UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Equine plasma gelsolin : from sequence to structure Koepf, Edward Kurt 1996

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1996-147754.pdf [ 9.25MB ]
JSON: 831-1.0059486.json
JSON-LD: 831-1.0059486-ld.json
RDF/XML (Pretty): 831-1.0059486-rdf.xml
RDF/JSON: 831-1.0059486-rdf.json
Turtle: 831-1.0059486-turtle.txt
N-Triples: 831-1.0059486-rdf-ntriples.txt
Original Record: 831-1.0059486-source.json
Full Text

Full Text

EQUINE PLASMA GELSOLIN: FROM SEQUENCE TO STRUCTURE by EDWARD KURT KOEPF B.Sc, The University of Victoria, 1990 M.Sc, The University of British Columbia, 1992 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Department of Chemistry) We accept this thesis as confoiroing to the required standard THE UNIVERSITY OF BRITISH COLUMBIA SEPTEMBER 1996 © Edward Kurt Koepf, 1996 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. The University of British Columbia Vancouver, Canada DE-6 (2/88) Abstract Equine gelsolin, a protein involved i n regulat ing l inear ac t in assembly, has been analyzed at various hierarchical levels of protein organization. A t the genetic level, the coding message for the intracellular cytoplasmic form of gelsolin consists of 2196 nucleotide bases, w h i c h translates into a 731 amino acid protein with a molecular mass of 80 696 daltons. The amino acid sequence derived from the cloned D N A for the .equine protein shares 94.7% identity with that of cytoplasmic h u m a n gelsolin. The exon that codes for the stretch of amino acids found at the N - t e r m i n u s of horse cytoplasmic gelsolin is ident ica l i n size (206 nucleotides) and similar i n composition to h u m a n gelsolin exon 4. The form of horse gelsolin that is secreted into blood plasma is distinguished from the cytoplasmic protein by an extension of 25 amino acids at the N -terminus. Spectroscopic examination of the solution structure of the C a 2 + - f r e e form of horse plasma gelsolin estimates the a-helix and (3-sheet contents at 16% and 23%, respectively. Similar analysis with C a 2 + - b o u n d gelsolin yields an tx-helix content of 17% and 22% P-structure. The predicted conformational weights of C a 2 + - f r e e gelsolin agree favorably wi th values derived from the crystal structure of the protein, which shows 18% a -hel ix and 23% P-structure. Moni tor ing the loss of these secondary s tructural elements during the course of unfolding experiments reveals that the behaviors of both native and chemically modified horse plasma gelsolin towards chemical denaturants depend greatly on the surface charge of the protein. Interaction with C a 2 + and guanidinium cations alters the charge on gelsolin's surface, which not only affects protein structure, i i i but modifies the pathway of unfolding. The path to the denatured state does not follow a simple two-state cooperative mechanism, but likely passes through various intermediate conformations which reflect the multi-domain configuration of gelsolin. Whole horse plasma gelsolin purified in the course of this project was used to grow crystals suitable for x-ray diffraction analysis. The 3-dimensional structure (Burtnick, L.D., Robinson, R.C., and Koepf, E.K., 1996, Biophysical Journal, 70, A14), which confirms the existence of six similar domains, S1-S6, was solved in the absence of bound C a 2 + . Superposition of Ca 2 + -free horse SI onto Ca 2 + -bound human SI produces an almost exact overlay, suggesting that C a 2 + does not exert its regulatory activity on gelsolin by altering the folding within an individual domain in a significant way. Modeling studies with gelsolin and F-actin provide a means of relating the structure of whole gelsolin to its biological activities. Docking of whole horse gelsolin with a model for the human SI-F-actin complex demonstrates how both capping and nucleating activities may come about, but does not adequately explain either the severing function or why Ca 2 + - free gelsolin does not bind to F-actin. A model in which C a 2 + binding induces shifts in the relative positions of the six domains would seem to be required. Table of Contents Abstract u TabteofOantents iv List of Tables viii ^ of Figures ix Abbreviations xiii Acknowledgpaenks xvi CHAPTER I Introduction 1. ACTIN, ACTIN BINDING PROTEINS, AND GELSOLIN 1 I Project Overview 1 ii. Actin 2 iii. Monomelic G-actin 2 iv. Polymeric F-actin 3 u Converting G to F-actin 6 vt Non-muscle Actin Binding Proteins 7 vii. Gelsolin 10 viii. Calcium Regulated Interactions with Actin 11 ix. Functional Role of Cytoplasmic Gelsolin 12 x. Gelsolin in Plasma 13 xi. Localization of Actin Binding Sites 14 xii. Gene Structure of Gelsolin 16 xiii. Crystallization of Gelsolin 19 xiv. Structural Comparison of Acting Binding Proteins 22 2. RECOMBINANT DNA METHODOLOGY 25 I Methods of Gene Cloning. 25 it. Plasmids and Phages 26 V iii. Cloning Vectors: pBluescript II KS+/- and XEMBL3 29 iv. Standard Enzymes Utilized in Molecular Biology 31 v. Polymerase Chain Reaction 33 vL Cloning DNA from Libraries or with RT-PCR 35 vii. Screening by Nucleic Acid Hybridization 38 viii. Sequencing with the Chain-Termination Method 39 ix. PCR Techniques Designed to Generate cDNA Ends 40 3. CIRCULAR DICHROISM SPECTROSCOPY 46 L Principles of Circular Dichroism 46 ii. Circular Dichroism Spectroscopy of Proteins 48 iii. Prediction of Secondary Structure 49 iv. Convex Constraint Analysis 50 4. DENATURATION STUDIES 52 I Protein Folding 52 ii. Equilibrium Unfolding Curves 53 5. PROTEIN CRYSTALLOGRAPHY 54 L Producing Protein Crystals 54 ii. Solving a Protein Crystal Structure 57 CHAPTER n Materials and Methods 1. RECOMBINANT DNA METHODOLOGY 59 I Bacterial Strains, Vectors, and Media 59 ii. Electrophoresis 60 iii. Oligodeoxyribonucleotide Design and Synthesis 61 iv. Extraction of Genomic DNA 61 v. Genomic Library Construction/Phage Packaging 63 vi. Library Titering 63 vii. Plaque Lifts/DNA Blotting 64 viii. Preparation of Radiolabeled Hybridization Probes 65 v i ix. Library Screening 66 x. RNA Extraction 67 xi. First Strand cDNA Synthesis 69 xii. Polymerase Chain Reaction 69 xiii. cDNA Cloning/Plasmid Transformations 70 xiv. Isolation of Double Stranded Plasmid DNA 72 xv. Isolation of Single Stranded DNA 72 xvi. Isolation of Recombinant Lambda Phage DNA 73 xvil Endonuclease Restriction Analysis 74 xviii. DNA Sequencing 75 xix. N-terminal Protein Sequencing 75 2. PROTEIN PREPARATIONS 76 I Purification of Actin 76 ii. Purification of Gelsolin 77 iii. Gelsolin Purity and Severing Activity 78 iv. Preparation of FITC-labeled.Gelsolin , 79 3. OPTICAL TECHNIQUES 82 I Absorbance measurements 82 ii. Fluorescence Measurements 82 iii. Circular Dichroism 82 iv. Convex Constraint Analysis 83 4. DENATURATION STUDIES 84 L Chemical Denaturations 84 ii. Thermal Denaturations 84 5. CRYSTALLIZATION TRIALS I Crystallization of calcium free gelsolin ii. Crystallization of calcium loaded gelsolin 86 86 87 CHAPTER LU Results and Discussion 1. CLONING AND SEQUENCE ANALYSIS L G e l s o l i n C l o n i n g S t r a t e g y ii. c D N A C l o n e s o f H o r s e C y t o p l a s m i c G e l s o l i n iii. I n S e a r c h o f t h e 5 ' P l a s m a E x t e n s i o n iv. I s o l a t i o n o f H o r s e G e l s o l i n G e n o m i c C l o n e s v. I d e n t i f i c a t i o n o f t h e 5 ' - e n d i n G e n o m i c C l o n e s vL c D N A S e q u e n c e o f C y t o p l a s m i c H o r s e G e l s o l i n vii. P l a s m a G e l s o l i n N - t e r m i n a l A m i n o A c i d S e q u e n c e viii. A m i n o A c i d S e q u e n c e o f H o r s e P l a s m a G e l s o l i n ix. P a r t i a l E x o n O r g a n i z a t i o n o f P l a s m a G e l s o l i n 2. CD AND SECONDARY STRUCTURE ANALYSIS i . C h a r a c t e r i z a t i o n b y C D ii. S e c o n d a r y S t r u c t u r e A n a l y s i s 3. EQUILIBRIUM UNFOLDING STUDIES I C h e m i c a l D e n a t u r a t i o n w i t h G u a n i d i n e - H C l ii. C h e m i c a l D e n a t u r a t i o n w i t h U r e a iii. T h e r m a l D e n a t u r a t i o n S t u d i e s 4. CRYSTALLIZATION TRIALS I C r y s t a l s o f H o r s e P l a s m a G e l s o l i n 5. SUMMARY i . S u m m a r y o f R e s u l t s i i . S u g g e s t i o n s f o r F u t u r e S t u d i e s REFERENCES APPENDIX I i . H i g h R e s o l u t i o n S t r u c t u r e o f P l a s m a G e l s o l i n i i . S t r u c t u r e - F u n c t i o n R e l a t i o n s h i p o f G e l s o l i n v i i i List of Tables Table Page 1. Synthetic oligodeoxyribonucleotide primers used in generating 62 the sequence of horse plasma gelsolin. 2. Comparison of horse cytoplasmic gelsolin cDNA to the human, 101 pig, and mouse sequences. ix List of Figures Figure Page 1. High resolution structure of G-actin. 4 2. Helical arrangement of the two polymer strands in an actin 5 filament. 3. Schematic diagram of actin binding proteins and their 8 interactions with monomelic and polymeric forms of actin. 4. Comparison of human gelsolin proteolytic fragments to the 15 gelsolin domain structure (S1-S6) derived from porcine cDNA. 5. Exon arrangement of the 5'-end of cytoplasmic and plasma 19 gelsolin cDNA. 6. Structure of recombinant human S1 solved in a complex with 21 rabbit skeletal muscle actin. 7. NMR solution structure of villin 14T. 23 8. NMR solution structure of severin domain 2. 24 9. Schematics of the phagemid vectors pBluescript II KS+/-. 30 10. >»EMBL3 high capacity replacement vector used in construction 31 of genomic libraries. 11. Outline of DNA amplification with the polymerase chain 34 reaction. 12. Construction of a genomic library with a phage replacement 37 vector. X 13. Schematic representation of the 5'-RACE system. 42 14. Illustration of the SLIC-PCR method used in 5'-end cloning. 43 15. Schematic representation of inverse PCR. 45 16. Illustration of the origins of the ellipticity angle OP) and 46 rotation angle (a) for an originally plane polarized light beam that has emerged from an optically active absorbing medium. 17. Illustration of hanging and sitting drop experiments used 56 to survey protein crystallization conditions. 18. Xho I and Xba I restriction digests of three cDNA and one 90 genomic clone that together code for horse cytoplasmic gelsolin. 19. Schematic representation of the 4 overlapping clones that 91 produced the sequence for horse cytoplasmic gelsolin. 20. Plaque purification of 1 of the 6 AJEMBL3 genomic clones 95 after three consecutive rounds of screening. 21. 5'-end characterization of ?iEMBL3 genomic clones with a 95 mixture of 32p_ e n c i labeled gelsolin #7 and #8 oligodeoxy-ribonucleotide primers. 22. Complete cDNA sequence for the cytoplasmic isoform of horse 100 gelsolin. 23. Sequence comparison of the N-terminal plasma extension 102 for horse, human and pig plasma gelsolins. 24. Alignment of horse and human plasma gelsolin protein 105 sequences. 25. Exon-intron arrangement of the portion of the horse gelsolin 108 gene that codes for residues Pro22 to Leu89 in plasma gelsolin. 26. Far UV CD spectrum of horse plasma gelsolin in a C a 2 + free 109 and C a 2 + saturated solution. 27. Near UV CD spectrum of horse plasma gelsolin in both a C a 2 + 110 free and C a 2 + saturated environment. 28. Reconstruction of the far UV CD spectrum of horse plasma 111 gelsolin with the convex constraint algorithm. 29. Denaturation of native and FITC-labeled gelsolin with Gnd-HCl 114 in C a 2 + free and C a 2 + saturated conditions . Denaturations were monitored at 215 nm. 30. Denaturation of FITC-labeled gelsolin with Gnd-HCl in the 116 presence of C a 2 + , monitored at 276 nm. 31. Denaturation of unlabeled gelsolin with Gnd-HCl in both C a 2 + 117 free and C a 2 + deficient solutions, monitored at 276 nm 32. Denaturation of FITC-labeled and unlabeled gelsolin with urea, 120 monitored at 215 nm. 33. Thermal denaturation of unlabeled gelsolin, monitored at 215 nm, in both C a 2 + deficient and C a 2 + saturated environments, 122 and in both the presence and absence of 0.3 M Gnd-HCl. 34. Thermal denaturation of FITC-labeled gelsolin, monitored at 215 nm, in both C a 2 + deficient and C a 2 + saturated environments, 124 and in both the presence and absence of 0.3 M Gnd-HCl. 35. Photograph of horse plasma gelsolin crystals grown in a calcium 126 deficient solution x i i A l . Schematic representation of the six domains of gelsolin 139 oriented in approximately the same way. A2. (a) Structure of whole gelsolin illustrating the relative positions 141 of the six individual domains, (b) Orientation of SI-S3 and S4-S6 on the end of the Heidelberg model of the actin filament, illustrating a structure for the A2G nucleation complex. A3. The structure of whole gelsolin positioned on the human SI 143 F-actin model. x i i i Abbreviations A adenine base Acc I Restriction enzyme from Acinetobacter calcoaceticus ADP Adenosine 5'-diphosphate A M V Avian myeloblastosis virus reverse transcriptase A T P Adenosine 5'-triphosphate B Barbed (+) end on an actin filament BamH 1 Restriction enzyme from Bacillus amyloliquefaciens H Buffer A 2 mM Tris-HCl, 1 mM DTT, 0.2 m M ATP, 0.2 m M C a C l 2 , pH 7.6 C cytosine base CCA Convex constraint analysis cDNA Complementary DNA CD Circular dichroism Cys Cysteine amino acid D Denatured protein state DBP Vitamin-D-binding protein d d N T P 2',3'-dideoxynucleoside 5'-triphosphate D E A E Diethylaminoethyl DEPC Diethyl pyrocarbonate Denhardt's 50X concentrated-1% Ficoll Type 400 (Pharmacia), 1% BSA, and 1% polyvinylpyrrolidone. dH20 Deionized water DMSO Dimethyl sulfoxide DNA Deoxyribonucleic acid d N T P 2'-deoxynucleoside 5'-triphosphate dsDNA Double stranded deoxyribonucleic acid D T T D,L-dithiothreitol E. coli Escherichia coli EcoR I Restriction enzyme from Escherichia coli RY13 EcoR V Restriction enzyme from Escherichia coli B946 E D T A Ethylenediaminetetraacetic acid E G T A 1,2-Di(2-aminoethoxy)ethane-N,N,N',N'-tetraacetic acid xiv F-actin Polymerized filamentous actin FITC Fluorescein-5-isothiocyanate G guanine base G-actin Monomelic globular actin Glucose Buffer 50 mM glucose, 25 mM Tris-HCl, pH 8.0, 10 m M E D T A Gnd-HCl Guanidine hydrochloride GSP Gene specific primer Hepes N-2-Hydroxyethylpiperazine-N'-2-ethanesulfonic acid IPTG I sopropyl- (3 -D - thiogalactopyrano side Klenow Buffer 10X concentrated-0.5 M Tris-HCl, pH 7.2, 0.1 M MgS04, ImM dithiothreitol, and 500 ug/mL BSA L Left circularly polarized light LB Luria broth growth medium MCS Multiple cloning site M - M L V - R T Moloney murine leukemia virus reverse transcriptase mRNA Messenger ribonucleic acid N Native protein state Nde II Restriction enzyme from Neisseria dentrificans NRCC 31009 NZ amine 5 g NaCl, 2 g MgS047H20, 10 g NZ Amine, pH 7.5 per 1.0 L of water. NZY media Constituents identical to NZ amine with the addition of 10 g of yeast extract One-phor-all 10X concentrated buffer-100 m M Tris-acetate, 100 mM magnesium acetate, 500 m M potassium acetate P Pointed (-) end of a filament PAGE Polyacrylamide gel electrophoresis PCR Polymerase chain reaction P E Plasma extension P E G Polyethylene glycol Pi Phosphate-inorganic PIP Phosphatidylinositol 4-phosphate PIP 2 Phosphatidylinositol 4,5-bisphosphate PMSF Phenylmethylsulfonyl fluoride R Right circularly polarized light X V RACE Rapid amplification of cDNA ends RNA Ribonucleic acid RNA Gel buffer 5X concentrated-0.2 M morpholinopropanesulfonic acid, pH 7.0, 50 mM sodium acetate, 5 mM EDTA RT-PCR Reverse transcription-polymerase chain reaction S1-S6 Domain segments of gelsolin, SI through S6. Sal I Restriction enzyme from Streptomyces albus G SDS Sodium dodecyl sulfate SLIC Single stranded ligation to cDNA ends SM buffer 0.1 M NaCl, 14.5 mM MgS0 4 , 50 mM Tris-HCl, pH 7.5, and 0.01% (w/v) gelatin SP Signal peptide 20X SSC buffer 3 M NaCl, 0.3 M sodium citrate, pH 7.0 ssDNA Single stranded deoxyribonucleic acid 20X SSPE buffer 3 M NaCl, 0.2 M NaH2P0 4H 20, and 0.02 M EDTA, pH 7.4 T thymine base TAE buffer 40 mM Tris-HCl, pH 7.5, 20 mM acetic acid, and 1 mM EDTA Taq DNA polymerase from Thermus aquaticus T d T Terminal deoxynucleotidyl transferase TE buffer 10 mM Tris-HCl, pH 8.0, 1 mM EDTA T E M E D N.N.N.'N'-tetramethylethylenediamine Tris-HCl Tris(hydroxymethyl)aminomethane hydrochloride U T Untranslated sequence UV Ultraviolet light Xbal Restriction enzyme from Xanthomonas campestris X-gal 5-bromo-4-chloro-3-indoyl-p-D-galactopyranoside Xho I Restriction enzyme from Xanthomonas campestris xvi Acknowledgments I would like to thank Dr. L.D. Burtnick for his guidance and support during the course of this project, and for helpful discussions on the preparation of this thesis. Additional thanks go to Dr. S. Withers for his helpful comments and suggestions. Many thanks are extended to Dr. R.T.A. MacGillivray, who offered us the use of his laboratory for the duration of the gelsolin cloning project. All the people in the MacGillivray lab were tremendously helpful in teaching molecular biology to me. Special thanks to Mr. J. Hewitt, whose expertise and knowledge were instrumental to the success of this project. I also wish to thank Ms. A. Maxwell and Mr. H. Vo for their help during the course of this project. 1 CHAPTER I INTRODUCTION 1. ACTIN, ACTIN BINDING PROTEINS, AND GELSOLIN i. Project Overview The focus of this research project is centered around gelsolin, a multifunctional actin binding protein that regulates actin filament assembly. Biochemical, biophysical, and molecular biological studies have delineated some of gelsolin's biological activities, but a detailed description of the structure-function relationship of this versatile protein has not been completed. The studies presented in this thesis detail a structural characterization of equine plasma gelsolin. The first level of protein structure was elucidated through the acquisition of gelsolin's primary amino acid sequence. The sequence of a protein reveals much about its evolutionary history, and is particularly informative if it is considered with other chemical and physical data. The relationship of amino acid residues that are close to one another in the linear sequence defines the secondary structure of a protein. CD studies provided an estimate of the secondary structural conformations found in gelsolin, in addition to providing insight into how gelsolin unfolds, aggregates, and precipitates. X-ray diffraction studies with crystals of horse plasma gelsolin produced the tertiary structure of the protein. Data collection and structure determination were done at Oxford University by L. D. Burtnick and R. C. Robinson. With the high resolution structure of gelsolin, we are able to present models of how gelsolin interacts with actin. 2 u. Actin Initially identified as a major constituent of muscle tissue over fifty years ago, research in the actin field is now progressing at atomic resolution (Straub, 1942; Mommaerts, 1992). Historically, actin was studied in the context of muscle contraction, a process where muscle cells convert chemical energy into mechanical movement. During the mid-1960s, isolation of highly pure non-muscle actin from slime molds put this protein into a new perspective. As participation in the field of non-muscle actin research increased, it quickly became evident that all eukaryotic cells contain significant amounts of this protein. Some cells devote up to 20% of their total protein production to actin (Korn, 1982). Actin is not only a very abundant cellular protein, sequence analysis has revealed that it is also one of the most conserved proteins identified to date. Such a degree of sequence conservation implies the existence of an evolutionary pressure which serves to maintain actin's tertiary conformation (Hennessey et al, 1993). Certain regions of the actin structure are critical recognition sites for the many regulating factors that associate with this molecule, sites which evolution strives to preserve. UL Monomeric G-actin In the absence of salt, actin isolated from rabbit skeletal muscle exists as a 42 kDa monomer composed of a single polypeptide chain of 375 amino acids (Korn, 1982). Due to its relatively globular shape, monomeric actin is referred to as G-actin. This protein contains a single high affinity nucleotide binding site, along with one for a divalent metal 3 ion, either C a 2 + or M g 2 + . In addition to the high affinity site, actin exposes a series of lower affinity sites capable of accepting either mono or divalent cations. The natural tendency of actin to polymerize instead of crystallize was an inherent problem that had to be resolved by structural biologists. Actin was crystallized in the presence of bovine pancreatic DNase I, an enzyme known to inhibit the polymerization process. Crystals of a 1:1 actin-DNase I complex grown with C a 2 + and ATP produced the three dimensional structure of the monomer to 2.8 A resolution (Kabsch et al, 1990). The electron density map was solved using a combination of molecular replacement for the DNase I moiety and with two isomorphous heavy metal derivatives. The crystallographic model of G-actin reveals a globular molecule with one large and one slightly smaller domain, each of which is divided into two subdomains (Figure 1). The core structural element of subdomains 1 and 3 is a central 5-stranded p-sheet surrounded by four a-helices. The cleft between the two main domains is occupied by a single ATP molecule and a closely associated C a 2 + ion which is chelated by the P and y phosphate groups of the nucleotide. The Ca 2 + -ATP moiety resides in a hydrophilic pocket which is shielded from bulk solvent. Stability is imparted to actin by the nucleotide and C a 2 + ion which help to hold the two main domains together. The overall dimensions of the molecule are approximately 55 X 55 x 35 A . iv. Polymeric F-actin The unique ability of actin to polymerize into non-covalent helical 4 filaments (F-actin) is fundamental to the biological activity of this protein in eukaryotic cells (Carlier, 1991). Actin filaments are linear polymers which are made up of two parallel actin strands that are slightly staggered with respect to one another and twisted into a right handed helix (Figure 2). Figure 1: Schematic representation of the G-actin structure (Kabsch et al, 1990). ATP and a calcium ion are located between the small (right) and large (left) domains. 5 Electron microscopic studies show that there are 13 monomer units in the two strands between a crossover point, with the distance between consecutive crossover points along the helical axis estimated at 36 nm (Korn, 1982). Although F-actin appears rod-like, the filaments do exhibit some torsional flexibility with fluctuation of 5-10 degrees. 36 nm Figure 2: Helical arrangement of the two staggered, parallel rows of actin polymers present in an actin filament. A model of the "F-actin filament has been generated by fitting the structures of individual G-actin molecules into experimental F-actin fiber diffraction data, with appropriate rotation of subdomains. (Holmes et al, 1990). The subunit conformation, as seen in the actin-DNase I complex, remains unchanged in the filament. In this model, the large domain of actin is placed near the central axis of the filament, with the smaller domain projected away from it. The main actin-actin contacts occur along the two start helix and are dominated by hydrophobic interactions. Additional contacts may be through electrostatic forces or by specific hydrogen bonds. 6 v. Converting G to F-actin Incorporation of G-actin into a growing filament may occur at either of the two ends of the polymer. Kinetic studies have demonstrated that the affinity and thus the rate of elongation differs at both ends (Korn, 1982; Mannherz, 1992). The end which displays preferential elongation is referred to as the barbed (B) or the (+) end, while the slower growing end is the pointed (P) or (-) end. The basis for the B and P nomenclature stems from electron micrographs of actin filaments decorated with myosin subfragment 1 which reveal a distinctive arrowhead pattern (Stryer, 1981). The two ends of the filament display this polarity because of the uniform head to tail subunit organization of the monomers. Many of the mechanistic details of actin polymerization have been thoroughly investigated, although not all aspects have been elucidated. The polymerization reaction is a highly cooperative condensation process which is fully reversible. In vitro actin polymerization may be induced in a number of ways, most commonly by addition of M g 2 + or KC1, or both to physiological levels (Pollard and Cooper, 1986). Our current understanding is that the polymerization process includes four stages: monomer activation, nucleation, elongation, and filament annealing (Pollard, 1990). Monomer activation is associated with a rapid change in G-actin conformation following ligand binding. After the high affinity ATP and metal binding sites have been filled, cations binding to the lower affinity sites induce a conformational change which facilitates polymerization. Oligomers which have a higher probability of elongating into filaments than dissociating back to monomers are the structures which nucleate filament assembly. This step is rate limiting and readily 7 observable during the time course of spontaneous polymerization (Korn, 1982). A plot of this time course is sigmoidal, showing an initial lag phase which is attributed to the nucleation process. Kinetic analysis of polymerization curves indicates that the stable nucleus is probably a trimer, one derived from a dimer and a monomer (Carlier, 1991). Bidirectional growth of the oligomers produces an F-actin polymer. Hydrolysis of actin-bound ATP, associated with polymerization, is not kinetically coupled to filament assembly. This process occurs in two discrete steps following monomer incorporation. ATP is initially hydrolyzed to ADP and inorganic phosphate (Pi), then the inorganic phosphate is slowly released from the polymer. This type of hydrolytic mechanism produces F-actin which has ATP monomers incorporated at its barbed end, with ADP monomers present in the middle and towards the pointed end of the filament. Conversion of ATP- to ADP-actin and release of Pi within a filament destabilizes individual actin-actin interactions, a process thought to facilitate depolymerization at some later time. The final stage of the polymerization reaction is filament annealing. Two existing filaments in the correct head to tail orientation may anneal together to form an even larger polymer. The average steady state filament at physiological conditions is approximately 3 \im long. vL Non-muscle Actin Binding Proteins The self assembly properties of G-actin only account for the formation of long, linear F-actin filaments. As actin in non-muscle cells is dynamic and known to exist in a variety of structural configurations, additional factors must be responsible for its architectural versatility. The 8 ability of actin to form diverse superstructures influences cell shape and function, and conversion from one structure to another is the basis for cell motility. Regulation of the various structural forms of actin is the responsibility of a group of proteins that are collectively known as actin binding proteins (Figure 3) (Stossel et al, 1985; Pollard and Cooper 1986; Hartwig and Kwiatkowski, 1991). By defining the three dimensional organization of actin structures, actin binding proteins indirectly influence a significant number of cellular responses. Figure 3: Actin binding proteins and their interactions with monomeric and polymeric forms of actin. Representative proteins of each category are shown in parentheses. 9 Characterization of their in vitro functional activities classifies actin binding proteins into three general categories: monomer binding proteins, capping and severing proteins, and crosslinking proteins. The monomer binding proteins, represented by DNase I, the vitamin D-binding protein, profilin and others, bind primarily to G-actin. Formation of a tight 1:1 stoichiometric complex effectively ties up actin monomers that would otherwise polymerize into filaments. They effectively raise the cellular G-actin critical concentration, which is defined as the minimum concentration of actin monomers required to initiate polymerization. Through inhibition of spontaneous polymerization, the availability of a pool of G-actin monomers is secured should it be required for a specific cellular response. The best characterized member of the severing and capping group of actin binding proteins is gelsolin. Related to gelsolin is villin from vertebrate epithelial cells, severin from Dictyostelium discoideum, and fragmin from Physarum polycephalum (Stossel et al, 1986). Actin severing is a non-proteolytic process in which the protein responsible for severing binds to an internal region of the F-actin polymer in a C a 2 + specific manner. Breakage of a filament occurs through the disruption of the non-covalent actin-actin bonds responsible for the integrity of the polymer. After filament cleavage, many severing proteins remain bound to the barbed filament end inhibiting monomer-polymer interactions at that site. This capping of actin filament ends prevents both polymer elongation and filament annealing. In addition to severing and capping, many of the proteins that belong to this group are able to promote pointed-end filament growth by generating an increased concentration of stable nuclei to which monomers 10 may add (Weeds, 1982). Spontaneous polymerization is accelerated by shortening the lag phase of the polymerization process. The third predominant actin binding protein group is made up of filament crosslinking proteins. The members of this family are a set of highly diverse proteins that have been isolated from a variety of species. Exemplified by a-actinin, filamin and spectrin, these molecules have the ability to link together at least two actin filaments to generate actin structures that vary from highly organized bundles to very disordered gels. Bundles or gels are capable of further association to form even higher levels of structure, or they may interact with membranes or organelles. viL Gelsolin Gelsolin is a calcium-dependent regulatory protein that controls linear actin assembly by severing and capping F-actin filaments, and by nucleating filament assembly (Yin et al, 1981; Yin, 1987). It was identified as the factor responsible for promoting the reversible gel to sol transformation of macrophage cytoplasmic extracts, hence the name gelsolin (Yin and Stossel, 1979). Around the time of this discovery, a protein with similar properties was found in mammalian plasma (Norberg et al, 1979; Harris et al, 1980, Harris and Schwartz, 1981). In the past fifteen years, gelsolin has been isolated and characterized from a wide variety of species, ranging from lower eukaryotes to mammals. A variety of cells synthesize gelsolin in the form of either an intracellular or secreted protein. The secreted isoform, once referred to as F-actin depolymerizing factor or brevin, was the protein that was initially discovered in blood plasma (Yin, 1987). 11 Cytoplasmic and plasma gelsolins are structurally and functionally very similar. The human plasma isoform, with a mass of 83 kDa, is slightly larger than the intracellular version at 80 kDa. The mass difference between these two proteins is attributable to a 25 amino acid residue extension found at the N-terminal of the plasma protein (Yin et al., 1984). Genomic analysis revealed that the origin of the two proteins stems from a single gene (Kwiatkowski et ah, 1986; 1988). Through an elaborate pattern of RNA splicing and alternative transcription initiation, a single gelsolin gene is able to code for both isoforms of the protein. viiL Calcium Regulated Interactions with Actin Gelsolin interacts with both monomeric and polymeric forms of actin in at least three ways, all of which participate in generating short actin filaments (Janmey et ah, 1985). The fiber length of an existing filament diminishes rapidly as a result of gelsolin's ability to break or sever the non-covalent bonds that hold the polymer together. Following severing, gelsolin caps the fast growing end of the cleaved filament. This inhibits filament elongation by obstructing the monomer binding site and by preventing reannealing of two severed filaments. Nucleation of actin filament assembly is the third functional activity associated with gelsolin. By creating stable nuclei, gelsolin promotes the formation of many small F-actin polymers in favor of a limited number of large ones. Two actin monomers (A) bind to gelsolin (G) in a C a 2 + specific manner to form a ternary A 2 G complex. Addition of excess E G T A dissociates one of the monomers, but leaves an intact binary A G complex that retains one C a 2 + ion (Bryan and Kurth, 1984; Harris, 1985). This so 12 called E G T A resistant A G complex exhibits nucleating and capping activities, but can not sever F-actin filaments (Coue and Korn, 1985). Dissociation of the A G complex was demonstrated by the addition of either phosphatidylinositol monophosphate (PIP) or phosphatidylinositol 4,5-bisphosphate (PIP2) (Janmey and Stossel, 1987; Janmey et al, 1987). In addit ion to dissociating the E G T A resistant complex, the polyphosphoinositides also unblocked gelsolin capped actin filaments and inhibited the severing of F-actin by gelsolin. Of the three functions of gelsolin, severing is the one which is most stringently controlled by calcium levels. Two distinct C a 2 + binding sites have been identified. One site with an affinity of 10-20 [iM is specific for C a 2 + , while the other site, with millimolar affinity, binds either C a 2 + or M g 2 + (Doi et al, 1990; Tellam, 1991). The affinity of the C a 2 + specific site correlates with the result that the half-maximal extent of severing occurs around 10 |xM C a 2 + (Lamb et al, 1993). Upon binding C a 2 + , the surface charge properties of gelsolin are significantly altered, indicative of a structural change which is essential to its severing activity (Hwo and Bryan, 1986; Doi et al, 1990). ix. Functional Role of Cytoplasmic Gelsolin The in vivo function of cytoplasmic gelsolin still remains somewhat unclear, although its role in rearranging the cytoplasmic actin skeleton in response to transient rises in cytosolic C a 2 + is slowly being elucidated. Based upon regulatory effects that P I P 2 and C a 2 + have on gelsolin, a functional model that incorporates these observation has been proposed (Yin, 1987). A n initial rise in intracellular C a 2 + levels, followed by P I P 2 13 hydrolysis, creates conditions favorable for severing. Subsequent removal of C a + 2 and an increase in P I P 2 inhibits severing, and uncaps the short oligomers to provide ready made nuclei for F-actin polymerization. Capped actin filaments may be key intermediates for eliciting actin assembly near locations where cell surface receptors cause appropriate changes in P I P 2 concentrations. x Gelsolin in Plasma Plasma gelsolin and the vitamin-D-binding protein (DBP) are two components that depolymerize F-actin in plasma. The combined action of these two proteins, both of which circulate at micromolar concentrations, forms an actin scavenging system (Ito et al, 1992). The ionic conditions of plasma are such that any actin introduced into it, whether by physiological or pathological cell death, would be expected to polymerize into filaments (Lind et al, 1986). Circulating filaments could potentially increase plasma viscosity, which would seriously impede blood flow through microcirculatory vessels. The clearance of actin from plasma transpires in two steps. The first of these is rapid and attributable to direct filament severing by gelsolin. The second step, which is slower than severing, involves actin-DBP complexation (Ito et al, 1992). DBP associates with monomeric actin to form a tight 1:1 stoichiometric complex, but does not interact with filaments (Lees et al, 1984; Coue et al., 1986). Actin-DBP complexes are rapidly cleared by the liver at a rate comparable to the clearance of actin from blood (Lind et al, 1986). Since DBP has a higher affinity for actin than does gelsolin, the gelsolin present in plasma at equilibrium is 14 expected to be free of actin. Preventing permanent actin-gelsolin complexation in plasma by preferentially forming actin-DBP complexes conserves the filament severing activity of plasma gelsolin. xL Localization of Actin Binding Sites Limited proteolytic digestion of human plasma gelsolin with chymotrypsin or subtilisin quickly cleaves the molecule into two fragments. Such studies revealed that gelsolin is composed of stable structural domains which are connected by proteolytically sensitive regions. The N-terminus produces a 45 kDa proteolytic fragment (CT45N), while the C-terminus is found in a 38 kDa (CT38C) fragment (Kwiatkowski et al, 1985; Chaponnier et al, 1986; Bryan and Hwo, 1986). At longer digestion times, the N-terminus peptide breaks down producing a 14 kDa (CT14N) and an adjacent 28 kDa (CT28) fragment. As the rate of proteolysis is enhanced in the presence of C a 2 + , it is thought that a C a 2 + induced conformational change opens up the protein structure and facilitates the protease's accessibility to the cleavage sites. The large CT45N fragment severs and caps filaments almost as effectively as native gelsolin, but it can not nucleate and is no longer regulated by calcium. The C-terminal fragment does not exhibit any severing or nucleating activity. Binding studies conducted with individual proteolytic fragments identified three gelsolin peptides which interact with actin. CT14N bound to G-actin independent of C a 2 + concentration, while CT38C rapidly formed a reversible complex with a single monomer but only in the presence of C a 2 + (Bryan, 1988). CT28 was found to contain a single calcium-independent F-actin binding site (Yin et al, 1988). These studies revealed 15 that one of the actin binding sites is hidden in the intact protein, and that the C-terminal domain imparts calcium regulation on the molecule. Comparison of peptide sequences with human gelsolin cDNA identified CT45N as residues 1-406 and CT38C as residues 407-755 (Kwiatkowski et al, 1986; Way et al, 1989). Similarly, the smaller CT14N fragment spanned residues 1-150 and CT28 consisted of residues 151-406. Using segmental nomenclature derived for pig plasma gelsolin cDNA (Figure 4), CT14N corresponds to segment SI , CT28 to S2-3, and CT38C to S4-6 (Way and Weeds, 1988). Expression and characterization of these and other gelsolin segmental combinations further delineated the critical domains within this molecule (Weeds and Maciver, 1993). SI contains a high affinity C a 2 + independent actin monomer binding site which is essential for severing. S2-3 is the segment combination which is responsible for targeting gelsolin to F-actin. Interaction of this domain with filamentous actin is not regulated by C a 2 + . S4-6 contains a single monomer binding site, and is responsible for regulating gelsolin's activities through binding calcium. 14-133 134-250 251-373 391-512 513-618 619-739 ' CT45N ' CT38C I 1 —I CT 14N CT28 Figure 4: Comparison of human gelsolin proteolytic fragments to the gelsolin domain structure (S1-S6) proposed from porcine cDNA (Way and Weeds, 1988). The stretch of residues that defines each domain is listed above every segment. 16 In the simplest model of severing, S2-3 orients SI into a correct position on the filament. The importance of both S1 and S2-3 for severing was demonstrated by a chimeric construct of SI and the F-actin binding domain of a-actinin, along with a second construct consisting of the SI equivalent of the capping protein gCap39 and S2-3 (Yu et al., 1991; Way et al, 1992). The chimeric protein with the a-actinin binding site showed that SI is essential for severing, while the gCap39 experiment induced previously non-existent severing activity by adding a F-actin binding site to a capping protein. Of two possible binding topologies, S2-3 binds along the filament axis between two monomers as opposed to across the filament to make contact with the other strand (Feinberg et at, 1995). xii. Gene Structure of Gelsolin Synthesis and secretion of plasma gelsolin is tissue specific, with skeletal, smooth, and cardiac muscles producing the largest amounts of plasma gelsolin mRNA (Nodes et al, 1987; Kwiatkowski et al, 1988a). These tissues dedicate approximately 0.5 to 3% of their total biosynthetic protein production to plasma gelsolin. High levels of synthesis and total mass makes skeletal muscle tissue the major source of plasma gelsolin. The first clones of gelsolin were isolated from a human HepG2 cDNA library (Kwiatkowski et al., 1986). The inferred amino acid sequence from full length cDNA produced a 755 amino acid protein that revealed both a signal peptide and the 25 residue plasma extension. Analysis of gelsolin's primary sequence showed a strong tandem repeat in the amino and carboxy halves of the molecule. Further investigation of the sequence revealed an even more elaborate repetitive structure of six similar 17 domains. Using probes derived from human plasma gelsolin, a series of overlapping porcine gelsolin cDNA clones were characterized from a pig liver cDNA library (Way and Weeds, 1988). The predicted primary sequence of porcine gelsolin contains features similar to those of human plasma gelsolin, namely elements of a signal peptide and the plasma extension. Based upon a comparison of inferred human and porcine N -terminal sequences, it is somewhat surprising that the pig protein as isolated is found with only a nine residue extension (Weeds et al, 1986). The difference between these two N-terminal sequences is probably due to protease activity that occurs after enzymatic removal of the signal peptide. Structural analysis of porcine plasma cDNA produced results that paralleled those obtained for human plasma gelsolin. The molecule consists of similar N and C terminus halves, each of which accommodates the three shorter stretches of repeating sequence at approximately equally spaced positions. Starting from the N-terminus, the segments are labeled sequentially beginning with SI and ending with S6. The greatest degree of identity is found between segments 1 and 4, followed by 2 and 5, then 3 and 6. Exact segmental boundaries are difficult to define by only sequence data, but previously mentioned limited proteolysis studies do support the existence of independent structural domains. The two large halves of repeating sequence found in both human and porcine gelsolin are suggestive of an evolutionary gene duplication event (Kwiatkowski et al, 1986; Way and Weeds, 1988). The smaller internal repeats detected in gelsolin, and in actin binding proteins related to it, may have developed from a monomeric actin binding precursor that consisted of 120 to 150 amino acids with a mass of about 14-15 kDa. 18 Analysis of genomic clones containing the human gelsolin gene confirmed previous observations that this single gene codes both cytoplasmic and plasma isoforms of gelsolin (Kwiatkowski et al, 1986, Kwiatkowski et al, 1988). cDNA clones representing both proteins showed that they are identical with the exception of their 5'-ends. A human genomic library screened with gelsolin cDNA probes revealed the origin of the two distinct messages. The gene that codes both gelsolin isoforms is made up of at least fourteen distinct exons which span a region of approximately 70 kb. The 5'-end arrangement of the cytoplasmic isoform is exon 1, intron, exon 2, intron, then exon 4. Together, exons 1 and 2, which are 13 kb apart in the gene, make up the unique 5' untranslated region of cytoplasmic gelsolin (Figure 5). The 5'-end of the plasma isoform differs from the cytoplasmic one as it is made up of exon 3, an intron, then exon 4. Exon 3, found 2.3 kb upstream from exon 4, comprises a small region of untranslated sequence and codes for both signal peptide and the first 21 residues of the plasma protein. Exon 4 is the first one common to both isoforms. The first five exons span a region of approximately 40 kb, while the remainder of the gene, with at least ten additional exons, spans another 32 kb. 19 U T (Exons 1 and 2) CYTOPLASMIC M (Exon 4) M PLASMA U T SP (Exon 3) P E Figure 5: Exon arrangement of the 5'-end of cytoplasmic and plasma gelsolin cDNA. UT represents untranslated sequence > SP is the signal peptide, and PE is the plasma extension. The methionine in the dark region represents the first common residue of both proteins. Initiation. of transcription. occurs at two distinct sites in the gelsolin gene, separated by 32 kb. Cytoplasmic mRNA is produced when transcription is initiated at exon 1, while the plasma initiation site is found in exon 3. To generate mature cytoplasmic message, exon 3 and the large adjacent intervening sequence must be spliced out during RNA processing, since this stretch is downstream to the cytoplasmic 5' start. By alternative transcription initiation and selective RNA processing, the single gelsolin gene is able to produce two distinct mRNA messages that code both forms of gelsolin. xiiL Crystal lization of Gelsolin A cDNA construct consisting of the leader peptide of X phage ell protein connected to human gelsolin by a Factor X proteolytic site was used to express recombinant protein in E. colt (McLaughlin and Gooch, 20 1992). Purified recombinant gelsolin was crystallized from a solution containing 33% (v/v) ammonium sulfate with either E D T A or C a 2 + present. Crystals grew in the pH range of 6.0 to 9.0, but the best conditions were found at pH 8.0. These conditions produced flat plates with typical dimensions of 0.5 x 0.5 x 0.2 mm that diffracted to 3.5 A resolution. The first atomic structure of a gelsolin domain was elucidated from a 1:1 stoichiometric complex consisting of rabbit skeletal muscle actin and recombinant human SI (Mannherz et al, 1992; McLaughlin et al, 1993). The purified complex crystallized into flat plates from a buffer containing 5-9% (w/v) polyethylene glycol 6000. The complex diffracted to a resolution of 2.5 A and was solved by locating the actin unit with molecular replacement. SI is composed of a three layered structure that is centered around a 4 stranded p- sheet. One face of the sheet is flanked by a parallel 4 turn a-helix, while the other side is covered with a smaller a-helix in an orientation perpendicular to the sheet (Figure 6). This fragment of gelsolin defines a globular domain with approximate dimensions of 13 x 26 x 25 A . The SI structure revealed that the conserved residues of the gelsolin sequence contribute to the closely packed apolar core domain. The majority of these residues are found in the two middle strands of the P-sheet and in the shorter a-helix. This structural motif is predicted for the other segments of gelsolin by virtue of repetition of its conserved sequence. 2 1 Figure 6: Structure of recombinant human SI solved in a complex with rabbit skeletal muscle actin (McLaughlin et al, 1993). Contacts between the 4 turn rx-helix of SI with the apolar cleft that forms at the interface of actin subdomains 1 and 3 hold the complex together. Located at the interface of the helix and cleft is a single C a 2 + ion. The metal is coordinated by both SI and actin ligands, and may represent the C a 2 + ion that remains trapped in the E G T A resistant actin:gelsolin complex. The structure also unveiled the location of a second C a 2 + ion which was located solely within the SI molecule. This ion is chelated by two buried carboxylates that are conserved in the sequence. Incorporation of the Sl:actin complex into the Heidelberg model of F-actin places the SI moiety tangentially along the filament. This orientation of SI creates steric clashes with subdomain 2 of the adjacent actin monomer in the same strand. Disruption of these critical actin-actin bonds may be the basis for filament severing. 2 2 xiv. Structural Comparison of Acting Binding Proteins Sequence and biochemical analysis of members belonging to the gelsolin family of actin binding proteins predicts the presence of an independent 125-150 amino acid core domain which is repeated throughout the protein. This sequence motif, which has been identified in actin binding proteins isolated from a very diverse group of phyla, is found in either a three or six fold tandem repeat (McLaughlin et al, 1993). One year after human gelsolin SI was solved, the structure of villin 14T was presented (Markus et al, 1994). Villin is an actin binding protein that behaves much like gelsolin, with the exception of having bundling activity. Peptide mapping and sequence data revealed six similar domains with an unrelated seventh domain (Arpin et al, 1988; Bazari et al, 1988). The latter domain forms a headpiece on the C-terminus of the protein which gives villin its bundling properties. The three dimensional solution structure of the N-terminal segment of this protein (Villin 14T) was determined by nuclear magnetic resonance spectroscopy (Markus et al, 1994). At this resolution, the structure of villin 14T reveals a central (3-sheet which is flanked on one side by two parallel running helices and by a small p-sheet and a-helix in a perpendicular orientation on the other side (Figure 7). 23 Figure 7: Ribbon diagram of the villin 14T structure (Markus et at, 1994). The central (3-sheet is surrounded by three helices and one short parallel (3-sheet. Subsequent to the structures of human SI and villin 14T, the structure of severin domain 2 was the next to be elucidated (Schnuchel et al., 1995). Like gelsolin and villin, this is a C a 2 + mediated actin binding protein that severs, nucleates, and caps actin filaments. Severin is made up of three similar domains believed to resemble the N-terminal half (Sl-S3) of gelsolin. Severin domain 2, which presumably corresponds to gelsolin domains 2 and 5, also contains a central 5-stranded P-sheet that is sandwiched between a 4 turn a-helix on one side and a smaller 2 turn helix on the other (Figure 8). 24 Figure 8: NMR solution structure of severin domain 2 (Schnuchel et al., 1995). Comparison of villin 14T, severin domain 2, and human gelsolin SI reveals that the conserved residues of each domain fold into a similar apolar core structure. The conserved residues of each repeating unit are not actin binding motifs as not all segments of gelsolin bind actin, but rather reflect the presence of the underlying structure. 25 2. RECOMBINANT DNA METHODOLOGY L Methods of Gene Cloning The process of cloning a gene essentially involves moving a specific fragment of genetic information from one DNA molecule to another. This is accomplished with a set of cutting and joining procedures that utilize specialized enzymes and shuttle vectors. Once cloned, quantities of recombinant DNA sufficient for sequence analysis, mutagenesis, protein production, and other genetic engineering applications are readily attainable. The two procedures which are commonly used to clone genes are library screening and reverse transcriptase-polymerase chain reaction (RT-PCR). Working with libraries presents two options as they may be constructed from either genomic or complementary DNA (cDNA). As with any procedure, both library screening and RT-PCR have inherent advantages and disadvantages. The direction that is taken to resolve a particular cloning problem is frequently determined by the type of information that is required upon completion of the experiment. If the organization of a gene or the site of transcription initiation is the focus of a cloning project, screening a genomic library is the method of choice. As genomic DNA is composed of coding exons and non-coding introns, clones from such libraries will contain pertinent organizational information. However, the presence of introns may manifest problems, especially if they are large. A significant amount of research time may be devoted to characterizing and sequencing regions of DNA that do not code for amino acids in the gene product. On 26 the positive side, large quantities of genomic DNA are readily available from virtually any tissue source. It is suited for cloning after a simple restriction endonuclease digest. cDNA libraries are created from complementary DNA. Such DNA is synthesized by a reverse transcriptase (RT) enzyme which incorporates nucleotide triphosphates into single stranded messenger RNA (mRNA) templates. After mRNA has been isolated, RT enzymes carry out the in vitro synthesis of the cDNA. As eukaryotic cells process their initial mRNA transcripts to remove intervening intronic sequences, clones from cDNA libraries do not show gene organization. However, as the nucleotide sequence of cDNA translates directly to the primary protein sequence, it is advantageous to use such clones in sequencing projects. Non-library screening approaches to cloning involve the polymerase chain reaction (Mullis et al, 1986). As PCR enzymes are able to use either genomic or cDNA templates, a variety of molecular biology methods have been developed to take advantage of this versatile reaction. The combination of RT-PCR has the capability of rapidly producing a target gene without having to complete the often arduous arid laborious task of constructing and screening a library. vL Plasmids and Phages Central to the advancement of recombinant DNA technology is the development of cloning and expression vectors; .Bacterial plasmids and bacteriophage viruses are infectious pieces of DNA that utilize host cell machinery to replicate. Having gained access to a cell, these extra chromosomal pieces of DNA can have pronounced effects on the living 27 host. Some may take control of a cell and kill it, while others may prove beneficial to the host. Plasmids are generally circular pieces of DNA which are found exclusively inside cells, while bacteriophages, which have their DNA surrounded by a protective protein coat, are capable of surviving outside host cells for extended time periods. Plasmids and phages may both be cut at sequence specific sites with any one of a number of specialized restriction enzymes. A foreign DNA fragment, with ends complementary to the cut vector, may be inserted or ligated into a linearized vector. For plasmids, the in vitro ligation is followed by a transformation, a process whereby the vector containing the D N A insert is introduced to competent host cells. Propagation of transformed cells produces large quantities of cloned DNA. The plasmids that are frequently utilized in recombinant DNA applications are relatively small (usually less than 5000 base pairs), circular, double stranded DNA molecules that occur naturally in bacterial cells. They contain an origin of replication or start signal which ensures that DNA polymerase will replicate it. Additionally, many plasmids contain a gene which confers antibiotic resistance to the host. Drug resistance turns out to be extremely important, as it allows us to select for only those bacteria that acquired a plasmid with the antibiotic resistance gene. Only transformed cells have the ability to grow in the presence of a toxic antibiotic. Bacteriophages (phages) are considerably more complex than plasmid vectors as they are intact bacterial virus particles capable of infecting host cells (Stryer, 1981). They contain a central core of DNA that codes for a series of proteins, including those found in the outer protective coat. Like plasmids, phages lack the machinery necessary to 28 reproduce, and are thus dependent on the replication services of a living bacterial host cell. Many common phages essentially consist of a piece of tightly wrapped DNA that is stuffed inside a protective head piece constructed of proteins. Attached to the head is a tail, also made of protein. When such a virus comes into contact with a bacterial cell, the phage tail adheres to a specific site on the host. The DNA inside the head is injected into the bacterium through the attached tail in a manner analogous to the action of a hypodermic syringe. Soon after phage infection, both head and tail proteins are actively being synthesized by bacterial ribosomes. In concert with protein production is phage DNA replication. Once all components have been synthesized, head proteins assemble and package the phage DNA, followed by tail attachment. Exit from the host cell is accomplished by synthesis of an enzyme that lyses bacterial cell walls. The viability and titer of a phage solution is assessed by a plaque assay. A bacterial culture infected with phage is plated out on a solid agar plate, then overlaid with soft agar. Bacteria in the softer top layer diffuse to form a lawn of cells. Areas which contain phage are evident by the formation of plaques, which are characterized as small clear spots on the lawn. Plaques are sites were phages have killed host cells, each of which originated from a single phage clone. To establish if a plaque contains a particular foreign D N A insert, a labeled hybridization probe complementary to a stretch of the desired DNA is used in a hybridization assay. 29 UL Cloning Vectors: pBluescript U KS+/- and XEMBL3 The plasmids, or, more appropriately, phagemid vectors, used throughout this study were pBluescript II KS+/- from Stratagene (Figure 9). These circular vectors of 2961 base pairs contain 21 unique restriction sites in their multiple cloning site (MCS), each capable of accommodating a DNA insert. With an f l filamentous origin of replication and by co-infection with a helper phage, these vectors may be packaged as single stranded phagemid particles (ssDNA). The (+) variant of the vector produces the sense or coding strand of the lacZ gene, while the (-) version produces the non-coding or antisense strand. Selection of recombinants is achieved with both antibiotics and with blue-white color selection. The ampicillin resistance gene selects for only those host cells that have been infected with a phagemid, whether recombinant or native. The presence of a portion of the lacZ gene provides the basis for color selection. As the M C S is in frame with the lacZ gene, insertion of a foreign piece of DNA into this region disrupts P-galactosidase, the protein product of the lacZ gene. A phagemid without an insert will produce enzyme that is capable of cleaving 5-bromo-4-chloro-3-indoyl-(_-D-galactopyranoside (X-gal) into a blue product (Horwitz et al, 1964). Recombinant phagemids, deficient in active P-galactosidase, produce white host colonies in the presence of X-gal, while colonies with only parental vector stain blue. Lambda E M B L 3 (Stratagene) is a genomic replacement vector capable of accepting BamH I compatible fragments that range in the size of 9 to 23 kb. Following double enzymatic digestion with BamH I and EcoR I, the vector produces a 19.3 kb left arm and a 9.2 kb right arm, both with BamH I ends (Figure 10). Liberated from the two arms is a stuffer fragment 30 of 13.7 kb. The basis for recombinant selection arises from active red and gam genes which reside on this stuffer fragment. Wild-type phages containing these genes are unable to grow on P2 phage lysogens. Replacement of the stuffer fragment with a DNA insert changes the phage's genotype to red- gam-, an alteration that allows the phages to grow on lysogenic strains. Nae1131 Figure 9: Schematics of cloning vectors pBluescript II KS+/- illustrating the multiple cloning site, and the ampicillin and lacZ genes. 31 - f o e — C o CD O O CO CO UJ O <D CD LU CD CO Left Arm Stuffer Fragment Right Arm (19.3 kb) (13.7 kb) (9.2 kb) Figure 10: A.EMBL3 high capacity replacement vector used in the construction of genomic libraries. A double digest with EcoR I and BamH I produces the left and right vector arms, both flanking BamH I sites. iv. Standard Enzymes Utilized in Molecular Biology Reverse transcriptase is a RNA dependent DNA polymerase which exhibits 5'->3' polymerase activity. Two different sources of this enzyme are commercially available. One form is isolated from an avian myeloblastosis retrovirus (AMV), while the other is produced from a cloned copy of the Moloney murine leukemia virus (M-MLV) reverse transcriptase gene. Both types are used to synthesize a complementary strand of D N A in the presence of primer and template. Reverse transcriptase and a DNA polymerase are the principal components of the versatile RT-PCR system. Taq DNA polymerase from the thermophilic eubacterium Thermus aquaticus is the thermostable enzyme most widely used in PCR experiments. This enzyme has optimum 5'->3' DNA polymerase activity in the 7 5 - 8 0 ° C temperature range, at a rate that approaches 150 32 incorporated nucleotides per second (Ehrlich, 1989). As thermostable polymerases can withstand repeated exposures to the high temperatures that are required for DNA strand separation (94-95°C) , these enzymes are ideally suited for automatic PCR applications. In addition to T. aquaticus, thermostable DNA polymerases have now been isolated from other Thermus and from several archaebacterial species. Enzymes that recognize and cleave DNA molecules at or near a specific sequence are known as restriction endonucleases. They recognize specific short stretches of nucleotides, often 4 or 6 base pairs long. With a large selection of readily available enzymes, it is possible to produce DNA fragments with discrete lengths and ends. Fragment ends after cleavage are either staggered with an overhang or blunt. Both may be cloned but with quite different efficiencies. When two DNA molecules with cohesive, complementary staggered ends overlap, they temporarily form base pairs. Under proper conditions, the break between the two DNA stands may be repaired by T4 DNA ligase. This enzyme in the presence of ATP catalyses the formation of a phosphodiester bond between a DNA strand with a 5'-phosphate terminus and one with a 3'-hydroxyl group (Weiss et al, 1967). As blunt-ended pieces of DNA can not overlap to base pair, cloning efficiencies with such molecules are reduced significantly. Two important enzymes used in modifying DNA molecules are T4 polynucleotide kinase and the Klenow large fragment of DNA polymerase I. Polynucleotide kinase catalyses the transfer of the y-phosphate group of ATP to the 5'-hydroxyl group of either DNA or RNA (Richardson et al, 1968). This enzyme is commonly used to end-label single or double stranded nucleic acids with 3 2 P - A T P and to phosphorylate synthetic oligomers. Cleavage of DNA polymerase I with subtilisin produces the 76 33 k D a large Klenow fragment which has traditionally been used for sequencing and for second-strand cDNA synthesis. It is also used to chew-back and fill-in 5'-overhangs during blunt-end production. Chewing back is accomplished with its 3 ' -»5 ' exonuclease activity, while filling in is due to its 5'->3' polymerase activity. Sequenase Version 2 .0® (United States Biochemical) is a genetic variant of T7 DNA polymerase which has been specially modified for sequencing applications. This enzyme has the 3'->5' exonuclease activity inherent to wild type T7 polymerase removed. This genetic alteration gives the enzyme the desirable properties of high speed and processing activity without slowing down or stalling as it progresses along the template strand. v. Polymerase Chain Reaction The discovery of thermostable DNA polymerases combined with knowledge acquired from DNA replication studies produced the polymerase chain reaction (PCR) amplification technique. With this ingenious process, a specific segment of DNA or a rare RNA transcript may be amplified millions of times (Mullis et al, 1986, Mullis and Faloona, 1987). In a one tube reaction, a sample of double stranded DNA (dsDNA) or a cDNA:mRNA hybrid is heated to a temperature which melts the two strands (Figure 11). After a brief cooling period, short, synthetic, oligonucleotides each complementary to a region on the two template strands anneal and prime the system for in vitro DNA synthesis. Following hybridization, the primers are extended or elongated by the enzymatic activity of a DNA polymerase such as Taq, which incorporates nucleotide 34 Region to be amplified i>um!|}sssj"jjBsgB28{^  >' Heat [a] » Strand 1 ; Strand 2 • Primer 1 -4 + Cool —"* Primer 2 rrrni • Strand 1 [b] Primer 2 + Primer 1 Strand 2 [C ] * 1 DNA polymerase and nucleotide triphosphates i i i i i i i i i i H i i i i u i i i i i i i i i i i i i i i i i i i i i i i i i i i H i i i i i i i i i i i i i i i i i i i i i i i • Strand 1 i " 1 " " 1 " " 1 1 1 1 1 1 1 1 1 " " " " l l i l l i i l i i l l l l l t + p= New D N A strands 4 1 Strand 2 [d] ^ Heat [c] Strand 1 Strand 2 Primer 1 L P r i T Cool ~ ~ • Primer 2 mm • Strand 1 4 U1UL. Strand 2 DNA polymerase and nucleotide triphosphates Amplification of DNA Figure 11: Amplification of DNA using the polymerase chain reaction. dsDNA is heated causing the two strands to separate, (a) Primers 1 and 2, each complementary to short regions on the template strands, are added to the reaction, (b) Cooling allows the primers to hybridize to the templates, (c) Addition of DNA polymerase and nucleotide triphosphates extends the primers to produce two new copies of the original template strands, (d) The first round of PCR products is heated a second time, creating four strands of DNA. (e) Upon cooling, primers hybridize to both old and new strands in the solution. They are extended once more with DNA polymerase. Subsequent cycles of heating, cooling and primer extension rapidly increase the relative abundance of the target DNA. 35 triphosphates into the templates. Complete extension of primed templates synthesizes a complementary strand of DNA. Amplification products from the first PCR cycle are then subjected to strand separation, followed by primer extension. If adequate amounts of primers and nucleotide triphosphates are used, many alternate cycles of heating, cooling and primer extensions may be carried out. Thermostable DNA polymerases with high fidelity rates make this extremely rapid process highly efficient. The amplified DNA may then be prepared for transfer into a cloning vector. vi Cloning DNA from Libraries or with RT-PCR Regardless of the experimental technique employed, the initial task in cloning any piece of foreign DNA is to acquire the particular template of interest. As virtually all cells in an organism contain the same genetic information, most tissues are good sources of genomic DNA. For isolating mRNA, the choice of tissue can turn out to be crucial, as transcription of certain messages may be tissue specific. To construct a genomic library, DNA that represents the entire genome of an organism is fragmented into shorter pieces with a restriction enzyme. These fragments are then ligated into a phage replacement vector such as ^EMBL3, and encapsulated in vitro to produce active phage particles capable of infecting bacterial cells (Figure 12). The infected cells are plated on agar plates and incubated until plaques are evident against a lawn of bacterial cells. The gene of interest is identified with a labeled probe that corresponds to a small portion of the desired gene in a hybridization assay. 36 Construction of a cDNA library commences with isolation of total RNA, which is subsequently enriched for mRNA. Synthetic primers such as oligo dT or random hexamers and reverse transcriptase convert the mRNA into first strand cDNA. Following second strand synthesis, the cDNA is cloned into an appropriate vector and transformed into bacterial host cells. As with genomic libraries, cDNA libraries are also screened with a hybridization probe. Perhaps the quickest and least labor intensive method of cloning a gene is with the RT-PCR method. With this procedure, target DNA is directly obtainable if gene specific priming is carried out in the reverse transcription reaction. The DNA that is amplified via PCR is cut with restriction endonucleases, ligated into a plasmid or phagemid vector, and transformed into competent host cells. Screening for recombinants is achieved with antibiotics and with blue-white color selection as described previously. Positive clones are readily identified by restriction analysis of purified plasmid or phagemid DNA. Cutting a recombinant plasmid with restriction enzymes that cleave around the fragment in the multiple cloning site will liberate the insert from the vector. 37 I Horse DNA (entire genome) Digested Horse genomic DNA Phage vector arms and T4 DNA ligase Recombinant phage DNA In vitro phage encapsulation I v Plaques I • Bacteria Hydridization assay to identify clone Figure 12: Construction of a horse genomic library with a phage replacement vector. 38 inL Screening by Nucleic Acid Hybridization Screening libraries for a particular piece of recombinant DNA is carried out by the Southern blotting technique (Southern, 1975). Transformed or phage infected cells, grown on agar, are blotted with a membrane made of a material which displays an affinity for these particles, such as nitrocellulose. Once placed into a solution containing sodium hydroxide, the cells or phages break open, which releases their cellular contents, including DNA that adsorbs to the blotting membrane. Alkali treatment of DNA has the secondary effect of denaturing the ds-DNA into single strands. Following neutralization and extensive washing to remove cellular debris, denatured DNA is coupled to the membrane by heat or with either U V or microwave radiation. The membrane with cross-linked DNA is incubated with a single stranded, radioactive hybridization probe. To locate and identify a clone with the target gene, the probe used in the screening process must contain a portion of that gene's sequence. Frequently, the hybridization probe is generated from amino acid sequence data that has been acquired from protein synthesized by the target gene. Probes are also designed from sequence data that has been determined for the gene of interest in a different species. In this study, primers corresponding to the sequence of human plasma gelsolin were made and used to elucidate the sequence of horse plasma gelsolin. As annealing of two DNA strands to form Watson-Crick base pairs requires the DNA initially to be single stranded, the labeled hybridization probe must also be in this form. If the target gene is present on the membrane, it will hybridize with the probe. Non-specifically bound probe is removed by extensively washing the membranes. 39 To pinpoint the locations of radioactivity, X-ray film is placed on top of the membranes. Wherever the radioactive probe has hybridized to the membrane, it will expose the film to produce a dark spot. These dark spots correspond to the colonies or plaques that contain the desired gene. Alignment of the developed film with the master agar plate determines the location of the colony or plaque that contains the cloned gene. Individual colonies or plaques are purified further by conducting additional rounds of screening at a much lower titer. vUL Sequencing with the Chain-Termination Method The chain termination sequencing method involves the in vitro synthesis of a complementary strand of DNA with a polymerase that utilizes ssDNA template (Sanger et'aL, 1977). As with reverse transcriptase and Taq polymerase, DNA synthesis is only initiated at sites where an oligonucleotide primer has annealed to the template. The strand is synthesized by incorporating 2'-deoxynucleoside 5'-triphosphates (dNTPs) until the enzyme encounters a 2',3'-dideoxynucleoside 5'-triphosphate (ddNTPs) chain termination molecule. As the ddNTPs lack hydroxyl groups at both the 2' and 3' positions on the ribose sugar, the polymerase can no longer catalyze the formation of the phosphodiester bond between adjacent nucleotides. Four independent enzymatic reactions, each with a different terminating ddNTP, are run for every segment of DNA that is sequenced. By adjusting the ratios of dNTPs to ddNTPs, only a fraction of the newly synthesized DNA population will be terminated at any one given position. A radioactive or fluorescently labeled nucleotide is included in the synthesis, allowing us to visualize the various 40 sized chains by autoradiography or optical detection following separation by high-resolution electrophoresis. Since the enzymes used for sequencing require a single stranded template, purified plasmid DNA has been subjected to a variety of conditions that denature the two strands prior to conducting a reaction. Alternatively, by using a vector such as pBluescript II, large quantities of clean ssDNA are attainable for sequencing. Single stranded DNA generally produces cleaner sequence bands on gels as compressions or stops in the polymerization reaction are minimized. ix. PCR Techniques Designed to Generate cDNA Ends A n often difficult task when working with cDNA, whether from libraries or RT-PCR experiments, is the isolation of intact, full-length clones. In many cases, the 5'-end of the clone is missing due to the inability of reverse transcriptase to read through the entire gene sequence. Both large mRNA transcripts and ones with extended secondary structures contribute to incomplete cDNA synthesis. This is especially problematic when working with transcripts that are extremely G C rich at their 5'-ends, as are many mammalian genes. To circumvent the time-consuming process of constructing and screening a library, specialized PCR-based methods have been developed to clone the 5'-ends of cDNA. The rapid amplification of cDNA ends or the R A C E method, developed by Frohman et al, 1988, is illustrated in Figure 13. A gene specific antisense primer (GSP 1), designed to a known portion of the desired DNA, is used to prime mRNA in the reverse transcription reaction. After degrading the remaining RNA and removing excess GSP 1, the cDNA 41 sample is tailed with terminal deoxynucleotidyl transferase (TdT). This enzyme generates a homopolymeric anchor on the ends of the cDNA to which a complimentary anchor primer can anneal, and therefore initiate PCR amplification. For the amplification reaction, the use of an additional primer (GSP 2), nested slightly upstream from GSP 1, is generally recommended. A modification of the R A C E technique termed SLIC, for single stranded ligation to cDNA ends, circumvents the homopolymeric tailing procedure (Edwards et al, 1991). This step is replaced by ligating a single stranded oligonucleotide anchor onto first strand cDNA ends with T4 RNA ligase (Figure 14). Following cDNA synthesis with GSP 1 and anchor ligation, the segment of unknown cDNA is amplified in a PCR reaction. A primer complementary to the ligated anchor (the anchor primer) and a nested GSP 2 primer are used in the amplification reaction. To prevent polymerization, the anchor is synthesized without a 3'-hydroxyl group. With both R A C E and SLIC techniques, GSP 1 and GSP 2 primers are chosen just 3' of the missing piece of sequence. This will maximize the probability of the reverse transcriptase enzyme reading through the template to the end of the clone. 42 mRNA 5" cDNA 5" (dC) Anchor Primer (dC) GSP V GSP 2 (A) Prime mRNA n with GSP 1 (A) Copy mRNA into cDNA (A) Degrade mRNA, n purify cDNA Tail purified cDNA with TdT and a dNTP, dCTP in this example Amplify dC-tailed cDNA with oligo dG and GSP 2 in a PCR experiment Figure 13: Outline of the 5'-RACE system. As the cDNA is tailed with 2'-deoxycytosine 5'-triphosphate in this example, the complementary anchor primer would be oligo dG. 43 GSP 1 mRNA 5' AAAAAAAA 3' (a) ss-cDNA V 5" (b) 5" H 2N-i P ^ T4 RNA ligase H-IM-™™ Anchor primer V H-N-l 5" GSP 2 (O (d) Amplified 5'-cDNA segment Figure 14: The SLIC PCR strategy, (a) Synthesis of first strand cDNA is carried out using GSP 1. (b) Excess primer is removed and RNA is hydrolyzed producing ssDNA. (c) Single stranded ligation of an anchor oligonucleotide is accomplished with T4 RNA ligase. (d) The region of DNA is then amplified with a PCR that is conducted with GSP 2 and the anchor primer. 44 A n alternative PCR based method, also designed to produce missing cDNA ends, is inverse PCR (Zeiner and Gehring, 1994). With the R A C E technique, success depends on the quality of the homopolymeric tailing reaction. As all cDNAs in the RACE mixture are tailed, combined with a relatively inefficient TdT reaction, the level of non-specific PCR products from such experiments may be high. For these reasons, the R A C E protocol frequently does not yield the desired products. Similarly, poor reaction efficiencies with RNA ligase and the need to conduct more than one round of PCR experiments all affect the outcome of the SLIC method. With inverse PCR, the steps required to anchor a tail onto the 5'-ends of the cDNA molecules are avoided. As outlined in Figure 15, second strand synthesis in inverse PCR is accomplished following reverse transcription of mRNA with GSP 1. Second strand cDNA is blunt-ended and circularized with T4 DNA ligase. Two primers, GSP 2 and GSP 3, are"subsequently used in the PCR amplification step. This methods amplifies unknown 5' sequence by exploiting regions of known sequence. Since the two primers are arranged in an orientation opposite to that normally employed for PCR, this method is referred to as inverse PCR. 45 GSP1 mRNA Reverse transcription Second strand synthesis cDNA J GSP 2 GSP 3 GSP 3 3' 5' GSP1 Circular DNA PCR i GSP 3 GSP1 GSP 2-AY (A)r ds DNA Figure 15: Schematic representation of inverse PCR. Following second strand synthesis, linear DNA is circularized with T4 DNA ligase, which is used as the template for PCR experiments. The desired 5'-end of the sequence is shaded. 46 3. CIRCULAR DICHROISM SPECTROSCOPY L Principles of Circular Dichroism Superposition of left (L) and right (R) circularly polarized light waves of equal amplitude and frequency produces a beam of linearly polarized light. Vectorial summation of unequal amplitude L and R circularly polarized components produced on passing plane polarized light through a circularly dichroic sample, generates light that is elliptically polarized (Cantor and Schimmel, 1980). As such light traces out an elliptical path, it may be characterized by an angle of ellipticity OF) and by the rotation angle (a) of the ellipse's major axis relative to the plane of polarization of the incident plane polarized light (Figure 16). Figure 16: Ellipticity angle OF) and rotation angle (a) for an originally plane polarized light beam that has emerged from an optically active absorbing medium. 47 Passing L and R circularly polarized light through an optically active medium will result in exiting electromagnetic waves that have propagated through the medium at different speeds and that have different amplitudes. The degree to which the L and R circular polarized components are differentially absorbed is described by the circular dichroism, while optical rotation measures the difference in wave speeds through the sample. Differential absorption expressed iri terms of extinction coefficients for L and R polarized waves, £j_ and £ R , defines the circular dichroism (Ae). A£ = eL- £ R Similarly, optical rotation is related to the difference in the refractive indices, n L and n R , of L and R circular polarized light passing through the medium. This difference in refractive indices, which is directly proportional to optical rotation, is known as the circular birefringence (An). A n = n L - n R Experimentally, CD spectrometers measure Ae, but the ellipticity (*?) is usually reported. This quantity is simply the arctangent of the minor to major axis ratio of the ellipse that characterizes elliptically polarized light. The dependence of a C D reading on path length and on concentration is eliminated by calculating the specific ellipticity [Y] at a given temperature and wavelength. This quantity, expressed in units of deg-cm 2-dag' 1 , is defined as [Y] = *F/l-c 48 where 1 is the light path length in dm and c is the concentration of the optical active substance in gem" 3 . Comparisons of molecules with different molecular masses produces the molar ellipticity [6], [0] = m-M/100 where M is simply the molar mass of the C D active compound. Molar ellipticity is expressed in units of deg-cm 2-dmol' 1 . Substitution of M in the above equation with a mean residual mass defines the mean residual ellipticity, a quantity which is frequently reported when working with protein samples. The relation between the molar ellipticity and the measured circular dichroism is [0] « 3 3 0 0 ( £ L - £ R ) . ii. Circular Dichroism Spectroscopy of Proteins Optically asymmetric conformational orientations and asymmetric atomic centers are the constituents that give rise to chiroptical properties of molecules. The predominant contributor to the far-UV circular dichroism signal of a protein is the amide chromophore that resides in a dissymmetric environment. Of the various conformational elements that are found in a protein, amides in an a-helix motif display the strongest CD absorbances. Right-handed oc-helical peptides and proteins are characterized by strong, negatively dichroic bands at 222 nm and 210 nm, with a strong positive peak near 190 nm (Johnson, 1990). A (3-sheet generally displays a negative band at 216-218 nm and a stronger positive band near 195 nm. In the near-UV spectral region (240 nm to 350 nm), aromatic amino 49 acids and disulfide bonds are the major contributors to the C D signal. The major tryptophan dichroic band is located around 300 nm, while those of tyrosine and phenylalanine occur at 280 nm and 260 nm, respectively. Mean residual ellipticities of the chromophores active in the near-UV spectral region are small compared to those of the amide groups. i UL Prediction of Secondary Structure Qualitative and quantitative analysis of a protein's secondary structure may be done using CD, as this form of spectroscopy is extremely sensitive to the proportions of a-helix, J3-sheet, and unordered conformations. A direct way of analyzing a CD spectrum of a protein is to fit it to C D spectra of the various secondary structures. In the simplest model, it is the helix, sheet, and unordered conformations that display the majority of protein circular dichroism. Assuming that the differential absorbances due to these secondary structures are additive, the fractional composition of each of these elements f H , fp,and fjjN may be estimated using the following relations, [Q]x = f H X H + fpXp + f U N X U N and Zfi = 1 where [9] is the observed mean residual ellipticity at a given wavelength, and X H ,Xp, and X T J N a r e the mean residual ellipticities of pure a-helical, (3-pleated sheet, and unordered structures, respectively. Pioneering contributions by Greenfield and Fasman produced the first estimates for X H ,Xp, and X U N by analyzing C D spectra of synthetic reference polypeptides that were presumed to-exist in only one of the three discussed conformations (Greenfield and Fasman, 1969). Since their 50 initial work, numerous CD methods that focused on deducing the conformational weight or percentage composition of the secondary structural elements of a protein have been published. iv. Convex Constraint Analysis Convex constraint analysis (CCA) is an algorithm that was developed to extract the chiral contributions of secondary structural elements directly from experimental C D curves (Perzcel et al, 1991, 1992, 1992a). This method differs from others as it does not make use of mean residual ellipticity values that originated from model peptides or proteins of known secondary structure. Analyzing a set of C D data simultaneously and extracting common features among them eliminates some of the problems that are associated with methods that rely on template sets. As with other algorithms, C C A assumes additivity of the C D contributions of all the chiral secondary conformations. The measured CD [Q]\ at any wavelength may be decomposed in the form [ < % = £ f i x i i=l where P represents the number of pure secondary structure components, Xi(^) is the CD of the i t h component at wavelength X, and ft is the weight or fractional composition of the i t h pure component. Additional constraints to the algorithm are that all conformational weights must be positive, and that the sum of the weights must equal 1. Appending experimental C D data 51 to a reference data bank of existing proteins (Yang et al., 1986) produces a 26 member data set that is deconvoluted simultaneously. Upon a priori input of the number of pure components expected to generate the protein's C D signal, the algorithm calculates a set of pure component spectra X\{\) along with a series of conformational weights f\. The validity of the conformational weights can be correlated directly with those obtained by other methods such as X-ray crystallography. 52 4. DENATURATION STUDIES L Protein Folding One of the most important problems in physical biochemistry is the elucidation of the mechanism by which proteins fold to their active three dimensional conformation. Since the discovery that the three dimensional fold of a protein is dictated by its primary amino acid sequence (Anfinsen, 1973), a number of investigators have challenged the protein folding problem using both thermodynamic and kinetic arguments. Experimentally, this problem is addressed by either inducing a native protein to unfold, or by monitoring refolding of a previously denatured protein. According to the principle of microscopic reversibility, the transition state for the folding reaction is the same as that for unfolding, thus both denaturation and renaturation experiments can be utilized to elucidate the protein folding problem. The event of unfolding a native protein has been analyzed to a much greater extent than has the reverse process of a denatured protein refolding to its native conformation. The most commonly used agents for studying the equilibrium unfolding process are guanidine hydrochloride (Gnd-HCl) and urea (Pace, 1986). Unfolding in the presence of a chaotropic agent can be explained by an increase in the affinity of internal residues for the denaturing solvent (Tanford, 1964). Solvation of those amino acid side chains that are buried in the folded conformation but exposed in the unfolded state stabilizes the denatured protein (Staniforth et al, 1993). In addition to the side chains, Gnd-HCl and urea molecules are also thought to solvate the amide groups of the protein. 53 ii. Equilibrium Unfolding Curves Unfolding a compact, globular protein is a cooperative process. A sigmoidal curve is usually observed with respect to the concentration of a denaturant or temperature as the native (N) state is converted to its denatured (D) state. Analysis of such curves produces several important parameters pertinent to the folding mechanism and conformational stability of proteins. If the denaturation event of a protein proceeds through a reversible two step mechanism, as that of many globular proteins has been observed to do, it is possible to get an estimation of a protein's conformational stability. The two step reversible process assumes interconversion of the N and D states (N<=>D), where only they are present at a significant concentration in the transition region. A l l other intermediate conformational states are only marginally populated. Conformational stability of a protein is defined as the change in free energy required to convert a native protein in H2O to the denatured state H 2 0 ( ^ G u )• This value may be estimated by assuming a linear dependence ° f A G U on the denaturant concentration in the transition region. A two state mechanism follows the relation H 2 0 A G U = A G u - m [denaturant] H 2 0 were A G u is the apparent free energy of unfolding and A G U is the apparent f r e e energy of unfolding in pure water. The slope of the line, m, is related to the number of denaturant molecules that bind to the protein upon denaturation (Pace, 1986): Extrapolation to zero denaturant concentration produces an estimate of the conformational stability. 54 5. PROTEIN CRYSTALLOGRAPHY £ Producing Protein Crystals Of the various physical and chemical methods that generate macromolecular structural information, X-ray diffraction analysis of single protein crystals undoubtedly provides the most detailed and precise description of a protein's three dimensional structure. At a resolution of 3 A or better, a protein crystal structure provides information on the location of peptide bonds and amino acid side chains, and suggests locations of hydrogen bonds, the degree of solvent accessibility, and positions of mobility or flexion in the molecule. At 2 A or better resolution, individual atomic positions become defined. The primary obstacle that limits the use of this powerful tool in the study of protein architecture is the ability to produce protein crystals. Once grown, the crystals must be of a quality suitable for high resolution X-ray diffraction analysis. The influence of environmental factors such as ionic strength, p H , temperature, along with others, on protein conformations can make these molecules extremely difficult to crystallize (McPherson, 1990). Structurally, proteins in solution are extremely complex as they are dynamic and made up of a heterogeneous population of slightly different structures. Adding to an already large array of variables that influence protein structure is our incomplete understanding of the many weak intermolecular forces that contribute to the crystallization phenomenon. The thermodynamic force that drives the crystallization process is obtained from a non-equilibrium or supersaturated solution returning to an 55 equilibrium state. In a solution saturated with respect to protein, both solid and soluble forms of the protein are in equilibrium. In this situation, a net increase in the proportion of the solid phase could not occur without an equivalent amount of existing solid dissolving. In order for crystal growth to occur, the protein solution of interest must initially be brought to a state of supersaturation, whereby its return to equilibrium drives the protein from solution to the solid crystalline phase. The stage prior to continuous crystal growth from supersaturated solutions is the formation of stable nuclei. Such nuclei are molecular aggregates of a specific size and physical composition such that solute molecules in solution will grow on their surfaces at a rate faster that the rate of dissolution of the nuclei. Experimental conditions which produce a limited number of nuclei are generally desired in order to promote the formation of crystals over those of random aggregates which lead to precipitates. The classical strategy employed for inducing protein crystal growth has been to gradually increase the concentration of a salt or other protein precipitating agent in a concentrated protein solution. Such agents effectively reduce the water content from the protein's hydration shells, which promotes the transfer of the protein from solution into the solid phase. Traditionally, salts such as ammonium sulfate have been used as "salting out" agents, but other substances such as organic solvents, long-chain polymers, low molecular weight polymers and non-volatile organics have all been used to induce protein crystallization (McPherson, 1990). Experimentally, a concentrated protein solution, in the form of a sitting or hanging drop, is brought to a state of saturation by vapor diffusion (Figure 17). Both these techniques rely on the transport of water 56 or solvent from the mother liquor of a microdrop to a much larger reservoir of buffer. Microdrops usually consist of a few microliters of a concentrated protein solution mixed together with a few microliters of the reservoir buffer that contains the precipitating agent. Once the microdrop is sealed from the atmosphere, solvent or water will migrate as vapor from the area of low precipitant (microdrop) to the area of high precipitant (buffer reservoir) until equilibrium has been reached. As the solution volume of the microdrop is continually being reduced, the concentration of the precipitating agent and of the protein increase to the point where a supersaturated solution capable of driving crystal growth is created. Inducing crystal growth in such small microdrops allows many crystallization conditions to be screened without using large quantities of protein sample. Protein 30Iution and precipitating agent suspended under cover 3lip Protein sol ution and precipitating agent supported by plastic micro bridge Hanging Drop Sitting Drop Figure 17: Hanging and sitting drop experiments used to survey protein crystallization conditions. Experiments are conducted in the wells of tissue culture plates which have been sealed with cover slips and vacuum grease. Drops usually consist of a few microliters of protein solution while the buffer reservoir volume is typically 1 mL. 57 it Solving a Protein Crystal Structure Although protein X-ray crystallography is a highly specialized field, a general sequence of experimental steps that will hopefully result in structural information is followed for most proteins under investigation. The ultimate goal of diffraction experimentation is the reconstruction of the crystal's asymmetric unit from the diffraction pattern. Having established reproducible protein crystallization conditions, data collection, phasing, electron density map interpretation and structure refinement are the crucial steps required to arrive at a crystal structure. Data collection is conducted with a single crystal which is irradiated with a continuous beam of X-rays of a wavelength comparable to the desired resolution of the structure, typically 1.54 A. As a single crystal is composed of a well-ordered array of individual molecules numbering in the billions, X-rays that are scattered by any single repeating unit in the lattice are additive to those scattered by all other copies found in the crystal. The intensity of each diffraction spot at any angle is the result of constructive and deconstructive interference from X-rays scattered by various atoms in the molecule. Each individual diffraction spot contains information on the entire atomic structure. Since crystals contain a periodic distribution of scattering matter, their diffraction data may be represented by a Fourier summation. Through this mathematical transformation, the electron density at any position in the asymmetric unit may be described by a particular combination of wave amplitudes and phases. Of these two independent physical parameters, amplitudes are obtained directly from intensities of the diffraction spots, while special techniques are required to make phase estimates (Glusker and Trueblood, 1985). 58 Reconstruction of a molecular structure from diffraction data requires values for both phase angles and amplitudes. As the data collection process with X-ray diffraction studies can only yield the amplitudes of structure factors, techniques have been developed to estimate phase angles. A n established method for phase estimation is with the isomorphous replacement technique. To solve the phase problem, a few sites on a protein molecule are derivatised with heavy metals. Introduction of these good X-ray scattering atoms changes the intensities of the diffracted X-rays without significantly changing the structure of the molecule. The differences between the diffraction intensities of the heavy atom derivative and of the native protein are used to reveal the positions of the heavy atoms. This information allows the phases of the diffracted X-ray waves to be calculated. Once both native and derivatized data sets have been collected and phasing estimates have been made, an electron density map of the repeat unit is calculated. A n initial model of the protein is then constructed by fitting the known amino acid sequence to the electron density of the unit cell. The initial structure is refined using established stereochemical restraints and by introducing bulk and ordered waters to produce the final three dimensional picture of the protein molecule in the crystalline state. 59 CHAPTER H MATERIALS AND METHODS 1. RECOMBINANT DNA METHODOLOGY L Bacterial Strains, Vectors, and Media The lysogenic E. coli host strain utilized in screening the horse genomic library was X L 1-Blue MRA (P2) from Stratagene {A(mcrA)183 A(mcrCB-hsdSMR-mrr)173 endAl supE44 thi-1 gyrA96 relAl lac}, while E. coli D H 5 a F ' {F ((>80dlacZAM15 A(lacZyA-argF)U169 e n d A l r e c A l h s d R 1 7 ( r K " m K + ) deoR thi-1 supE44 k-gyrA96 relAl} was used for all subcloning and DNA production purposes. A horse genomic library was established in the Lambda E M B L 3 replacement vector (Stratagene), which is supplied in a form of a left and right phage arm, along with an intervening stuffer fragment. Genomic DNA cut with restriction endonucleases that produce B a m H I compatible cohesive ends are cloneable into this lambda phage vector. Wild type lambda phages which contain viable red and gam genes are unable to grow on P2 lysogenic host strains. When the stuffer fragment is replaced by a foreign DNA insert, the recombinant E M B L 3 phage is converted to a genotype capable of growing on P2 lysogenic strains. Cloning of PCR fragments, production of single stranded DNA, and subcloning of genomic clones was achieved with the phagemid vectors pBluescript II K S + and KS" (Stratagene). Luria broth (LB) consisting of 5 g of yeast extract, 10 g of bacto-tryptone and 10 g of NaCl per litre at pH 7.5 was-used as the liquid growth medium for both XLl -Blue MRA (P2) and DH5aF' strains. The medium for 60 the X L 1-Blue MRA (P2) strain was supplemented with 10 m M MgSC>4 and 0.2% (w/v) maltose. Lambda phage infected E. coli cells were grown on NZY media which contained 5 g NaCl, 2 g MgS04-7H 2 0, 5 g yeast extract and 10 g NZamine per litre of water at pH 7.5. E. coli DH5ccF' cells containing recombinant plasmids were propagated on LB plates supplemented with 12.5 |j.g/mL ampicillin, 25 |ig/mL isopropyl-p-D-thiogalactopyranoside (IPTG), and 50 n g / m L 5-bromo-4-chloro-3-indoyl-p-D-galactopyranoside (X-gal) (Sambrook et al, 1989). The agar content in all plates was 1.5% (w/v), with a 0.7% (w/v) top agar overlay for the lambda phage library plates. ii Electrophoresis Samples of DNA were analyzed by agarose gel electrophoresis in the presence of 1 |ig/mL ethidium bromide under UV illumination. The size of the DNA fragment being analyzed determined the concentration of agarose in the gels. For most applications, the agarose concentration was between 0.7% to 2.0%. Samples were loaded in 10% glycerol, 0.1% xylene cyanol, 0.1% bromophenol blue, and run in IX T A E buffer (40 m M Tris-HCl, pH 7.5, 20 m M acetic acid, and 1 mM EDTA). Denaturing agarose gels (1%) in 40 m M morphopropanesulfonic acid, p H 7.0, 10 m M sodium acetate, 1 m M E D T A , and 0.44 M formaldehyde were used to assess the quality of RNA preparations. Prior to loading, 10 (iL of each RNA sample was heated to 68°C for 10 minutes with a solution of 2 \iL 5X RNA running buffer, 3.5 |xL formaldehyde, and 10 |iL formamide (Sambrook et al, 1989). Sequence data were obtained from 6% (w/v) polyacrylamide gels 61 that contained 50% urea. Polymerization was initiated by the addition of ammonium persulfate to 0.066% (w/v) and T E M E D to 0.04% (v/v). Gels with dimensions of 38.5 cm x 31.0 cm x 0.8 mm were run on a Model S2 Sequencing Apparatus (Gibco BRL) in 89 mM Tris-base, 89 m M boric acid, and 2.5 m M EDTA, pH 8.3. Electrophoretic run times were between 1.5 and 7.0 hours. Gels were dried under vacuum at 80°C, and exposed to Kodak XRP-1 film for 1-3 days. iit Oligodeoxyribonucleotide Design and Synthesis All oligodeoxyribonucleotides used in this study were designed with the aid of the computer program Oligo™ Version 4.0 (National Biosciences Inc.), and synthesized on an Applied Biosystems 391 DNA Synthesizer using phosphoramidite chemistryi Regions that displayed strong sequence identity between human and pig plasma gelsolin cDNAs were chosen as template sequences for synthetic primers (Kwiatkowski et al, 1986; Way and Weeds, 1988). Primer sequences are listed in Table 1, with numbers corresponding to the base pair positions on human gelsolin cDNA: GenBank Accession No. X04412. iv. Extraction of Genomic DNA A 5 g sample of frozen horse liver (Pel-Freez Biologicals) in liquid nitrogen was ground into a powder with a mortar and pestle. After removal of the liquid nitrogen, the minced tissue was mixed with 15 mL of a buffer that contained 10 m M Tris-HCl, pH 8.0, 100 m M EDTA, 0.5% SDS and 20 |ig/mL bovine pancreatic RNase A. The mixture was shaken until all visible material was in solution, then incubated for 1 hour at 37°C. Proteinase K 62 (Sigma) was added to the suspension of cells to a concentration of 100 |ig/mL. Proteolytic digestion of the sample continued for 3 hours at 50°C with occasional gentle Table 1: Synthetic oligodeoxyribonucleotide primers used in generating the sequence of horse plasma geteolin Primer Sequence Position Gelsolin 1 5'- GAA CAC C C C GAG TTC CTC AA-3' sense 177-196 Gelsolin 2 5' -TGC A C C ATT GGA GAC CTT GT-3' anti-sense 916-935 Gelsolin 3 5' -GCA A G C T G G CCA A G C TCT AC-3' sense 898-917 Gelsolin 4 5' -ACC CAG ACA AAG A C C T G G TC-3' anti-sense 2127-2146 Gelsolin 5 .5' -ACC GAT G C C TTT GTT CTG AA-3' sense 1785-1804 Gelsolin 6 5'-TCC T C A G G G A G C GAG CTC AG-3' anti-sense 2147-2166 Gelsolin 7 5' - G C C GTG T C G CCA CCA T G G CT-3' sense 1-20 Gelsolin 8 5' - T G G C G C TGT G C G C G C TGT CG-3' sense 61-80 Gelsolin 9 5'--CCG CTC T C A T C C T G G C T G CA-3' anti-sense 372-391 agitation. After cooling to room temperature, the solution was extracted with an equal volume of buffer saturated phenol. The two phases were separated by centrifugation at 10 OOOg for 10 minutes, and the DNA containing aqueous phase was decanted. Following two more phenol extractions, addition of 3 mL of 10 M ammonium acetate and 30 mL of 95% ethanol precipitated the DNA from the aqueous phase. The precipitate was pelleted with a spin at 5000g for 5 minutes, washed twice with 70% ethanol, then resuspended in 2 mL of 10 m M Tris-HCl, pH 8.0, and 1 m M EDTA. DNA resuspension was facilitated by placing the sample 63 on a rotating platform for 24 hours. High molecular weight genomic DNA was run out on 0.7% agarose gels . v. Genomic Library Construction/Phage Packaging Approximately 10 |ig of horse genomic DNA from liver tissue was digested for 5 minutes at 37°C with 1 unit of Nde II (Gibco BRL) in a reaction volume of 50 [ih. Electrophoretic analysis of the partially cut DNA on a 0.7% agarose gel revealed a uniform smear that extended from the top of the lane well down to the 500 base pair marker. The DNA that migrated between the 21 kb and 12 kb base pair markers was excised, electroeluted into a piece of dialysis tubing, precipitated, and finally resuspended in 10 m M Tris-HCl, pH 8.0 and 1 m M E D T A . A ligation consisting of 1 |J.g of B a m H I predigested A.EMBL3 vector with 0.3 |ig of size selected genomic DNA, 2 units of T4 DNA ligase (Gibco BRL) and 1 m M ATP was set up to a final volume of 5 |xl. The ligation reaction was conducted overnight at 15°C. Gigapack Plus® packaging extracts (Stratagene) were used for in vitro recombinant phage encapsulation. Immediately upon thawing, 1 |i,L of the overnight ligation mixture was added to a tube of Freeze-Thaw extract, followed by 15 |iL of Sonic extract. The sample was mixed and left undisturbed at room temperature. After two hours, packaged phage was suspended in 500 fiL SM buffer and 20 (iL chloroform, and stored at 4°C. vi Library Titering A fresh 5 mL overnight culture of E. coli X L l - B l u e MRA (P2) was 64 pelleted then resuspended in 10 mM MgSCU to an OD600 = 0.5. The stock phage library was diluted consecutively by a factor of ten with SM buffer, down to a final dilution of 10"5. Mixtures of 200 jiL of host strain and 1 pL of each library dilution were incubated at 37°C for 15 minutes. Following phage attachment, 3 mL of liquefied top agar at approximately 4 5 ° C were added to each sample and poured onto 10 cm NZY agar plates. After 12 hours of incubation, phage plaques were visible against a confluent lawn of bacterial cells. Quantitative analysis of the serial dilutions yielded a titer of approximately l x l O 9 plaque-forming units/mL of library. vii. Plaque Lifts/DNA Blotting Five samples, each consisting of 100 fiL of the stock horse genomic library and 2.5 mL of the resuspended lysogenic host strain, were incubated for 15 minutes at 37°C. Liquefied agar (50 mL) was added to these samples, which were then plated onto 22 cm x 22 cm NZY plates. The five plates were incubated until the first visual signs of phage plaques were evident, usually after six hours. The experiment was designed to produce approximately 400 000 independent plaques per plate, which yielded a total screen of about 2 x l 0 6 recombinants. Duplicate plaque lifts with 20 cm x 20 cm, 0.45 lum Protran™ nitrocellulose membranes (Schleicher & Schuell) were carried out on each plate. The first lift was done for 1 minute, and the second, with a new membrane, for 3 minutes. Membranes with attached phage were placed lift-side facing upwards onto fresh NZY feeder plates and incubated overnight, while the master plates were stored at 4°C. Plaques adhering to the nitrocellulose membranes were lysed by a 2 65 minute submersion in a solution of 1.5 M NaCl and 0.5 M NaOH. The membranes were then neutralized for 2 minutes in 1.5 M NaCl, 0.5 M Tris-HCl, pH 8.0 buffer, and finally rinsed with a solution of 0.2 M Tris-H C l , pH 8.0 and 2X SSC. After blotting the membranes with filter paper, adsorbed DNA was crosslinked to them with a 30 second stream of microwaves from an oven set to medium power (Angeletti et al., 1995). viii. Preparation of Radiolabeled Hybridization Probes Hybridization probes were labeled to a high specific activity using the T7 QuickPrime™ Kit from Pharmacia. 50 ng of dsDNA, heated to promote strand separation, were labeled with 50 |iCi [oc-32P]-dCTP using T7 DNA polymerase and random priming. Enzymatic incorporation of the "label was completed in 15 minutes at 3 7 ° C , after which 2 [iL of the reaction mixture were removed and precipitated with 1 mL of 10% trichloroacetic acid. The precipitate was filtered through a glass fiber membrane, and the extent of label incorporation analyzed quantitatively by scintillation counts. Multiple labeling reactions consistently produced probes that had specific activities in the range of l x l O 9 cprri/|ig of DNA. Prior to being used for hybridizations, the probes were first heated to 100°C for 5 minutes, then immediately cooled on ice. Single stranded synthetic oligonucleotides were phosphorylated by a modified procedure described by Sambrook et al, 1989. In a 10 fiL reaction, 1 p.L of stock T4 polynucleotide kinase (Pharmacia) with IX One-Phor-all buffer was added to 1 ug of primer, along with 7 |J,L of [y- 3 2P]-ATP at 3000 Ci /mmol . After 45 minutes at 37°C, the labeling reaction was terminated with the addition of 0.1 volume of 0.5 M E D T A and heat to 66 68°C for 10 minutes. A small sample of the reaction mixture was run out on a 3% agarose gel, which was subsequently exposed to XPR-1 film for 5 minutes. The developed autorad revealed a narrow fast migrating band, indicative of successful labeling. Unincorporated nucleotides were not removed from these preparations. be Library Screening A 100 mL prehybridization solution of 6X SSPE, 0.5% SDS, 5X Denhardt's, and 100 |ig/mL denatured salmon sperm DNA was added to the ten nitrocellulose membranes which had been placed inside a plastic bag. Overnight prehybridization in a 55°C water bath was carried out after the bag had been heat sealed. Following prehybridization, the membranes were first washed with a hybridization buffer of 6X SSPE, 0.5% SDS, 2.5X Denhard's, and 100 fig/mL denatured salmon sperm DNA, then incubated once more in a sealed plastic bag with 100 mL of hybridization solution and 50 (iL of [a-P 3 2 ] -dCTP labeled hybridization probe. The hybridization step was also conducted overnight in a 55°C water bath. Non-specifically bound probe was removed from the nitrocellulose membranes by consecutive 5 minute washes with 2X SSPE and 0.1% SDS at temperatures of 50°C (IX), 55°C (4X), 60°C (IX), and at 65°C (3X). A final stringent wash was carried out with IX SSPE and 0.1% SDS at 65°C. Blotted membranes were wrapped with cellophane, sandwiched between two Cronex Lightning Plus intensifying screens (DuPont) with a sheet of X-ray film, and left to expose for two days at -70°C. Positive plaques, as identified by aligning the developed autorads with the master plates, were excised as solid plugs from the agar and mixed with 500 \ih of SM buffer and 20 fiL of 67 chloroform. A second and third round of refined screening at a much lower plaque density was conducted for each potential positive clone as described previously, with the exception of using 15 cm diameter NZY plates and appropriately sized circular membranes. These additional rounds of screening did not require the feeder-plate amplification step. After the third screen, potential positive phage clones were propagated in lysogenic E. coli to produce DNA for blotting experiments. Following electrophoresis of purified phage DNA, the agarose gel was soaked in 0.2 M HC1 for 10 minutes to promote strand nicking. It was then transferred to a denaturing solution of 1.5 M NaCl, and 0.5 M NaOH for 45 minutes with gentle agitation. After several water rinses, the gel was neutralized with a solution of 1 M Tris-HCl, pH 7.4 and 1.5 M NaCl for one hour. A final 30 minute soak in 10X SSC conditioned the DNA in the agarose gel for transfer. DNA was transferred to nitrocellulose using the protocol given by Sambrook et al, 1989 for the simultaneous transfer of DNA from a single agarose gel to two nitrocellulose filters. DNA crosslinking, along with the prehydridization and hybridization procedures, were identical to protocols previously described. One of the two nitrocellulose membranes was probed with an [cc-P 3 2]-dCTP labeled gelsolin cDNA fragment, while the other was hybridized with a mixture of P 3 2 end-labeled gelsolin #7 and #8 primers. The membranes that were screened with end-labeled oligonucleotide probes were washed three consecutive times at 37°C with 2X SSPE and 0.1% SDS. x. RNA Extraction Uterus tissue {Equus caballus) was harvested from a freshly 68 slaughtered horse and frozen on site at - 1 9 6 ° C (Pel-Freez Biologicals). Total RNA was extracted from this tissue using a modified version of the protocol developed by Chomczynski and Sacchi, 1987. All appropriate precautions were taken to prevent endogenous RNAse contamination, including treating water and all disposable plastic tips with diethyl pyrocarbonate (DEPC) (Leonard et al, 1970). 5 g of frozen uterus were ground into a powder under liquid nitrogen, and homogenized at high speed for 30 seconds with a Tissumizer (Tekmar) in a 50 mL solution of 4 M guanidinium thiocyanate, 25 mM sodium citrate, pH 7.0, 0.5% sarcosyl, and 0.1 M [3-mercaptoethanol. The homogenate had 5 mL of 2 M sodium citrate added to it, along with 50 mL of buffer saturated phenol. This was followed by 10 mL of a 24:1 mixture of chloroform:isoamyl alcohol. The sample was shaken vigorously, and put on ice for 15 minutes. Phase separation was facilitated by a 20 minute spin at 10 OOOg, after which the aqueous phase was decanted and mixed with 50 mL of isopropanol. Precipitation of the sample was carried out for 60 minutes at -20°C. The RNA precipitate was recovered by centrifugation at 10 OOOg for 20 minutes. Pelleted RNA was resuspended in 3 mL of the above extraction buffer, then reprecipitated with the addition of 6 mL of isopropanol. Again, precipitation was performed at -20°C for a 60 minute period. Following centrifugation at 10 OOOg for 10 minutes, the pellet was first washed with 75% ethanol, then spun one additional time. The purified RNA sample was dissolved in DEPC-treated water and stored at -70°C. The quality of the isolated RNA was assessed by formaldehyde-agarose electrophoresis (Sambrook et al, 1989). 69 xi First Strand cDNA Synthesis Enzymatic synthesis of first strand cDNA was carried out with methylmercury (II) hydroxide denatured RNA and Moloney murine leukemia virus (M-MLV) reverse transcriptase (Gibco BRL). Aliquots of 1 |ig of total uterus RNA were primed with either oligo-dTis, random hexamers, or with gelsolin specific primers under conditions similar to those described by the manufacturer. The RNA in 5 |iL was mixed with 5 JLLL of a 60 m M methylmercury (II) hydroxide solution and with 5 |nL of D E P C treated d H 2 0. The sample was left at room temperature for 10 minutes, then placed at -70°C for 15 additional minutes. Manufacturer's 5X reverse transcription buffer (10 \iL), 10 m M stock dNTP solution (1 jiL), 1.0 M DTT (1 fiL), primer at 1 n g / | i L (1 ^L) and two hundred units of enzyme were added to the denatured RNA to a final volume of 50 \iL. The reaction proceeded for 60 minutes at 37°C, then for 30 minutes at 42°C, followed by 10 minutes at 50°C. Two hundred additional units of M-MLV reverse transcriptase were added to the reaction after the initial 60 minutes of incubation. Synthesized cDNA was purified with the GlassMax™ DNA Isolation Spin Cartridge System (Gibco BRL) according to the manufacturer's recommendations. xii. Polymerase Chain Reaction Approximately 50 ng of first strand cDNA were amplified in 67 m M Tris -HCl , p H 8.8, 16.6 m M ( N H 4 ) 2 S 0 4 , 10 m M p-mercaptoethanol, 1.5 m M M g S 0 4 , 20 pmoles each of two primers, 1- unit of Taq polymerase (Perkin-Elmer Cetus), and 100 m M each of dATP, dCTP, dGTP, and dTTP 70 (Innis et al, 1990). Each PCR reaction of 50 |iL was overlaid with an equal volume of mineral oil. Gelsolin fragment 1 was amplified with proof-reading V e n t ® DNA polymerase according to the manufacturer's recommendations (New England Biolabs). Thermocycling reactions were carried out in a Perkin-Elmer Cetus Model 480 DNA Thermocycler. After an initial 5 minute denaturation period at 94°C, reactions were conducted with the following cycling profiles: Gelsolin fragment 1, 9 4 ° C for 45 seconds, 63°C for 30 seconds, and 72°C for 30 seconds; Gelsolin fragment 2, 9 4 ° C for 60 seconds, 59°C for 60 seconds, and 72°C for 60 seconds; Gelsolin fragment 3, 94°C for 45 seconds, 55°C for 45 seconds, and 72°C for 60 seconds. A total of 35 amplification cycles were performed with each of the above conditions, which were then followed by a final 10 minute extension at 72°C. xiiL cDNA Cloning/Plasmid Transformations Calcium chloride treated E. coli D H 5 a F ' cells were used to transform recombinant plasmid DNA (Messing, 1993). Host cells were prepared from a fresh 50 mL inoculum that was grown to mid-log phase (OD600 = 0.5) at 37°C. Following centrifugation at 4000g for 10 minutes at 4°C, the pellet was resuspended in a cold 50 m M CaC_2 solution (25 mL) and placed on ice for 1 hour. The treated cells were centrifuged once more for 10 minutes, then resuspended in 5 mL of cold 50 m M C a C l 2 . Cells were aliquotted into 200 (J.L portions, frozen in a dry ice-acetone bath, and stored at -70°C until required. For each gelsolin PCR fragment that was cloned, the DNA from six independent amplification reactions was pooled and gel purified on a 1% 71 agarose gel. Purified DNA in 10 |iL of T E buffer was added to a solution containing 2 |J.L of 10 mM DTT, 2 jxL of 10X Klenow buffer, and 1 uL of stock Klenow DNA polymerase I (Gibco BRL) in a final reaction volume of 20 (xL. The blunt-ending reaction of Klenow proceeded for 15 minutes at 37°C. Overhangs created by the enzyme were repaired by the addition of 1 UL of a 10 m M dNTP solution and 10-20 additional minutes of incubation. Blunt-ended DNA was precipitated with 0.1 volume of 7.5 M ammonium acetate, pH 7.5 and 1 volume of isopropanol, then spun at 10 OOOg for 10 minutes. The pellet was washed with 200 |iL of ethanol, then resuspended in 20 fiL of water. Blunt-ended gelsolin cDNA was added to EcoR V cut pBluescript II K S + or KS" vector at approximately a 3 to 1 ratio, followed by the addition of 1.5 units of T4 DNA ligase (Gibco BRL) and ligation buffer to a final volume of 15 p.L. Ligations were incubated at 15°C until the following day. The entire ligation reaction was added to 50 [ih of thawed competent host cells and placed on ice for 30 minutes. After heat-shocking the cells for 2 minutes at 42°C, the samples were mixed with 100 fiL of LB broth and incubation at 37°C for 45 minutes. Transformed cells were then spread onto ampicillin, Xgal, and ITPG supplemented LB plates. After overnight incubation at 37°C, blue-white color selection was utilized to identity recombinants. DNA fragments that were subcloned from the horse genomic library were produced by random restriction endonuclease digestion. Following XEMBL3 DNA restriction digestion with various six base pair cutting endonucleases, the reaction samples were precipitated with 0.1 volume of 7.5 M ammonium acetate, pH 7.5, and 1 volume of isopropanol. After pelleting and resuspending the cut DNA, a ligation into linearized pBluescript II K S + or KS" was set up as described previously. 72 xiv. Isolation of Double Stranded Plasmid DNA Double stranded plasmid DNA was isolated from 5 m L overnight cultures of E. coli grown in LB and 12.5 jig/mL ampicillin according to the basic procedure described by Birnboim and Doly, 1979. Cells were pelleted at 6000g for 10 minutes in a Sorvall RT6000B benchtop centrifuge, then resuspended in 200 (iL of glucose buffer. After a few minutes at room temperature, the cells were lysed with the addition of 400 pJL of 0.2 M NaOH and 1% SDS. Samples were gently mixed and the viscous solution put on ice for 5 minutes. Precipitation of bacterial genomic DNA was completed by the addition of 300 fiL of 3 M sodium acetate, pH 4.8. Cellular debris and genomic DNA were removed by centrifugation at 10 OOOg for 5 minutes. Clarified supernatant (750 |iL) was transferred to a new Eppendorf tube and mixed with 1 m L of isopropanol. Plasmid DNA was harvested from this solution with a 10 minute spin at 10 OOOg. The sample was washed with 200 |iL of 70% ethanol, and resuspended in 100 |iL T E buffer. xu. Isolation of Single Stranded DNA A sterile 25 mL solution of LB broth with 12.5 |ig/mL ampicillin was inoculated with 100 |iL of an overnight E. coli culture harboring a K S + or KS" recombinant phagemid vector. After 4 hours of incubation at 37°C, M13K07 helper phage (50 (xL), with a titer of approximately l x l O 9 pfu/mL (Gibco BRL), was added to the culture. Following an additional 60 minute Incubation period, selection for infected cells was accomplished by addition of kanamycin to 70 fig/mL. Phagemid infected bacterial cultures 73 were grown overnight at 37°C with vigorous shaking. A 10 minute spin at 6000g removed cell particles, while phage in the remaining supernatant precipitated upon addition of 2.5 mL of both 5 M NaCl and 50% P E G 8000 (Vieira and Messing, 1987). After 30 minutes on ice, precipitated phage particles were recovered with a spin at 10 OOOg for 10 minutes. The clear, translucent pellets were resuspended in 500 [ih of 0.3 M sodium acetate, pH 4.8 and 1 m M EDTA, then extracted twice with an equal volume of phenol and once with phenol-chloroform. Single stranded D N A precipitated upon addition of 0.1 volume of 7.5 M ammonium acetate, pH 7.5 and 1 volume of isopropanol. After centrifugation at 10 OOOg for 10 minutes, the purified ssDNA pellet was resuspended in 30 |J.L of T E buffer. xvL Isolation of Recombinant Lambda Phage DNA Small-scale isolation of lambda phage DNA was carried out from a 20 mL inoculum of LB broth supplemented with 10 m M M g S 0 4 and 0.2% maltose. Typically, 200 ^ L of an overnight culture of X L l - B l u e MRA (P2) was mixed with 10 |iL of stock phage solution and incubated at 37°C for 15 minutes. Infected cells were then transferred to 20 mL culture tubes and shaken vigorously at 37°C until bacterial lysis was evident (5-7 hours). Transformation of the cloudy cell culture to a solution which was clear with floating strands of cell debris signified cell lysis. Complete lysis was achieved with the addition of 200 \ih of CHCI3 and 10 additional minutes of incubation. Cellular debris was removed by centrifugation at 10 OOOg for 10 minutes, and the clear supernatant was transferred to a 50 mL polycarbonate centrifuge tube that contained 6 mL of 50% P E G 8000 and 3 mL of 5 M NaCl. After precipitating overnight at 4°C, phage particles 74 were pelleted by centrifugation at 10 OOOg for 10 minutes, then resuspended in 500 [ih of DNase I buffer (50 m M Tris-HCl, pH 7.5, 5 m M magnesium chloride, and 0.5 mM calcium chloride). DNase I to 10 |j,g/mL (Calbiochem) and bovine pancreatic RNase A to 0.1 |j,g/mL (Sigma) were added to each phage sample and incubated at 37°C for 30 minutes. At this stage, residual debris was removed by centrifugation for 5 minutes at 10 OOOg. Phage particles in the supernatant were lysed with 0.2 |j.g/mL of Proteinase K (Sigma) in a buffer of 10 mM Tris-HCl, pH 7.8, 5 m M E D T A and 0.5% SDS. Proteolytic digestion was conducted for 60 minutes at 6 5 ° C . Samples were extracted twice with an equal volume of buffer saturated phenol and once with an equal volume of phenol-chloroform. Phage DNA precipitated upon addition of 0.1 volume of 7.5 M sodium acetate, pH 7.5 and 1 volume of isopropanol. After cooling to -70°C for 10 minutes, DNA was recovered by a 10 OOOg spin for 10 minutes and then resuspended in 50 \ih of T E buffer. xviL Endonuclease Restriction Analysis Vectors that contained a horse gelsolin cDNA insert were identified by restriction endonuclease analysis using Xho I and Xba I (Pharmacia). These enzymes where chosen for their abilities to cleave around the EcoR V cloning site without disrupting gelsolin cDNA inserts. A 20 |iL digest was set up with 1 fig of purified plasmid DNA, 1 iug/|iL RNase A, 5 (ig/|iL bovine serum albumin, and with 0.5 (iL of each stock restriction enzyme in buffer supplied by the manufacturer. Digestion was conducted for 60 minutes at 37°C, which was then terminated with a 15 minute heating period at 68°C. Approximately one third of each 20 \iL digest was loaded 75 on an agarose gel and run out. Positive clones were viewed to liberate a single DNA insert from the vector fragment. xviii. DNA Sequencing The chain termination method developed by Sanger and co-workers was employed in acquiring the nucleotide sequence of horse gelsolin (Sanger et al, 1977). A minimum of three clones were sequenced for all fragments that were generated by PGR. Sequencing reactions with both single and double stranded templates were conducted using Sequenase Version 2 .0® according to the manufacturer's recommended protocol (United States Biochemical). Oligonucleotides used in priming the sequencing reactions were Ty, T3, forward, and reverse from the vector, along with various gelsolin gene specific primers. Autoradiography was carried out for 1-2 days with Kodak XRP-1 film. xix. N-terminal Protein Sequencing Affi-Gel purified horse plasma gelsolin, dialyzed extensively against 10 m M Tris-HCl, pH 8.0, and 1 mM N a N 3 , was submitted to the Nucleic Acid Protein Service (N.A.P.S) at the Biotechnology Laboratory, University of British Columbia for N-terminal sequencing. A total of 26 cycles were completed on an Applied Biosystems 476A Sequencer. After the first 19 cycles, identification of the derivatized amino acids became increasingly difficult due to non-specific peptide hydrolysis. Following residue 22, the sequence of horse plasma gelsolin was determined entirely by cDNA. 76 2. PROTEIN PREPARATIONS £ Purification of Actin A n acetone-dried skeletal muscle powder was prepared from minced rabbit tissue (Pel-Freez Biologicals) from which actin was purified according to a procedure based on the method of Spudich and Watt, 1971. A 200 mL solution of 2 m M Tris-HCl, 1 m M DTT, 0.2 m M ATP, and 0.2 mM C a C l 2 , pH 7.6 (buffer A) was added to 10 grams of rabbit muscle powder. The muscle tissue was placed on ice for 30 minutes and extracted with constant stirring. The mixture was filtered through a double layer of cheesecloth, and the remaining solid residue re-extracted with an additional 100 mL of buffer A followed by a second filtration. The supernatants from the two cheesecloth filtrations were combined and filtered through Whatman #3 paper. Centrifugation for 1 hour at 80 OOOg removed all insoluble muscle powder residue. The clarified supernatant was made up to 150 m M KC1 and 2 mM M g C l 2 , and left undisturbed to polymerize overnight at 4 ° C . The resulting F-actin solution was supplemented with solid KC1 to 0.8 M and stirred gently for 1.5 hours. Centrifugation for 3 hours at 80 OOOg pelleted the F-actin polymer. The translucent F-actin pellet was resuspended in 30 mL of buffer A with gentle homogenization, then converted back to G-actin by extensive dialysis against buffer A at 4°C. After dialyzing for 3 days, G-actin was clarified with a final spin at 80 OOOg for 3 hours, frozen in liquid nitrogen, then stored at ~20°C until needed. Concentrations of solutions containing G-actin were determined spectrophotometrically using an absorption coefficient at 290 nm of 0.63 77 m L m g - ^ c m 1 (Gordon et al, 1976). Actin preparations from 10 g of muscle powder usually yielded around 90 mg of pure protein. vi Purification of Gelsolin A procedure modified from that initially described by Bryan, 1988 was used to purify gelsolin from previously frozen horse plasma. Typically, 1.0 L of frozen plasma was thawed under cold running water. Protease inhibitors, leupeptin dissolved in d H 2 0 and pepstatin in D M S O , were added to a final concentration of 0.2 mg/L, while phenylmethylsulfonyl fluoride (PMSF) in 1 mL of methanol was added to 0.2 m M . Once thawed, the plasma sample was dialyzed at 4°C against 10 L of 25 m M Tris-HCl, pH 7.5 and 0.5 mM C a C l 2 . Three changes of buffer were made over the next three days. Centrifugation at 10 OOOg for 20 minutes clarified the dialyzed plasma, which was then adjusted to 35 mM NaCl and 1 mM N a N 3 . A batch ion exchange procedure initially fractionated a significant portion of the plasma proteins. The clarified plasma was added to approximately 2 L of settled D E A E Sephadex A50 (Pharmacia), which had been equilibrated against 50 m M NaCl, 25 m M Tris-HCl . pH 7.5, 1 m M NaN3, and 0.5 m M C a C l 2 - Proteins adsorbed to the resin for 2 hours at 4°C with occasional gentle stirring. The slurry was filtered through Whatman #3 filter paper and the supernatant made up to 10 m M EDTA, pH 7.8. At this stage, the protein solution was clear, while the ion exchange resin retained the characteristic color of the plasma. A 36 cm x 6 cm column of D E A E Sephadex A50, equilibrated at 1 mL/minute with 50 m M NaCl, 25 m M Tris-HCl, pH 7.8, 1 mM N a N 3 and 0.1 m M EGTA, was set up at 4°C. The E D T A adjusted protein solution was applied to the column, followed 78 by a 2 bed volume wash with equilibration buffer. Proteins were eluted from the resin by the application of a linear 50 m M to 300 m M NaCl gradient in equilibration buffer. Protein content was monitored in the eluant fractions via absorbance readings at 280 nm, and gelsolin localized in the broad eluant peak by polyacrylamide gel electrophoresis (PAGE) in the presence of sodium dodecyl sulfate (SDS) and 2-mercaptoethanol according to the method of Laemmli, 1970. Gelsolin containing fractions were pooled and concentrated to approximately 1 m g / m L by ultrafiltration with an Amicon YM-30 membrane. A n additional chromatographic step with Affi-Gel Blue was implemented on those occasions that gelsolin preparations showed signs of contamination at this stage. Concentrated gelsolin was extensively dialyzed against 25 m M Tris-HCl, pH 8.0 and 1 m M E D T A at 4°C, then applied at 1 mL/minute to a 5 mL Econo-Pac® Affi-Gel Blue Cartridge column (BioRad Laboratories). After a two bed volume wash with equilibration buffer, gelsolin was biospecifically eluted with 1 m M ATP in equilibration buffer (Ito et al, 1990). Starting from 1 L of horse plasma, a yield of 25-30 mg of pure protein was generally attained. The concentration of gelsolin solutions was measured spectrophotometrically using an extinction coefficient determined in this laboratory of 1.4 n _ m g - 1 - c m - 1 at 280 nm (Ruiz Silva and Burtnick, 1990). UL Gelsolin Purity and Severing Activity The purity of gelsolin preparations was assessed by electrophoresis on 10% polyacrylamide gels (BioRad Mini-Protean II). A single band with 79 an apparent molecular mass of 90 000 daltons was visible on Coomassie stained gels. Viscosity measurements were conducted to evaluate the actin severing activity of each gelsolin preparation (Reid et al, 1993). A 25 |j.M solution of F-actin was prepared in 150 mM KC1, 2 m M Tris-HCl, 1 m M DTT, 0.2 m M C a C l 2 , and 0.2 mM ATP, pH 7.6, to which gelsolin was added in amounts yielding an actin to gelsolin mole ratio of 50:1. After a 60 minute incubation period at room temperature, a Cannon-Manning semi-micro viscometer (size 100) was used to measure the actin-gelsolin solution viscosities. In a thermostatically controlled water bath at 27°C, each 1 m L sample was analyzed 3 consecutive times. The flow time for buffer was determined at 103.5 seconds with a standard deviation of + 0.4 seconds. Actin together with gelsolin at a 50:1 mole ratio yielded a relative viscosity of 1.09, while actin and FITC-labeled gelsolin at the same ratio yielded a value of 1.12 (Koepf and Burtnick, 1993). The relative viscosities of F-actin and G-actin solutions, both at 25 |iM, were 1.49 and 1.04, respectively. iv. Preparation of FITC-labeled Gelsolin The amine selective reagent fluorescein-5-isothiocyanate (FITC isomer 1, Molecular Probes) was used to alkylate exposed amino groups on gelsolin by a modified procedure given for G-actin (Burtnick, 1984). Relative to highly reactive sulfonyl chlorides and to slower reacting succinimidyl esters, isothiocyanates are of intermediate reactivity towards nucleophilic attack. Isothiocyanates react with aliphatic amines such as those from lysines to form stable thioureas in aqueous environments. The 80 reaction occurs best above pH 8.5, where a significant fraction of protein amine groups are unprotonated. FITC Thiourea Approximately 2 mg of the fluorescent reagent were dissolved in 100 |iL of 0.1 M NaOH and added to 10 mL of a 1 mg/mL plasma gelsolin solution that had dialyzed extensively against 150 m M KC1, 50 m M Hepes, and 1 m M EDTA, pH 8.5. The reaction mixture was stirred at 4°C in a dark environment for 24 hours. All additional dialysis with labeled protein was conducted in the dark to prevent photobleaching and at 4°C to enhance protein stability. The majority of unincorporated reagent was removed by an initial 24 hours dialysis against 1 L of 150 mM KC1, 25 m M Tris-HCl, and 1 m M EDTA, pH 8.0. The sample was then passed through a 1.3 cm x 45 cm Bio-Gel P2 size exclusion column (BioRad Laboratories) that was equilibrated and run with the above buffer. Labeled gelsolin fractions were pooled, concentrated with a Centricon™ 30 micro-concentrator (Amicon) and dialyzed over the next two days with a fresh change of buffer each day. Independent measurements of the fluorophore and the protein concentrations were made to determine the average number of fluorescein molecules that were incorporated into the gelsolin molecule. 81 ^ r T t. i . [Fluorescein! Degree of Labeling = [Gelsolin] The fluorescein content was analyzed spectrophotometrically using a molar absorption coefficient of 74 500 M ^ c m - 1 at 493 nm, while the BioRad microassay, which is based upon the observation that the absorbance maximum of Coomassie Brilliant Blue G-250 shifts from 465 nm to 595 nm in protein solutions, was used to determine the protein concentration (Bradford, 1976; Bernhardt et al, 1983). 3. OPTICAL TECHNIQUES 82 I Absorbance Measurements Proteins were detected in column eluant fractions at various stages during the purification procedure by online absorbance measurements taken at 280 nm. Concentrations of purified protein preparations were determined with a Lambda 4B U V / V i s Spectrophotometer (Perkin Elmer). The same instrument was used to conduct the BioRad assay and to determine DNA concentrations. it Fluorescence Measurements Fluorescence measurements were performed with a LS-5B Luminescence Spectrometer (Perkin-Elmer) that was equipped with a circulating water bath (Haake). Due to the high absorptivity of ATP in the U V spectral region, eluant fractions from the Affi-Gel Blue column were analyzed via fluorescence. Aromatic amino acids in the protein containing fractions were excited at 280 nm and their emission monitored at 340 nm. ATP did not contribute to any interfering fluorescence in this wavelength region. tit Circular Dichroism Fixed wavelength measurements were carried out with a rebuilt J A S C O J20 instrument that has been modified for computer control and data acquisition (Landis Instrument), while CD spectra were recorded with 83 a J A S C O 710 spectropolarimeter. Both spectropolarimeters were equipped with N 2 -purged cell chambers that were thermostatically controlled by circulating water baths. CD spectra in the accompanying figures are averages of 4 consecutive scans corrected by subtracting signal from a buffer blank. Fixed wavelength measurements in the far U V spectral region were made with 0.5 mm Suprasil quartz cells (Hellma), while 1.0 cm pathlength cells were used in the near U V region. The protein concentration in all C D experiments was 1.0 mg/mL. Calibrations of the instruments were made with (+)-10-camphorsulfonic acid and D-(-)-pantolactone (Tuzimura et ah, 1977). iv. Convex Constraint Analysis The convex constraint algorithm (CCA), which deconvolutes a protein C D spectrum into its pure component curves and conformational weights was used to analyze the secondary structure of gelsolin (Perczel et al., 1991, 1992, 1992a). The data that was analyzed by the C C A method was measured in the 178 to 260 nm region in steps of 0.2 nm, then truncated to the 195-240 nm wavelength range by averaging it through a moving window of 5 points. The 46 resulting points obtained at each unit wavelength were appended to the reference data set of 25 proteins which was provided by the program. Deconvolutions were conducted with various values of P, the number of expected pure components, and with differing numbers of iterations. Five curves (P=5) and 40 iterations were required to deconvolute the experimental CD spectrum satisfactorily. 84 4. DENATURATION STUDIES £ Chemical denaturations C D denaturation studies were performed with electrophoresis grade urea from Fisher Biotech, and with ultrapure molecular biology grade guanidine hydrochloride from Sigma. Stock solutions of 10 M urea and 8M G n d - H C l were prepared fresh, and passed through 0.45 \im Millipore filters prior to use. Samples of 1.0 mg/mL native or FITC-labeled gelsolin were prepared at 25°C, to which aliquots of the stock denaturants were added. The extent of denaturation was monitored by C D at either 215 nm or 276 nm. A linear decrease in the CD signal upon volume increase was assumed as corrections were made for dilutions. Each representative data point in figures 29 to 34 is the average of 500 readings taken over a time interval of 100 seconds. The standard deviation of the data set was less than 3 percent of the signal reading for virtually every case, and thus is generally smaller than the symbols used in the accompanying figures. u. Thermal Denaturations Thermal denaturation studies of gelsolin were performed at a fixed wavelength of 215 nm. A n equilibration period of 15 minutes was allowed at each new temperature prior to a reading being taken. Samples were monitored for visible signs of precipitation above temperatures of 3 0 ° C . Precipitation was found to coincide with a large decrease in signal to noise. Data acquisition and processing was as described for the chemical denaturation studies. 85 All of the data points or spectra represented in the results section displayed no visible signs of turbidity. Because light scattering introduces artifacts into measured CD, it is mentioned in the text where precipitation was observed but such data are not presented in the figures. 86 5. CRYSTALLIZATION TRIALS £ Crystallization of calciumjree gelsolin Crystals of horse plasma gelsolin were grown by the hanging and sitting drop vapor diffusion methods. Crystallization conditions similar to those reported for human recombinant gelsolin were utilized as a starting point for the pH and ( N H 4 ) 2 S 0 4 screen (McLaughlin and Gooch, 1992). Initial trials screened the pH from 7.0 to 8.5, and ( N H 4 ) 2 S 0 4 from 30% to 40% saturation. Affi-Gel purified gelsolin was dialyzed extensively against 10 m M Tris-HCl, pH 8.0, 1 m M EDTA, and 0.5 mM N a N 3 , then concentrated to approximately 20 m g / m L with Centr icon™ 30 microconcentrators (Amicon). The protein along with all other solutions used in the crystallization trials were filtered through 0.22 |im Millipore filters. 100 m M T r i s - H C l buffers at each indicated pH were prepared in 24 well Linbro® tissue culture plates (Flow Laboratories). Saturated ( N H 4 ) 2 S 0 4 was then added to each of the wells along with d H 2 0 to a final reservoir volume of 1 mL. Crystallization drops were prepared by mixing 1 \ih Of well buffer with 2-3 |iL of concentrated protein solution, and placed on either coverslips for hanging drops or in microbridges for sitting drops. Each well top was surfaced with a layer of vacuum grease and sealed from the atmosphere with either a new cover slip or one with a hanging drop on it. Duplicate trials were conducted at 2 0 ° C and at 4 ° C . Crystallization conditions similar to these but with the inclusion of 2 m M C a C l 2 in the protein dialysis buffer produced no observable crystals. 87 ii. Crystallization of calcium loaded gelsolin Crystallization trials in the presence of calcium ions were set up with buffered sodium potassium tartrate solutions. Purified protein loaded with C a 2 + was prepared by dialysis against 25 mM Tris-HCl, pH 8.0, and 10 mM C a C l 2 . Protein concentration and hanging drop preparation was conducted as described above. In these experiments, the tissue culture plate wells contained 100 mM Tris-HCl solutions that ranged from pH 5.25 to pH 8.0. The precipitating agent, sodium potassium tartrate, was screened from 20% to 31% saturation. - \ 88 CHAPTER m RESULTS AND DISCUSSION 1. CLONING AND SEQUENCE ANALYSIS L Gelsolin Cloning Strategy Comparison of human and pig plasma gelsolin cDNA sequences reveals a remarkable degree of identity (Kwiatkowski et al., 1986; Way and Weeds, 1988). This led to the assumption that the cDNAs coding for horse and human plasma gelsolins should also share a high level of sequence identity. Therefore, using human gelsolin cDNA as our reference template, PCR experiments were designed to synthesize the cDNA for horse gelsolin, with two overlapping clones. Initially, the focus was on acquisition of the cytoplasmic portion of the whole gelsolin message. Experiments were then expanded to try to incorporate the 5' plasma extension. Two pairs of oligodeoxyribonucleotides of 20 bases in length, designed to span the cytoplasmic coding region of the gelsolin message, were synthesized. The first pair was located at position 167 on the coding or sense strand of the dsDNA and at position 1516 on the complementary non-coding or anti-sense strand. The primers of the second set were designed to positions 1446 (sense) and to 2390 (anti-sense). These primer sets were intended to generate fragments of 1369 and 964 base pairs, respectively. The overlapping region of 90 bases contained a unique Acc I restriction site at position 1506, one which was required for constructing full-length clones. Extensive PCR trials with various preparations of first strand cDNA failed to produce the desired products. Controls conducted concurrently 89 with human liver cDNA yielded similarly negative results. PCR variables such as primer, dNTP, M g + 2 , and enzyme concentrations were all examined meticulously, but after an exhaustive search of reaction conditions that numbered in the thousands, still no product was obtained. This predicament necessitated implementation of a secondary cloning strategy with different primer combinations. it cDNA Clones of Horse Cytoplasmic Gelsolin The first of three overlapping cDNA clones was generated via RT-PCR from total horse uterus RNA with Vent® DNA polymerase and gelsolin primers #1 and #2 (Table 1). Extension of these primers produced a single 759 base pair PCR product, gelsolin fragment 1. This cDNA was cloned into the EcoR V site of pBluescript K S + , and subsequently analyzed by restriction digests (Figure 18, Lane D). Simultaneous digestion of plasmids containing fragment 1 with Xho I and Xba I liberated a single DNA insert. PCR experiments conducted with gelsolin primers #3 and #4 yielded gelsolin fragment 2, a 1249 base pair stretch of cDNA. The third gelsolin fragment of 582 base pairs was obtained with gelsolin #5 and #6 primers. Fragments 2 and 3 were also cloned into the EcoR V cloning site of pBluescript K S + , and analyzed by restriction analysis as described above (Figure 18). Gelsolin fragments 1 and 2 overlap a region of 38 nucleotides, while 363 base pairs are shared between fragments 2 and 3. The polymerase that generated fragments 2 and 3 was Taq. Vent® DNA polymerase with exonuclease proofreading activity did not produce any products with these primer combinations. The overlap of these three gelsolin cDNA fragments, together with nine bases obtained from a 6.7 kb 90 h o r s e g e n o m i c c l o n e ( F i g u r e 18 , L a n e B ) , e n c o m p a s s e d a l l 2 1 9 6 n u c l e o t i d e b a s e s o f h o r s e c y t o p l a s m i c g e l s o l i n ( F i g u r e 19). O f t h e s e 2 1 9 6 n u c l e o t i d e s , 2 1 9 3 e n d u p c o d i n g f o r a m i n o a c i d s i n t h e t r a n s l a t e d p r o t e i n . T h e 6 . 7 k b g e n o m i c s u b c l o n e w a s g e n e r a t e d f r o m ? i E M B L 3 p h a g e s t h a t c o n t a i n e d 15 k b h o r s e g e l s o l i n g e n o m i c i n s e r t s . I n i t i a l l y , s i x g e l s o l i n c o n t a i n i n g p h a g e c l o n e s w e r e i s o l a t e d b y h y b r i d i z a t i o n s c r e e n i n g ( F i g u r e 20) , b u t u p o n f u r t h e r h y b r i d i z a t i o n c h a r a c t e r i z a t i o n , o n l y t w o c l o n e s w e r e i d e n t i f i e d a s h a v i n g a g e l s o l i n 5 ' - e n d i n s e r t ( F i g u r e 2 1 , l a n e s B a n d E ) . D N A f r o m t h e s e t w o c l o n e s w a s s u b j e c t e d to Xba I r e s t r i c t i o n d i g e s t i o n , f o l l o w e d b y c l o n i n g i n t o p B l u e s c r i p t K S + to y i e l d t h e 6 . 7 k b i n s e r t p r e s e n t i n F i g u r e 18 , l a n e B . a b c d e Figure 18: Xho I a n d Xba I r e s t r i c t i o n d i g e s t s of t h r e e c D N A a n d o n e h o r s e g e l s o l i n 5 ' - e n d g e n o m i c c l o n e t h a t t o g e t h e r c o d e f o r h o r s e c y t o p l a s m i c g e l s o l i n . T h e 1 k b l a d d e r i n l a n e A w a s u s e d a s a s i z e m a r k e r , w h i l e l a n e B c o n t a i n s a 6 . 7 k b g e n o m i c c l o n e . T h e D N A i n l a n e s C , D , a n d E i l l u s t r a t e g e l s o l i n f r a g m e n t s 2 ( 1 2 4 9 b . p . ) , 1 ( 7 5 9 b . p . ) , a n d 3 ( 5 8 2 b . p . ) , r e s p e c t i v e l y . L i n e a r i z e d p B l u e s c r i p t v e c t o r m i g r a t e s a t 3 k b . 91 168 2366 ± : ; I Cytoplasmic Gelsolin 2196 base pairs (731 amino acids ) 177 F r a 8 m e n t l 935 898 Fragment 2 ^ Fragment 3 I 1785 2366 Figure 19: Schematic representation of the 4 overlapping clones that produced the sequence for horse cytoplasmic gelsolin. Bases 168 to 176 are derived from a genomic clone. The A T G start codon is located at bases 168-170, while the 2364 to 2366 triplet represents the T G A termination codon. The numbering scheme is in alignment with the sequence for human plasma gelsolin. iii. In Search of the 5' -Plasma Extension A stretch of 167 base pairs found 5' to the cytoplasmic A T G start codon distinguishes the human plasma isoform of gelsolin from the cytoplasmic one (Kwiatkowski et al, 1986). Since the vast majority of the cytoplasmic message was obtained by PCR with human primers, it was postulated that the 5'-end of the horse plasma message would also be attainable in this manner. PCR reactions were conducted with either gelsolin #7 or #8 sense primers and with a number of anti-sense primers 92 synthesized to various locations just 3' to the missing piece of sequence. Amplification of the target fragment was not achieved with any of the primer pair combinations tested. A number of specialized PCR techniques designed to amplify the 5'-region of cDNA were explored. The first of these was R A C E (Figure 13), which utilized known 5' sequence of gelsolin cDNA fragment 1. Following reverse transcription with a gene specific anti-sense primer (gelsolin #9), the cDNA was modified by the addition of a poly A tail onto the 5'-end with terminal deoxynucleotidyl transferase. Attempts to amplify the cDNA with PCR, using a complementary homopolymeric oligo(dT)i8 primer and a second nested gene specific primer, were not successful. R A C E - P C R produced smears, but discrete bands were not observed. A second technique that was explored in the quest for the 5'-end was SLIC, developed by Edwards et al, 1991 (Figure 14). This method exploits the ability of T4 RNA ligase to link oligodeoxyribonucleotides onto the ends of ssDNA. A 27mer oligonucleotide anchor was ligated onto the ends of cDNA that was synthesized from gelsolin #9 primed mRNA. A n oligonucleotide complementary to the anchor (anchor primer) and a nested gene specific primer were used in subsequent amplification reactions. SLIC-PCR experiments produced results similar to those observed with RACE. The method of inverse PCR (Figure 15) was the third technique employed in the search for the 5'-end (Zeiner and Gehring, 1994). Similar to R A C E and SLIC, this PCR method also utilized first strand cDNA that was generated with gelsolin #9. Following second strand synthesis, linear cDNA was circularized with T4 DNA ligase, and an inverse PCR experiment was set up with primers made to previously acquired sequence. As with 93 R A C E and SLIC, the inverse PCR technique did not amplify the stretch of cDNA that codes for the 5' extension found in plasma gelsolin. Sequence analysis of a variety of mammalian genes has shown that their 5'-ends usually contain a large proportion of G and C residues. The mRNA that is transcribed from these genes will in turn contain G C rich regions. These areas have a tendency to form such secondary structures as hairpin loops, which may obstruct or hinder the reverse transcriptase enzyme. To avoid prematurely terminated cDNA transcripts, reverse transcriptase must copy the mRNA through these stretches of sequence. The R A C E , SLIC and inverse PCR methods were repeated with rTth DNA polymerase (Perkin Elmer), a multifunctional enzyme that displays reverse transcriptase activity at 7 0 ° C in the presence of M n + 2 , and DNA polymerase activity in the presence of M g + 2 . The increase in the reverse transcription reaction temperature should have facilitated the production of full-length cDNAs, as a significant portion of RNA secondary structure melts at such a temperature. However, template cDNA synthesized with rTth DNA polymerase did not improve the results of the 5'-end search when used in either conventional or specialized PCR experiments. iv. Isolation of Horse Gelsolin Genomic Clones The quality of a cDNA library ultimately depends on the success of the reverse transcriptase reaction that transcribes mRNA into first strand cDNA. Although both human and pig gelsolin sequences were obtained from cDNA libraries, a large number of such clones are found terminated prematurely at their 5'-ends. Since the 5'-end of human plasma gelsolin has a G C residue content of over 81%, the decision was made to 94 circumvent the reverse transcription reaction by creating a horse genomic library in favor of one constructed from cDNA. From a total screen of approximately 2 x l 0 6 recombinants, 6 plaques hybridized with an (a- 3 2P]-dCTP labeled probe. The 215 base pair probe, used to screen the genomic clones, was synthesized via PCR from fragment 1 template with gelsolin primers #1 and #9. To maximize the probability of finding the 5'-end of the plasma message, the probe used to screen the genomic clones contained the 5'-terminus of the cytoplasmic sequence. A second and a third round of stringent hybridization screening with the same probe, but at much lower phage titers, yielded single uniform sized plaques that produced intense spots on X-ray film (Figure 20). Following the third screen, a single phage plaque was isolated for each of the 6 potential, positive genomic clones. Digestion of. recombinant phage DNA with Sal I, an enzyme that cleaves around the B a m H I cloning sites of the vector (Figure 10), produced an insert of approximately 15 kb for all clones analyzed. v. Identification of the 5'-end in Genomic Clones Additional characterization of the 6 potential positive clones isolated from the genomic library was conducted with 3 2 P - e n d labeled gelsolin #7 and #8 oligodeoxyribonucleotide primers. These correspond to human plasma gelsolin positions 1 and 61, respectively. From this 5'-end screen, two clones were identified as positives, and selected for further investigation (Figure 21, lanes B and E). Attempts to subclone these fragments as full-length 15 kb Sal I inserts into pBluescript K S + were not successful. Subcloning of a smaller 6.7 kb fragment was achieved (Figure 95 18, lane B) following Xba I restriction digestion of the two clones identified in Figure 21 (lanes B and E) as containing a portion of the 5'-end of the horse gelsolin gene. » Figure 20: Plaque purification of 1 of the 6 A.EMBL3 genomic clones after three consecutive rounds of screening. Hybridizations were conducted with a 215 base pair PCR fragment that contained the 5' sequence of horse cytoplasmic gelsolin. Spots from this screen represent individual plaques, all of which contain an identical insert. a b c d e f Figure 21: 5'-end characterization of X.EMBL3 genomic clones with a mixture of 3 2 P - e n d labeled gelsolin #7 and #8 oligodeoxyribonucleotide primers. Clones in lanes B and E contain portions of the horse gelsolin gene that code for the 5' plasma extension. The bands on this blot represent recombinant phage DNA that is approximately 42 kb long. 96 vl cDNA Sequence of Cytoplasmic Horse Gelsolin The DNA sequence of the 4 overlapping fragments that code for horse cytoplasmic gelsolin was obtained by the chain termination method of Sanger. Sequencing reactions were carried out with both single and double stranded templates. A minimum of three clones were sequenced for each of gelsolin fragments 1,2, and 3, while the sequence of only two genomic clones was acquired. Sequence data from clones that displayed errors due to the infidelity of Taq (PCR errors) was not retained. Both strands of the cDNA sequence that correspond to horse cytoplasmic gelsolin are presented in Figure 22. The message begins with the universal A T G start codon at positions 1-3, and terminates with the T G A triplet stop codon at positions 2194-2196. For the purpose of sequence comparisons, position 1 in the horse cytoplasmic sequence is equivalent to position 168 in human plasma gelsolin. The total cDNA sequence of 2196 base pairs has a nucleotide composition of 460 A, 659 C, 686 G, and 391 T residues, and codes for a 731 amino acid protein that has a predicted molecular mass of 80 696 daltons. 10 20 30 40 50 60 70 I I I I I I ATGGTGGTAG AACACCCCGA GTTCCTCAAG GCGGGGAAGG AGCCTGGCCT GCAGATCTGG CGCGTGGAGA TACCACCATC TTGTGGGGCT CAAGGAGTTC CGCCCCTTCC TCGGACCGGA CGTCTAGACC GCGCACCTCT 80 90 100 110 120 130 140 I I I I I I I AGTTCGACCT GGTGCCCGTG CCCCCCAACC TTTACGGAGA CTTCTTCACA GGTGACGCCT ATGTCATCCT TCAAGCTGGA CCACGGGCAC GGGGGGTTGG AAATGCCTCT GAAGAAGTGT CCACTGCGGA TACAGTAGGA 97 150 160 170 180 190 200 210 I I I I I . I I GAAGACGGTG CAGCTGAGGA ACGGGATCCT GCAGTACGAC C T C C A C T A C T GGCTGGGCAA TGAGTGCAGC C T T C T G C C A C G T C G A C T C C T TGCCCTAGGA CGTCATGCTG GAGGTGATGA C C G A C C C G T T A C T C A C G T C G 220 230 240 250 260 270 280 I I I I I I I CAGGATGAGA GCGGCGCGGC C G C C A T C T T C ACCGTGCAGC TGGATGACTA CCTGAACGGT CGGGCTGTGC G T C C T A C T C T CGCCGCGCCG GCGGTAGAAG TGGCACGTCG A C C T A C T G A T G G A C T T G C C A G C C C G A C A C G 290 300 310 320 330 340 350 I I I I I I I A G C A C C G C G A GGTCCAGGGC TTTGAGTCGG C C A C C T T C C T T G G C T A C T T C AAGTCTGGCC T C A A G T A C A A TCGTGGCGCT C C A G G T C C C G A A A C T C A G C C GGTGGAAGGA A C C G A T G A A G T T C A G A C C G G A G T T C A T G T T 360 370 380 390 400 410 420 I I I I I I I GAAAGGAGGT GTGGCATCAG G A T T C A A G C A TGTGGTGCCC AATGAGGTGG TGGTGCAGAG G C T C C T C C A G C T T T C C T C C A C A C C G T A G T C C T A A G T T C G T ACACCACGGG T T A C T C C A C C A C C A C G T C T C CGAGGAGGTC 430 440 450 460 470 480 490 I I I I I I I GTCAAAGGGC GGCGTGTGGT CCGTGCTACC GAGGTGCCCG TATCCTGGGA G A G C T T C A A C A A T G G C G A C T C A G T T T C C C G C C G C A C A C C A GGCACGATGG CTCCACGGGC A T A G G A C C C T C T C G A A G T T G T T A C C G C T G A 500 510 520 530 540 550 560 I I I I I I I G C T T C A T C C T GGACCTGGGC A A C A A C A T C T ATCAGTGGTG C G G C T C C A A A A G C A A C C G A T TTGAGAGGCT CGAAGTAGGA CCTGGACCCG TTGTTGTAGA T A G T C A C C A C GCCGAGGTTT T C G T T G G C T A A A C T C T C C G A 570 580 590 600 610 620 630 I I I I I I I G A A G G C C A C A CAGGTGTCCA AGGGCATCCG GGACAACGAG CGGAGCGGCC GGGCCCAAGT G T C C G T G T T T C T T C C G G T G T GTCCACAGGT TCCCGTAGGC C C T G T T G C T C GCCTCGCCGG C C C G G G T T C A C A G G C A C A A A 640 650 660 670 680 690 700 I I I I I I I GAGGAAGGCG CTGAGCCCGA GGCGATGCTC CAGGTGCTGG GCCCCAAGCC A A C T C T G C C C GAAGCGACTG C T C C T T C C G C GACTCGGGCT CCGCTACGAG G T C C A C G A C C CGGGGTTCGG TTGAGACGGG C T T C G C T G A C 98 710 720 730 740 750 760 770 I I I I I I I AGGACACAGT CAAGGAGGAT GCGGCCAACC GCAAGCTGGC CAAGCTCTAC AAGGTCTCCA ATGGCGCGGG TCCTGTGTCA GTTCCTCCTA CGCCGGTTGG CGTTCGACCG GTTCGAGATG TTCCAGAGGT TACCGCGCCC 780 790 800 810 820 830 840 I I I I I I I CCCCATGGTG GTCTCCCTTG TGGCTGATGA GAATCCCTTC GCCCAGGGGG CCTTGAGGTC AGAGGACTGC GGGGTACCAC CAGAGGGAAC ACCGACTACT CTTAGGGAAG CGGGTCCCCC GGAACTCCAG TCTCCTGACG 850 860 870 880 890 900 910 I I I I I I I TTCATCCTGG ACCACGGCAA AGATGGGAAA ATCTTTGTCT GGAAAGGCAA GCAAGCCAAC ATGGAGGAGA AAGTAGGACC TGGTGCCGTT TCTACCCTTT TAGAAACAGA CCTTTCCGTT CGTTCGGTTG TACCTCCTCT 920 930 940 950 960 970 980 I I I . I I I I GGAAGGCTGC CCTCAAAACG GCCTCCGACT TCATCTCCAA GATGGACTAC CCCAAGCAGA CCCAGGTCTC CCTTCCGACG GGAGTTTTGC' CGGAGGCTGA AGTAGAGGTT CTACCTGATG'GGGTTCGTCT'GGGTCCAGAG 990 1000 1010 1020 1030 1040 1050 I I I I I I I CGTCCTTCCC GAGGGCGGTG AGACCCCGCT GTTCAGGCAG TTCTTCAAGA ACTGGCGGGA CCCGGACCAG GCAGGAAGGG CTCCCGCCAC TCTGGGGCGA CAAGTCCGTC AAGAAGTTCT TGACCGCCCT GGGCCTGGTC 1060 1070 1080 1090 1100 1110 1120 I I I I I I I ACGGAAGGCC TGGGCTTGGC CTATCTCTCC AGCCACATCG CCCACGTGGA GCGCGTGCCT TTCGATGCCG TGCCTTCCGG ACCCGAACCG GATAGAGAGG TCGGTGTAGC GGGTGCACCT CGCGCACGGA AAGCTACGGC 1130 1140 1150 1160 1170 1180 1190 I I I I I I I CCACCCTGCA CACCTCCACT GCCATGGCCG CCCAGCATGG CATGGATGAC GACGGCACAG GCCAGAAACA GGTGGGACGT GTGGAGGTGA CGGTACCGGC GGGTCGTACC GTACCTACTG CTGCCGTGTC CGGTCTTTGT 1200 1210 1220 1230 1240 1250 1260 I I I I I I I GATCTGGAGA GTCGAAGGGT CCAACAAAGT GCCCGTGGAC CCCGCCACCT ACGGGCAGTT CTACGGTGGG CTAGACCTCT CAGCTTCCCA GGTTGTTTCA CGGGCACCTG GGGCGGTGGA TGCCCGTCAA GATGCCACCC 99 1 2 7 0 1 2 8 0 . 1 2 9 0 1 3 0 0 1 3 1 0 1 3 2 0 1 3 3 0 I I I I I I I G A C A G C T A C A T C A T T C T G T A C A A C T A C C G C C A C G G C A G C C G T C A G G G A C A G A T C A T C T A C A A C T G G C A G G C T G T C G A T G T A G T A A G A C A T G T T G A T G G C G G T G C C G T C G G C A G T C C C T G T C T A G T A G A T G T T G A C C G T C C 1 3 4 0 1 3 5 0 1 3 6 0 1 3 7 0 1 3 8 0 1 3 9 0 1 4 0 0 I I I I I . I I G C G C C C A G T C C A C C C A G G A T G A G G T C G C T G C G T C T G C C A T C C T G A C C G C C C A G C T A G A T G A G G A A C T G G G C G C G G G T C A G G T G G G T C C T A C T C C A G C G A C G C A G A C G G T A G G A C T G G C G G G T C G A T C T A C T C C T T G A C C C 1 4 1 0 1 4 2 0 1 4 3 0 1 4 4 0 1 4 5 0 1 4 6 0 1 4 7 0 I I I I I I . I A G G T A C C C C T G T C C A G A G C C G T G T G G T C C A A G G C A A G G A G C C T G C C C A T C T C A T G A G T C T A T T T G G C G G G T C C A T G G G G A C A G G T C T C G G C A C A C C A G G T T C C G T T C C T C G G A C G G G T A G A G T A C T C A G A T A A A C C G C C C 1 4 8 0 1 4 9 0 1 5 0 0 1 5 1 0 1 5 2 0 1 5 3 0 1 5 4 0 I I I I I I A A G C C C A T G A T C G T C T A C A A G G G T G G C A C C T C C C G C G A G G G T G G G C A G A C G G C C C C C G C C A G T A C T C G C C T T C G G G T A C T A G C A G A T G T T " C C C A C C G T G G ' A G G G C G C T C C " C A C C C G T C T G C C G G G G G C G G ' " T C A T G A G C G G 1 5 5 0 1 5 6 0 1 5 7 0 1 5 8 0 1 5 9 0 1 6 0 0 1 6 1 0 I I I T C T T C C A G G T C C G G G C C A G C A G C T C T G G A G C C A C C C G A G C C G T T G A G A T A A T C C C C A A G G C C G G C G C G C T A G A A G G T C C A G G C C C G G T C G T C G A G A C C T C G G T G G G C T C G G C A A C T C T A T T A G G G G T T C C G G C C G C G C G A 1 6 2 0 1 6 3 0 1 6 4 0 1 6 5 0 1 6 6 0 1 6 7 0 1 6 8 0 I I I I I I I G A A C T C C A A C G A T G C C T T T G T C C T G A A A A C T C C C T C G G C T G C C T A C C T G T G G G T G G G T G C A G G A G C C A G C C T T G A G G T T G C T A C G G A A A C A G G A C T T T T G A G G G A G C C G A C G G A T G G A C A C C C A C C C A C G T C C T C G G T C G 1 6 9 0 1 7 0 0 1 7 1 0 1 7 2 0 1 7 3 0 1 7 4 0 1 7 5 0 I I I I I I I G A G G C G G A G A A G A C G G G G G C C C A G G A G C T G C T C A G G G T G C T G C G G G C C C A A C C T G T G C A G G T G G C A G A A G C T C C G C C T C T T C T G C C C C C G G G T C C T C G A C G A G T C C C A C G A C G C C C G G G T T G G A C A C G T C C A C C G T C T T C 1 7 6 0 • 1 7 7 0 1 7 8 0 1 7 9 0 1 8 0 0 1 8 1 0 1 8 2 0 I I I . I . I I I G C A G C G A G C C A G A C A G C T T C T G G G A G G C C T T G G G T G G G A A G G C C A C C T A C C G C A C G T C C C C G C G G C T G A A C G T C G C T C G G T C T G T C G A A G A C C C T C C G G A A C C C A C C C T T C C G G T G G A T G G C G T G C A G G G G C G C C G A C T T 100 1830 1840 1850 1860 1870 1880 1890 I I I I I I I G G A C A A G A A G ATGGACGCCC A C C C T C C T C G C C T C T T T G C C T G C T C C A A C A AGATTGGACG T T T T G T G A T C C C T G T T C T T C TACCTGCGGG TGGGAGGAGC GGAGAAACGG ACGAGGTTGT T C T A A C C T G C A A A A C A C T A G 1900 1910 1920 1930 1940 1950 1960 I I I I I I I GAAGAGGTCC CTGGCGAGTT CATGCAGGAA GACCTGGCCA CTGATGACGT C A T G C T C C T G G A C A C C T G G G C T T C T C C A G G G A C C G C T C A A G T A C G T C C T T CTGGACCGGT G A C T A C T G C A GTACGAGGAC C T G T G G A C C C 1970 1980 1990 2000 2010 2020 2030 I I I I I I I A C C A G G T C T T TGTCTGGGTC GGAAAGGATT C T C A A G A C G A GGAAAAGACG GAAGCCTTGA C C T C T G C T A A T G G T C C A G A A A C A G A C C C A G C C T T T C C T A A GAGTTCTGCT C C T T T T C T G C C T T C G G A A C T GGAGACGATT 2040 2050 2060 2070 2080 2090 2100 I I I I I I I G C G G T A T A T C G A C A C A G A C C C A G C T C A T C G CGATAGGCGT A C C C C C A T C A C C G T C G T G A A G C A A G G C T T T C G C C A T A T A G CTGTGTCTGG GTCGAGTAGC G C T A T C C G C A TGGGGGTAGT GGCAGCACTT C G T T C C G A A A 2110 2120 2130 2140 2150 2160 2170 I I I I I I I G A G C C T C C G T CCTTCGTGGG C T G G T T C C T C GGCTGGGATG A C A G C T A C T G GTCTGTGGAT C C C T T G G A C A CTCGGAGGCA G G A A G C A C C C GACCAAGGAG C C G A C C C T A C T G T C G A T G A C C A G A C A C C T A GGGAACCTGT 2180 2190 I I GGGCCTTGGC TGAGCTGGCT GCCTGA C C C G G A A C C G A C T C G A C C G A CGGACT Figure 22: Complete cDNA sequence for the cytoplasmic isoform of horse gelsolin. The message of 2196 base pairs translates into a 731 amino acid protein. 101 Sequence alignment of horse and human cytoplasmic cDNAs reveals that 2010 out of 2196 residues are equivalent, yielding a sequence identity of 91.5%. Once translated to amino acids, the identity increases to 94.7% as 692 out of 731 residues are identical. Implementation of homologous amino acid substitutions, as described for porcine gelsolin by Way and Weeds 1988, increases the sequence identify to 96.7%. Comparison of horse cytoplasmic gelsolin cDNA with the equivalent found in the two other non-human mammalian species for which sequences have been determined, pig and mouse (Table 2), reveals a similar high degree of identity (Way and Weeds, 1988; Dieffenbach et al, 1989). A sequence identity of over 93% at the amino acid level for these three diverse mammals suggests the presence of evolutionary pressure that preserves the structure of gelsolin. Table 2: Comparison of horse cytoplasmic gelsolin cDNA to the human, pig, and mouse sequences Spec ies I d e n t i c a l I d e n t i c a l H o m o l o g o u s P r o t e i n M o l e c u l a r Nuc leo t ides A m i n o A c i d s Rep lacements M a s s (daltons) Human 2010 (91.5%) 692 (94.7%) Pig 2021 (92.0%) 690 (94.5%) Mouse 1939 (88.3%) 680 (93.2%) 15 (96.7%) 15 (96.6%) 24 (96.0%) 80 509 80 607 80 746 102 viL Plasma Gelsolin N-terminal Amino Acid Sequence The sequence of the N-terminal extension that distinguishes the horse plasma protein from the cytoplasmic form was determined by Edman protein degradation. Purified horse plasma gelsolin was submitted for analysis to the Nucleic Acid Protein Service (N.A.P.S) at the Biotechnology Laboratory, University of British Columbia. Phenyl-thiohydantoin amino acids produced in the reaction were separated by C i s reverse phase HPLC and detected by UV absorbance. After 19 cycles on the automated sequencer, non-specific hydrolysis became apparent. This resulted in an increase in background noise, which may have introduced some uncertainty to the assignments of residues 20 and 21 in the horse sequence (Figure 23). The cDNA sequence that codes for this stretch of amino acids could not be produced by PCR methods. 10 20 25 I I I HORSE A T A S R G A A Q A R A P Q G R A K P Q K P S S M I I I II I I I I II II I I I HUMAN A T A S R G A S Q A G A P Q G R V P E A R P N S M I I I I II I I I I II I I I I PIG A T A S R G A P Q A R A P Q G R V S P M R P S T M T N-terminal residue of secreted pig plasma gelsolin Figure 23: Sequence comparison of the N-terminal plasma extension for horse, human and pig plasma gelsolins. Residues 1-21 of the horse sequence were determined entirely by protein sequencing. The N-terminal sequences of the human and pig proteins were confirmed by cDNA clones. The actual N-terminus of purified pig gelsolin, a valine, is illustrated in boldface; identical residues are indicated by the ' I' symbol. 103 Multiple alignment of horse, human, and pig plasma gelsolin N-terminal sequences illustrates 16 identical residues (Figure 23). This 25 amino acid plasma extension is the area in the primary structure that exhibits the largest degree of sequence diversity between these species. Interestingly, both horse and human plasma gelsolins are proteolytically cleaved at the same locations in the N-terminus. Pig plasma gelsolin on the other hand is isolated with only a 9 residue plasma extension, whereas secreted gelsolin in chicken does not contain a plasma extension (Way and Weeds, 1988; Nodes et al, 1987). Since cDNA clones show that the N-terminus of pig plasma gelsolin is similar to that of the horse and human, it is likely that additional proteolytic degradation occurs in the pig protein, perhaps as a consequence of signal peptidase activity. The methionine amino acid at position 25 in the plasma sequences corresponds to the initiation residue of each of the cytoplasmic forms of gelsolin. This amino acid is hydrolyzed from the polypeptide chain at some time during protein translation, as it is not found in any purified cytoplasmic gelsolins. viii. Amino Acid Sequence of Horse Plasma Gelsolin The combination of protein and cDNA sequence data produced the entire primary structure for equine plasma gelsolin. Of the total 755 residues, the first 21 were identified by N-terminal protein sequencing. The rest were obtained from gelsolin cDNA fragments 1, 2, and 3. Protein and cDNA sequences overlap at residue 22. A n alignment of full-length horse plasma gelsolin with human plasma gelsolin is presented in Figure 24. The two sequences have 708 out of 755 104 residues in common, yielding a 93.8% sequence identity. The slight decrease in sequence similarity, relative to the horse and human cytoplasmic isoforms (94.7%), is the direct result of the amino acid diversity exhibited at the N-terminus of the two plasma proteins. 10 1 20 i 30 i 40 1 50 I 60 I 70 1 1 ATASRGAAQA * 1 RAPQGRAKPQ * * * * * 1 KPSSMWEHP * * 1 EFLKAGKEPG 1 LQIWRVEKFD 1 LVPVPPNLYG * i DFFTGDAYVI ATASRGASQA GAPQGRVPEA RPNSMWEHP EFLKAGKEPG LQIWRVEKFD LVPVPTNLYG DFFTGDAYVI 80 1 90 i 100 i 110 t 120 I 130 I 140 1 1 LKTVQLRNGI * 1 LQYDLHYWLG 1 NECSQDESGA 1 AAIFTVQLDD I YLNGRAVQHR I EVQGFESATF 1 LGYFKSGLKY LKTVQLRNGN LQYDLHYWLG NECSQDESGA AAIFTVQLDD YLNGRAVQHR EVQGFESATF LGYFKSGLKY 150 i 160 i 170 i 180 i 190 1 200 I 210 I 1 KKGGVASGFK 1 HWPNEVWQ 1 RLLQVKGRRV * I VRATEVPVSW 1 ESFNNGDCFI 1 LDLGNNIYQW * i CGSKSNRFER * * KKGGVASGFK HWPNEVWQ RLFQVKGRRV VRATEVPVSW ESFNNGDCFI LDLGNNIHQW CGSNSNRYER .,. 220 .230 . .240 .250. . • 260 270 280 I I I I I I I LKATQVSKGI RDNERSGRAQ VSVFEEGAEP EAMLQVLGPK PTLPEATEDT VKEDAANRKL AKLYKVSNGA * * * * LKATQVSKGI RDNERSGRAR VHVSEEGTEP EAMLQVLGPK PALPAGTEDT AKEDAANRKL AKLYKVSNGA 290 300 310 320 330 340 350 I I I I I I I GPMWSLVAD ENPFAQGALR SEDCFILDHG KDGKIFVWKG KQANMEERKA ALKTASDFIS KMDYPKQTQV * * * * * GTMSVSLVAD ENPFAQGALK SEDCFILDHG KDGKIFVWKG KQANTEERKA ALKTASDFIT KMDYPKQTQV 360 370 380 390 400 410 420 I I I I I I I SVLPEGGETP LFRQFFKNWR DPDQTEGLGL AYLSSHIAHV ERVPFDAATL HTSTAMAAQH GMDDDGTGQK * * * * SVLPEGGETP LFKQFFKNWR DPDQTDGLGL SYLSSHIANV ERVPFDAATL HTSTAMAAQH GMDDDGTGQK 430 440 450 460 470 480 490 I I I I I I I QIWRVEGSNK VPVDPATYGQ FYGGDSYIIL YNYRHGSRQG QIIYNWQGAQ STQDEVAASA ILTAQLDEEL * * QIWRIEGSNK VPVDPATYGQ FYGGDSYIIL YNYRHGGRQG QIIYNWQGAQ STQDEVAASA ILTAQLDEEL 500 510 520 530 540 550 560 I I I I I I I GGTPVQSRW QGKEPAHLMS LFGGKPMIVY KGGTSREGGQ TAPASTRLFQ VRASSSGATR AVEIIPKAGA * * * * * GGTPVQSRW QGKEPAHLMS LFGGKPMIIY KGGTSREGGQ TAPASTRLFQ VRANSAGATR AVEVLPKAGA 105 570 580 590 600 610 620 630 I I I I I I I LNSNDAFVLK TPSAAYLWVG AGASEAEKTG AQELLRVLRA QPVQVAEGSE PDSFWEALGG KATYRTSPRL * * * LNSNDAFVLK TPSAAYLWVG TGASEAEKTG AQELLRVLRA QPVQVAEGSE PDGFWEALGG KAAYRTSPRL 640 650 660 670 680 690 700 I I I I I I I KDKKMDAHPP RLFACSNKIG RFVIEEVPGE FMQEDLATDD VMLLDTWDQV FVWVGKDSQD EEKTEALTSA * * KDKKMDAHPP.RLFACSNKIG RFVIEEVPGE LMQEDLATDD VMLLDTWDQV FVWVGKDSQE EEKTEALTSA 710 720 730 740 750 I I I KRYIDTDPAH R D R R T P I T W KQGFEPPSFV GWFLGWDDSY WSVDPLDRAL AELAA * * * * KRYIETDPAN R D R R T P I T W KQGFEPPSFV GWFLGWDDDY WSVDPLDRAM AELAA Figure 24: Comparison of horse (top) and human (bottom) plasma gelsolin protein sequences. Non-identical amino acids are indicated by a '*' symbol. Horse plasma gelsolin has a predicted molecular mass of 83 201 daltons. Experimentally, electrospray mass spectrometry analysed a single time determined a mass of 83 844 ± 17. A total of 96 aspartic and glutamic residues, along with 87 positive lysine and arginine residues produces an estimated net charge of -8.0 at neutral pH, with the isoelectric point near 5.9. Assuming that the molar extinction coefficients of tyrosine, tryptophan, and phenylalanine residues in the protein are independent, they can be combined to give an estimation of the protein's extinction coefficient. At 280 nm, the calculated extinction coefficient of horse plasma gelsolin is 1.37 m L m g - i - c m 1 , while differential refractometry yielded an experimental value of 1.4 m L - m g ^ c m " 1 (Ruiz Silva and Burtnick, 1990). 106 ix Partial Exan Organization of Plasma Gelsolin The 6.7 kb horse genomic clone (Figure 18, lane B), that was isolated from the phage library, provided limited gene organization information. This clone was sequenced in both directions, using gelsolin #1 to advance to the 3'-end of the gene, and its complement to move towards the 5'-end. Gene walking was accomplished by making synthetic primers to those areas for which new sequence information had just been determined. Moving in the 3' direction starting from gelsolin primer #1, over 350 nucleotide bases were sequenced. Considerably more walking was done towards the 5'-end of the gene in hopes of finding the DNA that coded for the plasma extension. In this direction, a total of 2143 bases pairs were sequenced (entire data not shown). Analysis of this genomic sequence revealed the exon that codes for the amino terminal residues of the cytoplasmic horse protein (Figure 25) is identical in size to human exon 4, the first exon that codes for amino acids in human cytoplasmic gelsolin (Kwiatkowski et al, 1988). The non-coding sequence that flanks the exon follows the general observation that introns at their 5'-end begin with T G and terminate with A G at their 3'-end. The 206 nucleotides in this exon (underlined sequence) code for 68 residues in gelsolin. The amino acid sequence that is translated from this coding region begins with Pro22 and ends with Leu89 (Figure 24). The surrounding intronic sequence of the horse does not match the sequence that surrounds human exon 4. Within the 2143 bases that were directed towards the 5'-end of the horse gelsolin gene, no exonic sequence was detected. The last few 107 nucleotides of this clone constituted a Xba I restriction site. As this fragment was produced by Xba I digestion of recombinant E M B L 3 phage DNA, the 5'-end of this subclone had been located. Assuming that the gene organization of the horse equivalent of exon 3 is similar to the human, the D N A that codes for the horse plasma extension should be located approximately 200 bases upstream of the end of this genomic clone. Human gelsolin gene organization shows that exon 3, which contains the codons for the N-terminal plasma extension, is located 2.3 kb upstream from exon 4 (Kwiatkowski et al, 1988). Numerous attempts to generate an overlapping genomic subclone to this one were unsuccessful. 108 10 20 30 40 50 60 70 I I I I I I I TCAAGTGAGA TGCTCTTGGT GCTGGAAACC CCCCTCCCTC GTGCCCTGGG GTCTCTCCGC AAGCTCCTGC AGTTCACTCT ACGAGAACCA CGACCTTTGG GGGGAGGGAG CACGGGACCC CAGAGAGGCG TTCGAGGACG 80 90 100 110 120 130 140 I I I I . I I I CCTGCTCAGC CTCCGCCCCG TCTCTTCCAG CCCAGCAGCA TGGTGGTAGA ACACCCCGAG TTCCTCAAGG GGACGAGTCG GAGGCGGGGC AGAGAAGGTC GGGTCGTCGT ACCACCATCT TGTGGGGCTC AAGGAGTTCC 150 160 170 180 190 200 210 I I I I I I I CGGGGAAGGA GCCTGGCCTG CAGATCTGGC GCGTGGAGAA GTTCGACCTG GTGCCCGTGC CCCCCAACCT GCCCCTTCCT CGGACCGGAC GTCTAGACCG CGCACCTCTT CAAGCTGGAC CACGGGCACG GGGGGTTGGA 220 230 240 250 260 270 280 I I I I I I I TTACGGAGAC TTCTTCACAG GTGACGCCTA TGTCATCCTG AAGACGGTGC AGCTGAGGAA CGGGATCCTG AATGCCTCTG AAGAAGTGTC CACTGCGGAT ACAGTAGGAC TTCTGCCACG TCGACTCCTT GCCCTAGGAC 290 300 ' . 310 320 330 ' 340 350 I I I I I I I CAGTACGACC TCCACTACTG GCTGGGGTAG CCGGCCCTGC CCAGCCCCCT CCCCAGAAAG CACAGACCTC GTCATGCTGG AGGTGATGAC CGACCCACTC GGCCGGGACG GGTCGGGGGA GGGGTCTTTC GTGTCTGGAG 360 370 380 390 400 I I I I I TGAGGTGTGG TCCCCAGGAA GCACCCATGG GCGCTAAGTG ACTGTTGAGC ATCAAG ACTCCACACC AGGGGTCCTT CGTGGGTACC CGCGATTCAC TGACAACTCG TAGTTC Figure 25: Exon-intron arrangement of the portion of the horse gelsolin gene that codes for residues Pro22 to Leu89 in plasma gelsolin. This stretch, which includes all the N-terminal residues of cytoplasmic gelsolin, contains gene splicing sites that are located at identical positions as in human exon 4. Exon sequence which is underlined was identified by comparison with cDNA. Only 100 bases of intronic sequence that surround both sides of the exon are shown. Boldface nucleotides designate the start (GT) and end (AG) of the intronic sequence. 109 2. CD AND SECONDARY STRUCTURE ANALYSIS i. Characterization by CD In a calcium free environment, the far U V C D spectrum of horse plasma gelsolin shows a broad minimum centered at 209 nm, with a mean residual ellipticity of -8680 deg c m 2 d m o H (Figure 26). Addition of C a C l 2 to 2 m M diminished the magnitude of the CD reading nearly 15% to -7411 deg c m 2 d m o l 1 . A similar decrease in ellipticity was also evident near 190 nm. These changes support previous reports that C a 2 + ions are able to induce a conformational change, one which opens up the structure of the protein (Doi et al, 1990; Koepf and Burtnick, 1993). 10 Wavelength (nm) Figure 26: Far U V C D spectrum of 1.0 mg/mL horse plasma gelsolin in 150 m M KC1, 25 m M Tris-HCl, and 1.0 m M EDTA, pH 8.0 (bold line), and in 150 m M KC1, 25 m M Tris-HCl, 2 m M C a C l 2 and 1.0 m M EDTA, pH 8.0 (thin line). 110 Additional evidence for a C a 2 + induced conformational change was sought by monitoring the near UV CD of gelsolin (Figure 27). However, both the apo and holo forms of the protein yielded similar spectra between 240 and 310 nm. A single trough with an ellipticity of -125 deg c m 2 dmol" 1 was observed in both conditions near 275 nm, indicative of aromatic amino acids buried in chiral environments. Since the signal to noise ratio for the C D of aromatic and disulfide species is much lower than that for peptide bonds, the near UV signal may not be sensitive enough to report the conformational change detected in the far UV. Alternatively, the conformational change may have occurred at a site in the protein molecule that is deficient in near UV CD reporter groups. 0 -150 H • 1 < 1 1 1 1 • 1 > 1 240 260 280 300 320 340 360 Wavelength (nm) Figure 27: Near U V CD spectra of horse plasma gelsolin (1.0 mg/mL) in 150 m M KC1, 25 m M Tris-HCl 1.0 m M EDTA, pH 8.0 (bold line), and in 150 m M KC1, 25 mM Tris-HCl, 2.0 mM C a C l 2 , 1.0 m M EDTA, pH 8.0. (thin line). I l l ii. Secondary Structure Analysis A n estimation of the secondary structure composition of native gelsolin in the Ca 2 + - f ree form was made with the convex constraint analysis (CCA) method. Deconvolution of the horse plasma gelsolin far U V spectrum calculates 16% oc-helix, 23% (3-structure, 6% antiparallel P-sheets and p-turns, and 22% undefined structure. Aromatics and disulfides accounted for the remaining 33% of the C D signal. These secondary structural weights multiplied together with their resultant deconvoluted curves yield a reconstituted spectrum which is nearly identical to the experimental C D spectrum of gelsolin (Figure 28). o s s 6t • © © 190 200 210 220 230 240 250 W a v e l e n g t h ( n m ) Figure 28: Reconstruction of the far U V CD spectrum of horse plasma gelsolin with the convex constraint algorithm. (•), experimental C D data of 1.0 m g / m L gelsolin in 150 mM KC1, 25 mM Tris-HCl, and 1.0 m M EDTA, pH 8.0 entered into the algorithm. (O), reconstituted spectrum from C C A deconvoluted component curves and statistical weights. 112 CCA analysis of gelsolin CD data obtained in the calcium bound form suggested 17% cc-helix, 21% P-structure, 11% antiparallel p-sheets and p-turns, 22% undefined structure and 29% due to aromatics and disulfides. These results suggest that the C a 2 + induced conformational change leaves the core helices and P-sheet structures intact, but causes some rearrangement among the less well-defined contributors to the far U V C D . The secondary structure of C a 2 + free gelsolin, determined from its crystal structure, yields conformational weights which are analogous to those derived by the C C A method, 18% a-helix and 23% P-sheets (Kabsch and Sander, 1983). High resolution structural studies of gelsolin domain analogs, villin 14T and severin domain 2, have demonstrated that calcium binding does not induce global conformational rearrangement of the domain (Markus et al, 1994; Schnuchel et al, 1995). The changes in conformation were localized to a few residues. Similar observations are made when the polypeptide backbone of horse plasma gelsolin crystallized in the absence of calcium is superimposed on calcium bound human SI (Burtnick et al, 1996; McLaughlin et al, 1993). These two structures produce an almost exact overlay. The C a 2 + induced structural changes observed in the whole molecule are likely the result of metal binding at a different location, perhaps at the site localized in the carboxy-terminal half of the molecule (Hellweg et al, 1993). Alternatively, interaction with C a 2 + may induce domain shifts within the gelsolin molecule, with the linker regions between domains acting as pivots. 113 3. EQUILIBRIUM UNFOLDING STUDIES i Chemical Denaturation with Guanidine-HCl Guanidine-HCl (Gnd-HCl) induced denaturation of FITC-labeled and native gelsolin was accompanied by diminished ellipticity readings at 215 nm (Figure 29). In the presence of millimolar concentrations of CaCl2, the midpoint of the unfolding transition for FITC-labeled gelsolin occurred at approximately 1.2 M Gnd-HCl (Figure 29a), a value which decreased to about 0.7 M denaturant when conducted in a C a 2 + deficient environment. However, careful visual inspection of the sample revealed that the rapid drop in the CD signal for the Ca 2 + -free case was due to light scattering caused by the onset of protein precipitation. A faint, fine film of protein was visible on the cell surface at Gnd-HCl concentrations above 0.35 M . Therefore, these data were deemed artifactual and are not represented in Figure 29a. In comparison to FITC-labeled gelsolin, denaturation of unlabeled plasma gelsolin was not influenced greatly by the presence or absence of C a 2 + ions (Figure 29b). The half-unfolded situation for both cases occurred at a G n d - H C l concentration near 1.0 M, with the representative curves being very similar throughout the range of denaturant concentration employed. The apparent absence of a C a 2 + effect may be the result of two competing processes. Saturation of the Ca 2 + -binding sites on the gelsolin molecule may induce a conformational change, already mentioned, where the structure of the protein opens up. This should facilitate access of guanidinium ions to otherwise buried residues. Negating this effect may be electrostatic repulsions between the bound C a 2 + and positively charged 114 E e *s E w "a* 0 1 2 3 4 5 Concentration of Gnd-HCl (M) Figure 29: Denaturation with Gnd-HCl at 25°C, monitored at 215 nm. (a) FITC-labeled gelsolin (1.0 mg/mL) in 150 m M KC1, 25 m M Tris-HCl , 2 m M CaCl2 and 1.0 mM EDTA, pH 8.0 to which aliquots of G n d - H C l were added, (b) Aliquots of Gnd-HCl added to unlabeled gelsolin (1.0 mg/mL) in either 150 mM KC1, 25 m M Tris-HCl, and 1.0 m M EDTA, pH 8.0 (•), or in 150 m M KC1, 25 m M Tris-HCl, 2 m M C a C l 2 and 1.0 m M EDTA, pH 8.0 (O). 115 guanidinium ions. A positive change in protein surface charge as a result of bound C a 2 + may hinder the approach of the guanidinium ions, and essentially decrease the effective concentration of the denaturant. The net result could be a denaturation profile that is similar to the one exhibited by native gelsolin in a calcium free solution. At Gnd-HCl concentrations above 2 M , both native and FITC-labeled gelsolins are essentially devoid of any organized secondary structure. Denaturation experiments conducted with FITC-labeled gelsolin and monitored at 276 nm produced data which supported the conclusions based on ellipticities obtained at 215 nm. Again, trials involving fluorescently labeled protein in the absence of free C a 2 + always led to precipitation. Use of a 1.0 cm path length cell in place of the 0.5 mm one made visualization of the precipitate trivial. Initial signs of precipitation were evident at 0.35 M Gnd-HCl and remained so up to 1.6 M . A n increase in denaturant concentration beyond 1.6 M resulted in dissolution of the entire precipitate. Addition of up to 1.0 M KC1 as a screening agent did not prevent the formation of this precipitate, indicating that more than simple ionic strength effects are involved in this phenomenon. During similar denaturation experiments of labeled gelsolin monitored in the near UV, but with solutions adjusted to 2 m M CaCi2, no signs of precipitation were evident (Figure 30). The midpoint of the unfolding process occurred at approximately 1 M Gnd-HCl, similar to data obtained at 215 nm. However, comparison of these denaturation curves (Figures 29a and 30) shows that the far UV signal decreases rapidly upon addition of Gnd-HCl, whereas the near UV signal does not drop away until a concentration of 0.5 M G n d - H C l . The decrease in magnitude of the ellipticity at 215 nm reflects an unfolding of the peptide backbone of the 116 protein, and may represent the loss of secondary structural elements that are known to surround core domain structures. £ e a, ti > C3 "3 Concentration of G n d - H C l (M) Figure 30: Denaturation of 1.0 m g / m L FITC-labeled gelsolin at 2 5 ° C , monitored at 276 nm. Aliquots of stock 8.0 M Gnd-HCl were added to the protein in 150 mM KC1, 25 mM Tris-HCl, 2.0 m M C a C l 2 , and 1.0 m M E D T A , pH 8.0. The protein precipitated from solution if C a C l 2 was not added (data not shown). The plateau observed upon monitoring the C D at 276 nm may indicate that conformationally loose elements, devoid of aromatic residues, denature first. Higher levels of denaturant would access the hydrophobic core of the protein, and remove hydrophobic aromatic residues from chiral environments, resulting in a reduced CD signal. 117 A n additional set of denaturation experiments was conducted with native gelsolin and monitored at 276 nm (Figure 31). The experiments that were conducted in the presence of C a 2 + did not reveal any unusual features, P5 0.0-I ' 1 • 1 • 1 • 1 0 1 2 3 4 Concentration of Gnd-HCl (M) Figure 31: Denaturation of 1.0 mg/mL unlabeled gelsolin at 25°C in: (•), 150 m M KC1, 25 mM Tris-HCl, and 1.0 mM EDTA, pH 8.0, and in (o), 150 m M KC1, 25 m M Tris-HCl, 2 mM C a C l 2 and 1.0 m M EDTA, pH 8.0 to which aliquots of Gnd-HCl were added. and agreed favorably with the results obtained at 215 nm. When repeated in the absence of C a 2 + , the ellipticity values actually became more negative at low Gnd-HCl concentrations. The maximum increase occurred at 0.3 M denaturant, a value approaching the concentration of G n d - H C l that precipitated out FITC-labeled gelsolin. This observation suggests that Gnd-H C l , when present at low concentration, binds to gelsolin and induces localized folding of the polypeptide chain. Such folding shifts aromatic 118 residues into chiral environments. At higher concentrations, G n d - H C l behaves as a standard denaturant and unfolds the protein. Induced folding at low G n d - H C l concentrations has been noted by others (Mayr and Schmid, 1993; Hagihara et al, 1993, and Monera et al, 1994). Since the inclusion of C a 2 + ions in this experiment suppressed the folding of gelsolin, the guanidinium and the C a 2 + ions may compete for the same binding sites on the gelsolin molecule. vi Chemical Denaturation with Urea The denaturation process of gelsolin was also investigated using urea, an electrostatically neutral compound (Figure 32). Experiments conducted with this denaturant showed that modification of gelsolin with FITC did not alter the protein in either the absence (Figure 32a) or the presence (Figure 32b) of free C a 2 + . Both native and FITC-labeled gelsolin in the absence of C a 2 + yielded the half-unfolded situation at 2.1 M urea, a transition that shifted to 1.5 M urea with C a 2 + present. The consequence of opening up the structure of gelsolin upon binding C a 2 + allows urea greater access to internal residues, facilitating the denaturation process. The electrostatic influences that C a 2 + ions exerted on Gnd-HCl appear to be negligible with urea. By virtue of binding C a 2 + ions, the holo form of native gelsolin is thermodynamically more stable than the Ca 2 + -free protein. Accompanying the increase in stability following ligand binding is a structural rearrangement which significantly affects exposure of charged residues to solvent. Loading the gelsolin molecule with two C a 2 + ions induces a significant conformational change, made evident by exposure of actin 119 binding sites, a doubling of the hydrodynamic volume, an increased susceptibility to proteolytic enzymes, and an increased extent to which gelsolin is modified with alkylating reagents (Bryan and Hwo, 1986; Koepf and Burtnick, 1993; Patkowski et al., 1990). Although our interpretation of the half-unfolded situations in figure 32 is oversimplified, as plateaus indicate the real unfolding pathway to consist of numerous, possibly unrelated steps, the suggestion of facilitated access to denaturant upon binding C a 2 + is apparent. Analysis of these data with the assumption that the denaturation of gelsolin is a two-state process that involves a reversible equilibrium between native (N) and denatured (D) states proved unsatisfactory. A two state equilibrium would have followed the relation: H 2 0 . A G U = A G u - m [denaturant] H 2 0 where A G U is the apparent free energy of unfolding in pure water, and m is the slope of the line, which is related to the number of additional denaturant molecules that bind to a protein on denaturation. Significant deviations of our calculated data from this linear relationship (data not shown) suggest that the unfolding process even in urea does not follow a simple two-state mechanism, but likely reflects the formation of various intermediate conformations prior to the D-state. 120 0 1 2 3 4 5 Concentration of Urea (M) Figure 32: Denaturation of 1.0 mg/mL labeled and unlabeled gelsolin with urea at 25°C, monitored at 215 nm. (a) FITC-labeled (o) and unlabeled (•) gelsolin each in 150 m M KC1, 25 mM Tris-HCl, and 1.0 m M EDTA, pH 8.0, to which aliquots of 10 M urea were added, (b) Addition of aliquots of 10 M urea to FITC-labeled (o) and unlabeled (•) gelsolin each in 150 m M KC1, 25 m M Tris-HCl, 2 m M C a C l 2 and 1.0 mM EDTA, pH 8.0. 121 Deviations from a two state unfolding process with both G n d - H C l and urea concentrations may be the result of the molecular packing arrangement found in native gelsolin (Appendix I). The crystallographic model of Ca 2 + - free gelsolin reveals a highly organized structure with the six individual domains packed into a conformationally constrained globular structure. Initial structural unraveling may involve the SI-S3 portion of the molecule breaking contacts with the S4-S6. Viewed as a clam shell opening up, this structural change would enhance the exposure of the six domains and their linker regions to denaturant molecules. Once presented to solvent, the six domains of gelsolin may unfold independently, each perhaps exhibiting two state behaviour. If the pathway or extent of unfolding varies amongst the six domains, the signals we observe would not show simple two state behaviour, but would appear as a complex multi-state denaturation profile. Experimental evidence for independent two state models may eventually come from characterization of the denaturation profiles of individual gelsolin domains. UL Thermal Denaturation Studies A series of thermal denaturation experiments were conducted with solutions of gelsolin in 150 mM KC1, 25 mM Tris-HCl, 1 mM EDTA, pH 8.0 in either the presence or absence of 2 m M CaCi2, and in the presence and absence of 0.3 M Gnd-HCl . At 15°C, all solutions were optically clear. O n heating, unlabeled gelsolin started to precipitate at about 50°C, regardless if C a 2 + was present or not (Figures 33a and 33b). In 0.3 M G n d - H C l , precipitation was evident at 35°C for samples of unlabeled gelsolin, again independently of C a 2 + concentration. The ability of G n d - H C l to induce 122 localized folding at room temperature does not translate into structural stability at elevated temperatures. Intermolecular aggregation that leads to precipitation is enhanced with the presence of 0.3 M Gnd-HCl at elevated temperatures. 1.1 Temperature (°C) Figure 33: Thermal denaturation of 1.0 m g / m L unlabeled gelsolin, monitored at 215 nm. (a) Unlabeled gelsolin in 150 m M KC1, 25 m M Tris-H C l , and 1.0 m M EDTA, pH 8.0 with the absence (•), or presence (o), of 0.3 M Gnd-HCl . fb) Unlabeled gelsolin in 150 mM KC1, 25 m M Tris-HCl, 2 m M CaCl2 and 1.0 m M EDTA, pH 8.0 with the absence (•), or presence (o), of 0.3 M Gnd-HCl. 123 FITC-labeled gelsolin is considerably more resistant to thermally induced precipitation than is unlabeled gelsolin (Koepf and Burtnick, 1993). Replacing positively charged side chains on gelsolin with negatively charged fluorescein moieties likely hinders protein aggregation. Precipitation of the fluorescently labeled protein in a calcium and Gnd-HCl deficient buffer first became evident at 58°C. The temperature at which precipitation was visible decreased slightly to 5 5 ° C in the presence of C a 2 + . However, prior to precipitation, the structure of FITC-gelsolin in calcium containing solutions was considerably more unfolded (Figures 34a and 34b). The conformational change induced by calcium binding is clearly a net destabilizing effect. Thermal denaturation of FITC-labeled gelsolin was also investigated with a set of experiments analogous to those for native gelsolin. As was the case for native gelsolin, G n d - H C l at 0.3 M served to enhance protein aggregation (Figures 34a and 34b). With 0.3 M Gnd-HCl present, FITC-labeled gelsolin without C a 2 + started to precipitate near 35°C, similar to solutions that contained unlabeled gelsolin. Inclusion of C a 2 + raised the point of detectable turbidity to 4 2 ° C . This elevation in temperature, relative to the C a 2 + free state, may be due to alterations of gelsolin's surface charge resulting from the creation of new cation binding sites by the presence of the negatively charged fluorescein groups. 124 1.1-1.0" 0.9-0.8-s s 0.7" i H • 0.6-• 0.5-•«-» a • 1.0-0.9-CQ 0.8-0.7-0.6-0.5-o-a™...^ ^ ~x>—-a \ \ b i i —o-1 0 —i 1 1 1 1 • 1 • 1 ' — 20 30 40 50 60 70 Temperature (°C) Figure 34: Thermal denaturation of 1.0 m g / m L FITC-labeled gelsolin, monitored at 215 nm. (a) Labeled gelsolin in 150 m M KC1, 25 m M Tris-H C l , and 1.0 m M EDTA, pH 8.0 with the absence (•), or presence (O), of 0.3 M Gnd-HCl . (b) Labeled gelsolin in 150 m M KC1, 25 m M Tris-HCl, 2 m M CaCl2 and 1.0 m M EDTA, pH 8.0 with the absence (•), or presence (O), of 0.3 M Gnd-HCl. 125 4. CRYSTALLIZATION TRIALS i Crystals of Horse Plasma Gelsolin Conditions for crystallization of gelsolin produced in this laboratory were screened both here and in the Laboratory of Molecular Biophysics, Oxford University. At both locations, C a 2 + free horse plasma gelsolin crystals were grown successfully at room temperature by the vapor diffusion technique, from solutions that contained 30% to 40% saturated ammonium sulfate in the pH range from 7 to 8. Optimum conditions were at 34% saturation and pH 7.5, similar to conditions reported for human plasma gelsolin (McLaughlin and Gooch, 1992). The crystal morphology was a flat plate, with maximum dimensions of 1.0 mm x 0.5 mm x 0.5 mm (Figure 35). The primitive tetragonal unit cell contained 16 gelsolin molecules, two per asymmetric unit, with approximately 55% crystallization solvent (Burtnick et ah, 1996). These crystals diffracted to a resolution of 2.5 A . Identical crystallization conditions but with CaCl2 added to 2 m M did not produce any visible crystals. Solutions consisting of 20% to 24% saturated sodium potassium tartrate produced C a 2 + bound gelsolin crystals that were similar in size and shape to gelsolin grown in the absence of C a 2 + . Physical characterization and diffraction studies with crystals grown under these conditions have not been completed. 126 Figure 35: Crystals of horse plasma gelsolin grown in a C a 2 + free environment. 127 5. SUMMARY £ Summary of Results Equine gelsolin isolated from blood plasma is structurally and functionally very similar to human plasma gelsolin. A series of four overlapping DNA clones produced the entire coding region for the cytoplasmic form of the protein, while protein sequencing identified the first 21 amino acids of the 25 residue N-terminal plasma extension. Processing at the N-terminus of horse gelsolin occurs at a position equivalent to that in human, both proteins are isolated with a 25 residue plasma extension. In comparison, pig plasma gelsolin is isolated with only a nine residue extension, while no extension is found in secreted mouse gelsolin. Alignment of horse and human plasma gelsolin protein sequences shows that the two proteins are nearly 94% identical. Genomic sequence covering a limited region of the horse gelsolin gene maps favorably to human gelsolin exon 4, the exon that codes for all residues found at the N-terminus of the cytoplasmic protein (Pro22 to Lys89). The exon identified in the horse gene consists of 206 nucleotides, the exact number found in human exon 4. In human, the preceding exon, which codes for the N-terminal plasma extension, is located 2.3 kb upstream from exon 4 (Kwiatkowski et al, 1988). Although not mapped, the horse equivalent of exon 3 was not located in 2.1 kb of upstream sequence. It is likely that horse gelsolin exon 3 is structured in a manner similar to human exon 3, which may also be the case for the entire horse gelsolin gene. The behavior of purified horse plasma gelsolin towards chemical 128 denaturants depends greatly on the surface charge of the protein. Changes in surface charge are evident upon binding cations such as guanidinium + and C a 2 + , which not only affect the structure in solution, but also modify the pathway for unfolding, and moderate the onset of precipitation. Solutions of FITC-labeled gelsolin that were devoid of C a 2 + ions always precipitated upon addition of Gnd-HCl. Screening agents were not able to prevent this phenomenon, but G n d - H C l concentrations greater than 1.6 M did dissolve the precipitate. In comparison, native protein in G n d - H C l was not significantly affected by the presence or absence of C a 2 + when monitored at 215 nm. However, induced local folding of aromatic amino acids into chiral environments was evident at low G n d - H C l concentrations, a phenomenon which was suppressed by C a 2 + ions. Localized folding did not translate into structural stability at elevated temperatures. Chemically induced denaturation by electrostatically neutral urea was not influenced by the presence of the bulky fluorescein groups. Rather, results from these studies provided additional evidence for a C a 2 + induced conformational change, one where the structure of gelsolin becomes more exposed to the solvent, an overall net destabilizing effect. The crystal structure of whole horse plasma gelsolin contains 18% a-helix and 23% p-structure, values which agree favorably with secondary structure predictions made from C D data. The structure of the entire molecule gives us the first view of the six domains of gelsolin, initially predicted from sequence analysis. Generation of a model that incorporates the human SI G-actin structure into the Heidelberg F-actin filament lets us visualize how the structure of gelsolin relates to function. Placement of whole horse gelsolin 129 onto the model structure, so that human SI and horse SI segments overlap, provides a view of a gelsolin capped filament and the A 2 G nucleation complex, but does not adequately illustrate the severing function. For this dynamic process to occur, individual domain shifts may be required. Regulation of gelsolin's biological activity is attributed to a global protein conformational change induced upon binding C a 2 + . The structure of Ca 2 + - free gelsolin presented in Figure A2-a illustrates the organization of the six individual domains into a constrained globular molecule. As this form of gelsolin is unable to bind to either G or F-actin, the critical actin binding residues localized in domains SI, S2 and S4 must be masked from contact with actin in the surrounding solution. Exposure of the actin binding sites following interaction with C a 2 + likely reflects a rearrangement of segments SI-S3 with respect to S4-S6, perhaps with the 50 amino acid residue linker acting as a flexible tether. Such a structural rearrangement may produce an "open" gelsolin conformation similar to that presented in Figure A2-b. This "open" structure not only exposes previously hidden actin binding sites, it illustrates a molecular size sufficient to accommodate two actin monomers per gelsolin, consistent with the observed A 2 G complex. Experimental observations that binding C a 2 + doubles gelsolin's hydrodynamic volume, increases its susceptibility towards proteolysis, facilitates chemically induced denaturation, and enhances its reactivity toward chemical reagents are all consistent with a structure that is significantly larger and more exposed to solvent than the Ca 2 + - f ree form (Patkowski et al, 1990; Bryan and Hwo; 1986, Koepf and Burtnick, 1993). Deletion of human plasma gelsolin residues Phe733 to Ala755 130 eliminates C a 2 + regulation of activity (Kwiatkowski et al, 1989), due to removal of a high affinity C a 2 + - b i n d i n g site (Pope et al, 1995). In the crystallographic model of gelsolin (Figure A2-a), a portion of this stretch of sequence is found as a small helix that follows S6 (orange). This helix is intimately associated with the large helix of S2 (light green). Binding of C a 2 + at the C-terminus of S6 may lead to disruption of these helix-helix contacts and promotion of a more open structure. The interactions between these two helices may be the critical structural contacts that prevent Ca 2 + - f ree gelsolin from binding to and regulating actin filament assembly and disassembly. it Suggestions for Future Studies The high resolution structure of horse plasma gelsolin has given us the means to relate the physical structure of gelsolin to its biological functions. The next phase of our research is being directed towards growing gelsolin crystals from C a 2 + rich environments. The crystal structure of C a 2 + bound gelsolin should provide insight into the nature of the conformational change that is detected upon binding the metal ligand, as well as test the idea that domain shifts are important for function. Additional crystallization efforts should be made towards co-crystallizing whole gelsolin in a complex with actin, with an emphasis on elucidating the A 2 G nucleation complex. Construction and characterization of gelsolin mutants may provide evidence for our proposed mechanisms of function. If gelsolin segments SI-S3 and S4-S6 do function as independent entities, joined by only a polypeptide tether, segmental deletions or insertions in this region could 131 drastically alter gelsolin's activities. Alterations of the intervening stretches between halves or between individual domains should not affect core domain structures, but may impede their ability to shift positions relative to one another. In all likelihood, a combination of experimental approaches will be required to fully elucidate all of gelsolin's complex biological activities. These will include mutants of gelsolin, characterized by both X-ray and kinetic studies. 132 REFERENCES Anfinson, C. (1973) Science 181, 223-227 Angeletti, B., Battiloro, E. , Pascale, E . , and D'Ambrosio, E . (1995) Nucleic Acid Res. 23, 879-880 Arpin, M . , Pringault, E . , Finidori, J . , Garcia, A. , and Jeltsch, J . - M . (1988) J. Cell Biol. 107, 1759-1766 Bazari, W.L., Matsudaira, P., Wallek, M . , Smeal, T., Jakes, R., and Ahmed, Y., (1988) Proc. Natl. Acad. Set USA 85, 4986-4990 Bernhardt, R., Dao, N., Stiel, H. , Schwarze, W., Friedrich J . , Janig, G.-R., and Ruckpaul, K. (1983) Biochim. et Biophys. Acta 745, 140-148 Birnboim, H.C. , and Doly, J . (1979) Nucleic Acid Res. 7, 1513-1523 Bradford, M . (1976) Anal Biochem. 72, 248-254 Bryan, J . , and Kurth, M.C. (1984) J . Biol. Chem 259, 7480-7487 Bryan, J . , and Hwo, S. (1986) J. Cell Biol. 102, 1439-1446 Bryan, J . (1988) J . Cell Biol 106, 1553-1562 Burtnick, L .D. (1984) Biochim et Biophys. Acta 791, 57-62 Burtnick, L .D. , Koepf, E.K., Grimes, J . M . , Jones, E.Y., Stuart, D.I.H., McLaughlin, P.J., and Robinson, R.C. (1996) Submitted for publication Cantor, C , and Schimmel, P. (1980) "Biophysical Chemistry Part II: Techniques for the Study of Biological Structure and Function", W. H . Freeman and Company, San Francisco Carlier, M.-F . (1991) J . Biol Chem 266, 1-4 Chaponnier, C , Janmey, P.A., and Yin, H.L. (1986) J. Cell Biol 103, 1473-1481 Chomczynski, P., and Sacchi, N. (1987) Anal Biochem 162, 156-159 Coue, M . , and Korn, E .D. (1985) J . Biol. Chem 260, 15033-15041 Coue, M . , Constans, J . , and Olomucki, A. (1986) Eur. J. Biochem 160, 273-277 133 Dieffenbach, C.W., SenGupta, D.N., Krause, D., Sawzak, D., and Silverman, R.H. (1989) J. Biol Chem. 264, 13281-13288 Doi, Y., Kim, F., and Kido, S. (1990) Biochemistry 29, 1392-1397 Edwards, J .B. , Delort, J . , and Mallet, J . (1991) Nucleic Acid Res. 19, 5227-5232 Ehrlich, H.A. (1989) "Principles and Applications for DNA Amplification", M Stockton Press, New York Feinberg, J . , Benyamin, Y., and Roustan, C. (1995) Biochem Biophys. Res. Com. 209 426-432 Frohman, M.A., Dush, M.K., and Martin, G.R. (1988) Proc. Natl. Acad. Set (USA) 85, 8998-9002 Glusker, J.P., and Trueblood, K.N. (1985) "Crystal Structure Analysis" (2nd edition), Oxford University Press, New York Gordon, D.J . , Yang, Y.-Z., and Korn, E .D. (1976) J . Biol. Chem 251, 7474-7479 Greenfield, N. , and Fasman, G.D. (1969) Biochemistry 8, 4108-4116 Hagihara, Y., Aimoto, S., Fink, A.L. , and Goto, Y. (1993) J. Mol. Biol 231, 180-184 Harris, H .E . , Bamburg, J.R., and Weeds, A . G . (1980) FEBS Lett. 121, 175-182 Harris, D.A., and Schwartz, J . H . (1981) Proc. Natl. Acad. Set USA 78, 6798-6802 Harris, H . E . (1985) Biochemistry 24, 6613-6618 Hartwig, J . H . , and Kwiatkowski, D.J . (1991) Current Opinion in Cell Biology 3, 87-97 Hellweg, T. , Hinssen, H . , and Eimer, W. (1993) Biophysical Journal65, 799-805 Hennessey, E.S., Drummond, D.R., and Sparrow, J . C . (1993) Biochemical Journal 282, 657-671 Holmes, K., Popp, D. , Gebhard, W., and Kabsch, W. (1990) Nature 347, 44-49 Horwitz, J.P., Chua, J . , Curby, R.J., Tomson, A.J . , Da Rooge, M.A., Fisher, B.E. , Mauricio, J . , and Klundt, I. (1964) J . Med. Chem. 7, 574-575 134 Hwo, S., and Bryan, J . (1986) J . Cell Biol 102, 227-236 Innis, M.A., Gelfand, D.H. , Sninsky, J.J. , and White, T . J . (1990) "PCR Protocols, A Guide to Methods and Applications," Academic Press Inc., New York Ito, H . , Yamamoto, H. , Yoshihiro, K., Kambe, H . , Okochi, T., and Kishimoto, S. (1990) J . Chromatog. 526, 397-406 Ito, H . , Kambe, H . , Kimura, Y., Nakamura, H. , Hayashi, E. , Kishimoto, T., Kishimoto, S., and Yamamoto, H. (1992) Gastroenterology 102, 1668-1692 Janmey, P.J., Chaponnier, C , Lind, S.E., Zaner, K.S., Stossel, T.P. and Yin, H.L. (1985) Biochemistry 24, 3714-3723 Janmey, P.A., and Stossel, T.P. (1987) Nature 325, 362-364 Janmey, P.A., Iida, K., Yin, H.L., and Stossel, T.P. (1987) J. Biol Chem 262, 12228-12236 Johnson, W.C. (1990) PROTEINS, Structure, Function, Genetics 7, 205-214 Kabsch, W. and Sander, C. (1983) Biopolymers 22, 2577-2637 Kabsch, W., Mannherz, H.G. , Suck, D., Pai, E.F., and Holmes, K .C. (1990) Nature 347, 37-44 Koepf, E.K. , and Burtnick, L .D. (1993) Eur. J. Biochem 212, 713-718 Korn, E .D. (1982) Physiological Reviews 62, 672-737 Kwiatkowski, D.J . , Janmey, P.A., Mole, J .E . , and Yin, H.L. (1985) J. Biol Chem. 260, 15232-15238 Kwiatkowski, D.J . , Stossel, T.P., Orkin, S.H., Mole, J .E . , Colten, H.R., and Yin, H.L. (1986) Nature 323, 455-458 Kwiatkowski, D.J . , Mehl, R., and Yin, H.L. (1988) J . Cell Biol 106, 375-384 Kwiatkowski, D.J . , Mehl, R., Izumo, S., Nadal-Ginard, B., and Yin, H.L. (1988a) J . Biol. Chem 8239-8243 Kwiatkowski, D.J . , Janmey, P.A., and Yin, H.L. (1989) J. Cell Biol 108, 1717-1726 Laemmli, U.K. (1970) Nature 227, 680-685 135 Lamb, J.A., Allen, P.G., Tuan, B.Y., and Janmey, P.J. (1993) J. Biol Chem. 268, 8999-9004 Lees, A. , Haddad, J . G . , and Lin, S. (1984) Biochemistry 23, 3038-3047 Leonard, N.J. , McDonald, J.J. , and Reichmann, M.E . (1970) Proc. Nat. Acad. Set USA 67, 93-98 Lind, S.E., Smith, D.B. , Janmey, P.A., and Stossel, T.P. (1986) J. Clin. Invest. 78, 736-742 Mannherz, H . G . , (1992) J . Biol Chem 267, 11661-11664 Mannherz, H.G. , Gooch, J . , Way, A., Weeds, A . G . , and McLaughlin, P.J. (1992) J . Mol Biol, 226, 899-901 Markus, M.A., Nakayama, T., Matsudaira, P., and Wagner, G. (1994) Protein Science 3, 70-81 Mayr, L . M . and Schmid, F.X. (1993) Biochemistry 32, 7994-7998 McLaughlin, P.J., and Gooch, J . (1992) FEBS Lett. 302, 253-255 McLaughlin, P.J., Gooch, J.T., Mannherz, H . G . , and Weeds, A . G . (1993) Nature 364, 685-692 McPherson, A. (1990) Eur. J . Biochem 189, 1-23 Messing, J . (1993) Methods in Molecular Biology 23, 9-22 Mommaerts, W.F. (1992) Bioessays 14, 57-59 Monera, O.D. , Zhou, N.E. , Kay, C .M. , and Hodges, R.S. (1994) J. Biol Chem 268, 19218-19227 Mullis, K., Faloona, F., Scharf, S., Saiki, R., Horn, G. , and Erlich, H . (1986) Cold Spring Harbor Symposia of Quantitative Biology 51, 263-273 Mullis, K., and Faloona, F. (1987) Methods in Enzymology 155, 335-350 Nodes, B.R., Shackelford, J .E . , and Lebherz, H . G . (1987) J . Biol Chem 262, 5422-5427 Norberg, R., Thorstensson, R., Utter, G. , and Fagraeus, A. (1979) Eur. J. Biochem. 100, 575-583 Pace, C.N. (1986) Methods in Enzymology 131, 266-280 136 Patkowski, A., Seils, J . , Hinssen, H . and Dorfmuller, T H . (1990) Biopolymers 30, 427-435 Perczel, A. , Hollosi, M . , Tusnady, G. , and Fasman, G.D. (1991) Protein Engineering 4, 669-679 Perczel, A. , Park, K., and Fasman, G.D. (1992) Analytical Biochem. 203, 83-93 Perczel, A. , Park, K., and Fasman, G.D. (1992a) PROTEINS, Structure, Function, Genetics 13, 57-69 Pollard, T . D . , and Cooper, J .A. (1986) Ann. Rev. Biochem. 55, 987-1035 Polard, T . D . (1990) Current Opinion in Cell Biology 2, 33-40 Pope, B., Maciver, S., and Weeds, A. (1995) Biochemistry 34, 1583-1588 Reid, S.W., Koepf, E.K., and Burtnick, L.D. (1993) Arch. Biochem Biophys. 302, 83-93 Richardson, C.C. , Masamune, Y., Live, T.R., Jacquemin-Sablon, A., Weiss, B., and Fareed, G. (1968) Cold Spring Harbour Symposia of Quantitative Biology 33, 151-164 Ruiz Silva, B.E. , and Burtnick, L.D. (1990) Biochem Cell Biol. 68, 796-800 Sambrook, J . , Fritsch, E.F., and Maniatis, T. (1989) "Molecular Cloning: A Laboratory Manual," (2nd edition), Cold Spring Harbor Labortatory Press, Cold Spring Harbor, New York Sanger, F., Niklen, S., and Coulson, A.R. (1977) Proc. Nat. Acad. Set USA 74, 5463-5467 Schnuchel, A. , Wiltscheck, R., Eichinger, L., Schleicher, M . , and Holak, T.A. , (1995) J. Mol. Biol. 247, 21-27 Southern, E . M . (1975) J . Mol. Biol. 98, 503-517 Spudich, J .A., and Watt, S. (1971) J. Biol. Chem 246, 4866-4871 Staniforth, R.A., Burston, S.G., Smith, C.J. , Jackson, G.S., Badcoe, I.G., Atkinson, T. , Holbrook, J.J. , and Clarke, A.R. (1993) Biochemistry 32, 3842-3851 Stossel, T.P., Chaponnier, C , Ezzell, R.M. , Hartwig, J . H . , Janmey, P.A., Kwiatkowski, D.J . , Lind, S.E., Smith, D.B., Southwick, F.S., Yin, H.L. , and Zaner, K.S. (1985) Ann. Rev. Cell Biol 1, 353-402 Straub, F.B. (1942) Studies University of Szeged 2, 3-15 137 Stryer, L. (1981) "Biochemistry" Second Edition, W. H . Freeman and Company, New York Tanford, C. (1964) J. Am Chem Soc. 86, 2050-2059 Tellam, R.L. (1991) Archiv. of Biochem Biophys. 288, 185-191 Tuzimura, K., Konno, T., Meguro, H . , Hatano, M. , Murakami, T. , Kashiwabara, K., Saito, K., Kondo, Y., and Suzuki, T. (1977) Anal. Biochem. 81, 167-176 Vieira, J . , and Messing, J . (1987) Methods in Enzymology 153, 3-11 Way, M . , and Weeds, A. (1988) J. Mol Biol 203, 1127-1133 Way, M . , Gooch, J . , Pope, B., and Weeds, A . G . (1989) J. Cell Biol 109, 593-605 Way, M . , Pope, B., and Weeds, A . G . (1992) J. Cell Biol 119, 835-842 Weeds, A. (1982) Nature 296, 811-816 Weeds, A. , Gooch, J . , Pope, B., and Harris, H .E . (1986) Eur. J. Biochem 161, 69-76 Weeds, A. , and Maciver, S. (1993) Current Opinion in Cell Biology. 5, 63-69 Weiss, B. and Richardson C.C. (1967) J. Biol Chem 242, 4270-4278 Yang, J.T., Wu, C.-S., and Martinez, H.M. (1986) Methods in Enzymology 130, 208-269 Yin, H.L. , and Stossel, T.P. (1979) Nature 281, 583-586 Yin, H.L. , Hartwig, J . H . , Maruyama, K., and Stossel, T.P. (1981) J . Biol. Chem. 256, 9693-9697 Yin, H.L. , Kwiatkowski, D.J . , Mole, J .E . , and Cole, F.S. (1984) J . Biol. Chem. 259, 5271-5276 Yin, H.L. (1987) BioEssays 7, 176-179 Yin, H.L. , Iida, K., and Janmey, P.A. (1988) J . Cell Biol 106, 805-812 Yu, F.-X., Zhou, D. , and Yin, H.L. (1991) J . Biol Chem 266, 19269-19275 Zeiner, M . , and Gehring, U. (1994) BioTechniques 17, 1051-1053 138 APPENDIX I i High Resolution Structure of Plasma Gelsolin X-ray diffraction studies with crystals of horse plasma gelsolin produced the first high resolution structure of the whole gelsolin molecule. Data collection and structure elucidation were done by L. D. Burtnick and R. C. Robinson at the Laboratory of Molecular Biophysics at Oxford University (Burtnick et al, 1996). The structure of horse plasma gelsolin was solved by multiple isomorphous replacement methods (MIR) to a resolution of 2.5 A . Heavy metal derivatives were prepared by reacting solvent accessible cysteine residues with thiol specific mercury reagents. Two additional reactive sites were exposed after pretreatment of the protein with DTT. The model of the protein covers horse plasma gelsolin residues Val26 through Ala755. No interpretable electron density was found for the N-terminal plasma extension. The existence of six independent globular domains, initially suggested by proteolytic studies and sequence considerations, was confirmed by the crystal structure of whole horse plasma gelsolin. Each of the six domains shows a similar fold, which consists of a central 5 or 6 stranded J3-sheet that is sandwiched between a large 3.5 to 4.5 turn a-helix that runs parallel to the sheet, and a 1 to 2 turn helix that is oriented approximately perpendicular to the sheet (Figure A l ) . This general structural motif, present in each domain, was also identified in villin 14T and severin domain 2, and may be common to all members of the gelsolin family of actin binding proteins. 139 Figure A l : Schematic representation of the six domains of gelsolin oriented i n approximately the same way. Coloring of the domains is: S I red, S2 light green, S3 yellow, S4 pink, S5 dark green and S6 i n orange. This coloring scheme is used throughout the accompanying figures. 140 Superposition of individual gelsolin domains onto each other shows strong structural relationships between SI and S4, S2 and S5, and S3 and S6. These structural similarities, which are in agreement with observations based upon sequence data, may indicate that a precursor gene coding SI-S3 duplicated to produce whole gelsolin (S1-S6). The SI-S3 precursor may itself have been the product of gene triplication. A gene that coded for a primordial actin binding protein, with structural features similar to the domains of gelsolin, may have been the precursor to the multi-domain actin severing and capping proteins we find today. The six domains of gelsolin pack into a globular shape, with approximate dimensions of 85 A x 55 A x 36 A . Segments SI-S3 and S4-S6 appear to organize into independent halves (Figure A2-a). These are connected by a 50 amino acid residue linker, a stretch of polypeptide that may be of sufficient length to accommodate independent structural shifts. At the center of the molecule is domain S6 (orange), which makes contact with SI-S3 at a number of sites. Perhaps the most important contacts between SI-S3 and S4-S6 are manifested through the association of the long helix of S2 (light green) with the extra helix at the terminus of S6 (orange). Interactions between these two helices may be the basis for the S4-S6 half of the molecule regulating the activity of the SI-S3 half in a manner that depends on the binding of C a 2 + . it Structure-Function Relationship of Gelsolin Modeling studies with gelsolin and actin structures illustrate how gelsolin may cap the fast growing end of filaments. The orientation of horse gelsolin on F-actin was determined in context of the structure of 141 Figure A2: (a) S t ruc ture of whole gelsol in i l l u s t r a t i n g the relative posi t ions of the s i x i n d i v i d u a l d o m a i n s . C o l o r i n g of d o m a i n s i s b a s e d u p o n the s c h e m e def ined i n f igure A l . fb) O r i e n t a t i o n of SI -S3 a n d S4-S6 at the e n d of the H e i d e l b e r g m o d e l of the a c t i n f i l ament i l l u s t r a t e s a potent ia l s t r u c t u r e for the A 2 G n u c l e a t i o n complex . A c t i n m o n o m e r s are presented i n gray. 142 human SI-G-actin incorporated into the Heidelberg model (Holmes et al, 1990) of an actin filament. Whole gelsolin was placed into this composite structure by superimposing horse SI on human SI . This orients gelsolin to a position at the end of the actin filament that could effectively block addition of monomers (Figure A3). The overall shape of gelsolin and the volume it occupies could conceal actin binding contacts from monomers in solution. Severing of a filament by gelsolin is a complex event that requires independent domains to work in unison. The model of horse plasma gelsolin positioned on the human SI F-actin filament does not adequately explain the severing process. The major problem with this model is the position of S2 relative to the filament. S2 is essential for severing activity, but in the static model, this domain points away from the actin filament. For severing to occur, S2 must first bind to the side of the actin filament, which presumably orients SI into a position to destabilize and ultimately break actin-actin contacts. Such a mechanism requires a rearrangement of the individual domains in the whole gelsolin molecule, perhaps with the polypeptide linkers between domains serving as pivot points. Gelsolin nucleates actin polymerization by formation of an A 2 G ternary complex, thus presenting a preformed end onto which monomers may add. Using the human SI F-actin filament model, horse gelsolin SI and S4, within the larger domains SI-S3 and S4-S6, are each positioned on one of the two filament strand ends in the same orientation (Figure A2-b). Disregarding all other actin subunits in the filament, the last two actins, which are bound to gelsolin, may resemble the structure of the A 2 G nucleation complex. The linker that connects segments SI-S3 to S4-S6 is long enough to accommodate this butterfly-like opening of the structure. 143 Figure A3: The structure of whole gelsolin positioned on the human SI F-actin model to show capping. The F-actin binding domain, S2 (light green) points way from the filament i n this model. Individual actin subunits i n this filament are drawn i n gray. 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items