MECHANISM of ACTION of INSECTICIDAL CRYSTAL TOXINS from Bacillus Thuringiensis: BIOPHYSICAL and BIOCHEMICAL ANALYSES of the INSERTION

MECHANISM OF ACTION OF INSECTICIDAL CRYSTAL TOXINS FROM Bacillus thuringiensis: BIOPHYSICAL AND BIOCHEMICAL ANALYSES OF THE INSERTION

OF CRY1A TOXINS INTO INSECT MIDGUT MEMBRANES.

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree

Doctor of Philosophy in the Graduate School of the Ohio State University

MANOJ S. NAIR, M.S.

The Ohio State University

2008

Dissertation Committee

Dr. Donald H. Dean (Advisor) Approved By

Dr. David L. Stetson ______

Dr. Charles E. Bell Advisor

Biophysics Graduate Program

ABSTRACT

The most controversial step in the study of the mechanism of action of insecticidal crystal toxins is that of insertion of the toxin into insect brush border membranes. Conflicting models of insertion of toxin can be categorized into two groups; ones that propose that only certain alpha helices of domain I insert into insect brush border membrane vesicles (BBMV) (Umbrella,

Penknife and Serial Receptor Binding models) and others that propose that most of the toxin inserts into BBMV (Aronson, Buried Dragon and Unchanged Structure models).

Protease protection studies of cysteine mutations from all domains of the toxin showed protection of most of the toxin (a 60 kDa form) similar to the wild type Cry1A toxin, when inserted into insect brush border membranes. Studies on steady state fluorescence measurements of these cysteine residues when bound to artificial vesicles or natural brush border membrane vesicles (BBMV), and fluorescence energy transfer measurements in labeled artificial vesicles suggested that residues from all the domains of the toxin inserted into the membrane.

Residues in the loop 2 of Domain II of the toxin that played a vital role in the insertion of the toxin into insect BBMV were identified. Examination of receptor binding and insertion of mutants of these residues have shown that insertion of Cry1Ab into the membrane is dependent on specific residues at positions in this loop. Absence of phenylalanine or a closely related amino acid such as tryptophan at position 371 allowed initial binding of the toxin to the receptor but compromised the insertion of the toxin into insect membrane, thereby confirming that irreversible binding step of the toxin-BBMV interaction is the critical step in the mode of action

of the toxin where Domain II is a major candidate mediating the step. Fluorescence blue shift studies into artificial and natural membranes also indicated a difference in the partitioning of the toxin into artificial and natural membranes; thereby suggesting that physiological mode of action required the presence of receptors.

Analyses of the oligomeric states of Cry1A toxins have shown that oligomerization of the toxin before insertion into insect membranes is not essential for toxicity as both monomeric and oligomeric forms induced similar levels of toxicity. Oligomers (defined as prepores) isolated under specified conditions from artificial membranes included all three domains of the toxin and receptor derived peptides used to elicit the prepore, suggesting that all domains of the toxin are able to insert into the membrane even as they were bound to receptor peptides. We also add further evidence that binding of two receptors in a serial manner is not essential for toxicity of

Cry proteins.

Our studies present an alternate model where an entire toxin unit (of 60 kDa) binds to a receptor, based on availability and affinity of binding and inserts into insect brush border membrane either as a monomer or oligomeric form of the unit.

iii

This work is dedicated to my parents (Mr.Sadasivan Nair & Mrs. Sathidevi B) and my wife

(Mrs. Anju Gangadharan) and other family members for their unconditional support and

encouragement through the course of my graduate career.

ACKNOWLEDGEMENTS

I would like to express gratitude to my thesis advisor Dr. Donald H. Dean for his faith in my abilities, his encouragement in exploring new avenues, his unrelenting enthusiasm to study

Cry toxins and his wonderful mentorship.

I would like to thank my committee members Dr. David L. Stetson and Dr. Charles E.

Bell for their support in generating this thesis work and in trying times. I would also like to thank

Dr. Thomas L. Clanton, ex-director of Biophysics program for his support and help in my initial advising at the university.

My sincere gratitude to my laboratory members, both current and past for their discussions and questions that enabled me to be more productive. A special thanks to Dr. Xinyan

Liu, postdoctoral research associate in Dean Laboratory for extending her help in several aspects and experiments of my thesis work and for several fruitful discussions in my project. I would also like to express my gratitude to Dr. Oscar Alzate, Sean McClory, Praveen Ramalingam and

Andika Gunadi for their contributions to my work. Other members of the laboratory that I would like to thank include but are not limited to Dr. JongYul Roh, Dr. Pradip Biswas, Betina McNeil,

Yoshio Ikeda and Miao Xiao.

I would like to make a special note of gratitude to Dr. David L. Stetson and Dr. Mary

Chamberlin for their guidance with my electrophysiology experiments and to Dr. David. L.

Stetson for allowing me to use the voltage clamp instrument.

I would like to thank members of the laboratory of Dr. Doug Pfeiffer in the Department of Molecular and Cellular Biochemistry at OSU, especially Drs. Greg Steinbaugh and Warren

Taylor for help with synthesis of lipid vesicles. I would like to express my sincere gratitude to

Biochemistry Department for providing me with resources to perform several of my experiments.

For help with the prepore project, I would like to acknowledge my sincere thanks to Drs.

Alejandra Bravo, Mario Soberon and Isabel Gomez for provision of essential resources and for initial training on preparing the prepore complex.

VITA

1976………………………………………………… Born, Mavelikkara, Kerala, India

1997………………………… Bachelor of Sciences, University of Mumbai, Mumbai, India

1999…………………………. Master of Sciences, University of Mumbai, Mumbai, India

2005 ………………………... Master of Sciences, The Ohio State University, Columbus, Ohio

2003- current Graduate Research Associate, The Ohio State University, Columbus, Ohio

PUBLICATIONS

Research Publications:

2008 Nair MS, Liu XS & Dean DH. Membrane Insertion of the Bacillus thuringiensis Cry1Ab Toxin: Single Mutation in Domain II Block Partitioning into Insect Brush Border Membrane. Biochemistry 47(21) 5814-5822. (http://dx.doi.org/10.1021%2Fbi7014234)

2008 Nair MS & Dean DH. All 3 Domains of Cry1A Toxins Insert into Insect Brush Border Membrane Vesicles (J.Biol.Chem. Epub. July 17) (http://www.jbc.org/cgi/doi/10.1074/jbc.M802895200)

2008 Nair, Manoj S. Mode of Action of Bacillus thuringiensis Cry1Ab Toxin: Role of Domain II Residue in Insertion into Insect Brush Border Membranes. Proceedings of the 22nd Edward F. Hayes Graduate Research Forum. (http://hdl.handle.net/1811/32046)

vii

Abstracts/ Presentations:

2008 22nd Annual Meeting of the Protein Society, San Diego, CA: “All three Domains of Cry1A Toxins Insert into Insect Brush Border Membranes” Protein Science Vol.17supp (1), pp: 184.

2008 Contributed Presentation at the Central Regional Meeting of the American Chemical Society (CERMACS), Columbus, Ohio titled “All 3 Domains of Cry1A Toxins Insert into Insect Brush Border Membranes.”

2008 Biophysics Seminar Series, Physics Department, Ohio University, Athens, Ohio: “Domain II is involved in Insertion of Cry1Ab toxin into Insect Brush Border Membranes.” 2008 Annual Meeting of the American Association for Advancement of Sciences (AAAS): “Single Mutation in Domain II of Cry1Ab Toxin Blocks Insertion into Insect Brush Border Membranes.”

2007 40th Annual meeting of Society for Invertebrate Pathology at Universite Laval, Quebec City, Canada: “Domain II residue F371 is essential for Insertion of Cry1Ab toxin into Insect Brush Border Membranes.”

2003 Annual Colloqium of the Biophysics Program: “Quenching of Fluorescence of 1, 5- IAEDANS- labeled Cry1Ab Toxin upon Partitioning into Small Unilamelar Vesicles.”

1999 International Research Scholar‟s meet held at Institute of Science, Mumbai, India: “Effect of ultraviolet radiations on amino acid production by Bacillus subtilis.”

FIELD OF STUDY

Major Field: Biophysics

Protein-protein interactions; Protein-lipid interactions; Protein oligomerization; Membrane insertion; Toxin- membrane interactions.

viii

TABLE OF CONTENTS

ABSTRACT……………………………………………………………………………………..ii

DEDICATIONS…………………………………………………………………………………iv

ACKNOWLEDGEMENTS…………………………………………………………………….v

VITA……………………………………………………………………………………………vii

LIST OF TABLES………………………………………………………………………………xi

LIST OF FIGURES……………………………………………………………………………xiii

CHAPTER 1: LITERATURE REVIEW………………………………………………………1

1.1 Bacillus thuringiensis & its pesticidal proteins: The Journey:……...………………...... 1 1.2 Biology of Bacillus thuringiensis:……………...…………………………………………...2 1.3 Structure of the crystal toxin:….…………………………………………………………….4 1.4 Mechanism of Action of the toxin:……..……………………………………………………..4 1.5 Models for insertion of insecticidal Cry toxins into insect brush border membranes:…...…..8

CHAPTER 2: ANALYSES OF THE REGIONS OF CRY1A TOXIN THAT INSERTS INTO INSECT BRUSH BORDER AND ARITIFICIAL MEMBRANES………………….20

2.1 Introduction:………………………………………………………………………………20 2.2 Experimental Procedures:...………….……………………………………………………21 2.3 Results:…………………..…………………………………………………………………28 2.4 Discussion:…………………...……………………………………………………………..46

CHAPTER 3: IDENTIFICATION OF MUTATIONS IN DOMAIN II OF CRY1Ab THAT MEDIATE INSERTION OF TOXIN INTO INSECT BRUSH BORDER MEMBRANES………………………………………………………………………………….50

3.1 Introduction:……………………………………………………………………………….50 3.2 Experimental Procedures:...…….………………………………………………………….51 3.3 Results:…………………………..………………………………………………………….56

3.4 Discussion:………………………..…………………………………………………………68

CHAPTER 4: ANALYSES OF THE OLIGOMERIZED FORMS OF CRY1A TOXINS FORMED IN SOLUTION, UPON RECEPTOR BINDING AND IN MEMBRANE…….. 72 4.1 Introduction:……………….……………………………………………………………….72 4.2 Experimental Procedures:……..…..………………………………………………………..74 4.3 Results:…………………………..………………………………………………………….83 4.4 Discussion:……………………..…………………………………………………………..108

CHAPTER 5: DISCUSSION………………………………………………………………....113

5.1 Comments on the current understanding of the process of insertion of Cry1A Toxins:…….113 5.2. Future studies to improve understanding of the mechanism of insertion of toxin into insect membrane:…………………………………………………………………………………………115

BIBLIOGRAPHY……………………………………………………………………………..118

LIST OF TABLES

Table 2.1: Bioassay measurements of Cry 1A toxins and their mutants on first instar larvae of M. sexta using surface contamination method………………………………………………...35

Table 2.2: The rate of ion transport measured from the slope of the linear region of the

decrease in short circuit current remaining (Isc) in M. sexta midgut for Domain I mutants……..44

Table 3.1: Bioassay measurements of Cry1Ab and its mutants on 1st instar larvae of

M. sexta (tobacco hornworms) using surface contamination method…………………………...60

Table 3.2: Binding measurements of the toxin 1Abwt and its mutants to cadherin repeats

11 and 12 using surface plasmon resonance analysis (Biacore)…………………………………61

Table 4.1: Comparison of procedures used to prepare prepore oligomers……………………………...... 82

Table 4.2: Major proteins of known function from Bt crystal or insect receptors identified

using LC-MS/MS analysis of S200 purified prepore samples prepared by Method 1A………104

Table 4.3: Major proteins of known function from Bt crystals or receptors identified using LC-MS/MS analysis of S200 purified prepore samples prepared by Method 2…………105

Table 4.4: Sequence of major peptides (based on the homology scores) identified to

crystal proteins from the LC-MS analysis of the prepore from all tested samples of prepore.106

Table 4.5: Bioassay data for the monomer, solution oligomer, prepore oligomer and

membrane extracted forms of Cry1A toxins tested………………………………………….107

xii

LIST OF FIGURES

Fig.1.1: Three dimensional structures of Cry toxins determined by X-ray crystallography….11

Fig.1.2: The Penknife Model of Insertion of Cry toxins………………………………………12

Fig.1.3: The Umbrella Model of Insertion of Cry toxins………………………………………13

Fig.1.4: Orientation of synthetic helices of Domain I in artificial membranes………………14

Fig.1.5: Model of insertion of toxin as a dimer……………………………………………….15

Fig.1.6: Refined Tetramer Umbrella Model…………………………………………………...16

Fig.1.7: The Signal Transduction Model……………………………………………………….18

Fig.1.8: The Serial Receptor Binding Model……………………………………………………18

Fig.2.1: SDS PAGE gel showing the expression and purification of Cry1Ab toxin in E.coli

DH5 cells………………………………………………………………………………………27

Fig.2.2: Western blot analysis of purified Cry1A toxin used for proteinase K protection assays.30

Fig.2.3: Purified mutant proteins run on 8% SDS PAGE gel…………………...... 31

Fig.2.4: Western blot analysis of 8% SDS-PAGE run of proteinase K protection assay of Cry1A

Domain I mutants bound to BBMV…………………………………………………………..33

xiii

Fig.2.5: Western blot analysis of 8% SDS-PAGE run of proteinase K protection assay of Cry1A

Domains II & III mutants bound to BBMV…………………………………………………..34

Fig. 2.6: Percentage of quenching of fluorescence of 1, 5-IAEDANS tagged cysteine mutants..39

Fig. 2.7: Blue shift in the maximal emission wavelength of each acrylodan labeled mutant……40

Fig.2.8: Steady state fluorescence spectra for selected acrylodan labeled mutants……………41

Fig.2.9: Measurement of efficiency of FRET response………………………………………..42

Fig. 2.10: Voltage clamp response of Cry1Ab (●) to those of Domain I mutants……………...43

Fig. 2.11: Voltage clamp response of Cry1Aa (●) to those in Domain II mutants……………..44

Fig. 2.12: Time required for formation of ion channels in M. sexta midgut / Lag time (T0) determined from voltage clamp measurements………………………………………………….45

Fig 3.1: Proteinase K protection assay of Cry1Ab wt and its mutants………………………….62

Fig 3.2: Circular dichroism spectra of Cry1Ab mutants before () and after () labeling with acrylodan expressed in mol.deg-1.cm-1units……………………………………………….63

Figure 3.3: Steady state fluorescence spectra of Cry1Ab mutants labeled with acrylodan……..65

Fig. 3.4: Steady state fluorescence spectra of Cry1Ab G374C labeled with acrylodan…………66

Fig.3.5: Voltage clamp response of Cry1Abwt () compared to that of BBMV inserting mutant 1AbV171C () and non-BBMV inserting mutants 1AbF371C (▼) and

1AbF371A/V171C ()…………………………………………………………………………..67

Fig.4.1: Schematic procedure [Method 1A] for obtaining prepore from Cry1Ab toxin

using insect brush border membranes…………………………………………………………..79 xiv

Fig.4.2: Schematic procedure [Method 1B] for obtaining prepore from Cry1Ab toxin

in the absence of insect brush border membranes………………………………………………80

Fig.4.3: Schematic procedure [Method 2] for obtaining prepore from Cry1Ab toxin

using insect brush border membranes…………………………………………………………..81

Fig.4.4: SDS PAGE of oligomers and monomers of Cry1Ab toxin in 50 mM Na2CO3 pH10.5 stored in 1mM PMSF for 40-72 hours…………………………………………………..87

Fig. 4.5: Circular dichroism spectra of 2 M Cry1Ab toxin purified out as a monomer (●) and as a solution oligomer ()…………………………………………………………………...88

Fig. 4.6: Sedimentation velocity analyses of the Superdex 200 HR purified toxin

monomers (A) and oligomers formed in solution (B) run on Beckman XL-I………………...... 89

Fig. 4.7: Western blot of Cry1Ab toxin extracted from insect BBMV using n-octyl-β-D-glucopyranoside detergent and treated with 0.5% SDS final concentration……….90

Fig.4.8: Western blot of Cry1Ab toxin extracted from insect BBMV using n-octyl-β-D-glucopyranoside detergent and treated with 1.0% SDS final concentration……….91

Fig. 4.9: Superdex 200 HR gel filtration retention peaks of the monomer form of

Cry1Ab versus the membrane bound forms of the Cry1Ab toxin………………………………92

Fig.4.10: Sedimentation velocity run of the membrane bound fraction of Cry1Ab toxin extracted into β- OG detergent on Beckman XL-I analytical ultracentrifuge…………………. 93

Fig.4.11: CD spectra of Cry1Ab toxin purified into 50 mM Na2CO3 pH 10.5 + 0.5M NaCl (●), bound to BBMV (▼) and bound to BBMV and treated with 10 fold excess proteinase K ()...94

Fig.4.12: Detection of prepore extracted from several Cry protoxins using Method 1A of prepore extraction…………………………………………………………………………….95

Fig.4.13: Detection of prepore reaction for mutations of Cry1A toxins……………………….96

Fig. 4.14: Western blot of prepore reactions formed using modified method 2 of extracting prepore………………………………………………………………………………………….97

Fig.4.15: Coomasie staining of 8% SDS PAGE gel of prepore prepared from Cry1Ab protoxin in the presence of M. sexta brush border membrane using Method 2………………………….98

Fig.4.16: Superdex 200 HR 16/60 gel filtration column purification of the prepore complex..99

Fig.4.17: Sedimentation velocity analyses of Cry1Ab toxin prepore………………………..100

Fig.4.18: Western analyses of prepore obtained by Method 1………………………………101

Fig.4.19: Superdex 200 HR 16/60 retention peak of the prepore complex obtained using

Method 2 of prepore extraction from Cry1Ab…………………………………………………102

Fig.4.20: Regions of the Cry1Ab toxin that were identified based on peptide matched from

LC-MS/MS of digested Cry1Ab prepore samples……………………………………………...103

xvi

CHAPTER 1

LITERATURE REVIEW

1.1 Bacillus thuringiensis & its pesticidal proteins: The Journey:

It was the beginning of the 20th century when the first bacterium was isolated from dead silkworms in Japan. In 1915, the term Bacillus thuringiensis was first coined by a German entomologist Ernst Berliner, when he isolated the bacterium from a dead Mediterranean flour moth in the Free State of Thuringia [1-3]. Long after its isolation, the study and application of the crystal toxin produced by the bacterium did not gain momentum, primarily because of ready availability of synthetic pesticides. Until 1970s, only crystal toxins effective against caterpillars of moths and butterflies were known [4]. The most effective bacterium that produced these toxins was Bacillus thuringiensis var kurstaki (HD-1) active against a wide variety of caterpillars. Since early 1980s, several toxins active against other orders of insects like the dipterans and the coleopterans were discovered. Over the past decade, several thousands of strains of bacteria have been isolated that produced more than 200 different crystal proteins that have activity against more than 150 species of insects, nematodes and more recently against cancerous cells[5, 6]. Awareness of the harmful effects of synthetic pesticides arose in the 1960s,

[7] leading to widespread application of Bt toxins sold under brands of Dipel, Vectobac,

Thuricide, etc. While these sprays were effective against selective pests, their external mode of application on plant surfaces rendered them ineffective against insects that were cryptic. But development of molecular biology techniques provided a solution to this problem by introducing

the gene for the toxin produced by the bacterium into the plants affected by the insects, thereby allowing the insecticidal protein to be manufactured in every cell of the plant. The first successful genetically engineered plant was Bt corn in 1995. While Bt corn and cotton are being commercially produced across the globe, other crops such as soy, rice, tomato etc. are in their developmental phase.[8]

1.2 Biology of Bacillus thuringiensis:

Bacillus thuringiensis is a gram- positive sporulating bacterium that is unique in its capacity to produce intracellular parasporal inclusion bodies during its stationary phase of growth [5].

These inclusion bodies are composed of proteins active against a wide variety of insect larvae, especially from the Orders of Lepidoptera, Diptera and Coleoptera. As these alternative insect control agents are specific to the insects they target, they have been applied for more than 40 years without any harmful effects to other organisms and the environment.

The two major classes of protein toxins that are produced by the bacterium are the crystal

(Cry) and the cytolytic (Cyt) proteins. Upon reaching the stationary phase, these proteins accumulate in the bacterial cell. Approximately 20-30% of the dry weight of the cell in this phase is made up of crystal proteins [9]. Genes responsible for the expression of crystal toxins are usually located on extrachromosomal plasmids. Expression of most cry genes (eg: cry1Aa, cry2A, cry4A etc.) are well regulated in the sporulation phase of growth, as are the sporulation- specific genes in Bt [5, 10]. However, a few genes like cry3A are expressed during the vegetative phase of the bacterium and are shown to have promoters that can be recognized by the primary sigma factor during that phase [10, 11].

Studies have shown that several Cry proteins, when expressed in either E. coli or B. subtilis, produce 130 to 140 kDa protoxin molecules that retain their biological activity. The self- assembly of these 130 kDa proteins is spontaneous and is mediated primarily by the C-terminus of the protein [12]. The cysteine- rich carboxyl terminus that is highly conserved among lepidopteran -specific Cry proteins [13] generates a number of disulfide bonds that allow good crystal packing and also protects the toxin from proteases. Absence of cysteines has rendered the crystal protein more sensitive to the environment. One example is the Cry1Ab from B.t. subsp. kurstaki HD1-9, in which inclusions accumulate only when grown at a temperature below 25oC but not at 30oC even though mRNA for Cry1Ab is present at either temperatures [14]. The difference was attributed to the protease sensitivity of the crystals at 30oC, suggesting different crystal packing at the two temperatures. Smaller forms of protoxin proteins are produced for some toxins as in Cry3A, in which the absence of C-terminus generates a flat crystal and in

Cry2A, which generates a cubical crystal. These shortened forms are interesting as the lack of cysteine- rich C-terminus does affect its folding and crystal formation and there are probably other regulatory mechanisms involved in its folding process.

Solubilization of these crystals occurs only at a high pH of  10.0 as seen in the lepidopteran midgut [15-17]. Protoxin is processed in the midgut of the lepidopteran insects by several proteases like trypsin and chymotrypsin to produce the active form of the toxin [18, 19].

DNA bound to the protoxin core is found to play a role in the proteolytic processing of the C- terminus of the protoxin [20, 21]. The presence of reducing agents such as 1mM DTT or 2% β – mercaptoethanol enhances the formation of activated toxin. These reducing agents enable the disulfide bonds in the well-packed crystal to be reduced so that the trypsin (protease) can gain access to degrade the C-terminal half of the toxin, thereby generating the trypsin- resistant core 3

of the toxin (N-terminal half).This N-terminal half that mediates toxicity is digested by proteases at position 28 on the N-terminal end and position 623 or 637 at the C-terminal end [22, 23].

1.3 Structure of the Crystal Toxin:

Proteases like trypsin cleave the C-terminus of the protoxin, generating a trypsin-resistant core toxin. For most toxins, the core is approximately 600 amino acids long with a molecular weight of 65 kDa. X-ray crystallography has illuminated the 3 dimensional structures of several toxins. The first 3D structure determined was that of the coleopteran specific toxin Cry3A in

1991 [24]. The structures of many Cry toxins: Cry1Aa, Cry2A, Cry4Ba, Cry4Aa and Cry3Bb have also been reported [25-29]. The overall structure has been conserved across all the different toxins even though the homology of the amino acid sequence across these Cry toxins is only from 30- 50% and their target hosts are different. All structures described so far show a three domain conserved structure for the toxin as shown in Fig. 1.1. Domain I is made of seven amphiphilic -helices where the hydrophobic  helix 5 is positioned in the middle surrounded by

6 other amphiphilic helices. In case of  helix 2, the helix is split by a helix- breaking motif with a conserved Proline in position 70 (and glycine) into  helix 2A and 2B. Domain II is made of three anti-parallel beta sheets that form a “Greek Key” topology in a β prism fold with  loop regions. Domain III is a β sandwich with a “Jelly Roll” topology. The exception to this rule was first seen in Cry4Ba structure in which the N-terminus of Domain I lacked the first 2 alpha helices [25]. However, the overall structure of Cry4Ba is closer to Cry1Aa than Cry3A [28]. This led to the postulation of a common mechanism of action in all these toxins.

1.4 Mechanism of Action of the toxin:

Active toxin in the midgut is known to bind several receptors on the surface of the insect

epithelium on the brush border membranes. The initial binding incorporates a reversible

binding event of the toxin to the brush border membranes that has been well characterized by

competition binding studies of toxin to brush border membrane vesicles (BBMV). This

binding is shown to be highly specific based on the toxin and receptor involved in the

interaction. The number of binding sites on a toxin surface and their affinity has been shown

to be specific determinants in this toxin-receptor interaction [30]. The specific binding event

leads to the second event of irreversible association of the toxin with the membrane.

Irreversible binding has been shown to be a more consistent indicator of toxicity than binding

affinity [31-33]. This event is now more specifically addressed as the event of “insertion” of

the toxin into the membrane [34] leading to the formation of ion channels.

In case of Cry1Aa, Cry1Ab and Cry1Ac toxins, binding of the toxin to the receptor has

been well characterized by several studies. Factors that determine the specificity for each

toxin-receptor interaction were characterized. Ge et al. had determined regions between

residues 332-450 as the region specifying the toxicity of Cry1Aa towards Bombyx mori[35].

It was found that alanine substitutions in residues 365- 371 of Cry1Aa had an effect on both

its binding and toxicity to B.mori [36].

Individual receptors for Cry1A toxin binding were also characterized. Characterization of

the first receptor that bound Cry1Ab toxin was done by two independent groups who cloned

a 120 kDa aminopeptidase (APN) that bound Cry1Ac toxin [37], [38]. This was immediately

followed by isolation of a 210 kDa cadherin-like receptor, a membrane glycoprotein called

Bt-R1 by Vadlamudi et al. in 1994 [39]. The binding affinity of Cry1Aa toxin was determined to be close to 2.6 nM for Bt-R1 while it was much less (75 nM) to APN from B. mori as determined by surface plasmon resonance analysis [40]. In the case of another lepidopteran insect Manduca sexta, earlier studies in our laboratory and others [41, 42] have shown a univalent mode of binding of the toxin to Bt-R1 with binding affinities from 0.9- 2.6 nM. The APN isolated from M. sexta displayed multiple binding sites for Cry1Ac but only one site for Cry1Aa and Cry1Ab [43]. Other receptors that were characterized for Cry1A toxins include the BtR-270, a membrane bound glycoprotein from Lymantria dispar that bound to Cry1Aa and Cry1Ab with high affinity but less to Cry1Ac [44] and a glycolipid from M. sexta [45] and a glycosphingolipid [46], however their role as a functional receptor or their kinetics have not been shown.

Exhaustive studies have been carried out to map the regions of Cry1A toxin, especially

Cry1Aa and Cry1Ab that bind to Bt-R1 through mutagenesis and deletion analysis. It was found that two residues in the loop region of Domain II, R368 and R369, lost their binding to M. sexta APN without altering the competition binding to BBMV [47]. Alanine mutagenesis also revealed two residues R281 and R289 in the -8 loop of Domain II that lost binding to purified M. sexta APN and thereby losing toxicity to the insect [48]. In case of

Cry1Aa, receptor- binding studies correlated the binding of Domain II of the toxin to both cadherin and APN receptors with its toxicity to B. mori [40]. Jenkins et al. suggested a

“bivalent sequential” (two stage) binding model of Cry1Ac toxin to aminopeptidase N from

L. dispar in which Domain III residues were found to be important in binding to the first site on APN while Domain II residues were found to be vital in the later step of binding, the

tighter of the two [49]. More recently, a new cadherin -binding epitope around F328 residue in Domain II of Cry1Aa has been identified by mutagenesis [50].

While the reversible component of the binding (defined as the binding of the toxin to the receptor) is the initial major requirement for the protein to act, the irreversible component is emerging as the more important step in the final toxicity of the protein. Studies by

Wolfesberger [31] have shown that the initial binding step is not always correlated with toxicity. When he tested Cry1Ab and Cry1Ac against L. dispar, he found that while the former was 100 fold more toxic than the latter, Cry1Ac bound to the BBMV with higher affinity than Cry1Ab. Thus, he suggested that the process of binding is not a reversible phenomenon. Rather, there is an irreversible step that follows immediately after the binding, which represents the insertion and pore- formation by the toxin. This was shortly followed by work showing that irreversible binding was directly correlated to toxicity [32, 33]. Knowles and Ellar [51] had proposed that binding of the toxin to the receptor leads to insertion into plasma membrane, causing colloid osmotic lysis of the cell where the pore formation resulted in a net influx of ions, cell swelling and eventual lysis.

Upon binding to the receptors, the toxin has also been postulated to undergo conformational changes to form an aggregate or an oligomer [52, 53]. Some studies indicate that the oligomer formed in solution has a preferential binding for one receptor over another

[54]. They also suggest that the toxic form of Cry1Ab is the oligomeric form and not the monomeric form. Other studies have argued that the toxin aggregates within the membrane even when monomeric toxin was introduced to the brush border membranes [53].

The insertion of the toxin into the membrane leads to the formation of ion channels in the

brush border membranes of the columnar epithelial cells. These columnar epithelial cells

lining the midgut are joined together by tight gap junctions [55] and goblet cells [56]. They

are protected from direct exposure to the luminal environment by a lining of chitin and

glycoproteins that make up the peritrophic membrane [57, 58]. Potassium is the major ion

transported in the midgut of the lepidopteran larvae. The K+ pump consists of a primary H+

pump (V-ATPase) and the secondary K+/ H+ antiporter [59]. Most of the transport activity

occurs across the goblet cell apical membrane (GCAM) fraction. In the lepidopteran midgut,

+ transport of K ions account for about 90% of the Isc generated [60]. The measurement of

active ion transport to the lumen can be studied by measuring the short circuit current (Isc).

+ This Isc is generated when the concentration of the K ions on both sides of the tissue is equal

and the transepithelial potential is set to zero mV. The voltage clamp technique used to study

the inhibition of the short circuit current at a constant voltage serves as a measure of the

formation of ion channel by the toxin and an alternative approach to demonstrating toxicity

other than bioassays.

1.5 Models for insertion of insecticidal Cry toxins into insect brush border membranes.

Several models have been proposed for the mechanism by which the toxin inserted into

insect brush border membranes. The first one was the “Penknife” model in 1990 by

Hodgman and Ellar [61] who carried out a prediction of the hydrophobicity of the

amphipathic helices of Domain I based on the three dimensional X-ray structures of Cry3A

toxin. The sequence that was most conserved in Domain I was the region from  helix 5 to 

helix 6. They believed that these two helices were joined at the end of Domain I that was

furthest away from the membrane and therefore flipped out like a penknife opening. The model was also derived from a similar model that was proposed [62] for association of another toxin called colicin with membranes. The Penknife model was represented (Fig.1.2) by Barbara Knowles in her review [63].

The second model and the one that became prevalent was the “Umbrella” model proposed by Li et al. [24]. This model was also based on theoretical comparisons with other toxins that formed ion channels like the diphtheria toxin or colicin. The model proposed that a pair of alpha helices, either  helices 6 and 7 or  helices 4 and 5, whichever pair is closer to the membrane side, inserts like a hairpin. This would cause rearrangement only in Domain

I and allow the rest of the helices in the domain to fall onto the surface of the membrane like the ribs of an umbrella. This model too has been figuratively represented by Knowles in the review [63] ( Fig.1.3).

These initial models were tested by several groups using several experimental approaches. First among them were the reports by Shai [64] who synthesized each peptide in

Domain I separately and used fluorescence spectroscopy to examine their ability to partition into membranes. Based on the observation that peptides of  helix 4 and 5 can aggregate in the membrane and have transmembrane orientation, they predicted that these two helices are lining the ion channel, in support of the Umbrella model. The figure showing the orientation of each alpha helix in a membrane as predicted by them is depicted in Fig 1.4 and the Shai model of toxin insertion predicting a dimerized form of the toxin in the membrane is shown in Fig.1.5.

A refined version of the umbrella model was generated by Schwartz et al. [65]. Based on mutagenesis in  helix 4 residues, he proposed a model in which the Domain I swings away from the rest of the toxin. Following this, the most hydrophobic regions of the toxin, the C- terminus of  helix 4 and N-terminus of  helix 5, unfolds and forms a transmembrane hairpin. The inserted hairpins of four toxin molecules aggregate such that the hydrophilic faces of the four helices of 4 form the lumen of the pore. This tetramer has negatively charged side chains of Asp 129 and 136 lining the pore as represented in Fig. 1.6.

Fig. 1.1: Three dimensional structures of Cry toxins determined by X-ray crystallography

[24-27, 29]. Domain I is depicted in red, Domain II in green and Domain III in yellow for each of the toxins.

Fig. 1.2: The Penknife Model of Insertion of Cry toxins [63]

Fig. 1.3: The Umbrella Model of Insertion of Cry toxins [63]

Fig. 1.4: Orientation of synthetic helices of Domain I in artificial membranes [64]

Fig. 1.5: Model of insertion of toxin as a dimer [66]

Fig.1. 6: Refined Tetramer Umbrella Model [65]

Besides the Umbrella model for mechanism of insertion of the toxin, there are two more recent models emerging in the field. The first one is the “signal transduction” model [67] in which Zhang et al. hypothesize that the binding of a monomeric form of the toxin to one of the receptors: cadherin (BtR-1), is responsible for stimulating the G protein  subunit and thereby adenyl cyclase to induce production of cyclic AMP (Mg+2 dependent process). The cAMP in turn activates Protein Kinase A leading to destabilization of the cytoskeleton and disruption of ion channels in the membrane. This model downplays the role of membrane insertion of the toxin in mediating toxicity, suggesting that the insertion of the toxin is receptor independent. It portrays the oligomerization and membrane insertion event as a secondary side effect in the mechanism of action of the toxin. A schematic of the model as portrayed by Zhang et al. [67] is shown in Fig.

1.7.

The second model, the “serial receptor binding” model [54] of insertion of Cry toxin suggests that the toxin monomer binds Bt-R1 cadherin receptor first with an affinity of 1nM and loses  helix 1 to undergo a conformational change, leading to formation of an insertion- competent oligomer called “prepore”. This prepore form releases itself from the Bt-R1 and binds to an aminopeptidase [68] receptor with an affinity of 0.75 nM, causing the prepore to reach cholesterol- rich lipid rafts and insert into the membrane. This model necessitates the formation of an oligomer for insertion into the membrane and suggests that the tetramer is responsible for cellular toxicity. The model has been schematically represented by Bravo et al. [54]and reproduced in Fig. 1.8.

Fig. 1.7: The Signal Transduction Model [67]

Fig.1.8: The Serial Receptor Binding Model [54]

My work deals with characterization of the membrane- inserted form of the toxin to determine the following: 1) the regions of the toxin important for mediating insertion of the toxin, 2) the regions of the toxin that are inserted into the insect brush border membrane and 3) the size of the toxin in solution and in the membrane- inserted form. A combination of biochemical and biophysical techniques has been used to attain each of these goals and will be described in the successive chapters.

CHAPTER 2

ANALYSES OF THE REGIONS OF CRY1A TOXIN THAT INSERTS INTO INSECT

BRUSH BORDER AND ARITIFICIAL MEMBRANES

2.1 INTRODUCTION

A critical step in understanding the mode of action of insecticidal crystal (Cry) toxins from Bacillus thuringiensis is their partitioning into membranes, and the insertion of the toxin into insect brush border membranes, in particular. The Umbrella and Penknife Models of insertion of the toxin into insect brush border membrane as discussed in Chapter 1 indicate that only 2 alpha helices of Domain I, -4 and -5 or -5 and -6 insert into BBMV[63]. Evidence in support of the model has been increasing since its proposal in 1980s. However there was little analysis on other regions of the toxin for their role in insertion. Domain II and Domain III together constitute 60% of the bulk of the toxin. The understanding of the role these domains play in insertion is vital in generating a holistic understanding of the mode of action, thereby impacting the study of resistance mechanisms in insects and safe application of these toxins as pesticides.

Studies on protease protection of Cry1A toxins in insect membranes were independently carried out by several groups. The first published work in which Cry toxins were protected from a non- specific protease, Proteinase K , showed that a 60 kDa form of the toxin was protected in the membranes, suggesting that only a region equivalent to alpha helix 1 was susceptible to this

protease [69]. In a more recent study, Pronase digestion of Cry1Aa toxin for 24 hours resulted in peptide fragments from all three domains of toxin protected in the membrane [70].

The present study uses site- directed cysteine mutagenesis to examine the 3 domains of the toxin in order to identify specific regions that may be embedded into the insect membrane including the regions of the toxin proposed by the Umbrella model and beyond. Protease protection assay and steady- state fluorescence measurements using cysteine mutations in regions of 3 domains of Cry1Aa or Cry1Ab toxin show that all regions of the toxin studied except α-helix 1 are embedded within the membrane, although to different extents. Cry1Aa and

Cry1Ab were both investigated because they show 89% identity in their sequence [13] and target similar insects.

2.2 EXPERIMENTAL PROCEDURES

Site directed mutagenesis: An E.coli cell culture containing the Bacillus thuringiensis cry1Ab9-

033 [71] gene was obtained from T. Yamamoto (Sandoz Agro Inc., Palo Alto, CA). A similar culture for Cry1Aa was obtained from American Type Culture Collection [72]. Uracil- rich template for Cry1Ab was obtained by transforming these phagemids into E .coli CJ236 ung-dut- strain. Phagemids grown in this cell culture incorporated uracil into the DNA in place of thymine. When these cultures were superinfected with helper phage M13K07, the phage packs one strand of the phagemid into newly synthesized phage coat protein. These single stranded

DNA molecules, rich in uracil, are extracted from the phage coat using organic reagents like phenol, chloroform and isoamyl alcohol. Resulting DNA is used for mutagenesis using the

Kunkel method of mutagenesis [73]. Alternatively, double- stranded DNA obtained from E.coli

DH5 was used for mutagenesis using the QuikChange mutagenesis protocol (Stratagene Inc.).

Primers for site- directed mutagenesis were obtained from either Integrated DNA Technologies

Inc. or Bioneer Inc. Mutations were confirmed by sequencing double stranded DNA at the Plant

Molecular Genomics Facility (PMGF), The Ohio State University, Columbus, Ohio.

Expression and purification of the toxin mutants: Expression and preparation of the toxin was carried out as described earlier [30]. Crystals were solubilized in 50 mM Na2CO3, pH 10.5, to extract the protoxin and were digested with 1/50 w/w of trypsin/crystal protein to yield the activated toxin. The toxins were purified using Sepharose Q ion exchange chromatography,

Sephacryl S300 and Superdex S200 gel filtration chromatography in series.

Preparation of Small Unilamelar Vesicles [74]: Small Unilamelar Vesicles (SUV) were formed using artificial phospholipids, 1-Palmitoyl-2-oleyl-sn-glycerol-3-phophatidylcholine (POPC), 1-

Palmitoyl-2-oleyl-sn-glycerol-3-phosphatidylethanolamine (POPE) and cholesterol (Avanti Polar

Lipids Inc.) in the molar ratio of 7:2:1, similar in composition to lipids commonly in bilayer and vesicle formation of Cry toxins [75-77]. SUV were prepared from a preparation of Large

Multilamellar Vesicles (LMV) by using a Branson 2200 bath sonicator. The LMV mixture free of any chloroform was sonicated in the water bath for 10 min intervals during which the solution turned less opaque. The size class of the SUV was measured on a DynaPro light scattering instrument (Wyatt Technologies.). This procedure yielded an average size of 27.5± 2.5 nm which was reproducible from batch to batch analysis. Typical sizes of SUV vary from 15-50 nm when obtained using this protocol [78].

Preparation of Brush Border Membrane Vesicles (BBMV): Fourth instar larvae of Manduca sexta (Carolina Biologicals Supply Company) were dissected using procedures described elsewhere [79]. BBMV were prepared using the modified differential magnesium precipitation

method , originally used by Wolfersberger, M. et al.[80]. The final BBMV pellet was resuspended in a binding buffer (10 mM HEPES, 150 mM NaCl pH 7.4). Protein concentrations were estimated using Coomassie Protein Assay (Bradford) Reagent (Pierce Biotechnologies,

Inc.)

Proteinase K protection assays: Pure toxin was mixed with 10- fold excess of BBMV and incubated at 25oC for 30 min after which proteinase K at 10- fold excess concentration of the toxin was added and incubated at 37oC for 30 min. Phenylmethylsulfonylfluoride (PMSF) was added to stop the reaction. The mixture was centrifuged at 15,000 x g for 10 min. The pellet was washed with 10 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), 150 mM

NaCl, pH 7.4 and then solubilized into 1% n-octyl β-D-glucopyranoside (Sigma) and boiled for

3-5 min before loading onto an 8% SDS-PAGE gel. Proteins were transferred onto a polyvinylidene difluoride (PVDF) membrane and blotted using polyclonal anti Cry1A rabbit antisera at a dilution of 1 in 10,000 and horseradish peroxidase (HRP)-tagged goat anti-rabbit antisera at a dilution of 1 in 50,000 (BioRad). Blots were visualized using chemiluminescent

HRP substrate (BioRad).

Labeling of purified cysteine mutants with probes: 5-({2-[(iodoacetyl)amino]ethyl}amino)- napthalene-1-sulfonic acid (1,5-IAEDANS) and 6-acryloyl-2-dimethyl-aminonapthalene

(Acrylodan) were purchased from Invitrogen Inc. Purified cysteine mutants were incubated with

10 fold molar excess of the probes and incubated in the dark overnight. Unbound label was removed using Sephadex G50 gel filtration (GE Healthcare). Purity of the protein was checked on 8% SDS-PAGE gel and the efficiency of labeling was measured using the molar extinction coefficient of each probe.

Steady- state fluorescence quenching measurements: 50 µg of 1, 5- IAEDANS- labeled Cry1A toxin was mixed with 5mgs of SUV and incubated for 60 min. The bound toxin was separated from the unbound labeled toxin by passing the sample through a Sephadex G100 column (GE

Healthcare). Concentration of the SUV- bound protein was measured using the Bicinchonic Acid

(BCA) Protein Assay Kit (Pierce Biotechnologies Inc.) after removal of lipids from the proteoliposomes using Compat-Able Protein Assay Preparation Reagent Set (clean up kit)

(Pierce Biotechnologies Inc.). Steady- state fluorescence measurements of equal amounts of free and bound toxin were carried out in a Fluoromax-3 fluorimeter (JY Horiba). The sample was excited at a wavelength of 380 nm and an emission scan showed maximum fluorescence around

460 nm. The SUV- bound, labeled toxin was treated with increasing aliquots of potassium iodide

(KI) in thiosulfate (final concentration of 0.83M) to test if the label bound to the toxin was further susceptible to collisional quenching in the aqueous environment. The percentage of quenching of fluorophore was calculated as the percentage ratio of the difference in the quantum yield before and after partitioning of the labeled protein to the total fluorescence of that free labeled protein in buffer.

Fluorescence “blue shift” measurements: Acrylodan- labeled Cry1A toxin (50 µg) was allowed to bind with BBMV (500 µg) for 60 min. The bound toxin was separated from the unbound by centrifuging the sample at 15,000 x g. Proteinase K (500 µg) was added to the reaction and incubated for additional 30 min. The sample was centrifuged again at 15,000 x g and the resultant pellet was resuspended in binding buffer. Steady- state fluorescence measurements of the labeled toxin, bound toxin before and after proteinase K treatment were carried out on

Fluoromax-3 fluorimeter (JY Horiba). The samples were excited at 360 nm and the fluorophore

emission was read from 390 nm to 650 nm. Maximal emission wavelengths (max) of fluorescence of labeled toxin and toxin bound to BBMV before and after proteinase K treatment were recorded for each mutant. Each experiment was repeated for 3 times and the average value of the max was recorded.

Fluorescence Energy- Transfer(FRET) Measurements: Acrylodan- labeled Cry1A mutants (10

g) were mixed with 2 mg small unilamelar vesicles [74] prepared from a mixture of 1-

Palmitoyl-2-oleyl-sn-glycerol-3-phophatidylcholine (POPC) and 1-Palmitoyl-2-[12-[(7-nitro-2-

1,3-benzoxadiazol-4-yl)amino]dodecanoyl]-sn-Glycero-3-Phosphocholine (16:0-12:0 NBD-PC)

(1:1) in 10 mM HEPES +150 mM NaCl, pH 7.4 and incubated at 25oC for 60 min. Proteinase K

(100 g) was added to the reaction and incubated for further 30 min. The reaction was stopped by adding 1mM PMSF and was immediately centrifuged at 300,000 x g to pellet proteoliposomes. Proteoliposomes were resuspended in 10 mM HEPES +150 mM NaCl, pH 7.4.

Steady- state measurements of labeled protein, liposomes and proteoliposomes were carried out by exciting acrylodan at 360 nm and measuring the emission of 7-nitro-2-1,3-benzoxadiazole

(NBD)group attached to the fatty acid at 500 - 600nm (maximal emission at 550 nm). The energy transfer was calculated as an increase in the acceptor fluorescence. Efficiency of FRET was calculated as: [Ida (mutants) –Ia (wildtype)]/Ia (wildtype) where Ida (mutants) is the acceptor fluorescence in the presence of acrylodan- labeled mutants and Ia (wildtype) is the acceptor fluorescence in the presence of non- labeled wild- type toxin.

Toxicity Bioassays: Toxicity levels were determined by estimating the median lethal concentration (LC50) on M. sexta larvae using the diet surface contamination assay [30]. Sixteen first instar larvae were used for each concentration of the toxin and a total of 6 concentrations of

each toxin were used. Mortalities were recorded after 5 days. The LC50 for each toxin was calculated by probit analysis using SoftTox (WindowsChem Software, Inc.).

Voltage- clamp measurements of Cry1A mutants: Inhibition of short- circuit currents (Isc) was measured by clamping M. sexta midguts using procedures described earlier [79]. Briefly, 100 ng of purified proteins were added to the luminal side of the gut stabilized in the buffer [81].

Measurements were made on a DVC 1000 voltage/current clamp (World Precision Instruments,

Sarasota, FL) connected to MacLab- 4 (AD Instruments, Mountainview, CA). Data analysis was done using SigmaPlot v.10 (Systat Software Inc. Richmond, CA). The data points were normalized to scale the percentage of short- circuit current remaining. The slope of the linear region was used to measure the rate of pore formation in case of each mutant and the lag time

(T0) was also used as an indicator of the rate of partitioning of the toxins into BBMV [79, 82,

83].

1 2 3 4 5 6 7 8 9

Fig.2.1: 8% SDS PAGE gel showing the expression and purification of Cry1Ab toxin in E.coli

DH5 cells. Lane 1 = Cry1Ab protoxin (130 kDa) extracted from inclusion bodies obtained from

DH5 cells. Lane 2 = Trypsinized Cry1Ab toxin (65 kDa). Lane 3 = Biorad prestained molecular weight marker Lane 4 = Cry1Ab toxin purified using Sepharose Q column (0.15M NaCl). Lane

5 -7 = Cry1Ab toxin purified using Sephacryl S300 column (0.5M NaCl). Lane 8 = Cry1Ab toxin purified using Superdex S200 column (0.5M NaCl). Lane 9 = Cry1Ab purified toxin concentrated and stored (0.5M NaCl, 4oC).

2.3 RESULTS

Expression and purification of the mutant toxins: Cysteine- scanning mutagenesis of several residues in the 3 domains of the toxin was performed using the Kunkel method of mutagenesis

[73] where uracil- rich single- stranded templates were annealed to primers and elongated in the presence of T7 DNA polymerase and T4 DNA ligase. The resulting products were transformed into DH5 cells (containing dUTPase) and were screened for mutants. Proteins were expressed in DH5 cells under a „leaky‟ promoter. The resulting protoxins were run on an 8% SDS-PAGE gels to ensure the presence of a 130 kDa band. Expressed proteins were digested with trypsin to yield an active 65 kDa form that was purified using ion exchange and gel filtration chromatography. Expressed and purified products were run on an SDS PAGE gel to confirm their size and purity (Fig.2.1). The secondary structures of the mutants were compared to the wild type using a circular dichroism spectrophotometer. Only mutations with good expression level and whose secondary structure was similar to the wild- type toxin upon expression, as measured by circular dichroism, were used in this study. They are L40C, V171C, S191C, L199C,

L215C, S279C, S324C, S364C, F371C from Cry1Ab and D62C, E460C, K489C and I526C from

Cry1Aa. The mutations used span all the 3 domains of the toxin and most of the chosen residues are conserved across Cry1Aa and Cry1Ab. Purified mutant toxins (10µg) used for the proteinase

K protection assay has been blotted using anti Cry1A antibody shown in Fig. 2.2. All Cry1A toxins produced a doublet band on the SDS PAGE gel upon purification from the crystals.

Sequence analyses of the bands have shown that the doublet was a result of multiple trypsin sites close to each other, at the C-terminus of the active toxin, due to incomplete digestion of all

molecules. The toxin (Fig. 2.2) formed a band of high intensity in the Western, albeit multiple bands of toxin were seen from SDS PAGE (Fig. 2.3).

Toxicity of mutants: Biological activity of each mutant toxin was compared to that of the activity of the wild type toxins using a surface contamination method against first instar M. sexta larvae.

The activity is reported as an LC50 value (concentration of the toxin required to kill 50% of the larvae) shown in Table 2.1. All mutants that we used had an overlapping range of toxicity levels when compared to that of the wild- type Cry1Aa and Cry1Ab toxin, except F371C which compromised membrane partitioning as described in chapter 3 [34].

Proteinase K protection assays: To test if the mutant toxin has retained the ability to insert into

M. sexta BBMV, we digested the toxin bound to BBMV with proteinase K, a non- specific protease. Western blot analyses show that even after treatment with 10- fold excess of proteinase

K for 30 minutes, most of the mutants have retained an approximately 60 kDa form of the toxin

(Fig. 2.4 and Fig. 2.5). For the domain II residue F371 (Fig. 2.5), its mutation to tryptophan protected it from the protease; however its mutation to cysteine did not protect it. The effects of mutating F371 on insertion of the toxin have been published [34] and will be discussed in the next chapter in detail.

Fig. 2.2 A. Western blot analysis of purified Cry1A toxin used for proteinase K protection assays. Lane 1: Cry1Aa. Lane 2: Cry1Ab. Lane 3: Cry1AbL40C. Lane 4: Cry1AaD62C. Lane 5:

Cry1AbV171C. Lane 6: Cry1AbS191C. Lane 7: Cry1AbL199C. Lane 8: Cry1AbL215C.

Fig. 2.2 B. Lane 1: Cry1AbS279C. Lane 2: Cry1AbS324C. Lane 3: Cry1AbS364C. Lane 4:

Cry1AbF371W. Lane 5: Cry1AbF371C. Lane 6: Cry1AaE460C. Lane 7: Cry1AaK489C. Lane

8: Cry1AaI526C.

Fig. 2.3 A. Purified Domain I mutant proteins run on 8% SDS PAGE gel. Lane 1= Protein

Standard; Lane 2 = 1Ab L40C; Lane 3 = 1Aa D62C; Lane 4 = 1Ab V171C; Lane 5 =

1AbS191C; Lane 6 = 1Ab L199C; Lane 7 = 1Ab L215C.

Fig. 2.3 B. Purified Domain II and Domain III mutants run on 8% SDS PAGE gels. Lane 1 & 9

= Protein Standards; Lane 2 = 1Ab S279C; Lane 3 = 1Ab S324C; Lane 4 = 1Ab S364C; Lane 5

= 1Ab F371C; Lane 6 = 1AaE460C; Lane 7 = 1Aa K489C and Lane 8 = 1Aa I526C.

Labeling of Cry1A mutants with fluorophore: Purified cysteine mutants were labeled with acrylodan or IAEDANS and purified off free labels using gel filtration. The labeling efficiency was measured with each fluorophore and was found to be 95 ± 0.3 % for acrylodan and 99 ± 0.5

% for IAEDANS using their respective extinction coefficients. Retention of the secondary structure of the mutants was verified to be similar to the wild type toxin using circular dichroism spectrophotometry.

Quenching analysis of labeled mutants in artificial vesicles: Using artificial SUV made of POPC,

POPE and cholesterol, the quenching of each cysteine mutant labeled to a fluorophore of the aminonapthalene sulfonate (ANS) family, IAEDANS, were measured upon partitioning into the vesicles. The IAEDANS fluorophore has a very high dipole moment in an aqueous environment and hence an increased quantum yield of emission in polar solvents that gets quenched inside the SUV. Thus, upon partitioning of the toxin into SUV, the fluorescence emission of

IAEDANS- labeled mutant is quenched. The percentage of quenching for each of the mutants in our study has been reported in Fig. 2.6. Results show that mutants in Domain I (D62C, V171C,

L199C, L215C) , Domain II (S279C, S324C, S364C) and Domain III (E460C, K489C , I526C) all have about 50% or more of fluorescence quenched inside the SUV, while residues L40C and

S191C have less than 50% quenching. Only F371C displays almost no quenching when the toxin is incorporated into SUVs. Addition of KI solution to a final concentration of 0.83M quenched the fluorescence of labeled toxin in buffer completely at the volumes used, but when mixed with the SUV- bound labeled toxins, there was no further quenching of the already quenched fluorescence of the vesicle- bound label, indicating that the hydrophilic collisional quencher was not able to access the label in the SUV- bound form.

Fig. 2.4. Western blot analysis of 8% SDS-PAGE of proteinase K protection assay of Cry1A

Domain I mutants bound to BBMV. The mutants used are as follows: Lane 1: Cry1Ab. Lane 2.

1AbL40C. Lane 3: 1AaD62C. Lane 4: 1AbV171C. Lane 5: 1AbS191C. Lane 6: 1AbL199C.

Lane 7: 1AbL215C.

Fig. 2.5 Western blot analysis of 8% SDS-PAGE of proteinase K protection assay of Cry1A

Domains II & III mutants bound to BBMV. The mutants used are as follows: Lane 1: Cry1Aa.

Lane 2: 1AbS279C. Lane 3: 1AbS324C. Lane 4: 1AbS364C. Lane 5: 1AbF371W. Lane 6:

1AbF371C. Lane 7: 1AaE460C. Lane 8: 1AaK489C. Lane 9: 1AaI526C.

Toxin/ Mutant LC (ng/cm2) Toxin/Mutant LC (ng/cm2) 50 50 Wt/Domain I Domain II& III Cry1Aa 16.0 S279C 22.0 (8.0-25.3) (8.2-35.6) Cry1Ab 20.0 S324C 19.8 (7.5-31.7) (7.5-32.2) L40C 20.0 S364C 25.6 (7.0-33.4) (11.2-40.2) D62C 12.0 F371C >2000 (4-20.2) V 171C 20.0 F371W 13 (7.5 -31.7) (8 -20) S191C 28.2 E460C 14.2 (12.4- 44.2) (5.6-23.0) L199C 32.4 K489C 12.0 (16.2-48.6) (3.8-21.0) L215C 25.4 I526C 12.6 (11.2-40.4) (3.0-21.4) Table 2.1 Bioassay measurements of Cry 1A toxins and their mutants on first instar larvae of M. sexta using surface contamination method. 16 larvae were measured per concentration. Results were measured after 5 days of exposure to toxin and calculated as LC50 using probit analysis

(Softox).

“Blue shift” measurements of acrylodan labeled mutants: Acrylodan is an environment- sensitive fluorophore that reacts with thiol groups of cysteines to form a covalent conjugate.

Depending on the environment of the thiol group, there is a variation in the fluorescence emission from the molecule. Emission of the probe is low and at longer wavelengths (around 500 nm) in aqueous environments, while the emission occurs at much shorter wavelengths (around

460 nm) in a lipid environment like that of the BBMV. Emission from acrylodan- tagged mutants is dependent on its dipole moment and is therefore different for different mutants, depending on the exposure of the residue to aqueous environment.

All of our acrylodan- tagged mutants showed a maximal emission wavelength in aqueous environment around 480-500 nm. Upon binding to BBMV, the maximal emission wavelength shifted to a shorter wavelength for all of them, the extent of which was different for each mutant

(Fig. 2.7). This “blue shift‟ was retained for all the mutants even after treating the BBMV- bound labeled toxins with 100- fold excess of proteinase K.

Fig. 2.8 A, B and C are representative spectra of mutants in Domain I, II and III respectively indicating the wavelength shift from longer to shorter wavelength under different environments for the label. The extent of blue shift was not the same before and after proteinase

K treatment as indicated in Fig. 2.7. All these mutants displayed an increase in the intensity of acrylodan fluorescence as seen in the representative spectra. The only exceptions to this were

Cry1Ab L40C mutant, which showed decrease in its fluorescence emission (Fig. 2.8 D) and

Cry1Ab F371C mutant (to be discussed in Chapter 3), which did not partition into the membrane.

Energy transfer measurements for mutants: Fluorescence energy transfer between acrylodan donor and nitroxybenzdiazole (NBD) receptor fluorophores were measured for some of the mutants expressed. The increase in receptor fluorescence was measured in the presence and absence of donor fluorescence to calculate the efficiency of FRET response. Since the Förster

o Radius (R0) value for the two probes were in a wide range of 20-40A , accurate distance measurements between the probes were not possible. However the FRET efficiencies were calculated using the formula in methods section and have been plotted for each mutant in Fig.

2.9. Two additional mutations, 1Aa S98C and 1Aa S351C, have been studied here while not all the mutants used for blue shift measurements have been examined yet. Results show that mutant

S324C in Domain II of Cry1Ab showed maximum transfer efficiency while mutants from domains II and III were able to achieve similar levels or better response than Domain I.

Voltage- clamp measurements on Cry1A mutants: To test the pore- forming abilities of each mutant, we carried out voltage clamping of M. sexta midguts and measured the percentage of remaining short circuit current in the midgut after adding 100 ng of each toxin (Fig. 2.10 and Fig.

2.11). Slopes for the linear region of the drop in the Isc were calculated (Table 2.2). The rate of pore formation was measured as the slope of the linear region of the drop in short- circuit current for each of the mutations. We found that the rate of ion transport for mutants in Domain I were overlapping with those of mutants in Domain II and Domain III, indicating that pore formation was occurring at a similar rate. In addition, the time lag (T0) values of 100 ng of all mutants ranged from 4.5 min to 6.5 min for all the mutants, while that of the same amount of Cry1Aa and

Cry1Ab was 5.0 min (Fig. 2.12). This time lag is an indicator of the time the toxin takes, after adding to the membrane, to initiate pore formation. In other words, it is the time of partitioning of the toxin into BBMV [79, 82, 83] suggesting that all our mutants partitioned into the 37

membrane at a similar rate, except F371C, the non- inserting mutant, which will be discussed in

Chapter 3.

Fig. 2.6 Percentage of quenching of fluorescence of 1, 5-IAEDANS tagged cysteine mutants calculated as (Iaq– ISUV)/Iaq where Iaq is the quantum yield of fluorescence of the labeled mutants in aqueous buffer and ISUV is the quantum yield of fluorescence of the labeled mutants in SUV.

510

500

490

480

470

460

max of Acrylodan Emission max of Acrylodan

 450

D62C

I526C

L199C L215C

V171C S191C S279C S324C S364C E460C K489C Mutant Residue Position

Fig. 2.7 Blue shift in the maximal emission wavelength of each acrylodan labeled mutant. The

X- axis indicates the name of each labeled mutant studied and the Y- axis shows the maximal emission wavelength. Maximal emission wavelength of each mutant in aqueous buffer (50mM

Na2CO3 + 0.5M NaCl, pH 10.5) is indicated by (●), in BBMV without any protease treatment is indicated by () and in BBMV after Proteinase K treatment is indicated by (▼).

Fig 2.8 Steady state fluorescence spectra for the following acrylodan labeled mutants:

A. D62C (Domain I); B.S324C (Domain II), C. I526C (Domain III) and D. L40C (Domain I).

(-●-) represents the fluorescence in buffer, (-•) represents the fluorescence in BBMV before proteinase K treatment and (-▼-) represents the fluorescence in BBMV after proteinase K treatment. Y–axis represents the relative intensity of fluorescence of the labeled mutant in buffer versus membranes and not absolute values of intensity.

Fig.2.9 Measurement of efficiency of FRET response calculated as a measure of increase in acceptor (NBD) fluorescence in the presence and absence of acrylodan donor. Excitation of samples occurred at 360 nm while maximal emission intensity was recorded at 550 nm.

Fig. 2.10. Voltage clamp response of Cry1Ab (●) to those of Domain I mutants: L40C (),

D62C (■), S191C (), L199C () and L215C (▼). (V171C has been reported in Chapter 3. The arrow indicates the time at which toxin was added to the midgut.

Fig. 2.11. Voltage clamp response of Cry1Aa (●) to those in Domain II mutants: S279C ()

S324C (▼) S364C () E460C (■) K489C () and I526C (). F371C has been reported in

Chapter 3 and F371W has been reported earlier [84]. The arrow indicates the time at which toxin was added to the midgut.

Table: 2A

SAMPLE T0 (min) Slope (µA/min) Cry1Ab 4.0-5.0 -12.0± 3.2 1AbL40C 4.0-5.0 -9.42± 2.33 1AaD62C 6.0-7.0 -10.72± 5.2 1AbS191C 4.0-5.0 -13.5± 2.92 1AbL199C 5.0-6.0 -15.1± 5.0 1AbL215C 4.0-5.0 -11.45± 0.75

Table: 2B

SAMPLE T0 (min) Slope (µA/min) Cry1Aa 4.0-5.0 -11.0± 1.7 1AbS279C 5.0-6.0 -10.3± 3.0 1AbS324C 4.0-5.0 -9.12± 0.89 1AbS364C 5.0-6.0 -11.0± 1.2 1AaE460C 4.0-5.0 -10.67± 1.67 1AaK489C 5.0-6.0 -9.58± 1.67 1AaI526C 5.0-6.0 -11.66± 3.0

Table 2.2: The rate of pore formation inferred from the slope of the linear region of the decrease in short circuit current remaining (Isc) in M. sexta midgut for Domain I mutants (Table 2A) and Domain II and III mutants (Table 2B).

14 12

(MIN) 10

0 8 6 4

Lag Time T 2 0

L40C

D62C

I526C

L199C L215C

V171C S191C S279C S324C S364C E460C

K489C

Cry1Aa

Cry1Ab Toxin/Mutants

Fig. 2.12 Time required for formation of ion channels in M. sexta midgut / Lag time (T0) determined from voltage clamp measurements.

2.4 DISCUSSION

Studies on the insertion of Cry toxins for the past 2 decades have focused on the mechanism by which the toxin forms ion channels in the brush border membrane vesicles of the insect midgut. Based on the crystal structure of Cry3Aa, Li et al. [24] proposed the Umbrella

Model of insertion of the toxin, which predicts that only -helix 4 and 5 of Domain I of the toxin would insert into the membrane because of the hydrophobic nature of -helix 5. Subsequent studies on the toxin were focused extensively on the helices of Domain I, concluding that only those regions could partition into the membrane [64, 85] or line the pore [65]. These studies did not address the fate of regions of the toxin other than the -helices of Domain I once the toxin is inserted. However, proteinase K protection studies (typically used for detecting the regions of membrane proteins inside lipid bilayer) have shown that a 60 kDa form (or higher molecular weight aggregate) of the toxin is protected in membranes [53, 69, 70, 82]. This suggests that most of the toxin is likely to be embedded in the membrane. We mutated several residues across the 3 domains of the toxin to cysteine to determine if these residues (and thereby the region of the toxin around them) are embedded into the membrane. Using fluorescence quenching and/or blue shift measurements, our results indicate that regions in all 3 domains of the toxin partition into the membrane.

This study used 6 mutations that span Domain I, 4 mutations that span Domain II and 3 mutations that span Domain III. Toxicity data showed that none of these mutations compromised the biological activity of the toxin (Table 2.1) and voltage- clamp analysis (Table 2.2) further indicates that all the mutations formed ion channels at a similar rate to wild- type toxin.

Quenching data for IAEDANS-labeled mutants show that most of the labeled residues in all 3 domains of the toxin quenched their fluorescence upon partitioning into SUVs. Certain residues showed more quenching indicating that the label attached to those residues were in a relatively more hydrophobic environment as compared to those that showed less quenching. That the quenching was due to the label being in SUV was confirmed by the lack of any further quenching by the hydrophilic quencher KI added to the SUV-bound, labeled toxin. An alternative possibility for the quenching of IAEDANS-labeled toxin is that the quenching could be due to the hydrophobic environment generated by the toxin molecules themselves, upon self- aggregation or self- oligomerization. We have examined several IAEDANS-labeled toxins using in vitro self- aggregation in buffers lacking ionic strength. None of these labeled toxins showed significant quenching upon aggregation (data not shown). A second possibility for the quenching of IAEDANS-labeled toxin in SUV is the possibility that SUVs do not mimic the natural BBMV environment where receptors play a role in determining the regions of the toxin that would be buried in the membrane. However, when IAEDANS-labeled toxin was used in natural BBMV, all labeled positions were quenched in the 80-90% range (data not shown). The result that the label attached to residues in all 3 domains was quenched in SUV suggests that more than two helices of Domain I of the toxin are bound to the vesicles in a hydrophobic environment.

To verify the above observations with IAEDANS, we choose a chromophore, acrylodan that undergoes enhanced fluorescence and a spectral “blue shift” upon entering a hydrophobic environment. Each cysteine mutant used in the study except L40C (on -helix 1) and F371C showed a “blue shift” in its fluorescence upon binding to BBMV. We observed that not only regions of Domain I insert into the brush border membrane, as predicted by the models in question, but that regions of Domain II and Domain III used in the study also inserted into the

membrane. In these experiments, we were able to circumvent any blue shift that might have occurred due to self- aggregation or oligomerization of the toxin outside the membrane by incorporating an additional step by measuring the fluorescence of toxin- treated vesicles before and after treating them with proteinase K for each mutant. This treatment enabled us to confirm that the region of the toxin to which the label was bound was inserted into the membrane bilayer and was not in a hydrophobic environment outside the membrane. The extent of “blue shift” seen with each of the mutants before and after the proteinase K treatment was different. The “red” end of the spectrum (the environment that the labeled toxin is exposed to in the buffer) varies for each mutant, depending on the polarity of the environment for that residue of the toxin. The extent of “blue shift” that each toxin undergoes indicates the change in the polarity of the environment that the toxin is exposed to upon insertion into the membrane. Domain I residues

D62C and V171C had the greatest shift while residues in Domain II (S324C) and Domain III

(I526C) also underwent a significant blue shift in their fluorescence even after proteinase K treatment of the vesicles. Blue shifts of fluorescence in all other proteinase K- treated mutants indicate that regions of the toxin around those residues were also buried in the membrane. The quantum yield of fluorescence in all these mutants was also increased when the toxin was in

BBMV compared to when they were in buffer. In addition, FRET response measurements of certain mutants (Fig. 2.9) show that mutations from each domain generate energy transfer with similar efficiencies, suggesting a similar environment for each of these regions when inserted into the membrane. Proteinase K treatment of labeled residue L40C showed a complete loss in fluorescence upon binding to BBMV, indicating that the region of α helix 1 is not present in the membrane. SDS-PAGE of the proteinase K- protected mutant L40C showed an intact 60 kDa form of the toxin, indicating that only the region around that residue (α-helix 1) was vulnerable

to the protease and the rest of the toxin in this mutant was protected intact inside the membrane.

Voltage clamping of the mutant also generated a similar rate of formation of ion channels as the wild type (Fig. 2.10).

Our fluorescence partitioning data is complemented by electrophysiological analysis of all the mutants using voltage clamping of M. sexta midguts. Measurements of the rate of partitioning (T0) and the rate of ion channel formation (µA/min) for each mutant from Domains

I, II or III showed that all mutants were able to partition and form ion channels at a similar rate.

The data suggest that the entry of the toxin into brush border membranes may be more likely at the same rate for each domain, i.e., the entire toxin molecule rather than isolated regions, may partition into the membrane as a unit.

Our observations do not support the Umbrella Model of insertion of Cry1A toxin into brush border membrane vesicles, since we show that residues from Domains I, II and III insert into the membrane. In summary, this work supports an alternative model of insertion [34, 83, 86] that proposes almost the entire toxin of about 60 kDa inserts into the insect brush border membrane to mediate toxicity.

CHAPTER 3

IDENTIFICATION OF MUTATIONS IN DOMAIN II OF CRY1Ab THAT

MEDIATE INSERTION OF TOXIN INTO INSECT BRUSH BORDER MEMBRANES

3.1 INTRODUCTION

The study of binding of insecticidal crystal (Cry) toxins to insect brush border membranes has elucidated a number of insect specificity- determining sites on the toxin [30, 87].

Insect specificity was correlated to the regions between amino acids 280- 550 by several studies, at least for the Cry1A series of toxins [35, 88]. However, several studies have indicated that the initial binding of the toxin to the brush border membranes was not the most critical determinant of toxicity in Cry1A toxins. Garzynski et al. reported that even though BBMV of Spodoptera frugiperda bound Cry1Ac toxin with high affinity, Cry1Ac was non- toxic to the insects [89].

Wolfersberger demonstrated, using competition- binding experiments, that Cry1Ab and Cry1Ac bound tightly to Lymantria dispar BBMV even though the two toxins were non- toxic to the insect [31]. These studies clearly demonstrate additional steps after initial binding of the toxin to receptors on insect brush border membrane that would be determinants of toxicity. Ihara et al. have categorized the binding of Cry1Aa toxin to B. mori BBMV into two components, a reversible and a irreversible component[32]. They conclude that the irreversible binding of

Cry1Aa toxin was much higher to BBMV of these insects as compared to Cry1Ab toxin, rendering Cry1Aa more toxic to the insect. Studies carried out by Rajamohan et al. measuring the dissociation rate of Cry1Ab mutants bound to BBMV from M. sexta (constructed in the

hypervariable loop region of Domain II of Cry1Ab) indicated that 75- 80% of the wild type and toxic mutants were not dissociated from the membrane while only 45 -60% of the non- toxic mutants were irreversibly associated with the BBMV[90]. Single amino acid substitutions of residue F371 to tryptophan, tyrosine, leucine, valine, serine, cysteine and alanine using site- directed mutagenesis have shown that while these substitutions do not affect competition binding of the toxin to M. sexta BBMV, a significant loss in their toxicity was observed in correlation to the hydrophobicity of the substituted residues [84]. The authors concluded that this could have been due to the hydrophobic residue playing a role in tighter binding of the toxin to BBMV or due to the regions of domain II, including loop 2, being inserted into the apical membrane of the gut.

The following study is aimed at proving that the mutations were affecting the irreversible binding of the toxin or the insertion of the toxin into insect BBMV. This study also aims to define the irreversible component as the membrane- inserted form or membrane- inserted species of the toxin and thereby demonstrate that Domain II is critical in the insertion of the toxin into insect BBMV.

3.2 EXPERIMENTAL PROCEDURES

Site-directed mutagenesis. The E.coli cell culture containing the B. thuringiensis -endotoxin gene for Cry1Ab (cry1Ab9033) was obtained from T. Yamamoto (Sandoz Agro Inc., Palo Alto,

CA). Uracil-containing template of Cry1Ab was obtained as described[73]. Primers for site- directed cysteine and alanine mutagenesis were obtained from Integrated DNA Technologies,

Inc. Site-directed mutagenesis was carried out using Kunkel mutagenesis [73] as described in section 2.3. The resulting mutations were confirmed using DNA sequencing performed at the

Plant Microbe Genomics Facility (PMGF), The Ohio State University, Columbus, Ohio.

Expression and protein purification. Expression and purification of the Cry1Ab wild-type and mutant toxins were carried out as described elsewhere [30]. Proteins, in protoxin form, were extracted from inclusion bodies by dissolving the crystals in 50 mM sodium carbonate, pH 10.5 at 37C with shaking for 2 hours. The presence of the protein was assessed in SDS -PAGE gels

(8%). The total concentration of the proteins was estimated using Coomassie Protein Assay reagent (Pierce Biotechnologies, Inc.). Activated toxins were obtained by digesting the protoxin in 1/100 (trypsin/protoxin) at 37C for 30 min. The toxin was purified by anion exchange using a Sepharose Q column (GE Healthcare) followed by size- exclusion chromatography using

Sephacryl S300 and Superdex 200 columns (GE Healthcare) in series. The eluting solvent was

50 mM carbonate buffer, pH 10.5 at a flow rate of 1.0 mL/min and 2.0 mL/min for the 2 gel filtration columns.

Preparation of small unilamellar vesicles. 1-Palmitoyl-2-oleyl-sn-glycerol-3-phophatidylcholine,

1- Palmitoyl-2-oleyl-sn-glycerol-3-phosphatidylethanolamine and cholesterol (Avanti Polar lipids Inc.) were used in the ratio of 7:2:1 dissolved in chloroform and dried in a stream of nitrogen. These phospholipids were the predominant ones from studies that profiled the lipid environment in the insect midgut [91] and have been used in several earlier studies on insertion of Cry toxins into lipid bilayers or artificial vesicles to mimic the insect gut environment[75-77,

92]. After complete removal of organic solvent under vacuum, the lipids were then slowly hydrated with 10mM HEPES, 150mM NaCl (pH 7.4) to form multilamellar vesilces (MLV) based on the physical properties of the fatty acids used. The resulting MLV were sonicated in a

Branson Sonifier water bath for 30 min to form small unilamellar vesicles (SUVs). Vesicles were subjected to light scattering to ensure a uniform size class among each batch of preparation [78].

Preparation of BBMV. M. sexta eggs (Carolina Biological Supplies, Inc.) were hatched and larvae were raised to the fifth instar on artificial diet (Bio Serv, Inc.). Dissection of the insect gut is as described elsewhere[79]. BBMV were prepared by a modified differential magnesium- precipitation method [80]. The final BBMV pellet was resuspended in binding buffer (10mM

HEPES, 150mM NaCl, pH 7.4). Protein concentration was estimated using Coomassie Protein

Assay reagent (Pierce Biotechnologies, Inc.)

Proteinase K protection assays. Pure toxin (10g) in 50 mM Na2CO3 (pH 10.5) was mixed with

100g of BBMV and incubated at 25oC for 30 min. After incubation, 100g of proteinase K

(Recombinant Grade; Roche Biochemicals) was added to the mixture and incubated at 37o C for

30 min. At the end of incubation with proteinase K, 1 mM PMSF was added to stop any further reaction. The mixture was centrifuged at 15,000 g for 10min. The pellet was washed in 10 mM

HEPES, 150 mM NaCl (pH 7.4), was treated with 1% n-octyl--D-glucopyranoside (Sigma) to dissolve the pellet and was boiled for 3-5 mins before loading on SDS-PAGE gels. Proteins were transferred from the gel onto a PVDF membrane and Western blotting was performed using rabbit polyclonal anti-1A antibody (1:5000) and goat anti-rabbit HRP tagged antibody (BioRad).

Blots were visualized using Immune –HRP substrate (BioRad).

Labeling of purified cysteine mutant toxins. Purified cysteine mutants were mixed with 10-fold molar excess of 6-acryloyl-2-dimethylaminonapthalene (Acrylodan) (Invitrogen Inc.) and incubated in the dark overnight. The labeled protein was purified off the free label using a desalting Sephadex G25 column (GE Healthcare). Purity of the labeled protein was checked on an 8% SDS-PAGE gel and the degree of labeling was estimated using the molar extinction coefficient of acrylodan. Labeled protein (50 g) was mixed with 500g BBMV or 5 mgs of

SUV and incubated for 60 min. Bound forms of the labeled protein were separated from the unbound labeled proteins by either centrifuging the BBMV pellet down at 21000 x g or by passing the SUVs through a Sephadex G100 column (GE Healthcare). The BBMV or SUV was treated with 500 g of proteinase K and incubated for 30 min at 37o C. 1 mM PMSF was added to stop the reaction. The reaction was then spun down at 21000 x g for 10 min to recover the

BBMV or passed through Sephadex G 100 column (GE Healthcare) to recover the SUV.

Fluorescence measurements. Steady- state fluorescence measurements were carried out on

Fluoromax 3 fluorimeter (JY Horiba Instruments). The labeled proteins were excited at 360 nm and emission intensity was measured from 390 nm to 600 nm. The labeled protein in solution, bound to BBMV and proteinase K- treated, were measured simultaneously to avoid any instrumentation error. Spectra were corrected for background from buffer and /or the vesicles.

Each experiment was performed three times. Emission spectra were plotted using relative fluorescence of the mutants in buffer to that in the membrane before and after proteinase K treatment.

Surface plasmon resonance analysis. Surface plasmon resonance experiments using a BIAcore

3000 instruments were performed for kinetic analysis. BIAcore‟s carboxymethylated dextran matrix (CM5) sensor chip was used. The analysis temperature was set to 25o C. Around 15000

RU of anti-maltose binding protein (MBP) IgG was immobilized on the surface of flow cell 2 on the CM5 sensor chip using an EDC/NHS-mediated amine coupling procedure [93-95]. A freshly prepared solution of 50 mM NHS (N-hydroxysuccinimide) and 0.2M EDC [1-ethyl-3-(3- dimethylaminopropyl) carbodiimide] was injected for 7 min to activate the flow cell. Anti-MBP

IgG was reconstituted in 10 mM NaCOOH, pH 5.0 and injected at a flow rate of 10 l/min.

Excess activated ester groups on the surface were deactivated using a 7 minute injection of 1 M ethanolamine-HCl, pH 8.5. Flow cell 1 was activated with 50 mM NHS and 0.2 M EDC and blocked with 1M ethanolamine-HCl, pH 8.5 without immobilization of anti-MBP IgG, serving as the reference surface.

A fusion of maltose binding protein to the receptor cadherin protein ( MBP-CAD-D) was constructed as a subclone of the M. sexta cadherin (CAD) gene provided to us by the Adang lab in pMECA vector [41] including CAD regions 11 and 12. MBP-CAD-D at 50 g/ml was injected at a flow rate of 5l/min to interact with IgG on the chip. Over a 3 min period of time, approximately 100 RU of MBP-receptor was captured. The (anti-MBP IgG)-(MBP-CAD-D) surface was allowed to stabilize for 1 min, before toxin at various concentrations was injected.

Buffer only was included as a blank. The flow rate for toxin injections was at 30 l/min. The association phase was 3 min and the dissociation phase was 10 min. Regeneration was achieved by two 30-second injections of 10 mM glycine, pH 1.8 at 100 l/min. The control flow cell that was activated and blocked without immobilization of the antibody had both MBP-receptor fusion protein and toxin flowing through each cycle. Running buffer in all experiments was HBS-P buffer (0.01 M HEPES, pH 7.4, 0.15 M NaCl, 0.005% surfactant). MBP-CAD-D fusion protein was purified in the same buffer while toxin was prepared in sodium carbonate buffer and filter- dialyzed into the HBS-P buffer.

Toxicity Bioassays. Toxicity levels were determined on first- instar M. sexta larvae. The median lethal concentration (LC50) was estimated by surface- contamination assays where toxin was overlaid on the surface of the diet and the larvae were allowed to feed on the contaminated diet.

First- instar larvae were confined in 24-well sterile dishes (Falcon) containing solidified artificial

diet surface-contaminated with activated toxin. Five to six toxin concentrations were prepared for each assay. Twenty M. sexta larvae were used for each concentration. Mortalities were scored after 5 days. The LC50 for each toxin was calculated by Probit analysis using SoftTox

(WindowChem Software, Inc.).

Secondary structural studies by circular dichroism (CD). Purified proteins (labeled or unlabeled with acrylodan) were used for CD analysis at 2 µM concentration in 50mM Na2CO3 buffer, pH

10.5. Protein samples (500l) were used in a 1 cm pathlength quartz cuvette (Starna Cells Inc).

The spectra were collected in an AVIV CD2 spectropolarimeter at 25oC. Ellipticity was measured as a function of wavelength from 300 to 200 nm in 1 nm increments and plotted using

SigmaPlot 2000 (Jandel Scientific Co.). Each CD spectrum shown is the average of ten scans.

Voltage clamp analysis. Voltage clamp analysis was performed as described earlier [79]. After stabilization of the midguts in the buffer [81], 100 ng of each toxin was added into the luminal side of the chamber. The volume of the luminal chamber is 3.75 ml. The inhibition of short circuit current (Isc) was measured with a DVC-1000 voltage/current clamp (World Precision

Instruments, Sarasota, FL) connected to a MacLab-4 (AD Instruments, Mountain View, CA).

Data analysis was performed with SigmaPlot 2000 (Jandel Scientific Co.). Recorded data was normalized to the percentage of Isc remaining. Each experiment was repeated at least 3 times.

3.3 RESULTS

Production of stable toxins: The desired cysteine and alanine mutations were obtained as described in methods and were verified using DNA sequencing. The wild- type and mutant proteins were expressed in E. coli DH5 as 130 kDa protoxin molecules. All proteins used in

this study were digested with trypsin to yield 65 kDa toxin molecules. The protoxin and toxin molecule sizes were verified on an 8% SDS–PAGE gel. The trypsin- activated mutant toxins displayed the same stability as the wild- type activated toxin. All activated toxins were purified through an ion exchange and 2 gel filtration columns to remove all smaller fragments and higher order molecules generated in the process of solubilization and/or activation. All toxins showed a

65 kDa band on 8% SDS-PAGE gels post purification. The secondary structures of the purified mutants were compared to that of the wild type Cry1Ab using CD spectrometry.

Toxicity Bioassays: The biological activity of each toxin was measured using a surface contamination method against M. sexta larvae and the results reported as LC50 (concentration required to kill 50% of the larva tested) as shown in Table 3.1. The toxin with a mutation in

Domain I (V171C) retained toxicity at similar levels to wild-type but a mutation in Domain II residue F371 to either cysteines or alanine made the protein non-toxic to the larvae.

Surface plasmon resonance analysis: The binding of the mutant toxins to the toxin- binding cadherin repeats 11 and 12 in the BT-R1 sequence was compared to that of Cry1Ab wild- type toxin. Table 3.2 shows that compared to the wild type (KD ~ 18 nM), the 1AbF371C (KD ~

48nM) and the double mutant 1AbF371A/V171C (KD ~ 66 nM) have only a minor decrease in binding.

Proteinase K protection assays: Protease protection assays were carried out on toxin- bound

BBMV to determine if each toxin inserted into the BBMV. Western blot analysis showed that while the wild- type toxin and the mutant 1AbV171C were protected from 10- fold excess of proteinase K after 30 min of incubation at 37o C, seen as a 60 kDa band on the gel, the domain II mutants 1AbF371C, 1Ab G374C and the double mutant 1AbV171C/F371A were completely

digested by the non- specific protease in the same conditions (Fig. 3.1). The results suggest that the change in residue F371 to cysteines or alanine and G374 to cysteine has an effect on the insertion process.

Circular dichroism analysis: The cysteine mutants in the activated form were labeled with acrylodan and the secondary structures of the labeled mutants were checked using circular dichroism spectrophotometry. CD spectra of the cysteine mutants showed that the labeling had not affected the global structure due to any of the mutations. The CD spectrum of 1AbV171C differs from the wild-type but the labeled and unlabeled spectra are similar (Fig.3.2).

Fluorescence measurements: The fluorescence emission of acrylodan is highly sensitive to the environment of the fluorophore, with the fluorescence maxima in a hydrophilic environment of >

480 nm and < 460 nm in a hydrophobic environment [96], [97]. Our studies with the label show that while the excited- state emission of the acrylodan bound to the protein is very low in solution, there is a dramatic blue shift in fluorescence emission when the labeled protein is in either artificial SUVs or in BBMVs. Because the dipole moment of the label is highly sensitive to the environment, the emission of each labeled mutant is different in the free form itself, depending on the location of the mutation [96]. While a blue shift is the predominant indication of the change in the environment, the intensity of the emission is also a representation of the environment of the label [97].

The acrylodan-labeled, toxic, mutant protein 1AbV171C shows a blue shift in maximal emission wavelength from 500  10 nm in free solution to 462   nm in BBMV and to about the same value (462  11 nm) in BBMV treated with proteinase K (Fig. 3.3A). That the shift was not from hydrophobic effects of the receptors outside the vesicles was ensured by the 58

emission spectra from the proteinase K- treated BBMV or SUVs bearing the labeled toxin, since receptors are not present in the case of SUVs, and are removed from BBMV by proteinase K.

The intensity of emission of the spectra both before and after proteinase K treatment was similar for this mutant, suggesting that the particular region of the toxin was embedded into the vesicles.

The acrylodan-labeled domain II mutant, 1AbF371C mixed with vesicles (both BBMVs and

SUVs) showed blue shift in the spectra, but upon proteinase K treatment, showed a significant drop in fluorescence intensity (Fig. 3.3C and 3.3F). However for the double mutant (Fig. 3.3B), where the protein was labeled with acrylodan in domain I (position 171), the protein was able to insert into the SUVs and was also protected from proteinase K in these artificial vesicles, as seen by protection of the label in the protease- treated vesicles. However, upon proteinase K treatment of the toxin bound to BBMV, there was a loss of fluorescence intensity. Cry1AbG374C showed a blue shift upon binding to BBMV, but lost all fluorescence when the bound toxin was treated with proteinase K, similar to F371C (Fig. 3.4). In the case of 1Ab G374C, we did not perform the blue shift assays with SUV.

Voltage- clamping analysis: To confirm the lack of pore formation of the non –inserting mutants, we carried out voltage- clamping of M sexta guts and measured the percentage of remaining short circuit current in the gut upon addition of Cry1Ab and its mutant proteins used in this study. The results as shown in Fig. 3.5 demonstrate that while the domain I mutant V171C can form ion channels better than Cry1Ab wild type, the domain II mutant F371C and the double mutant proteins have completely lost their ability to form pores in the midgut membrane.

Rajamohan et al. have also shown the same for domain II mutant G374C, which lost pore- forming abilities [90].

2) Sample LC50 M. sexta (ng/cm

Cry1Ab 20.0 [7.5-31.7]

1AbV171C 40.3 [26.6-53.4]

1AbF371C >2000

1AbV171C/F371A >2000

1AbG374C >2000

Acrylodan Labeled 1Ab V171C 36.6 [24.6-48.2]

Acrylodan Labeled 1Ab F371C >2000

Acrylodan Labeled 1Ab V171C/F371A >2000

Table 3.1: Bioassay measurements of Cry1Ab and its mutants on 1st instar larvae of M. sexta

(tobacco hornworms) using surface contamination method. Eight larvae were used per concentration of toxin. The results were measured after 5 days of incubation and calculated as

LC50 using Probit analysis (Softox).

2 Sample ka (1/Ms) kd (1/s) KD (M) 

1Abwt: exp1 6.51E+04 1.17E-03 1.79E-08 1.08

exp 2 6.42E+04 1.21E—03 1.89E-08 0.859

1AbV171C: exp1 7.64E+04 1.84E-03 2.41E-08 6.02

exp 2 7.04E+04 1.55E-03 2.21E-08 1.1

1AbF371C: exp1 3.66E+04 1.79E-03 4.89E-08 5.38

exp 2 3.83E+04 1.83E-03 4.78E-08 6.32

1AbF371A/V171C: exp 1 1.83E+04 1.25E-03 6.80E-08 6.58

exp 2 1.83E+04 1.17E-03 6.41E-08 2.15

Table 3.2: Binding measurements of the toxin 1Abwt and its mutants to cadherin repeats 11 and

12 using surface plasmon resonance analysis (Biacore). The results are expressed as KD values

(ratio of the kd:ka) and are expressed in Molar units. Sets of measurements for each sample were performed twice, as shown. SPR measurements for G374C were not performed. “exp 1” and

“exp 2” in column 1 indicate two independent sets of experiments for each mutant.

1 2 3 4 5 6 7 8 9 10

Fig. 3.1. Proteinase K protection assay of Cry1Ab wt and its mutants. Reaction was run on 4-

20% SDS-PAGE gels and the membranes were blotted using anti-1A polyclonal antibody and

HRP tagged anti rabbit secondary antibody. Lane1: Pure Cry1Abwt (10g). Lane 2: proteinase K treated BBMV bound to Cry1Ab wt. Lane 3: Pure 1AbV171C (10g). Lane 4: proteinase K treated BBMV bound to 1AbV171C. Lane 5: Pure 1AbF371C (10g). Lane 6: proteinase K treated BBMV bound to 1AbF371C. Lane 7: Pure 1AbF371A/V171C (10g). Lane 8. proteinase

K treated BBMV bound to 1AbF371A/V171C. Lane 9. 1AbG374C (10g). Lane 10. proteinase

K treated BBMV bound to 1AbG374C.

Fig 3.2 A. Circular dichroism spectra of Cry1Ab V171C before () and after () labeling with acrylodan expressed in mol.deg-1.cm-1units.

Fig 3.2 B.Circular dichroism spectra of Cry1Ab F371C before () and after () labeling with acrylodan expressed in mol.deg-1.cm-1units.

Fig 3.2 C Circular dichroism spectra of Cry1Ab 1Ab V171C/F371A before () and after () labeling with acrylodan expressed in mol.deg-1.cm-1units

20 D

-10

-20

-30 180 200 220 240 260 280 300 320 Wavelength (nm)

Fig 3.2 D Circular dichroism spectra of Cry1AbG374C before () and after (▼) labeling with acrylodan expressed in mol.deg-1cm-1 units. Spectrum of 1Abwt is also shown (●).

Figure 3.3: Steady state fluorescence spectra of Cry1Ab mutants labeled with acrylodan. The relative fluorescence (Y-axis) is expressed in arbitrary units. Correction of the spectra was made against either a buffer blank for the free protein in solution or against SUV or BBMV for the protein bound to the respective vesicles. (______) represents the spectra of the purified labeled protein in solution. (………….) represents the spectra of the pure labeled protein bound to SUV or

BBMV before proteinase K treatment. (------) represents the spectra of labeled protein bound to BBMV or SUV after proteinase K treatment. Figure 3A: Cry1AbV171C treated with

BBMV.Figure 3B: Cry1AbF371A/V171C treated with BBMV. Figure 3C: Cry1AbF371C treated with BBMV. Figure 3D: Cry1AbV171C treated with SUV. Figure 3E:

Cry1AbF371A/V171C treated with SUV. Figure 3F: Cry1AbF371C treated with SUV.

-1

-2

-3

Relative fluorescence intensity (A.U) intensity fluorescence Relative -4 400 450 500 550 600 650 Wavelength (nm)

Fig. 3.4 Steady state fluorescence spectra of Cry1Ab G374C labeled with acrylodan. The relative fluorescence (Y-axis) is expressed in arbitrary units. Correction of the spectra was made against either a buffer blank for the free protein in solution or against BBMV for the protein bound to the vesicles. (●) represents the spectra of the purified labeled protein in solution. () represents the spectra of the pure labeled protein bound to SUV before proteinase K treatment. (▼) represents the spectra of labeled protein bound to BBMV after proteinase K treatment.

Fig.3.5 Voltage clamp response of Cry1Abwt () compared to that of BBMV inserting mutant

1AbV171C () and non-BBMV inserting mutants 1AbF371C (▼) and 1AbF371A/V171C ()

3.4 DISCUSSION

The process of insertion of Cry toxins has been studied actively since 1994. Most studies have limited their focus to the fate of domain I of the toxin in the inserted state in membranes, despite work that has shown protection of the whole toxin inside the membrane [53, 69, 70, 82].

This study examines mutants in domain II of the Cry1Ab toxin, at position F371 that allow receptor binding but prevent insertion of the toxin into the membrane. Proteinase K protection combined with steady state fluorescence measurements of labeled toxin molecules effectively demonstrates that the residue F371 plays a role in the mechanism of insertion, at least in Cry1Ab.

Rajamohan et al. [84] reported that mutating residue F371 to a number of amino acids did not affect its competition binding to M. sexta BBMV, but affected toxicity and “irreversible binding” in a manner inversely related to the hydrophobicity of the replacement amino acid. At the time of that study, it was unclear whether “irreversible binding” was due to tightness of binding to the receptor or proficiency of insertion into the membrane. Our binding studies of the mutant protein 1Ab-F371C and the double mutant protein 1Ab-F371A/V171C to cadherin receptor (Dorsch TBR sequence of repeats 11 and 12 [98]) using surface plasmon resonance support the view that these mutants do not suffer significant loss in binding to M. sexta cadherin when compared to the wild- type protein. However, the data from toxicity bioassays and the voltage- clamping studies reflect the inability of the mutant forms to retain their toxicity or ion- channel activity. This indicates a role for the residue F371 in associations of the toxin with the

BBMV in membrane insertion. Steady- state fluorescence measurements of the bound and protected toxin show a blue shift in the maximal wavelength of the domain I mutant protein, suggesting the displacement of the label to a hydrophobic (membrane) environment [96]. The hydrophobicity could hypothetically be due to binding of toxin to receptor (assuming the

receptor is sufficiently hydrophobic), however, because binding is not greatly affected, this seems unlikely. In addition, one would expect that proteinase K treatment would be able to access and digest the region of the receptor exposed outside the membrane and thereby also digest the toxin, given the incubation conditions of 30min. Furthermore, when V171C, a domain

I residue that is not believed to make contact with the receptor, is labeled with acrylodan, we observe that the label migrates to a more hydrophobic environment in BBMV (Figure 3.3A).

Our data from the fluorescence assays for the domain I mutant protein 1AbV171C and other mutants dispersed across the toxin confirm that the label is in an environment not accessible by Proteinase K, suggesting that it is embedded into the bilayer. However, when the label is attached to the cysteine on position 371 or 374, the toxin was unable to enter the vesicles, leaving the protein exposed to proteinase K even after binding to the receptors on BBMV. This is also the case for the double mutant where the fluorescent label is on domain I residue but also incorporates a mutation in position 371 in domain II. Non receptor-mediated residual partitioning of the regions of domain I (around position 171) into the vesicles occurs in BBMV, explaining the residual fluorescence associated with the proteinase K- treated BBMV, in the case of double mutant 1Ab-F371A/V171C. SDS- PAGE gels show a 60 kDa protected form of toxin for the wild- type and 1AbV171C, which is also seen with several other mutations in 1Ab toxin. But an absence of the protected 60 kDa form for the F371C, G374C and the double mutant on the gels corroborates the pattern of fluorescence data.

An interesting observation in these studies is that partitioning of Cry1Ab into BBMVs is different than its partitioning into SUVs. Labeling the toxin on -helix 5 of domain I (residue

171) with acrylodan showed a blue shift in both BBMV and SUV when domain II residue 371 is

wild-type (Phe). When residue 371 is mutated to alanine, the acrylodan-labeled domain I undergoes a blue shift in SUV and is protected from proteinase K, but not in BBMV. This indicates that domain I, at least residue 171 or -helix 5, is able to enter the artificial SUV membrane but not the BBMV membrane, even if residue 371 is alanine or cysteine. The choice of the lipids in our study was similar to phospholipids (and ratios) used in earlier studies on Cry toxin or synthetic peptide- insertion into lipid bilayers or vesicles [65, 66, 99, 100], in support of the Umbrella Model. As early as 1988, Yunovitz and Yawitz [101] showed that phosphatidyl choline, phosphatidyl ethanolamine and cholesterol allowed Cry proteins to partition into liposomes. Other combinations of lipids did not. Our results indicate that domain I alone of Cry proteins can partition into artificial SUVs, but not into BBMVs when F371C and F371A mutants block partitioning of domain II. This demonstrates that domain I can spontaneously partition into artificial lipids but not into native membranes. This observation calls into question the conclusions of previous studies using artificial vesicles as support for the Umbrella and Penknife

Models.

In summary, the data presented in this study indicate a role for the domain II residue, phenylalanine 371, in a post- receptor-binding step in the mechanism of action of the toxin.

Based on the available data from studies of the crystal toxin, after receptor binding, the toxin undergoes insertion into the vesicles to form ion channels. We postulate that phenylalanine 371 participates directly in the membrane insertion step, or in some post-binding step that leads to insertion.

Our observation that the mutant 1AbF371C is able to bind to the cadherin receptor, Bt-

R1, but has lost all measurable toxicity to M. sexta has implications on other current models for

the mechanism of action of Cry toxins. It is not in agreement with the conclusions of Zhang et al.[67], who propose that cytotoxicity is correlated to binding of Cry toxin to Bt-R1. In their model, cell death is mediated only by a Mg+2-dependent cellular response. The difference between our studies and those of Zhang et al. [102] possibly lie in the cellular system utilized.

Our results were gathered on M. sexta BBMV for insertion studies. The insects used in the bioassay were M. sexta and the electrophysiological studies were also done on M. sexta midguts, while Zhang, et.al [102] have used a High Five (Invitrogen, Inc.) cell line from Trichoplusia ni expressing a Bt-R1 fragment.

The larger goal of this study is to determine the mechanism of insertion of the toxin. The ability of domain II mutant to partition into the membranes of these vesicles is in agreement with our hypothesis proposed in our recent paper [83] that while -helices from domain I are able to insert, the mechanism of insertion is not based solely on individual helices entering the membrane. Other studies [86] suggest a buried “unchanged structure” model for the toxin where almost all of the toxin is buried into the membrane. Our data support the view that the mode of entry of the toxin into the membrane is in the form of an intact 60 kDa monomer or oligomeric toxin that lacks -helix 1 at its N-terminus.

CHAPTER 4

ANALYSES OF THE OLIGOMERIZED FORMS OF CRY1A TOXINS FORMED IN

SOLUTION, UPON RECEPTOR BINDING AND IN MEMBRANE.

4.1 The process of oligomerization of crystal toxins

An important step in understanding the mechanism of action of crystal toxins is the step of oligomerization of these toxins. The process of oligomerization of the toxin is reported at several steps during the mechanism of action. The toxin has been suggested to form several orders of high molecular weight species upon activation to 65 kDa monomer. As early as 1994,

Feng and Becktel had shown that at neutral pH values, Cry1A toxins prefer to stay as a monomer and the conversion from monomer to oligomer is slow[103]. However at high pH values of 10.0 or greater as seen in the midgut, the equilibrium shifts greatly to an oligomer that they predicted was due to a conformational change induced by pH as predicted by their CD spectra analysis in which the helical content of Cry1A toxins increased as pH changed from 6.0 to 11.0. They predicted that the increase in helical content with rising pH was a result of oligomerization of the toxin. They also suggest that the molecular weight of the oligomeric form is close to 220 kDa, indicating a likely tetramer. Another study by Guereca and Bravo [104]on the toxin oligomers in buffer generated a 60 kDa and a greater than 600 kDa form of the toxin. They suggested that the toxin forms a decamer in solution but that it was not the functional form (they however presented no data on toxicity). They concluded that the toxin only formed an oligomer once in the

membrane. However, the Bravo group has altered their view of oligomers from this original view as will be reported below. Another study by Masson et al. [105] performed dynamic light scattering studies to show that Cry1Ac, when incubated in low pH or under low salt conditions, rapidly aggregated to form an oligomeric form of the toxin in 50 mM Na2CO3 pH 10.5 as early as 3 hours after incubation and that the equilibrium shifts to the oligomeric form.

Arthur Aronson, in his studies of toxins extracted from brush border membrane vesicles, indicated that most toxins aggregate within the membrane to a 200 kDa form [53].

Immunoblotting analysis had shown that the aggregate was composed of only toxin without any of the major receptors bound to it. He also indicated that a mutation in - helix 5 of Domain I on residue H168 caused the toxin to lose its ability to oligomerize but retained toxicity (2-3 fold lower than wild type) and its ability to bind to BBMV.

The role of - helix 4 in oligomerization was confirmed by studies [106] on residue

N135 which, when mutated to glutamine, rendered Cry1A toxins non- toxic and also caused the toxin to stay as a monomer when bound to BBMV, as determined from SDS –PAGE analysis of toxins heated to 60oC.

More recently in 2004, Bravo et al. put forth a new model of the mechanism of toxicity of

Cry1A toxins where oligomerization of the toxin was an obligate intermediate in the process [54,

107]. Their model proposed that the toxin undergoes conformational change on binding to the cadherin receptor due to loss in  helix 1 at the N-terminus and forms a modified “prepore”.

They proposed that the prepore form of toxin loses binding to cadherin and binds to aminopeptidase more tightly. Binding to aminopeptidase results in the insertion of toxin into cholesterol- rich bilayer rafts of membrane, in agreement with the finding that membrane- bound 73

toxin is enriched in the lipid rafts of insect midgut membranes [91]. Their studies include mutagenesis of residues on  helix 3 of Domain I of the Cry1Ab toxin [107] that prevented the toxin from forming ion channels but retained binding of the toxin to BBMV. The structural integrity of these mutations was not discussed.

Quenching studies of tryptophan fluorescence with both KI and acrylamide carried out by

Bravo et al. [108] on monomers and prepore oligomers in solution and membrane- bound oligomers of Cry1Ab suggested that the tryptophans are at the lipid-water interface and are sequestered at the boundary of the membrane. They performed quenching of the tryptophan fluorescence of oligomeric toxin with lipids containing dibromo phosphatidylcholine, the bromines located at positions 6, 7 and 11, 12 of the fatty acid chain and showed that only 6,7- dibromo PC (closer to the head group of the fatty acid) was able to quench tryptophan residues as the molar concentration of the lipids increased. Their conclusion was that the toxin was able to interact with the membrane as an oligomer rather than as a monomer. The work did not examine quenching of tryptophan fluorescence from monomers inside labeled vesicles so a comparison of the rate of monomer and oligomer partitioning was not possible.

Mutational analysis of tryptophan residues in Cry1Ab toxin done by the Bravo group in collaboration with our lab [109] showed that competition binding of mutants of residues W316,

W219 and W455 were affected. These results were interpreted by the Bravo group as a result of lack of structure of the mutants (no structural analysis data was shown) or as a result of loss of binding. Irreversible binding analyses were not performed on any of the mutants.

In the present work, we have independently investigated the oligomeric state of the toxin in solution and in the membrane. This study is also aimed at verifying the serial- receptor - binding model [54] of toxicity of Cry proteins.

4.2 EXPERIMENTAL PROCEDURES

Preparation and purification of “in solution” oligomers of Cry1Ab toxin: Insecticidal crystal toxins were expressed in E. coli DH5 cells using a “leaky” T7 promoter in pBlueScript KS- vector. Cells were harvested after incubation at 25oC for 72-96 hours by centrifuging at 8000 rpm for 10 min. Crystal inclusion bodies were obtained from the cells by treating the cells with

1mg/ml of lysozyme and sonication followed by washes with 2% Triton X detergent and0.5M

NaCl. The crystals were solubilized in 50 mM Na2CO3 (pH 10.5) in the presence of 2% β- mercaptoethanol. Protoxin was activated using trypsin (1:50 w/w of trypsin: toxin). Activated toxins were purified using anion exchange chromatography (Sepharose Q column from GE

Healthcare) and gel filtration chromatography (Sephacryl S300 column from GE Healthcare).

The resultant proteins were purified further using a 90 ml Superdex 200 column (GE Healthcare) to obtain pure monomeric toxin. Monomers were concentrated to 2 M and stored in 1mM

PMSF (diluted from a stock of 100 mM PMSF prepared in isopropanol) for 4-6 days. The stored solution was further run on S200 gel filtration column and oligomer peaks were collected and concentrated to 1 M of oligomeric toxin used for characterization of these oligomers.

Characterization of the “in solution” oligomers: The oligomers obtained above were resolved on

6% SDS-PAGE gels. Sedimentation velocity experiments were performed by Dr. James Lear

(Department of Biochemistry and Biophysics, University of Pennsylvania) to determine the size of the oligomers in a Beckman XL-I analytical ultracentrifuge (Beckman Coulter). 75

Extraction of oligomeric forms of toxin from BBMV: Purified toxin monomers were mixed with insect brush border membranes in a ratio of 1:10 and incubated for 60 min. The BBMV were separated from free toxin by centrifuging the pellet at 21000 x g. BBMV were solubilized using

1% n-octyl-β-D-glucopyranoside in 10 mM HEPES pH 7.5. Low SDS loading dye (0.25%SDS) was added to the sample and warmed at 60oC for 5-10 min. Samples were loaded on 6% SDS

PAGE gels and blotted against anti Cry1A antibody (1:10,000). Samples were also boiled for 5 min and loaded on SDS-PAGE gels and blotted against anti-Cry1A antibody. This method is hereafter referred to as Aronson method after the principal author [53].

Characterization of oligomers extracted from BBMV: Samples extracted from BBMV were also analyzed on a Superdex 200 gel filtration column. The size of the membrane bound form of the toxin in their native state was determined using sedimentation velocity experiments in a

Beckman XL-I analytical ultracentrifuge.

Preparation and purification of “prepore” toxin oligomers: Prepore oligomers were prepared as described by Gomez et al. [110]or by the method described by Jiminez-Juarez et al. [107]. The two procedures are schematically represented in Figs. 4.1, 4.2 and 4.3 and a comparison of the methods is tabulated in Table 4.1. Method 1 has been split into two sub methods based on the components used (Figs.4.1 and 4.2). Briefly in method 1A (Fig. 4.1), [110], 1 g protoxin was mixed with 10 g insect brush border membrane vesicles (BBMV) in the presence of 50 l of solubilization buffer (50 mM Na2CO3 pH 10.5 + 0.2% β-mercaptoethanol) and incubated at room temperature (25-28oC) for 15 min. Alternatively, in the absence of brush border membranes, oligomers were obtained by using method 1B (Fig.4.2) [110]. In this method, 50 g of crystals were incubated with 200 g of scFv73, a synthetic single chain antibody that mimics a cadherin

fragment generated using a phage display method (courtesy of Drs. A. Bravo and M. Soberon), in the presence of 0.5 l of midgut juice (M. sexta) and 50 l of solubilization buffer at 37oC for

60 min. The incubations by either method were immediately followed by addition of 1mM

PMSF and centrifugation of the mixture at 21,000 x g for 15 min. The resulting supernatant (20

l) was treated with loading dye, boiled for 3-5 min, loaded on SDS-PAGE gel and blotted against an antibody generated by immunizing rabbits with the prepore reaction (referred to anti-

Cry1Ab prepore antibody in this manuscript) to observe oligomeric forms of the toxin.

Method 2 (Fig. 4.3) [107] is a modification of method 1B, where 200 ng of protoxin was incubated with scFv73 in a 1:2 mass ratio and 5% midgut juice was added in 100 l of solubilization buffer containing 60 M small unilamelar vesicles [74] made of 1,2-Dioleoyl-sn-

Glycero-3-phosphocholine (DOPC). The mixture was incubated at 37oC for 1 hr, followed by precipitating the membrane bound toxin at 400,000 x g in a Beckman L8 ultracentrifuge. A control reaction which lacked any SUV was used to show no precipitate formation in the absence of SUV. The pellet was resuspended in 50 l of buffer in presence of 10% n-octyl-β-D- glucopyranoside and clarified by centrifugation, treated with loading dye and boiled for 3-5 min.

Western blot analysis of the protein was performed using polyclonal anti Cry1Ab prepore antibody (1:50,000; 1hr) (courtesy of Dr. Bravo) and secondary HRP antibody (1:10,000; 1hr) and was detected using chemiluminescence substrate (Bio-Rad).

Purification of prepore oligomers of 1Ab toxin: Cry1Ab prepore oligomers prepared by either of the above methods were purified using a Superdex 200 HR 16/60 (GE Healthcare) on an AKTA

Explorer 100 (GE Healthcare). Several prepore samples were collected and pooled to yield 1-5

g of sample. Sample volume of 2 ml was loaded into the column for every run. The flow rate

was 1 ml/min and fractions of 2 ml were collected in a mobile phase containing protease inhibitors and 0.01% n-octyl-β-D-glucopyranoside. Purified toxins were analyzed using Western blots with polyclonal anti- Cry1Ab prepore antibody.

Characterization of prepore proteins: Purified 1Ab prepore proteins obtained from the two methods were subjected to sedimentation velocity analysis using a Beckman XL-I analytical ultracentrifuge to determine their size in native conditions. In order to determine the identity of the proteins in the purified prepore complex, we subjected the SDS-PAGE- resolved Cry1Ab prepore to LC-MS/MS analysis, performed at the W.M. Keck Facility, New Haven, CT. Briefly, the entire lane from the gel, containing the resolved prepore bands, was cut into 1 mm slices and digested with trypsin using in-gel digestion and the resulting samples were desalted in a 100 micron ID C18 column (Waters) in a gradient of 2-98% acetonitrile in the presence of 0.01% trifluoroacetic acid. Protein was considered identified if two or more peptides matched to the same protein accession numbers in the database (MASCOT analysis).

Solubilization buffer (50 Crystals/ Brush Border Membrane mM Na2CO3 pH 10.5 + 0.2% protoxin (1 g) vesicles (10 g) from M.sexta β-mercaptoethanol)

Incubate at 25 -28oC for 15 min

Add 1mM PMSF to inhibit/stop further reaction

Centrifuge reaction mixture at 21000 x g for

15 min. Supernatant contains prepore.

Prepore purified by gel filtration chromatography (Superdex 200 HR 16/60)

Fig.4.1 Schematic procedure [Method 1A] for obtaining prepore from Cry1Ab toxin [110] using insect brush border membranes.

Crystal/ protoxin single chain peptide M.sexta midgut Solubilization    50 g (scFv73) 200 g juice (0.5 L) buffer (50 L)

o Incubate @ 37 C for 60 min

Add 1mM PMSF to inhibit further reaction

Centrifuge the reaction mixture

at 21000 x g for 15 min.

Supernatant contains low

amounts of prepore.

Fig.4.2 Schematic procedure [Method 1B] for obtaining prepore from Cry1Ab toxin [110] in the absence of insect brush border membranes.

Toxin/ protoxin single chain peptide M.sexta midgut Solubilization Small 200 ng (scFv73) 400 ng juice (5%) buffer (50 L) Unilamellar Vesicles SUV (60 M)

Incubate at 37oC for 60 min on rotary shaker

Add 1 mM PMSF to inhibit further reaction

Centrifuge reaction mixture at 400,000 x g for 60 min. Prepore is in the membrane pellet.

Prepore extracted from membrane vesicles by addition of 1% n-octyl-β-D- galactopyranoside detergent.

Prepore purified by gel filtration using Superdex 200 HR 16/60

Fig.4.3 Schematic procedure [Method 2] for obtaining prepore from Cry1Ab toxin [107] using insect brush border membranes.

METHOD 1A* METHOD 1B* METHOD 2**

1. Source of Cry protein Bacillus Bacillus Bacillus thuringiensis or E. coli thuringiensis thuringiensis 2. Reaction components 1 g 50 g 200 ng protoxin/ 200 ng toxin protoxin/crystals crystals/protoxin + 400 ng scFv73 + 5% midgut + 10 g BBMV + + 200 g scFv73 juice + 100 l solubilization 50 l + 0.5 l midgut buffer + 60 M small solubilization juice +50 l unilamelar vesicles [74] buffer solubilization buffer 3. Incubation conditions 25oC -28oC / 15 37oC /60 min 37oC / 60 min min 4. Stopping the reaction Addition of 1mM PMSF to the reaction mixture 5. Collection of sample Centrifugation at 21000 x g / 15 min Centrifugation at 400,000 x g / 60 min 6. Sample Prepore sample is collected from Prepore sample is collected the supernatant from pellet 7. Extraction of sample N/A Membrane solubilized with 1% n-octyl-β-D-glucopyranoside 8. Purification of sample Purified complex through Superdex 200 HR 16/60 (GE Healthcare)

Table 4.1 Comparison of procedures used to prepare prepore oligomers.

* = Method 1A and 1B based on published protocol : Gomez et al. FEBS 2002 [110]

** = Method 2 based on published protocol: Jiminez-Juarez et al. JBC 2007 [107]

 = Solubilization buffer composition: 50 mM Na2CO3 pH 10.5 + 0.2% β-mercaptoethanol

4.3 RESULTS

Characterization of „in solution‟ oligomers of Cry1Ab toxin: The oligomers of Cry1Ab formed by allowing the toxin to be stored in 1 mM PMSF for 4-5 days were purified through Superdex

200 HR 10/90 column and resolved on 8 % SDS-PAGE gel. The fractions of the column that corresponded to the oligomer and monomer were displayed on the gel separately and stained as shown in Fig. 4.4. These oligomer fractions were pooled and secondary structures of the oligomer and monomer forms of 1Ab toxin were examined using circular dichroism spectrophotometry. No significant differences in the structure were observed between the two forms at pH 10.5 (Fig. 4.5). SedFit analyses of the monomeric and oligomeric forms (Fig. 4.6) using sedimentation velocity experiments show that the monomer slowly aggregated to the oligomeric form while the oligomer sample formed aggregates of higher order but showed no monomer in it, indicating that the equilibrium shifted in one direction: from monomer to oligomer.

Characterization of membrane- bound oligomers of Cry1Ab toxin: The membrane- bound oligomers of Cry1Ab were produced using the Aronson method as described [53]. When these toxins were extracted from the membrane using detergent, were incubated with low amounts of

SDS (0.5%) at 60oC for 10 min and were resolved on SDS PAGE gels, protein bands of high molecular weight (~250 kDa) aggregates / oligomers of Cry1Ab toxin were obtained as shown in

Fig. 4.7. However, when the same samples were boiled for 5 min in presence of 1% SDS in the loading buffer, the toxin molecules ran as a 60 kDa monomer (missing  helix 1) on SDS PAGE as seen in Fig. 4.8. Both Figs. 4.7 and 4.8 are western blots using anti- Cry1Ab antibody as the primary antibody. Gel filtration analyses showed monomeric and oligomeric peaks of the toxin extracted from the membrane as shown in Fig. 4.9. The size of the molecules as determined by 83

analytical ultracentrifugation experiments showed that the toxin existed in monomeric, dimeric, trimeric and tetrameric states in the membrane (Fig. 4.10). The secondary structure of the toxin in BBMV however did not reveal significant structural changes as demonstrated by CD spectrophotometry (Fig. 4.11).

Formation of prepore of Cry1Ab toxins: Cry1Aa and Cry1Ab were tested for prepore formation using the published protocol and materials provided to us by the Bravo group who first proposed formation of prepore oligomers [54]. Using a polyclonal antibody that was generated by immunizing rabbits with the prepore reaction mixture that was provided to us by Bravo, we were able to reproduce the prepore forms of the toxin. It is noteworthy that polyclonal antibody against native or denatured Cry1Ab did not detect the prepore. It was detected only with anti- prepore antibody.

Using Method 1A (Fig. 4.1) [111] we were successful in generating oligomeric prepore forms against toxins from Bacillus thuringiensis crystals (Fig. 4.12). This method produced abundant prepore when brush border membranes were used, however little to no prepore product was obtained in our hands when using scFv73 peptides in the absence of brush border membranes. This method was used to examine prepore formation in mutations of F371 that affected toxin insertion (Chapter 3) [34] . All three mutants used in that study formed prepore oligomer (Fig.4.13). Method 1B using single chain peptide antibody failed to provide detectable amounts of prepore complex in our hands. The modified protocol (Method 2, Fig. 4.3) of prepore synthesis involving addition of small unilamelar vesicles [74] [107] resulted in better recovery of the 250 kDa prepore than the former method (Method 1A, Fig. 4.1)and generated oligomers in the absence of BBMV (Fig. 4.14).

Characterization of the prepore complex: Cry1Ab prepore extracted by method 1A (Fig. 4.1) was purified using Superdex 200 HR 16/60 gel filtration column as a single peak (Fig. 4.16) and was subjected to sedimentation velocity measurements to verify the size of the oligomer.

However, the results of the analytical ultracentrifugation showed that the complex was of several size classes, including 250 kDa and higher forms (Fig. 4.17). Prepore complex made using toxin and BBMV was subjected to Western blot using anti Bt-R1 antibody (provided to us by Dr. Lee

A. Bulla Jr.,). The antibody showed strong reactivity to the prepore complex as a predominant 75 kDa band (Fig.4.18). Samples that reacted to both anti-1A prepore antibody and anti Bt-R1 antibody were subjected to mass spectrometry analyses. Initial a shotgun approach of the entire reaction mixture was performed which identified only Cry1A toxin in the complex. However, when prepore toxin samples were separated on 8% SDS-PAGE gels, digested using trypsin and subjected to mass spectrometry (LC-MS/MS) analyses, several proteins were identified (Table

4.2). Prominent among them were cadherin and aminopeptidase from M. sexta.

The modified protocol (Method 2 Fig.1) [107] yielded prepore complexes from E. coli as reported by Bravo [107]. Fig. 4.14 shows the prepore isolated and blotted against anti- Cry1A prepore antibody. The result of the reaction was a mixture of monomers and oligomers of the toxin on SDS PAGE. Purification of the prepore through a Superdex 200 HR 16/60 gel filtration column resulted in generation of a single peak that eluted in the void volume of the column

(Fig.4.19).The prepore generated using this method did not show a reactivity to anti-BtR1 antibody. Mass spectrometry (LC-MS/MS) from samples on trypsinized gel slices showed the presence of synthetic single chain peptides like scFv73 and a fragment of the variable region of

IgG in addition to Cry toxins as proteins in the purified complex (Table 4.3). Peptides identified from the prepore complex were mapped to the structure of the toxin to determine the regions of 85

the toxin in the complex that partitioned into artificial vesicles. We confirmed peptides from all three domains of the toxin present in the complex (Table 4.4). In fact, more peptides (with better homology or identity score) were identified from Domains II and III in the LC-MS/MS method

(Fig. 4.20), indicating that all three domains were present in the complex.

Bioassays using the gel filtration- purified Cry1Ab prepore complex was performed against 1st instar M. sexta larvae. Results showed that the toxicity of the prepore complex had an overlapping range of toxicity with that of the purified monomer complex as seen in Table 4.5.

The toxicity of F371, the mutant that lost its ability to partition into the membrane was able to form a prepore. The prepore form of the mutant was also non- toxic.

1 2 3 4 5 6 7 8 9 10

250

150

Fig.4.4: SDS PAGE of oligomers and monomers of Cry1Ab toxin in 50 mM Na2CO3 pH10.5 stored in 1mM PMSF for 40-72 hours. Lane 1= Cry1Ab protoxin, Lane 2 = Cry1Ab trypsinized toxin. Lane 3 -7 = Fractions eluted from Superdex 200 HR corresponding to the oligomer peak of Cry1Ab. Lane 8-10= Fractions eluted from Superdex 200HR corresponding to the monomer peak of Cry1Ab.

-10

mdeg -20

-30

-40 180 200 220 240 260 280 300 320 Wavelength (nm)

Fig. 4.5 Circular dichroism spectra of 2 M Cry1Ab toxin purified out as a monomer (●) and as an oligomer formed in solution ().

~65kDa ~1020 kDa

0 50 100

Sed. Coefficient (Svedbergs)

~1200 kDa B

5 20 50 100

Sed.coefficient (Svedbergs)

Fig. 4.6: Sedimentation velocity spectra obtained by SedFit analysis of the Superdex 200 HR purified toxin monomers (A) and oligomers formed in solution (B) run on Beckman XL-I. Peaks of the spectra were graphed against a Nomogram to measure range of molecular sizes in kDa. The size of some of the oligomer peaks could not be calculated using Nomogram method.

1 2

250

150

100

25 Fig. 4.7: Western blot of Cry1Ab toxin extracted from insect BBMV using n-octyl-β-D- glucopyranoside detergent and treated with 0.5% SDS final concentration. Samples were preheated to 60oC for 5 min before loading on the gel. Polyclonal anti Cry1Ab antibody was used at a dilution of 1 in 10,000. Lane 1 = Trypsinized Cry1Ab purified on Superdex 200 column and used for treating BBMV using above conditions. Lane 2= Toxin extracted from BBMV treated using above conditions.

1 2 250

150

100

Fig.4.8: Western blot of Cry1Ab toxin25 extracted from insect BBMV using n-octyl-β-D- glucopyranoside detergent and treated with 1.0% SDS final concentration. Samples were boiled at 100oC for 5 min before loading the gel. Polyclonal anti Cry1Ab (1 in 10000) used. Lane 1=

Trypsinized Cry1Ab toxin purified on S200 column used to treat BBMV. Lane 2 = Cry1Ab toxin extracted from BBMV and treated using above conditions.

800

600 Monomer toxin purified on S200

280 400

milliA 200

-200 0 10 20 30 40 50

Volume (mL)

100 Membrane 80 bound toxin purified on S200 60

280 40

milliA 20

0 10 20 30 40 50

Volume (mL)

Fig. 4.9: Superdex 200 HR gel filtration retention peaks of the monomer form of Cry1Ab versus the membrane bound forms of the Cry1Ab toxin.

0 10 20 30 40 Sed.coefficient (Svedbergs)

Fig.4.10 Sedimentation velocity run of the membrane bound fraction of Cry1Ab toxin extracted into β- OG detergent on Beckman XL-I analytical ultracentrifuge.

mdeg 0

-5

-10

-15 180 200 220 240 260 280 300 320 Wavelength (nm)

Fig.4.11 CD spectra of Cry1Ab toxin purified into 50 mM Na2CO3 pH 10.5 + 0.5M NaCl (●), bound to BBMV (▼) and bound to BBMV and treated with 10 fold excess proteinase K (). The

BBMV were separated from free toxin and/ or proteinase K by centrifugation at 21000 x g and resuspended in 10 mM HEPES pH 7.4 + 150 mM NaCl buffer.

Fig.4.12. Detection of prepore extracted from several Cry protoxins using Method 1A (Fig 4.1) of prepore extraction [110]. 100ngs of the reaction loaded on 8% SDS PAGE and blotted using polyclonal anti Cry1A prepore antibody (1 in 50000). Lane 1 = Prepore from Cry1Abwt produced in BtHD1-19. Lane 2 = Prepore from Cry1Ab F371C mutant produced in Bt4Q7. Lane

3 = Prepore from Cry1Aawt produced in Bt sotto 4E3. Lane 4 = Cry1Ab produced in E.coli

DH5. Lane 5= Cry1Aa produced in E.coli DH5.

1 2 3 4 5

Fig.4.13. 100ng of prepore reaction (Method 1A) from each of the samples below were loaded on 6% SDS PAGE gel and blotted against anti Cry1A prepore antibody (1 in 80000). Lane 1 =

Cry1Ab wild type from Bt HD1-19. Lane 2 = Cry1Ab V171C from Bt4Q7. Lane 3 = Cry1Ab

F371C from Bt4Q7. Lane 4 = Cry1Ab V171C protoxin control. Lane 5 = Cry1Ab F371C protoxin control.

1 2 3 4

250

150

100

Fig. 4.14. Western blot of prepore reactions formed using modified method 2 of extracting prepore [107]. Lane 1 = Cry1Ab trypsinized toxin control. Lane 2 = Cry1Ab toxin + 0.5 % midgut juice (No scFv control). Lane 3 = Cry1Ab toxin +0.5% midgut juice in presence of scFv73 peptide and SUV (pellet fraction). Lane 4 = Cry1Ab toxin +0.5% midgut juice in presence of scFv73 peptide and SUV (supernatant fraction).

Fig.4.15. Coomasie staining of 8% SDS PAGE gel of prepore prepared from Cry1Ab protoxin in the presence of M.sexta brush border membrane using Method 2 [110]. 3 g of total protein from purified prepore complex is loaded in lanes 2 and 3 of the gel. Lane 1 is the molecular weight marker.

800

Monomer 600

400

280

milliA 200

-200 0 10 20 30 40 50

Volume (mL)

Prepore 50 Oligomer

280 30

milliA 20

0 10 20 30 40 50 Volume (mL)

Fig.4.16. Superdex 200 HR 16/60 gel filtration column purification of the prepore complex. The two chromatograms show the elution profile of monomer toxin versus the prepore complex as indicated.

~600 kDa

~120 kDa ~400 kDa ~1200 kDa

10 20 30 40 50 60 70

Sed.coeff (Svedberg)

Fig.4.17. Sedimentation velocity spectra of Cry1Ab toxin prepore obtained by Method 1 [110].

Molecular weights indicated on each peak were deciphered using a Nomogram to an approximation.

100

1 2 3

250

150

100

Fig.4.18. Western analyses of prepore obtained by Method 1. Lane 1, 2 and 3 show 50, 100 and

200 ng of prepore reaction mixture loaded on 6% SDS PAGE gel and blotted against anti Bt-R1 antibody at a dilution of 1: 10000 (courtesy of Dr. Bulla LA Jr).

101

800

Monomer 600

400 280

milliA 200

-200 0 5 10 15 20 25 30 35 40 45 50 Volume (mL) 80

Prepore 60

280

milliA 20

0 10 20 30 40 50 Volume (mL)

Fig.4.19. Superdex 200 HR 16/60 retention peak of the prepore complex obtained using the

Method 2 [107] of prepore extraction from Cry1Ab. The two chromatograms show the elution profile of monomer toxin versus the prepore complex as indicated.

102

Fig.4.20. Regions of the Cry1Ab toxin that were identified based on peptide matched from LC-

MS/MS of digested Cry1Ab prepore samples. Red color indicates peptides matched to Domain I residues, green color indicates peptides matched to Domain II residues and yellow color indicates peptides matched to Domain III residues.

103

METHOD 1A: Cry1A prepore formed by treatment with M.sexta BBMV

GI accession # of Match Protein Identified Maximum no. of peptides

identified

gi|40255 Insecticidal crystal protein 13

gi|43099 Insecticidal crystal protein 13

gi|20465244 Cadherin from M.sexta 5

gi|2499901 APN-like protein (Membrane 32

alanyl aminopeptidase

precursor Cry1Ac receptor)

gi|8488965 Aminopeptidase 2 24

gi|20279109 Aminopeptidase 3 18

Table 4.2 Major proteins of known function from Bt crystal or insect receptors identified using

LC-MS/MS analysis of S200 purified prepore samples prepared by Method 1A [110].

104

METHOD 2: Cry1A prepore formed by treatment with M.sexta midgut juice and scFv73

peptide

GI accession # of Match Protein Identified Maximum no. of peptides

identified

gi|40255 Insecticidal crystal protein 21

gi|43099 Insecticidal crystal protein 16

gi|1902832 Single chain (scFv) antibody 3

Mol.wt. 26267

gi|106429 Ig heavy chain V region 2

(alpha-phOx15)-human

fragment

Mol.wt. 13696

Table 4.3 Major proteins of known function from Bt crystals or receptors identified using LC-

MS/MS analysis of S200 purified prepore samples prepared by Method 2 [107].

105

Sequence of Peptides Region of Toxin Residue positions on toxin identified DVSVFGQR Domain I (alpha 5) D174-R181 TLSSTLYR Domain II (beta) T361-R368 LSHVSMFR Domain II (beta) L430- R437 WYNTGLER Domain I (loop between alpha 6 & 7) W210-R217 TSPGQISTL Domain III T502-R511 GSAQGIEGSIR Domain II G282-R292 GPGFTGGDILR Domain III G490-R500 VNITAPLSQR Domain III V512-R521 IVAQLGQGVYR Domain II I350 -R360 WGFDAATINSR Domain I (loop between alpha 5 & 6) W182 - R192 EWEADPTNPALR Domain I (loop between alpha 3 & 4) E116 -R127 EIYTNPVLENFDGSFR Domain II E266 - R281 IEFVPAEVTFEAEYDLER Domain III end (C-term of active toxin) I602- R619 LEGLSNLYQIYAESFR Domain II L100-R115 ELTLTVLDIVSLFPNYDSR Domain I (alpha 7) E235-R253 SAEFNNIIPSSQITQIPLTK Domain III S458- K477 SPHLMDILNSITIYTDAHR Domain II S293- R312

Table 4.4 Sequence of major peptides (based on the homology scores) identified to crystal proteins from the LC-MS analysis of the prepore from all tested samples of prepore.

106

Sample LC50 Source organism for toxin (ng/cm2) production Cry1Aa Monomer 16.0 E.coli (8.0-25.3) Cry1Ab Monomer 20.0 B.thuringiensis & E.coli (7.5-31.7) Cry1Ab oligomers formed in 28.0 E.coli solution (10.0- 46.5) Cry1Ab toxin extracted 25.0 E.coli from membrane** (5.0-35.0) Cry1Aa prepore oligomer 17.1 E.coli* (6.2 -35.3) Cry1Ab prepore oligomer 26.2 B.thuringiensis & E.coli* (3.9-40.8) Cry1Ab V171C Monomer 20.0 E.coli (7.5-31.7) Cry1Ab V171C Prepore 44.4 B.thuringiensis & E.coli* (22.3 -76.0) Cry1Ab F371C Monomer >2000 E.coli

Cry1Ab F371C Prepore >2000 B.thuringiensis & E.coli* Oligomer Cry1Ab V171C/F371A >2000 E.coli Monomer Cry1Ab V171C/F371A >2000 B.thuringiensis & E.coli* Prepore Oligomer

Table 4.5 Bioassay data for the monomer, solution oligomer, prepore oligomer and membrane extracted forms of Cry1A toxins tested [* indicates that these prepore oligomers were tested by the modified protocol [107] and ** indicates that these toxins were extracted from membranes

(BBMVs) and retested for toxicity to insect larvae.]

107

4.4 DISCUSSION

This study is aimed at characterizing oligomers of insecticidal crystal toxins produced by different methods and under different conditions. We obtained Cry toxin oligomers that were formed by aggregation of toxin in solution, oligomers of toxin formed in the membrane and oligomers of toxin that were formed by binding to receptors but before the step of insertion based on the methodologies published for each [53, 54]. We were able to characterize each of these oligomers for their polydispersity and their components. Most of the methods rendered multiple- sized toxin structures on 6% or 8% SDS- PAGE gels. However it is interesting to note that SDS was incapable of disrupting all the oligomers, especially the ones formed after binding to the receptor (prepore form). In case of oligomers formed in the membrane, lowering the SDS content retained them as oligomers on SDS- PAGE when heated to 60oC but converted them to monomers when boiled in 1% SDS suggesting non covalent forces that are resistant to lower

SDS concentrations (Figs. 4.7 and 4.8).

The serial- receptor- binding model for the mechanism of action of insecticidal crystal toxins by Bravo et al. [54] proposed that binding of the toxin to Bt-R1 is the first step that a monomeric toxin undergoes, leading to cleavage of - helix 1 by a membrane- bound protease.

This cleavage induces a „conformational change‟ that forms a tetrameric toxin that is „insertion- competent‟. The tetrameric toxin is released from the cadherin to bind to APN driving it into lipid rafts in membrane. We examined this hypothesis at several steps.

We reproduced the prepore using the conditions and materials provided to us by the

Bravo group as mentioned in experimental procedures (Fig. 4.1- 4.3). The oligomer form of the toxin was generated in the presence of the cadherin receptor and/or the scFv73 peptide that

108

mimicked the cadherin receptor. However, as the toxin prepore was in a mixture of materials that was incorporated into it, including the proteases, receptors and the- non oligomerized crystals, we purified the prepore moiety. Purification showed generation of a >250 kDa single peak on the

Superdex 200 HR column, suggesting that the prepore was a complex. However, reactivity to the prepore antibody demonstrated that the single peak had the 250 kDa moiety of toxin in it. But

SDS- PAGE analysis showed smaller moieties below 50 kDa (Fig.4.15). Mass- spectrometric analysis of the purified complex confirmed that the prepore was a mixture of components that had, in addition to the tetramer toxin of 250 kDa, the cadherin and APN receptors or the scFv73 peptide bound to it. The confirmation that the toxin was not a pure 250 kDa form came from the sedimentation velocity analyses that suggested that the toxin was capable of binding the receptors or receptor mimics irreversibly, thereby suggesting that the prepore that was purified out as a single peak remained bound to the cadherin receptor. This contradicts the serial receptor- binding model which suggests that the prepore oligomer form releases from the cadherin to bind to the APN. The conformational change studied by the authors of the serial binding model is based on change in exposure of tryptophan to aqueous environment when forming the prepore

[108, 112]. However, this event could be mediated by several factors. First, the authors do not consider any form of oligomer beyond the tetramer. Second, if the toxin and cadherin are bound in the prepore, as we see in our mass spectrometric analysis of the prepore, it is possible that the tryptophan exposure to the aqueous environment is being protected by the cadherin or its mimics in the complex. Considering the 1nM affinity of the toxin to cadherin receptor [39] and the lack of evidence of a substantial conformational change in any study so far, it is difficult to fathom that the toxin can release itself from a 1 nM binding of cadherin to bind to APN with a 0.75 nM affinity as suggested by the serial receptor- binding model. While the formation of oligomers is

109

indisputable, the toxicity values of prepore are not remarkably higher in our hands as compared to that of the monomers.

Mutational analyses of Cry1Ab toxin that affected prepore formation [107] were performed by the Bravo group. They focused on Domain I residues and concluded that this region alone is the insertion competent region in agreement with the Umbrella Model [63]. Our earlier studies that focused on other regions of the toxin have shown a 60 kDa form of the toxin to be capable of inserting into the membrane [113, 114] and a domain II residue F371 to be participating in insertion of the toxin into membrane. This mutant is capable of forming the prepore (Fig.4.13) but the prepore form of F371C is non- toxic (Table 4.5). This suggests a role for Domain II in a step that follows receptor binding and prepore formation (if essential) to insert the toxin into the membrane bilayer of the insect gut.

In addition, results of the MS/MS analyses of the prepore prepared by Method 2 (Fig.4.3)

[107] have shown peptides (Table 4.4) from all three domains of the toxin to be present in the prepore form obtained from lipid vesicles. There were 5 peptides identified from Domain I, 7 from Domain II and 5 from Domain III. This suggests further that all 3 domains of the toxin were capable of partitioning into membranes even when they were bound to the receptors, in this case scFv73.

It has been shown that the prepore toxin induces a high stable conductance in artificial lipid bilayers [108]. We speculate that the complex mixture of toxin and receptor might be responsible for the increase in the ability to form ion channels. Another recent work suggested a synergistic enhancement of toxicity in the presence of some regions of the receptor [115]. We are currently testing the effect of toxin-scFv peptide complex on insect toxicity.

110

Studies on partial unfolding of Cry1Ab [112] have indicated that in the prepore form of

Cry1Ab, domains II and III unfold before domain I and are more sensitive to temperature and chaotropic agents like urea, suggesting that only Domain I inserts into the membrane in the oligomeric prepore structure. These studies concluded that this “flexible conformation” was induced by the oligomerization process, causing it to insert into the membrane. Our circular dichroism spectra demonstrate that the toxin does not suffer heavy structural alterations in the membrane before and after proteolysis (Fig.4.11), suggesting that the aforementioned conformational change may not be necessary for the toxin to insert into the membrane. Evidence for the serial receptor binding of the toxin in Gomez et al. [116] suggests that the domain II and

III suffer a conformational change upon binding to Bt-R1, causing them to dissociate from Bt-R1 and bind to APN. Our study indicates that the scFv73 mimic peptide of Bt-R1 is bound to the toxin prepore as shown by the mass spectrometric analysis of the prepore. In addition, if domains

II and III were more flexible as per the claim of Gomez et al., trypsin treatment of the prepore extracted from SUV and run on SDS-PAGE, would identify only Domain I peptides, upon LC-

MS/MS of the complex. In contrast our mass spectrometry data (Table 4.4) indicate that almost equal number of peptides with very high homology or identity scores from each of the domains

(5 from Domain I, 7 from Domain II and 5 from Domain III) were identified, suggesting the presence of entire toxin molecule in the prepore.

Our data suggest that oligomer formation of insecticidal crystal toxins is a possibility both in solution and in membrane and that these oligomers are also capable of producing toxicity within the same range as the monomer toxin. However, the the Bravo model requires that the toxin should bind to receptors in a serial manner to mediate toxicity which is not observed by our data. We saw that the toxin bound tightly to the receptor resulting in its insertion, either as a 111

monomer or oligomer, but the requirement of multiple receptors for insertion was not necessary.

Recent results [50]showed that when the binding of the Cry1Aa toxin to B. mori cadherin was reduced by 23- fold, there was a negligible 4- fold loss in its toxicity, indicating that in the absence of one receptor, binding to other available receptors might be sufficient to mediate toxicity. We also speculate that this is the reason why modified toxins (lacking alpha helix-1) generate toxicity in cadherin- silenced insects [117]; we are investigating this further. However, the present data are strong evidence to suggest that both monomers and oligomers of the toxin are able to insert into the membrane as a whole unit of toxin and all three domains of the toxin insert into the membrane.

112

CHAPTER 5

DISCUSSION

5.1. Comments on the current understanding of the process of insertion of Cry1A Toxins:

The topic of insertion of crystal toxins into insect brush border membranes has been studied since the inception of the first crystal structure of Cry 3A. However, the complexity of the interactions between the toxin, receptors and the membrane bilayer has resulted in a lack of a complete understanding of the path of toxicity mediated by these toxins. Based on the current studies evaluating toxicity mechanisms, there seems to be more than a single mechanism by which the toxicity seems to be mediated. Some studies indicate a path of toxicity where the cadherin receptor molecule, upon binding of the toxin, triggers a pathway resulting in the death of the cell by an apoptosis- like mechanism [42, 67, 102] . The process of insertion of the toxin clearly plays an important role in mediating toxicity as evidenced by the formation of a pore in toxic mutants and by studies showing that non-BBMV inserting mutants of Cry1Ab were not toxic [34, 114]. The role of oligomerization in mediating the process of insertion of crystal toxins is another step that requires further inquiry.

Most models predicting mechanism of toxicity have focused only on one route for the insertion and therefore toxicity [63, 67, 117] . The likelihood of more than a single mechanism has seldom been addressed [118]. A perfect example of this is the signal transduction model that provides the best understanding of the cellular signaling events that occur when the toxin is introduced into the gut and bound to the cadherin receptor [67]. However the fact that this model 113

does not recognize insertion into the membrane bilayer as a likely mode of insertion has undermined its significance. Non-toxic mutants of domain II residue F371 were examined for its ability to bind BBMV, cadherin and APN and have shown to retain binding to all the receptors but have lost ability to insert into brush border membranes in contrast with the predictions of the signal transduction model [34].

Other models predicting the insertion of Cry toxins have also followed suit by suggesting a single route by which these toxins mediate toxicity. The serial receptor binding model is a good example of this. The authors propose that a conformational change of the toxin to form a tetramer form of the toxin is an obligate requirement for the toxin to insert [107]. According to the authors, insertion of monomer toxins into the membrane is insufficient to mediate the toxicity. When we compare the toxicity of oligomer and monomer forms of the toxin (Table 4.5), we observe that the toxicity values are in an overlapping range. The requirement that the toxin has to bind in a serial manner is contradicted by several studies where loss of binding to a receptor retained toxicity values and vice versa that loss of toxicity was observed even when binding to the receptors were intact [50, 89]. Mutant F328 of Cry 1Aa had lost binding to cadherin from Bombyx mori while retaining toxicity to the insect. The mutants of Cry1Ab F371 also retain binding to cadherin and form prepore oligomer even though they are non toxic. These examples definitely hint of the likelihood that several mechanisms are in operation to mediate toxicity by the toxins depending on the toxin and its specific target host. Thus a single model of toxicity of insertion for all toxins is not supported by published data.

114

5.2. Future studies to improve understanding of the mechanism of insertion of toxin into insect membranes:

Current analyses on the insertion process into natural vesicles (BBMV) and artificial vesicles (SUV) have not upheld the Umbrella Model for insertion of Cry toxins [34, 114]. Many more aspects of insertion need to be studied to gain a complete understanding of the orientation of the toxin regions in the membrane and its associations within the membrane.

One of the first steps is to understand the role of F371 mutants and thereby the region of the toxin in mediating toxicity. Our studies have postulated that the bulk group attached to the side chain of the residue at position 371 seems to allow or mediate partitioning and insertion of the toxin into membranes. Testing for this biological phenomenon where the toxin is driven by a bulky group to mediate insertion into membranes may be carried out using fluorescent spectroscopy involving use of heterobifunctional cross linkers attaching hydrophobic and / or bulky residues on the side chain of non- toxic mutant of F371. Mutants cross linked to bulk side chain maybe tested for restoration of their pore forming abilities using voltage clamp analyses.

One of the urgent questions that needs addressed since the observation that all three domains insert into membranes is the orientation of the toxin when inserted in the membrane vesicles. While fluorescence energy transfer (FRET) measurements are underway in examining the extent of insertion of residues in different domains, the current method does not provide a quantitative measurement of the depth of insertion into membrane due to the wide range of the

Förster radius values for energy transfer between the acrylodan and 7-nitro-2-1, 3- benzoxadiazol-4-yl (NBD) labels used in these experiments. A more quantitative approach to understand the depth of the toxin in insect membranes is to use paramagnetic spin labels attached

115

to cysteines mutants [119-122]. This site directed spin labeling process examines the depth of the labeled residue by measuring the frequency of collision of the spin label with polar and non polar reagents that are paramagnetic themselves using power saturation -electron paramagnetic resonance (EPR) spectroscopy. A very good example of a non polar reagent used in such studies is oxygen while that of a polar reagent is Ni-EDDA. Using the cysteine mutants available in all the three domains of Cry1A toxins we can examine the depth of these mutants using the power saturation of the spin in presence of these reagents. Alzate et al. (manuscript in preparation) have initiated such experiments using a spin label attached successfully to mutant S176C. The most widely used spin label to attach to cysteines: methanethiosulfonate (MTS) was used in these studies indicating that the residue was not deeply buried in the membrane. However, a complete profile generated by spin labeling of residues in all the domains of the toxin will provide specific depths of insertion for each domain and the orientation of the sheets and helices in these domains.

Besides these biophysical approaches, we are also pursuing another direct approach of obtaining crystals of the Cry1Ab toxin in membranes using artificial lipids called bicelles, a complex of phospholipids and detergents [123] in order to obtain the structure of the membrane- bound form of Cry1Ab toxin using X-ray crystallography analyses. Initial trials of crystallization have generated crystals that diffract at 5.0Åor lower resolution (Lee M and Chan MK, personal observation). Further trials will be performed to optimize generation of crystals in bicelles with the aim of high resolution diffraction. Also obtaining crystals of the toxin and receptor in membranes is another approach that will provide structural insights into the understanding of the insertion mechanism of the toxin.

116

Another important step is the examination of the affinity of prepore form of Cry toxins to its receptors. The prediction of the model that only a monomeric form of the toxin can bind the cadherin with high affinity but that an oligomeric form cannot, needs further analyses. Binding parameters of the prepore form to cadherin and APN receptors would enable evaluation of the serial binding model. We are investigating the increased toxicity reported for the toxin prepore form [107] or the modified toxin that was toxic to cadherin silenced forms of Manduca sexta

(using RNAi experiments) [117], which could be due to presence of the cadherin receptor in the prepore complex that substituted for the silenced cadherin in the latter study.

As we examine the insertion process of the toxin, another goal that needs more attention is the engineering of toxins with increased toxicity and having more widespread application. One of the mutants used in my studies Cry1Ab V171C (generated by O. Alzate;) was found to be more efficient in forming pores as compared to the rate of pore formation by the wild type

Cry1Ab toxin (Alzate et al. unpublished observation). This observation has generated enthusiasm for toxins that would have better performance in terms of pore formation and thereby toxicity.

Mutations likely to enhance pore formation needs to be investigated by voltage clamping analyses and bioassays. Besides mutants with better pore forming abilities, generation of Cry toxins with a wider spectrum of insecticidal activity are also a need for future. Mosquitocidal activity against Culex pipiens was introduced into otherwise lepidopteran Cry1Aa toxin by rational design around loop 2 of Domain II of the toxin. Studies with a similar goal may also allow suppression of resistance to the toxin.

117

BIBLIOGRAPHY

1. van Frankenhuyzen, K., The challenge of Bacillus thuringiensis, in Bacillus thuringiensis an Environmental Biopesticide: Theory and Practice, P.F. Entwistle, et al., Editors. 1993, Wiley: New York. p. 1-36.

2. Beegle, C.C. and T. Yamamoto, History of Bacillus thuringiensis Berliner research and development. Can. Ent., 1992. 124: p. 587-616.

3. Berliner, E., Uber de schlaffsucht der Mehlmottenraupe. Zeitschrift fur das Gesamstadt, 1911. 252: p. 3160-3162.

4. Goldberg, L.J. and J. Margalit, A bacterial spore demonstrating rapid larvicidal activity against Anopheles sergentii, Uranotaenia unguiculata, Culex univitattus, Aedes aegypti, and Culex pipiens. Mosquito News, 1977. 37: p. 355-358.

5. Schnepf, E., et al., Bacillus thuringiensis and its pesticidal crystal proteins. Microbiol. Mol. Biol. Rev., 1998. 62: p. 775-806.

6. Crickmore, N., et al., Revision of the nomenclature for the Bacillus thuringiensis pesticidal crystal proteins. Microbiol. Mol. Biol. Rev., 1998. 62(3): p. 807-813.

7. Carson, R., Silent Spring. 1962, Cambridge, MA: The Riverside Press. 368.

8. Bates, S.L., et al., Insect resistance management in GM crops: past, present and future. Nat Biotechnol, 2005. 23(1): p. 57-62.

9. Agaisse, H. and D. Lereclus, How does Bacillus thuringiensis produce so much insecticidal crystal protein? J. Bacteriol., 1995. 177: p. 6027-6032.

10. Sekar, V., The insecticidal crystal protein gene is expressed in vegetative cells of Bacillus thuringiensis var. tenebrionis. Curr. Microbiol., 1988. 17: p. 347-349.

11. Malvar, T., C. Gawron-Burke, and J.A. Baum, Overexpression of Bacillus thuringiensis HknA, a histidine protein kinase homolog, bypasses early Spo- mutations that result in CryIIIA overproduction. J. Bacteriol., 1994. 176: p. 4742-4749.

118

12. Choma, C.T. and H. Kaplan, Folding and unfolding of the protoxin from Bacillus thuringiensis: Evidence that the toxic moiety is present in an active conformation. Biochemistry, 1990. 29: p. 10971-10977.

13. Höfte, H. and H.R. Whiteley, Insecticidal crystal proteins of Bacillus thuringiensis. Microbiol. Rev., 1989. 53: p. 242-255.

14. Minnich, S.A. and A.I. Aronson, Regulation of protoxin synthesis in Bacillus thuringiensis. J. Bacteriol., 1984. 158: p. 447-454.

15. Hofmann, C., et al., Binding of the delta-endotoxin from Bacillus thuringiensis to brush-border membrane vesicles of the cabbage butterfly (Pieris brassicae). Eur. J. Biochem., 1988. 173: p. 85- 91.

16. Aronson, A.I., et al., The solubility of inclusion proteins from Bacillus thuringiensis is dependent upon protoxin composition and is a factor in toxicity to insects. Appl. Environ. Microbiol., 1991. 57: p. 981-986.

17. Du, C., P.A.W. Martin, and K.W. Nickerson, Comparison of disulfide contents and solubility at alkaline pH of insecticidal and noninsecticidal Bacillus thuringiensis protein crystals. Appl. Environ. Microbiol., 1994. 60: p. 3847-3853.

18. Choma, C.T., et al., Secondary structure of the entomocidal toxin from Bacillus thuringiensis subsp. kurstaki HD-73. J. Protein Chem, 1990. 9: p. 87-94.

19. Chestukhina, G.G., et al., The main features of Bacillus thuringiensis delta-endotoxin molecular structure. Arch. Microbiol., 1982. 132: p. 159-162.

20. Bietlot, H.P., et al., Evidence that the CryIA crystal protein from Bacillus thuringiensis is associated with DNA. J. Biol. Chem., 1993. 268: p. 8240-8245.

21. Clairmont, F.R., et al., Role of DNA in the activation of the Cry1A insecticidal crystal protein from Bacillus thuringiensis. J. Biol. Chem., 1998. 273(15): p. 9292-9296.

22. Bietlot, H.P., et al., Isolation of carboxyl-terminal peptides from proteins by diagonal electrophoresis: application to the entomocidal toxin from Bacillus thuringiensis. Anal. Biochem., 1989. 181: p. 212-215.

23. Nagamatsu, Y., et al., A toxic fragment from the entomocidal crystal protein of Bacillus thuringiensis. Agric. Biol. Chem., 1984. 48: p. 611-619.

24. Li, J., J. Carroll, and D.J. Ellar, Crystal structure of insecticidal -endotoxin from Bacillus thuringiensis at 2.5 Å resolution. Nature, 1991. 353: p. 815-821.

25. Boonserm, P., et al., Crystal structure of the mosquito-larvicidal toxin Cry4Ba and its biological implications. J. Mol. Biol., 2005. 348: p. 363-382.

119

26. Grochulski, P., et al. 3-D crystal structure of lepidopteran-specific delta-endotoxin Cry1A(a). in Proc. VIth Internatl. Coll. Invert. Path. Micro. Control. 1994. Montpellier, France.

27. Morse, R., T. Yamamoto, and R.M. Stroud, Structure of Cry2Aa suggests an unexpected receptor binding epitope. Structure, 2001. 9: p. 409-417.

28. Boonserm, P., et al., Structure of the functional form of the mosquito larvicidal Cry4Aa toxin from Bacillus thuringiensis at a 2.8-angstrom resolution. J Bacteriol, 2006. 188(9): p. 3391-401.

29. Galitsky, N., et al., Structure of the insecticidal bacterial delta-endotoxin Cry3Bb1 of Bacillus thuringiensis. Acta Crystallogr D Biol Crystallogr, 2001. 57((Pt 8)): p. 1101-1109.

30. Lee, M.K., et al., Location of a Bombyx mori receptor binding region on a Bacillus thuringiensis - endotoxin. J. Biol. Chem., 1992. 267: p. 3115-3121.

31. Wolfersberger, M.G., The toxicity of two Bacillus thuringiensis -endotoxins to gypsy moth larvae is inversely related to the affinity of binding sites on midgut brush border membranes for the toxins. Experientia, 1990. 46: p. 475-477.

32. Ihara, H., et al., Specific toxicity of -endotoxins from Bacillus thuringiensis to Bombyx mori. Biosci. Biotech. Biochem., 1993. 57: p. 200-204.

33. Liang, Y., S.S. Patel, and D.H. Dean, Irreversible binding kinetics of Bacillus thuringiensis CryIA - endotoxins to gypsy moth brush border membrane vesicles is directly correlated to toxicity. J. Biol. Chem., 1995. 270: p. 24719-24724.

34. Nair, M.S., X.S. Liu, and D.H. Dean, Membrane insertion of the Bacillus thuringiensis Cry1Ab toxin: single mutation in domain II block partitioning of the toxin into the brush border membrane. Biochemistry., 2008. 47(21): p. 5814-22.

35. Ge, A.Z., N.I. Shivarova, and D.H. Dean, Location of the Bombyx mori specificity domain on a Bacillus thuringiensis -endotoxin protein. Proc. Natl. Acad. Sci. USA, 1989. 86: p. 4037-4041.

36. Lu, H., F. Rajamohan, and D.H. Dean, Identification of amino acid residues of Bacillus thuringiensis -endotoxin CryIAa associated with membrane binding and toxicity to Bombyx mori. J. Bacteriol., 1994. 176: p. 5554-5559.

37. Sangadala, S., et al., A mixture of Manduca sexta aminopeptidase and phosphatase enhances Bacillus thuringiensis insecticidal CryIA(c) toxin binding and 86Rb+-K+ efflux in vitro. J. Biol. Chem., 1994. 269: p. 10088-10092.

38. Knight, P.J.K., N. Crickmore, and D.J. Ellar, The receptor for Bacillus thuringiensis CryIA(c) delta- endotoxin in the brush border membrane of the lepidopteran Manduca sexta is aminopeptidase N. Mol. Microbiol., 1994. 11(3): p. 429-436.

120

39. Vadlamudi, R.K., et al., Cloning and expression of a receptor for an insecticidal toxin of Bacillus thuringiensis. J. Biol. Chem., 1995. 270: p. 5490-5494.

40. Jenkins, J.L. and D.H. Dean, Binding specificity of Bacillus thuringiensis Cry1Aa for purified, native Bombyx mori aminopeptidase N and cadherin-like receptors. BMC Biochem., 2001. 2(12): p. 1-7.

41. Hua, G., J.L. Jurat-Fuentes, and M.J. Adang, Bt-R1a Extracellular Cadherin Repeat 12 Mediates Bacillus thuringiensis Cry1Ab Binding and Cytotoxicity. J. Biol. Chem., 2004. 279: p. 28051 - 28056.

42. Griko, N.B., et al., Univalent binding of the Cry1Ab toxin of Bacillus thuringiensis to a conserved structural motif in the cadherin receptor BT-R1. Biochemistry, 2007. 46(35): p. 10001-7.

43. Masson, L., et al., Kinetics of Bacillus thuringiensis toxin binding with brush border membrane vesicles from susceptible and resistant larvae of Plutella xylostella. J. Biol. Chem., 1995. 270: p. 11887-11896.

44. Valaitis, A.P., et al., Isolation and Partial Characterization of Gypsy Moth BTR-270, an Anionic Brush Border Membrane Glycoconjugate that Binds Bacillus thuringiensis Cry1A Toxins with High Affinity. Archives of Insect Biochemistry and Physiology, 2001. 46(4): p. 186 - 200.

45. Griffitts, J.S., et al., Glycolipids as receptors for Bacillus thuringiensis crystal toxin. Science, 2005. 307: p. 922-925.

46. Dennis, R.D., et al., Thin layer chromatography overlay technique in the analysis of the binding of the solubilized protoxin of Bacillus thuringiensis var. kustaki to an insect glycosphingolipid of known structure. Biomed. Chromatogr., 1986. 1: p. 31-37.

47. Lee, M.K., et al., Role of two arginine residues in domain II, loop 2 of Cry1Ab and Cry1Ac Bacillus thuringiensis -endotoxin in toxicity and binding to Manduca sexta and Lymantria dispar aminopeptidase N.Mol. Microbiol., 2000. 38(2): p. 289 - 98.

48. Lee, M.K., et al., Mutations at the arginine residues in 8 loop of Bacillus thuringiensis - endotoxin Cry1Ac affect toxicity and binding to Manduca sexta and Lymantria dispar aminopeptidase N. FEBS Letts., 2001. 497: p. 108-112.

49. Jenkins, J.L., et al., Bivalent sequential binding model of a Bacillus thuringiensis toxin to gypsy moth aminopeptidase N receptor. J. Biol. Chem., 2000. 275: p. 14423-14431.

50. You, T.H., et al., Blocking binding of Bacillus thuringiensis Cry1Aa to Bombyx mori cadherin receptor results in only a minor reduction of toxicity. BMC Biochem, 2008. 9: p. 3.

51. Knowles, B.H. and D.J. Ellar, Colloid-osmotic lysis is a general feature of the mechanism of action of Bacillus thuringiensis -endotoxins with different insect specificity. Biochim. Biophys. Acta, 1987. 924: p. 509-518.

121

52. Rausell, C., et al., Unforlding events in the water-soluble monomeric Cry1Ab toxin during transition to oligomeric pre-pore and to membrane inserted pore channel. J. Biol. Chem., 2004. 279(53): p. 55168 - 75.

53. Aronson, A.I., C. Geng, and L. Wu, Aggregation of Bacillus thuringiensis Cry1A toxins upon binding to target insect larval midgut vesicles. Appl. Environ. Microbiol., 1999. 65(6): p. 2503- 2507.

54. Bravo, A., et al., Oligomerization triggers binding of a Bacillus thuringiensis Cry1Ab pore-forming toxin to aminopeptidase N receptor leading to insertion into membrane microdomains. Biochem. Biophys. Acta, 2004. 1667: p. 38-46.

55. Lane, N.J., J.G. Harrison, and W.M. Lee, Changes in microvilli and golgi-associated membranes of lepidopteran cells induced by an insecticidally active bacterial -endotoxin. J. Cell Sci., 1989. 93: p. 337-347.

56. Billingsley, P.F. and M.J. Lehane, Structure and ultrastructure of the insect midgut. Biology of the Insect Midgut, ed. M.J.L.a.P.F. Billingsley. 1996, London: Chapman & Hall. 3-30.

57. Adang, M.J. and K.D. Spence, Surface morphology of peritrophic membrane formation in the cabbage looper, Trichoplusia ni. Cell Tissue Res, 1981. 218(1): p. 141-7.

58. Richards, A.G. and P.A. Richards, The peritrophic membranes of insects. Annu Rev Entomol, 1977. 22: p. 219-40.

59. Wieczorek, H., et al., A vacuolar-type proton pump energizes K+/H+ antiport in an animal plasma membrane. J Biol Chem, 1991. 266(23): p. 15340-7.

60. Cioffi, M. and W.R. Harvey, Comparison of potassium transport in three structurally distinct regions of the insect midgut. J. Exp. Biol., 1981. 91: p. 103-116.

61. Hodgman, T.C. and D.J. Ellar, Models for the structure and function of the Bacillus thuringiensis -endotoxins determined by compilational analysis. DNA Sequence. Map, 1990. 1: p. 97-106.

62. Lakey, J.H., et al., The membrane insertion of colicins. FEBS Letts., 1992. 307: p. 26-29.

63. Knowles, B.H., Mechanism of action of Bacillus thuringiensis insecticidal -endotoxins. Adv. Insect Physiol., 1994. 24: p. 275-308.

64. Gazit, E., et al., The structure and organization within the membrane of the helices composing the pore-forming domain of Bacillus thuringiensis d-endotoxin are consistent with an "umbrella- like" structure of the pore. Proc. Nat'l. Acad. Sci., 1998. 95: p. 12289-12294.

65. Masson, L., et al., Helix 4 of the Bacillus thuringiensis Cry1Aa toxin lines the lumen of the ion channel. J. Biol. Chem., 1999. 274: p. 31996-32000.

122

66. Gazit, E. and Y. Shai, The assembly and organization of the 5 and 7 helices from the pore- forming domain of Bacillus thuringiensis -endotoxin. J. Biol. Chem., 1995. 270: p. 2571-2578.

67. Zhang, X., et al., A mechanism of cell death involving an adenylyl cyclase/PKA signaling pathway is induced by the Cry1Ab toxin of Bacillus thuringiensis. Proc Natl Acad Sci U S A, 2006. 103(26): p. 9897-902.

68. Bittner, M. and D. Vapnek, Versatile cloning vectors derived from the runaway-replication plasmid pKN402. Gene, 1981. 15: p. 319-329.

69. Aronson, A., Incorporation of protease K into larval insect membrane vesicles does not result in disruption of function of the pore-forming Bacillus thuringiensis -endotoxins. Applied and Environmental Microbiology, 2000. 66: p. 4568-4570.

70. Tomimoto, K., T. Hayakawa, and H. Hori, Pronase digestion of brush border membrane-bound Cry1Aa shows that almost the whole activated Cry1Aa molecule penetrates into the membrane. Comp Biochem Physiol B Biochem Mol Biol, 2006. 144(4): p. 413-22.

71. Lee, M.K., et al., Involvement of two amino acid residues in the loop region of Bacillus thuringiensis Cry1Ab toxin in toxicity and binding to Lymantria dispar. Biochem. Biophys. Res. Commun., 1996. 229: p. 139-146.

72. Schnepf, H.E. and H.R. Whiteley, Cloning and expression of the Bacillus thuringiensis crystal protein gene in Escherichia coli. Proc. Natl. Acad. Sci. USA, 1981. 78: p. 2893-2897.

73. Kunkel, T.A., Rapid and efficient site-specific mutagenesis without phenotypic selection. Proc. Natl. Acad. Sci. USA, 1985. 82: p. 488-492.

74. Suvorov, A., J. Kok, and G. Venema, Transformation of group A streptococci by electroporation. FEMS Microbiol. Lett., 1988. 56: p. 95-100.

75. Grochulski, P., et al., Bacillus thuringiensis CryIA(a) insecticidal toxin: crystal structure and channel formation. J. Mol. Biol., 1995. 254: p. 447-464.

76. Peyronnet, O., et al., Ion channels induced in planar lipid bilayers by the Bacillus thuringiensis toxin Cry1Aa in the presence of gypsy moth (Lymantria dispar) brush border membrane. J. Membrane Biol., 2001. 184: p. 45-54.

77. Schwartz, J.-L., et al., Lepidopteran-specific crystal toxins from Bacillus thuringiensis form cation- and anion-selective channels in planar lipid bilayers. J. Membrane Biol., 1993. 132: p. 53-62.

78. Pitcher, W.H., 3rd and W.H. Huestis, Preparation and analysis of small unilamellar phospholipid vesicles of a uniform size. Biochem Biophys Res Commun, 2002. 296(5): p. 1352-5.

79. Liebig, B., D.L. Stetson, and D.H. Dean, Quantification of the effect of Bacillus thuringiensis toxins on short-circuit current in the midgut of Bombyx mori. J. Insect Physiol., 1995. 41: p. 17-22.

123

80. Wolfersberger, M., et al., Preparation and partial characterization of amino acid transporting brush border membrane vesicles from the larval midgut of the cabbage butterfly (Pieris brassicae). Comp. Biochem. Physiol., 1987. 86A: p. 301-308.

81. Chamberlin, M.E., Developmental changes in midgut ion transport and metabolism in the tobacco hornworm (Manduca sexta). Physiological Zoology, 1994. 67: p. 82-94.

82. Arnold, S., et al., The role of a proline-induced broken-helix motif in -helix 2 of Bacillus thuringiensis -endotoxins. FEBS Letts., 2001. 490: p. 70-74.

83. Alzate, O., et al., Effects of disulfide bridges in domain I of Bacillus thuringiensis Cry1Aa delta- endotoxin on ion-channel formation in biological membranes. Biochemistry, 2006. 45(45): p. 13597-605.

84. Rajamohan, F., et al., Role of domain II, loop 2 residues of Bacillus thuringiensis CryIAb - endotoxin in reversible and irreversible binding to Manduca sexta and Heliothis virescens. J. Biol. Chem., 1996. 271: p. 2390-2397.

85. Gazit, E. and Y. Shai, Structural and functional characterization of the 5 segment of Bacillus thuringiensis -endotoxin. Biochemistry, 1993. 32: p. 3429-3436.

86. Loseva, O.I., et al., Structure of Cry3A -Endotoxin within Phospholipid Membranes. Biochemistry, 2001. 40: p. 14143-14151.

87. Van Rie, J., et al., Specificity of Bacillus thuringiensis -endotoxin: Importance of specific receptors on the brush border membrane of the mid-gut of target insects. Eur. J. Biochem., 1989. 186: p. 239-247.

88. Ge, A.Z., et al., Functional domains of Bacillus thuringiensis insecticidal crystal proteins: refinement of Heliothis virescens and Trichoplusia ni specificity domains on CryIA(c). J. Biol. Chem., 1991. 266: p. 17954-17958.

89. Garczynski, S.F., J.W. Crim, and M.J. Adang, Identification of putative insect brush border membrane-binding molecules specific to Bacillus thuringiensis -endotoxin by protein blot analysis. Appl. Environ. Microbiol., 1991. 57: p. 2816-2820.

90. Rajamohan, F., et al., Single amino acid changes in domain II of Bacillus thuringiensis CryIAb - endotoxin affect irreversible binding to Manduca sexta midgut membrane vesicles. J. Bacteriol., 1995. 177: p. 2276-2282.

91. Zhuang, M., et al., Heliothis virescens and Manduca sexta lipid rafts are involved in Cry1A toxin binding to the midgut epithelium and subsequent pore formation. J. Biol. Chem., 2002. 277: p. 13863-13872.

92. Schwartz, J.L., et al., Restriction of intramolecular movements within the CrylAa toxin molecule of Bacillus thuringiensis through disulfide bond engineering. FEBS Letts., 1997. 410: p. 397-402.

124

93. Malmqvist, M., Biospecific interaction analysis using biosensor technology. Nature, 1993. 361: p. 186-187.

94. Malmqvist, M., Surface plasmon resonance for detection and measurement of antibody-antigen affinity and kinetics. Curr Opin Immunol, 1993. 5(2): p. 282-6.

95. Jönsson, U., et al., Real-time biospecific interaction analysis using surface plasmon resonance and a sensor chip technology. BioTechniques, 1991. 11(5): p. 620-627.

96. Valeva, A., et al., Staphylococcal alpha-toxin: repair of a calcium-impermeable pore in the target cell membrane. Mol Microbiol, 2000. 36(2): p. 467-76.

97. Prendergast, F.G., et al., Synthesis, spectral properties, and use of 6-acryloyl-2- dimethylaminonaphthalene (Acrylodan). A thiol-selective, polarity-sensitive fluorescent probe. J Biol Chem, 1983. 258(12): p. 7541-4.

98. Dorsch, J.A., et al., Cry1A toxins of Bacillus thuringiensis bind specifically to a region adjacent to the membrane-proximal extracellular domain of BT-R1 in Manduca sexta: Involvement of a cadherin in the entomopathogenicity of Bacillus thuringiensis. Insect Biochem. Mol. Biol., 2002. 32: p. 1025-1036.

99. Vie, V., et al., Lipid-induced pore formation of the Bacillus thuringiensis Cry1Aa insecticidal toxin. J. Membrane Biol., 2001. 180: p. 195-203.

100. Peyronnet, O., et al., Effect of Bacillus thuringiensis toxins on the membrane potential of lepidopteran insect midgut cells. Appl. Environ. Microbiol., 1997. 63: p. 1679-1684.

101. Yunovitz, H., et al., A new sensitive method for determining the toxicity of a highly purified fraction from delta-endotoxin produced by Bacillus thuringiensis var. entomocidus on isolated larval midgut of Spodoptera littoralis (Lepidoptera, Noctuidae). J. Invert. Pathol., 1986. 48: p. 223-231.

102. Zhang, X., et al., Cytotoxicity of Bacillus thuringiensis Cry1Ab toxin depends on specific binding of the toxin to the cadherin receptor BT-R1 expressed in insect cells. Cell Death Different., 2005. 12: p. 1407-16.

103. Feng, Q. and W.J. Becktel, pH induced conformational transitions of CryIA(a), CryIA(c) and CryIIIA -endotoxins in Bacillus thuringiensis. Biochemistry, 1994. 33: p. 8521-8526.

104. Guereca, L. and A. Bravo, The oligomeric state of Bacillus thuringiensis Cry toxins in solution. Biochim. Biophys. Acta, 1999. 1429: p. 342-350.

105. Masson, L., et al., Polydispersity of Bacillus thuringiensis Cry1 toxins in solution and its effect on receptor binding kinetics. Biochim. Biophys. Acta, 2002. 1594: p. 266-275.

125

106. Tigue, N.J., J. Jacoby, and D.J. Ellar, The -helix 4 residue, Asn135, is involved in the oligomerization of Cry1Ac1 and Cry1Ab5 Bacillus thuringiensis toxins. Appl. Environ. Microbiol., 2001. 67(12): p. 5715-5720.

107. Jimenez-Juarez, N., et al., Bacillus thuringiensis Cry1Ab mutants affecting oligomer formation are non-toxic to Manduca sexta larvae. J Biol Chem, 2007. 282(29): p. 21222-9.

108. Rausell, C., et al., Tryptophan spectroscopy studies and black lipid bilayer analysis indicate that the oligomeric structure of Cry1Ab toxin from Bacillus thuringiensis is the membrane-insertion intermediate. Biochemistry, 2004. 43(1): p. 166-74.

109. Padilla, C., et al., Role of tryptophan residues in toxicity of Cry1Ab toxin from Bacillus thuringiensis. Appl Environ Microbiol, 2006. 72(1): p. 901-7.

110. Gómez, I., et al., Cadherin-like receptor binding facilitates proteolytic cleavage of helix -1 in domain I and oligomer pre-pore formation of Bacillus thuringiensis Cry1Ab toxin. FEBS Lett., 2002. 25809: p. 1-5.

111. Gómez, I., et al., Cadherin-like receptor binding facilitates proteolytic cleavage of helix -1 in domain I and oligomer pre-pore formation of Bacillus thuringiensis Cry1Ab toxin. FEBS Lett., 2002. 513(2-3): p. 242 -6.

112. Rausell, C., et al., Unfolding events in the water-soluble monomeric Cry1Ab toxin during transition to oligomeric pre-pore and membrane-inserted pore channel. J Biol Chem, 2004b. 279(53): p. 55168-75.

113. Nair, M.S., X.S. Liu, and D.H. Dean, Membrane Insertion of the Bacillus thuringiensis Cry1Ab Toxin: Single Mutation in Domain II Blocks Partitioning of the Toxin into the Brush Border Membrane. Biochemistry, 2008. 47(21): p. 5814 - 5822.

114. Nair, M.S. and D.H. Dean, All three domains of Cry1A Toxins Insert into Insect Brush Border Membranes. Journal of Biological Chemistry, 2008.Epub July 17.

115. Chen, J., et al., Synergism of Bacillus thuringiensis toxins by a fragment of a toxin-binding cadherin. Proc Natl Acad Sci U S A, 2007. 104(35): p. 13901-6.

116. Gomez, I., et al., Specific epitopes of domains II and III of Bacillus thuringiensis Cry1Ab toxin involved in the sequential interaction with cadherin and aminopeptidase-N receptors in Manduca sexta. J Biol Chem, 2006. 281(45): p. 34032-9.

117. Soberon, M., et al., Engineering modified Bt toxins to counter insect resistance. Science, 2007. 318(5856): p. 1640-2.

118. Jurat-Fuentes, J.L. and M.J. Adang, Cry toxin mode of action in susceptible and resistant Heliothis virescens larvae. J Invertebr Pathol, 2006. 92(3): p. 166-71.

126

119. Oh, K.J., et al., Organization of diphtheria toxin T domain in bilayers: a site-directed spin labeling study. Science, 1996. 273(5276): p. 810-2.

120. Zhan, H.J., et al., Interaction of the isolated transmembrane domain of diphtheria toxin with membranes. Biochemistry, 1995. 34: p. 4856-4863.

121. Altenbach, C., et al., A collision gradient method to determine the immersion depth of nitroxides in lipid bilayers: Application to spin-labeled mutants of bacteriorhodopsin. Proc. Nat'l. Acad. Sci. U.S.A., 1994. 91: p. 1667-1671.

122. Hubbell, W.L., and and C. Altenbach, Site-directed spin labeling of membrane proteins, in Membrane Protein Structure: Experimental Approaches., W.S. H., Editor. 1994, Oxford University Press: London. p. 224-248.

123. Faham, S. and J.U. Bowie, Bicelle crystallization: a new method for crystallizing membrane proteins yields a monomeric bacteriorhodopsin structure. J Mol Biol, 2002. 316(1): p. 1-6.

127