1 MEMBRANE BINDING AND LIPID EXTRACTION STUDIES OF THE GM2 ACTIVATOR PROTEIN By STACEY ANN BENJAMIN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2012
2 2012 Stacey Ann Benjamin
3 To Toto and my sister Camille Benjamin
4 ACKNOWLEDGMENTS I thank my advisor Dr. Gail E. Fanucci for all her assistance and guidance throughout my ent ire graduate career. She has expertly taught me more th an I could ever imagine learn ing during my time in her research group She has facilitated my growth as a scientist through motivation, challenge and support. I am most grateful that she allowed me to pursue my own goals and offered unselfish advice regarding my career interests whenever I needed it. I am pa rticularly appreciative for her playing a huge role in me overcoming my fear of, and becoming a lover of dogs. I am grateful to Dr. Daniel R. Talham for adopting me in his lab for the latter part of my graduate career. He always made me feel like a part of his group, and never hesitated to offer suggestions when I asked. The Talham lab became a second home for me and I will always be grateful to him and h is students for the accommodation and making me feel welcomed. I would like to thank past and current Fanucci and Talham group members for numerous scientific conversations and for providing a desirable work environment in which to work. I would esp ecially like to thank Dr. Yong Ran, Dr. Jamie Kear, Dr. Mandy Blackburn, Dr. Jordan Mathias, Dr. Jeff Carter, Dr. Roxane Fabre, Hao Liu, and Captain Emily Pollard for their unwavering support both in and outside the lab. The Chemistry department on the who le has provided support during my time here. I am grateful for the close friendships I developed while in Gainesville Dr. Marilyn Prieto, Whitney Stutts, Robert Menger, Pascale Attal lah, Dr. Danniebelle Haase, Matt Baker Dr. Richard Farley, Dr. Jared Ly nch, Joey Lott, and Ms. Glennis Brown have all enriched my graduate experience and they hold a special place in my heart. I
5 thank Stacy Ann Stephenson, Michelle Jurcak, and Dr. Latanya Fisher for being amicable roommates who have grown to be my good friend s. I am most thankful for my teaching mentor Dr. Jeff Keaffaber for inspiri ng me to become the best instructor I can be, and for encouraging me to constantly seek ways in which to perfect my craft. I am grateful to Dr. Phil Brucat for allowing me to teach my own undergraduate class on more than one occasion an invaluable experience that I will always treasure. I thank Dr. Ben Smith for always being available and for engaging me in thought provoking conversations I am most grateful to Ms. Lori Clark for sharing my joys, and constantly going above and beyond the duties of her responsibilities to assist me with any problems I may face. I have developed a great sense of respect and love for her, and will treasure our friendship well beyond my years at the U nive rsity of Florida Last and by no means least, w ithout the love and support of my family and close friends, graduate school would ha ve been difficult. I am appreciative of my parents, Sonia and Georgie Benjamin my aunt Sharon, and my sister Camille, fo r always being there for me when I need them the most. It is because of them that I am able to complete my degree. I thank my vacation buddies T amar, Lamonde, and Tracey Ann for always keeping me entertained and indulging with me in my second love fine d ining I would also like to acknowledge Wendy Batchelor and Dominic Watson for being very special pers ons in my life, and for encouraging me to be the best version of me. Finally, I will always remember Toto Benjamin for being my best friend and for teachi ng me what it means to love unconditionally.
6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 9 LIST OF FIGURES ................................ ................................ ................................ ........ 10 LIST OF ABBREVIATIONS ................................ ................................ ........................... 14 ABSTRACT ................................ ................................ ................................ ................... 16 CHAPTER 1 INTRODUCTION AND BIOLOGICAL RELEVANCE ................................ .............. 18 Gangliosides ................................ ................................ ................................ ........... 18 Function and Catabolism of Gangliosides ................................ ........................ 19 GM2 Gangliosidoses ................................ ................................ ........................ 21 Tay Sachs disease ................................ ................................ .................... 21 Sandhoff disease ................................ ................................ ....................... 22 AB Variant ................................ ................................ ................................ .. 23 Hexosaminidase A ................................ ................................ ............................... 23 Sphingolipid Activator Proteins ................................ ................................ ............... 24 Saposins A D ................................ ................................ ................................ ... 25 GM2 Activator Protein ................................ ................................ ...................... 26 Biosynthesis and posttranslatio nal modifications ................................ ....... 27 GM2AP structure ................................ ................................ ....................... 27 GM2AP function ................................ ................................ ......................... 28 Glycer ophospholipids ................................ ................................ .............................. 29 Research Overview ................................ ................................ ................................ 32 2 THEORY OF TECHNIQUES ................................ ................................ .................. 42 Circular Dichroism Spectroscopy ................................ ................................ ............ 42 Polarized Light ................................ ................................ ................................ .. 43 Principle of CD ................................ ................................ ................................ 43 Fluorescence Spectroscopy ................................ ................................ .................... 45 Principle of Fluorescence Spectroscopy ................................ ........................... 46 Fluorescence Components and Configuration ................................ ................. 48 Biological Fluorescent Probes ................................ ................................ .......... 48 Quenching of Protein Fluorescence ................................ ................................ 49 Sensitivity of Fluorescence Spectroscopy ................................ ........................ 51 Surface Plasmon Resonance Enhanced Ellipsometry ................................ ............ 52 Ellipsometry ................................ ................................ ................................ ............ 53 Generation of Elliptically Polarized Light ................................ .......................... 54
7 Reflection of Light off Surfaces ................................ ................................ ......... 54 Ellipsometric Parameters and Definitions ................................ ......................... 57 Ellipsometric Components and Nulling Ellipsometry Configuration .................. 58 Surface Plas mon Resonance Spectroscopy ................................ ........................... 61 Total internal reflection ................................ ................................ ............... 61 Kretschmann configuration of SPR ................................ ............................ 62 Components of SPR measurement and sensitivity of the technique .......... 63 SPREE Measurements ................................ ................................ ........................... 64 3 CHARACT ERIZATION OF GM2 ACTIVATOR PROTEIN CONSTRUCTS USING INTRINSIC TRYPTOPHAN FLUORESCENCE ................................ .......... 78 Introduction ................................ ................................ ................................ ............. 78 Materials and Methods ................................ ................................ ............................ 80 Site Directed Mutagenesis of GM2AP Tryptophan to Alanine Constructs ........ 81 Expression and Purification of GM2AP TRP to ALA Constru cts ....................... 82 Circular Dichroism (CD) Spectroscopy Measurements ................................ .... 85 Intrinsic Fluorescence Quenching Measurements ................................ ............ 86 Results and Discussion ................................ ................................ ........................... 87 Site Directed Mutagenesis ................................ ................................ ............... 87 Protein Expression and Purif ication ................................ ................................ .. 88 Intrinsic Fluorescence Quenching ................................ ................................ .... 90 Conclusions ................................ ................................ ................................ ............ 92 4 VESICLE BINDING AND LIPID EXTRACTION STUDIES OF GM2 ACTIVATOR PROTEIN VARIANTS ................................ ................................ ........................... 112 Introduction ................................ ................................ ................................ ........... 112 Materials and Methods ................................ ................................ .......................... 115 GM2 Activator Protein Expression and Purification ................................ ........ 115 Lipid Preparation ................................ ................................ ............................ 116 Fluorescence Spectroscopic Measurements ................................ .................. 116 Dansyl DHPE Extraction Assay ................................ ................................ ...... 117 GM2 Extraction ................................ ................................ ............................... 117 Results and Discussion ................................ ................................ ......................... 118 Conclusions ................................ ................................ ................................ .......... 124 5 SURFACE PLASMON RESONANCE ENHANCED ELLIPS OMETRY STUDIES TO STUDY LIPID BILAYER INTERACTIONS BY GM2 ACTIVATOR PROTEIN 131 Introduction ................................ ................................ ................................ ........... 131 Materials and Methods ................................ ................................ .......................... 134 GM2 Activator Protein Expression and Purification ................................ ........ 134 Gold Slide Preparation ................................ ................................ ................... 135 Surface Modification with Zirconium Phosphonate ................................ ......... 135 Lipid Preparation ................................ ................................ ............................ 136 Formation of Lipid Supported Layers ................................ ............................. 136
8 Surface Plasmon Resonance Enhanced Ellipsometry (SPREE) Measurements ................................ ................................ ............................ 137 Results and Discussion ................................ ................................ ......................... 138 Interaction of GM2AP with Phospholipid/ODM Hybrid Bilayers ...................... 138 Interaction of GM2AP with Zirconium Octadecylphosphonate modified SLBs 140 Conclusions ................................ ................................ ................................ .......... 141 6 CONCLUSIONS AND FURTHER DIRECTIONS ................................ .................. 149 LIST OF REFERENCES ................................ ................................ ............................. 151 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 160
9 LIST OF TABLES Table page 1 1 Selected naturally occurring fatty acid chains ................................ .................... 40 1 2 Common glycerophospholipids with their net charge at acidic lysosomal pH .... 40 3 1 Ste rn Volmer quenching cons tants and fraction of total tryptophan accessible to acrylamide and potassium iodide at acidic and neutral pH ........................... 111 4 1 Extinction coefficients of GM2AP wildtype and its W to A variants .................. 126 4 2 Half lives and extraction efficiencies of GM2AP variants and their ratios with respect to WT protein ................................ ................................ ....................... 130
10 LIST OF FIGURES Figure page 1 1 Structure o f GD1a ganglioside ................................ ................................ ............ 34 1 2 Degradation pathway of multisialogangliosides to form GM1 ............................. 35 1 3 Lysosomal catabolism of gangliosid es and other glycosphingolipids ................. 36 1 4 GM2 Activator protein assisted hydrolysis reaction that converts GM2 to GM3 37 1 5 Ribbon diagrams showing three different structural conformations of GM2AP (PDB ID 1G13) within one un it cell ................................ ................................ ..... 38 1 6 Chemical str ucture of sn glycerol 3 phosphate and the gen eral anatomy of glycerophosphol ipids ................................ ................................ .......................... 39 1 7 Biological membrane models ................................ ................................ .............. 41 2 1 Sc hematic representation of an electromagnetic wave showing the electric field as a function of position at constant time ................................ .................... 67 2 2 Peptide bond region of protein backbone showing electronic energ y transitions associated with the absorption of amide chromophores .................... 68 2 3 Far UV circular dichroism spectra showing the various types of se condary structure ................................ ................................ ................................ ............. 68 2 4 Jablonski diagram showing the energy level transitions involved in absorption and fluorescence emission ................................ ................................ ................. 69 2 5 Representative absorption and fluorescence emission spectra .......................... 69 2 6 Block diagram illustrating the general schematic of a spectrofluorometer .......... 70 2 7 Chemical structures o f the three intrinsic fluorescent amino acids ..................... 71 2 8 Chemical structure of the extrinsic fluorescent probe, dansyl amine .................. 71 2 9 Modified Jablonski diagram showing the energy level transitions involved in collisional quenching ................................ ................................ .......................... 72 2 10 Reflection of a polarized li ght bean from a surface ................................ ............. 73 2 11 Interaction of light with a material at a single interface wi th complex index of refraction ................................ ................................ ................................ ............ 74
11 2 12 Interaction of light with a material showing refl ections and transmissions through two interfaces ................................ ................................ ........................ 74 2 13 Nulling ellipsometry conf iguration ................................ ................................ ....... 75 2 14 Kretschmann configuration of SPR showing a prism metallic coati ng substrate layer interface ................................ ................................ ..................... 76 2 15 Illustration of the experimental set up used in surface plasmon resonance spectroscopy ................................ ................................ ................................ ...... 76 2 16 Schematic of a SPREE experimental setup ................................ ....................... 77 2 17 Typical SPREE sensorgram of protein adsorbing to a lipid bilayer that is functionalized on a me tallic thin film ................................ ................................ ... 77 3 1 pET16b vector map ................................ ................................ ............................ 94 3 2 E. coli codon optimized DNA and amino acid sequence of GM2AP wild type protei n ................................ ................................ ................................ ................ 95 3 3 Sample agarose gel picture of GM2AP variants after DpnI digestion ................. 96 3 4 Sample agarose DNA gel of pET16b GM2AP varia nts after plasmid purifica tion ................................ ................................ ................................ .......... 96 3 5 E. coli codon optimized DNA and amino acid seq uences of GM2AP W5A ........ 97 3 6 E. coli codon o ptimized DNA and amino acid sequences of GM2AP W5AW63A ................................ ................................ ................................ .......... 98 3 7 E. coli codon optimized DNA and amino acid sequences of GM2AP W5AW131A ................................ ................................ ................................ ........ 99 3 8 E. coli codon optimized DNA and amino acid sequences o f GM2AP W63AW131A ................................ ................................ ................................ .... 100 3 9 E. coli codon optimized DNA and amino acid sequences of G M2AP W5AW63AW131A ................................ ................................ ............................ 101 3 10 Sample column chromatographs during GM2AP purification ........................... 102 3 11 18% SDS PAGE gel of 15 L samples of purified GM2AP protein and variants after size excl usion chromatography ................................ ................... 103 3 12 Circular dichroism spectra of GM2AP wild type (black) and 0.5 mg/mL samples of GM2AP vari ants ................................ ................................ ............. 104
12 3 13 Fluorescence emission spectra of 1 M GM2AP W5A showing results from the titration of increasing amounts of ac rylamide ................................ .............. 105 3 14 Fluorescence emission spectra of 1 M GM2 AP W5AW63A showing results from the titration of increasing amounts of acrylamide ................................ ..... 106 3 15 Fluorescence emission spectra of 1 M GM2AP W5AW131A showing results from the titration of increasin g amounts of acryla mide ................................ ..... 107 3 16 Fluorescence emission spectra of 1 M GM2AP W63AW131A showing results from the titration of increasing amounts of acryla mide .......................... 108 3 17 Stern Volmer plots of GM2AP variants ................................ ............................. 109 3 18 Modified Stern Volmer plots of GM2AP variants ................................ ............. 110 4 1 Ribbon structure of GM2AP showing the modeled binding modes o f GM2 and PG ................................ ................................ ................................ ............. 125 4 2 Ribbon structure of GM2AP (PDB ID IG13) showing the location of the three native tryptop han r esidues ................................ ................................ ............... 126 4 3 Fluoresce nce emission spectra of 4:1 POPC:dansyl DHPE and 1:1 wild type GM2AP:dansyl DHPE complexes ................................ ................................ .... 127 4 4 Ch anges in relative transfer as dansyl is being sequestered from 1 mM 4:1POPC:dansyl DHPE vesicles at 484 n m ................................ ..................... 128 4 5 Proposed model of the membr ane bound orientation of GM2AP ..................... 129 4 6 Elution profiles showing the extraction efficiency of GM2 extraction by a series of W to A GM2AP variants ................................ ................................ ..... 130 5 1 A schematic illustrati on of a hybrid lipid bilayer ................................ ................ 143 5 2 A schematic representation of a zirconium octadecylphosphonate modified surface for the formation of supported lipid bilayers ................................ ......... 143 5 3 SPREE experime nt al set up showing the adsorption of GM2AP on the octadecyl mercap tan/phospholipid hibrid bilayer ................................ .............. 144 5 4 SPREE experime nt al set up showing the adsorption of GM2AP on the zirconium octadecylphosphonate supported phospholipid b ilayer .................... 144 5 5 SPREE sensorgrams s howing POPC extraction by GM2AP ............................ 145 5 6 SPREE sensorgra ms showing the binding o f GM2AP to octadecylmercaptan phospholipid hybrid bilayers ................................ ................................ ............. 146
13 5 7 SPREE sensorgra ms showing the binding of GM2AP to zirconium octadecy lphosphonate modified SLB s as a function of POPG concentration ... 147 5 8 SPREE sensorg rams showing the binding of varying concentrations of GM2AP to zirconium octadecylphosphonate modified SLB s ............................ 148
14 LIST OF ABBREVIATION S A Alanine Asp Aspartic acid BMP Bis(monoacylglycero)P hosphate Cer Ceramide D Aspartic acid Dansyl DHPE N (5 dimethylaminonaphthalene 1 sulfonyl) 1, 2 dihexadecanoyl sn glycero 3 phosphoethanolamine DMSO dimethylsulfoxide dNTP deoxynucleotide triphosphate E Glutamic acid EPR Electron paramagnetic resonance Fu c Fucose Gal gGalactose GalNAc N acetylgalactosamine Glc Glucose GlcNAc N acetylglucosamine Glu Glutamic acid GM2AP GM2 Activator Protein GSLs Glycosphingolipids L Left handed LB Langmuir Blodgett IPTG Isopropyl D thiogalacatopyranoside NeuNAc N acetylne uraminic acid Hex A beta Hexosaminidase A n Refractive index
15 ODM Octadecyl mercaptan ODPA Octadecylphosphonic acid PA Phosphatidic acid PAF Platelet activating factor PC Phosphatidylcholine PCR Polymerase chain reaction PE Phosphatidylethanolamine PG Phosp hatidylglycerol PI Phosphatidylinositol PMT Photomultiplier tube PS Phosphatidylserine R Right handed SAM Self assembled monolayer SEC Size exclusion chromatography sn stereospecific numbering SPR Surface plasmon resonance SPREE Surface plasmon enhanced el lipsometry TIRE Total internal reflection ellipsometry W Tryptophan
16 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosop hy MEMBRANE BINDING AND LIPID EXTRACTION STUDIES OF THE GM2 ACTIVATOR PROTEIN By Stacey Ann Benjamin August 2012 Chair: Gail E. Fanucci Major: Chemistry GM2AP is an accessory protein that solubilizes the GM2 ganglioside from intralysomal vesicles in ne uronal cells for hydrolytic cleavage by HexA to form GM3. This non enzymatic protein functions also a lipid transfer protein. The precise molecular interactions and method of extraction of GM2 and other lipids from vesicles are unknown. GM2AP contains thre e native tryptophan residues (W5, W63 and W131), with two of these (W63 and W131) located in the putative membrane binding loops of the protein. In this report, we use fluorescence spectroscopy and surface plasmon resonance enhanced ellipsometry (SPREE) to investigate the interaction of GM2AP with lipids as a function of protein electrostatics and hydrophobicity. Utilizing fluorescence spectroscopy, d ansyl labeled phospho lipids were used to monitor the changes in the rates of lipid extraction and transfer b y GM2AP from liposomes as a function of both pH and a s eries of tryptophan to alanine substituted constructs of the protein The ability of GM2AP to bind and/or extract dansyl labeled lipids from liposomes was affected with increased pH of the lipid enviro nment with optimal lipid extraction efficiency occurring at pH 4.8. A mino acid substitutions from tryptophan to alanine in the putative membrane binding loops of the protein resulted in slower lipid extraction rates, suggesting the
1 7 relevance of these resid ues for membrane binding by GM2AP. Additionally, a resorcinol based GM2 extraction assay provided results suggesting that though extraction rates slowed, total ganglioside extraction efficiency was not affected by the W to A substitutions. SPREE analysis allowed us to study the interaction of GM2AP with lipids on surface supported lipid bilayers. When PG, a negatively charged lipid was added to the lipid bilayer, GM2AP was able to adsorb onto the surface possibly due to electrostatic interaction s between t he protein and the immobile lipid bilayer on the functionalized surface. Protein adsorption was not affected by increased concentrations of PG to the lipid bilayer but adsorption increased with increasing protein concentration. This provided us with a syst em with which we could study GM2AP membrane binding in an attempt to gain a better understanding of the molecular protein int eractions involved in the GM2AP lipid membrane bindi ng process.
18 CHAPTER 1 INTRODUCTION AND BIO LOGICAL RELEVANCE Gangliosides Gangl iosides are a group of glycosphingolipids(GSLs) found primarily i n the outer leaflet of eukaryotic neuronal cell plasma membranes. 1 Gangliosides are comprised of a sialic acid containing oligosaccharide chain and a ceramide (sphingosine linked to a fatty acid) moiety. 2 Sy nthesis occurs in the endoplasmic reticulum and g olgi apparatus where ceramide is first made, followed by the stepwise addition of sugar and sialyl groups to an oligosaccharide chain by specific glycosyltransferases. 2,3 The glycosyltranferases involved in synthesis determine the particular sequence, linkage positions, and configurations of gangliosides. Six monosaccharides, namely, fucose (Fuc), glucose (Glc), galactose (Gal), N acetyl glucosamine (GlcNAc), N acetyl galactosamine (GalNAc), and sialic acid N acetylneuraminic acid (NeuNAc), are the only carbohydrates that appear in the oligosaccharide component of gangliosides characterized from vertebrate cells and tissue to date. 2 All ganglioside names begin with the letter G for ganglioside, followed by one of the letters M (mono), D (di), T (tri) or Q (quarto) indicating the number of sialic acids that are in its structure. The name ends with a number x, where 5 x depicts how many non placed after the number in higher gangliosides to indicate that there are one or two sialic acid residues linked to the galactose residue nearest to the ceramide moiety of the lipid respectively. For example, Figure 1 1 shows a structure of GD1a, a ganglioside t hat consists of two sialic acid residues linked to four other monosaccharides where one
19 of the sialic acid residues in connected to the galactose residue closest to the ceramide moiety. Function and Catabolism of Gangliosides Gangliosides are ubiquitous co nstituents of tissues and cells, and their oligosaccharide groups undergo alterations with cell differentiation, cellular development, and ontogenesis. 2 Often times, the sialic acid residue(s) in gangliosides serve as recognition sites and allow these GSLs to act as receptors for bacteria, viruses, bacterial toxi ns, and adhesive proteins. 2 They also function as antigens mediators of cell adhesion and modulators of signal transduction, 4 as well as suppo rt the functions of membrane bound receptors and enzymes. 5 Additionally, gangliosides are essential in homeostasis of biological functions, they participate in transmembrane cell signaling events, cell cell interactions and prevent ing inappropriate degradation by forming a protective layer on biologi cal membranes. 3,6 Gangliosides and other components of membranes are degraded by endosomal/lysosomal membrane digestion. This process is essential for cellular membrane stability 7 Once degraded, the products are either recycled and re used in salvage processes, or are further degraded 8,9 First, components of plasma membranes are endocytosed by coated pits, and are trafficked in the f orm of vesicles through the endosomes where lipids are sorted and sent to the Golgi apparatus, lysosome, or back to the plasma membrane 10 During endocytosis the lipid composition of internal endosome and then to the l ysosome. The lysosome exhibits a decrease in membrane stabilizing cholesterol and other sterols, and an increase in the anionic lipid, bis(monoacylglycero)phosphate ( BMP ), and an acidic pH of 4.5 7,10
20 Once in the ac idic lysosomal cell compartment, gangliosides and other glycosphingolipids are catabolized by a stepwise cleavage of sialic acid and monosaccharide groups from the nonreducing end of the oligosaccharide chain by water soluble exohydrolases. For multi sialog angliosides, degradation begins with the sequential removal of sialic acid residues by neuraminidase until GM1 is formed (Figure 1 2) 11 In vivo, membrane bound monosialogangliosides with four or fewer sugar head groups are not sufficiently accessible to the exohydrolases so sphingolipid activator proteins (SAPs) are required for cleavage 3,7,9,10 GM1 is degraded to GM2 by galactosidase in the p resence of SAP B of GM2 Activat or Protein (GM2AP) 12 GM2 is further degraded by hexosaminidase A (Hex A) in the presence of G M2AP to form GM3 13 Sialidase in the presence of Sap B degrades GM3 to form l actosylceramide 14 which is further degraded to g lucosylceramide by GalCer galactosidase or GM1 galactosidase in the presence of Sap B and Sap C 15 Glucosylceramide catabolises to form ceramide by glucosylceramid glucosidase in the presence of Sap C and then finally sphingosine is form ed by acid ceramidase in the presence of Sap D 16 Figure 1 3 shows a flow chart of the catabolism of the above mentioned gangliosides as well as other GSLs. In the absence of SAPs or detergents, exohydrolases are not able to degrade membrane bound gangliosides with four or fewer sugar head groups because they are not far enough into the aqueous space and away from the lipid core 17 Defects in gangl ioside degradation due to genetic mutations of exohydrolases or SAPs can lead to a buildup of gangliosides in the lysosome resulting in apoptosis and a wide range of
21 lysosomal storage diseases called gangliosidoses. This dissertation will further discuss t he GM2 gangliosidoses. GM2 Gangliosidoses The degradation of GM2 to GM3 relies on the exohydrolase, Hex A. This enzyme is a heterodimer with and products are required for the degradation of GM2. 3 Since gangliosides are found primarily in neuronal cells, gangliosidoses are diseases of the nervous system. GM2 gangliosido s es are inherited, genetic disorders with incidences of about 1 in every 310,000 bir ths 18 When there is a genetic mutation in the Hex A subunit, the catabolism of GM2 ganglioside is inhibited, resulting in Tay Sachs disease. Similarly, defects in the subunit of Hex A and GM2AP causes Sandhoff disease and AB Variant respectively, due to a buildup of GM2 in the cell. 3 Tay Sachs disease Tay Sachs disease is the most prevalent of the GM2 gangli osidoses and is inherited as a Mendelian autosomal recessive trait 19 Infantile Tay Sachs disease is caused by the absence or catalytic defect of Hex A due to mutations of the gene encoding the subunit of the enzyme. If the mutations allow Hex A some degree of residual activity, milder forms of the disease, namely Juvenile Tay Sachs and Adut Tay Sachs disease, arise with a later onset. The gene coding for this subunit is on chromosome 15 and mo re than 75 mutations have been reported 2 0 The disease is more prevalent among persons of Ashkenazi Jewish and French Canadian descent s 21 Tay Sachs disease was first reported in 1881 by the British ophthalmologist Warren Tay 22 and fi fteen years later by Bernard Sachs, an American neurologist 23 The classical infantile form of the disease appears usually six months after birth with
22 progressive psychomotor retardation, regression and loss of mental skills 19 By age one, patients experience seizures, blindness and the in ability to crawl or stand. Afte r age two, persons with Tay Sachs disease develop spastic quadriplegia and reach a decerebrate stage, which leads to death by age four 19,21 Sandhoff disease Genetic defects in the subunit of Hex A give rise to Sandhoff disease, a similar and almost indistinguishable neurological condition from Tay Sachs. Konrad Sandhoff and his colleagues first differentiated this disease from Tay Sachs when they discovered a massive accumulation of glycosphingolipids and other glycoprotein fragments with a terminal hexosamine residue both in the central nervous system and systematic organs 24 The increase in glycosphingolipid accumulation is due to defects also in hexosaminidase B, a homodimer with subunits of hexosaminidase 25 The gene coding for the been reported 3 Unlike Tay Sachs disease mutations in the subun it is more pre valent in the Creole/Spanish community of Cordoba, Argentina, the Maronite community of Cyprus, and the Metis Indians of Saskatchewan 21 Though clinically and neurologically similar to Tay Sachs, Sandhoff disease also causes nonneurologic e vents including the simultaneous enlargement of the liver and spleen, occasional foamy histocytes in the bone marrow, and the occurrence of N acetylglucosamine containing oligosaccharides in urine 3 The most common neuropathological findings are related to delayed myelination or demyelination and the degree of GM2 accumulation is more severe than in Tay Sachs disease 21 Persons with infantile Sandhoff disease usually die by age three though late onsets variants of the
23 disease have been reported where the symptoms are delayed for two to ten years (juvenile) or eve n into late adult life 3 AB Variant Persons who exhibit an accumulation of GM2 in neuronal cells, despite showing normal activities of Hex A are said to have AB variant 26 This rarely diagnosed variant is due to a deficiency of the GM2 Activator protein, the accessory protein necessary for the hydrolysis of GM2 by Hex A 3 The gene coding for GM2AP is located on chromosome 5 and only five mutations in the gene have been reported 26 AB variant presents in similar fashion to classical infantile Tay Sachs, where in fants lose motor skills, develop seizures and, vision and hearing loss, paralysis and ultimately death during early childhood 26 no cases of lat e onset variants have been described. Hexosaminidase A hexosaminidase A (Hex A) is one of three isoenzymes of lysosomal hexosaminidase B (Hex B), hexosaminidase S (Hex S) hexosaminidases hydrolyze the glycosidic bond of N acetylglucosamine (GlcNAc) and N acetylgalactosamine (GalNAc) residues from glycoproteins, oligosaccharides, and glycosphingolipids 3 Each isoenzyme is composed of two noncovalently linked subunits, and which differ in their substrate specificity 27,28 Hex A is a heterodimer, while Hex B and Hex S are and homodimers respectively. Both the and subunits of the hexosamindases contain active sites but dimerization is required for catalytic cleavage 3 The active site on the subunit of hexosaminidases cleaves neutral, water soluble oligosaccharide chains with terminal GlcNAc and GalNAc terminal residues, but the less
24 active site on the subunit can also cleave the sugar groups from negatively charged substrates 29 Because of this, only Hex A is able to cataboli ze GM2 in the lysosome. GM2 consists of four sugar head groups, hence steric hindrance prevents Hex A from direc tly hy drolyzing the membrane bound GM2 ganglioside. Therefore, GM2AP, a sphingolipid activator protein, is required to solubilize GM2 from the membrane and present the ganglioside in the proper orientation for cleavage. The accepted mechanism is that GM2AP binds to the ceramide moiety of GM2 forming a 1:1 complex. The protein:lipid complex then interacts with the subunit of Hex A and forms a ternary complex where the terminal GalNAc group is cleaved, forming GM3 3,30 Sph ingolipid Activator Proteins Sphingolipid Activator Proteins (SAPs) are a group of small, heat stable glycoproteins that are essential for the degradation of gangliosides and other glycosphingolipids with short (four or fewer) oligosaccharide headgroups 13,31 When glycosphingolipids with a short carbohydrate chain reach the lysosome for catabolism after endocytosis, the terminal sugar residue is not situated far enough from the vesicular membrane for cleavage by exoh ydrolases at the water lipid interface. SAPs facilitate this process by binding to the lysosomal membrane vesicles, solubilizing the glycosphingolipids and making the substrate accessible to its specific enzyme for degradation 10,13 There are five known SAPs to date and they are encoded by two genes. One gene codes for a precursor protein, prosaposin, which is proteolytically processed to form four highly homologous proteins called Saposin A, B, C and D 32 34 The second gene codes the other SAPs 25,35 37
25 Saposins A D Saposin s A, B, C and D are water s oluble, lipid membrane binding, and transfer proteins. They are derived from the proteolytic processing of the 70 kDa precursor glycoprotein, prosaposin in the late end o somes and lysosomes of cells 7,32 First, the N terminal peptide preceding the saposin A domain is cleaved, followed by the release of saposin A, which results in the formation of a saposin B D trimer. The trimer is then cleaved via sap B/C and C/D dimers, forming mature proteins of about 80 amino aci ds each, weighing 8 11 kDa 10,13 The x ray crystallographic structures of the human recombinant forms for all four saposins have been determined 38 40 Saposins A D are structurally homologous showing a conserved N glycosylation site and six highly conserved cysteine residues that have the same pairings to form three disulfide bonds 41 They contain an helical bundle of five helices where several hydrophobic residues serve as the internal structure. The structure is stabilized by the disulfide bridges which are thought to be necessary for protein function, and also responsible for the high level of s tability against heat, acid, and proteolytic enzymes 13,42,43 The saposins differ in specificity despite their high degree of homology. In vivo saposin A is required for the degradation of galactoceramide by galac toceramide galactosidase 9 and a genetic defect in the protein results in juvenile and sometimes a late onset of Krabbe disease 10 In vitro saposin A is shown to bind to GM1 and GM2, and to stimulate the hydrolysis of glucosyl and galactocylceramide 44 Saposin B was the first saposin to be discovered and it seems to be the least specific of the enymes 25,31 In vivo, it is required for the catabolism of a number of GSLs including sulfatide, globotriaosylceramide, digalactosylceramide, sphingomyelin and GM1 (Figure
26 1 3). Due to this, saposin B is said to behave like a general, physiological detergent 9,25 Saposin C is specific for the degradation of glucosylceramide by glucosylceramide gucosidase and a deficiency of this protein leads to a juven ile form of Gaucher disease 45,46 Saposin D participates in the degradation of ceramide by acid ceramidase both in cultured cells 47 and in vitro 44 It has also been reported that saposin D binds to, and solubilized vesicles that contain negatively charged lipids 48 GM2 Activator Protein GM2AP, the fifth SAP, is essential for the degradation of the ganglioside GM2 by GM2 from intralysomal vesicle membranes. It is believed that GM2AP recognizes the hydropho bic ceramide moiety 49 the sialic acid moiety, and the N acetyl galactosamine moiety of GM2 50 The protein lipid complex leaves vesicles and is recognized by HexA where GM2 is hydrolyzed in solution, and GM3 is released. 1,51,52 Speci fically, the terminal GalNAc sugar residue is cleaved from the GM2 tetrasaccharide (Figure 1 4). 1,53 It has been determined from in vitro sedimentation experiments that less than 15% of the activator protein is membrane associated and BMP, which is found in the lysosome in increased concentrations, enhances the extraction efficiency of GM2AP 54 GM2AP is structurally different from saposins A D. It is larger, carries one n glycosylation, contains eight cysteine residues and adopts a cup topology instead of helical secondary structure seen in the saposins 13 A detailed description of GM2AP will be given below.
27 Biosynthesis and posttranslational m odifications GM2AP is synthesized on the endoplasmic reticulum as a 193 amino acid prepro polypeptide. The pre sequence is a 23 amino acid signal peptide that directs both the protein synthesi s and its extrusion into the lumen; it is believed that the pre sequence is cleaved from the newly formed protein cotranslationally 25 As was previously mentioned GM2AP contains eight cysteines. The oxidizing environment of the ER helps the formation of four disulfide bridges. The GM2AP sequence contains a N glycosylation site at Asn63 Val Thr with the initiating Met being considered residue 1 (another style start s numbering the final protein at residue 32 due to the preprosequence). The preassembled oligosaccharide is made of many mannose groups where one or more is phosphorylated in the endoplasmic reticulum and g olgi after glycosylation so that the protein can be de livered to the lysosome via the mannose 6 phosphate receptor 3,55 Once in the lysosome, the 8 amino acid prosequence is removed during processing of the N terminus, leaving a mature, 162 amino acid protein weig hing between 20 and 27kDa depending on its oligosaccharide composition 25 The deglycosylated form of GM2AP weighs approximately 18 kDa 7 GM2AP s tructure The crystal structure of the nonglycosylated form of GM2AP purified in Escherichia coli has been solved and shows that there are three distinct monomers of the protein in th e 11 monomer unit cell (Figure 1 5). 53 The properly folded protein forms a single globular domain with dimensions 45 x 28 x 25 53 The secondary structure of the protein comprises of an eight cup fold forming a hollow hydrophobic pocket which includes approximately half of the total amino acid residues in the protein 53 The dimensions of the cavity are 12 x 14 x 22 and together with the hydrophobic
28 residues that line the cavity, it is suitable for binding the acy l chains of lipids like the 53 The hydrophobic pocket is accessible from one end of the protein only 56 and is lined by surface loops and a 2.5 turn alpha helix at its rim, which is thought to be the interaction site with HexA to allow GM2 degradation 53 The four independent disulfide bonds are located at surface regions of the protein, where they connect flexible loops at the rim of the hydrophobic cavity and are thought to be the igh stability 53 Loop regions in GM2AP were identified after structural refinement analysis of the crystals at sites spanning residues 58 78, 87 97, and 120 133 53 The residues in these regions exhibited high B factors and were found in different conformations for different monomers within a unit cell, with the loop region of residues 120 133 being the most flexible 53 Figure 1 5 shows crystal structures of the three monomers found in the 11 monomer cell. Monomers A and B shows the position of the loop region of residues 120 hydropho bic cavity. The entrance to the hydropho bic cavity of GM2AP differs by 3 with the two loop conformations, suggesting the region may facilitate lipid extraction 57,58 GM2AP f unction In vivo GM2AP participates in the lysosomal catabolism of GM2 to GM3 by HexA. Without GM2AP, HexA is not able to hydrolyze the terminal GalNAc group of GM2 in liposomes 49 Like the other SAPs, one known function of GM2AP is to act as a l e complexes with glycolipids 59 Because of this, GM2AP is kn own to function as a general glycosphingolipid transfer protein, w here it is known to extract other GSLs like GM1 and
29 GA1 from liposomes or micelles and transfer them as soluble 1:1 complexes between membranes 25,56 58,60,61 Utilizing a fluorescence dequenching assay, the binding affinity and specificity of GM2AP were determined 50 The results indicated that the binding affinity to selected gangliosides is GM2 >> GT1b >> GM1 Crystal structure analyses of ligands bound in the hydrophobic pocket of GM2AP shows that the protein accommodates ligands in different binding modes one for gangliosides, and another for glycerophospholipids 57 It has been shown that GM2AP binds GM2 in such a way that the 18 carbon acyl chains are within the hydrophobic pocket while the tetrasaccharide head group of the ligand sticks out of the protein, into the aqueous environment 57 While the fatty acid tails of phosphatidylglycerol (PG) binds in a similar position to the acyl chains of GM2, the head gr oup of the head group on the other hand is buried in the hydrophobic pocket 57 Other physiological roles have been credited to GM2AP for example, it has been thoug ht to act as a factor that stimulates and enhances the association between phospholipase D and enzyme activators 62,63 GM2AP is also thought to participate in the regulation of pr oton pumps in intercalated kidney ce lls 64,65 Additionally, GM2AP was shown to bind platelet activating factor (PAF), inhibiting its action 66 A GM2AP:lysoPAF complex was observed via X ray crystallography, and introd uced the possibility that GM2AP may display some hydrolase activity towards PAF 60 Glycerophospholipids Glycerophospholipids, also known as phosphoglycerides, are the most common group of structural lipids that constitute eukaryotic cel l membranes. They are a group of glycerol containing lipids with a phosphate headgroup, that are derived from sn 3 glycero phosphate(Figure 1 6A) 11 Glycerophospohlipid anatomy and nomenclature
30 utilizes the stereospecific numbering (sn) system where two fatty acid chains, R1 and R2, are attached via an ester linkage to the first (sn 1 position) and se cond (sn 2 position) carbons of the glycerol backbone 67,68 The R groups consist of saturated or cis unsaturated acyl tails of varying length (Table 1 1), and make up the hydrophobic portion of the lipid 11,69 The polar po rtion of the lipid consists of the glycerol backbone and a highly polar or charged functional group, Y (Table 1 2), attached via a phosphodiester linkage to the third (sn 3 position) carbon of glycerol 67,68 Figure 1 6B depicts the general anatomy of glycerophosp holipids. The specific functional group along with the nature and chain length of fatty acids determine physiochemical properties and structural and functional roles of glycerophospholipids in cell membranes 11,70 Selected l ipid s ystems as b iological m embrane m odels : Lipid systems have been used in a number of scientific research fields to mimic biological membranes. One of the first and still the most common lipid system used as biological membrane models is the li pid vesicle, also called the liposome (Figure 1 7A). Liposomes are spherical in shape and are typically composed of amphiphilic molecules such as phospholipids. They provide a closed, stable and regular bilayer membrane and have been utilized to study prot ein and DNA interactions with lipids 71 the permeability of ions and drugs 72,73 and other molecular biological process es. Liposomes may exist as small unilamellar vesicles with diameters less than 50 nm, as large unilamellar vesicles with diameters between 100 and 1000 nm, or as multilamellar vesicles consisting concentric, multiple vesicular bilayers.
31 Under physiological conditions, membrane lipids can exist as two lipid monolayers forming a two dimensional sheet, called a bilayer (Figure 1 7B) 70 Bilayer formation occurs more readily with glycerophospholipids and sphingolipids, where the cross sectional areas of the head groups and acyl tails are similar 67 Because the hydrophobic regions at the edges of bilayers are transiently in contact with water, the sheets are unstable and tend to spontaneously fold back on itself to form liposomes 70 Lipid bilayers on solid supports solve this problem. Tamm and McConnell developed the first solid supported lipid bilayer system when they deposited lipid membranes separated by a thin aqueous layer on quartz, glass, and oxidi zed silicon 74 These bilayers on solid supports resemble cell membranes bec ause they retain and represent fluidity and lateral mobility ; 75 and provide a system with which to investigate molecular biological processes like protein lipid interactions. Substrates used to support phospholipid bilayers with high lipid mobility and little or no defects, need to be hydrophilic, smooth and clean 75 Fused silica, borosilicate glass, mica, oxidized silicon, as we ll as thin films like titanium IV oxide, indium tin oxide, silver, gold, and platinum have been utilized as supports for lipid bilayers 74,75 Lipid layers tethered to metal surface offer a means to study biological membrane processes via electrochemical or optical means 76 It has also been reported that inorganic supports like zirconium phosphonate may also be used to support phospholipid bilayers 77 Langmuir Blodgett (LB) technique and Langmuir Schaefer procedures 74 along with lipid adsorption and fusion 78 are the usual methods used for the formation of supported phospholipid bilayers on substrate supports. The Langmuir deposition methods involve controlled dipping or pulling of a support through an organic amphipatic monolayer,
32 while lipid adsorption and vesicle fusion involve exposing the hydrophilic support to liposomes. Hybrid bilayers are another approach to mimicking c ell mem branes 76,79,80 This process involves the use of metal supported alkanethiol self assembled monolayer (SAM), and monolayer of a phospholipid (Figure 1 7C). The term hybrid is used because the bilayer consists of nat ural and synthetic parts. The SAM layer is formed by incubating a clean gold substrate with an alkanethiol solution in ethanol and allowing the alkanethiol to self assemble on the gold surface, rendering the surface hydrophobic. The resulting covalent inte raction with the gol d surface is not chemically aff ected by chan ges in pH, ionic strength, lipid composition, or type of buffer 80 The phospholipid mo nolayer is then added to the SAM either by vesicle fusion 79 or lipi d transfer from an air water interface 81 Research Overview This project aims to investigate membrane binding and extractio n of lipids by GM2 AP in order to determine the precise molecular interactions involved in these processes. Studies on the interaction of GM2AP with its specific ligand G M2 have been performed using several techniques such as sucrose density ultracentrifuga tion, sucrose density isoelectric focusing, polyacrylamide gel electrophoresis, circular dichroism, and steady state fluorescence spectroscopy 51,52 as a lipid transfer protein has also be en investigated using gel filtration chromatography, thin layer chromatography (TLC), fluorescence resonance energy transfer (FRET), fluorescence dequenching assays, surface plasmon resonance and a dansyl based fluorescence assay 27,51,52,82 84
33 Despite the fact that the crystal structure and function of GM2AP are known, the orientation of the protein when interacting with lipid bilayers and the specific mechanism of interacti on with vesicular membranes, followed by extraction and transfer of lipids/ligands are still being determined. Due to the intrinsic tryptophan fluorescence of GM2AP, and the fact that two of the three tryptophan residues are located in regions thought to be involved in binding to the vesicular membrane, fluorescence spectroscopy is deemed a useful technique to resolve these questions. Additionally, gel filtration chromatography has also been proven to be convenient for studying the interactions of GM2AP with GSLs 51 First, a series of tryptophan to alanine mutations were constructed via site directed mutagen esis to determine the fraction of accessible intrinsic fluorophore to charged and neutral quencher in solution. G el filtration and dansyl based fluorescence assays were empl o yed to determine th e changes in lipid binding and/or extraction as a function of pH and hydrophobicity as a result of substituting A for W in the putative membrane bind ing loops of the protein The membrane perturbing properties and function of SAPs including GM2AP have bee n shown to be dependent on, or greatly increased in the presence of acidic lipids like BMP, PI, PS and PG in lipid membranes 27,42,43,85 Surface plasmon resonance enhanced ellips ometry (SPREE) was used to further st udy interactions between GM2AP and select phospholipids (PC and PG ) on a solid supported membrane model system. Additionally, a newly developed zirconium octadecyl phosphonate surface based system 77 was investigated to determine its suitability for studying glycerophospholipid me mbrane binding by GM2AP.
34 Figure 1 1. Structure of GD1a ganglioside. Gangliosides consists of a ceramide tail an d oligosaccharide head group, of which two of the sugar residues are sialic acid s
35 Figure 1 2. Degradation pathway of multisialogangliosid es to for m GM1
36 Figure 1 3. Lysosomal catabolism of gangliosides and other glycosphingolipids showing the necessary exohydrolases and sphingolipid activator proteins required for degradation
37 Figure 1 4. GM2 Activator protein assisted hydrolysis reaction that converts GM2 to GM3. The terminal N acetylgalactose (GalNAc) monosaccharide is cleaved by hexosaminidase A (Hex A)
38 Figure 1 5 Ribbon diagrams showing three different structural conformations of GM2AP (PDB ID 1G13) within one unit c ell. The mobile loops of the protein are highlighted in gold Chain A Chain B Chain C
39 Figure 1 6. Chemical structure of A) sn glycerol 3 phosphate B) the gen eral anatomy of glycerophosphol ipids
40 Table 1 1. Selected naturally occurring fatty acid chains Chain Chemical For mula Systematic (Common) Name 14:0 CH 3 (CH 2 ) 12 COOH n Tetradecanoic (Myristic) acid 16:0 CH 3 (CH 2 ) 14 COOH n Hexadecanoic (Palmitic) acid 18:0 CH 3 (CH 2 ) 16 COOH n Octadecanoic (Stearic) acid 24:0 CH 3 (CH 2 ) 22 COOH n Tetracosanoic (Lignoceric) acid 16:1 ( 9 ) CH 3 (CH 2 ) 5 CH=CH(CH 2 ) 7 COOH cis 9 Hexadecanoic (Palmitoleic) acid 18:1 ( 9 ) CH 3 (CH 2 ) 7 CH=CH(CH 2 ) 7 COOH cis 9 Octadecanoic (Oleic) acid 18:2 ( 9,12 ) CH 3 (CH 2 ) 4 CH=CHCH 2 CH= cis cis 9,12 Octadecanoic (Linoleic) CH(CH 2 ) 7 COOH acid Table 1 2 Common glycerop hospholip ids with their net charge at acidic lysosomal pH Functional Group (Y) Chemical Formula Net Charge at pH 4.5 Hydroxyl OH 1 Choline CH 2 CH 2 + N(CH 3 ) 3 0 Ethanolamine CH 2 CH 2 + NH 3 0 Serine CH 2 CH(COO ) + NH 3 1 Glycerol CH 2 CH(OH) CH 2 OH 1 Inositol C 6 H 6 (OH) 6 1
41 Figure 1 7. Biological membrane models. A) Lipid vesicles ( liposomes ). B) Lipid bilayer. C) Hybrid lipid bilayer
42 CHAPTER 2 THEORY OF TECHNIQUES Circular Dichroism Spectroscopy Circular dichroism (CD) spectroscopy is a powerful and technique used to monitor and study the secondary structure of proteins in solution. Using far UV absorption, CD spectra is extremely sensitive to the analysis of alpha h elix, beta sheet, turn, and 86 Advance s in molecular biology have allowed proteins to be produced in a number of host systems in their native form, as site directed mutants, or being engineered synthetically Other techniques like X ray cryst allography and nuclear magnetic resonance spectroscopy are also capable of giving higher resolution structural information on prote ins. However, CD is non destructive, is able to study protein structure under a variety of experimental conditions, and good s pectra can be obtained on less than 0.1 mg of samples in 30 minutes or less making it a useful technique for monitoring the structure of these proteins. 87,88 Similar to ultraviolet visible (UV Vis) spectroscopy, CD spectroscopy is based on the absorption of light as a function of wavelength. 86 Absorption wavelength depends on the type of chromophore and/ or arrangement of atoms in the sample. Proteins for example, absorb strongly in the UV region of the electromagnetic spectrum due to peptide bonds, amino acids w ith aromatic side chains (tyrosine, phenylalanine, and tryptophan), disulfide bonds, and any prosthetic groups. 89 For our purposes, we focus on contributions to CD spectra in the far UV spectral region (190 250 nm) where secondary structural information can be obtained due to the electronic absorption of peptide bonds in the protein backbone. 87
43 Polarized Light Light is an electromagnetic wave comprising an electric, E, and a magnetic, B, field vect or which are mutually perpendicular and also perpendicular to the direction of propagation of the wave. 90 We specifically consider the strength and direction of E because the specification of E completely determines B and it has a stronger interaction with matter than B. At any point in an electric field, light of a single wavelength can be resolved into three oscillations along an x, y, z coordinate system. Figure 2 1 illustrates a light wave as a plane wave travelling alo ng the z axis. The electric field is orthogonal to the z axis and the oscillations have the same frequency, but usually different amplitudes and phases. Polarization is defined as the behavior with time of a vector field at a fixed point in space. Unpolar ized light emits light that has components with electric fields oriented in all positions perpendicular to the direction of travel. However, if all of the photons in a light beam have the electric field oriented in the same direction, the light is said to be polarized. If the phase of the x and y oscillations are the same, the polarization is linear. If the phases differ by +/ 90 the polarization is circular. Light is elliptically polarized in all other cases where the phases are different. In fact, line arly and circularly polarized light are specific cases of the more general elliptically polarized state of light. Principle of CD Absorbance, A of electronic transitions in species is measured according to Beer Lambert law A( ) = ( )lc (2 1) Where A has no units, is the molar extinction coefficient in liter mol 1 cm 1 at wavelength l is the pathlength of the cell in cm, and c is the molar concentration of the
44 sample in mol liter 1 Since is dependent on A or can be plotted vs. to produce an absorption spectrum. CD of a molecule is defined as the difference between the absorption of left and right circular ly polarized light by a sample. 88 Plane polarized light is made up for two circularly polarized components with equal amplitudes : left handed, L (rotating counter clockwise), and right handed, R (rotating clockwise). After passing through a chiral chromophore, L and R are absorbed to diff erent extents thus yielding unequal molar coefficients. The resulting radiation no longer traces a circle, but now possesses elliptical polarization which can be monitored Proteins are chiral molecules with different ellipticity values for L and R circul arly polarized light, L and R 87,88 The difference between these quantities, = L R as a function of is plotted to produce a CD spectrum. CD bands may be either positive or negative depending on which type of light is absorbe d more strongly. Ellipticity is historically the unit reported for the CD of samples, but others like mean residue ellipticity and delta epsilon have become more popular. 88 Ellipticites are typically in the range 10 mDeg. A CD signal is observed only if the chromophore in the sample molecule is intrinsically chiral, if it is covalently linked to a chiral center in the molecule, or if it is in an asymmetric environment. 88 As was previously stated, CD signal due to the amide chromophores along protein backbones is sensitive to the various types of secondary structure. 87 Electronic ab sorption in the far UV region are due to the peptide bond which begins wit h a weak but broad n at 210 220 nm, followed by more intense transition at 190 nm (Figure 2 2 ). CD signals that correspond to negative bands at 222 nm and 208 nm, and
45 a positive band at 198 nm are characteristic of helical proteins, whereas spectra for proteins with anti parallel sh eet structures exhibit negative and positive bands at 218 nm and 195 nm respectively. 91 Disordered proteins with random coil conformations have very low ellipticity above 210 nm and negative bands near 195 nm. 91 Figure 2 3 shows the characteristic CD spectra of different secondary structural elements found in proteins. In our work, we use CD spectroscopy analysis to compare the secondary structure of our site directed protein variants to the spectrum of published wild type, 92 to en sure that the protein is still properly folded. Fluorescence Spectroscopy Most biological processes with the exception of the storage of genetic information, involve proteins. 93 Proteins are extremely diverse in their secondary structure and function and there always is ongoing interest in the relationsh ip between the two. Specifically, researchers are interested in how proteins fold, how they recognize other molecules, and the mechanism involved when carrying out their particular function. While many analytical and biophysical techniques can be used to p robe protein structure and function, no single one can provide a complete picture. X ray crystallography and nuclear magnetic resonance provide detailed information about proteins that are essentially static, but fluorescence spectroscopy has the advantage of being able to study proteins in a more realistic dynamic state. Fluorescence spectroscopy is a sensitive and selective optical technique that has been extensively used for many years in all scientific fields. With regard to protein fluorescence, the t echnique has been used to gain insight into the polarity of different regions of a protein, the flexibility of proteins, and the conformational transitions proteins
46 undergo. 93 In this work fluorescence quenching experiments were performed to determine the nature of the environment of tryptophan residues i n GM2AP, and their accessibility to the quencher. Additionally, a dansy l based fluorescence assay was used to monitor the changes in the rates of lipid extraction by GM2AP from liposomes as a function of both pH and a s eries of tryptophan to alanine substi tuted constructs of the protein Principle of Fluorescence Spectroscopy Photoluminescence a process that occurs in fluorimetric analysis is the emission of light from an atom or molecule in an electronically excited state Depending on whether the analyt e is in a singlet or triplet excited state, fluorescence or phosphorescence occurs. In a singlet excited state, the electron in the excited orbital is paired to an electron in the ground state of opposite spin. This permits return of the excited state elec tron to the ground state rapidly (10 8 s 1 ) by emission of a photon. 94 E xcitation to the singlet state followed by radiative relaxation of an electron between orbitals of different energies is illustrated by the Jablonski diagram (Figure 2 4). First, a molecule absorbs an incident photon of sufficient energy to promote an electron to a higher electronic state ( for example S 2 ). Next, the molecule releases excess energy through vibrational relaxation and goes to the lowest energy of the excited state, S 1. The molecule then returns to the ground state (S 0 ) in a number of wa ys: 1. it fluoresces by emitting a photon, 2. It returns to the ground state nonradiatively through internal conversion, or 3. It transitions to the triplet state through intersystem crossing where it returns to the ground state via phosphore scence. 93 Phosphorescence is not discussed in this chapter.
47 Fluo rescence emission can be characterized by its lifetime, quantum yield, anisotropy, or simply by its emission spectrum. Fluorescence lifetime is defined as the average time a molecule spends in the excited state before returnin g to the ground state and is defined by: (2 2) nr are both rate constants. Fluorescence lifetimes are usually close to 10 ns. 94 Fluorescence quantum yield Q, is the ratio of the number of photons emitted to the number of photons absorbed and is given by: (2 3) Fluorescence anisotropy relates the orientation of the exci tation light polarization to the absorption and emission dipoles which provides information about the size and shape of proteins, or the rigidity of various molecular environments. An emission spectrum is a plot of fluorescence intensity vs. wavelength, a nd the spectrum is generally independent of excitation wavelength because vibrational rule). 94 Other nonradiative relaxation processes like solvent relaxation, complex formation, energy transfer, and excited state reactions lead to emitted photons of less energy than the energy associated with excitation. The photons subsequently emit at longer wavelength s (Figure 2 5) 95 Note too that fluorescence absorption and emission spectra are generally symmetric because the vibrational energy levels in the ground and exited states a re similar, and transitions being involved in both absorption and emission are also the same. 94
48 Fluorescence Components and Configuration All spectrofluorom eters contain a light source, excitation and emission monochromator s a sample holder, and a detector. Figure 2 6 shows a block diagram with the essential componen ts of a spectrofluorometer. The light source used must produce a constant photon output a t all wavelengths Common light sources include mercury lamps, xenon lamps, halogen lamps, light emitting diode sources, and laser diodes. The instrument used for this work is equipped with a xenon arc lamp mounted vertically in an attempt to increase stability and the useful life of the source. For excitation and emission wavelength selection, monochromators must be able to pass all photons of all wavelengths with the same efficiency and their efficiency must not be related to polarization 94 Diffracting grating monochromators are suitable for this purpose. Photomultiplier tubes (PMTs) are used as the detector in commercial fluorescence instruments. PMTs are best described as current sources that detect individual photons that can be counted individually or as an average signal. Ideal photomultiplier tubes should be able to detect photons at all wavelengths with equal efficiency. Biological Fluorescent Probes Fluorescent chromophores used in biological studies can be intrinsic, extr insic, or coenzymic. In proteins, intrinsic fluorophores are the amino acids tryptophan, tyrosine, and phenylalanine, with the indole group of tryptophan being the most dominant due to a higher quantum yield (0.13 in water), and a larger absorption cross s ection that the other fluorescent amino acids. 96 Tryptophan fluorescence is extremely dependent on polarity of its e nvironment and its emission blue shifts as polarity decreases, while its quantum yield increases. 94 In our work, tryptophan is excited at 295 nm and maximum
49 fluorescence emission is observed at 346 nm. There are some proteins that contain fluorescent conenzymes like nicotinamide adenine dinucleotide (NADH) and pyridoxa l phosphate. Reports on studies of these chromophores have provided information about the structure and interactions of a number of proteins. 93 Often times, analytes of interest do not possess fluorophores, and if they do, they are not located in the region of interest. To solve this problem, extrinsic fl uorophores are bound or covalently attached to regions of interests in proteins or other biomolecules. In our work, we use a dansyl chromophore (Figure 2 8) covalently attached to a phosphatidylethanolamine lipid headgroup, to monitor protein lipid extract ion by GM2AP. When selecting extrinsic fluorophores special care should be taken in terms of ensuring that the chromophore should not disturb the structure or function of the molecule of interest, and that the probe is sensitive to its environment so that definitive interpretations can be made. 97 Dansyl is excited at 340 nm, where proteins do not absorb; hence there is no inter ference by tryptophan emission. The emission spectrum of dansyl is also highly sensitive to polarity of its environment with emission maxima at 518 nm when the chromophore is solvent exposed. Quenching of Protein Fluorescence Solute fluorescence quenching has widely been used in biophysical studies as a source of information for proteins. Polypeptide chains fold, forming secondary structure of protein, which results in some amino acid residues aqueous exposed on the surface, and others buried and inaccessible to the polar solvent. One method commonl y used to st udy the structure of proteins in solution, is to identify amino acid residues that are aqueous exposed vs those that are buried via quenching of protein fluorescence. 1,98 Fluorescence quenching is defined as the decrease in fluorescence intensity of a
50 sample as a result of excited state molecular interactions, molecular rearrangements, energy transfer, ground state complex formation, or collisional quenchi ng. 94 Since most proteins usually contain tryptophan residues, fluorescence quenching is employed to probe for the exposure of this intrinsic fluorophore. Quenching experiments are very popular in pro tein studies because of the value of the information that can be obtained, and the relative ease with which the experiments can be performed. Various low molecular weight substances for example, o xygen, iodide ion, nitrate ion, amines, halogens, and cesium ion, act as quenchers of fluorescence. However acrylamide (neutral) and iodide (charged) have been successfully used as efficient quenchers of tryptophan fluorescence. 99 Iodide, being a charged and heavily hydrated quencher, can only quench aqueous exposed tryptophanyl residues on the surface of proteins, 100 while acrylamide, being an uncharged and polar quencher can quench any excited state tryptophanyl residue it collides with regardless of the nature of its environment. 101 Collisi onal or dyanamic quenching involves the transient collisional interacton between a ground state quencher and an excited state fluorophore; u pon contact, the fluorophore returns to the ground state without emitting a photon A modified Jablonski dia gram is shown in Figure 2 9 to illustrate this process Static quenching, on the other hand, involves the formation of a ground state, nonfluorescent complex between the fluorophore and quencher. 94 Th e quenching experiements described in this dissertation utilizes collisional quenching. In general, fluorescence quenching occurring between the excited state of the indole ring in tryptophan, M*, and a quencher is described by: (2 4)
51 where is the fluorophore quencher complex, k d is the diffusional rate constant and k i is the dissipation rate constant. The Stern Volmer equation, used to describe and interpret collisional quenching, is given by: (2 5) where F 0 and F are the fluorescence intensities in the absence and presence of quencher, 0 and are the fluorescence lifetimes in the absence and presence of quencher and K sv is the collisional quenching constant, which i s equal to k q 0 where k q is the bimolecular rate constant for the quenching process. Quenching data are usually plotted as F 0 /F vs [Q] and the slope of the line gives K sv the degree of quenching of fluorophore. The inverse of K sv is the quencher concent ration at which 50% of the fluorescence intensity is quenched. 102 Proteins can sometimes contain multiple tryptophan residues located in different environments, thus posses sing unequal accessibility to quencher. A modified Stern Volmer plot describes the fraction of total fluorescence accessible to quencher and is given by: (2 6) where f a is the fraction of the initial fluorescence accessible to quencher. Plots of F 0 / F vs 1/[Q] give a y incercept of f a 1 Sensitivity of Fluorescence Spectroscopy Flu orescence spectroscopy is inherently as much as 1000X more sensitive than absor ption spectroscopy 95 One reason for high sensitivity of the technique is that intrinsic fluorescence is not common to most chemical species so fluorescence is
52 generally measured against a low or zero background. 95 The overall sensitivity of fluorescence is dependent on the fluorophore as well as the instrument. The molar absorptivity and quantum yield of a f l and these factors cannot usually be controlled. Additionally, fluorescing power is dependent on protein concentration and the power of the source, unlike other absorb ance methods where concentration is related to the ratio of the source power before and after sample interaction. 103 In terms of the instrument, sensitivity can be expressed as the ratio of signal to noise at a particular set of conditions, or by determining the minimum detectable quantity of a fluorophore at a particular set of conditions. The components of the instrument that contribute to sensitivity include the intensity of the source, the efficiency of the optical system, the spectral ba ndpass of the monochromators, and the efficiency of the detector. 104 Surface Plasmon Resonance Enhanced Ellipsometry The study of bio logical molecules by way of biosensors is a k ey area of research in biology, chemistry and engineering. S everal te chniques have been employed to qualitatively and quantitatively detect biomolecular processes such as antibody antigen, lipid protein and protein protein interactions. A majority of tech niques used today involve labels like fluor ophores and radioactive isotopes to detect biomolecular interactions Though s ensitive, these techniques sometimes present c oncerns like the biomolecules not being able to react in t heir native states, the prepara tion time and expense associated with labeling, problems with photobleac h ing, and the precautions involved with handling radioactive material. 105
53 T here has been growing interest in employing label free optical methods to monitor int eractions because they are usually nondestructive and noninvasive techniques Affinity optical biosensors are commonly used where biomolecules are immobilized on surfaces which are irradiated with light and then the reflected light is analyzed. The basis o f analysis is the detection of changes in the refractive index of the sample(s) of interest ne ar the reflecting surface as a result of the particular inter actions. 105 Techniques like interferometry, ellipsometry, and surface plasmon r esonance (SPR) are all capable of analyzing very thin films in the pm and low nm range ( ranges of interest for the thickness of biomolecular layers ) 106,107 Surface plasmon enhanced ellipsometry (SPREE) is an opti cal technique that utilizes the combination of surface plasmon resonance (SPR) and ellipsometry to monitor and analyze adsorption and desorption of molecules on thin films. It is a lso referred to as total i nternal reflection ellipsometry ( TIRE ), which tech nique is based on the principles of ellipsometry under total internal reflection conditions. 108 With the use of appropriate thin metal films, the SPR effect gives high sensitivity to the real time monitoring of binding events, bond formations, and structural changes of biomolecules, among other things. 109 Here the theory behind both ellipsometry and SPR spectroscopy will be explained and the SPREE instrumentati on used to combine these two techniques will be described SPREE was used in this work to study phospholipid binding and e xtraction by GM2AP. Ellipsometry Ellipsometry is an old but very sensitive nondestructive, and widely used optical technique used to study surfaces and thin films. 110 The technique of ellipsometry is based on the change in the polarization state of elliptically polarized light reflected from
54 a surface. 107 When a surface is co vered with one or more la yers of thin film and substrate the optics of the entire system changes the polarization state. An ellipsometer measures ellipsometric angles and which can provide informati on about film thickness down to angstrom resolution and index of r efraction out to the thousandth position. Generation of Elliptically Polarized Light Elliptically polarized light can be generated when linearly polarized light is passed through an optical element called a compensator which consists of a fast axis and a slow axis which are perpendicular to each other and the direction of the propagation of the wave. When linearly polarized light passes through a compensator, the two components that were in p hase, will subsequently be out of phase, producing elliptically polarized light. Also, when linearly polarized light is reflected off a metal surface it causes u nequal phase shifts of both components that are perpendicular and parallel to the plane of inc idence. The change in polarization results in elliptically polarized light and the amount of ellipticity depends on the physical and optical properties of the surface, and substrates being measure d Reflection of Light off Surfaces Ellipsometry involves th e reflection of light off surfaces. Figure 2 10 illustrates a light beam being reflected from a film substrate. The incident beam, reflected beam, and the di rection normal to the surface are perpendicular to the surface. This is known as the plane of incid ence. The direction of x and y components are parallel and perpendicular to the plane of incidence, and are referred to as p and s components of the electric field respectively. When light is reflected by the surface, the sample may comprise various layers of different optical properties, resulting in a complex optical
55 system. The p and s components of the electric field will experience altered phase shifts, changing the state of polarization. The incident (E in ) and reflected (E out ) electric vectors are con nected by the sample reflection matrix R: (2 7 ) When an incident beam interacts with a material, some of the light is reflected and some is transmitted at the interface between medium 1 and medium 2 (Figure 2 11 ). The complex index of refraction, is the parameter used to describe the interaction of light with the material. is a combination of a real part (index of refraction, n) and an imaginary part as given by the equation: (2 8 ) w here k is the extinction coefficient and i is an imaginary number, the square root of 1. The theory for ellipsometry is contained in the Fresnel formalism. The Fresnel reflection coefficient, r, is the ratio of the amplitude of the reflected wave to that of the incident wave. r can be determined for the s and p components of the electric field as light travels from medium 1 to 2 by: (2 9 ) (2 10 ) where 1 and 2 are the complex indices of refraction f or mediums 1 and 2 respectively, and 1 2 law. 111 The reflectance ratios, and are defined as the ratios of the reflected
56 intensity of light to that of the incident intensity, and is the squar e of the magnitude of the Fresnel reflection coefficients: (2 11 ) (2 12 ) For light incident on one interf ace the light transmitted through the interface is neglected and the above is true In this work however, as well as many real world situations, more than one interface is present. Figure 2 11 shows ligh t interacting with three media The resultant reflected light from medium 1 comprises a combination of the light reflected from the first in terface and all the beams transmitted from medium 2 back into medium 1. The components of light will have different phases depending on the additional optical distances they travel, and for multiple interfaces, each successive transmission back to medium 1 is smaller. The ratio of the amplitude of the outgoing resultant beam to the amplitude of the incoming beam is now denoted as the total reflection coefficient, R R is analogous to r and is given by: 107 (2 13 ) (2 14 ) where r 12 and r 23 are the Fresnel reflection c oefficients for the interface between media 1 and 2, and 2 and 3 respectively. and are generally complex numbers and is the film phase thickness and: (2 15 )
57 Where d is the film thickness. When multiple interfaces are present, R parallel and perpendicular to the plane of incidence now becomes: (2 16 ) (2 17 ) Ellipsometric Parameters and Definitions The ellipso metric angles and Recall that incident p polarized light and s polarized light are generally not in the same phase. When each component makes a reflection, the resulting phase shift for the different waves may not be the same. The phase difference between the p and s 1 and the difference after reflection is 2 is the phase shift that occurs after reflection, determined as ( 2 18 ) Similar to a phase shift, the amplitudes of the p and s components may change after reflection. Given that the total reflection coefficient for p and s is defined as the ratio of the reflected beam amplitude to that of the incident beam amplitude, is related to change for the p and s wave components: (2 19 ) In ellipsometry the two reflection coefficients and describe and and are used to determine the complex reflectance ratio, of the total reflection coefficients
58 (2 20 ) The information about the system of interest is contained in and the parame ters and are measured directly by ellipsometers. Subsequently, th e primary equation of ellipsometry is or (2 21 ) The ellipsometric parameters are related to the optical properties of the sample, that is, film thickness (due to mass change on the surface) and complex refractive indices, and are calculated from computerized optical models. 107 Ellipsometric Components and Nulling Ellipsometry Configuration The components used in ellipsometry include a monochromatic light source, polarizer, compensator, analyzer and a detector. The most common light sources used are laser s, arc lamps, or polychromatic source s and filtering Typically, lasers are preferred because of their high output intensities as well collimated Gaussian beams. 90 In o ur work, a n unpolarized laser beam at 532 nm laser is used In cases where spectroscopic measurements are taken, arc lamps may be used because they have an output over a very broad wavelength range. Arc lamps how ever, are much less intense than lasers, res ulting in decreased signal to noise ratios. Polarizers are optical components that are used to convert any polarization state of a light beam to a single known and desired polarization state. The polarizer used in our work converts unpolarized light form o ur laser source, to linearly po larized light after transmission through the polarizer Polarized light is produced when the electric field component of the light beam travels parallel to the optical axis of the polarizer. Beams
59 perpendicular to the optica l axis are extinguished. In ellipsometry, polarizers are used in two different ways. When a polarizer is used to convert unpolarized light to linearly polarized light, it is known as a polarizer, with its angle position, P, located between the polarizing a xis and the plane of incidence. Polarizers are also used to determine the polarization state of the already polarized reflected beam, and are called analyzers. In nulling ellipsometry, an analyzer locates the null, and its angle position is denoted A. Ano ther optical component of ellipsometry is the compensator, which is also known as a retarder or quarter wave plate. Compensators alter the phase of one polarization component of a light beam with respect to another 110 Compensators have a fast axis and a slow axis, both of which are perpendicular to the direction of propagation and each other. The difference in speeds leads to a phase shift of 90 in the components of the electric field along these axes. Sin ce both components of linearly polarized light are in phase, the component of the wave aligned with the fast axis passes through the compensator faster than that of the slow axis, they will emerge out of phase, producing elliptically polarized light. The t hree most common types of optical detectors used in ellipsometric measurements are photomultiplier tubes, semiconductor diodes, and charge coupled device (CCD) cameras. Photomultiplier tubes are significantly sensitive to polarization sensitivity but they require high voltage power supplies and display nonlinearity. 90 Semi conductor diodes are inexpensive and linear over a broad range of intensities, but in our work, we utilize a CCD camera as our detection tool. There is o ne basic ellipsometer configuration (light source polarizer sample analyzer detector) for all ellipsometric measurements. However, variation in
60 configurations can arise in the polarizing and analyzing regions of the incident and reflected beams res pectively. In our work we employ the rotating analyzer nulling ellipsometry configuration. First monochromatic linearly polarized light is passed through a compensator to produce elliptically polarized light. E lliptically polarized light incident on the sa mple will result in the reflected light exhibiting a linear state of polarization. This is necessary because linearly polarized light is much easier to analyze after reflection. After reflection the linearly polarized beam is extinguished by rotating the analyzer to a 90 position with respect to the axis of linear polarization. This is called finding the null. The CCD camera then detects the minimum in the signal. The polarize r compensator sample anal yzer (PCSA) configuration is used because the rel ationship between the settings of the o p ti cal elements and and is simpler. 110 Figure 2 13 illustrates the set up for nulling ellipsometry w h ere we use a PCSA arrangement to determ ine the null. The polarizer compensator combination produces elliptically polari zed light incident on the sample, and the reflected linearly polarized light passes through the analyzer which rotates to find the null which is detected by the CCD camera. The angle combination s will change to determine the null depending on the change in the polarization state after r eflection from different systems on the surface. Therefore calculations for and made and based on optical models, information about biological interactions at the surface of our film may be acquired One advantage of nulling ellipsometry is that in measuring angles, problems of stability of the light source, and nonline arity of detectors is avoided. 108
61 Surface Plasmon Resonance Spectroscopy Surface plasmon resonance spectroscopy is an optical technique used to study interactions occurring at surface and interfaces. 112 SPR is based on the excitation of charge density oscillations called surface plasmons by light in thin layers of noble metals for example, gold, silver, and copper. Since the discovery of using SPR to characterize thin films and to monitor processes at metal interfaces in the late 1970s, the technique has been used in optical sensing devices for measu ring physical, chemical, and biological quantities 113 For biological systems, binding events of biomolecules can be studied using SPR which often requires the immobilization of the biomolecules on the surface of the meta l via surface f unctionalization The advantages of SPR are that it is extremely sensitive, and it is a label free technique SPR provides a powe rful tool to investigate the dynamics and structural changes that are occurring in real time during biological interactions by monitoring the refractive index (n) changes at the surface of thin film s that support surface plasmons due to total internal refl ection Total internal reflection Total internal reflection (TIR) occurs w hen light is reflected at an interface where the refractive index of the incident medium (n 0 ) is greater than the refractive index of the reflecting medium (n 1 ) 108 The incident light is usually passed through a dielectric medium like a prism (n = 1.72), and the reflecting medium is most like ly a water based liquid (n = 1.33). As the angle of incidence is increased, the transmitted beam approaches a minimum angle called the critical angle, where the transmitted light is parallel to the interface, resulting in no energ y crossing the interface When n 0 is greater than n 1 TIR occurs if the angle of incidence is larger than and is described as
62 (2 22) a nd the beam is propagated back into the incident medium. Al though the reflected beam does not transfer energy across the interface, an electrical field intensity called an evanescent field wave is leaked into the reflecting medium. The evanescent wave decreases exponentially with increasing distance from the surfa ce of the interface decaying over a distance of about one light wavelength. 114 If the TIR interface is coated with a conducting meta llic layer like gold with an appropriate thickness, the p polarized component of the evanescent field wave penetrates the layer and excites electromagnetic surface plasmon waves within the layer that is in contact with the reflecting medium. Since gold is non magnetic, the surface plasmon wave will also be p polarized, and due to the nature of its propagation, an enhanced evanescent wave will be created. Since the electric field penetrates a short distance into the reflecting medium, SPR conditions are sens itive to the changes in refractive index at the gold surface. Kretschmann configuration of SPR There are three main configurations of SPR: grating coupled systems, optical waveguide systems, and prism coupled attenuated total reflection systems. In our wor k, the latter is used in the Kretschmann arrangement (Figure 2 14). The Kretschmann arrangement is the most widely used geometry in SPR as it has been found to be very suitable for sensing the change in refractive index on thin films 106 In this configuration, a light wave passes through the high refractive index prism and is completely reflected at the interface between a prism and a thin m etal layer. The evanesce nt wave that leaks
63 into the reflecting medium excites surface plasmon wave s at the outer boundary of a metal. The high refractive index prism modifies the wave vector of the light by decreasing the phase velocity of the photons. A matching condition occur s when the wave vector component of the incident beam parallel to the conductor surface, is equal to the wave vector of the surface plasmons which are bound to the conductor surface. The wave vector of the incident beam can be tuned to be equal to the surf ace plasmon wave vector by varying either the angle of incidence or the wavelength of the beam. If the wavelength of the beam is fixed, the reflectance of p polarized light can be measured as a function of angle of incidence, and a sharp drop in the reflec ted light intensity will be observed and a phenomenon called surface plasmon resonance occurs. Components of SPR measurement and sensitivity of the technique The principal elements involved in SPR measurements are shown in Figure 2 15. First, a source whi ch is converted to p polarized light is used to excite the plasmon wave in the attenuated total reflectance configuration. Also, a prism is needed to couple light photons to the plasmons, and a thin metallic or semiconducting film is required on the surfac e of which plasmons can be excited. Finally, the SPR signal is detected by a CCD camera. SPR is dependent on different parameters of the reflecting system which affects the spectral resolution and sensitivity of measurement. The parameters are: the refract ive index ( n), the extinction coefficient (k), and thickness of the different layers (d) on the surface. For improved sensitivity, the thickness of the metal should be a fraction of the wavelength of the incident beam. Thus, special consideration should be taken into account when selecting the type of metal and its thickness. For our purposes, the
64 metal chosen was gold with a thickness of 28.5 nm coupled to an adhesive 2 nm layer of chromium on SF10 glass slides. The incident beam is at a wavelength of 532 nm. Additionally, an increase in temperature can also affect the sensitivity of the technique. 106 SPR is very sensitive to molecular adsorption to the metal inter face in the distan ce of the evanescent field wave, thus high sensitivity is achieved when monitoring the molecular adsorption of biomolec ules on the surface of the gold SPREE Measurements Under total internal reflection conditions, ellipsometric measureme nts and parameters obtain enhanced sensitivity and detection of interactions occurring at or near the surface of metallic (gold in this work) thin films. The increased se nsitivity associated with the optical response of the technique comes from the efficie ncy of the collective excitation of conduction electrons near the metal surface. The resolution of SPREE is on the order of 5 x 10 7 refractive index units compared to 10 5 and 2 x 10 6 to 10 5 for ellipsometry and SPR respectively. 115 In this work, SPREE was used to measure the adsorption and de sorption of lipid vesicles to form bilayers on functionalized gold surfaces. GM2AP was then flowed over the surface, and binding and/or extraction of lipid was observed by monitoring the change in the ellipsometric parameter, The c hange in resonance angle is proportional to the change in refractive index and thickness of the layers on the surface since the change in p amplitude. The minimum detectable change in angle of the instrument used in this work i s about 10 mD eg which corresponds to a change in thickness of 0.1 nm. 77 Figure 2 16 illustrates the experimental set up of SPREE. Measurements were taken using a commercial EP3 SW imaging system (Nanofilm Surface Analysis,
65 Germany). A SF10 glass slide coated with an adhesive chro mium layer then gold layer is assembled on top of a 70 L sample cell. A Peltier temperature control system (not shown) was linked to the sample cell to minimize baseline drift due to temperature fluctuations with time as data is being collected. A 60 SF1 0 prism is attached to the back side of the slide using diiodomethane as the refractive index matching fluid. A laser beam with wavelength of 532 nm produces elliptically polarized light incident on the sample by passing through a linear polarizer flowed b y a compensator. The angles of the polarizer and compensator are fixed and set in a way to ensure that the reflected beam is linearly polarized. Once reflected off the sample, the beam is passed through a 10x working distance objective and the analyzer, wh ich rotates until the null condition is obtained. This minimum light intensity is detected by the CCD camera and the angles of the different components are related to the optical properties of the sample. The angle is plotted a s a function of time in our experiments. Figure 2 17 shows a typical sensorgram of a protein being adsorbed to a lipid bilayer that was fused on a was used to fit the experimenta l data to a simpl e Langmuir model Measurements began with injecting buffer in a flow cell and allowing it to come in contact with the surface (a). Next, lipid vesicles were injected and through vesicular fusion, lipid bilayers were adsorbed on the surface, resulting in an increase in (b). After the adsorption process was allowed to equilibrate, buffer was flowed through the cell to remove any free lipids. At (c) protein was injected into the flow cell, which resulted in adsorption to the surface, followed by a buffer was h to monitor desorption (d). In each adsorption or
66 desorption step, sufficient tim e was given for equilibration of substrates, and to allow for stable signal to be acquired.
67 Figure 2 1 Schematic representation of an electromagnetic wave showing the e lectric field as a function of position at constant time
68 Figure 2 2 Peptide bond region of protein backbone showing electronic energy transitions associated with the absorption of amide chromophores Figure 2 3 Far UV circular dichroism spectra showing the various types of secondary structure. Spectra show helical (black), sheet (blue), and random coiled (green) structures of proteins. Figure was rep rinted from Besley et al. ( http://besley.chem.nottingham.ac.uk/research/research prospec.html ) Accessed in May 2012
69 Figure 2 4 Jablo nsk i diagram showing the energy lev el transitions involved in absorption and fluorescence emission The solid lines represent radiative transitions while broken lines represent nonradiative transitions in the singlet state. Figure 2 5 Representative absorption and fluorescence emission spectra
70 Figure 2 6 Block diagram illustrating the general schematic of a spectrofluorometer
71 Figure 2 7. Chemical structures of the three intrinsic fluorescent amino acids: tryptophan, tyrosine, and phenylalanine Figure 2 8. Chemical structu re of the extrinsic fluorescent probe, dansyl amine
72 Figure 2 9. Modified Jablonski diagram showing the energy level transitions involved in collisional quenching
73 Figure 2 10 Reflection of a polarized light bean from a surface. The plane of inci dence contains the incoming beam, the reflected beam, and the normal to the surface. E p and E s are the amplitudes of the electric field parallel and perpendicular to the plane of incidence
74 Figure 2 11 Interaction of light with a material at a single interface with complex index of refraction, 2 Figure 2 12 Interaction of light with a material showing reflections and transmissions through two interfaces
75 Figure 2 13 Nulling ellipsometry configuration. The fixed polarizer compensator com bination before the reflection gives an elliptical polarization of light incident on the sample. The reflected beam is linearly polarized and the analyzer is rotated to find the null
76 Figure 2 14. Kretschmann configuration of SPR showing a pri sm metallic coating substrate layer interface. is the angle of incidence Figure 2 15. Illus tration of the experimental set up used in surface plasmon resonance spectroscopy
77 Figure 2 16. Schematic of a SPREE experimental setup. Ellipsometric measurements are taken under SPR conditions. The pri sm and surface are mounted on a flow cell for measurements under solution conditions Figure 2 17. Typical SPREE sensorgram of protein adsorbing to a lipid bilayer that is functionalized on a metallic thin film First, a buff er baseline is acquired (a) followed by the adsorption of lipid vesicles for bilayer fusion (b). Protein is then adsorbed (c), then buffer is flowed to wash away unbound protein a b c d
78 CHAPTER 3 CHARACTERIZATION OF GM2 ACTIVATOR PROTEIN CONSTRUCTS USING INTRINS IC TRYPTOPHAN FLUORE SCENCE Introduction Glycosphingolipids are ubiquitous constituents of cell membranes in eukaryotic cells. 56 The catabolism of glycosphingolip ids is an essential cellular process that occurs in the lysosome of cells. GM2 Activator Protein (GM2AP) is one of five nonenzymatic sphingolipid activator protein s (SAPs) required for the hydrolytic degradation of glycosphingolipids that contain four or f ewer sugar head groups. 13 The other four SAPs, Saposins A D, are encoded by a single gene mapped to chromosome 10 and are de rived from the proteolytic processing of the precursor glycoprotein, prosaposin. 7,32 Saposins A D are structurally homologous: existing as homodimers, sharing three disulfide bonds with the same connectivity, and po ssessing secondary structures of a predominantly alpha helical nature. 33,41 Conversely, GM2AP is encoded by an unrelated gene on chromosome 5 and is synthesized as a 193 amino acid prepro polypeptide. The GM2AP seq uence consists of an Asn linked glycosylation site at Asn63 Val Thr which targets the protein to the lysosome from the endoplasmic reticulum 50 In it s mature and deglycosylated form, GM2AP is an 18 kDa protein consisting of 162 amino acids 25 GM2AP is structurally different from saposins A D: it exists a s a mo nomer, contains 4 disulfide bridges and adopts an eight stranded cup topology instead of the helical secondary structure seen in the saposins. 13,25 In vivo, GM2AP is required for the lysosom al catabolism of GM2 to GM3 by the hexosaminidase A (Hex A). extracting GM2 from intralysosomal vesicular membranes, 59 forming a protein lipid complex where the ceramide moiety of the lipid is buried in the hydrophobic pocket of
79 the protein, and the bulky carbohydrate head group protrudes from the binding pocket and into the aqueous environ ment 57 GM2 is then presented in the proper orientatio n to terminal sugar group GalNAc, for cleavage. 59 No other and its physiological significance is evident by the occurrence o f a fatal neurological disorder, AB variant of GM2 gangliosidoses, caused 116,117 Physiologically, GM2AP has also been found to be involv ed in CD1 mediated lipid antigen presentation to T cells 118 to act as a factor that stimulates and enhances the association between phospholipase D and enzyme act ivators, 62,63 and to participate in the regulation of proton pumps in intercalated kidney cells. 119,120 In vitro, GM2AP acts as a general lipid transfer protein, where it is known to bind and extract other glycolipids and phospholipids. 25,57,121 GM2AP has been isolated from natural sources and has been expressed in its glycosylated form from insect and yeast cells 122,123 as well as in its deglycosylated form from E. coli. c ells 36,124 The revealed cup topology from x ray crystallography of declycosylated protein provides a 12 x 14 x 22 hollow pock et lined with hydrophobic residues to carry out its function. 53 A 2.5 turn alpha helix at the rim of the hydrophobic cavity and two flexible loop regions were also identified in the resolved crystal structure of GM2AP. 53 precise molecular interactions of GM2AP are not yet well understood. GM2AP contains three tryptophan residues in its 162 amino acid sequence one residue (W5) is located on the backside of the binding pocket of the protein, and t wo residues (W63 and W131) are located in putative membrane binding loops of the
80 protein. Because tryptohan is an intrinsic fluorophore, fluorescence can be employed to study GM2AP. Fluorescence spectroscopy is a sensitive optical technique that has long been used to study proteins. Unlike other biophysical techniques, fluores cence spectroscopy has the advantage of studying proteins as dynamic systems as opposed to in their static form. Often times however, in order to use this technique extrinsic fluorescent probes are needed for protein analysis either because the native pro tein does not contain tryptophan residues, or if they do, the residues are not located in regions of interest in the protein. Here we report our findings from the characterization of the intrinsic fluorescence of the tryptophan residues in GM2AP. A series of tryptophan to alanine substi tutions were constructed via site directed mutagenesis of the recombinant protein. Fluorescence quenching experiments were employed to determine the nature of the environment of the tryptophan residues and their accessibilit y to the quencher. Tryptophan fluorescence of solvent exposed residues can be used to monitor GM2AP when investigating the regions of the protein involved in lipid binding and extraction Furthermore, because W63 and W131 are natively located in regions th ought to interact with the membrane surface the need to attach extrinsic fluorescent labels would be eliminated when studying GM2AP interactions with lipid membranes via fluorescence spectroscopy. Materials and Methods Primers were obtained from Integrate d DNA Technologies, Inc ( Coralville, IA). 1kb standard DNA ladder was purchased from New England Biolabs (Ipswich, MA). SDS PAGE Molecular Weight Standards broad range, and Biorad Criterion pre cast 18% SDS PAGE gel were purchased from BioRad (Hercules, CA ). The Quikchange TM site
81 directed mutagenesis kit was purchased from Stratagene (Santa Clara, CA). The QIAprep spin miniprep kit was purchased from Quiagen (Valencia, CA) Factor Xa Cleavage Capture Kit was purchased from Novagen (Gibbstown, NJ). The Ni NTA resin column, HiTrap Q HP Anion exchange column, HiPrep 26/10 desalting column, and Sephacryl S 200 High Resolution column were purchased from GE Healthcare (Pittsburg, PA). All other reagents and supplies were obtained from Fisher Scientific (Pittsburg, PA) and were used as received. Site D irected Mutagenesis of GM2AP Tryptophan to Alanine Constructs The gene encoding the sequence for wild type GM2AP, a gift from Christine Schubert Wright 53 was constructed in pET16(b+) vector. The map of the plasmid pET16(b+) vector, containing cDNA sequence and the amino acid sequence of GM2AP wild type a re given in Figure 3 1 and Figure 3 2 respectively. A series of tryptophan (W) to alanine (A) variants designated as W5A, W 5AW63A, W5AW131A, W63AW131A, and W5A/W63A W131A were engineered via several rounds of single point site directed mutagenesis. First, a single amino acid mutation of tryptophan at position 5 to alanine using pE T16b GM2AP as the template was performed and the variant was designated W5A. Similarly, W63A was obtained by the single amino acid mutation of tryptophan at position 63 to alanine. The W5A variant was used to engineer additional amino acid mutations at res idues 63 and 131 to produce W5AW63A, W5A W1 31A and W5A W63AW131A. T he W63A va riant was used to engineer W63A W131A. Primers for each mutation site were designed using the software, PrimerX ( http://www.bioi nformatics.org/primerx/ ) and were obtained from Integrated DNA Technologies Inc.
82 Mutagenesis reactions were performed using the QuikChange TM Site Directed Mutagenesis Kit and protocol (Stratagene, La Jolla, CA). Each buffered reaction contained 20 ng or 50 ng of template DNA, 125 ng each of forward and reverse primer, deoxynucleotide triphosphate ( dNTP) mix, 2.5 units ( U ) PfuTurbo DNA polymerase, and 3 5% (v/v) dimethylsulfoxide. The reaction mixtures were vortexed and centrifuged for a few seconds before subjection to polymerase chain reaction (PCR). PCRs were conducted in an Eppendorf (Hauppauge, NY) automated thermal cycler, for seventeen cycles Each cycle consisted of 30 seconds at 95C for denaturation of the double stranded DNA, one minute at 55C for annealing of the primers, and a six minute extension at 68C. After PCR, the reaction mixture was treated with 10 U Dpn I endonuclease for one hour at 37C in order to digest the original, unmutated DNA template on the basis of methylation. After dige stion, each reaction was checked for mutated DNA via a 1% polyacrylamide gel electrophoresis using 10 L of the product The resulting nicked vector DNA which contained the desired mutated sequence was then transformed into E. coli XL1 Blue supercompetent cells via the heat shock method. The plasmid containing the mutated DNA sequence was subsequently isolated and purified using a QIAprep Spin Miniprep Kit (Qiagen, Valencia, CA), an d submitted for DNA sequencing. Expression and Purification of GM2AP TRP to ALA Constructs Recombinant GM2AP expression and purification procedures were modified from the protocol reported by Wright et a l 53 and originally published by Wu et al. 124 Plasmids containing the W to A mutations were transformed in E. coli BL21(DE3) cells using the heat shock method. 5mL sterile pre culture of Luria Bertani (LB) media (tryptone, yeast extract and sodium ch loride) with 100 mg/mL ampicillin were each inoculated with a
83 single colony of cells and grown for approximately 5 hours (to OD 600 = 0.6) at 37C, while shaking at 250 rpm. The pre cultu re was then used to inoculate 1L LB media containing 100 mg/mL ampicil in and the cells were allowed to grow (OD 600 ~0.8) with shaking at 250 rpm at 37C for 3 hours. The temperature was reduced to 20C and expression was induced with 0.8 mM isopropyl D thiogalacatopyranoside (IPTG), while sha king at 250 rpm overnight for 8 hours. Following induction, c ells were pelleted by centrifugation at 5 000 rpm for 10 minutes at 4 C using a Sorvall RC6 centrifuge with SLA 3000 rotor. The cell pellet was resuspended in 25 mL lysis buffer (0.1M phosphate, 0.1M NaCl, 1% Triton X 100, 0.3M PMSF, 0.5mM Benzamidine,pH 8.0). The cells were tip sonicated for two minutes and then passed through a French pressure cell (Thermo Scientific, Waltham, MA) three times at 20,000 psi. The lysed cells wer e centrifuged at 12,000 rpm for 20 minutes at 4 C The pellet was resuspended in lysis buffer, after which the previously men tioned sonication, f rench pressure, and centrifugation steps were repeated to pellet inclusion bodies and other cell debris. Inclusion bodies in the pellet were denatured by resu spension in 5 M urea b uffer (5M urea, 0.1M phosphate, 10mM Tris HCl, 5% glycerol, TCEP, pH 8.0), homogenized using a Dounce Tissue glass homogenizer, tip sonicated for 5 minutes (5 second on and off time interva ls), and centrifuged at 12,000 rpm for 20 mi nutes at 4C The supernatant containing denatured GM2AP was passed through a prepacked Ni NTA resin column (GE Healthcare, Piscataway, NJ), pre equilibrated with urea buffer. The N terminal fused histidine tag on GM2AP allows for separation of the protei n by Ni chelating chemistry. The column was then washed with 10 column volumes (CV)
84 5 M pH urea buffer pH 8.0; 10CV 6M Guanidine HCl pH 8.0 followed by 10 CV 6M Guanidine HCl pH 7.0 (6M GuHCl, 0.1M phosphate, 10mM Tris HCl, 5% glycerol). GM2AP was eluted w ith 6M Guanidine HCl pH 4.5. GM2AP eluate fractions were pooled and the protein was refolded by drop wise dilution overnight in a 20 fold volume of slow stirring refolding buffer (50mM Tris base, 1mM EDTA, 0.05% T ween 20, 10% glycerol, 2mM Gluta thione redu ced, 0.2mM Glutathione oxidized pH10) at 4C. The solution containing the protein was subsequently buffer exchanged by way of dialysis (25mM Tris base, 0.05% Tween 20, 2.5% ; glycerol pH 8.0) for 2 days to get rid of guanidine hydrochoride, after which the protein was concentrated by anion exchange chromatography by loading the sample unto a 5 mL HiTrap TM Q Column (GE Healthcare, Piscataway, NJ) and eluting q column elution buffer ( 200mM NaCl, 25mM Tris base 2.5% glycerol, 0.05% Tween 20; pH 8.0 ) It has been reported that the histidine tag fused to GM2AP alters the intrinsic fluorescence spectra under acidic conditions 125 so the tag was removed by Factor Xa cleavage. The Q column eluate containing GM2AP fractions was buffer exchanged into Factor Xa cleavage buf fer (100mM NaCl, 50mM Tris HCl, 5mM CaCl 2 pH 8.0) using a HiPrep TM 26/10 desalting column that was pre equlibrated with Factor Xa cleavage buffer. His tag cleavage was performed using a Factor Xa Cleavage Capture Kit (Novagen, Gibbstown, NJ). Briefly, Fac tor Xa enzyme was mixed with GM2AP in cleavage buffer at a 2:1 ratio (units of enzyme: mg protein) and the mixture was incubated at 4C for 48 hours. The enzyme was subsequently quen ched and removed using Xarrest a garose at a ratio of 100 L Xarrest agaros e : 8 units Factor Xa. GM2AP
85 was separated from Xarrest Agarose and cleaved enzyme by centrifugation at 1000 x g for five minutes. Cleaved protein was separated from uncleaved protein by passing the supernatant through a Ni NTA resin column, with the flow through containing cleaved protein. As a final purification step, GM2AP was concentrated to less than 5 mL and was loaded to a Sephacryl S 200 High Resolution column for size exclusion chromatography at a flow rate of 0.5 mL/min Here residual detergent w as removed and the protein was equilibrated in the appropriate buffer for spectroscopic measurements. The isolated sample was concentrated using Amicon Ultra 15 centrifugal filter units (EMD Millipore, Billerica, MA) and protein concentration was determin ed by Brad ford assay and by UV absorption at 280 nm on a Cary 50 Bio UV Visible spectrophotometer (Varian, Palo Alto, CA) using the appropriate extinction coefficient depending on the particular variant. Protein purity was evaluated by SDS PAGE using 18% p re casted polyacrylamide gels and Coomassie Brilliant Blue detection. Post purification, protein was stored at 20C. Circular Dichroism (CD) Spectroscopy Measurements Circular dichroism spectra of 0.5 mg/mL of each GM2AP variant was obtained to ensure pro per protein folding and to compare the relative secondary structure of the variants to that of the reported wild type spectrum. 92 All measurements were collected on an Aviv 215 CD spectrometer using Hellma cuvettes with a 1 cm path length. Spectra were reported from 260 190 nm, collecting data at every nanometer. Temperature was 25 C and each spectrum is the average of four scans, corrected for buffer absorption and normalized to = 0 at 280 nm.
86 Intrinsic Fluorescence Quenching Measurements Steady state fluorescence experiments were carried out on a FluoroMax 3 fluorimeter (Jobin Yvon Horiba, NJ) with a temperature co ntrolled sample cell (at 20 0.1 C) and a H AAKE K20 temperature controller (Thermo Electron Corporation, Waltham, MA). Measurements were made using a 4 x 4 mm light path quartz cuvette (Starna, Atascadero, CA). The excitation and emission slits were set to 2 nm and 5 nm respectively for optimal int ensity measurements. 126 All s amples were excited at 295 nm in order to eliminate contributions from amino acid residues other than tryptophan, and also to minimize absorbance by acrylamide. Emission spectra were col lected between 300 nm and 450 nm for 1 M protein in 50 mM ammonium acetate buffer (pH 4.5 and pH 7.0), Collisional quenching experiments with acrylamide (neutral) and iodide (charged) were employed to analyze tryptophan accessibility of the GM2AP variant s. 4 M stock solutions of KI and acrylamide were made fresh, with the addition of 0.1 M sodium thiosulfate to the KI stock solution to p revent triiodide formation given it absorbs in the same region that tryptophan fluoresces. 127 Emmission spectra were collected for 1M GM2AP variant solutions in 50 mM sodium acetate at pH 4.5 and 7.0 in the absence of and after each titration of quencher. All measurements were collected in triplicate. Quenching data was analyzed u sing the Stern Volmer equation (3 1) where F 0 and F are the fluorescence intensities in the absence and presence of quencher respectively, [Q] is the quencher concentration, and K SV is the Stern Volmer quenching constant. The Stern Volmer quenching constant was obtained from plots of
87 the ratio of fluorescence intensity at 344nm in the absence of quencher to fluorescence in the presence of quencher vs quencher concentration, and it provided information about the degree of qu enching of tryptophan fluorescence. The fraction of total fluorophore accessible to quencher was determined using the modified Stern Volmer equation ( 3 2) where f a is the fraction of the initial fluorescence accessible to qu encher, K a is the modified quenching constant and F 0 F and Q represent the same variables described in equation 3 1. Plots of F 0 /F 0 F vs the reciprocal of quencher concentration were generated and the reciprocal of the y intercept gave f a Results and Dis cussion Site Directed Mutagenesis The QuikChange site directed mutagenesis kit was used in this work to construct a series of tryptophan to alanine substitutions. W5A, W5AW63A, W5AW131A, and W63AW131A were constructed using pET16b GM2AP wild type as the DN A template. The mutagenesis method employed the use of mutagenic primer directed GM2AP gene amplification using polymerase chain reaction. Agarose gel electrophoresis was performed after Dpn I digestion of the methylated parent plasmid and selection of muta tion containing amplified DNA 128 to verify successful DNA amplification. Verification of site directed mutagenesis was obtained by the observation of the Dpn I digested PCR products of W5AW63A (Lanes 1, 2, and 3) and W5AW131A (Lanes 4, 5, 6, and 7) analyzed in 1% agarose gel depicted in Figure 3 3 Lane 8 shows pET16b GM2AP wild type as a ref erence. The bands are approximately 6kb in size which
88 represent plasmid DNA containing nicks, and are at the expected pos i tion for linear pET16b plasmid with the GM2AP insert. The absence of b ands in lanes 2 and 7 were indicative of a failed PCR amplificat ion reaction. Adding DMSO to the mutagenesis reaction mixture resulted in positive results as it prevented primers and the DNA template from annealing to itself by disrupting base pairing. 129 After PCR, the resulting mixtures from mutagenesis were transformed into XL 1 blue E. coli cells in order t o seal nicks in the plasmid DNA and to enhance the stability of the gene in the plasmid. The cells were grown on ampicilin containing LB agar plates, to allow for the selection of cells that have taken up plasmid possessing our gene of interest. Colonies o n the agar plates were selected and the DNA was harvested using a QIAprep spin miniprep kit. Confirmation of amino acid mutations was done by sequencing facility. Figure 3 4 shows a sample agarose gel picture of the purified products which were submitted. Results from DNA sequencing show the DNA and corresponding amino acid sequences of the GM2AP variants in Figures 3 5, 3 6, 3 7, 3 8, and 3 9 Tryptophan was successfully muta ted to alanine as depicted by TGG GCG conversions at the respective positions in the sequences highlighted in red Sequence analysis showed that all desired variants we re obtained, and their corresponding plasmids containing the mutated genes were transfor med into E. coli BL21 DE3 cells (a host which allows for high level protein expression and easy induction) for protein expression and purification. Protein Expression and Purification GM2AP expression and purification were performed based on modifications reported by Wright et al. 53 and were optimized in our lab. During the purification
89 process, guanidine HCl and urea buffers were prepared fresh because urea decomp oses to form cyanate ions, which can covalently modify primary amines on the target protein For urea buffer any degradation products were removed by ion exchange with AG501 X8 mixed ion bed resin to prevent protein carbamylation. Glutathione oxidized and reduced were added to the refolding buffer to assist with the formation of the 4 disulfide bonds in the protein. Figure 3 10 shows sample chromatographs after the Ni affi nity (A ), anion exchange Q (B ), b uffer exchange (C) Ni after his tag cleavage (D ), an d size exclusion (E ) columns throughout the purification process. Protein was detected by monitoring UV absorption at 280 nm (blue traces). Conductivity of the solution was also monitored (brown traces). The purity and molecular weight of the various GM2AP variants were monitored by running SDS PAGE, and comparing the bands to known molecular weight standards (Figure 3 11) Bands appeared near the 20 kDa marker which corresponds closely to the 18kDa molecular weight of GM2AP. The W5AW63A variant in lane 4 s howed another band of approximate size 15 kDa This band possibly sug gests the presence of degraded protein. The protein sample was re run through a size exclusion column in efforts to separ ate properly folded protein from degraded protein. Circular dichro ism spectra were collected to ensure that the secondary structure of GM2AP variants was conserved after the tryptophan to alanine mutations. Figure 3 12 shows CD spectra of all the protein variants overlaid and compared to the published spectrum for GM2AP wild type. 92 All variants show a similar profile in agreement with that of the published wild type, where the secondary structure of the protein predominantly co n sists of sheets.
90 Intrinsic Fluorescence Quenching Native GM2AP is instrinsically fluorescent because it contains three tryptophan, three tyrosine, and five phenylalanine residues. Our goal was to determine the environment of the tryptophan residues in GM2AP to ex plore the potential of studying the mechanism of this protein using intrinsic protein fluorescence. For multi tryptophan proteins, fluorescence studies can be complicated because each fluorescing residue may have its own fluorescence yield, spectral positi on and may not contribute equally to the observed fluorescence emission 130 A series of tryptophan to alanine amino acid substitutions were constructed, allowing for single (W5AW63A, W5AW131A, W63AW131A ) and one double (W5A) tryptophan residues to remain in the amino acid sequence of GM2AP. T he degree of quenching of the GM2AP variants was investigated using neutral and charged quenchers. Two (W63 and W131) of the three tryptophan residues are located in the putative membrane binding regions of GM2AP. If the residues are accessible for fluorescence quenching, then the sites are exposed and suitable for monitoring changes in protein fluorescence as a result of lipid binding and/or extraction. To selectivel y excite the tryptophan residues and not phenylalanine or tyrosine an excitation wavel ength of 295 nm was used and emission peaks were obtained Maximal emission wavelength was observed at 346 nm for each variant (Figures 3 13 to 3 16 ). Total concentrati ons of 0 0.31 M neutral quencher acrylamide and charged quencher KI, were titrated in GM2AP so lutions. Figure 3 13 shows the total emission spectra at 20 C associated with quenching of tryptophan residues in GM2AP W5A at pH 4.5 (A, C) and pH 7.0 (B,D). Similar results were observed for all other variants (Figures 3 14 to 3 16 ) F luorescence intensity decreases with increasing amounts of both quenchers It
91 must be noted that the overall quenching of protein by acrylamide was greater than that by KI. Tota l tryptophan fluorescence intensity was quenched on average by 80% with acrylamide, and 55% by iodide. Acrylamide and iodide are known to quench the fluorescenc e of indole derivatives that have aqueous exposure but the charge and large size of KI due to h ydration, restricts it from accessing buried residues in hydrophobic environments. 98,131 Thus, being a charged species electrostatic effects may affect the quenching action of KI. Stern Volmer plots (Figure 3 17 ) we re generated and quenching constants (Table 3 1 ) were obtained to determine the degree of tryptophan quenching by acrylamide and iodide. The classic interpretation F 0 /F vs. [Q] di ctates that upward curvature or linear plots indicate that all tryptophan res idues are homogeneous and are nearly equally accessible to quencher while downward curvature results from tryptophan residues are in different environments. 131 Plots at acid ic and neutral pH for both quenchers suggested that the tryptophan residues in the GM2AP variants were accessible to quencher, though not to the same degree. For iodide, it was observed that at neutral pH the Stern Volmer quenching constants for a ll protein variants were lower than those at acidic pH. This may be explained by the glutamic acid ( Glu ) and aspartic acid ( Asp ) residues located in the putative membrane binding loops of the protein the same regions where W63 and W131 are also present. At neutral pH Glu and Asp are expected to be negatively charged (side chain pKa values of 4.07 and 3.90 respectively) which should generate electrostatic repulsions with the negatively charged iodide ions, resulting in lower Ksv values.
92 Conversely, trypt ophan yl quenching by acrylamide was lower at acidic pH than at neutral pH. The resolved crystal structure of GM2AP shows two conformations of the mobile loop containing W131 one where the loop is aqueous exposed, and another where the loop was tucked int o the protein. 53 The d ecrease in degree of quenching by acrylamide at acidic pH conformation at acidic pH of the region of GM2AP where W131 resides, as was evident by a lower K SV value when W131 is present (W5AW63A) in the protein. GM2AP W63AW131A displayed lower degrees of quenching by both acrylamide and iodide. W5A is located on the backside of the protein, in the region of the hydrophobic binding pocket. Predictably, W5A was not as accessible to quencher as W63 and W131 and this site did not show a pH dependence for acrylamide Figure 3 18 shows m odifie d Stern Volmer plots which by calculating the reciproc al of the y intercept for each plot, information about the fraction of accessible fluorophore to quencher at acidic and neutral pH can be obtained. T able 3 1 sum m arizes the fractional accessibility, f a to trypto phan fluorescence of the GM2AP variants The accessibility of the tryptophan residues to both quenchers were approxim ately 1 .This indicates that the residues were accessible for quenchi ng regardless of quencher or pH. However, it must be noted that while f a tells the fraction of total fluorescence ac cessible to quencher, it does not mean that all tryptohan residues are equally accessible for quenching 98 Conclusions Here we described the site directed mutagenesis and protein expression and purification methods W to A constructs of GM2AP. CD spectroscopy revealed that the amino acid substitutions did not alter proper protein folding. The variants were then
93 used to characterize the intrinsic tryptophan fluorescence of the protein. Quenching results suggest that while the tryptophan residues were accessible to acrylamide and iodide, the degree of quenching varied. W63 seemed to be most accessible, followed by W131, and W5 was not as accessible due to its location in the hydrophobic region of GM2AP These findings insinuate that since W63 and W131 are located in putative membrane binging regions of the protein, and are solvent exposed, intrinsic tryptophan fluorescence may be a useful technique for the further investigation of the precise molecular mechanism of GM2AP functio n without the need for attaching fluorescent probes to the protein
94 Figure 3 1. pET16b vector map. The T7 promoter region is depicted by the thick black arrow. This region contains the site for insertion of the gene encoding GM2AP DNA Figure was repr inted from https://wasatch.biochem.utah.edu/chris/links/pET16b.pdf. Accessed May 2012
95 1 1 76 26 151 51 226 76 301 101 376 126 451 151 AGTAGCTTTTCCTGG GATAACTGTGATGA A GGGAAGGACCCTGCG GTGATCAGAAGCCTG ACTCTGGAGCCTGAC S S F S W D N C D E G K D P A V I R S L T L E P D CCCATCGTCGTTCCT GGAAATGTGACCCTC AGTGTCGTGGGCAGC ACCAGTGTCCCCCTG AGTTCTCCTCTGAAG P I V V P G N V T L S V V G S T S V P L S S P L K GTGGATTTAGTTTTG GAGAAGGAGGTGGCT GGCCTCTGGATCAAG ATCCCATGCACAGA C TACATTGGCAGCTGT V D L V L E K E V A G L W I K I P C T D Y I G S C ACCTTTGAACACTTC TGTGATGTGCTTGAC ATGTTAATTCCTACT GGGGAGCCCTGCCCA GAGCCCCTGCGTA CC T F E H F C D V L D M L I P T G E P C P E P L R T TATGGGCTTCCTTGC CACTGTCCCTTCAAA GAAGGAACCTACTCA CTGCCCAAGAGCGAA TTCGTTGTGCCTGAC Y G L P C H C P F K E G T Y S L P K S E F V V P D CTGGAGCTGCCCAGT TGGCTCACCACCGGG AACTACCGCATAGAG AGCGTCCTGAGCAGC AGTGGGAAGCGTCTG L E L P S W L T T G N Y R I E S V L S S S G K R L GGCTGCATCAAGATC GCTGCCTCTCTAAAG GGCATA G C I K I A A S L K G I Figure 3 2. E. col i codon optimized DNA and amino acid sequence of GM2AP wild type protein
96 Figure 3 3 Sample agarose gel picture of GM2AP variants after Dpn I digestion. Lanes 1, 2, 3 W5AW63A; 4, 5, 6, 7 W5AW131A; 8 pET16b GM2AP wild type as a reference; 9 1 kb s tandar d DNA ladder from NEB ( Ipswich, MA) Figure 3 4 Sample agarose DNA gel of pET1 6b GM2AP variants after plasmid purification Lanes 1, 2, 3 W5AW63A ; 4, 5, 6 W5AW131A; 7 1 k b standard DNA ladder from NEB (Ipswich, MA)
97 1 1 76 26 151 51 226 76 301 101 376 126 451 151 AGTAGCTTTTCC GC G GATAACTGTGATGAA GGGAAGGACCCTGCG GTGATCAGAAGCCTG ACTCTGGAGCCTGAC S S F S A D N C D E G K D P A V I R S L T L E P D CCCA TCGTCGTTCCT GGAAATGTGACCCTC AGTGTCGTGGGCAGC ACCAGTGTCCCCCTG AGTTCTCCTCTGAAG P I V V P G N V T L S V V G S T S V P L S S P L K GTGGATTTAGTTTTG GAGA AGGAGGTGGCT GGCCTCTGGATCAAG ATCCCATGCACAGAC TACATTGGCAGCTGT V D L V L E K E V A G L W I K I P C T D Y I G S C ACCTTTGAACACTTC TGTGATGTGCTTGAC ATGT TAATTCCTACT GGGGAGCCCTGCCCA GAGCCCCTGCGTACC T F E H F C D V L D M L I P T G E P C P E P L R T TATGGGCTTCCTTGC CACTGTCCCTTCAAA GAAGGAACCTACTCA CTGCCCA AGAGCGAA TTCGTTGTGCCTGAC Y G L P C H C P F K E G T Y S L P K S E F V V P D CTGGAGCTGCCCAGT TGGCTCACCACCGGG AACTACCGCATAGAG AGCGTCCTGAGCAGC AGTGGGAAGCG TCTG L E L P S W L T T G N Y R I E S V L S S S G K R L GGCTGCATCAAGATC GCTGCCTCTCTAAAG GGCATA G C I K I A A S L K G I Figure 3 5 E. coli codon optimized DNA and amino acid sequence s of GM2AP W5A. Th e amino acid mutation site is colored red
98 1 1 76 26 151 51 226 76 301 101 376 126 451 151 AGTAGCTTTTCC GC G GATAACTGTGATGAA GGGAAGGACCCTGCG GTGAT CAGAAGCCTG ACTCTGGAGCCTGAC S S F S A D N C D E G K D P A V I R S L T L E P D CCCATCGTCGTTCCT GGAAATGTGACCCTC AGTGTCGTGGGCAGC ACCAGTGTCCCCCTG AGTTCTCC TCTGAAG P I V V P G N V T L S V V G S T S V P L S S P L K GTGGATTTAGTTTTG GAGAAGGAGGTGGCT GGCCTC GC GATCAAG ATCCCATGCACAGAC TACATTGGCAGCTGT V D L V L E K E V A G L A I K I P C T D Y I G S C ACCTTTGAACACTTC TGTGATGTGCTTGAC ATGTTAATTCCTACT GGGGAGCCCTGCCCA GAGCCCCTGCGTACC T F E H F C D V L D M L I P T G E P C P E P L R T TATGGGCTTCCTTGC CACTGTCCCTTCAAA GAAGGAACCTACTCA CTGCCCAAGAGCGAA TTCGTTGTGCCTGAC Y G L P C H C P F K E G T Y S L P K S E F V V P D CTGGAGCTGCCCAGT TGGCTCACCACCGGG AACTACCGCATAGAG AGCGTCCTGAGCAGC AGTGGGAAGCGTCTG L E L P S W L T T G N Y R I E S V L S S S G K R L GGCTGCATCAAGATC GCTGCCTCTCTAAAG GGCATA G C I K I A A S L K G I Figure 3 6. E. coli codon optimized DNA and amino acid sequences of GM2AP W5AW63A. The amino acid mutation sites are colored red
99 1 1 76 26 151 51 226 76 301 101 376 126 451 151 AGTAGCTTTTCC GC G GATAACTGTGATGAA GGGAAGGACCCTGCG GTGATCAGAAGCCTG ACTCTGGAGCCTGAC S S F S A D N C D E G K D P A V I R S L T L E P D CCCATCGTCGTTCCT GGAAATGTGACCCTC AGTGTCGTGGGCAGC ACCAGTGTCCCCCTG AGTTCTCCTCTGAAG P I V V P G N V T L S V V G S T S V P L S S P L K GTGGATTTAGTTTTG GAGAAGGAGGTGGCT GGCCTCTGGATCAAG ATCCCATGCACAGAC TACATTGGCAGCTGT V D L V L E K E V A G L W I K I P C T D Y I G S C ACCTTTGAACACTTC TGTGATGTGCTTGAC ATGTTAATTCCTACT GGGGAGCCCTGCCCA GAGCCCCTGCGTACC T F E H F C D V L D M L I P T G E P C P E P L R T TATGGGCTTCCTTGC CACTGTCCCTTCAAA GAAGGAACCTACTCA CTGCCCAAGAGCGAA TTCGTTGTGCCTGAC Y G L P C H C P F K E G T Y S L P K S E F V V P D CTGGAGCTGCCCAGT GC GCTCACCACCGGG AACTACCGCATAGAG AGCGTCCTGAGCAGC AGTGGGAAGCGTCTG L E L P S A L T T G N Y R I E S V L S S S G K R L GGCTGCATCAAGATC GCTGCCTCTCTA AAG GGCATA G C I K I A A S L K G I Figure 3 7. E. coli codon optimized DNA and amino acid sequences of GM2AP W5AW131A. The amino acid mutation sites are col ored red
100 1 1 76 26 151 51 226 76 301 101 376 126 451 151 AGTAGCTTTTCCTGG GATAACTGTGATGAA GGGAAGGACCCTGCG GTGATCAGAAGCCTG ACTCTGGAGCCTGAC S S F S W D N C D E G K D P A V I R S L T L E P D CCCATCGTCGTTCCT GGAAATGTGACCCTC AGTGTCGTGGGCAGC ACCAGTGTCCCCCTG AGTTCTCCTCTGAAG P I V V P G N V T L S V V G S T S V P L S S P L K GTGGATTTAGTTTTG GAGAAGGAGGTGGCT GGCCTC GC GATCAAG ATCCCATGCACAGAC TACATTGGCAGCTGT V D L V L E K E V A G L A I K I P C T D Y I G S C ACCTTTGAACACTTC TGTGATGTGCTTGAC ATGTTAATTCCTACT GGGGAGCCCTGCCCA GAGCCCCTGCGTACC T F E H F C D V L D M L I P T G E P C P E P L R T TATGGGCTTCC TTGC CACTGTCCCTTCAAA GAAGGAACCTACTCA CTGCCCAAGAGCGAA TTCGTTGTGCCTGAC Y G L P C H C P F K E G T Y S L P K S E F V V P D CTGGAGCTGCCCAGT GC GCTCACCACCGGG A ACTACCGCATAGAG AGCGTCCTGAGCAGC AGTGGGAAGCGTCTG L E L P S A L T T G N Y R I E S V L S S S G K R L GGCTGCATCAAGATC GCTGCCTCTCTAAAG GGCATA G C I K I A A S L K G I Figure 3 8. E. coli codon optimized DNA and amino acid sequences of GM2AP W63AW131A. The amino acid mutation sites are colored red
101 1 1 76 26 151 51 226 76 301 101 376 126 451 1 51 AGTAGCTTTTCC GC G GATAACTGTGATGAA GGGAAGGACCCTGCG GTGATCAGAAGCCTG ACTCTGGAGCCTGAC S S F S A D N C D E G K D P A V I R S L T L E P D CCCATCGTCGTTCCT GGAAATGTGACCCTC AGTGTCGTGGGCAGC ACCAGTGTCCCCCTG AGTTCTCCTCTGAAG P I V V P G N V T L S V V G S T S V P L S S P L K GTGGATTTAGTTTTG GAGAAGGAGGTGGCT GGC CTC GC GATCAAG ATCCCATGCACAGAC TACATTGGCAGCTGT V D L V L E K E V A G L A I K I P C T D Y I G S C ACCTTTGAACACTTC TGTGATGTGCTTGAC ATGTTAATTCCTACT GGGG AGCCCTGCCCA GAGCCCCTGCGTACC T F E H F C D V L D M L I P T G E P C P E P L R T TATGGGCTTCCTTGC CACTGTCCCTTCAAA GAAGGAACCTACTCA CTGCCCAAGAGCGAA TTCGTTGTG CCTGAC Y G L P C H C P F K E G T Y S L P K S E F V V P D CTGGAGCTGCCCAGT GC GCTCACCACCGGG AACTACCGCATAGAG AGCGTCCTGAGCAGC AGTGGGAAGCGTCTG L E L P S A L T T G N Y R I E S V L S S S G K R L GGCTGCATCAAGATC GCTGCCTCTCTAAAG GGCATA G C I K I A A S L K G I Figur e 3 9 E. coli codon optimized DNA and amino acid sequences of GM2AP W5AW63AW131A. The amino acid mutation sites are colored red
102 Figure 3 10 Sample column chromatographs during GM2AP purification. UV absorption at 280 nm (blue traces) was used to detect protein af ter (A) Ni NTA column, (B) anio n exchange Q column, (C) buffer exchange column, (D) Ni NTA column after histag cleavage, and (E) S200 size exclusion column The brown traces show solution conductivity
103 Figure 3 11 18% SDS PAGE gel of 15 L samples of purified GM2AP protein and variants after size exclusion chromatography : Lane 1 Wild type ; 2 W5A; 3, 4 W5AW63A ; 5, 6 W5AW131A ; 7 W63AW131A The two MW weight lanes correspond to p recision plus protein standards from BioRad (Hercules CA)
104 Figure 3 12 Circular dichroism spectra of GM2AP wild type (black) and 0.5 mg/mL samples of GM2AP variants: W5A (red), W5AW63A (blue), W5AW131A (green), W63AW131A (fuchsia), and W5AW63AW131A (olive)
105 Figure 3 13 Fluorescence emission spectra of 1 M GM2AP W5A showing results from the titration of increasing amounts of acrylamide at A) pH 4.5, B) pH 7.0, and KI at C) pH 4.5 and D) pH 7.0 as described in the text. Emmision scans were recorde d using 2 nm and 5 nm excitation and emission slits while exciting at 295 nm A B C D
106 Figure 3 14 Fluorescence emission spectra of 1 M GM2AP W5AW63A showing results from the titration of increasing amounts of acrylamide at A) pH 4. 5, B) pH 7.0, and KI at C) pH 4.5 and D) pH 7.0 as described in the text. Emmision scans were recorded using 2 nm and 5nm excitation and emission slits while exciting at 295 nm A C B D
107 F igure 3 15 Fluorescence emission spectra o f 1 M GM2AP W5AW131A showing results from the titration of increasing amounts of acrylamide at A) pH 4.5, B) pH 7.0, and KI at C) pH 4.5 and D) pH 7.0 as described in the text. Emmision scans were recorded using 2 nm and 5nm excitation and emission slits while exciting at 295 nm A B C D
108 Figure 3 16 Fluorescence emission spectra of 1 M GM2AP W63AW131A showing results from the titration of increasing amounts of acrylamide at A) pH 4.5, B) pH 7.0, and KI at C) pH 4.5 and D) pH 7.0 as described in the text. Emmision scans were recorded using 2 nm and 5nm excitation and emission slits while exciting at 295 nm A B C D
109 Figure 3 17 Stern Volme r plots of GM2AP A) W5A, B) W5A W63A, C)W5AW131A, D) W63AW131A quenching by 0 0.31 M acrylamide a t pH 4.5 (black square); and pH 7.0 (red circle); KI at pH 4.5 (blue triangle) and pH 7.0 (green inverted triangle). All titrations were performed at 20C with an initial protein concentration of 1 M; excitation and emission wavelength of 295 nm and 346 n m respectively. Data calculations were corrected for dilution and experiments were done in triplicate
110 Figure 3 18 Modified Stern Volmer plots of GM2AP A) W5A, B) W5A/W63A, C)W5AW131A, D) W63AW131A quenching by 0 0.3M acrylamide at pH 4.5 (black square) and pH 7.0 (red circle); 0 0.31 M KI at pH 4.5 (blue triangle) and pH 7.0 (green inverted triangle). All titrations were performed at 20C with an initial protein concentration of 1 M; excitation and emission wavelength of 295 nm and 346 nm resp ectively. Data calculations were corrected for dilution and experiments were done in triplicate
111 Table 3 1 Stern Volmer quenching constant s (K SV ) and f raction of total tryptophan accessible ( f a ) to acrylamide and potassium iodide at acidic and neutral pH ACRYLAMIDE pH 4.5 pH 7.0 Tryptophan Mutant K SV (M 1 ) f a K SV (M 1 ) f a W5A 10.5 0.1 0.97 0.01 16.7 0.3 0.95 0.01 W5A W63A 8.4 0.1 0.81 0.02 12.9 0.1 0.90 0. 0 2 W5A W131A 9.5 0.1 1.10 0.02 13.8 0.1 0.99 0. 0 2 W63A W131A 3.42 0.03 1.03 0.03 3.3 0.10 1.01 0.05 IODIDE W5A 5.6 0.1 1.00 0. 0 2 3.5 0.1 1.11 0.2 W5A W63A 6.5 0.1 1.00 0. 0 2 4.1 0.1 1.00 0.2 W5A W131A 6.5 0.1 1.21 0. 0 2 3.4 0.1 1.20 0.2 W63AW131A 1.38 0.01 1.3 0 0. 0 1
112 CHAPTER 4 VESICLE BINDING AND LIPI D EXTRACTION STUDIES OF GM2 ACTIVATOR PROTEIN VARIANTS Introduction Gangliosides are a group of sialic acid containing glycosphingolipids (GSLs) that are primarily found in the outer leaflet of neuronal cell membranes, where they function as cell markers and are available for cell signaling events. 1,53 Ganglioside catabolism takes place in the lysosomal compartments of cells and degradation which entails a stepwise cleavage of oligosaccharide groups, occurs by acid ic hydrolases 59 For GSLs with fewer than four oligosaccharide head group s, hydrolases require non enzymatic accessory proteins called sphingolipid activator proteins (SAPs) for cleavage. 13,132 GM2 Activator Protein (GM2AP), one of the five SAPs, is an essential accessory protein for th e degradation of the ganglioside GM2 to form GM3 (which is further broken down into reusable components of sphingosine by numerous other enzymes) GM2AP function involves extraction of GM2 from intralysosomal vesicles, followed by a protein lipid leaving t hexosaminidase A (HexA) for cleavage in solution. 1,51,52 Disruptions in the degradation process lead to the accumulation of harmful quantities of GM2 gangliosi de in the neuronal cells, resulting in a collection of diseases called GM2 gangliosidoses disorders that cause physical and mental disorders in infants that usually resulting in death by age four. 3 The mature form of GM2AP has been isolated from natural sources and expressed as a glycosylated protein from insect 133 and yeast 56 cells. The nonglycosylated protein has been expressed in Escherichia coli 124 and will be the form used in this study. GM2AP is an 18 kDa protein whose crystal structure of has been
113 resolved via X ray crystallography. Studies of the wild type protein have revealed three distinct monomers in the 11 monomer unit cell. 53 The secondary structure of GM2AP comprises an eight cup fold forming a hollow hydrophobic pocket suitable for binding GM2 and contains approximately one half of the 162 total amino acid residues in the protein. 53 The cavity is accessible from one end of the protein only and is lined by surface loops and a 2.5 turn alpha helix at its rim. 53 Analyses of the protein conformation s in crystal structures of GM2AP show that one of the two putative membrane binding loops of GM2AP containing W131 and referred to as the flexible loop has one conformation where the loop is more structured with W131 tucked into the protein, and two othe r unstructured conformations where W131 is aqueous exposed 53 In addition to taking part in the con version of GM2 to GM3, GM2AP also functions as a lipid transfer protein. The protein is known in vitro to bind and extract several GSLs and phospholipids such as phosphatidylglycerol (PG) and phopsphatidylcholine (PC) from micelles or liposomes and transfer them as soluble 1:1 complexes between membranes. 1,57,60 Phospholipids bind to GM2AP in a different orientation from GM2. GM2 binds in a way that the ceramide tails are tucked in the hydrophobic pocket of the protein and the bulky carbohydrate head group protrudes from the surf ace of the protein and into the aqueous environment. Phospholipids like PG however, bind in such a way that the entire lipid is buried in the hydrophobic cavity of GM2AP. 60 Figure 4 1 shows a ribbon structure of GM2AP showing the model ed binding modes of GM2 and PG adopting different orientations within the hydrophobic pocket of the protein.
114 Despite the fact that the crystal structure and function of GM2AP have been determined, little is known about the specific mechanism involving the interaction with vesicular surfaces, extraction of lipids, and transfer of these lipids between membranes. Studies on the interaction of GM2AP with its specific ligand GM2 have been performed utilizing several techniques such as sucrose density ultracentri fugation, sucrose density isoelectric focusing, polyacrylamide gel electrophoresis, circular dichroism, and steady state fluorescence spectroscopy. 51,52 protein has als o been investigated using gel filtration chromatography, thin layer chromatography (TLC), fluorescence resonance energy transfer (FRET), fluorescence dequenching assays and surface plasmon resonance. 27,51,57,82 GM2A P contains three native tryptophan residues (Figure 4 1), two of which are located in mobile regions of the protein thought to be involved in vesicular membrane binding. We can therefore mutate the more hydrophobic tryptophan residues to a less hydrophobic residue in effo rts to modulate effective membrane pa rtitioning and protein activity in order to further understand the molecular mechanism of GM2AP binding and extraction. Here, a series of sing le, double and triple point tryptophan (W) to alanine (A) mu tations have been engineered in GM2AP to investigate the changes in lipid extraction from vesicles as a result of decreasing the hydrophobicity of the regions of the protein where the residues are located. The ability of each tryptophan variant to extract and transfer phospholipids from liposomes was investigated by utilizing a fluorescenc e assay containing dansyl label ed phospholipids. 134 Additionally, the extraction efficiency of GM2AP in a POPC:GM2 mixture was d etermined by physical separation and direct measurement of the amount of protein bound to GM2 in solution
115 after separation. 135 Ch anges in extraction efficiency and by extension, the ability of GM2AP to bind and extract GM2 was investigated for each protein variant containing the W to A substitutions. Materials and Methods 1 Palmi toyl 2 Oleoyl sn Glycero 3 Phosphocholine (POPC) in chloroform was purchased from Avanti Polar Lipids (Alabaster, AL) and used without further purification. N (5 dimethylaminonaphthalene 1 sulfonyl) 1, 2 dihexadecanoyl sn glycero 3 phospho ethanolamine (Dan syl DHPE ) was purchased from Molecular Probes (Eugene, OR) in the form of powder. Monosialoganglioside GM2 was purchased from Sigma Aldrich (St. Louis, MO) as a powder. Unless otherwise stated, all other reagents were purchased from Fisher Scientific (Pitt sburg, PA). GM2 Activator Protein Expression and Purification The DNA of recombinant GM2AP was cloned in pET16b vector and site directed mutagenesis was performed to acquire the desired protein variants. The protein was over expressed in BL21 (DE3) E. col i cells and purified as previously described. 134 A more detailed description of protein purification can be found in Chapter 3. Purity of final protein samples was identified via SDS PAGE using 18% pre casted poly acrylamide gels and Coo massie Brilliant Blue detection To ensure the protein was properly folded, circular dichroism spectra were obtained to verify secondary structure was analogous to published circular dichroism spectra of GM2AP wildtype 92 Protein concentration was determined by Bradford assay and by UV absorption at 280 nm on a Cary 50 Bio UV Visible spectrophotometer (Varian, Palo Alto, CA) The appropriate extinction coefficient for each protein variant was provided by the ExPASy website ( http://expasy.org ) and is listed in Table 4 1.
116 Lipid Preparation M ixtures of the appropriate volumes of each lipid in chloroform were pre pared and dried under a stream of nitrogen gas. The mixture was further dried overnight for at least 8 hours in a desiccator to produce a film. For fluorescent vesicles (those containing dansyl DHPE), the tubes containing the films were wrapped with alumin um foil due to fluorophore sensitivity to light. The dr ied films were rehydrated into appropriate volumes of 50 mM NaOAc, pH 4.8 at room temperature, followed by vortex mixing for one hour and five freeze thaw cycles using liquid nitrogen Each lipid mixtu re was extruded by 55 passes through 100 nm polycarbonate filters using an Avanti handheld mini extruder (Avanti Polar Lipids) to produce large unilamellar vesicles (LUVs). Fluorescence Spectroscopic Measurements Steady state fluorescence experiments were carried out using a FluoroMax 3 fluorimeter (Jobin Yvon Horiba, NJ) with a temperature controlled sample cell (at 20 0.1C) and a HAAKE K20 temperature controller (Thermo Electron Corporation, Waltham, MA). Measurements were made using a 4 mm light path q uartz cuvette (Starna, Atascadero, CA). The excitation and emission slits were set to 2 nm and 5 nm respectively and GM2AP in solution was excited at 295 nm. For experiments involving the use of lipids, the excitation and emission polarizers were configure d to 90C and 0C respectively in order to provide maximal suppression of scattering artifacts associated with lipid vesicles, which can affect emission intensity and spectral measurements 126 Fluorescence intensities were corrected for dilution before final analyses
117 Dansyl DHPE Extraction Assay 1 mM stock samples of 4:1 molar ratio POPC: d ansyl DHPE LUVs were diluted with 50 mM sodium acetate buffer pH 4.8 to g ive a final concentration of 1 M LUVs in 300 L solution. Emission spectra were collected between 400 and 600 nm with a dansyl excitation of 340 nm and excitation and emission slit widths set to 8nm and 5 nm respectively. Protein was added to the LUVs to give a final concentration of 5 M and the fluorescence emission intensity and maximal wavelength change as the GM2AP:dansyl complex was being formed was monitored over time Spectra were collected until the maximum intensity was constant at or near 484 n m. Transfer rates of DANSYL DHPE by the different W to A variants monitored at 484 nm and were determined according to Equation 4 1 and plots of relative transfer of dansyl DHPE from liposomes as a function of time were generated. Relative transfer of dans yl DHPE = (4 1) where is the fluorescence emission intensity at 484 nm at time t, is the fluorescence emission intensities at 484 nm in the absence of protei n, and and is the constant maximum fluorescence intensity at 484 nm after dansyl DHPE had been sequestered from liopsomes GM2 Extraction 10 nmol GM2AP variants were mixed with 300 nmol POP C: GM2 (9:1) vesicles in 50 mM sodium a cetate buffer pH 4.8. Mixtures of GM2AP variants with vesicles had a final volume of 100L and were incubated at room temperature for 60 minutes before loading onto a 1.6 x 500 mm Sephacryl S 200 self packed column ( GE Healthcare, Piscataway, NJ). Wit h the aid of a peristaltic pump the vesicles and GM2AP GM2
118 complexes were separated by size exclusion chromatography (SEC). l eluate fractions were collected. GM2AP concentrations in fractions that d id not contain vesicles (did not scatter light) were determined by measuring UV absorbance at 28 0 nm in a 1 cm path length cuvette on a Cary 50 Bio UV Vis spectrophotometer (V arian, Palo Alto, CA) GM2 concentrations in each fraction were measured by a resorcinol assay previously described. 18,136 Breifly, the collected fractions were mixed with an equal volume of resorcinol reagent and then the mixture was incubated in boiling water for 15 minutes. The tubes containing the fractions were then cooled and 60 L of 85:15 n butyl acetate:n butanol were added to each tube. The mixture was centrifuged at 4000 rpm for 2 minutes using a benchto p centrifuge (Beckman, Fullerton, CA) to separate the phases. GM2 ganglioside concentration s from the GM2 reso rcinol complex were determined by measuring UV absorbance at 580 nm, using a 1 cm path length cuvette and an extinction coefficient of 5700 mol 1 c m 1 The amount of GM2AP variant boun d to the vesicles and hence, extraction efficiency of the protein was determined by subtracting the concentration of the protein in fractions that did not contain vesicles from the total protein concentration in the mix ture prior to gel filtration. Results and Discussion In addition to extracting and forming a complex with its functional ligand GM2, GM2AP is known to bind and extract other lipid ligands 12,25,60 Unlike GM2, where the sugar head group of the ligand is solvent exposed while the acyl chains are buried inside the hydrophobic cavity of the protein, GM2AP has been crystallized with phosphoglycerol lipids showing the entire lipid structure oriented in the binding pocket of the protein 60,137 It has been shown that GM2AP forms a 1:1 protein ligand complex with N (5 Dimethylaminonapthalene 1 sulfonyl) 1, 2 dihexadecanoyl sn glycero 3
119 phosphoethanolamine ( DANSYL DHPE ), a dansyl labele d glycerophospholipid, which is suspected to adopt a bindin g mode similar to that of PG 134 When the dansyl fluorophore is in a non polar environment, the wavelength of maximal emission is blue shifted when compa red to being in a polar environment 97 Fluorescence intensity is also increased. Figure 4 3 shows how this fluor escent property of dansyl is used to observe a wavelengt h shift from 518 nm when dansyl DHPE is in corporated in 100 nm large unilamellar vesicles consisting of 4:1 POPC:dansyl DHPE vesicles to 484 nm, when dansyl DHPE is sequestered by GM2AP for form the protein lipid complex 134 GM2AP wildtype DHPE (4:1) vesicles and emission spectra were obtained every 15 seconds until there was no change in wavelength shift. The molar ratio of GM2AP to total dansyl DHPE was 25:1 in order to ensur e complete sequestration of dansyl DHPE fr om the outer lipid bilayer. P lot s showing the change in wavelength intensity at 484 nm as a function of time while GM2A P was extracting dansyl DHPE were recorded a nd fit to appropriate curves (Figure 4 4B ) Gangl ioside degradation occurs in the acidic lysosomal compartment of cells where pH is a round 4.8. Figure 4 3A shows the rate of transfer at 484 nm a s wild type GM2AP extracts dansyl DHPE over time at different pHs. Data collected for pHs 4.8 and 5.5 were fit to single exponential first order curves (Equation 4 2), the data for pH 6.0 was fit to a Boltzmann function curve (Equation 4 3), and the data for pHs 6.5 and 7 were fit to linear curves. (4 2) (4 3)
120 (4 4) For data fit by first order exponential curves, the half lives, defined as the time required for one half of the total dansyl DHPE to be sequestered by GM2AP were determined. At pH 4.8, the half life was fou nd to be 1.5 0.2 minutes (Figure 4 4A ). The rate of lipid extraction decreases with increasing pH At pH 5.5 the observed time for one half of total dansyl DHPE extraction was 4 .3 0.3 minutes and at pH 6. 0, 16.4 1 mi nutes were required For data coll ected at pH 6.0, we defined the half life as the time when half of the total amount was extracted, where the total value was taken as the equilibrium value after 60 minutes. No appreciable lipid extraction was observed at pH 6.5 or 7.0 Although GM2AP can function as a lipid transfer protein where binding, extraction, and transfer of lipids other than GM2 can occur, acidic pH conditions (optimal at 4.8) seem to be required for extraction to proceed. Note, we have found that ligand binding is not pH dependen t, but membrane binding, and thus lipid extraction and transfer are. The pH dependence of membrane binding can be understood by considering the electrostatic potential of the surface of GM2AP as the pH changes. There are a number of aspartic acid (D) and g lutamic acid (E) residues around the binding pocket of GM2AP and at acidic pH, these amino acid residues likely become protonated. We can postulate that protonati on of D and E reduces the charge lining the pocket and exposes the hydrophobic residues in the binding cavity to accommodate lipid binding and extraction. Figure 4 4 B shows plots of the relative transfer of dansyl DHPE from liposomes to GM2AP monitored at 4 84 nm vs. time for the W to A variants in comparison with wild
121 type (WT) GM2AP. Plots for W5A W5A W63A, W5AW131A, W5AW63AW131A were fit to single exponential first order curves (Equation 4 2), and the W63W131A plot was fitted to a SGompertz sigmoidal curve (Equation 4 4). When tryptophan is substituted with a lanine at amino acid position 5 (W5A) similar extractio n rates to WT were observed where the time for one half of total dansyl DHPE extraction for each were 1.5 0.2 minutes and 2.2 0.1 minutes ; respectively This result is not unexpected given that W5 is located on the back side of t he pro and not in the region of the protein thought to be involved in binding to the lipid membrane surface for ligand extraction An amino acid residue substituti on at that site should not affect GM2AP transfer rates For W5A W131 A a half life of 2.7 0.2 minutes was observed, which is approximately 2x slower than WT. Resolved crystal structures of GM2AP show two different conforms of the W131 flexible loop of the protein in the 11 monomer unit cell, where the loop is solvent exposed in one con formation and tucked in the other. 53 Based upon results of cross linking studies of this loop (Jeff Carter PhD dissertation 2012), we believe that the flexibility of this loop regulates access to the hydrophobic cav ity where extracted lipid resides. Interestingly, the alanine substitution at this site does not dramatically alter the rate of ligand extraction; implying that the hydrophobicity of this loop does not regulate membrane binding. In comparison, for all con structs containing W63A, lipid extraction rates werefour or more times lower compared to WT. For the W5A/W63A variant a half life of 6.0 0. 3 minutes was observed. With the W63AW131A and W5AW63A W131A constructs the kinetics of dansyl DHPE sequestration slowed further, to 8. 0 0.1 and 10 0 4
122 m inutes ; respectively. Based upon X ray structures and results from our lab using power saturation spin labeling electron paramagnetic resonance (EPR), a model of how GM2AP interacts with the membranes to extract l ipids has been proposed. From X ray loop that contains W63 protruding into the membrane surface. Contrary to this orientation, powere saturation data suggest a model wher e the face of GM2AP has the opening to the lipid cleft, which includes the flexible loop containing W131 juxtaposed on the bilayer surface. 138 Results from these W/A mutagenesis studies indicate that it is likely that GM2AP first binds to the membrane via the hydrophobic loop, with then rotation to the orienta tion suggested by EPR results (Figure 4 5 ). Hence the hydrophobicity of W63 is critical for the initial membrane binding event, whereas W131 is likely only involved in modulating the mobility of the flexible loop. These results indicate d that although the W/A variants are still able to extract dansyl DPHE from vesicles, removal of the tryptophan residues from the apolar and flexible loops decreased initial extraction rates providing insights into the molecular steps of membrane binding, lipid recognition, and extraction. Wimley and White developed a hydrophobicity scale to determine the contribution of whole amino acid residues to the partitioning of lipid membranes. 139 The scale ranks the interfacial hydrophobicity for the 20 natural amino acids and it is shown that aromatic residues like tryptophan are more favored at the membrane interface due to high hydrophobic interactions. The hydrophobicity scale sup ports our claim that substituting alanine residues for tryptophan would disrupt the hydrophobic interactions at the membrane interface weakening GM2A P
123 ty to bind lipid membranes and extract modified phospholipid. Given that the binding pocket in GM2AP differs for phospholipids and gangliosides, we investigated the effects of the W/A substitution on the ability of GM2AP to mobilizes its functional ligand, GM2 from vesicles To determine how GM2 extraction efficiency was altered as a result of the W to A amino acid substitutions, an equilibrium gel filtration assay was used Extraction efficiency is described as the ratio of the molar amount of GM2 relative to GM2AP recovered after gel filtration in non vesicle fractions 125 Figure 4 6 shows the elution profiles of the amounts of GM2 and GM2AP measured in the collected fractions after gel filtration GM2AP was allowed to incubate for 60 minutes with 100 nm extrude d 9:1 POPC : GM2 vesicles in 50 mM NaOAc buffer at pH 4.8. The measured concentrations of GM2 and GM2AP were used to determine the extraction effi ciency of each protein variant. Due to light scattering resulting in high absorbance readings at 280 nm protein concentrations could not be determined in eluted fractions containing vesicles. However, the total protein concentration was known for each experiment, so the total amount of protein bound to GM2 was determined by subtracting the protein concentration mea sures from fractions without vesicles from the initial total protein concentration. WT and W5A constructs showed the largest ratio of about 0.5, meaning that approximately one half of the protein added to vesicles formed a complex with GM2. The W5A/W131A variant had a ratio of 0.43, while W5A/W63A and W63A/W131A had ratios of 0.40 and 0.37 respectively. Within error, these ratios are similar indicating that the equilibrium extraction of GM2 is not affected by making W to A substitution in the
124 putative mem brane binding loops of the protein. Results from this assay support the idea that the point mutations made in the mobile loops of the protein did not hinder lipid ng to the vesicular membrane prior to lipid extraction. Conclusions GM2AP is essential for GM2 degradation in neuronal cells and also functions as a phospholipid transfer protein As shown here, a dansyl based fluorescence assay to determine changes in lipid transfer, and a gel filtration assay to determine lipid extraction efficiency of a series of W to A GM2AP constructs was used to study the the mobile loops of GM2AP resulted in slower lip id b inding, but did not affect lipid extraction. Binding to lipid vesicles becomes mor e difficult because the hydrophobic characteristic of tryptophan residues located in the apolar and hydrophobic loops of GM2AP is favored for hydrophobic interactions at the lipid membrane interface. Our results corroborate previous findings that it is likely that GM2AP first binds to the membrane via the hydrophobic loop, with then rotation to the orientation that has the opening to the lipid cleft, which includes the fle xible loop containing W131 juxtaposed on the bilayer surface, allowing lipid to be extracted into the binding pocket of the protein. Additionally, while GM2AP also functions as a lipid transfer protein, an aci dic environment (near pH 4.8) seems to be requi red for optimal protein function.
125 Figure 4 1. Ribbon structure of GM2AP showing the modeled binding modes of GM2 and PG adopting different orientations within the hydrophobic pocket of the protein. The carbohydrate head group of GM2 protrudes from t he surface of the protein, while PG is entirely buried in the hydrophobic cavity. Reprinted from Journal of Molecular Biology 331 Wright, C.S.; Zhao, Q.; Rastinejad, F., Structural Analysis of Lipid Complexes of GM2 Activator Protein, 951 964., Copyright (2003), with permission from Elsevier
126 Figure 4 2 Ribbon structure of GM2AP (PDB ID IG13) showing the location of the three native tryptophan residues (green) and the two putative binding loops (yellow highlights). F igure made via VMD Table 4 1 Extinction coefficients of GM2AP wild type and its W to A va riants Protein Variant Extinction coefficient (M 1 cm 1) WT 23000 W5A 17460 W5AW63A 11960 W5AW131A 11960 W63AW131A 11960 W5AW63AW131A 6460 Mobile loop Apolar loop
127 Figure 4 3 Fluoresc ence emission spectra of 4:1 POPC:dansyl DHPE (black solid line) and M 1:1 wild type GM2AP :dansyl DHPE (grey dashed line ) complexes after 16 minutes. The blue shift in maximal emission wavelength and increase i n intensity suggests the dansyl headgroup is located in a more hydrophobic environment. Both protein and lipid vesicles were in 50 mM NaOAc, pH 4.8. The spectra were taken at 20C on a FLUOROMAX 3 with an excitation wavelength of 340 nm. The excitation pol arizer was set to 90 and the emission polarizer was set to 0 Relative Intensity
128 Figure 4 4 Changes in relative transfer as dansyl is being sequestered from 1 mM 4:1POPC:dans yl M WT at different pHs as a function of time M protein var iants at pH 4.8 as a function of time A B
129 Figure 4 5. Proposed model of the membra n e bound orientation of GM2AP. GM2AP first binds to the membrane v ia the hydrophobic loop (A),then rotates to the orientation that has the opening to the lipid cleft, which includes the flexible loop containing W131 juxtaposed on the bilayer surface (B) A B
1 30 Figure 4 6 Elution profiles showing the extraction efficiency of GM2 extraction by a series of W to A GM2AP variants 10 nmol protein was mixed with 300 nmol POPC:GM2 (9:1) vesicl es in 50 mM NaOAc pH 4.8 buffer and allowed to react for 60 minutes before separation via SEC. The average fraction size columns show the amount of GM2 (determined by resorcinol as say) in each fraction. The grey columns show the amount of protein (determined by UV Vis) in each fraction Fractions containing vesicles were not ass ayed for protein concentration Table 4 2. Half lives and extraction efficiencies of GM2AP variants and their ratios with respect to WT protein Variant Half life Ratio of half life to WT Extraction efficiency Ratio of extraction efficiency to WT WT 1.5 0.2 0.50 0.05 W5A 2.2 0.1 1. 5 0.3 0.47 0.07 1.0 0.2 W5AW63A 6.0 0.3 4.0 0.7 0.40 0.13 0.80 0.34 W5AW131A 2.7 0.2 1.8 0.4 0.43 0.09 0. 90 0.27 W63AW131A 8.0 0.1 5.3 0.8 0.37 0.04 0.7 0 0.15 W5AW63AW131A 10 0.4 7 1
131 CHAPTER 5 SURFACE PLASMON RESONANCE ENHANCED ELLIPSOMETRY STUDIES TO STUDY LIPID BILAYER INTERACTIONS BY GM2 ACTIVATOR PROTEIN Introduction GM2 activator protein (GM2AP) is an 18 kDa nonenzymatic sphingolipid activator protein (SAP) required for the lysosomal degradation of GM2 ganglioside by the exohydrolase, hexosaminidase A (Hex A) in neuronal cells 3,25 In vivo, the protein bind s to and extract s GM2 from intralysomal vesicles where the tetrasaccharide headgroup of the lipid is made accessible to Hex A for hydrolyti c cleavage of the N acetyl glucosamine ( GalNAc ) group, to form GM3. 35,59 A deficiency or defect of GM2AP causes a buildup of GM2 in cells w hich leads to the fatal neurological disorder, AB variant a form of GM2 gag liosidoses. 3 The mature form of GM2AP has been isolated fro m natural sources and has been expressed in its glycosylated form from insect and yeast cells. 122,140 Th e recombinant form of GM2AP has been expressed in and purified from Escherichia coli 36,141 Resolved crystal structure of nonglycosylated GM2AP reveals a hollow, hydrophobic binding pocket for lipids, formed from eight strands in the protein 53 Lipid extraction studies have been performed on GM2AP in vitro, and it is known that the protein is not specific for GM2 only, but it can bind and extract other nonspecific glycolipids from micelles and lipid vesicles, forming 1:1 complexes i n solution. 121 GM2AP can also bind and transfer phospholipids such as phosphatidylc holine (PC), phosphatidylglycerol (PG) and a dansyl labeled phosphatid ylethanolamie between membranes. 57,60,142 A dditionally, a n u mber of crystal structures have been published and deposited in the protein data ban k showing binding modes of GM2AP lipid complexes with GM2 and other phospholipids. 57,60 As such, GM2AP is referred to as a
132 lipid transfer protein. Despite knowledge about the crystal structure and function of GM2AP, details of the precise molecular process by which the protein binds and extracts lipids from vesicles are not well understood. Most GM2AP interactions with li pids have been investigated where the lipids exist in solution as mi celles or liposomes before ex traction and transfer A biomolecular interaction analysis of GM2AP via surface plasmon resonance on a Biacore Pioneer L1 chip has been reported, but the surface was functionalized with a hydrophilic dextran matrix and lipids were immobilized on the sensor chip in the form of large unilamellar vesicles (LUVs) for analysis. 27,92 I t was discovered from sucrose loaded sedimentation assays that during lipid extraction l ess than 15% of GM2AP remained on the surface of lipo somes in the presence and absence of GM2. 125 R esults from the assay explain the reason why it is difficult to solely investigate the membrane binding step involved in the overall interaction of GM2AP with lipids This finding coupled with the fact that GM2AP al so functions a lipid transfer protein lead us to pursue another system with which to study GM2AP: lipid membrane associated interactions planar supported lipid bilayers (SLBs). A typical SL B consists of a phospholipid bilayer that covers the surface of a planar solid support. Different solid sub s trates (glass, silicon, quartz, and mica) and surfaces (metal films and polymers ) have been used to support the formation of the bilayers ; 75 there a re three main m ethods by which SL Bs are formed: 1) The classical Langmuir Blod gett technique where lipid monolayers are transferred onto a surface ; 143,144 2) the adsorption of liposomes on a surface for membrane ves icle fusion ; 74 and 3) a combination of the two previous methods where the inner lipid monolayer is transferred
133 to a surface via the Langmuir Blodgett method, followed by vesicle fusion of the outer lipid monolayer. 145 SLBs serve as membrane mimics to stud y a number of biological processes for example cellular signaling events, 146 cell adhesion, 147 ligand receptor interactions, 148 membrane insertion of proteins, 149 and protein lipid binding. 150 In this work, planar supported lipid bilayers and surface plasmon resonance enhanced ellipsometry (SPREE) were used to examine the interaction of GM2AP with membranes. A lkanethiol self assemb led monolayer (SAM) and zirconium octadecylphosphonate modified surfaces on a thin gold film was used to form a lipid bilayer (the former forms a hybrid bilayer) by vesicle fusion. 77,79 T he gold film provided an op tically active and physiologically compatible co ating for SPREE analysis of the protein lipid membrane interactions. 151 Surface plasmon resonance enhanced ellipsometry (SPREE) is an optical technique that utilizes the combination of surface plasmon resonance (SPR) and ellipsometry to monitor and analyze adsorption and desorption of molecules on thin films. The addition of the SPR effect gives high sensitivity to the real time monitoring of binding events of proteins to SPBs. The tech nique is non destructive and does not require labeling probes unlike other surface analytical methods. This chapter reports on the use of phosphatidylcholine (PC) and phosphatidylglycerol (PG) lipids in varying ratios to construct model membranes. T he eff ect of pH, hydrophobicity and anionic lipid with the supported hybrid bilayer is examined here. A d ditionally, the effect of anionic lipid and suppo rted bilayer was also determined The results reveal ed that in the absence o f PG,
134 GM2AP extracts PC from the hybrid bilayer. Protein binds to the membrane in the presence of anionic lipid on both surfaces, and on the zirconium octadecylphosphonate surface, binding increases with increasing GM2AP concentration. Detailed results are discussed. Materials and Methods 1 Palmitoyl 2 oleoyl sn glycero 3 p hosphocholine (POPC) and 1 palmitoyl 2 oleoyl sn glycero 3 [phosphor rac (1 glycerol] (POPG) in chloroform were purchased from Avanti Polar Lipids (Alabaster, AL) and used without further purification. Octadecyl mercaptan, 98% was purchased from Aldrich Chemical Company Inc. (Milwaukee, WI). Diiodomethane and zirconyl chloride octahydrate, 98% were purchased from S igma Aldrich (St. Louis, MO). SF10 glass slides (25 mm x 25 mm) were purchased from Schott glass (Duryea, PA) then were subjected to metal deposition of a 4 nm chromium adhesive layer, and a 28.5 nm gold layer (LGA Thin Films, Santa Clara, CA). Unless oth erwise indicated, all reagents and supplies were obtained from Fisher Scientific (Pittsburg, PA) and were used as received. GM2 Activator Protein Expression and Purification Recombinant GM2AP wild type was over expressed in E. coli BL21 (DE3) cells and pur ified as previously described in Chapter 3 and elsewhere 125 A more detailed description of protein purification can be found in Chapter 3. Purity of final protein samples was identified via SDS PAGE using 18% pre casted polyacrylamide gels and Coomassie Brillia nt Blue detection The structural integrity of final protein samples was verified by circular dichroism spectra and comparison to published circular dichroism spectra of GM2AP wildtype 92 Protein protein concentration was determined by Bradford
135 assay and by UV absorption at 280 nm on a Cary 50 Bio UV Visible spectrophotometer (Varian, Palo Alto, CA) using an extinction coefficient of 23000 M 1 cm 1 Purity of fina l protein samples was identified via SDS PAGE. To ensure the protein was properly folded, circular dichroism spectra were obtained to verify secondary structure analogous to published circular dichroism spectra of GM2AP wildtype 92 Protein concentration was measured by Bradford assay and by UV absorbance at 280 nm absorption and an extinction coefficient of 23000 M 1 cm 1 using a Cary 50 Bio UV Vis Spectrophotomete r (Varian, Palo Alto, CA). Gold Slide Preparation Gold slides were initially cleaned in a 3:3:15 v/v solution of ammonium hydroxide : hydrogen peroxide : nanopure water (18.1 M ) for 1 minute at 60 C then were rinsed with water and dried under a flow of nitrogen. The slides were further plasma cleaned for 10 minutes using a Harrick Plasma sterilizer. Slides were exposed to a 1 mM solution of octacecylmercaptan (ODM) in ethanol overnight to form a self asse mbled alkanethiol monolayer, rendering the slides hydrophobic. The slides were rinsed with water and dried under nitrogen before use or further modification. Surface Modification with Zirconium Phosphonate Hydrophobic glass sl ides were modified with zirconium octadecyl phosphonate monolayers using a Langmuir Blodgett (LB) deposition technique. 152 Monolayers were transferred to the gold slides using a KSV 3000 Teflon coated LB trough with hydrophobic barriers (KSV Instruments, Stratford, CT) The aqueous subphase comprised of 2.6 mM calcium chloride at pH 7.6. The surface pressure was measured by Wilhelmy plate filter paper. Concentrations of 0.3 mg/mL of octadecylphosphonic acid (ODPA) in chloroform were spread on the subphase in the trough and the solvent
136 was allowed to evaporate for 10 minutes. The monolayer was compressed linearly at a rate of 8 mN/min at room temperature to a target surface pressure of 20 mN/min. The monolayers were transferred by dipping the s lides at a rate of 8 mN/min through the monolayer film and into a vial that was submerged in the subphase. The vial containing the slide was removed from the trough and a final concentration of 3 mM zirconyl chloride solution was added to the vial in orde r to self assemble Zr 4+ ions at the organic template. The slides were incubated in the ZrOCl 2 solution for 4 days, then were rinsed with and stored in water until used. Lipid Preparation Lipid mixtures of the appropriate ratios in chloroform were prepared and dried under a stream of nitrogen gas. The mixture was further dried overnight for at least 8 hours in a desiccator to remove residual chloroform The dr ied films were rehydrated into appropriate volumes of 50 mM NaOAc, pH 4.8 at room t emperature, to ob tain a final concentration of 1.0 mg/mL. The mixtures were mixed to yield multilamellar vesicles and then subjected to five freeze thaw cylcles using liquid nitrogen Each lipid mixture was extruded by 55 passes through 100 nm polycarbonate filters using an Avanti handheld mini extruder (Avanti Polar Lipids) to produce large unilamellar vesicles (LUVs). Formation of Lipid Supported Layers Alkanethiol phospholipid supported bilayers were formed by adding 1 mg/mL lipid vesicle solution to the ODM monolayered gold slides. Similarly, supported phospholipid bilayers were formed by adding 1 mg/mL lipid vesicles to the zirconium octadecylphosphonate surface. Through adsorption and rupture, the vesicles were
137 allowed to self assemble onto the surface in a flow cell forming either a hybrid bilayer with ODM (Figure 5 1), or a lipid bilayer on zirconated surfaces (Figure 5 2). Surface Plasmon Resonance Enhanced Ellipsometry (SPREE) Measurements Ellipsometry measurements under total internal reflection conditions were p erformed on a commercial EP3 SW imaging system (Nanofilm Surface Analysis, Germany). Briefly, an ellipsometer was coupled with a SPR cell which was mounted on top of a flow cell. The instrument was equipped with a frequency doubled Nd.YAG laser (adjustable power up to 20 mW) at 532 nm. By way of a polarizer and quarter wave plate, elliptically polarized light was directed through a 60 equilateral prism coupl ed to each functionalized gold slide in the Kretschmann configuration. To keep good optical contact between the prism and slide, diiodomethane an index matching fluid was used For maximu m sensitivity, the angle of inc idence (AOI) was set to 64 Buffer lipid vesicles, and GM2AP were inject ed into the flow cell via a peristaltic pump at a flow rate of 50 L/min. The sample cell was kept at a constant temperature of 24 C using a peltier temperature control system linked to the cell. After the addition of vesicles (to form supported lipid layers) and protein (to monitor membrane binding or extraction) Figures 5 3 and 5 4 illustrate how typical SPREE measurements are taken for both types of SLBs. The change in the ellipsometric parameter was recorded as a fu n ctio n of time as each substrate was flowed over the surface. The time step in recording for all experiments was 30 seconds. The minimum detectable signal change of the instrument used in this work was about 10 millidegrees which gives a thickness precision of 0.1 nm. The experimental data for adsorption and desorption of layers on the surface w as fit
138 by the curve fitting analysis prog ram AnalysR (Nanofilm, Germany) to a Langmuir model called the law of mass action where the equations for change in ellipsometric signal ( f) are given by: Baseline equation: (5 1) Adsorption equation: (5 2) Desorption equation: (5 3) Results and Discussion The interaction of GM2AP with SLB s was successfully monitored by SPREE. Lipid vesicles were prepared and adsorbed to the solid supports by ves icle fusion prior to addition of protein. Two types of solid supports were used to mon itor lipid membrane interaction a phospholipid/ODM hybrid b ilayer, and a zirconium oc tadecylphosphonate modified SLB. Both solid supports are said to form stable solid supports with which to study protein lipid interactions. 76,77,79 Reported m onolayer studies of protein lipid interaction show ed that GM2 AP is surface associated with phospholipid bilayers and does not penetrate into the lipid bilayer 1 so the motivation behind using stable, planar SLBs to monitor interactions with GM2AP was based on the hopes of finding a stable system with which to study membrane binding. Interaction of GM2AP with Phospholipid/ODM Hybrid Bilayers Hybrid bilayers were formed by functionalizing gold sl ides with a self assembled ODM monolayer and by adsorbing phospholipid vesicles by vesicle fusion in the flow cell of the instrument. One advantage of using an alkanethiol modified surface to form bilayers is that alkanethiols can form a complete hydropho bic surface on metals, thus providing the driving force for vesicle fusion to form complete bilayers on the solid
139 supports. 76 It was reported that the covalent association between the ou ter phospholipid monolayer with the inner ODM layer is insensitive to changes in buffer, pH, and ionic strength, 79 so any changes in the ellipsometric parameter, psi can be interpreted to be as a result of adsorption of GM2AP. Figure 5 5 shows SPREE sensorgrams of the interaction of GM2AP with POPC/ODM hybrid bilayers. For all experiments, upon addition of GM2AP to the membrane, lipids were extracted from the surface. This result was not surprising though disappointing, as GM2AP is known to be a phospholipid trans fer protein. 153 At pH 4.8, approximately 55% of lipid was extracted from the bilayer (Figure 5 5A), while only about 20% of lipid was extracted at pH 5.5 (Fi gure 5 5B). Consistent with the dansyl based assay described in Chapter 4, increasing the pH of the protein environment changes the degree of interaction with lipids. Additionally, the GM2AP W63W131 variant produced 33% lipid extraction from the surface a decrease in percentage when compared to wildtype protein. The fact that similar results to other assays performed allowed us to be confident with the SPREE results that were obtained. POPG was added to the lipid mixture to investigate how the incorporat ion of anionic lipid would affect protein interaction with the membrane. In vitro studies revealed that in the presence of an anionic lipid, BMP, lipid extraction by GM2AP was enhanced. 83 SPREE results after incorporating different concentrations of POPG to the outer lipid monolayer are shown in Figure 5 6. With increasing POPG concentrations, the rates of lipid adsorption to the functionalized surface decreased. Sim ilar results were observed elsewhere. 77,154 The slower rates of lipid adsorption onto t he surface can be
140 attributed to different stiffness of the vesicles formed from different lipids, leading to thinning of the li pid layers Another interesting observation from incorporating POPG to the outer lipid monolayer was that i nstead of an enhanced extraction of lipid due to the incorporation of anionic lipid protein bound to the membrane. The increase in binding signal w as similar regardless of the concentration of anionic lipid in the membrane. The reason for membrane binding is currently not definitive, though it is believed that electrostatic interactions are playing a role in this phenomenon. Interaction of GM2AP with Zirconi um Octadecylphosphonate modified SLBs A zirconium octadecylphosphonate modified substrate was developed, characterized, and shown to be a n effective and stable solid support for lipid bilayers. 77 Herein, we took adv antage of this stable SLB to investigate the binding of G M2AP to lipid bilayers. Figure 5 7 shows SPREE sensorgrams of the adsorption of GM2AP to POPC/POPG bilayers consisting of varying percentages of POPG on the zirconium octadecylphosphonate modified surfaces. Protein adsorbed to the lipid bilayers but was de sorbed off the bilayer that contained no POPG upon rinsing with buffer (Figure 5 7A) Unlike the observation for 100% POPC lipid layers on the hybrid bilayer, GM2AP was unable to extract lipid from the solid support. For bilayers containing POPG, the chan ge in psi that corresponds with adsorption of protein to the membrane, were similar regardless of percent PG in the membrane. These results suggest that while the zirconium modified surfaces are stable enough to not be extracted from the membrane, some ani onic lipid is needed for protein to remain bound to the bilayer. Z irco nium ions are responsible for the stability of lip id layers formed on the surface and their restricted mobility. 77 Zirconium octadecylphosphonate surfaces are more stable than the hybrid
141 bilayers in this regard though a lipid membrane containing only POPC is not sufficient for GM2AP to bind. O ther studies of protein interactions with SLBs report that membrane binding increases with increasing concentrations of anionic lipids in the membrane though that was not the case in our experiments. F urther experiments on the zirconium modified surfaces, with combinations of other anionic lipids need to be performed to definitively say why addition of POPG induces membrane binding, especially since the amount of protein b ound did not vary with the different con centrations of POPG The addition of POPG to zirconium octadecylphosphonate modified SLBs induce binding of GM2AP to the surface. The system allows for the investigation of GM2AP binding to planar membranes, which ar e more closely related to the cell membrane than lipsomes. Furthermore, because GM2AP also functions as a lipid transfer protein, membrane binding studies have not been successful with lipid vesicles in solution. 83 Different concentrations of GM2AP were adsorbed onto the lipid bilayers Figure 5 8 shows that with increasing protein concentrations, the binding signal increased For 5 uM protein the binding signal was 0.11 and the signal increased to 0.27 for binding of 25 uM GM2AP. Similar results were observed for 50 uM and 100 uM protein with binding signals of 0.45 and 0.60 respectively. These preliminary results are encouraging, showing that membrane bind ing events can be monitored on these surfaces by SPREE. Conclusions In this study, we were able to use planar solid supported lipid bilayers to monitor the interactions of GM2AP with lipid membranes. Because GM2AP functions as a phospholipid transfer prot ein with vesicles in solution these systems provide a means
142 by which membrane binding can be studied using SPREE. The POPC/ODM hybrid bilayer allowed for lipid extraction from the membrane, but upon addition of POPC, and anionic lipid the membrane, GM2AP bound to the membrane. Membrane binding was also observed with zirconium octadecylphosphonate modifies SLBs. While binding signals were similar for protein regardless of POPG concentration with both supports, an increase in signal was observed with increas ing GM2AP concentration on the zirconium modified surfaces.
143 Figure 5 1. A schematic illustration of a hybrid lipid bilayer. An octadecylmercaptan monolayer is functionalized on a gold surface and a hybrid bila yer was formed by ve sicular fusion of phospholipids Figure 5 2. A schematic representation of a zirconium octadecylphosphonate modified surface for the formation of supported lipid bilayers
144 Figu re 5 3. Illustration of the SPREE experime n t al set up showing the layers involved with the adsorption of GM2AP on the octadecyl mercaptan/phospholipid hibrid bilayer. The angle of incidence was set to 64 and psi values were r ecorded in 30 second increment s Figure 5 4. Ill ustration of the SPREE experime n t al set up showing the layers involved with the adsorption of GM2AP on the zirconium octadecylphosphonate supported phospholipid bilayer. The angle of incidence was set to 64 and psi values were r ecorded in 30 second increments
145 Figure 5 5. SPREE sensorgrams showing POPC extraction by 50 uM GM2AP at (A) pH 4.5 and (B) pH 5.5; and by (C) 50 uM GM2AP W63AW131A at pH 4.8 on octadecylmercaptan POPC hybrid bilayers. Lipid concentration was 1 mg/L The an gle of incidence was 64 and measurements were taken at 24 C A B C
146 Figure 5 6. SPREE sensorgrams showing the binding of 50 uM GM2AP to octadecylmercaptan phospholipid hybrid bilayers. The outer lipidla yer consisted of POPC/POPG lipids containing (A) 5%, (B) 10% and (C) 20% POPG. L ipid concentrations we re 1 mg/L The angle of incidence was 64 and measurements were taken at 24 C A B C
147 Figure 5 7. SPREE sensorgrams show ing the binding of 50 uM GM2AP to zirconium octadecylphosphonate modified SLBs. The lipid bilayers consisted of POPC/POPG lipids containing (A) 0%, (B) 5%, (C) 10%, and (D) 20% POPG. L ipid concentrations were 1 mg/L The angle of incidence was 64 and mea surements were taken at 24 C A B C D
148 Figure 5 8. SPREE sensorgrams showing the binding of (A) 5 uM GM2AP to zirconium octadecylphosphonate modified SLBs. The lipid bilayers consisted of POPC/POPG lipids containing 20% POPG. The inset shows theprotein bin ding portion of (B) 25uM, (C) 50 uM, and (D) 100uM protein only. L ipid concentrations were 1 mg/L The angle of incidence was 64 and measurements were taken at 24 C A B C D
149 CHAPTER 6 CONCLUSION S AND FURTHER DIRECTIO NS The site directed mutagenesis and protein expression and purification methods W to A constructs of GM2AP were successfully performed. CD spectroscopy revealed that the amino acid substitutions did not alter proper protein folding. The variants were then used to characterize the intrinsic tryptoph an fluorescence of the protein. Quenching results suggest that while the tryptophan residues were accessible to acrylamide and iodide, the degree of quenching varied. W63 seemed to be most accessible, followed by W131, and W5 was not as accessible due to i ts location in the hydrophobic region of GM2AP. These findings insinuate that since W63 and W131 are located in putative membrane binging regions of the protein, and are solvent exposed, intrinsic tryptophan fluorescence may be a useful technique for the f urther investigation of the precise molecular mechanism of GM2AP function without the need for attaching fluorescent probes to the protein. GM2AP is essential for GM2 degradation in neuronal cells and also functions as a phospholipid transfer protein. As shown here, a dansyl based fluorescence assay to determine changes in lipid transfer, and a gel filtration assay to determine lipid extraction efficiency of a series of W to A GM2AP constructs was used to study the osomes. Removal of tryptophan residues from the mobile loops of GM2AP resulted in slower lipid binding, but did not affect lipid extraction. Binding to lipid vesicles becomes more difficult because the hydrophobic characteristic of tryptophan residues loca ted in the apolar and hydrophobic loops of GM2AP is favored for hydrophobic interactions at the lipid membrane interface. Our results corroborate previous findings that it is likely that GM2AP first binds to the
150 membrane via the hydrophobic loop, with then rotation to the orientation that has the opening to the lipid cleft, which includes the flexible loop containing W131 juxtaposed on the bilayer surface, allowing lipid to be extracted into the binding pocket of the protein. Additionally, while GM2AP also functions as a lipid transfer protein, an acidic environment (near pH 4.8) seems to be required for optimal protein function. We were able to use planar solid supported lipid bilayers to monitor the interactions of GM2AP with lipid membranes. Because GM2AP functions as a phospholipid transfer protein with vesicles in solution, these systems provide a means by which membrane binding can be studied using SPREE. The POPC/ODM hybrid bilayer allowed for lipid extraction from the membrane, but upon addition of PO PC, and anionic lipid the membrane, GM2AP bound to the membrane. Membrane binding was also observed with zirconium octadecylphosphonate modifies SLBs. While binding signals were similar for protein regardless of POPG concentration with both supports, an in crease in signal was observed with increasing GM2AP concentration on the zirconium modified surfaces. Other experiments should be performed, namely varying lipid compositions, and ionic strengths of the system, to confirm the suitability of stable support ed lipid bilayers to studying membrane binding by GM2AP. Once confirmed, molecular interactions between GM2AP and lipid membranes can be investigated by mutating amino acid residues in the protein to probe regions of the protein involved in membrane bindin g.
151 LIST OF REFERENCES (1) Giehl, A.; Lemm, T.; Bartelsen, O.; Sandhoff, K.; Blume, A. European Journal of Biochemistry 1999 261 650 658. (2) Nagai, Y.; Iwamori, M. In Biology of the sialic acids Rosenberg, A., Ed.; Plenum Press: New York, 1995, p 197 241. (3) Gravel, R. A.; Clarke, J. T. R.; Kaback, M. M.; Mahuran, D.; Sandhoff, K.; Suzuki, K. In The Metabolic and molecular bases of inherited disease 7th ed.; Scriver, C. R., Ed.; McGraw Hill, Health Professions Division: New York, 199 5, p 2839 2879. (4) Hakomori, S. Curr Opin Hematol 2003 10 16 24. (5) Kielczynski, W.; Harrison, L. C.; Leedman, P. J. Proceedings of the National Academy of Sciences 1991 88 1991 1995. (6) Thompson, T. E.; Tillack, T. W. Annu Rev Biophys Bio 1985 14 361 386. (7) Kolter, T.; Sandhoff, K. Febs Letters 2010 584 1700 1712. (8) Tettamanti, G.; Bassi, R.; Viani, P.; Riboni, L. Biochimie 2003 85 423 437. (9) Sandhoff, K.; Kolter, T. Philos T Roy Soc B 2003 358 847 861. (10) Schulze, H.; Kolter, T.; Sa ndhoff, K. Biochimica et Biophysica Acta (BBA) Molecular Cell Research 2009 1793 674 683. (11) Sastry, P. S. Prog Lipid Res 1985 24 69 176. (12) Wilkening, G.; Linke, T.; Uhlhorn Dierks, G.; Sandhoff, K. J Biol Chem 2000 275 35814 9. (13) Schuette, C. G.; Pierstorff, B.; Huettler, S.; Sandhoff, K. Glycobiology 2001 11 81R 90R. (14) Fingerhut, R.; van der Horst, G. T.; Verheijen, F. W.; Conzelmann, E. European Journal of Biochemistry 1992 208 623 9. (15) Zschoche, A.; Furst, W.; Schwarzmann, G.; Sanhoff, K. European Journal of Biochemistry 1994 222 83 90. (16) Linke, T.; Wilkening, G.; Sadeghlar, F.; Mozcall, H.; Bernardo, K.; Schuchman, E.; Sandhoff, K. J Biol Chem 2001 276 5760 8. (17) Meier, E. M.; Schwarzmann, G.; Furst, W.; Sandhoff, K. J ournal of Biological Chemistry 1991 266 1879 1887.
152 (18) Kaback, M. M. Prog Clin Biol Res 1977 18 1 7. (19) Suzuki, K. In Biology of the Sialic Acids ; Rosenberg, A., Ed.; Plenum Press: New York, 1995, p 337 361. (20) Myerowitz, R. Hum Mutat 1997 9 195 208. (21) Barranger, J. A.; Cabrera Salazar, M. A.; Charria Ortiz, G.; Springer US: 2007, p 229 256. (22) Tay, W. Trans. Opthalmol. Soc. UK 1881 1 3. (23) Sachs, B. J. Nerv. Ment. Dis. 1887 14 13. (24) Sandhoff, K.; Andreae, U.; Jatzkewitz, H. Pathol Eur 1968 3 278 85. (25) Mahuran, D. J. Bba Lipid Lipid Met 1998 1393 1 18. (26) Sandhoff, K. Febs Letters 1969 4 351 354. (27) Werth, N.; Schuette, C. G.; Wilkening, G.; Lemm, T.; Sandhoff, K. J Biol Chem 2001 276 12685 12690. (28) Kytzia, H. J.; S andhoff, K. J Biol Chem 1985 260 7568 72. (29) Hepbildikler, S. T.; Sandhoff, R.; Kolzer, M.; Proia, R. L.; Sandhoff, K. Journal of Biological Chemistry 2002 277 2562 72. (30) Yadao, F.; Hechtman, P.; Kaplan, F. Biochim Biophys Acta 1997 1340 45 52. (31) Mehl, E.; Jatzkewitz, H. Hoppe Seylers Z Physiol Chem 1964 339 260 76. (32) Furst, W.; Machleidt, W.; Sandhoff, K. Biol Chem Hoppe Seyler 1988 369 317 28. (33) Hiraiwa, M.; O'Brien, J. S.; Kishimoto, Y.; Galdzicka, M.; Fluharty, A. L.; Ginns, E. I .; Martin, B. M. Arch Biochem Biophys 1993 304 110 6. (34) Vielhaber, G.; Hurwitz, R.; Sandhoff, K. J Biol Chem 1996 271 32438 46. (35) Li, S. C.; Wu, Y. Y.; Sugiyama, E.; Taki, T.; Kasama, T.; Casellato, R.; Sonnino, S.; Li, Y. T. J Biol Chem 1995 27 0 24246 51. (36) Klima, H.; Klein, A.; van Echten, G.; Schwarzmann, G.; Suzuki, K.; Sandhoff, K. Biochemical Journal 1993 292 ( Pt 2) 571 6.
153 (37) Klima, H.; Tanaka, A.; Schnabel, D.; Nakano, T.; Schroder, M.; Suzuki, K.; Sandhoff, K. Febs Lett 1991 289 260 4. (38) Ahn, V. E.; Leyko, P.; Alattia, J. R.; Chen, L.; Prive, G. G. Protein Science 2006 15 1849 57. (39) Ahn, V. E.; Faull, K. F.; Whitelegge, J. P.; Fluharty, A. L.; Prive, G. G. Proc Natl Acad Sci U S A 2003 100 38 43. (40) Popovic, K.; Priv e, G. G. Acta Crystallogr D Biol Crystallogr 2008 64 589 94. (41) Kishimoto, Y.; Hiraiwa, M.; O'Brien, J. S. Journal of Lipid Research 1992 33 1255 67. (42) Vaccaro, A. M.; Salvioli, R.; Barca, A.; Tatti, M.; Ciaffoni, F.; Maras, B.; Siciliano, R.; Zap pacosta, F.; Amoresano, A.; Pucci, P. J Biol Chem 1995 270 9953 60. (43) Vaccaro, A. M.; Ciaffoni, F.; Tatti, M.; Salvioli, R.; Barca, A.; Tognozzi, D.; Scerch, C. J Biol Chem 1995 270 30576 80. (44) Linke, T.; Wilkening, G.; Lansmann, S.; Moczall, H.; Bartelsen, O.; Weisgerber, J.; Sandhoff, K. Biol Chem 2001 382 283 90. (45) Bar, J.; Linke, T.; Ferlinz, K.; Neumann, U.; Schuchman, E. H.; Sandhoff, K. Hum Mutat 2001 17 199 209. (46) Christomanou, H.; Aignesberger, A.; Linke, R. P. Biol Chem Hoppe S eyler 1986 367 879 90. (47) Klein, A.; Henseler, M.; Klein, C.; Suzuki, K.; Harzer, K.; Sandhoff, K. Biochem Biophys Res Commun 1994 200 1440 8. (48) Ciaffoni, F.; Salvioli, R.; Tatti, M.; Arancia, G.; Crateri, P.; Vaccaro, A. M. Journal of Biological Chemistry 2001 276 31583 9. (49) Conzelmann, E.; Burg, J.; Stephan, G.; Sandhoff, K. European Journal of Biochemistry 1982 123 455 464. (50) Smiljanic Georgijev, N.; Rigat, B.; Xie, B.; Wang, W.; Mahuran, D. J. Biochim Biophys Acta 1997 1339 192 202. (51) Li, S. C.; Hama, Y.; Li, Y. T. Methods Enzymol 2003 363 230 41. (52) Ravasi, D.; Masserini, M.; Vecchio, G.; Li, Y. T.; Li, S. C. Neurochem Res 2002 27 785 792. (53) Wright, C. S.; Li, S. C.; Rastinejad, F. J Mol Biol 2000 304 411 422.
154 (54) Alb iach Marti, M. R.; Robertson, C.; Gowda, S.; Tatineni, S.; Belliure, B.; Garnsey, S. M.; Folimonova, S. Y.; Moreno, P.; Dawson, W. O. Molecular plant pathology 2010 11 55 67. (55) Xie, B.; Mcinnes, B.; Neote, K.; Lamhonwah, A. M.; Mahuran, D. Biochem Bio ph Res Co 1991 177 1217 1223. (56) Wendeler, M.; Hoernschemeyer, J.; Hoffmann, D.; Kolter, T.; Schwarzmann, G.; Sandhoff, K. European Journal of Biochemistry 2004 271 614 627. (57) Wright, C. S.; Zhao, Q.; Rastinejad, F. J Mol Biol 2003 331 951 964. (58) Wright, C. S.; Li, S. C.; Rastinejad, F. J Mol Biol 2000 304 411 22. (59) Furst, W.; Sandhoff, K. Biochim Biophys Acta 1992 1126 1 16. (60) Wright, C. S.; Mi, L. Z.; Lee, S.; Rastinejad, F. Biochemistry Us 2005 44 13510 21. (61) Wendeler, M.; Ho ernschemeyer, J.; John, M.; Werth, N.; Schoeniger, M.; Lemm, T.; Hartmann, R.; Kessler, H.; Sandhoff, K. Protein Expr Purif 2004 34 147 57. (62) Nakamura, S.; Akisue, T.; Jinnai, H.; Hitomi, T.; Sarkar, S.; Miwa, N.; Okada, T.; Yoshida, K.; Kuroda, S.; K ikkawa, U.; Nishizuka, Y. Proc Natl Acad Sci U S A 1998 95 12249 53. (63) Jinnai, H.; Nakamura, S. Kobe J Med Sci 1999 45 181 90. (64) Mundel, T. M. Annals of Anatomy Anatomischer Anzeiger 2001 183 11 11. (65) Mundel, T. M.; Heid, H. W.; Mahuran, D. J.; Kriz, W.; Mundel, P. Journal of the American Society of Nephrology 1999 10 435 443. (66) Rigat, B.; Reynaud, D.; Smiljanic Georgijev, N.; Mahuran, D. Biochem Bioph Res Co 1999 258 256 259. (67) Lehninger, A. L.; Nelson, D. L.; Cox, M. M. Lehninger principles of biochemistry ; 4th ed.; W.H. Freeman: New York, 2005. (68) Gennis, R. B. Biomembranes : molecular structure and function ; Springer Verlag: New York, 1989. (69) van Meer, G.; Voelker, D. R.; Feigenson, G. W. Nat Rev Mol Cell Biol 2008 9 112 2 4.
155 (70) Yeagle, P. The membranes of cells ; 2nd ed.; Academic Press: San Diego, 1993. (71) Edwards, K. A.; Baeumner, A. J. Analytical and Bioanalytical Chemistry 2006 386 1613 23. (72) Foldvari, M.; Faulkner, G. T.; Mezei, C.; Mezei, M. Cell Mater 1992 2 67 85. (73) Foldvari, M.; Gesztes, A.; Mezei, M. J Microencapsul 1990 7 479 489. (74) Tamm, L. K.; McConnell, H. M. Biophysical Journal 1985 47 105 13. (75) Castellana, E. T.; Cremer, P. S. Surface Science Reports 2006 61 429 444. (76) Plant, A. L. Langmuir 1999 15 5128 5135. (77) Fabre, R. M.; Talham, D. R. Langmuir 2009 25 12644 52. (78) Brian, A. A.; McConnell, H. M. Proc Natl Acad Sci U S A 1984 81 6159 63. (79) Plant, A. L. Langmuir 1993 9 2764 2767. (80) Plant, A. L.; Gueguetchkeri, M. ; Yap, W. Biophys J 1994 67 1126 33. (81) Meuse, C. W.; Krueger, S.; Majkrzak, C. F.; Dura, J. A.; Fu, J.; Connor, J. T.; Plant, A. L. Biophysical Journal 1998 74 1388 98. (82) Schwarzmann, G.; Wendeler, M.; Sandhoff, K. Glycobiology 2005 15 1302 131 1. (83) Ran, Y.; Fanucci, G. E. Biophysical Journal 2009 97 257 66. (84) Kuwana, T.; Mullock, B. M.; Luzio, J. P. Biochem J 1995 308 ( Pt 3) 937 46. (85) Morimoto, S.; Kishimoto, Y.; Tomich, J.; Weiler, S.; Ohashi, T.; Barranger, J. A.; Kretz, K. A.; O 'Brien, J. S. J Biol Chem 1990 265 1933 7. (86) Voet, D.; Voet, J. G. Biochemistry ; J. Wiley & Sons: Hoboken, NJ, 2004. (87) Johnson, W. C., Jr. Annu Rev Biophys Biophys Chem 1988 17 145 66. (88) Kelly, S. M.; Jess, T. J.; Price, N. C. Biochim Biophys Acta 2005 1751 119 39.
156 (89) Barrett, G. C.; Elmore, D. T. Amino acids and peptides / G.C. Barrett and D.T. Elmore ; Cambridge University Press: Cambridge, U.K. ; New York, 1998. (90) Tompkins, H. G.; McGahan, W. A. Spectroscopic ellipsometry and reflectom etry : a user's guide / by Harland G. Tompkins, William A. McGahan ; Wiley: New York, 1999. (91) Greenfield, N. J. Nat. Protocols 2007 1 2876 2890. (92) Wendeler, M.; Werth, N.; Maier, T.; Schwarzmann, G.; Kolter, T.; Schoeniger, M.; Hoffmann, D.; Lemm, T .; Saenger, W.; Sandhoff, K. Febs Journal 2006 273 982 991. (93) Stryer, L. Science 1968 162 526 533. (94) Lakowicz, J. R. Principles of fluorescence spectroscopy / Joseph R. Lakowicz ; 3rd ed. ed.; Springer: New York, 2006. (95) Schulman, S. G.; Sharma A. Introduction to fluorescence spectroscopy / Stephen G. Schulman and Ashutosh Sharma ; John Wiley & Sons, Inc: New York, 1999. (96) Demchenko, A. P. Ultraviolet spectroscopy of proteins / Alexander P. Demchenko ; Rev. and enl. translation of the Rus sian ed. ed.; Springer Verlag: Berlin ; New York, 1986. (97) Waggoner, A. S.; Stryer, L. Proc Natl Acad Sci U S A 1970 67 579 89. (98) Eftink, M. R. Methods Biochem Anal 1991 35 127 205. (99) Eftink, M. R.; Ghiron, C. A. Anal Biochem 1981 114 199 227 (100) Lehrer, S. Biochemistry Us 1971 10 3254 3263. (101) Eftink, M. R.; Ghiron, C. A. Biochemistry Us 1976 15 672 680. (102) Eftink, M.; Lakowicz, J., Ed.; Springer US: 2002; Vol. 2, p 53 126. (103) Skoog, D. A.; Crouch, S. R.; Holler, F. J. Princip les of instrumental analysis ; 6th ed. / Douglas A. Skoog, F. James Holler, Stanley R. Crouch. ed.; Brooks/Cole: Belmont, CA, 2007. (104) Chowdhury, M. H.; Aslan, K.; Malyn, S. N.; Lakowicz, J. R.; Geddes, C. D. J Fluoresc 2006 16 295 9. (105) Westphal, P .; Bornmann, A. Sensors and Actuators B Chemical 2002 84 278 282.
157 (106) Homola, J.; Yee, S. S.; Gauglitz, G. Sensor Actuat B Chem 1999 54 3 15. (107) Azzam, R. M. A.; Bashara, N. M. Ellipsometry and polarized light ; North Holland ; sole distributors fo r the USA and Canada, Elsevier Science Publishing Co., Inc.: Amsterdam ; New York, 1987. (108) Arwin, H.; Poksinski, M.; Johansen, K. Applied Optics 2004 43 3028 3036. (109) Poksinski, M.; Arwin, H. Optics Letters 2007 32 1308 1310. (110) Tompkins, H. G. A user's guide to ellipsometry / Harland G. Tompkins ; Academic Press: Boston, 1993. (111) Born, M.; Wolf, E. j. a. Principles of optics : electromagnetic theory of propagation, interference and diffraction of light / by Max Born and Emil Wolf, with cont ributions by A. B. Bhatia ... [et al.] ; 6th ed. ed.; Pergamon Press: Oxford ; New York, 1980. (112) Salamon, Z.; Tollin, G. Spectrosc Int J 2001 15 161 175. (113) Salamon, Z.; Macleod, H. A.; Tollin, G. Bba Rev Biomembranes 1997 1331 131 152. (114) Coo per, M. Analytical and Bioanalytical Chemistry 2003 377 834 842. (115) Iwata, T.; Maeda, S. Applied Optics 2007 46 1575 1582. (116) Conzelmann, E.; Sandhoff, K. Proc Natl Acad Sci U S A 1978 75 3979 83. (117) Burg, J.; Banerjee, A.; Sandhoff, K. Biol Chem H S 1985 366 887 891. (118) Zhou, D.; Cantu, C., 3rd; Sagiv, Y.; Schrantz, N.; Kulkarni, A. B.; Qi, X.; Mahuran, D. J.; Morales, C. R.; Grabowski, G. A.; Benlagha, K.; Savage, P.; Bendelac, A.; Teyton, L. Science 2004 303 523 7. (119) Mundel, T. M. Ann Anat 2001 183 11. (120) Mundel, T. M.; Heid, H. W.; Mahuran, D. J.; Kriz, W.; Mundel, P. J Am Soc Nephrol 1999 10 435 43. (121) Conzelmann, E.; Burg, J.; Stephan, G.; Sandhoff, K. Eur J Biochem 1982 123 455 64. (122) Wendeler, M.; Hoernschemey er, J.; John, M.; Werth, N.; Schoeniger, M.; Lemm, T.; Hartmann, R.; Kessler, H.; Sandhoff, K. Protein Expres Purif 2004 34 147 157.
158 (123) Wendeler, M.; Lemm, T.; Weisgerber, J.; Hoernschemeyer, J.; Bartelsen, O.; Schevers, U.; Sandhoff, K. Protein Expre s Purif 2003 27 259 266. (124) Wu, Y. Y.; Lockyer, J. M.; Sugiyama, E.; Pavlova, N. V.; Li, Y. T.; Li, S. C. J Biol Chem 1994 269 16276 83. (125) Ran, Y.; Fanucci, G. E. Biophys J 2009 97 257 266. (126) Ladokhin, A. S.; Jayasinghe, S.; White, S. H. A nal Biochem 2000 285 235 245. (127) Li, X. M.; Malakhova, M. L.; Lin, X.; Pike, H. M.; Chung, T. W.; Molotkovsky, J. G.; Brown, R. E. Biochemistry Us 2004 43 10285 10294. (128) Weiner, M. P.; Costa, G. L.; Schoettlin, W.; Cline, J.; Mathur, E.; Bauer, J. C. Gene 1994 151 119 23. (129) Zheng, L.; Baumann, U.; Reymond, J. L. Nucleic Acids Research 2004 32 e115. (130) Burstein, E. A.; Vedenkina, N. S.; Ivkova, M. N. Photochem Photobiol 1973 18 263 79. (131) Eftink, M. R.; Ghiron, C. A. Biochemistry U s 1976 15 672 80. (132) Wendeler, M.; Hoernschemeyer, J.; Hoffmann, D.; Kolter, T.; Schwarzmann, G.; Sandhoff, K. Eur J Biochem 2004 271 614 27. (133) Wendeler, M.; Lemm, T.; Weisgerber, J.; Hoernschemeyer, J.; Bartelsen, O.; Schepers, U.; Sandhoff, K. Protein Expr Purif 2003 27 259 66. (134) Ran, Y.; Fanucci, G. E. Anal Biochem 2008 382 132 4. (135) Rao, C. S.; Chung, T.; Pike, H. M.; Brown, R. E. Biophysical Journal 2005 89 4017 28. (136) Svennerholm, L. Biochim Biophys Acta 1957 24 604 11. (1 37) Wright, C. S.; Zhao, Q.; Rastinejad, F. J Mol Biol 2003 331 951 64. (138) Mathias, J. D.; Ran, Y.; Carter, J. D.; Fanucci, G. E. Biophysical Journal 2009 97 1436 1444. (139) Wimley, W. C.; White, S. H. Nat Struct Biol 1996 3 842 8. (140) Wendeler M.; Reilaender, H.; Hoernschemeyer, J.; Schwarzmann, G.; Kolter, T.; Sandhoff, K. Methods Enzymol 2003 363 476 89.
159 (141) Hama, Y.; Li, Y. T.; Li, S. C. Journal of Biological Chemistry 1997 272 2828 33. (142) Ballester, A.; Cervera, M.; Pena, L. Plant cell reports 2008 27 1005 15. (143) Langmuir, I. T Faraday Soc 1920 15 0062 0074. (144) Blodgett, K. B. J Am Chem Soc 1934 56 495 495. (145) Kalb, E.; Frey, S.; Tamm, L. Faseb Journal 1992 6 A290 A290. (146) Qi, S. Y.; Groves, J. T.; Chakraborty, A. K. P Natl Acad Sci USA 2001 98 6548 6553. (147) Tendian, S. W.; Lentz, B. R.; Thompson, N. L. Biochemistry Us 1991 30 10991 10999. (148) Yang, T. L.; Baryshnikova, O. K.; Mao, H. B.; Holden, M. A.; Cremer, P. S. J Am Chem Soc 2003 125 4779 4784. ( 149) Ramsden, J. J.; Schneider, P. Biochemistry Us 1993 32 523 529. (150) Kalb, E.; Engel, J.; Tamm, L. K. Biochemistry Us 1990 29 1607 1613. (151) Luth, D.; Moore, G. A. Plant Cell, Tissue and Organ Culture 1999 57 219 222. (152) Byrd, H.; Pike, J. K.; Talham, D. R. Chem Mater 1993 5 709 715. (153) Mahuran, D. J. Biochim Biophys Acta 1998 1393 1 18. (154) Pandey, A. P.; Haque, F.; Rochet, J. C.; Hovis, J. S. Biophysical Journal 2009 96 540 551.
160 BIOGRAPHICAL SKETCH Stacey Ann Benjamin was b orn in Linstead, St Catherine on the island of Jamaica Catholic High School for girls in Kingston where she graduated in 2000. Stacey opted to return to high school for an additional two years at Hampton School for girls in St. Elizabeth where she took advanced level classes and sat the University of Cambridge A level exams in c hem istry, mathematics and general p aper. After graduating high school in 2002, Stacey Ann left Jamaica for undergraduate studies at Wesleyan College in Macon, GA She participated in two Summer Research programs in 2004 and 2005 at the Florida State University where she was introduced to, and developed an interes t in c hemistry research. During her senior year, Stacey Ann did a co op at J.M. Huber Corporation in Macon, GA where she assisted in the design of standard procedures for determining the surface area and pore size of kaoilin clay samples. In 2006, Stacey Ann receive d a Bachelor of Arts degr ee in c hemistry with a minor in Mathematics from Wesleyan College. Stacey chemistry program in the fall of 2006 and soon after she joined the Fanucci research group. After a few yea rs of being a teaching assistant for general chemistry courses, she developed a passion for teaching and decided that she wanted to commit her academic career primarily to undergraduate chemistry education. Stacey Ann received a Masters of Science in teach ing in May 2011 and a Ph.D. in c hemistry in August 2012 from UF.