZO surface through elution. The strong interaction betw een phage and IZO surface was disrupted using a low pH
57 buffer solution. The elution procedure i s described as follow s. After the last washing step, 1 ml of low pH elution buffer solution was added i nto the IZO sheet with bound phages in a microfuge tube. T he m icrofuge tube containing the IZO sheet with bound phages and low pH elution buffer was put the on an agitator for 8 minutes to elute the bound phages from IZO surface. The supernatant including eluted bound phages was transferred into a fresh microfuge tube Then, one hundred l supernatant was transferred into E. coli ER2738 culture (LB Lennox medium: E. coli ER2738 overnight culture = 100:1) and it was incubated 4.5 hours at 37C and 250 rpm. The remaining 900l supernatant was neutralized by adding 40 l of Tris, pH 9.1. One ml of low pH elution buffer was put into microfuge tube. The Microfuge tube which contains IZO and low pH elution buffer was put the on an agitator for 8 minutes to elute the phage from IZO surface again. Subsequently, t he supernata nt was transferred into a microfuge tube. One hundred l supernatant into was transferred into E. coli ER2738 culture (LB medium: E. coli ER2738 overnight culture = 100:1) and it was incubated for 4.5 hours at 37C and 250 rpm. The remaining 900l supernat ant Then, one ml of low pH elution buffer solution was put for the third time into microfuge tube. The Microfuge tube which contains IZO and low pH elution buffer solution was put the on an agitator for 8 min to elute the phage from IZO surface. The supernatant was transferred into m icrofuge tube and 100l supernatant was transferred into E. coli ER2738 culture ( LB Lennox medium: E. coli ER2738 overnight culture = 100:1) and it was incubated fo r 4.5 hours at 37C and 250 rpm The Next, one ml of low pH buffer solution was put for the forth time into microfuge tube. The Microfuge tube which contains IZO and low pH buffer solution was put the on an agitato r for 8 min to elute the phage from IZO surface. The supernatant was t ransferred into microfuge tube and 100l supernatant
58 was transferred into E. coli ER2738 culture ( LB Lennox medium: E. coli ER2738 overnight culture = 100:1) and it was incuba ted for 4. 5 hours at 37C and shaked at 250 rpm.Finally, t he remaining 900l supernatant was ne Amplification and p urification s teps : Eluted phage samples were infected into the host strain E. coli ER2738 and amplified. Before the beginning of amplification period, E. coli host strain ER2738 was cultured to r each the OD600 ~ 0.5 (the best phage host strain propagation period) and then eluted phage solutions were transferred to bacteria culture. The incubation period was approximately 4.5 hours at 37C and phage E. coli host strain solution was shaken at 250 r ounds per minute ( rpm ) on a shaker (Max 2000, Barnstead Lab Line) E. coli host strain ER2738 used in amplification is a robust F+ strain with a rapid growth rate and is particularly well suited for M13 propagation. ER2738 is a recA+ strain and the F facto r of ER2738 contains a minitransposon, which confers tetracycline resistance. After amplification eluted phages were purified from host cell according to the procedure below. First, E.coli phage culture was transferred to 50 ml sterilized centrifuge tub es after 4.5 hours of the growth period. Then, s amples were centrifuged at 8000 rpm for 10 min, and the s upernatant was transferred to 50 ml sterilized centrifuge tubes. PEG/NaCl was added (1:6) into supernatant to precipitate phage and it was le ft overnig ht at 4 Samples were cent rifuged at 10000 rpm for 20 min. Then, the s upernatant was discarded and a phage pellet was resuspended with 5 ml PC buffer (no detergent) by pipetting to remove any remaining E.coli ER2738. Next, s amples were centrifuged at 10000 rpm for 10 min. After that, the s upernatant was transferred to 50 ml sterilized centrifuge tubes. PEG/NaCl was added (1:6) into the solution to precipitate phage and the solution was le ft for 2 hours at 4 Samples were centrifuged at 10000 rpm for 10 m in. The s upernatant were discarded and phage pellet was resuspended by pipetting with 1 ml PC buffer (no detergent) to
59 remove E.coli ER2738. Samples were centrifuged at 10000 rpm, for 10 min and supernatant were transferred to sterilized microfuge tubes. P EG/NaCl solution was added (1:6) into the microfuge tube to precipitate phage, sample was vortex ed 5 sec, and leaved this solution in the air 10 min. Samples were centrifuged at 13200 rpm, 3 min to compact the phage s The s upernatant was discarded and phag e pellet was resuspended with 0.2 ml PC buffer (no detergent) by pipetting gently. Samples were centrifuged at 13200 rpm for 3 min. The s upernatant was transferred to sterilized microfuge tubes and it was stored at 20 188.8.131.52 Blue white screening This e xperiment was performed to estimate the phage titers at the end of each biopanning round. The b lue white screening experiment has three fundamental steps, namely preparation of Xgal/IPTG plates, serial dilution of eluted phage samples, and estimation of phage titers for each round. Preparation of LB Xgal/IPTG p lates : 187 l Xgal/IPTG was added 150 ml liquid warm LB agar in 150 ml glass medium flask and it was poured to 60 mm plastic sterile petri dish. Plates were wrapped with parafilm an d Aluminum foil and stored at 4 in the dark for a maximum of 1 month. Serial dilution of phage s amples : 190 l PC buffer (without detergent) and 10 l phage was added into A1 well of 96well plate 180 l PC buffer (without detergent) was put into wells from A2 to A12. Ten fold dilutions were made from A1 to A12 by taking 20 l samples from preceding wells (Figure 26) Calculation of Phage Titers : After serial dilutions, LB Xgal/IPTG plates that were prepared previously and stored at 4 were put at room temperature. Three ml melt top agar (for 60mm petri dish) or 5 ml (for 90 mm petri dish) was aliqo t ted into 15 ml cell culture tube and
60 these tubes were put in 55 water bath to prevent the solidification until other processes carried out. E. coli ER2738 (1:250) from overnight culture were inoculated into 5 ml LB medium in 50 ml falcon tube. The culture was incubated until midlog phase (OD600 ~ 0.5) at 37C and 250 rpm. After incubation period, 200 l E. coli ER2738 culture and 10 l d iluted phage sample were put into 1.5 ml eppendorf tube. Phage E. coli ER2738 mixture was added into 3 ml melt top agar and poured onto LB Xgal/IPTG petri dishes. All petri dishes were kept up side down and incubated at 37 for overnight. After incubation period, blue plaques were obtained (Figure 18 (b)) .The plate with 30~100 plaques from each eluted page solution (E1E4) were chosen to calculate the amount of phage with the equation as follow: The amount of phage ( pfu) = The number of plaques The volume of diluted phage solution (ml) X Dilution factor pfu: plaque forming unit For example, 10l diluted phage solution from A9 well was added into 3 ml top agar. A ter votexing, the mixed solution was pour onto a LB Xgal/IPTG plate. A fter the incubation overnight, 100 phage plaques were counted on this LB Xgal/IPTG plate. The calculation of the amount of phage was as follows. (100 phage plaques/0.01 ml diluted phage solution) x 109 = 1.00 x 1013 pfu/ml According to phage titers at each round, phage amount was determined to generate a phage pool for the next biopanning round. Saving P hage C lones for S equencing : Prep aration procedure of saving clones for sequencing i s described in the following subsections. First, 150 l, 0.02 % PC buffer was put each well of 96 well ELSA plate. Each phage plaque from big plates was picked and put into different well of ELSA plate. Twenty four plaques were picked per each elution step. At the end
61 of the fourth round, 576 plaques were pi cked. Next, 96 well plate containing phage clones was placed into the incubator at 60 for 45 min. and 96 well plate was left at 4 for overnight. After that, 60 l sterilized 80% glycerol solution was put each well in two sets of 96 well plates (Overall g lycerol concentration was kept as 50 % in stocks). Subsequently, 50 l of each phage clone was added from storage plate to glycerol containing plates. Plates were covered by parafilm and stored in 80 After saving clones, phages DNA were isolated for DNA sequencing. DNA isolation was performed using the QIAprep Spin M13 kit (QIAGEN) procedure. The procedure is described below in detail Ten l sample was taken from the glycerol stock of a single phage clone, and was added into the 3ml E. coli ER2738 cul ture (LB medium: E. coli ER2738 overnight culture = 100:1) which was incubated until mid log phase (OD600 ~ 0.5). After this step, culture was incubated 4.5 hours for phage E. coli ER2738 infection. Next, culture was centrifuged at 5000 rpm for 15 min. at room temperature supernatant containing M13 bacteriophage was transferred to a fresh reaction tube. During transferring the supernatant, bacterial pellet was not disturbed. Any carryover of bacterial cells will result in contamination of the M13 precipita tion with bacterial chromosomal DNA or double stranded bacteriophage RF DNA. Buffer MP was added 1/100 volume (i.e. 10l per 1 ml of phage supernatant) to the supernatant in reaction tube. It was mixed by vortexing and incubated at room temperature for at least 2 min. During this step, bacteriophage particles were precipitated from the culture medium. A QIAprep spin column was placed in a 2 ml microcentrifuge tube and applied 0.7 ml of the sample to the QIAprep spin column. Reaction tube was centrifuged for 15 sec. at 8000 rpm and discarded flow through from collection tube. During this step, intact bacteriophage was retained on the QIAprep silicagel membrane. The last step was repeated until all supernatant passed through QIAprep spin column. 0.7 ml MLB buffer was added for M13
62 lysis and binding, to the QIAprep spin column and centrifuged for 15 sec. at 8000 rpm. This step creates appropriate conditions for binding of the M13 DNA to the QIAprep silica gel membrane. When bacteriophage lysis begins a nother 0.7 ml MLB buffer was added the QIAprep spin column and was incubated 1 min. at room temperature to lyse bacteriophage completely. QIAprep spin column was centrifuged for 15 sec. at 8000 rpm. M13 singlestranded DNA is released from bacteriophage particles and adsorbed to the QIAprep to the silica gel membrane. Buffer PE 0.7 ml was added and centrifuged for 15 sec. at 8000 rpm. In this step residual salt is removed. Buffer PE was discarded from collection tube and centrifuged QIAprep spin column for 15 sec. a t 8000 rpm to remove residual buffer PE. It is important to dry the QIAprep membrane with quick microcentrifugation step. This prevents residual ethanol from being carried over into subsequent reactions. QIAprep spin column was placed in a clean 1.5 ml microcentrifuge tube. 100l EB buffer (10 mM Tris.Cl, pH 8.5) was added to the center of the column membrane to elute the DNA. Incubation of elution buffer in the QIAprep spin column significantly increases the recovery of single stranded M13 DNA molecules, which adsorb tightly to the silica membrane. The DNA can also be eluted with water. When using water for elution, the pH of water should be in the range 7.0 8.5. Elution efficiency is dependent on pH and the maximum elution efficiency is achieved within th is pH range. DNA s equencing : The Interdisciplinary Center for Biotechnology Research ( ICBR ) in the University of Florida provided service in PCR and DNA sequencing for single strain DNA of M13 phage clones. 96 gIII sequencing primer ( 5 HOCCC TCA TAG TTA GCG TAA CG 3, 100 pmol, 1 pmol/ l) purchased from New England Biolabs Ph.D.C7C Phage Display Peptide Library Kit was used in PCR. DNA samples were sequenced by using Genome Sequencer 20 System ( Roche Applied Science).
63 184.108.40.206 The d etermination of e xpressed 12 mer p eptides from M13 p h age DNA For the DNA of M13 Phage in Ph.D.12 phage library, the DNA codes are the same except 12mer peptide gIII fusion for each phage (Figure 27) In order to find out the 12 mer peptide gIII fusion, two segments of DNA sequences in the immediate vicinit y of gIII fusion domain were used and aligned with any DNA sequence of phage clones by using Nucleotide alignment in Basic Local Alignment Search Tool (BLAST) fromNational Center for Biotechnology Information (NCBI): http://blast.ncbi.nlm.nih.gov/ After obtaining the DNA sequence of gIII fusion, DNA codes of gIII fusion can be translated into amino acid sequences of 12mer peptides with the tool for DNA to protein translation in the website as follow s : http://bio.lundberg.gu.se/edu/translat.html An example is shown below: T he DNA sequence for a M13 phage clone is aligned by two segments of DNA sequences in the immediate vicinity of the gIII fusion domain a s follow: 1. TTATTCGCAATTCCTTTAGTGGTACCTTTCTATTCTCAC TCT 2.GGTGGAGGTTCGGCCGAAACTGTTGAAAGTTGTTTAGCAAAATCCCATACAGAA After the alignment using Nucleotide BLAST, the DNA sequence of gIII fusion (green domain) will be confined between two sets of aligned DNA domains (yellow dmain): ACCCCGAGCCGCAGTCATAGAAAAAAGAAGAGGCG Subsequently, this DNA sequence is translated into amino acid sequence. 1 CGTCTTTCCA GACGTTAGTA AATGAATTT T CTGTATGGGA TTTTGCTAAA 51 CAACTTTCAA CAGTTTCGGC CGAACCTCCA CC ACCCCGAG CCGCAGTCAT 101 AGAAAAAAGA AGAGGCGT AG AGTGAGAATA GAAAGGTACC ACTAAAGGAA 151 TTGCGAATAA TAATTTTTTC ACGTTGAAAA TCTCCAAAAA AAAGGCTCCA 201 AAAGGAGCCT TTAATTGTAT CGGTTTATCA GCTTGCTTTC GAGGTGAATT 251 TCTTAAACAG CTTGATACCG ATAGTTGCGC CGACAATGAC AACAACCATC 301 GCCCACGCAT AACCGATATA TTCGGTCGCT GAGGCTTGCA GGGAGTTAAA 351 GGCCGCTTT T GCGGGATCGT CACCCTCAGC AGCGAAAGAC AGCATCGGAA 401 CGAGGGTAGC AACGGCTACA GAGGCTTTGA GGACTAAAGA CTTTTTCATG 451 AGGAAGTTTC CATTAAACGG GTAAAATACG TAATGCCACT ACGAAGGCAC
64 501 CAACCTAAAA CGAAAGAGGC AAAAGAATAC ACTAAAACAC TCATCTTTGA 551 CCCCCAGCGA TTATACCAAG CGC GAAACAA AGTACAACGG AGATTTGTAT 601 CATCGCCTGA TAAATTGT DNA to protein translation: ACCCCGAG CCGCAGTCATAGAAAAAAGAAGAGGCGT TPLLFSMTAARG 220.127.116.11 Immunofluorescence (IF) m icroscopy e xperiment At the beginning of the fluorescence microscopy experiment, bot h negative and positive control experiments were carried out to decide on the right procedure during the fluorescence experiment. In the negative control experiment, IZO sheet was incubated with Anti M13 pIII monoclonal antibody (Amersham Bioscience) ( Ant i M13 monoclonal antibody: PC buffer = 1: 500) as the primary antibody which is specific to the M13 pVIII protein in 1ml PC buffer for 20 minutes on an agitator. Then, discarded supernatant and use 1 ml PC with suitable concentration of detergent washed IZ O sheet twice (0.1 %, 0.3 %, and 0.5 % for phage clones from 1st, 2nd, and 3nd biopanning respectively). Subsequently, anti mouse IgG FITC as secondary (SigmaAldrich) antibody (anti mouse IgG FITC : PC buffer = 1 : 100) was incubate with IZO in 1 ml PC buffer for 20 minutes on an agitator. After 20 minute incubation, added 1 ml PC buffer with suitable concentration of detergent to wash IZO three times. IZO was visualized at 20X magnification and 2 sec exposure time by the fluorescence microscope (NikonEcl ipse E600) with WIB filter. In the positive control experiment (Figure 28), IZO was incubated with 10 clone (1011 pfu/ml) in 1 ml PC buffer on an agitator for 1 hour. Subsequently, discarded supernatant and washed IZO with 1 ml PC with suitable concentration of detergent three times. Next, anti M13 pIII monoclonal antibody in 1 ml PC buffer (1:500) was incubated IZO for 20 minutes on an agitator. After incubation period, discarded supernatant and washed IZO with 1 ml PC with suitable concentration of detergent twice. Then, anti mouse IgG FITC in 1 ml PC buffer
65 (1:100) was incubated with IZO for 20 minutes on an agitator. Subsequently, discarded supernatant and washed IZO with 1 ml PC with suitable concentration of detergent three times to remove the residue of unbound anti mouse IgG FITC from IZO. Finally, these complexes were visualized at 2 0X magnification and 2 sec exposure time by fluorescen ce microscope with WIB filter. In order to evaluate the specificity of IZO phage binders to IZO, all obtained good binders were analyzed by using the sapphire (0001), silicon (100), and silicon oxide (a morphous) for cross specificity at 20X magnification and 2 sec exposure time by fluorescence microscopy with W IB filter. 18.104.22.168 Enzyme linked immunosorbent assay (ELISA) A phage single clone (107 pfu/ml) in 1 ml PC buffer was incubated with IZO, sapphire (0001), silicon (100), silicon oxide (amorphous) sheet for 1 hour on an agitator respectively. After incubation, discarded supernatant and washed substrates with 1 ml PC buffer with suitable concentration of detergent three times. Then, monoclonal anti M13:HRP (GE Health ) which is the secondary antibody conjugated to horseradish peroxide (HRP) for M13 phage detection (anti M13:HRP : PC buffer = 1:2500) with inorganic sheets mentioned above for 20 minutes on an agitator. Subsequently, discarded supernatant and washed inorganic sheets with 1 ml PC with suitable concentration of detergent three times to re move unbound anti M13:HRP. Substrate for the development reaction for HRP conjugated secondary enzymes was prepared by dissolving one capsule of phosphate c itrate buffer with sodium perborate (Sigma Aldrich) in 100 ml distilled waster (0.05 M phosphate citrate buffer pH5.0, 0.3% (w/v) sodium perborate). A 10 mg tablet of 3, 3, 5, 5 tetramethylbenzidine (TMB) substrate (Sigma Aldrich) was added to 10 ml of t he buffer to give a final concentration of 1 mg/ml. 1 ml substrate solution was added into microcentrofuge tube containing inorganic sheet respectively and was developed for 20 minutes
66 (Figure 2 9). After that, supernatant from each tube was transferred into a 96 well plate. The plate was read in an ELx 800UV plate reader (Bio Tek) at 630 nm. 22.214.171.124 Calculation of surface coverge of a phage clone on substrate sheets Image processing software, Image J, was used to measure the surface coverage of phage clo nes on substrate (Figure 210). A fluorescence image of phage clones on IZO sheet (Figure 210(a)) was con v erted to a 32 bit black and white image (Figure 210 (b)). Make a threshold of F igure 210 (b) to F igure 210(c). Finally, measure the ratio of black a rea to the whole picture area. This ratio measures surface coverage of phage clones on substrate sheets. 2.4 Results and Discussions In this study, Indium Zinc Oxide (IZO) was as a target material to select phage clones with 12 mer expressed peptides show ing binding affinity to IZO determined from the phage display process. After each biopanning, phage titers were necessary to calculate the concentration of eluted phages. Based on the concentration of eluted phages, the volume of each el ution solution was determined based on mak ing the same contribution of the phage amount to phage pool for the succeeding round. The phage titer result with using l ow pH elution buffer is show n in T able 21. Based on these result s the level of phage concentration was over 1011 pfu/ml. This indicates that amplification efficiency of phages within a host bacterial E. coli. ER2738 strain, was promising. After phage titers, some phage clones were picked for DNA sequencing to deduce the IZO binding peptides sequences. Figure 2 11 show s the expressed 12 mer peptide sequences translated from the DNA of phage clones selected from IZO. For amino a cid sequences show n in F igure 211, most of the peptide sequences contained a block of hydrophobic amino acids, and one or two basic amino acids. In addition, approx.4 to 5 polar amino acids were distributed within 12 mer expressed peptides separately.
67 It can be seen that the IZO substrate led to a consensus sequence; in fact two sequences were fully duplicated with 6 clones for one (TKNMLS LPVGPG), and two clones for the other (MNRPSPPLPLWV). Although inorganic surfaces dont always evolve to a consensus sequence, the amorphous nature of the IZO may have provided a very uniform surface. This has been observed in Sanos work with TiO2 particl es with of amorphous oxide surface layer, where 33 of the 43 clones had an identical sequence  However, a phage clone with expressed consens us 12mer peptide sequences do es not mean it could be a good binder for a tar get material because a consensus peptide sequence just represents the expressed fusion peptide from the most en rich ed phage clones observed after three or more biopannings. Thus, immunofluorescence (IF) analysis and enzyme linked immunosorbent assay (ELISA ) were used to evaluate the properties of phage clones in binding affinity and specificity to a target material. In ELISA, the secondary antibody conjugated to horseradish peroxide (HRP), anti M13:HRP, was used to detect M13 phages absorbed onto the surfa ce of inorganic substrates semi quantitatively. When HRP catalyze the enzymatic reaction of 3, 3, 5, 5 tetramethylbenzidine substrate, the oxidized product of TMB has a deep blue color. A deep blue color indicates a higher density of phages on those subs trates. In addition, the degree of blue color can be quantified by measuring UV Visible absorbance at 630 nm. The seven representative phage clones eluted in low pH buffer were first evaluated for their binding affinity and specificity to IZO by using ELI SA (Figure 212 and Figure 213 ). The phage clone with the consensus amino acid sequence (TKNMLSLPVGPG) preferentially bound to IZO based on the UV absorbance, but also displayed a high binding affinity to sapphire. With respect to another phage clone with the other consensus amino acid sequence
68 (MNRPSPPLPLWV), it mainly tended to bind onto sapphire, even if IZO was chosen as target material. The reason is that IZO was sputtered onto the top and bottom surfaces of a sapphire plate. Thus, some thin edges of sapphire were exposed to a phage library during biopanning. This caused a few of the clones to be selected for this region. The phage clone with expressed peptide (MNRPSPPLPLWV) was probably selected from the sapphire edge due to its strong binding affinit y to sapphire. In addition, other phage clones with different type s of express ed 12 mer peptides were also explored for their binding affinity to IZO in ELISA results as shown in Figure 212 and Figure 213. For example, ASQITHFPRPPW contained less polar a mino acid residues than TKNMLSLPVGPG, but they have similar properties of peptide seuqences TEAHRQSMTLTW was comprised of an acidic amino acid residue which is not common in IZO binding peptides Furthermore, ASQITHFPRPPW mostly consisted of polar amino a cid residues. In order to visualize the phage binding on inorganic substrates, immunofluorescence analysis (IF) was used by utilizing a fluorescence tag (FITC) conjugated with anti mouse IgG to indicate the location of M13 phage on the inorganic substrate surface under fluorescence microscopy. However, Anti mouse IgG FITC may have a binding affinity to som e inorganic substrate s directly. Thus, it was necessary to test anti mouse IgG FITC to see if it has the ability to bind onto IZO, sapphire, Si, and SiO2 in the absence of M13 phage and anti M 13 primary antibody. Figure 214 show s that antimouse IgG FITC did not tend to bind onto those inorganic substrates because no green spot s were observed without loading M13 phage and anti M13 primary antibody under f luorescence light. The phage clone TKNMLSLPVGPG exhibited preferential binding to IZO i n ELISA (Figure 2 13a). In IF analysis, this phage clone had high binding affinity to IZO, but showed
69 minor binding affinity to the other materials (Figure 2 15). For t he phage clone: MNRPSPPLPLWV it showed preferential binding affinity to sapphire, rathe r than IZO (Figure 216). This also proved this clone was selected from sapphire edge. The results in IF analysis corresponded to the one s in ELISA (Figure 213). In ad dition, I ZO was etched using elution buffer with low pH (Figure 217). Thus, some phages with strong binding for I ZO were released. This is not a benefit for the development of electroactive peptide linker. Thus, high pH or high salt elution buffer may be considered better to use in the elution step. In high salt elution, high salt elution buffer solution, 5M MgCl2.6H2O did not cause the etching of the IZO thin film on the sapphire substrate. In addition, high salt elution buffer also elute similar phage amount compared to low pH elution buffer solution (Table 22). This indicated high salt elutio n buffer also could provide enough diversity of eluted phage clone for the subsequent biopa nning. Figure 218 show s that some expressed 12 mer peptide sequences eluted from high salt elution. High salt elution solution evolved into three consensus peptide sequence: AGFPWSTHSSWL, SHAPDSTWFALF, and TNSSSQFVVAIP. In most of peptides sequences from high salt elution, they were basically comprised of a block of hydrophobic amino acids (2~5 residues) which also appeared in peptides sequences from low pH elution. In addition, a block of polar amino acids (3~5 residues) were usually followed by a block of hydrophobic amino acids. In addition, Figure 2 11 and Figure 218 sh owed that the amino residue, histidine, frequently appeared in most IZO binding peptide sequences. Some metalloproteins have been proved to coordinate to zinc cations through basic amino acid residue such as histidine [108, 109] Thus, histidine in IZO binding peptides may function as binding sites to interact with zinc components in IZO through coordination binding.
70 The evaluation of representative phage binding affinity through ELISA is shown in Figure 219 and Fig ure 2 20. M ost of the representative 12 mer expressed peptide sequences including three consensus peptide sequences display significantly higher preferential binding to IZO except one of consensus peptide sequence, AGFPWSTHSSWL. Figure 221 show s that the phage clone: SHAPDSTWFALF displayed high preferential binding affinity to IZO. In addition, it also displayed lower surface coverage on sapphire than phage clones from low pH elution. On the whole, the binding affinity phage clones eluted from high sal t elution was more selective to IZO than low pH elution. This indicated high salt elution indeed avoided the selection of phage clones with strong binding affinity to different inorganic substrates. However, the surface coverage of phage clones from high s alt elution is obviously lower than low pH elution. After evaluating the binding of phage clones to IZO, the phage clones with good surface coverage on IZO surface were chosen to test electro releasing ability using IZO devices. The phage clones from high salt elution showed insufficient surface coverage on IZO surface even if they have better selectivity to IZO. Thus, it is difficult to observe the difference of fluorescence intensity between phage binding and phage releasing on IZO device surface. Thus, we chose phage clones from low pH elution because of their high surface coverage on IZO surface. Although they still show obvious binding affinity to sapphire, sapphire is not a candidate for a transducer surface material due to its low conductivity. In this study we mixed three phage clones with good surface coverage on the IZO thin film from low pH elution to incubate with the IZO device (Figure 2 22(a)). Figure 222 (b) showed those phage clones also had good surface coverage on the IZO device surfac e. After applying 1400 mV for 5 minutes, the releasing area, the bending c hannels between electrodes, turned darker than the electrode area covered with IZO thin film (Figure 2 2 2(b)). The decrease
71 of fluorescence light is mainly due to the releas e of phag e clones, rather than the degradation of fluoresce tags caused by applying a voltage. In order to prove that voltage is not the main factor to cause the decrease of fluorescence intensity, t he phage clone NMTMSFPTYPIA that had irreversible binding was used to examine the change of fluorescence intensity after applying a voltage. Figure 2 23 showed fluorescence intensity of the IZO device was not changed even if a voltage was applied onto the IZO device. The se electro releasing results could be envisioned th e potential of peptide s designed by phage display technique to serve as linkers for the development of re furbishabl e biosensors.
72 Table 21. P hage titers in the I ZO system with low pH elution buffer Elution F irst round S econd round T hird rou nd Cum -p (pfu/ml) Cam -p (pfu/ml) Vtatoal Cum -p (pfu/ml) Cam -p (pfu/ml) Vtatoal Cum -p (pfu/ml) Cam -p (pfu/ml) E1 (1st elution) 2.7x105 3.3x1014 75 1.0x105 2.0x1017 1 8.0x104 3.4x1014 E2 (2nd elution) 1.9x104 3.8x1016 2 1.0x104 7.0x1013 150 1.5x104 7.1x1014 E3 (3nd elution) 2.0x104 1.2 x1015 25 5.0x104 1.0x1013 175 4.5x102 1.8x1011 E4 (4nd elution) 1.0x104 2.2x1015 150 1.1x104 1.0x1014 150 2.0x102 8.2x1011 Total 252 476 Cum p: The concentration of u m amplified phages Cam p: The concentration of amplified phages Vtatoal: The total taken volumn of eluted phages Table 22. P hage titers in the I ZO system with high salt elution solution Elution F irst round S econd round T hird round Cum -p (pfu/ml) Cam -p (pfu/ml) Vtatoal ( Cum -p (pfu/ml) Cam -p (pfu/ml) Vtatoal Cum -p (pfu/ml) Cam -p (pfu/ml) E1 (1st elution) 1.2x105 1.8x1018 1 2.1x105 1.3x1013 75 8.2x104 1.0x1013 E2 (2nd elution) 2. 3 x1 04 3.9x1018 1 6.1x104 6.7x1012 150 3.3x104 8.4x1012 E3 (3nd elution) 4.0x104 1.2x 1017 10 5.6x104 1.6x1016 1 1.2x104 8.1x1012 E4 (4nd elution) 1.0x104 1.4x1015 100 1.3x104 1.2x1015 1 1.1x104 5.2x1012 Total 112 227 Cum p: The concentration of umamplified phages Cam p: The concentration of amplified phages Vtatoal: The total take n volumn of eluted phages
73 Figure 21. S chematic of the electro releasing mechanism using electro activated peptides. T he clogged receptors are depicted as brown spots. T he electroactive peptide linker s are red curved lines. Medical Environmental BioDefense Pharmaceutical Food & Beverage Market in 2003 $7.3 B Figure 22. S chematic of the market size and potential application s of biosensors in the worldwide market
74 Figure 23. S chematic of the main components of biosensor devices Figure 2 4. S chematic for the deposition of IZO on a sapphire sheet Figure 2 5. D esign of the device for releasing the phage clones Biosensing MaterialsEnzyme Antibodies Cells Microbes DNA probes Tissues TransducersElectrochemical Optical Piezoelectric Magnetic Calorimetric LabelsFluorescence Nanoparticles Enzymes Metal particlesImmobilized (Peptide linker) Electrical signal QuantificationDisplay Data processing
75 10-210-1110-1210-1310-910-1010-810-610-710-510-410-3 Figure 26. Serial dilution of phage samples Expressed 12 mer foreign peptide Eag I 28 sequencing primer 96 sequencing primer Plll leader sequence KpnI 5 TTA TTC GCA ATT CCT TTA GTG GTA CCT TTC TAT TCT CAC TCT 3 AAT AAG CGT TAA GGA AAT CAC CAT GGA AAG ATA AGA GTG AGA Leu Phe Ala Ile Pro Leu Val Val Pro Phe Tyr Ser His Ser 5 TTA TTC GCA ATT CCT TTA GTG GTA CCT TTC TAT TCT CAC TCT 3 AAT AAG CGT TAA GGA AAT CAC CAT GGA AAG ATA AGA GTG AGA Leu Phe Ala Ile Pro Leu Val Val Pro Phe Tyr Ser His Ser NNK NNK NNK NNK NNK NNK NNK NNK NNK NNK NNK NNK GGT GGA GGT NNM NNM NNM NNM NNM NNM NNM NNM NNM NNM NNM NNM CCA CCT CCA Xxx Xxx Xxx Xxx Xxx Xxx Xxx Xxx Xxx Xxx Xxx Xxx G1y G1y G1y NNK NNK NNK NNK NNK NNK NNK NNK NNK NNK NNK NNK GGT GGA GGT NNM NNM NNM NNM NNM NNM NNM NNM NNM NNM NNM NNM CCA CCT CCA Xxx Xxx Xxx Xxx Xxx Xxx Xxx Xxx Xxx Xxx Xxx Xxx G1y G1y G1y TCG GCC GAA ACT GTT GAA AGT TGT TTA GCA AAA TCC CAT ACA GAA AGC CGG CTT TGA CAA CTT TCA ACA AAT CGT TTT AGG GTA TGT CTT Ser A1a Glu Thr Val Glu Ser Cys Leu Ala Lys Ser His Thr Glu TCG GCC GAA ACT GTT GAA AGT TGT TTA GCA AAA TCC CAT ACA GAA AGC CGG CTT TGA CAA CTT TCA ACA AAT CGT TTT AGG GTA TGT CTT Ser A1a Glu Thr Val Glu Ser Cys Leu Ala Lys Ser His Thr Glu AAT TCA TTT ACT AAC GTC TGG AAA GAC GAC AAA ACT TTA GAT TTA AGT AAA TGA TTG CAG ACC TTT CTG CTG TTT TGA AAT CTA Asn Ser Phe Thr Asn Val Trp Lys Asp Asp Lys Thr Leu Asp AAT TCA TTT ACT AAC GTC TGG AAA GAC GAC AAA ACT TTA GAT TTA AGT AAA TGA TTG CAG ACC TTT CTG CTG TTT TGA AAT CTA Asn Ser Phe Thr Asn Val Trp Lys Asp Asp Lys Thr Leu Asp CGT TAC GCT AAC TAT GAC GGC 3 GCA ATG CGA TTG ATA CTC CCG 5 Arg Tyr Ala Asn Tyr Glu Gly CGT TAC GCT AAC TAT GAC GGC 3 GCA ATG CGA TTG ATA CTC CCG 5 Arg Tyr Ala Asn Tyr Glu Gly Figure 2 7. N terminal sequence of random 12mer peptide g III fusion for M13 phage DNA in Ph.D.12 phage library
76 Figure 28. S chematic of immunofluorescence analysis. Im m unoFluoresence (IF) analysis uses an antibody raised against the phage, but in this case the antibody contains a docking domain for attachment of a secondary antibody that contains a fluorescent tag (FITC). Fl uorescence analysis can be done qualitatively on a fluorescence microscope to compare binding to select particles, or quantified on an overall surface using a fluorimeter. E E E E E E E E E E E E Add substrate WashingE: Horseadish peroxidase (HRP) : Enzymatic Substrate, 3,3 ,5,5 tetramethylbenzidine : Monoclonal anti M13 antibody : Blue product in enzymatic reaction with HRP Figure 29. S chematic of enzyme linked immunosorbent assay. Enzyme Linked ImmunoSorbent Assay ( ELISA) uses commercially available antibodies which have been raised to bind to this specific M13 phage. The antibodies have an attached enzyme (HRP) which can provide a quantitative measure of the amount of enzyme (and therefore phage) that are bound to a surface. The reaction product of the enzyme with a particular substrate (TMB) produces a blue color from which the absorbance can be measured at the wavelength of 630 nm
77 (a) (b) (c) Figure 210. M easurement of the surface coverage of a phage clon e on the IZO surface. (a) The fluorescence image of phage clones on substrate sheet ; (b) convert ed from color image to black and white image using image J ; (c) The threshold from (b) Figure 211. 12 mer amino acid sequences of selected phages from InZnO with low pH elution
78 (a) InZnO Sapphire SiO2Si STTLNNTTWRLY TFKYSHELESRG TKNMLSLPVGPG MNRPSPPLPLWV (b) InZnO Sapphire SiO2Si TEAHRQSMTLTW GNHSTTNMHPPL ASQITHFPRPPW Figure 212. I mage s of ELISA plate containg phage clones eluted from IZO with low pH elu t ion buffer. (a) An image of ELISA plate containing 4 phage clones from the third biopanning with low pH buffer elution after 10 minut e development time: these phage clones with expressed 12 mer peptides (STTLNNTTWRLY, TFKYSHELESRG, TKNMLSLPVGPG, and MNRPSPPLPLWV) (b) An image of ELISA plate containing 3 phage clones from the second biopanning with low pH buffer elution after 10 minute development time: these phage clones with expressed 12 mer peptides (TEAHRQSMTLTW, GNHSTTNMHPPL, and ASQITHFPRPPW).
79 (a) 0.000 0.010 0.020 0.030 0.040 0.050 0.060 0.070 0.080 0.090 STTLNNTTWRLY TFKYSHELESRG TKNMLSLPVGPG MNRPSPPLPLWV 12-mer expressed peptide seuqences UV absorbance (a.u.) InZnO Sapphire SiO2 Si (b) 0.000 0.010 0.020 0.030 0.040 0.050 0.060 0.070 0.080 0.090 0.100 TEAHRQSMTLTW GNHSTTNMHPPL ASQITHFPRPPW 12-mer expressed peptide sequences UV absorbance (a.u.) InZnO Sapphire SiO2 Si Figure 2 13. UVVisible absorbance of the enzymatic substrate solution (a) from ELSA plate in Figure 2 12 ( a ) (b) from ELISA pl ate in Figure 2 12( b)
80 Figure 214. Images of the immuofluorescence analysis for t he negative control experiment O nly monoclonal anti M13 antibody and monoclonal anti mouse IgG FITC were incubated with IZO, sapphire, Si, and SIO2 respectively. No fluorescence light was Figure 215. IF analysis for t he phage clone TKNMLSLPVGPG. It showed binding affinity to the both of IZO and sapphire. However, this phage clone pr eferentially b ound to IZO because the intensity of fluorescence from IZO is markedly stronger than sapphire. (c) Si (100) (d) SiO 2 (amorphous) (a) InZnO (amorphous) (b) Sapphire ( 0001 )
81 Figure 216. IF analysis of t he phage clone MNRPSPPLPLWV It showed preferential binding affinity to sapphire, rather than IZO. This indi cated this phage clone might be Before Elution After elution 4 times Figure 217. Photography of t he etching InZnO with using the low pH elution buffer (a) (b) (c) (d)
82 Figure 218. 12 mer amino aci d sequences of selected phages from InZnO with high salt elution Figure 219. I mage of ELISA plate containing 6 phage clones from the third biopanning with high salt elution buffer after 10 minute development time
83 Table 2 1. List of styles inluded i n Version 9.0 of the MS Word Formatting Figure 2 20. UVVisible absorbance of the enzymatic substrate solution from ELISA plate in Figure 2 19 Figure 221. IF image s for t he phage clone SHAPDSTWFALF from the low pH elution (a) InZnO (amorphous) (b) Sapphire (0001) (c) Si (100) (c) Si (100) (d) SiO 2 (d) SiO 2 ( amorphous ) 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 AGFPWSTHSSWL TNSSSQFVVAIP HPKPSLDAARLV SHAPDSTWFALF TLMYAQPHQSKT ALDDLRARFLPP 12-mer expressed peptide sequences UV adsorbance (a.u.) InZnO Sapphire Si SiO2
84 (a) (b) Figure 222. R eleasing test for the mixture of phage clones. (a) A picture of the electro active device for releasing phage clones. (b) Th e electro releasing test for phage mixture comprised of phage clones with expressed peptide sequences FNGRHGTTDHPT, TNPLSSWTFPTY, and ASQITHFPRPPW obtained from low pH elution. White light Fluorescence light 0 mV 0 mV 1400 mV 1400 mV
85 Figure 223. E lectro rele a sing test for the phage clone NMTMSFPTYPIA. It was not rel eaed from the IZO device by applying voltage. The fluo rescence intensity on this phge clone also was not changed. White light Fluorescence light 0 mV 0 mV 1400 mV 1400 mV
86 CHAPTER 3 PARTICLE SEPARATION USING INORGANIC BIND ING PEPTIDES DESIGNE D BY PHAGE DISPLAY TECHNI QUES 3.1 Motivation Phosphate rock is the main r a w material in the production of fertilizers. In the whole world the United State s has been the biggest phosphate producer. Florida phosphate contributes 80% of the phosphate production in the United State s Current ly, the phosphate rock in Florida phosphate reserves is of low concentrate grade due to high dolomite contamination. The MgO content in dolomite affects the filtration of phosphoric acid, an important intermediate during the production of fertilizers. In general, the MgO content in a phosphate concentrate must be less than 1% for commercial applications. Currently, some conventional flotation a gents such as fatty acid s can float francolite which is a fluoroaptite ( Ca5(PO4)3 (F, OH) ), at a high recovery ra te. However, they also float dolomite. Thus, there are no commercial flot a tion agents that can differentiate francolite from dolomite (CaMg(CO3)2) very well due to their similar surface properties such polarity The goal of this study is to demonstrate th e feasibility of using a biotechnology approach to particle based applications that may benefit from a high degree of specificity achieved by molecular recognition between peptides and inorganic surfaces. More specifically, a biopanning approach based on phage display will be used to screen for peptides that have high binding affinity and selectivity for francolite particles P hage display is a combinatory approach that utilizes a commercially available library of phage (virus) particles that have a set of foreign oligonucleotides encoding short random peptide inserts into the protein tails (pIII) of the phage. I f a high degree of selectivity can be demonstrated in this study, inorganic binding peptides based on phage display techniques have the potential to lead to a design for a peptide amphiphile to accomplish particle separation by flotation. For this, the inorganic binding
87 peptides c ould be covalently linked to a hydrocarbon tail ( e.g. a short hydrocarbon chain or a short hydrophobic peptide chain) using a terminal amino acid with either amine or cysteine functionality, creating a peptide amphiphile. It is anticipated that the hydrophobic tail of the peptide surfactant would segregate to the air bubbles, such that selective flotation occurs with the specifically targeted particles attached to the froth (Figure 3 1) 3.2 Background and Significance In nature, a particular mineral is usually found in association with one type of rock. For instance, cassiterite usually accompanies granitic rocks. With mechanical and chemical weathering, a mineral deposit may be con v erted to ore consist ing of the extractable mineral of interest and gangue, extraneous rocky materials. Within an ore, the mineral of interest is found in a sufficien t concentration for recove ry In the mining industry, a main issue is to extract a valuable mineral from gangue. In addition, m ining process es called ore dres sing, or milling, mainly reduce the volume of waste minerals (e.g. gangue minerals) to concentrate valuable minerals which red uc es the shipping and handling costs. 3.2.1 Liberation The m ining process contains two parts: liberation of valuable minerals and separation of these valu abl es from gangue minerals. During liberation of valuable minerals from gangue minerals, comminution is necessarily involved for grinding or crushing lumps of ore into a mixture of relatively clean particles of valuable minerals and gangue minerals. The degree of liberation is the key toward the efficient recovery in the mining process. The degree of liber ation is defined as the percentage of free mineral p articles in total ore content. In general, v aluable minerals usually strongly bind with gangue minerals. Thus, different constituen t s are contained in free mineral particles to cause much middling and a l ow degree of liberation when an ore is
88 ground. Wills et al . developed a new approa ch in which breaking stresses w ere directed at mineral crystal boundaries to crush rock without breaking grain boundaries. 3.2.2 Concentration After liberating mineral particles from ore minerals by crushing and grading, th e following processing separates free valuable mineral particles from gangue mineral particles. This process is known as concentration. The efficiency of concentration can be quantified by measuring the ra tio of concentration: the ratio of weight of f r eed to the weight of concentrate. The ra tio of concentration is relat ed to the grade of concentrate which is defined as the content of end valuable mineral products of interest in the materials. The ratio of concentration usually increases with the grade of co ncentrate. In addition, recovery is another measure in the efficiency of concentration: the percentage of the total valuable minerals in the ore that is recovered in the concentrate. There is the inverse relationship between recovery and the grade of conce ntrate (Figure 3 2). After liberating mineral particles from gangue minerals, the ore is subsequently classified into two or more products in a concentration process. During the concentration process, separation of mineral particles of interest from gangue minerals is usually base d on some difference in physical or chemical surface properties between valuable minerals and gangue minerals. In the mining industry, some physical separation methods have been widely adopted for concentrating ore  including photometric sorting  gravity concentration  magnetic concentration  electrical con ductivity concentration  and f ro th flotation [111, 114120] For the separation of mineral particles, f roth flotation is currently the most widely adopted method in the concentration of mineral particles of interest and the method of interest in this study Froth flotation is based on the difference in physi ochemical surface properties between
89 valuable mineral particles and gangue mineral particles  By adjusting the chemical environment within a flotation pulp, an aqueous ground ore suspension, it is possible to enable froth flotation to achieve a specific separation of valuable mineral s from a complex ore. Originally, froth flotation was aimed at the separation of sulphides of copper, lead, and zinc. However, various chemical agents were designed to enhance the specific separation of minerals in froth flot ation. Thus, froth flotation ha s been exp a nded to separate many kinds of minerals from an ore  (e.g. oxides: hematite and cassiterite; oxidized minerals: malachite and cerussite; non metallic minerals: phosphates, fluorites, and fine coal). In the froth flotation process, an air stream is pumped into a pulp solution to create air bubbles. Gangue minerals within a flotation pulp tend to display aerophobic properties and are kept in the pulp solution. However, valuable minerals with areophilic properties attach to air bubbles and are lifted to the surface of the liquid phase isolate d from gangue minerals (Figure 33  ) Large particles may not work well in froth flotation because their gravity force is larger than buoyant force. In this case, larger particles can not be lifted by air bubbles. Thus, froth flotation is mainly suitable for fine particles (5  In addition, most minerals show the tendency toward aerophobic properties in their natural state. Thus, the interaction between mineral particles and air bubbles is too weak to lift mineral particles in the froth phase. In order to enhance the efficiency of flotation concentration, a collector, a kind of surfactant, is added into the pulp to be absorbed onto the valuable mineral particle surface to render them areophil ic ity which can help mineral particles adhere onto air bubbles  As an i deal collector, it needs to have complete wet ting for the valuable mineral particles to adhere to the air bubbles. That means that there is the high work of adhesion at the interface between mineral particle and the air bubble: the work of adhesion indicates the required energy
90 that separates min eral particles from the air bubbles. The wetting activity of a collector on a particle can be quantified by measuring the contact angle of the mineral particle surface/bubble interface (Figure 34  ) The relationship between the work of adhesion and the contact angle is expressed as followed: Ws/ a w/ a (1 Where Ws/a: the work of adhesion w/ a: the surface energy between water and air Based on the equation above, mineral particles become more and more a e r ophilic (increase in Ws/a) with increasing contact angle. The interaction between the mineral particle and the air bubble also become s strong to promote the flotation ability of mineral particles due to the enhancing surface tension. Collector molecules c an be classified into two types: nonionizing and ionizing compounds [111, 119, 120] Nonionizing collectors are insoluble in an aqueous medium and render mineral aerophilic abilities by forming a thin film on the mineral surface. Ionizing collectors are b asically composed of polar head goups and hydr ophobic tails. Collectors can adsorb to the particle sur face through chemical, electr ostatic or physical interaction between polar groups and the mineral particle surf ace. Then, the hydrophobic tails are segregated into air bubbles. Ionizing collectors can be further divided into the anionic type and the cationic type based on the ionic properties of their polar head groups (Figure 3 5  ). 3.2.3 C hallenges of C urrent M ining Processing T echnologies The objective of mining processing is to separate valuable minerals from gangue minerals regardless of the concentration methods applied. However, the current techni ques of mineral particle separation are never perfect because they always isolate two or more mineral into concentrate. Thus, the big issue has been to improve the purity of the concentrate. Furthermore,
91 the concentration efficiency of current techniques i s still low in the treatment of fine size particles. This often causes the loss of valuable minerals. For example, 30% of phosphate mined in Florida was lost due to the discard of fine particles  With the development of froth flotation, it ha s become the most important and powerful concentration method widely used in the mining industry. Froth f lotation can be used to treat various kinds of ores such as oxides and metal oxid es. In addition, it also can selectively separate valuable mineral particles by changin g the chemical environment of the flotation pulp in the large scale. Even though froth flotation ha s been prove n a s a useful application in the concentration of mineral particles, it still has a limitation in the separation of two minerals with similar su rface properties. For instance, dolomite is the main impurit y in Florida phosphate ore which is an indispensible material during the production of fertilizers. However, dolomite severely affects the formation of phosphoric acid, an intermediate of fertiliz ers. Dolomite has not been effectively removed from phosphate ore during the conventional flotation process due to their similar surface properties. Several flotation techniques were developed to reduce the content of dolomite in the phosphate ore s  However, there are some disadvantages that limit their commercial application in flotation processes. First, those flotation processes involve a secondary separation process in order to remove dolomite from the final phosphate concentrate. Secondly, many flotation reagents are not cost effective due to high dose s and high price. Finally, the most complex problem is the chemical environment that is changed from the alkaline to the acidic medium. This made the flotation operation very complex. El Midany et al proposed reactive flotation to improve the separation efficiency of dolomite from phosphate ore in one step  : high dolomitic phosphate pebbles are dipped
92 with a surfactant composed of polyvinyl alcohol (PVA) to form a coating around an individual particle. Subsequently, the particles with PVA coating are immersed into an acidic me dium. When dolomite particles with PVA coating are exposed to the acidic solution, CO2 is produced through chemical reaction between carbonate and acid, and is confined between a particles surface an d a PVA coating layer When the average density of the dolomite particle with the CO2 layer around it is smaller than the acidic medium, the dolomite particles will float. However, reactive flotation only works well for dolomite particles in the scale of mm. In addition, the composition of francolite also contai ns carbonate which has the ability to produce CO2 in an acidic solution. This may reduce the specificity of the reactive flotation because the acid ic solution reacts with both francolite and dolomite. Furthermore, the acidic solution also causes a harsh en vironment 3.2.4 Role of Inorganic Binding Peptides in Separation of M inerals In the past two decades, phage display techniques have proved t he ability to screen for peptides with highly binding affinity for inorganic surfaces, including semiconductors ( CdS, GaN) [2, 18, 24, 26, 46, 47] ceramics (SiO2, TiO2) [2, 18, 28, 30] metals (Au, Ag, Pt) [2, 18, 21, 23, 32, 33, 35, 3942] and minerals (CaCO3, Fe2O3, hydroxyapatite) [2, 18, 31, 53] .The whole procedure of screening for inorganic binding peptides is called biopanning (Fugre 17). After three biopanning rounds peptide binding peptide seque nces determined by the DNA of selected phage clones di s play not only strong binding affinity to a target material but also the ability of the molecular recognition to it. In this study, francolite pe bbles provided by Mosaic, Inc. are cho sen as a target material to pan for 12mer francolite binding peptide sequence Those peptide sequences are expected to display strong and specific binding affinity to francolite, and are considered as the hydrophilic head group of a collector. If the francolite binding peptides were modified with hydrocarbon tails which could render francolite particles hydrophobicity, the
93 specific separation of francolite could be achieved from phosphate ore by using those peptides amphiphilies in the flotation process. 3.3 Materials and Methods 3.3.1 Materials Francolite p articles (Ca5(PO4)3(F OH) ) : Francolite pebbles with the size of mm were provided by Mosaic, I nc. from their pl ants at two different locations, Four Corners Francolite and South Fort Meade Francolite Dolomite r ock (CaMg(CO3)2) : Dolomite rock was purchased from Wards Natural Science Ingredients for p hage d isplay s creen ing : All materials involved in phage display are the same as the one used in chapter 2 3.3.2 M ethod s Grinding : Dolomite rock (Wards Natural Science) was first crushed into small piece s less than 5 mm in size using a hammer. Subsequently, planetary ball milling which consist s of 250 ml ZrO2 bowl and 40 Zirconium balls (FRITSCH Pulverisette 5) was used to ground small dolomite pieces into powder under 150 rpm for 30 minutes. For francolite pebble (Mosaic, Inc.), planetary ball milling was also applied to cr ush francolite pebble s into powder under the same operation conditions as dolomite. All grinding process es w ere operated by Dr. Wolfgang Sigmunds group in the department of Materials Science and Engineering, University of Florida. C leaning p rocedure of t arget p articles : One hundred mg powder was weighed and loaded into 1.5ml microfuge tube. 1 ml CH3OH/acetone mixture (1:1) was added into the tube. The p owde r pellet was dispersed by pipetting and then was vortexed for 5 10 minutes Subsequently, the powder was sonicated for 20 minutes in ultrasonic bath to break the clumps. After sonication, t he powder was vortexed quickly to re disperse. The powder was cent rifuged at
94 13200 rpm for 1minute The supernatant was removed and 1ml 70 % ethanol was added onto the powder for sterilization. The p owder was vortexed for 510 minutes, and then was sonicated for 20 minutes in ultrasonic bath. A fter sonication, the powde r was cen trifuged at 13200 rpm for 1 min ute. The supernatant was removed and 1ml 0.5 % PC buffer was added. The powder was sonicated for 30 minutes The powder was vortexed quickly to resuspend. The powder was centrifuged at 13200 rpm for 1 min. The supern atant was removed and 1ml 0.5 % PC buffer was added. Then, 100 l of powder solution was aliquoted into each sterile 1.5ml microfuge tube. Samples were centrifuged at 13200 rpm for 1minute After centrifugation, supernatant was removed from each tube. Samples were washed twice with 1 ml sterilized DI water, and rinsed with 1 ml ethanol. Samples were centrifuged at 13200 rpm for 1 min., and then removed supernatant from each tube Samples were dried under vacuum. Phage d isplay p rotocol : The experimental steps of the phage display protocol are the same as the description in C hapter 2 except for the target materials. Binding a ffinity assay : Image processing software, Image J, was used to measure the surface coverage of phage clones on francolite powders. In this method, the ratio (R1) of particle surface area to picture a rea was first calculated in white light (after white light image was set as the threshold), and then the ratio (R2) of fluorescence area to picture area (after fluorescence image is threshold). Phage surface coverage on powders is defined as the ratio of R 2 to R1. The following example shows how phage surface coverage is calculated. Threshold
95 R1=0.0327 (The ratio of black area to the whole picture under white light ) R2=0.0315 (The rat io of black area to the whole picture area under fluorescence light ) Surface cov erage of phage on powder = R2/R 1 = (0.0315/0.0327) x 100% = 96.9% M easurement of z eta p otential : The measurement of the zeta potential was mainly used to determine the net sur face charge of particles. In this measurement, dolomite/francolite particles were dispersed respectively in PC buffer at pH 2.0, 4.0, 7.0, 8.0, and 10.0. Then, Zeta Reader Mark 21 ( in Particle Engineering Research Center, University of Florida) was utilize d to determine the net surface charge of particles. 3.4 Results and Discussions In this chapter, we mainly explore the effect of M13 phage with expressed inorganic binding peptide s on the separation of francolite particles from dolomite particles. For th e conventional flotation stra t eg ies it has been a tough issue to concentrate francol ite particles from dolomitic ph osphate ore in Florida due to the similar surface properties such as polarities be tween the fran colite and dolomite particle s Phage display techniques ha ve demonstrated the potential in the selection of inorganic binding peptides with specific binding affinity to a target material in the past two decades [2, 1429, 3136, 3942, 4449, 51, 53, 56, 57, 6063, 128] However, selected inorganic binding peptides based on phage display techniques may display cross binding affinity to some inorganic m a terials in addition to the target material Thus we first pan ned for the dolomite binding peptides using phage display techniques for Threshold
96 understanding the properties of their peptide sequences i n the beginning of this study After panning for the francolite binding peptides some of the francolit e binding peptide sequ ences similar to dolomite binding peptide sequences could be excluded from the sequence data base of francolite binding peptides by applying bioinormatics, the alignment tool of amino acid sequences. Table 31 indicate s that the concentration of eluted phage was in the scale of 105 pfu/ml in each biopanning round. That means the diver si ty of phage clones is enough to amplify the phage pool for the subsequent biopanning. After selecting the eluted phage with binding affinity to dolomite powder, the chemical properties (such as polarity) of some dolomite binding peptides sequences are summari z ed in Figure 36. Basically, o ne notable feature is that acidic amino acids are common in dolomite binding peptides In addition, the 3rd round dolomite binding peptide sequences are always accompan ied by a block of hydrophilic amino acids, but do not seem to contain a very large hydrophobic block region compared to 2rd round dolomite binding peptides. The binding afffinty of dolomite binding peptides are evaluated as the surface coverge of phage clones with expressed 12 mer binding peptide sequence s with IF analysis. Generally, images taken at a high magnification is more accurate than at a low magnification. However, the drawback using a high magnification is to cause a smaller view of an image. In order to compromise the accuracy and view of an image at different magnifications, images of phage clone ADYFTARPGPIT selected from a dolomite system were taken at an objective magnification 20X and 40X respectively (Figure 37). At the objective magnification 20X, the surface coverage of this phage clone was 94.0% + 3.0% which is no significant different from 96.4% + 3.6% at the objective magnification 40X statistically. Thus, surface coverage of phage clones were taken IF ima ges to calculate their surface coverage on mineral surfaces at the
97 objective magnification 20X. Figure 3 8 show s that the binding affinity of expressed 12mer dolomite binding was classfied into strong, medium, and weak binders. Those binders, especially strong binders, could be input data in bioinformatics for excluding the similar peptide sequences from francolite binding peptides. Figure 3 9 descibes the difference in the surface coverage of phage clones with strong, medium, and weak binders respectivel y. The binding behavior of binding peptides could be affected by the conformation of the peptides, the chemical properties of amino acid residu es of peptides, and the interaction between inorganic subsrates and peptide binders, especially electrostatic in teraction. For expressed 12 mer inorganic peptides, their chain length is too short to display an obvious conformation. Thus, the electrostatic interaction may play an important role in the determination of peptide binding affinity to a target material. I n the exploration of electrostatic interaction between inorganic materials and peptide binders, the charge character of the inorganic surfa ce and peptides need to be conf i rmed first For the charge character on inorganic surface s the z e ta poteial c an be applied to determine this. Figure 310 indicated the ze ta potential of dolomite /francolite particle surface as it varies with pH The isoelectric point is at pH 4.3. After pH 4.3, the ze ta potential on the dolomite particle s surface became more and more negative with the increase of pH. Otherwise, the ze ta potential on dolomite particles was more and more positive with the decrease of pH. The charge character of the inorganic peptides is dependent on pKa of amino acid residues and pH of the aqueous medium ac cording to Henderson Hasselbalch as follow  : pH = pKa + log10([base]/[acid]) The charge property of carboxy late group and amino group as side groups of amino acid residues with pH of aqueous medium can be described in the form of Henderson Hasselbalch below  : RCOOH = RCOO1 + H+, pH = pKa + l og10[RCOO1]/[RCOOH]
98 RNH3 + = RNH2 + H+, pH = pKa + log10[RNH2]/[RNH3 +] For carboxy late groups, they tend to bear negative cha r ge due to the abundance of RCOOwhen the pH of the aqueous medium is larger than the pKa. Furthermore, amino groups easily accumu late positive charge while the p H of the aqueous medium is smaller than the P Ka. For instance, the phage clone: QTLPLPLTIA H P showed higher surface coverage at pH 7.4 ( 71.7% + 7.5% ) than pH 4.0 (19.0% + 4.5% ) and pH 10.0 (3.6% + 0.8% ) (Figure 311). In the expressed 12mer amino acid sequence, QTLPLPLTIA H P, t he most of amino residues did not bear charge in the range of p H 4.010.0 except histidine. Histine contains a tertiary amine side g roup with P Ka 6.0. At pH 7.4, some amine groups of histidine w ere partia lly protonated to bear positive charge, and dolomite surface showed the negative zeta potential (Table 32 ). In this case QTLPLPLTIA H P tend ed to adhere onto dolomite particle surface s due to the electrostatic attraction force (Figure 3 11) At pH 10.0, no charge was produced on QTLPLPLTIA H P. Thus, the interaction between this peptide and dolomite surface was from the weak Van der Waals force which caused low surface coverage ( 3.6% + 0.8% ) (Figure 311 ) A t pH 4.0, e ach h istidine was fully protonted to be a r a +1 charge. However, the dol omite surface tended to have the positive zeta potential to repell the peptide s with the posive charges to cause lowe r surface coverage than the one at pH7.4. Although the dolomite surface contained the positive charge s in high percentage, little negative charge still existed to attact peptide via electrostatic interaction. That i s why the surfa ce coverage of this peptide at pH 4.0 is higher than pH 10.0 (Figure 311). With repesct to the phage clone: GFASDPSSSPWT, aspartic acid (Asp, D) with P Ka 3.9 dominated the adhesion of the peptide s on dolomite surface. This peptide sequence display ed higher surface coverage ( 78.0% + 6.7% ) on the dolomite surface at pH 4.0 than pH 7.4 and pH
99 10.0 because the partial dissociated asparti c acid with little negative charge was adhered onto the dolomite surface with the highly positive zeta potential because of the electrostatic attraction force (Table 3 3 and Figure 312) At pH 7.4 and pH 10.0, aspartic acid was fully dissociated to bear a 1 charge. In addition, negative charge density is higher and higher on dolomite surface with increasing pH. Thus, the electrostatic repulsive force is stronger at pH 10.0 than pH 7.4. Thus, the surface coverage of expressed peptide GFASDPSSSPWT at pH 7.4 was higher than at pH10.0 (Figure 312). In this study, our goal was to use biopanning techiques to achieve the separation of francolie particles from dolomite paticles specifically. In order to select binding peptides with suitable binding affinity to dolomitic phosphate ore in Florida, we chose Four Corner (FC) and South Fort Meade (SFM) pure francolite particles provided by Mosaic, Inc to pan for francolie binding peptides. Table 3 4 and Table 3 5 showed the phage titers for FC and SFM fr a ncoli te: the concentration of eluted phage was in the scale of 106 pfu/ml Thses results reflected that the diversity of phages with binding affinity to francolite particles was satisfied: a n acceptable diversity of bound phages is usually above 104 pfu/ml Alt hough those two target materials are all francolites, their purity is different, and thus their overall morphology/structure may display some differ ences in surface characteristic. Thus, we anticipated that a consensus amino acid sequence of francolite bin ding peptides from those samples may be different, but also may have some similarities in chemical characteristics. T he francolite binding peptide sequences from FC and SFM samples are listed in Figure 3 13 and Figure 314 after three biopanning rounds, r espectively. The peptide sequences of phage clones from FC and SFM were roughly comprised of a block of polar and a block of hydrophobic amino acids. In addition, one or two basic amino acids were also found in most of
100 the peptide sequences. Although some similarities can be seen between the chemical properties of the 12mer peptide sequences from FC and SFM, they were not identical (as expected), and a complete consensus sequence was not found at 3 rounds of panning for either one. This is not unusual for inorganic materials, particularly powders, which contain many different surface characteristics and thus binding sites. The differences between the peptide sequences for FC and SFM samples are roughly quantified in Table 36 with respect to block le ngths a nd basic amino acids. In addition, the 3rd round dolomite binding peptide sequences do not seem to contain a very large hydrophobic block region compared to francolite binding peptide sequence (Table 3 7). The examination of the binding affinity of franc olite binding peptides was show n in Figure 3 15. M ost of the francolite binding peptides display ed preferential binding affinity to francolite, even if there were similar surface properties between the fran colite and dolomite particles e specially, the peptide WSITTYHDRAIV which displayed highly preferential binding a ff inity to francolite (Figure 3 16) Thus, francolite binding peptides have the potential to function as hydrophilic head groups linked with hydrophobic molecules to float francolite particles from phosphate ores specifically.
101 Table 3 1. Phage titers for the phages selected from dolomite particles Elution F irst round S econd round T hird round Cum -p (pfu/ml) Cam -p (pfu/ml) Vtatoal Cum -p (pfu/ml) Cam -p (pfu/ml) Vtatoal Cum -p (pfu/ml) Cam -p (pfu/ml) E1 (1st elution) 4.5x 105 2. 7x 1013 50 2. 7 x105 1.4 x 1013 50 5.3 x105 1.5 x 1012 E2 (2nd elution) 1.2x 104 5.0 x 1011 75 5.2 x104 4.6 x 1012 75 4.7 x104 1.0 x 1012 E3 (3nd elution) 1.3x 104 5. 2 x 1010 125 3.2 x104 1.4 x 1011 125 2.3 x104 1.3 x 1011 E4 (4nd elution) 3.3x 103 6. 5x 108 145 4.8 x103 3. 3x 1010 100 1.1x104 8.8 x 109 Total 375 350 Table 3 2. C harge of Histidine in peptide sequence QTLPLPLTIAHP and zeta potential of dolomite surfaces in 3.3% PC buffer at pH 4.0, pH 7.4, and pH10.0 pH 4.0 pH 7.4 pH10.0 Zeta potential on dolomite surface s (mV) +2.2 7.6 15.0 His, H +1 + 0 Interaction force Electrostatic repulsive Electrostatic attraction Van der waals Table 3 3. C harge of A spartic acid in peptide sequence GFASDPSSSPWT and Zeta potential of dolomite surfaces in 3.3% PC buffer at pH 4.0, p H 7.4, and pH10.0 pH 4.0 pH 7.4 pH10.0 Zeta potential on dolomite surface s (mV) +2.2 7.6 15.0 Asp D 1 1 Interaction force Electrostatic attraction Electrostatic repulsion Electrostatic repulsion -: Partial negative charge
102 Table 3 4. P hage t iters for phage selected from Forur Corner francolite pebbles Elution F irst round S econd round T hird round C um p (pfu/ml) C am p (pfu/ml) V tatoal C um p (pfu/ml) C am p (pfu/ml) V tatoal C um p (pfu/ml) C am p (pfu/ml) E1 (1st elution) 2.0 x 106 1.9x 1013 27 3.9 x106 8.2x 1012 100 3.3 x106 1. 9x 1013 E2 (2nd elution) 1.4 x 106 2.1x 1013 25 1.7 x106 9.7x 1012 85 1.2 x106 2.4x 1013 E3 (3nd elution) 1.7x 106 6.3x 1013 83 1.8 x105 1. 0x 1013 82 1.7 x106 9.1x 1013 E4 (4nd elution) 5.4x 105 5.2x 1012 100 4.2 x105 1.0x 1013 82 1. 3 x105 6.9x 1012 Total 235 349 Cum p: The concentration of umamplified phages Cam p: The concentration of amplified phages Vtatoal: The total taken volumn of eluted phages Table 3 5. P hage titers for phage selected from South Fort Meade francol ite pebbles Elution F irst round S econd round T hird round C um p (pfu/ml) C am p (pfu/ml) V tatoal C um p (pfu/ml) C am p (pfu/ml) V tatoal C um p (pfu/ml) C am p (pfu/ml) E1 (1st elution) 6.1x 105 3.7x 1012 100 1.6 x107 1. 0x 1013 74 3.0 x106 5.0x 1012 E2 ( 2nd elution) 7.9x 105 6.1x 1012 61 1.0 x107 7.3x 1012 100 1.2 x106 8.3x 1012 E3 (3nd elution) 2.8x 105 6.6x 1012 56 5.3 x106 9.5x 1012 79 8.9 x106 5.4x 1012 E4 (4nd elution) 6.1x 105 7 5x 1012 50 2.9 x106 8.7x 1012 84 6.9 x105 5.7x 1012 Total 267 337 Cum p: The co ncentration of umamplified phages Cam p: The concentration of amplified phages Vtatoal: The total taken volumn of eluted phages
103 Table 3 6. C hemical properties of peptide sequences selected from FC and SFM francolite pebbles Basic amino acid (amino aci d# / peptide) Block of Polar amino acid (amino acid# / block) Block of N on polar amino acid (amino acid# / block) FC 1.7 Long (3.2) Short (nonobvious, 2.1) SFM 1.5 Short (2.3) Long (4.1) Table 3 7. The comparison of peptide sequences selected from f rancolite and dolomite powders Basic amino acid (amino acid# / peptide) acidic amino acid (amino acid# / peptide) Block of Polar amino acid (amino acid# / block) Block of N on polar amino acid (amino acid# / block) Francolite 1.6 Low (0.5) Long (2.8) 3 D olomite 1.5 High (1.0) Short (1.6) Non obvious (1.7) Figure 31. Schematic of a peptide amphiphile as a flotation agent. In a suspension containing the target particles of francolite, along with other undesired impurities, the pe ptide selected for francolite binds specifically to those particles, which are then collected at the interface of air bubbles due to the hydrocarbon tails, enabling flotation separation. Air bubble F F F F F F F F F Hydrophobic tail Francolite binding peptide
104 RecoveryGrade of concentrate RecoveryGrade of concentrate Figure 32. A curve representing the relationship between recovery and gra de of concentrate Flotation Cell PulpAir bubbles Agitator Air Mineralised froth Mineral with a surfactant coating Flotation Cell PulpAir bubbles Agitator Air Mineralised froth Mineral with a surfactant coating Figure 33. S chematic for the flotation process Agitator
105 Figure 34. Contact angle between bubble and particle in an aqueous medium Cation is water repellent. Based on pentavalent nitroge Dithiophosphates Carboxylic Non ionising Liquid, nonpolar hydrocorbons which do not dissociate in water Ionising Cationic Anionic Oxyhydryl Sulphydryl Based on organic and sulpho acid groups Based on bivalent sulphur Xanthates Sulphonates Sulphates O O C S O O O O S O O O S S C O O O S S PCation is water repellent. Based on pentavalent nitroge Dithiophosphates Carboxylic Non ionising Liquid, nonpolar hydrocorbons which do not dissociate in water Ionising Cationic Anionic Oxyhydryl Sulphydryl Based on organic and sulpho acid groups Based on bivalent sulphur Xanthates Sulphonates Sulphates O O C S O O O O S O O O S S C O O O S S P Non ionising Liquid, nonpolar hydrocorbons which do not dissociate in water Ionising Cationic Anionic Oxyhydryl Sulphydryl Based on organic and sulpho acid groups Based on bivalent sulphur Xanthates Sulphonates Sulphates O O C S O O O O S O O O O S O O O S O O O S S C O S S C O O O S S P O O S S P Figure 35. T ypes of collectors used in the flotation concentration
106 Clone # The first round No consensus Clone # The second round Clone 1 A D Y F T A R P G P I T Clone 2 M P N P H L A L P H G S Clone 3 S P N P P A N A V I T N Clone 4 Q T L P L P L T I A H P Clone 5 A N D G L A T R P R D L Clone 6 N I Q T T H L F P L P R Clone 7 G M E L H S K L P I Y R Clone # The third round Clone 1 Q H H T L S T A P Y L Y Clone 2 Q Q N Y L T Q N I G R A Clone 3 Q L T V D N N H Q G N D Clone 4 H Y T E A S F D I R T R Clone 5 H T E P A N W Y P H T H Clone 6 D T N F V K A P R Q P N Clone 7 S T D M S P S P M S H S Clone 8 T S E N N Y A V E S F H Clone 9 A P K G L T N T S Q L M Clone # The first round No consensus Clone # The second round Clone 1 A D Y F T A R P G P I T Clone 2 M P N P H L A L P H G S Clone 3 S P N P P A N A V I T N Clone 4 Q T L P L P L T I A H P Clone 5 A N D G L A T R P R D L Clone 6 N I Q T T H L F P L P R Clone 7 G M E L H S K L P I Y R Clone # The third round Clone 1 Q H H T L S T A P Y L Y Clone 2 Q Q N Y L T Q N I G R A Clone 3 Q L T V D N N H Q G N D Clone 4 H Y T E A S F D I R T R Clone 5 H T E P A N W Y P H T H Clone 6 D T N F V K A P R Q P N Clone 7 S T D M S P S P M S H S Clone 8 T S E N N Y A V E S F H Clone 9 A P K G L T N T S Q L M Figure 3 6. S ummary of expressed 12 mer dolomite binding peptide sequences. The letters correspond to the commonly used one letter code for the 20 amino acids. Figure 3 7. Images of the phage clone ADYF TARPGPIT on dolomite particles at the objective magnification 20X and 40X. (a) Image under white light at 20X; (b) image under fluorescence light at 20X; (c) image under white light at 40X; (d) image under fluorescence light at 40X (Surface coverage at 20X : 94.0% + 3.0%; surface coverage at 40X: 96.4% + 3.6%) Hydrophilic, basic Hydrophobic Polar, uncharg ed Hydrophilic, acidic
107 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 SLNCSLASSACR STDMSPSPMSHS QLTVDNNHQGND SNITPQTSTPSL ADYFTARPGPIT DTNFVKAPRQPN TSENNYAVESFH MPNPHLALPHGS YVAHEVINLHHT HLGGSIARIPEQ LPTMMNNNWNQR NPIPDTRNHRLV LPSRVQELWWPA TANWHPARTLLT GFAHHSWAPDRA APVAHAFPQAMM QTLPLPLTIAHP EFQTPLRANVSF ANDGLATRPRDL GMELHSKLPTYR TVLHNKSPDQSQ NIQTTHLFPLPR HTEPANWYPHTH SPNPPANAVTTN QIPNAVDLYWSP SPNIGIAKNMLY QHHTLSTAPYLY NFDELTMPNYRT APKGLTNTSQLM THYTRGLSPFSL MNDTKWAAPQGL VDIHSGTWPLSY HYTEASFDIRTR NQNYDAEQLITP QQNYLTQNIGRA GFASDPSSSPWT Expressed 12-mer peptides Surface Coverage (%) Red: Strong binder s Yellow: Medium binders Green: Weak binders Figure 3 8. C lassification of phage s with expressed 12mer peptide sequences into three categories based on the surface coverage on dolomite Figure 3 9. The binding affinity of three representatives of phages on dolomite particles The definition of binding affinity is the surface coverage of phages on dolomite powder surface (R2/R1); (a) The phage clone SNITPQTSTPSL exhibit s strong binding ( 94.3% + 3.5% ) ; (b) The phage clone SPNPPANAVTTN exhibit s moderatebinding affinity (7 1.7% + 7.5% ) ; (c) The phage clone THYTRGLSPFSL exhibit s weak binding affinity (47.3 % + 6.3% ) (a) Strong binding affinity (b) Moderate binding affinity (c) Weak binding affinity Medium binder Fluorescence light Weak binder Fluo rescence light S trong binder White light Medium binder White light Weak binder White light S trong binder Fluorescence light
108 -40.0 -30.0 -20.0 -10.0 0.0 10.0 20.0 30.0 40.0 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 pH Zeta potential (mV) Francolite Dolomite Figure 3 10. Zeta potential f or francolite/dolomite powders in the PC buffer Figure 3 11. Images of the p hage clone QTLPLPLTIAHP on dolomite surfaces (a)~(c) I mages under white light at pH4.0, pH 7.4, an d pH10.0 respectively; (d)~(e) i mages under fluo rescence light at pH4.0, pH 7.4, and pH10.0 respectively. The surface coverge of this phge clone on dolomite surfaces is 19.0% + 4.5% 71.7% + 7.6% and 3.6% + 0.8% at pH 4.0, pH 7.4, and pH 10.0 respectively. (b) White light, pH 7.4 (c) White light, pH 10.0 ( e ) Fluorescence light, pH 7.4 (a) White light, pH 4.0 (d) Fluorescence light, pH 4.0 ( e ) Fluorescence light, pH 10.0
109 Figure 3 12. Images of p hage clone GFASDPSSSPWT on dolomite surfaces (a) ~ (c) I mage s under white light at pH4.0, pH 7.4, and pH10.0 respectively; (d)~(e) images under fluorescence light at pH4.0, pH 7.4, and pH10.0 respectively. The surface coverge of this phge clone is 78.0% + 6.7% 9.7% + 2.5% and 2% + 0.7% at pH 4.0, pH7.4, and pH10.0 respectively. Clone # The first round No consensus Clone # The second round 1 Y S Q P T L W A L T S R 2 T N H T F W F P A E F G 3 T S P P Q V A Y P T L S 4 S H V G N P Y I S A T L 5 E H W Q D N W M R W I T 6 E K I S D Y A W P Y R T 7 D H R S I S A F P N P P Clone # The third round 1 G S N G I W F N L A H R 2 T N S N W T P F W P L P 3 W S I T T Y H D R A I V 4 S S M T H Q H A R V D T 5 A S L Q H T A L L N Q N 6 H Y G V Q A P H N S N S 7 W S Y V P F A R Q V N Q 8 M E Q F Q S A G N P G W 9 M E Q T Y P S S H R P G 10 N I G H R V N S P F P Q 11 A P R L L S D N T Y N V 12 A H L E D I T V H D G S 13 H M P H H V S N L Q L H Clone # The first round No consensus Clone # The second round 1 Y S Q P T L W A L T S R 2 T N H T F W F P A E F G 3 T S P P Q V A Y P T L S 4 S H V G N P Y I S A T L 5 E H W Q D N W M R W I T 6 E K I S D Y A W P Y R T 7 D H R S I S A F P N P P Clone # The third round 1 G S N G I W F N L A H R 2 T N S N W T P F W P L P 3 W S I T T Y H D R A I V 4 S S M T H Q H A R V D T 5 A S L Q H T A L L N Q N 6 H Y G V Q A P H N S N S 7 W S Y V P F A R Q V N Q 8 M E Q F Q S A G N P G W 9 M E Q T Y P S S H R P G 10 N I G H R V N S P F P Q 11 A P R L L S D N T Y N V 12 A H L E D I T V H D G S 13 H M P H H V S N L Q L H Hydrophilic, acidic Hydrophilic, basic Polar, uncharged Hydrophobic Figure 3 13. Four Corner francolite binding peptide sequences ( e ) Fluorescence light, pH 7.4 (a) White light, pH 4.0 (b) White light, pH 7.4 (c ) White light, pH 10.0 (d) Fluorescence light, pH 4.0 ( e ) Fluorescence light, pH 10.0
110 Clone # The first round No consensus Clone # The second round 1 Y Q S T R T H A E A S P 2 H S V Q T Y A R P L P S 3 H G L T V Q R P E Q M M 4 G H V V T N S V W M L P 5 I D Y S A P S R Y A N S 6 D P F P Q R V N Y L K R 7 A A S F Q H S A T A N L Clone # The third round 1 Y I G S Q T N E R Y S P 2 Q N L I N W P P P R F S 3 V S H S E Y N R A A T Y 4 A S D N R T M V L M F P 5 Q G Y T M F V A A E P L 6 Y S L P R H L V S L P P 7 H Y N P E M P S S H N A 8 H S M P H M G T Y L I T 9 T I T P S Y L L A H G P 10 T K N M L S L P V G P G 11 H H H Q T L R P A P F A 12 A V P H R V G G L H S L Clone # The first round No consensus Clone # The second round 1 Y Q S T R T H A E A S P 2 H S V Q T Y A R P L P S 3 H G L T V Q R P E Q M M 4 G H V V T N S V W M L P 5 I D Y S A P S R Y A N S 6 D P F P Q R V N Y L K R 7 A A S F Q H S A T A N L Clone # The third round 1 Y I G S Q T N E R Y S P 2 Q N L I N W P P P R F S 3 V S H S E Y N R A A T Y 4 A S D N R T M V L M F P 5 Q G Y T M F V A A E P L 6 Y S L P R H L V S L P P 7 H Y N P E M P S S H N A 8 H S M P H M G T Y L I T 9 T I T P S Y L L A H G P 10 T K N M L S L P V G P G 11 H H H Q T L R P A P F A 12 A V P H R V G G L H S L Hydrophilic, acidic Hydrophilic, basic Polar, uncharged Hydrophobic Figure 3 14. S outh F ort M eadefrancolite binding peptide binding sequences 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 AHLEDITVHDGS APRLLSDNTYNV TNHTFWFPAEFG WSITTYHDRAIV VHFRIATPYFSP TPPPSEITTSPP MEQFQSAGNPGN ASLQHTALLNQQ MEQTYPSSHRPG GDDVNTMRARPL NIGHRVNSPFPQ SAHGTSTGVPWP EKISDYAWPYRT DLFYDANNVHAG DRAPLIPFASQH YLAHSSNNKILF FANTSSPVVHPF DIRTEPHNTSNS YSQPTLWALTSR YNLTPLPKGNAM SPTSLLPTQAHY WSYVPEARQVNQ QQYVAYPIMKAL LAPVRPIFSMEV AQINLDNHARWF TNSNWTPFWPLP HYGVQAPHNSNS EHWQDNWMRWTT SSMTHQHARVDT LSASSPTTTATW Expressed 12-mer peptide seuqences Surface coverage (%) Francolite Dolomite Figure 315. The surface coverage of phage clones with expressed fr ancolite binding peptides on fra nc olite and dolomite respectively
111 Figure 316. Images of t he pha ge clone with expressed peptide WSITTYHDRAIV on francolite particles and dolomite partic les. (a) A schematic for the glass slide containing francolite powder and dolomite powder ; (b) francolite particles under white light ; (c) francolite particles under fluorescence light ; (d) the interface between francolite and dolomite under white light ; (e) the interface between francolite and dolomite particles under fluorescence light ; (f) dolomite particles under white light (g) dolomite particles under fluorescence light.
112 CHAPTER 4 M13 PHAGE AMPHIPHILE AS COLLECTORS FOR TH E SPECIFIC SEPARATIO N OF MI N ERAL PARTICLES 4.1 Motivation In the Florida phosphate industry, froth flotation is the most effective method in separatin g phosphate particles from contamination by other minerals However, it is difficult to remove dolomite contamination from Florida dolomitic phosphate ores because the characteristics of Florida dolomitic phosphate rock are CO2 substitution, porous structure, and cryptocrystalline structure in which the crystalline nature is vague and can only be characterized by a polarizing light mic roscope In the work here, we hypothesized that t he s elected phage clones with expressed 12 mer francolite binding peptides would act as bio amphiphilic flotation agents because some of the domains of coat proteins on M13 phage bodies are hydrophobic, havi ng aerophilic ability to attract them to air bubbles (Figure 41). For example, i n the froth flotation process (Figure 3 1) francolite binding peptides would display preferential binding affinity to francolite particles, and with the attached phage, thes e particles would be attracted to the air bubbles Then, w hen air bubbles with attached phage particle s rise to the top of the flotation cell, the francolite particles will be come separated from dolomite particles that remain on the bottom of the flotation c ell 4.2 Background and Significance In the froth flot a tion process, the concentrate of valuable minerals is largely de pen dent of the collectors. Currently, the conventional collectors are usually organic surfactants such as fatty acid s. Those chem ical reagents are not environmentally friendly. In addition, the mining industry may also face the fluctuation of price due to the shortage of certain flotation agent s A side from using organic surfactants as a collector in 1993 Smith et al first explore d Mycobacterium phlei a kind of bacterium, that could be applied as collectors to concentrate hematite in the flotation
113 process as collectors  T he main composition on the surface of Mycobacterium phlei is lipid s which have hydrophobic tails  In the aqueous medium at pH 2.5 ( the isoelectric point of these lipids ), the head region s of those lipid s tend to bear a a negative charge and easily adhere onto the surface of mineral particles bearing neutral, or positive charge s The hydrophobic domain on Mycobacterium phl ei render ed the h ematite particles hydrophobic to enhance the adh e sion with air bubble s In the froth flotation using Mycobacterium phlei the specific separation of hematite particles could be achieved Florida contributes 80% of USA phosphate products. H owever, phosphate reserves in Florida usually contain high MgO contamination due to a high quantity of dolomite  As mentioned before, one of the biggest challenges is to remove dolomitic carbonate from the phosphate ore s using the froth flotation technology In the usage of microorganism s as flotation agents, Bacillus licheniformis JF 2 was found to have the ability to interact with the metal cations through their anionic cell walls composed of teichuronic acid, teichoic acid, and peptidoglycan  Beveridge et al. proved cationic metal ions were mainly captured in teichuronic acid and teichoic acid domains  In addition, it was also confirmed th at Bacillus licheniformis show ed a stronger binding affinity to Mg2+ than Ca2+ [132, 135] This selective binding of Bacillus licheniformis to Mg2+ had been demonstrated in the separation of dolomite particles from phosphate ore by Misra et al.  Thus, Misra i utilized Bacillus licheniformis as a collector for removing the dolomite particles from phosphate ore s At pH 10.012.0, Bacillus licheniformis displayed preferential binding affinity to Mg2+. However, the fraction of floated dolomite is below 20% which is far lower than the 60% commercial floated fraction  That means Bacillus licheniformis is not a good collector for floating dolomite powder because of the competition of OHcations. Besides viable microorganism s
114 freeze dr ied bacteria also can be used as flotation collectors  Smith et al used freeze dr ied Staphylococcus carnosus as a collector to float apatite, calcite, and quartz at pH 9.0 respectively. The recovery rate of apatite wa s 50% 70% with microbe dosage. However, Staphylococcus carnosus also can float calcite with the recovery rate of 25% 55%. Thus, Staphylococcus carnosus is a good collector, but lacks of high recognition cap ability in the differen tiation between apatite and calcite. As mention ed above, microorganism s ha ve been considered as alternative flotation agents in mining processing due to the molecular recognition of their cell walls and nontoxicity compared to chemical flotation reagents. However, all of these bacteria only display preferential binding affinity to a c ertain metal cation, such as Mg2+, in a low pH or high pH solution which is still harsh to the environment. Furthermore, the floated fraction of mineral particles of interest is usually too low to satisf y the requirement of commercialization in a neutral e nvironment. Finally, the required quantity of microorganisms is usually 10100 times greater than the conventional surfactants in order to reach a satisfied recovery rate of valuable minerals. Thus, it is desirable that the dose of microorganism is reduced to the level of the chemical agents as flotation agents for commercial applications In this study, phage display techniques  were adopted to select the phage clones with expressed inorganic binding peptides. The amino acid sequences of inorganic binding peptides can display strongly specific binding affinity to a certain inorganic substrate based on their chemical composition, surface morphology, particle size, pH, and crystal structure. Thus, phage display biopanning can be performed to select inorganic bindng peptides with strong preferential binding affinity to minerals of inte rest at pH 7.0. I n addition, some domains of the coat proteins on M13 phage bodies may be hydrophobic. Thus, th e M13 phage clones acting as
115 bioreagents could be used under mild environment al conditions where they have the potential to be used as collectors for floating mineral particles of interest in one step within a neutral aqueous medium. In addition, the M13 phage is a kind of filamentous virus with the expressed inorganic binding peptides at one end of the phage body If the adhesion of M13 phages on the mineral surface is through those inorganic binding peptides the binding dom ain of selected M13 phage clones may be in the scale of n ano m eters far smaller than other microorganisms in the scale of micro meter s In this case, the dose of M13 phages might be anticipated to be less, and possibly close to the dose used with chemical surfactants 4.3 Materials and Methods 4.3.1 Materials Phage d isplay i ngredients : See Chapter 2 M 13 p hage c lones : Three selected phage clones with expressed 12 mer peptide sequences WSITTYHDRAIV TNSNWTPEWPLP and SSMTHQHARVDT Fluorapatite : Ca5(PO4)3 (F OH) : Crystalline fluorapatite with green to brown color was purchased from Ward s Natural Science, Inc. Dolomite : CaMg(CO3)2: Crstallized dolomite with large and gray cleavages was purchased from Ward s Natural Science, Inc. Florida f rancolite pebbles : Ca5(PO4)3 (F, OH) : Francolite pebbles were provided by Mosaic, Inc. after the flotation process. 4.3.2 Methods Immunofluorescence (IF) a nalysis : See chapter 3 3.3.2 methods Flotation t est : In the bench top flotation experiment, a flotation column (Figure 42(a)) was used to float mineral particle s in the scale of 1 to 10 gram. A plastic tube served as the connection between the bottom of the flotation column and a n air gas cylinder. A n air stream adjusted by the gas valve on the bottom of the flotation colu mn was pumped into the aqueous medium to create air bubbles in the flotation column. Before pumping the air stream into the
116 flotation column, 1 gram of mineral particle such as francolite (size: 50250 m ) dolomite (size: 2060 m ) or a mixture of francolite/dolomite was incubate d with the desired phage clone ( 0.96 mg per gram mineral ) in 5 ml PC buffer solution at PH 7.4 on an agitator for 1 hour. After that, the mixture of mineral particle/phage solu tion and a magnetic stir bar were added to fill into the flotation column, and then 145 ml distilled water (at p H 7.4) was filled into the flotation column. When the mineral particles were precipitat ed (Figure 42 (b)), the air stream was pumped in to the fl otation column and the magnetic stir bar was allowed to agitate the mineral particles for 3 minutes. Floated particles on the top of the flotation column were collected b y filtration and dried to weigh the floated mineral particles (Figure 42 (c)). Meas urement of contact a ngle s : In the measurement of contact angles a mineral rock was first cutted into mineral sheets with approximate dimension s of 2 cm x 2 cm x 2 mm by using a precision sectioning saw (Isomet 1000, Buehler, Inc.) with a 4 15HC diamond w affering blade (Buehler, Inc.). The surface of each mineral sheet was po lished using 600 and 1200 grift SiC sand papers (Alleid, Inc.). Subsequently, the mineral sheets were rubbed over a wet polishing cloth on which a 0.05m alumina suspensions was applied Then, the mineral sheets were washed using distilled water for removing th e residue of alumina suspension on their surfaces. In this method, the captive bubble tec hnique was adopted to determine the advancing and recedi ng contact angles for dolomite and fluorapatite sheets using a Contact Angle Goniometer image system equipped with an auto pi pettin g system (Rame Hart, Inc., USA). T he polished mineral sheets were attached on the sample holder (Figure 4 3 ( a) ) Then, they were immersed into an optical acrylic cell containing distilled water or 3.3% PC buffer solution at pH7.4 (Figure 43 ( b) ) An air bubble created by the auto pipetting system was attached to the mineral surface
117 thro ugh a plastic needle (Figure 4 3 ( c) ) Th e sample holder was raised until the area between the air bubble and the mineral surface was in contac t. In this situation, the in flection point could be observed. In order to explore the effect of M13 phage clones on the contact angle of minerals, minera l sheets were immersed in 3 ml phage/PC solution (pha g e concentration: 1011 pfu/ml in 3ml PC buffer), and agitated for 50 minutes. Subsequently, mineral sheets were rinsed with the solution which is the same as the liquid in the acrylic cell. The measurement was made after 10 minutes after the mineral coated with M13 phage clones was immersed into the solution in the acrylic cell. At least f ive measurements were made for each condition, and the angle of each side of the air bubble was recorded at 23. Ind uctively coupled p lasma atomic e missi on spectroscopy (ICP AES) a nalysis : In this study, ICP AES (Optima 3200KL, Perkin Elmer) was used to determine the content of Magnesium/Calcium in the minerals. In this analysis, 20 mg mineral sample was added into the 20 ml concentrated HCl solution, and was stirred for approximately 24 hours in order to prepare 1000 mg/L (ppm) mineral solution. Subseque ntly, 1000 mg/L (ppm) was diluted down to 100 mg/L (ppm) using distilled water. The emission of calcium was detected at 315.89 nm, 317.93 nm, and 393.37 nm. T he emission of magnesium was detected at 279.77 nm and 285.23 nm. A calibration curve of Magnesium/Calcium was established from 100 ppm to 0.1 ppm before the measurement of the samples: the 1000 ppm standard Magnesi um/Calcium solution (RICCA Chemical Company) was diluted into 100 ppm, 10 ppm, 1 ppm, and 0.1 ppm respectively. After that, the mineral solution with 100 mg/L (ppm) was loaded into ICP and the magnesium/ calcium content was determined in the mi neral sample. The concentration of Mg and Ca was co nverted into the content of MgO and CaO based on the calculation as follows:
118 The content of MgO (%) = (CMg/Cdissovle d mineral ) x (MMgO / MMg ) x 100% The content of CaO (%) = (CCa/Cdissovled mineral ) x (MCa O / MC g ) x 100% Where CMg : The measured concentration of Mg (mg/L or ppm) CCa : The measured concentration of Ca (mg/L or ppm) Cdissovle d mineral : The concentration of mineral dissolved in HCl MMgO : The molecular weight of MgO (g/mol) MCa O : The molecular weight of CaO (g/mol) MMg : The atomic weight of Mg (g/mol) MC g : The atomic weight of Ca (g/mol) 4.4 Results and Discussions A successful f lotation process is often dependent on the choice of a collector with the molecular configuration like a surfactant where the hydrophilic head selectively adsorbs on the valuable mineral and the hydrophobic tail renders valuable mineral hydrophobicity to reduce the work of adh es ion to an air bubbl e. Phage dis play techniques have been proven as a potential in the selection of phage clones with the expressed inorganic binding peptides that have specific binding affinity to a target material. In addition, the hydrophobic domains of M13 coat proteins m ay have the potential to function as the hydrophobic tail of a collector. In this study, M13 phage clones with expressed francolite binding peptides were used to test the hypothesis that selective flotation could be achieved, whereby francolite particles w ould be concentrated from the mixed mineral containing the dolomite contamination. In this study, there were three phage clones selected from the francolite pebbles after three biopannings for exploring the effect of expressed 12mer francolite bindi ng peptides on the performace of the selective flotation Their expressed francolite binding pept ide sequences are WSITTYHDRAIV, TNSNWTPFWPLP, and SSMTHQHAVDT respectively. In IF analysis
119 (Fig ure 4 4 45), the phage clone WSITTYHDRAIV ws found to prefere ntially adhere onto francolite particles, but not dolomite particles. However, phage cloneTNSNWTPFWPLP tended to bind onto dolomite particles, rather than francoli te. With respect to phage clone SSMTHQHARVDT it did not seem to differentiate between francolite versus dolomite The surface co verage of these phage clones is summari z ed in Table 4 1. In the beginning of the flotation test, a fatty acid that is widely used as a conventional collector used widely for the concentration of francolite from phosphat e ores in the mining industry, was first used to evaluate the recovery rate of pure mineral s s uch as francolite and dolomite in the benchtop flotation process (Figure 42). Here, the recovery rate of pure minerals (a positive control) was defined as follo ws: Recovery rate (%) = ( Wfloated minerals / W feed minerals) x 100% Where Wfloated minerals: The weight of the floated minerals Wfeed minerals: The weight of feed minerals Figure 4 6 shows that the fatty acid didn t differentiate francolite f rom dolomite due to their similar recovery rate. In the mining industry, the dose of fatty acid is usually 0.5mg per gram mineral to achieve a satisfactory recovery rate of francolite (approximately 100%) In the benchtop flotation test, the recovery rat e of francolite was 52.5% in the dis tilled water solution with p H 7.4 using this suggested dose of fatty acid but the recovery rate of dolomite was as high as 60% When the selected phage clone WSITTYHDRAIV was used as the collector in the flotation proce ss, the recovery rate of francolite was clearly higher than dolomite (Figure 4 7) At 52.7% recovery rate of francolite, the dose of this phage clone is 0.96 mg per gram mineral which is twice as much as fatty acid (0.5 mg per gram mineral in 52.5% recover y rate of francolite). The
120 dose of M13 phage clone is far lower than the microorganisms described early, in which the dose of the general microorganisms is 10100 times higher than chemical agents. The low required quantity of M13 phages could possibly b e explained by the adhered domain on the mineral surface. As mentioned in chapter1, the configuration of M13 phage is like a flexible filamentous rod (Figure 1 5) The strong interaction between M13 phage and a target material is usually through the expres sed inorganic binding peptides at the end of the pIII proteins [30, 35] The dimension of PIII protein is less than 10 nm. For other microorganism s used as collectors in the fro th flotation techniques, their binding area is in the scale of micrometer on the mineral surface. That may be why the dose of M13 phage is in the same scale as chemical agents. Figure 4 8 shows another phage clone TNSNWTPFWPLP that can still differentiate francolite from dolomite, but it tend ed to float dol omite even though this clone was selected from francolite pebbles. However, the phage clone SSMTHQHARVDT behaved like a fatty acid, with limited specificity, but also with a low recov ery rate of minerals (Figure 4 9). Figure 4 10 summarize s the recovery rate of francolite by comparing those three selected phage clones as collectors These results support my hypothesis that a hi gh surface coverge of phages can lead to a high recovery rate of francolite. It was considered that the high recovery rate of mineral might be relat ed to the high hydrophobicity on the mineral surface after being coated with the phage The hydrophobicity can be evaluated with the measurem ent of contact angle (Table 4 2 to 4 3); the higher the contact angle, the more hydr ophobic the mineral surface. The p hage clone WSITTYHDRAIV increased the contact angle of francolite from 12.5to 50in 3.3% PC buffer (Table 4 2 and Figure 411) due to 99% surface coverage of this phage clone on the francolite surface. In contrast this p hage clone displayed weak binding affinity to dolomite (11% surface coverage). As was expected the
121 contact angle of dolomite modified with this phage clone just was increased only around 2, as com pared to bare dolomite Thus, high surface coverage of pha ge clones can render the minerals to become more hydrophobic If the solution medium was changed from 3.3% PC buffer to distilled water, the contact angle of the francolite surface modified with this phage clone was still maintained at 50(Table 4 2 to 43) This can be explained by the hydrophobic coat protein of M13 phages which shields the hydrophilic mineral surface from the influence of hydrophilic solution medium. Thus, I believe the hydrophobic domains of M13 coat protein rendered the francolite hy drophobic and decreased the work of adhesion between the francolite surface and the air bubble. That may be why this phage clone tends to float francolite, rather than dolomite. The phage clone TNSNWTPFWPLP that displayed strong binding affinity to dolomite was found to increase the contact angle of dolomite from 15.6 to 48.4 ( in 3.3% PC buffer ) but the contact angle of francolite was not changed very much. Thus, this phage clone would be anticipated to preferentia lly float dolomite as was observed With respect to the phage clone SSMTHQHARVDT it didn t increase the contact of mineral of either mineral presumably due to its weak binding a ffinity to both of the two miner a ls (Table 4 2 to 43) Thus, this would explain why this phage clone did not have a good ability to float either of minerals. In addition, the contact angle of the bare minerals was lower in 3.3% PC solution than distilled water medium. Especially, the contact angle of the bare dolomite surface which was reduced from 28.7 to 15.6when distilled water was substituted with 3.3% PC buffer ( Table 4 2 to 43). This res ult also reflected the decrease i n the recovery rate of the bare dolomite (from 38.5% to 25.4%). Thus, PC buffer is able to function as a depressant to dolomite. The most important requirement of a collector is to concentrate valuable mineral from ore particles. Althou gh M13 phages had shown the ability of floating the pure minerals it is
122 necessary to validate their effectiveness in the separation of francolite f rom phosphate ores containing the dolomite contaminations In this study, we simulated the dolomitic Florida phosphate ore by mixing francolite with dolomite in the weight rati os of 1:1, 3:1, a nd 1:3. Subsequetly, the selected M13 phage clones were applied to concentrate the francolite from the mixed minerals. The purity of francolite can be evaluated using ICP AES in the floated minerals. In order to determine the content of francolite in the floated minerals, the content of magnesium and calcium are neces sary to be measured in the pure francolite and dolomite resepectively using ICP AES. Table 4 4 showed the result of ICP AES measurement for pure francolite and dolomite. The content of magnesium mainly appeared in the pure dolomite. Thus, the result of th e ICP AES measurement shown in table 4 4 can be used to calculate the percentage of dolomite in the floated mixture of minerals after the conte nt of magnesium is determined in the mixed floated minerals. If the concentration of dissolved minerals is the sa me, the content of magnesium is calculated as follows: 11.91 XDolomite + 0.17 XFrancolite = 11.91 XDolomite + 0.17 (1XDolomite) = CMg mixed mineral XDolomite = ( CMg, mixed mineral 0.17)/10.74 XFra ncolite = 1 XDolomite W here XDolomite: th e percentage of dolomite in the floated mixed mineral XFrancolite: th e percentage of francolite in the floated mixed mineral CMg, mixed mineral: the concentration of magnesium in the floated mixed mineral Futhermore, the recovery rate of francolite and dolomite is also deduced in the mixed minerls based on XDolomite as folloes:
123 Floated percentage of dolomite = (XDolomite* WFloated mix ed mineral) / WTotal dolomite Flotated percentage of francol ite = (XFrancolite* Wfloated mix ed mineral) / Wtotal Francolite = ((1 XDolomite) Wfloated mix ed mineral) / Wtotal Francolite Where WFloated mix ed mineral: the weight of floated mixed mineral WTotal dolomite: the weight of total dolomite in the mixed mineral before the flotation process Wtotal Francolite: the weight of total francolite in the mixed mineral before the flotation process When the mixed mineral (francolite: dolomite = 1:1) wa s floated using fatty acid and selected M13 phage clones as collectors respectively, the ICP AES analysis for the floated mixture is shown in Table 4 5. The floated mineral contained 10.5% MgO conta mination using fatty acid as a collector In this case, th e content of dolomite occupied 52.3% of the floated minerals. This also indicates that the fatty acid didn t enable differentiation francolite over dolomite. When the phage clone WSITTYHDRAIV (phage #1) was cho sen to float the mixed minerals, the content of MgO contamination was lowered from 10.5% to 6.7% in the floated minerals and the content of CaO indicated the puri ty of francolite was increased from 35.2% to 36.8%. T he percentage of dolomite was reduced from 5 2.3% to 39.3% in the floated mixture of minerals When phage #2 functioned as a collector, it behaved similarly to the fatty acid due to its similar binding affinity to francolite and dolomite. The only difference between phage #2 and fatty acid is that phage #2 caused low er recovery rate of mine ral than fatty acid due to its weak binding affinity to mineral s With respect to phage #3, it tended to concentrate dolomite, rather than francolite. Thus, the content of MgO caused by phage #3 was 11.6 % even higher than the MgO content of 10.5% obtained using fatty acid as a collector. In addition, the intrinsic recovery
124 rate of dolomite was 21.3% in 3.3% PC buffer in the absence of collectors (See Figure 4 10) For the mixed mineral (francolite: dolomite = 1:1) the minimum content of dolomite contaminat ion is around 27.6% due to the intrinsic recovery rate of dolomite (21.3%) even if the selected phage clone did not ds i play binding affinity to dolomite. The recovery rate of francolite/dolomite floated from mixed minerals (Table 4 5) matched with the res ults from the flotation of pure minerals (Figure4 6 to Figure 49) When the ratio of francolite to dolomite was changed from 1:1 to 1:3 or 3:1, the selected phage clones versus fatty acid fo llow ed basically the same trend of flotation propotional to the f eed mixture of minerals composed of francolite and dolomite (Table 46 and Table 4 7, respectively ). In particular the ability of concentrating francolite was enhanced gradually using phage #1 versus the fatty acid as a collector with decreasing the rati o of francolite to dolomite in the fe e d mixture Overall the preferential binding affinity of M13 phages to francolite and the amphiphilic properties of M13 phages can envision that M13 phages has a great potential to achieve the concentration of francol i te from dolomitic phosphate ores.
125 Table 4 1. Summary for the surface coverage of selected phage clones on francolite and dolomite The expressed 12 mer peptide sequences on selected phage clones The surface coverage (%) Francolite Dolomite WSIT TYHDRAIV 94.1 + 5.7 15.7 + 4.0 TNSNWTPFWPLP 61.7 + 6.5 96 .0 + 2.6 SSMTHQHARVDT 5 0.7 + 3.1 46.5 + 4.0 Table 4 2. C ontact angle on francolite and dolomite in 3.3% PC buffer solution at pH7.4 Contact angle Francolite ( D egree) Contact angle Dolomite ( D egr ee) Advancing Receding Average Advancing Receding Average No P hage 11.2 + 1.7 14 .0 + 3.7 12.5 + 2.7 15.6 + 3.2 15.5 + 2.1 15.6 + 2.7 WSITTYHDRAIV 49.0 + 3.4 51.0 + 3.6 50.0 + 3.5 17.0 + 4.7 17.4 + 5.2 17.2 + 5.0 TNSNWTPFWPLP 14.2 + 2.3 15.3 + 2.8 14.8 + 2.6 47.6 + 5.6 49.2 + 6.2 48.4 + 6.0 SSMTHQHARVDT 12.0 + 2.3 12.5 + 3.1 12.3 + 2.7 18.5 + 2.9 19.4 + 3.2 19.0 + 3.1 Table 4 3. C ontact angle on francolite and dolomite in distilled water at pH7.4 Contact angle Francolite ( D egree) Contact angle Dolomite ( D egree) Advancing Receding Average Advan cing Receding Average No P hage 13.5 + 2.9 18.0 + 0.5 15.8 + 1.7 28.3 + 1.5 29.1 + 3.2 28.7 + 2.4 WSITTYHDRAIV 49.3 + 3.6 50.5 + 3.8 50.0 + 3.7 28.9 + 3.1 30.2 + 3.4 29.6 + 3.3 TNSNWTPFWPLP 15.6 + 3.5 15.9 + 4.1 15.8 + 3.8 49.8+7.2 49.7 + 7.4 49.8 + 7.3 SSMTHQHARVDT 12.5 + 4.2 13.3 + 3.8 12 .9 + 4.0 29.5 + 2.0 29.2 + 3.3 29.4 + 2.7
126 Table 4 4. ICP AES analysis for the pure minerals Mineral Conc. of dissolved minerals (mg/L) Mg (mg/L) Ca (mg/L) MgO (%) CaO (%) Francolite 100 0.17 + 0.01 33.40 + 0.33 0.28 46.81 Dolomite 100 11.91 + 0.02 19.96 + 0.11 19.85 27.94 Table 4 5. ICP AES analysis of mixed minerals (Francolite:Dolomite=1:1) at pH7.4 Collector Percentage of floated mineral (%) MgO (%) CaO (%) Distribution (%) Recovery rate (%) Francolite Dolomite Francolite Dolomite Fatty acid 53.5 10.5 35.2 47.7 52.3 51.0 56.0 Phage #1 41.0 6.7 36.8 60.7 39.3 50.8 30.9 Phage # 2 34.0 9.5 36.1 52.9 47.1 36.0 32.0 Phage # 3 45.0 11.6 33.2 42.0 58.0 37.8 52.2 Phage #1:WSITTYHDRAIV (F: 99%, D: 11%) Phage #2: SS MTHQHARVDT (F: 52%, D: 41%) Phage #3: TNSNWTPFWPL P (F: 66%, D: 100%) Table 4 6. ICP AES analysis of mixed minerals (Francolite:Dolomite= 3:1) at pH7.4 Collector Percentage of floated mineral (%) MgO (%) CaO (%) Distribution (%) Recovery rate (%) Francolite Dolomite Francolite Dolomite Fatty acid 52.4 5.7 40.7 72.2 27.8 50.5 58.4 Phage #1 43.8 3.7 44.4 83.0 17.0 48.4 29.8 Phage # 2 33.4 5.9 39.2 71.4 28.6 31.7 38.2 Phage # 3 34.0 7.1 37.5 65.8 48.3 31.2 48.3 Phage #1:WSITTYHDRAIV (F: 99%, D: 11%) Phage #2: SS MTHQHARVDT (F: 52%, D: 41%) Phage #3: TN SNWTPFWPLP (F: 66%, D: 100%)
127 Table 4 7. ICP AES analysis of mixed minerals (Francolite:Dolomite= 1: 3) at pH7.4 Collector Percentage of floated mineral (%) MgO (%) CaO (%) Distribution (%) Recovery rate (%) Francolite Dolomite Francolite Dolomite Fatt y acid 55.0 15.2 31.9 24.0 76.0 52.8 55.7 Phage #1 33.7 11.9 32.7 41.8 58.8 56.1 26.2 Phage # 2 35.1 15.0 31.2 27.2 72.8 38.3 34.1 Phage # 3 42.2 18.0 28.2 12.6 87.4 21.2 49.1 Phage #1:WSITTYHDRAIV (F: 99%, D: 11%) Phage #2: SS MTHQHARVDT (F: 52%, D: 41%) Phage #3: TNSNWTPFWPLP (F: 66%, D: 100%) Figure 41. S chematic of froth flotation using M13 phage with expressed francolite binding peptides as collectors for separation of francolite particles from dolomite particles
128 (a) (b) (c) Figure 42. The set up for th e bench top flotation. (a) Design of a flotation tube (while empty) ; (b) the tube filled with 1.00 gram mineral/5 ml PC buffer in 145 ml distilled water; (c) an air stream is pumped into solution to create air bubbles for 3 minutes, and then floated minera ls can be collected on the top of the tube and dried for characterization of the powder. (a) (b) (c) Figure 4 3. The method of the contact angle measurements. (a) The mineral sheet was attached on the sample holder ; (b) t he sample holder was placed side down, and was immersed into t he solution in the acrylic cell; (c) a n a ir bubble was attached to the mineral surface through a small plastic needle.
129 Figure 4 4. Images of francolite particles modified with phage clones selected from francolite. (a)(d) T he phage C lone WSITT YHDRAIV; (b)(e) the phage clone TN SNWTPFWPLP; (c)(f) the phage clone SSMTHQHARVDT; under white light (a) to (c) and fluorescence light (d) to (f) Figure 4 5. Images of dolomite particles modified with phage clones selected from dolomite. (a)(d) the phage C lone WSITTY HDRAIV; (b)(e) the phage clone TNSNWTPFWPLP; (c)(f) the phage C lone SSMTHQHARVDT; under white light (a) to (c) and fluorescence light (d) to (f) (a) ( b ) ( c ) ( d ) ( e ) ( f ) ( d ) ( e ) ( f ) (a) ( b ) ( c )
130 0 10 20 30 40 50 60 70 80 90 100 0 0.5 1.5 2.5 5 Dose of fatty acid (mg /1.0g mineral) Recovery rate (%) Francolite Dolomite Figure 4 6. The recovery rate of pure francolite and dolomite particles versus dose of fatty acid as a collector in di stilled water at pH 7.4 0 10 20 30 40 50 60 70 80 90 100 0.00000 0.00960 0.09600 0.96000 9.60000 Dose of phage (mg/1.0 g mineral) Recovery rate (%) Francolite Dolomite Figure 4 7. The recovery of pure francolite and dolomite particles versus dose of phage clone WSITTYHDRAIV as a collector in 3.3% PC buffer at pH 7.4
131 0 10 20 30 40 50 60 70 80 90 100 0.00000 0.00960 0.09600 0.96000 9.60000 Dose of phage (mg/ 1.0g mineral) Recovery rate (%) Francolite Dolomite Figure 4 8. The recovery of pure francolite and dolomite part icles versus dose of phage clone TNSNWTPFWPLP as a collector in 3.3%PC buffer at pH 7.4 0 10 20 30 40 50 60 70 80 90 100 0.00000 0.00960 0.09600 0.96000 9.60000 Dose of phage (mg/1.0g mineral) Recovery rate (%) Francolite Dolomite Figure 4 9. The recovery of pure francolite and dolomite particles versus dose of phage clone SSMTHQHARVDT as a collector in 3.3%PC buffer at pH 7.4
132 0 10 20 30 40 50 60 70 80 90 100 0.00000 0.00960 0.09600 0.96000 9.60000 Dose of phage (mg/1.0g mineral) Recovery rate (%) WSITTYHDRAIV TNSNWTPFWPLP SSMTHQHARVDT Figure 4 10. Comparison of recovery rate of francolite showing all three representative phage clones as co llectors respectively at pH 7.4 (a) 0 10 20 30 40 50 60 Francolite Dolomite Contact angle (degree) WSITTYHDRAIV No phage (b) 12.5 12.5 50.0 50.0 Figure 4 11. The hydrophobicity of francolite through phage clone WSITTYHDRAIV in 3.3%PC buffer solution at pH7.4 (a) Phage clone WSITTYHDRAIV rendered francolite hydrophobicity as judginh by the increase d contact a ngle from 12.5 to 50 ; (b) Images of an air bubble on francolite surface in 3.3%PC buffer solution (left) no phage coating (right) the addition of phage clone WSITTYHDRAIV. Phage clone WSITTYHDRAIV Enhaced hydrophobicity
133 CHAPTER 5 CONCLUSIONS In this dissertation inorganic binding peptides selected by phage display techniques were used to demonstrate the feasibility of the approach for two applications, reversibly electroactive peptides and specific separation of minerals. In the first case, amorphous IZO was used as a target material to pan for IZO binding pepti des as electro activated peptide linkers for the development of refurbishable bioe snsors. Amorphous IZO led to some consensus binding peptide sequences which I attribute to its homogeneous surface ( i.e. lack of crystal facets that express various crystal lographic planes) It was found that the IZO i norganic binding peptides showed preferential binding affinity to the amorphous IZO as compared to sapphire, Si, and SiO2. Thus, although not studied here, t he specificity of inorganic bindi ng peptides may provide spatial control with multicomponent systems. In the electro releasing test, the relea sing phenome non of the selected phage c l ones with expressed IZO binding peptides was demonstrated that an electro releasing device. These preliminary studies are pro mising suggesting that this technique can contribute to the development of self cleaning biosensors. In the future a n electro biopanning protocol in which an electric field substitutes the traditional low pH or high salt buffers may aid in the selectio n of binding peptides more sensitive to an electric field In the second project of this dissertation, the separation of mineral particles, it w a s found that the phage clone WSITTYHDRAIV displayed preferential binding affinity to francolite The data suggests that this occurs through electrostatic interaction between the expressed binding peptide and the francolite surface Aft er the adhesion of phage s onto the fran colite surface, the coat protein of M13 phages re ndered the francolite particles hydrophobic Furthermore, the body of the phage appeared to serve as an amphiphili c tail while the peptide head was selective for the par ticles of interest, allowing for the separation of francolite particles from dolomite prticles
134 I n addition, M13 phage amphiphiles can lead a satisfactory re covery rate of minerals in a neut ral environment. Thus, it could be valuable to subs tit ute organic chemical agents with M13 phage amphiphile s for reducing the toxicity to our environment. In a cost effective ness consideration, the dose of the filamentous M13 amphiphile is in the same scale as chemical surfactants. In addition, M13 phages can be replicated by infecting host cells. Thus, there is no problem in the shortage of chemical agents. In the amplification process of M13 phages, the main expense is the cost of nutrient media such as the Luria Bertani (LB) Lennox The reason is that the quantity of M13 phage clone and host cell stock solution is negligible compared to the LB media and M13 phages and host cells are reusable For the longevity of M 13 phages, their viability can be maintained over two years at 70C in a refrigerator These advantages suggest that M13 phages are more feasible as flotation agents than other microorganisms as collectors in an industrial froth fl otation process Althoug h most of the inorganic binding peptides selected with phage display techniques can show preferential binding affinity to a target material, they still displayed minor binding affinity to other inorganic m a terials. In the future the bioinformatics approach could potentially improve the specificit y of inorganic binding peptides In this approach, a set of experimentally selected peptides are categorized for their binding affinities (such as quantifying the surface coverage of bound phage s on an inorganic surface by immunofluoresence microscopy). Those experimental pept ide sequences serve as the input data for defining of scoring matrices which include similarities within strong binding sequences and the differences between the st rong and weak binders. N ew computational ly designed peptide sequences have the potential to display enhanced specificity to a target material, depending on the accuracy of the computational prediction.
135 For the commercial 12 mer random peptide used in this dissertation, the diversity of peptide sequences 109 sequences is only a small portion of all possible sequences based probability theory, 2012 sequences Thus, some possible binding peptide sequences may not be obtained from the commercial phage display library. Thus, we may consider extending the diversity of expressed peptide sequences by genetically modifying phage libraries for selecting more experimental binding peptide sequences This could also provide more input data in bioinformatics where the prediction accuracy of new ly designed peptide s can be enhanced further. For understanding the binding mechanism of phage clones with expressed inorganic binding peptides on an inorganic surface, it is necessary to determine if th e binding of phages to the inorganic surface is via the expressed inorganic binding peptides or the phage bod y. This can be done using quartz crystal microbalancedissipation ( QCM D ) with the aid of AFM image analysis. For the electroactive peptides project, r eleasing kinetics could also be explored for comparison among the different reversibly binding clones using QCM D or surface plasmon resonance (SPR). In addition, it would be necessary to determine if the surrounding phage coat proteins and overall mass of phage particles i mpacts the releasing characteristics by comparing phage clones containing the genetically expressed targeted peptides with chemically synthesized peptides of corresponding amino acid sequences. In sum mary the biotechnology approach of phage display can be applied to these types of advanced materials application s or can even enhance the productivity of commodity industries, such a s the mining industry demonstrated in this work.
136 LIST OF REFERENCES 1. Mann S. Molecular recognition in biom ineralization. Nature 1988 Mar;332(6160):119124. 2. Tamerler C, Sarikaya M. Molecular biomimetics: Utilizing naturels molecular ways in practical engineering. Acta Biomater 2007 May;3(3):289299. 3. Schler D. Formation of magnetosomes in magnetotactic bacteria. Journal of Molecular Microbiology and Biotechnology 1999;1(1):7986. 4. CattaneoVietti R, Bavestrello G, Cerrano C, Sara M, Benatti U, Giovine M, et al. Optical fibres in an Antarctic sponge. Nature 1996 Oct;383(6599):397398. 5. Sarikaya M, Fong H, Sunderland N, Flinn BD, Mayer G, Mescher A, et al. Biomimetic model of a sponge spicular optical fiber mechanical properties and structure. J Mater Res 2001 May;16(5):14201428. 6. Katti DR, Katti KS, Sopp JM, Sarikaya M. 3D finite element modelin g of mechanical response in nacre based hybrid nanocomposites. Comput Theor Polym Sci 2001;11(5):397404. 7. Mayer G, Sarikaya M. Rigid biological composite materials: Structural examples for biomimetic design. Exp Mech 2002 Dec;42(4):395403. 8. Harmat V, Naray Szabo G. Theoretical aspects of molecular eecognition. Croat Chem Acta 2009 Nov;82(1):277282. 9. Tompa P. Intrinsically unstructured proteins. Trends BiochemSci 2002 Oct;27(10):527533. 10. Tsuchiya Y, Kinoshita K, Nakamura H. Analyses of homooligomer interfaces of proteins from the complementarity of molecular surface, electrostatic potential and hydrophobicity. Protein Eng Des Sel 2006 Sep;19(9):421429. 11. Kroger N, Deutzmann R, Sumper M. Polycationic peptides from diatom biosilica that di rect silica nanosphere formation. Science 1999 Nov;286(5442):11291132. 12. Liou YC, Tocilj A, Davies PL, Jia ZC. Mimicry of ice structure by surface hydroxyls and water of a beta helix antifreeze protein. Nature 2000 Jul;406(6793):322324. 13. Schonbrun J, Wedemeyer WJ, Baker D. Protein structure prediction in 2002. Curr Opin Struct Biol 2002 Jun;12(3):348354. 14. Willats WGT. Phage display: practicalities and prospects. Plant MolBiol 2002 Dec;50(6):837 854. 15. Smith GP, Petrenko VA. Phage display. C hem Rev 1997 Mar Apr;97(2):391410.
137 16. Kehoe JW, Kay BK. Filamentous phage display in the new millennium. Chem Rev 2005 Nov;105(11):40564072. 17. Hosse RJ, Rothe A, Power BE. A new generation of protein display scaffolds for molecular recognition. Prote in Sci 2006 Jan;15(1):1427. 18. Sarikaya M, Tamerler C, Jen AKY, Schulten K, Baney F. Molecular biomimetics: nanotechnology through biology. Nat Mater 2003 Sep;2(9):577585. 19. Yoo PJ, Nam KT, Qi JF, Lee SK, Park J, Belcher AM, et al. Spontaneous assem bly of viruses on multilayered polymer surfaces. Nat Mater 2006 Mar;5(3):234240. 20. Sanghvi AB, Miller KPH, Belcher AM, Schmidt CE. Biomaterials functionalization using a novel peptide that selectively binds to a conducting polymer. Nat Mater 2005 Jun;4(6):496502. 21. Braun R, Sarikaya M, Schulten K. Genetically engineered goldbinding polypeptides: structure prediction and molecular dynamics. J Biomater Sci Polym Ed 2002;13(7):747757. 22. Naik RR, Brott LL, Clarson SJ, Stone MO. Silica precipitating peptides isolated from a combinatorial phage display peptide library. J Nanosci Nanotechnol 2002 Feb;2(1):95100. 23. Naik RR, Stringer SJ, Agarwal G, Jones SE, Stone MO. Biomimetic synthesis and patterning of silver nanoparticles. Nat Mater 2002 Nov;1(3 ):169172. 24. Whaley SR, English DS, Hu EL, Barbara PF, Belcher AM. Selection of peptides with semiconductor binding specificity for directed nanocrystal assembly. Nature 2000 Jun;405(6787):665668. 25. Lee SW, Mao CB, Flynn CE, Belcher AM. Ordering of quantum dots using genetically engineered viruses. Science 2002 May;296(5569):892895. 26. Seeman NC, Belcher AM. Emulating biology: Building nanostructures from the bottom up. Proc Natl Acad Sci U S A 2002 Apr;99:64516455. 27. Flynn CE, Mao CB, Hayhurs t A, Williams JL, Georgiou G, Iverson B, et al. Synthesis and organization of nanoscale II VI semiconductor materials using evolved peptide specificity and viral capsid assembly. J Mater Chem 2003;13(10):24142421. 28. Eteshola E, Brillson LJ, Lee SC. Sel ection and characteristics of peptides that bind thermally grown silicon dioxide films. Biomol Eng 2005 Dec;22(56):201204. 29. Gaskin DJH, Starck K, Vulfson EN. Identification of inorganic crystal specific sequences using phage display combinatorial library of short peptides: A feasibility study. Biotechnol Lett 2000 Aug;22(15):12111216.
138 30. Chen HB, Su XD, Neoh KG, Choe WS. QCM D analysis of binding mechanism of phage particles displaying a constrained heptapeptide with specific affinity to SiO2 and TiO2. Anal Chem 2006 Jul;78(14):48724879. 31. Li CM, Botsaris GD, Kaplan DL. Selective in vitro effect of peptides on calcium carbonate crystallization. Cryst Growth Des 2002 SepOct;2(5):387393. 32. Yu LT, Banerjee IA, Matsui H. Direct growth of shape c ontrolled nanocrystals on nanotubes via biological recognition. J Am Chem Soc 2003 Dec;125(48):14837 14840. 33. Zhang SG, Marini DM, Hwang W, Santoso S. Design of nanostructured biological materials through self assembly of peptides and proteins. Curr Opi n Chem Biol 2002 Dec;6(6):865871. 34. Kase D, Kulp JL, Yudasaka M, Evans JS, Iijima S, Shiba K. Affinity selection of peptide phage libraries against single wall carbon nanohorns identifies a peptide aptamer with conformational variability. Langmuir 2004 Sep;20(20):89398941. 35. Sano KI, Shiba K. A hexapeptide motif that electrostatically binds to the surface of titanium. J Am Chem Soc 2003 Nov;125(47):1423414235. 36. Morita Y, Ohsugi T, Iwasa Y, Tamiya E. A screening of phage displayed peptides for t he recognition of fullerene (C60). J Mol Catal B Enzym 2004 Jun;28(46):185190. 37. Huang Y, Chiang CY, Lee SK, Gao Y, Hu EL, De Yoreo J, et al. Programmable assembly of nanoarchitectures using genetically engineered viruses. Nano Lett 2005 Jul;5(7):14291434. 38. Allred DB, Zin MT, Ma H, Sarikaya M, Baneyx F, Jen AKY, et al. Direct nanofabrication and transmission electron microscopy on a suite of easy to prepare ultrathin film substrates. Thin Solid Films 2007 May;515(13):53415347. 39. Kantarci N, Tamerler C, Sarikaya M, Haliloglu T, Doruker P. Molecular dynamics simulations on constraint metal binding peptides. Polymer 2005 May;46(12):43074313. 40. Nam KT, Lee YJ, Krauland EM, Kottmann ST, Belcher AM. Peptide mediated reduction of silver ions on engineered biological scaffolds. ACS Nano 2008 Jul;2(7):14801486. 41. Oren EE, Tamerler C, Sarikaya M. Metal recognition of septapeptides via polypod molecular architecture. Nano Lett 2005 Mar;5(3):415419. 42. Tamerler C, Duman M, Oren EE, Gungormus M, X iong XR, Kacar T, et al. Materials specificity and directed assembly of a gold binding peptide. Small 2006 Nov;2(11):13721378. 43. Tomczak MM, Slocik JM, Stone MO, Naik RR. Biofunctionalized nanoparticles and their uses. MRS Bull 2008 May;33(5):519 523.
139 44. Zin MT, Munro AM, Gungormus M, Wong NY, Ma H, Tamerler C, et al. Peptide mediated surface immobilized quantum dot hybrid nanoassemblies with controlled photoluminescence. J Mater Chem 2007 Mar;17(9):866872. 45. Krauland EM, Peelle BR, Wittrup KD, Be lcher AM. Peptide tags for enhanced cellular and protein adhesion to single crystal line sapphire. Biotechnol Bioeng 2007 Aug;97(5):1009 1020. 46. Peelle BR, Krauland EM, Wittrup KD, Belcher AM. Design criteria for engineering inorganic material specific peptides. Langmuir 2005 Jul;21(15):69296933. 47. Sweeney RY, Mao CB, Gao XX, Burt JL, Belcher AM, Georgiou G, et al. Bacterial biosynthesis of cadmium sulfide nanocrystals. Chem Biol 2004 Nov;11(11):15531559. 48. Tamerler C, Sarikaya M. Molecular biomi metics: Genetic synthesis, assembly, and formation of materials using peptides. MRS Bull 2008 May;33(5):504510. 49. Fang Y, Poulsen N, Dickerson MB, Cai Y, Jones SE, Naik RR, et al. Identification of peptides capable of inducing the formation of titania but not silica via a subtractive bacteriophage display approach. J Mater Chem 2008;18(32):38713875. 50. Nam KT, Kim DW, Yoo PJ, Chiang CY, Meethong N, Hammond PT, et al. Virus enabled synthesis and assembly of nanowires for lithium ion battery electrodes Science 2006 May;312(5775):885888. 51. Thai CK, Dai HX, Sastry MSR, Sarikaya M, Schwartz DT, Baneyx F. Identification and characterization of Cu2O and ZnO binding polypeptides by Escherichia coli cell surface display: Toward an understanding of metal oxide binding. Biotechnol Bioeng 2004 Jul;87(2):129137. 52. Donatan S, Yazici H, Bermek H, Sarikaya M, Tamerler C, Urgen M. Physical elution in phage display selection of inorganic binding peptides. Mater Sci Eng C Biomimetic Supramol Syst 2009 Jan;29(1) :1419. 53. Gungormus M, Fong H, Kim IW, Evans JS, Tamerler C, Sarikaya M. Regulation of in vitro calcium phosphate mineralization by combinatorially selected hydroxyapatite binding peptides. Biomacromolecules 2008 Mar;9(3):966973. 54. Yuca E, Karatas A Y, Mustafa G, Sarikaya M, Tamerler C. GFP based bifunctional proteins for real time monitoring of biomineralization. Febs J 2008 Jun;275:374374. 55. Yang XS, Jiang ZB, Song HT, Jiang SJ, Madzak C, Ma LX. Cell surface display of the active mannanase in Ya rrowia lipolytica with a novel surface display system. Biotechnol Appl Biochem 2009 Nov;54:171176.
140 56. Ueda M. Novel highthroughput system for production of new medicines integration and combination with molecular display and combinatorial bioengineerin g. Yakugaku Zasshi J Pharm Soc Jpn 2009 Nov;129(11):12771284. 57. Lin HN, Cornish VW. Screening and selection methods for large scale analysis of protein function. Angew Chem Int Edit 2002;41(23):44034425. 58. Kim SY, Sohn JH, Pyun YR, Choi ES. A cell surface display system using novel GPI anchored proteins in Hansenula polymorpha. Yeast 2002 Sep;19(13):11531163. 59. Weaver RF. Molecular biology, forth edition 2008. 60. Oren EE, Tamerler C, Sahin D, Hnilova M, Seker UOS, Sarikaya M, et al. A novel knowledge based approach to design inorganic binding peptides. Bioinformatics 2007 Nov;23(21):28162822. 61. Brown S, Sarikaya M, Johnson E. A genetic analysis of crystal growth. J Mol Biol 2000 Jun;299(3):725735. 62. Kim E, Xia YN, Whitesides GM. Polymer microstructures formed by moldling in capillaries. Nature 1995 Aug;376(6541):581584. 63. Sano KI, Shiba K. In aqua manufacturing of a three dimensional nanostructure using a peptide aptamer. MRS Bull 2008 May;33(5):524529. 64. Yablonovitch E. Inhibite d spontaneous emission in solidstate physics and electronics. Phys Rev Lett 1987 May;58(20):20592062. 65. Vlasov YA, Yao N, Norris DJ. Synthesis of photonic crystals for optical wavelengths from semiconductor quantum dots. Adv Mater 1999 Jan;11(2):165169. 66. Lee SW, Lee SK, Belcher AM. Virus based alignment of inorganic, organic, and biological nanosized materials. Adv Mater 2003 May;15(9):689692. 67. Lee SW, Wood BM, Belcher AM. Chiral smectic C structures of virus based films. Langmuir 2003 Mar;19(5):15921598. 68. Fuji Keizai USA I. Biosensor market, R&D and commercial implication Kaketa Pacific 2004. 69. S. Hauck CK, S. Drost, H. Wolf. Biosensor System for the Determination of Ochratoxin A. In Proceedings of the 12th European Conference on Soli dState Transducers and the 9th UK Conference on Sensors and their Applications, Sensors Series 1998;1:515518. 70. Datta AK. Novel chemical and biological sensors for monitoring and control of food processing operations. Journal of Food Engineering 1990; 12(3):223 238.
141 71. Bilitewski U, Drewes W, Neermann J, Schrader J, Surkow R, Schmid RD, et al. Compasrison of different biosensor systems suitable for bioprocess monitoring. J Biotechnol 1993 Dec;31(3):257 266. 72. Jenkins DM, Delwiche MJ. Manometric bios ensor for online measurement of milk urea. Biosens Bioelectron 2002 Jun;17(67):557563. 73. Smith DB, Roddick JG, Jones JL. Potato glycoalkaloids: Some unanswered questions. Trends Food Sci Technol 1996 Apr;7(4):126131. 74. Sananikone K, Delwiche MJ, BonDurant RH, Munro CJ. Quantitative lateral flow immunoassay for measuring progesterone in bovine milk. Trans ASAE 2004 Jul Aug;47(4):13571365. 75. Su XL, Li YB. A self assembled monolayer based piezoelectric immunosensor for rapid detection of Escherichia coli O157 : H7. Biosens Bioelectron 2004 Jan;19(6):563574. 76. Tombelli S, Minunni A, Mascini A. Analytical applications of aptamers. Biosens Bioelectron 2005 Jun;20(12):24242434. 77. Jayasena SD. Aptamers: An emerging class of molecules that rival antibodies in diagnostics. Clin Chem 1999 Sep;45(9):16281650. 78. Clark SL, Remcho VT. Aptamers as analytical reagents. Electrophoresis 2002 May;23(9):13351340. 79. Luzi E, Minunni M, Tombelli S, Mascini M. New trends in affinity sensing: aptamers for ligand binding. Trac Trends Anal Chem 2003 Dec;22(11):810818. 80. You KM, Lee SH, Im A, Lee SB. Aptamers as functional nucleic acids: In vitro selection and biotechnological applications. Biotechnol Bioprocess Eng 2003 Mar Apr;8(2):64 75. 81. Potyrailo RA, Conrad RC, Ellington AD, Hieftje GM. Adapting selected nucleic acid ligands (aptamers) to biosensors. Anal Chem 1998 Aug;70(16):34193425. 82. Rogers KR, Mascini M. Biosensors for field analytical monitoring. Field Anal Chem Technol 1998;2(6):317331. 83. Bousse L. Whole cell biosensors. Sens Actuator B Chem 1996 Aug;34(13):270275. 84. D'Souza SF. Microbial biosensors. Biosens Bioelectron 2001 Aug;16(6):337353. 85. Lei Y, Chen W, Mulchandani A. Microbial biosensors. Anal Chim Acta 2006 May;568(1 2):200210. 86. Karube I. Microbial sensor. J Biotechnol 1990 Aug;15(3):255 265.
142 87. Vandyk TK, Majarian WR, Konstantinov KB, Young RM, Dhurjati PS, Larossa RA. Rapid and sensitive pollutant defection by induction of heat shock gene bioluminescence gene fusions. Appl Environ Microbiol 1994 May;60(5):14141420. 88. Ben Israel O, Ben Israel H, Ulitzur S. Identification and quantification of toxic chemicals by use of Escherichia coli carrying lux genes fused to stress promoters. Appl Environ Microbiol 1998 Nov;64(11):43464352. 89. Niculescu M, Nistor C, Frebort I, Pec P, Mattiasson B, Csoregi E. Redox hydrogel based amperometric bienzyme electrodes for fish freshness monitoring. Anal Chem 2000 Apr;72(7):15911597. 90. Ruan C, Wang H, Li Y. A bienzyme elec trochemical biosensor coupled with immunomagnetic separation for rapid detection of Escherichia coli O15 : H7 in food samples. Trans ASAE 2002 JanFeb;45(1):249255. 91. Liu GD, Lin YH. Biosensor based on self assembling acetylcholinesterase on carbon nan otubes for flow injection/amperometric detection of organophosphate pesticides and nerve agents. Anal Chem 2006 Feb;78(3):835 843. 92. Shah J, Wilkins E. Electrochemical biosensors for detection of biological warfare agents. Electroanalysis 2003 Feb;15(3) :157167. 93. Bakker E. Electrochemical sensors. Anal Chem 2004 Jun;76(12):32853298. 94. Watson LD, Maynard P, Cullen DC, Sethi RS, Brettle J, Lowe CR. A microelectronic conductimetric biosensor. Biosensors 1987;3(2):101115. 95. Wilson WD. Analyzing b iomolecular interactions. Science 2002 Mar;295(5562):2103+. 96. Yang VC, Ngo TT. Biosensors and Their applications. New York NY:Kluwer Academic 2000:195212. 97. Tschmelak J, Proll G, Gauglitz G. Optical biosensor for pharmaceuticals, antibiotics, hor mones, endocrine disrupting chemicals and pesticides in water: Assay optimization process for estrone as example. Talanta 2005 Jan;65(2):313323. 98. Palmqvist L, Holmberg K. Dispersant adsorption and viscoelasticity of alumina suspensions measured by quartz crystal microbalance with dissipation monitoring and in situ dynamic rheology. Langmuir 2008 Sep;24(18):99899996. 99. Balz M, Therese HA, Li JX, Gutmann JS, Kappl M, Nasdala L, et al. Crystallization of vaterite nanowires by the cooperative interacti on of tailor made nucleation surfaces and polyelectrolytes. Adv Funct Mater 2005 Apr;15(4):683688.
143 100. Weiss M, Welsch W, Von Schickfus M, Hunklinger S. Viscoelastic behavior of antibody films on a shear horizontal acoustic surface wave sensor. Anal Che m 1998 Jul;70(14):28812887. 101. Lange K, Bender F, Voigt A, Gao H, Rapp M. A surface acoustic wave biosensor concept with low flow cell volumes for label free detection. Anal Chem 2003 Oct;75(20):55615566. 102. Lange K, Grimm S, Rapp M. Chemical modif ication of parylene C coatings for SAW biosensors. Sens Actuator B Chem 2007 Aug;125(2):441446. 103. Baselt DR, Lee GU, Natesan M, Metzger SW, Sheehan PE, Colton RJ. A biosensor based on magnetoresistance technology. Biosens Bioelectron 1998 Oct;13(78): 731739. 104. Richard P. Buck, William EH, Mirtha Umana, Bownden EF. Biosensor Technology: Fundamentals and Applications. Marcel Dekker, Inc, New York 1990. 105. Griffiths D, Hall G. Biosensors what real progress is being made. Trends Biotechnol 1993 A pr;11(4):122130. 106. Tang C, Feller L, Rossbach P, Keller B, Voros J, Tosatti S, et al. Adsorption and electrically stimulated desorption of the triblock copolymer poly(propylene sulfide bl ethylene glycol) (PPS PEG) from indium tin oxide (ITO) surfaces Surf Sci 2006 Apr;600(7):15101517. 107. Yeh PYJ, Kizhakkedathu JN, Madden JD, Chiao M. Electric field and Vibrationassisted nanomolecule desorption and anti biofouling for biosensor applications. Colloid Surf B Biointerfaces 2007 Sep;59(1):6773. 108. Carfi A, Pares S, Duee E, Galleni M, Duez C, Frere JM, et al. The 3 D structure of a zinc metallo betalactamase from Bacillus cereus reveals a new type of protein fold. Embo J 1995 Oct;14(20):49144921. 109. Alexander DWCaRS. Carboxylatehistidine zinc interactions in protein structure and Function. Journal of American Chemical society 1988;111(16):64126419. 110. Wills BA. Developments and research needs in mineral processing. Proc 1st Int in Proc Symp, Izmir, Turkey 1986:1. 111. Wills BA. Mineral pr ocessing technology fourth edition : an introduction to the practical aspects of ore treatment and minera. Pergamon press, Oxford/ New York/ Beijing/ Frankfurt/ Sao Paulo/ Sydeney/ Tokyo/ Toronto 1988. 112. Paterson OD. How many low grade ores are now bei ng recovered by gravity concentration. World mining 1977;30:44.
144 113. Taggart AF. Handbook of mineral dessing Wiley, New York 1945. 114. Leja J. Chemistry of froth flotation Plenum Press, New York 1982. 115. King RP. The principles of fotation. S Afr I M M 1982. 116. Ives KJ. The scientific basis of flotation. Martinus Nijhoff Publishers, The Hague 1984. 117. Schulze HJ. Physico chemical elememtary processes in flotation. Elservier Science Publishing Co, Amsterdam 1984. 118. Ranny MW. Flotaion agents a nd processes technology and application. Noyes Data Corp, New Jersey 1980. 119. Soto H, Iwasaki, I. Flotation of apatite from calcareous ores with primary amines. Minerals and Metallurgical Processing 1985;2:160. 120. Somasundaran P. Interfacial chemis try of particulate flotation. AIChESymposium Series No 150 1975;71:1. 121. Somasundaran P. An overview of the ultrafine problem, in mineral processing at a crossroads problems and prospects Martinus Nijhoff Publishers, Dordrecht 1986:1. 122. Moudgil BM, Chanchani, R. Selective flotation of dolomite from francolite using two stage conditioning. Miner Metall Process 1985:1925. 123. Lehr JR, Hsieh, S.S. US patent 4 287 053 1981. 124. Anazia I, Hanna, J. New flotation approach for carbonate phosphate separ ation. Miner Metall Process 1987;4:196202. 125. Davis BE, Liewellyn, T.O., Smith, C.W. Continuous benefication of dolomitic phophate rocks. USBM, RI 8903 1984. 126. Snow RW. US patent 4 144 969 1979. 127. El Midany A, El Shall H, Stana R. Mechanisms in volved in reactive flotation of dolomite. Miner Metall Process 2009 May;26(2):94100. 128. Sarikaya MaA, Ilhan A. Biomimetics: design and processing of materials. AIP Press, Woodbury, NY 1995:12. 129. Smith RW, Misra M, Chen SZ. Adsorption of hydrophobic bacterium onto hematite implications in the frinth flotation of the mineral. J Indust Microbiol 1993 Feb;11(2):6367.
145 130. Vanloosdrecht MCM, Lyklema J, Norde W, Schraa G, Zehnder AJB. The role of baterial cell wall hydrophobicity in adhesion. Appl Environ Microbiol 1987 Aug;53(8):18931897. 131. Lawver JE, Wiegel RL, Snow RE, Hwang CL. Phosphate reserves enhancement by beneficiation. Mining Congress Journal 1982;68(12):2731. 132. Beveridge TJ, Forsberg CW, Doyle RJ. Major sites of metal binding in B acillus Licheniformis walls. J Bacteriol 1982;150(3):1438 1448. 133. Jenneman GE, McInerney MJ, Knapp RM, Clark JB, Feero JM, Revus DE, et al. A halotolerant, biosurfactant prodcing bacillus species potentially useful for enhanced oil recovery. Developmen ts in Industrial Microbiology 1983;24:485492. 134. Javaheri M, Jenneman GE, McInerney MJ, Knapp RM. Anaerobic production of biosurfactant by bacillus ligheniformis JF 2. Appl Environ Microbiol 1985;50(3):698700. 135. Beveridge TJ, Murray RGE. Uptake and retension of metals by cell walls of bacillus subtilis. J Bacteriol 1976;127(3):15021518. 136. Misra M, Mehta, R.K., Zheng, X., Smith R. W. Bacteria as flotation reagents for the flotation of a dolomitic phophate rock. Florida Institute of Phosphate Re search 1997. 137. Smith RW, Miettinen M. Microorganisms in flotation and flocculation: Future technology or laboratory curiosity? Miner Eng 2006 May Jul;19(68):548553.
146 BIOGRAPHICAL SKETCH ChihWei Liao was born in Yilan, Taiwan in 1977. His family is a standard nuclear family: father, mother, younger sisters, and he. His parents strongly emphasized the importance of education even if they just had the degree of elementary school. Thus, his parents always en c ourage him to purs ue his ow n dream through a good education. H e p u rsued his master s degree in the D epartment of A pplied C hemistry, National C hiao Tung University, Taiwan, from 2000 to 2002, his research mainly fo cus the synthesis of polyfluorene with blue emitting light and construct the devices for polymer light emitting diodes (PLED). After getting his master s degree, he went to military for one and half a year for finishing his obligation as a Taiwan citizen. Then, he served as research assistant in the center for condensed mater science, National Taiwan University one year. In 2005, he decided to go to the United States for pursu ing his P h.D. In the United States his major was materials science and engin e e ring at University of Florida. His speci a l ty was to utilize the molecular recognition of peptides based on phage display techniques in the application of electroactivated reversible biosensors and the specific separation of mineral particles under Dr. Laurie Gower group. He bel i eves that people have the right to live with the high quality of health. He expects himself to apply his speci a l ties to develop the technologies for improving peoples life quality as his life goal.