University of New Hampshire University of New Hampshire Scholars' Repository

Doctoral Dissertations Student Scholarship

Fall 2003

Investigations into the digestive system proteases of Argiope aurantia (Araneae: Araneidae), hemolytic agents of venoms, and the role of both in necrotic arachnidism

Matthew Joseph Foradori University of New Hampshire, Durham

Follow this and additional works at: https://scholars.unh.edu/dissertation

Recommended Citation Foradori, Matthew Joseph, "Investigations into the digestive system proteases of Argiope aurantia (Araneae: Araneidae), hemolytic agents of spider venoms, and the role of both in necrotic arachnidism" (2003). Doctoral Dissertations. 178. https://scholars.unh.edu/dissertation/178

This Dissertation is brought to you for free and open access by the Student Scholarship at University of New Hampshire Scholars' Repository. It has been accepted for inclusion in Doctoral Dissertations by an authorized administrator of University of New Hampshire Scholars' Repository. For more information, please contact [email protected]. INVESTIGATIONS INTO THE DIGESTIVE SYSTEM PROTEASES OF

ARGIOPE AURANT1A (ARANEAE: ARANEIDAE), HEMOLYTIC AGENTS OF

SPIDER VENOMS, AND THE ROLE OF BOTH IN NECROTIC ARACHNIDISM

BY

MATTHEW JOSEPH FORADORI

BS, Indiana University of Pennsylvania, 1996

MS, University of New Hampshire, 1999

DISSERTATION

Submitted to the University of New Hampshire

in Partial Fulfillment of

the Requirements for the Degree of

Doctor of Philosophy

In

Zoology

September, 2003

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UMI Number: 3097784

UMI

UMI Microform 3097784 Copyright 2003 by ProQuest Information and Learning Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code.

ProQuest Information and Learning Company 300 North Zeeb Road P.O. Box 1346 Ann Arbor, Ml 48106-1346

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. This dissertation has been examined and approved.

Dissertation Director, Edward K. Tillinghast, Professor of Zoology, Emeritus

u , \ o4a)i?uA... Ji. S\^ahJbC% Larry G. Harris) Professor of Zoology

7 'flArt \j Marian K. Litvaitis, Associate Professor of Zoology

Myung-Jin Moon, Professor of Biological Sciences (Dankook University, South Korea)

Wells, DVM, MS, DACVP -'Senior Veterinary Pathologist

Z\ 2.063 Date

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DEDICATION

For my parents, Alexander and Janice Foradori, who encouraged me to follow

dreams.

For my wife, Renee, without whose understanding and love I could not have

finished.

111

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ACKNOWLEDGEMENTS

I truly believe that I would not have gotten this far without the guidance of Dr.

Robert S. Prezant at the Indiana University of Pennsylvania. I had floundered at

Duquesne University, leaving with a bruised ego and battered self esteem. I had turned

things around at IUP. The real problem was, I was unsure of where my career was

heading, and Dr. Prezant advised me, “If you don’t apply anywhere, you wont have any

options.” It was a difficult time in my life. My father had just been diagnosed with Non-

Hodgkin's Lymphoma and fought insurmountable odds to beat it. I was graduating soon,

and I really didn’t want to leave home and my family. My mother and father, of.^ourse,

told me to go and do what I had to do to continue my education. I would like to thank

Dr. Prezant for his insistence that I get some graduate school applications sent out.

I really had no idea where I wanted to go school. I new that I would like the

northeast. I had been to Boston, Massachusetts and Bath, Maine and I really like the

area. The problem was, I couldn’t find a potential advisor that worked with . It

wasn’t until Renee showed me a guy from the University of New Hampshire who looked

like a priest in the picture. I hadn’t thought much about New Hampshire before, let alone

going to the university there. If Renee had not shown me that graduate catalog from the

Zoology Department, my life would have most certainly been different.

I wrote Dr. Tillinghast a letter, and he responded quickly, sending me reprints and

asking me how soon I could come and visit. I quickly applied to work with him. Renee

and I decided to use our spring break and visit some of the schools we applied to, and

iv

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. along the way, we came to UNH. It was picturesque; a beautiful campus covered in

snow. Shortly after we returned home I found out I was accepted into the Zoology

graduate program.

Dr. Tillinghast took me, a person with no molecular lab skills, and turned me into

an electrophoresis machine. My master’s took three and a half years, but it was good

work. We answered quite a few questions about the egg case of Argiope aurantia in

those years; it is work I am truly proud of. I decided to say for my Ph. D. after he asked

me to. We had gotten into some excited new research at the end of my master’s degree

that I wanted to stay and work on. I would like to thank Dr. T for believing in me and my

work. I would also like to thank him for taking me on those two week SDF power-

collecting trips. I have seen parts of the southeastern United States I would not have seen

otherwise.

While here at the University of New Hampshire, I have made many friendships,

but there are a few in particular I would like to acknowledge. Dr. Patrick Danley and I

have been friends since the first day we met during the TA training sessions, back in

1996. Pat is one of the most sincere and sarcastic people I know. A true friend, even

though he moved away, we still keep in touch and see each other from time to time. Dr.

James Sulikowski is another great person. James is one of the most driven people I

know. His athletic ability is amazing to me, and yet he truly breaks the “dumb jock”

stereotype. Dr. Rick Hochberg rounds out my close group of colleagues. He is one of

the few people I have ever met that can make me laugh all the time. These three men

have made this experience worthwhile for me.

v

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. There are a few other people I would like to mention. Lauren Keil, one of the

most meticulous lab workers I have ever met taught me how to be organized. Max Diem,

one of the most enthusiastic people I have ever met was always ready for a day in the

field with his “right on!” attitude. Last but not least, in one of the bumps I experienced

during my dissertation experience, Marian Litvaitis, told me to stick it out. I was so close

to being done that I could taste it. I’m glad I took her advice.

Thank you all.

vi

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TABLE OF CONTENTS

DEDICATION ...... iii

ACKNOWLEDGEMENTS ...... iv

LIST OF TABLES...... ix

LIST OF FIGURES ...... ;...... x

ABSTRACT...... xii

CHAPTER PAGE

INTRODUCTION...... 1

I. AN EXAMINATION OF THE POTENTIAL ROLE OF SPIDER DIGESTIVE PROTEASES AS A CAUSATIVE FACTOR IN NECROSIS ...... 7

Abstract...... 7

Introduction ...... 7

Materials and Methods...... 9

Results ...... 14

Discussion ...... 22

II. SURVEY FOR POTENTIALLY NECROTIZING SPIDER VENOMS, WITH SPECIAL EMPHASIS ON MILDEI...... 27

Abstract ...... ,.27

Introduction ...... 28

Materials and Methods ...... 29

vii

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Results ...... 33

Discussion ...... ,...... 42

HI. FRACTIONATION OF SPIDER DIGESTIVE FLUID AND PARTIAL CHARACTERIZATION OF THE PROTEASES AND PROTEASE INHIBITORS...... 47

Abstract...... 47

Introduction ...... 47

Materials and Methods...... 49

Results ...... 54

Discussion ...... 65

LIST OF REFERENCES ...... 71

APPENDIX...... 80

viii

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF TABLES

TABLE PAGE

2.1. Survey of the venom of selected spider species for their ability to cause hemolysis...... 34

3.1. N-terminal amino acid sequence of isolated protease from the digestive fluid of Argiope aurantia ...... 63

ix

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST GF FIGURES

FIGURE PAGE

1.1. Proteases in the digestive fluid (SDF) of A. aurantia using casein as a substrate. ...15

1.2. Effect of A. aurantia digestive fluid on (a) Type III collagen and (b) gelatin ...... 16

1.3. Effect of A. aurantia digestive fluid on (a) fibrinogen and (b) fibronectin ...... 17

1.4. Effect of A. aurantia digestive fluid on (a) fibrin and (b) elastin...... 18

1.5. The effect of EDTA on A. aurantia digestive fluid (SDF) proteases ...... 19

1.6. Inhibition of (a) A. aurantia digestive fluid (SDF) proteases and (b) L. reclusa venom proteases by A. aurantia hemolymph as shown by diffusion in agarose/ casein gels...... 20

1.7. Comparison of the effects of A. aurantia digestive fluid and L. reclusa venom on the skin of the rabbit 18 hours post-injection...... 21

1.8. A histological comparison of the effects of intradermal injection ofArgiope aurantia digestive fluid (II a and b; from site II, Fig. 7) and Loxosceles reclusa venom gland homogenate (III a and b; from site III, Fig. 7), 72 hours post­ injection ...... 23

2.1. The effects of spider venom gland homogenates (VGH) on RBC ghosts...... 37

2.2. The effect of venom gland homogenate (VGH) on individual purified phospholipids...... 38

2.3. The effect of Cheiracanthium mildei venom gland homogenate (VGH) on phosphatidylcholine (/PC) fluorescently labeled on the sn-l (D-3771) or sn-2 (D-3803) position...... •...... 39

2.4. The effect of Cheiracanthium mildei venom gland homogenate (VGH) on the skin of New Zealand white rabbits ...... 40

2.5. Histology of the skin of New Zealand white rabbit 48 hours after intradermal injection of 79pg protein Cheiracanthium mildei venom gland homogenate.*...... 41

x

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2.6. Calcium dependency of Cheiracanthium mildei venom gland homogenate hemolytic factor...... 43

3.1. Fractionation of SDF by anion exchange chromatography ...... 55

3.2. Fractionation of fractions 1-10 on Sephacryl S-100 HR column...... 57

3.3. Fractionation of fractions 48-54 on cation exchange chromatography...... 58

3.4. Fractionation of fractions 26-32 on a Sephadex G-75 column ...... 59

3.5. Separation of SDF protease inhibitors by natural PAGE ...... 60

3.6. Separation of SDF protease inhibitors by SDS-PAGE ...... 62

3.7. Comparison of spider and crayfish digestive proteases...... 64

3.8. Fractionation of SDF by ultracentrifugation ...... 66

3.9. Whole SDF fractionation by sucrose density gradient ultracentrifugation examined by (A) zymography and (B) SDS-PAGE...... 67

xi

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ABSTRACT

INVESTIGATIONS INTO THE DIGESTIVE SYSTEM PROTEASES OF

ARGIOPE AURANTIA (ARANEAE: ARANEIDAE), HEMOLYTIC AGENTS OF

SPIDER VENOMS, AND THE ROLE OF BOTH IN NECROTIC ARACHNIDISM

By

Matthew J. Foradori

University of New Hampshire, September, 2003

Necrotic arachnidism, or tissue necrosis following spider bites is a widespread

problem. Knowledge of the exact mechanism that produces such lesions is still

incomplete. Here I have examined spider digestive fluid proteases and spider venom

enzymes (e.g., sphingomyelinase D and phospholipase A2) for their potential to cause

necrotic lesions, as both are thought to be primary agents.

The digestive fluid of Argiope aurantia was examined for its ability to cleave a

variety of extra-cellular matrix proteins, including collagen, elastin, and fibronectin.

Having confirmed that the fluid has collagenases, the digestive fluid was injected into the

skin of rabbits to observe whether it would cause necrotic lesions. It did not. The data do

not support the suggestions that spider digestive collagenases have a primary role in

spider bite necrosis.

In an effort to design a simple, inexpensive screening method for identifying

with necrotizing venoms, venom gland homogenates of a variety of spider species

were examined for their ability to cause hemolysis. Those venoms that were positive

xii

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. were further examined for the presence of sphingomyelinase D, and the ability to evoke

necrotic lesions in the skin of rabbits. Of 45 species examined, only the venom of

Loxosceles reclusa and Cheiracanthium mildei caused hemolysis. Unlike L. reclusa

venom, however, C. mildei venom did not possess sphingomyelinase D nor did it cause

necrotic lesions in the skin of rabbits. In an effort to determine the hemolytic agent in C.

mildei venom, I found evidence for a phospholipase A2.

Through a combination of ion exchange chromatography, size exclusion

chromatography, ultracentrifugation, and electrophoresis, an attempt was made to isolate

several digestive fluid proteases and serine protease inhibitors in milligram quantities

from the digestive fluid of Argiope aurantia. The proteases, however, eluted

anomalously during the size exclusion chromatography and ultracentrifugation trials. I

have hypothesized that spider proteases can undergo a reversible aggregation to account

for this behavior. I was able to isolate several proteases in quantities sufficient for N-

terminal amino acid sequences. These data revealed that spider proteases are related to

astacins and meprins, which are in the astacin family of metalloproteases.

xiii

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. INTRODUCTION

Digestion is initiated externally in spiders. They are incapable of ingesting solid

food as they lack a chewing apparatus and have only a slender pharynx connecting with

the stomach. Moreover, bristles bordering the mouth opening serve as a sieve to prevent

particles in excess of 1 pm from clogging the opening. Their food must therefore be

liquefied. Typically, spiders envenomate prey and begin to crush it in their chelicerae.

Spiders then regurgitate digestive fluid from the intestinal tract onto the prey, wait a few

moments, then suck the solubilized products into their gut using the muscular pharynx

and sucking stomach (Foelix, 1996). The process is repeated until the prey is reduce to a

small chitinous mass.

Although there has been longstanding interest in the enzymatic content of spider

digestive fluid (SDF), research has been sporadic. One of the first demonstrations of

proteolytic activity from spiders was that of Plateau (1877). Using abdominal caeca

extracts from European araneomorph spiders, he discovered proteolytic, amylytic, and

lipolytic activities. Betkau (1885) confirmed the presence of proteolytic and other

enzyme activity. A quantitative investigation of (Avicularia) digestive enzymes

by Schlottke (1936) revealed a potent proteinase with caseinolytic ability, as well as

moderate amounts of peptidase and amylase activity. Pickford (1942) found fibrinase,

gelatinase (pH 1.5-12), amylase, phosphatases, lipase (pH 6-11) and proteinases in the

digestive fluids of Araneus trifolium, Araneus marmoreus, and Larinioides comutus.

Poliak (1966) and Houghton et al. (1967) found evidence of esterase activity in several

1

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. species of spider homogenates and histological sections, respectively. Couch and Benton

(1971) found general esterase and phosphatase activity in two species of spiders.

Tillinghast and Kavanagh (1977) confirmed the presence of alkaline proteases,

collagenase, and elastase activity in the digestive fluid of Argiope aurantia. They were

also the first to demonstrate spider web digestion in vitro. Mommsen’s (1979a, b, c)

systematic studies of the digestive enzymes of Tegenaria atrica is one of the most

complete inventories of spider digestive fluid to date. He found a single protease that

could hydrolyze native as well as denatured proteins, an amidase, and chymotrypsin-like

activity (Mommsen, 1979a), a chitinase and several a-amylases (Mommsen, 1979b), as

well as alkaline and acid phosphatases (Mommsen, 1979c). A later study (Mommsen,

1980) revealed the presence of chitinase and {3-N-acetylglucosaminidase in Cupiennius

salei. Rekow et al. (1983) found protease and lipase activity in the abdominal

homogenates of Loxosceles reclusa, but lacking in cephalothorax homogenates. Herrero

and Odell (1988) examined the digestive secretions of two Costa Rican and

determined that they had high proteolytic activity. More recently, Joo et al. (2002) found

a prothrombin-activating protease in the isolated digestive tract homogenate ofNephila

clavata.

There are also numerous reports of proteases in spider venoms (see review by

Perret, 1977). However, Perret observed that proteases were absent from those venoms

which had been collected by methods which took precautions to avoid contamination by

SDF. Mebs (1970, 1972) for example, reported protease in tarantula (Pamphobetus

roseus) venom that he received from the Instituto Butantan (Sao Paolo, Brazil). He

isolated three proteases of approximately 11 kDaltons, that had pH optima of 8.5 and

2

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. cleaved casein, but did not cleave trypsin substrates, and were equally unaffected by

serine protease inhibitors. These properties are identical to those later reported for

proteases from SDF (Tillinghast and Kavanagh, 1977; Tugmon and Tillinghast, 1995). It

seems likely that the venom was contaminated by the spider’s digestive fluid.

The work I conducted for my master’s thesis examined the mechanism by which

Argiope aurantia spiderlings exit the egg case. The egg case is constructed of a tightly

woven outer cover, composed primarily of large-diameter cylindrical gland fibers and

small-diameter fibers, likely of aciniform gland origin; It must be sufficiently robust to

resist degradation for the many months between its formation in early fall and the

emergence of the spiderlings in late Spring. I demonstrated that spiderlings escape from

the egg case by a combination of enzymatic digestion and mastication with their

chelicerae to form a communal hole in the outer cover. I recorded the spiderlings

chewing the edges of the hole as they exited the cocoon using video equipment.

Electrophoretic analysis revealed that the proteases of the spiderlings were identical to

those found in washes of the edges of the exit holes (Foradori et al., 2002).

In the first chapter of this dissertation, I examined the possible involvement of

SDF proteases in the formation of lesions following spider bites. Atkinson and Wright

(1992) observed that gut extracts of spiders, including those of mygalomorphs and

araneomorphs, produced cellular disruption of mouse skin in cell culture. From their

research, they concluded that it was digestive proteases released at the area of the bite

rather than the constituents of the venom that were the cause of necrotic lesions. As gut

extracts contain a variety of tissues as well as digestive enzymes, I have, examined the

hypothesis of Atkinson and Wright (1992) using SDF alone. First, SDF was examined

3

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. for its ability to cleave a variety of connective tissue proteins. Substrates including

casein and gelatin, and more importantly the basement membrane proteins fibrin,

fibrinogen, fibronectin, elastin, collagen were all cleaved by the proteolytic enzymes.

SDF was then injected into the skin of rabbits under the assumption that it would cause

necrotic lesions, but it evoked no response. By comparison, 3 pg of brown recluse

venom (Loxosceles reclusa) resulted in severe lesions. Thus, it does not appear that SDF

has a primary role in necrotic arachnidism (Foradori et al., 2001).

The venom of L. reclusa contains sphingomyelinase D, an enzyme that has been shown

to be responsible for the necrotic lesions which often follow this spider’s bite

(Kurpiewski, 1981).

In the second chapter of this dissertation, I attempted toidentify other spider

species that might cause necrotic lesions by analyzing venoms for sphingomyelinase D

activity. Sphingomyelinase D is known to lyse red blood cells (RBQ in addition to

causing necrotic lesions (Forrester et al., 1978). To identify other spiders that might

cause such lesions, I assayed homogenates of venom glands (VGH) for their ability to

cause RBC lysis (hemolysis). Venom gland homogenates of more than 260 spiders

representing 55 spider species from 18 families were examined. The VGH ofL. reclusa

always causes hemolysis, as reported by Forrester et al. (1979), and was therefore, used

as a positive control. Of the remaining 55 species, onlyCheiracanthium mildei VGH

caused hemolysis, however, further research revealed that it did not contain

sphingomyelinase D. To identify the agent causing RBC lysis, I subjected RBC ghosts to

C. mildei VGH. Chromatographic analysis of the products revealed that phospholipids

(glycerophospholipids) had been cleaved. I then evaluated the VGH for its ability to

4

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. cleave a variety of natural and synthetic phospholipids. These studies demonstrated the

presence of a phospholipase A2.

Cheiracanthium mildei is a common spider on the University of New Hampshire

campus, and the question arose as to whether its venom would cause necrotic lesions. A ./ review of the literature revealed that C. mildei venom caused necrotic lesions in some

laboratory (Spielman and Levi, 1970; Newlands et al., 1980; Hughes, 1981).

However, when I injected VGH of C. mildei into rabbits, no lesions resulted. Most

reports suggest that C. mildei venom produces only a mild effect in humans (Krinsky,

1987; Minton, 1972). Which was consistent with my own observations. Even Spielman

and Levi (1970), who reported necrosis in following a C. mildei bite, reported nothing

more than mild redness and swelling in humans.

The ability of the digestive enzymes of Argiope aurantia to cleave connective

tissue proteins suggests they would be valuable tools for medical research and a possible

therapeutic use. Previous efforts to isolate spider proteases, however, have resulted in

very low yield (Mebs, 1972; Kavanagh and Tillinghast, 1983). In the third chapter of this

dissertation I attempted to develop a more effective protocol for the purification of the

SDF proteases.

Digestive fluid was fractionated by a combination of ion-exchange

chromatography, gel filtration chromatography, and sucrose density gradient

ultracentrifugation. In none of theses investigations was I able to isolate an individual

protease excepting pl2 (a protein with a mass of 12 kDaltons). Protease 12 (pl2),

however, was found in all other fractions as well. These results led me to propose that

the spider digestive proteases are reversibly associated into higher aggregates. This

5

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. assumption explains the anomalous elution times that I and Mebs (1972) observed with

gel filtration, and explains why earlier efforts to obtain individual proteases resulted in

low yields of a single protease (p i2).

6

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 1

AN EXAMINATION OF THE POTENTIAL ROLE OF SPIDER DIGESTIVE

PROTEASES AS A CAUSATIVE FACTOR IN SPIDER BITE NECROSIS1

ABSTRACT

Tissue necrosis following spider bites is a widespread problem. In the continental

United States the brown recluse {Loxosceles reclusa), hobo spider {Tegenaria agrestis),

garden spider {Argiope aurantia), and Cheiracanthium species among others reportedly

cause such lesions. The exact mechanism producing such lesions is controversial. There

is evidence for both venom sphingomyelinase and spider digestive collagenases.

I have examined the role of spider digestive proteases in spider bite necrosis. The

digestive fluid of A. aurantia was assayed for its ability to cleave a variety of connective

tissue proteins, including collagen. Having confirmed that the fluid has collagenases, the

digestive fluid was injected into the skin of rabbits to observe whether it would cause

necrotic lesions. It did not. The data do not support the suggestions that spider digestive

collagenases have a primary role in spider bite necrosis.

1.1. Introduction

Tissue necrosis resulting from spider bites is a widespread problem (Bettini and Brignoli,

1978). Whereas Loxosceles species are the best known spider producing

1. Foradori, M.J., Keil, L.M., Wells, R.E., Diem, M., Tillinghast, E.K., 2001. An examination of the potential role of spider digestive proteases as a causative factor in spider bite necrosis. Comp. Biochem. Physiol. 130 C: 209-18.

7

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. necrotic lesions, other spider species are thought to do so as well. Thus Vest (1987) has

reported that the bite of Tegenaria agrestis (hobo spider) in the Pacific Northwest is

responsible for necrotic lesions, and Spielman and Levi (1970) have reported that the bite

of Cheiracanthium inclusum is also capable of producing such lesions. Russell and

Gertsch (1983) have compiled a list of other spider species suspected of causing necrotic

arachnidism.

At this time there is a lack of agreement as to the exact mechanisms that cause

tissue necrosis (Futrell, 1992). Among the existing theories, Forrester et al. (1978) and

Kurpiewski et al. (1981) have proposed that the venom enzyme sphingomyelinase D is

the responsible agent. Atkinson and Wright (1992) observed that gut extracts of 13

different spider species, including mygalomorphs and araneomorphs, produced cellular

disruption of mouse skin in cell culture. They also observed high collagenase activity in

the gut extracts of these spiders. They concluded that the digestive system collagenases,

rather than any venom component, were responsible for the development of tissue

necrosis. As gut homogenates very likely contain a variety of enzymes and

pharmacologically active compounds, I have tested their hypothesis using spider

digestive fluid.

Argiope aurantia was chosen to test the hypothesis of Atkinson and Wright

(1991, 1992) because preliminary research from the laboratory detected collagenases in

the digestive fluid of this species. In addition, the digestive fluid is easily collected from

A. aurantia in large quantity (an average of 100 pi per spider) and thus I was able to

examine the fluid for its ability to cleave a variety of connective tissue proteins.

Moreover, A. aurantia hemolymph contains inhibitors that effectively block digestive

8

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. fluid proteases (Tugmon and Tillinghast, 1995). This provided me with a tool to block

spider digestive proteases and thereby test the hypothesis that the digestive proteases

have a role in spider bite necrosis. Had the proteases caused necrosis, hemolymph

inhibitors would have provided insight into the mechanisms that cause necrosis as well as

serve as a potential antidote.

My results confirm the observation of Atkinson and Wright (1992) that the gut of

spiders contains collagenases. In addition, spider digestive fluid had the ability to cleave

a variety of other connective tissue proteins such as elastin and fibrinogen. In spite of the

presence of these proteases, intradermal injections of this fluid into rabbits did not result

in necrosis. My results do not support the hypothesis that the spider digestive proteases

have a primary role in spider bite necrosis.

1.2. Materials and Methods

1.2.1. Collection of spider fluids

A. aurantia were collected in Maine, New Hampshire, North Carolina, South

Carolina, Georgia and Virginia during the summers of 1998/99. Twenty microliter

micropipettes were used to collect spider digestive fluid (henceforth referred to as SDF),

which spiders regurgitated freely when handled. Approximately 4 ml was collected from

100 spiders. SDF was pooled, placed on ice, and frozen within 24 hours of its collection.

Hemolymph was collected fromA. aurantia narcotized by carbon dioxide. The tip of the

fourth left leg was cut off with surgical scissors and the exuding hemolymph was

collected with a micropipette. Hemolymph was pooled and stored at -20° C.

9

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1.2.2. Zymography

To visualize the major proteases in the digestive fluid of A. aurantia as well as

establish their apparent molecular weights. SDF from three A. aurantia spiders (0.5 pi

quantities) were examined separately for caseinolytic activity. The samples where

electrophoresed on 4-16% Zymogram (blue casein) gels (Invitrogen, EC# 6415) under

non-reducing conditions. Molecular weight standards ran concurrently were from Gibco-

BRL (Benchmark prestained protein ladder #10748-010). Following the electrophoretic

separation, gels were renatured and developed according to the manufacturer's

recommendations.

1.2.3. Degradation o f extracellular matrix proteins

The degradation of collagen was visualized by the method of Phillips and Dresden

(1973). Three milligrams of collagen (Sigma, C-3511, type in calf skin) was solubilized

in 1.5 ml 0.1 mM acetic acid for two days at 15° C then dialyzed overnight against 0.01M

Tris, pH 7.8 containing 0.15 M NaCl and 0.5 mM CaCl2. Ten microliters SDF was then

added, mixed, and incubated at 20° C. Thirty microliter aliquots were removed at 0, 0.5,

2, 4, and 23 hours. These were mixed immediately with an equal volume of sample

buffer, and boiled to stop the reaction. Controls lacking SDF were incubated

concurrently for 23 hours. The samples were electrophoresed (SDS-PAGE) on 4-20%

Tris Glycine gels (Invitrogen, EC6028) under reducing conditions with constant 1,25 V

and stained with a 0.02% Coomassie R-250,40% Methanol (v/v), and 1% Acetic acid

(v/v) solution overnight to observe degradation products. The degradation of gelatin was

performed in an identical manner except that the collagen (Sigma, C-3511, calf skin) was

heated to 70° C for 15 minutes and cooled to 20° C prior to the addition of SDF.

10

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The effect of SDF on extracellular matrix proteins was visualized by the method

of Feitosa et al. (1998). Two microliters of SDF (89 pg protein) was added to 0.5 mg of

fibrinogen (Sigma, F-8630) or fibronectin (Sigma, F-4759) suspended in 200 pi 25 mM

Tris buffer, pH 8.1, containing 5 mM CaCl2 (henceforth referred to as buffer A) at 37° C.

Thirty microliter aliquots of the reaction mixtures were removed at 0 and 30 min, 1, 2, 4,

8, and 23 h and frozen promptly. Controls lacking SDF were incubated for 23 h. The

fibrinogen and fibronectin samples were electrophoresed (SDS-PAGE) on 10% Tris-

Glycine gels (Invitrogen, EC# 6078) under non-reducing and reducing conditions,

respectively, with constant 125 V and stained with a 0.02% Coomassie Blue R-250, 25%

2-propanol, and 10% acetic acid solution to observe degradation products.

The degradation of elastin was examined by the radial-diffusion method of

Schumacher, and Schill (1972). A gel was prepared by adding 100 mg agarose to 6 ml

buffer A and boiling for 5 min. Upon cooling to 60°C, 50 mg of elastin (Sigma, E-1625)

suspended in 4 ml buffer A was added and poured into a diffusion plate. When the gel

solidified at room temperature, a 6 mm diameter well was placed in the center. A 20 pi

sample containing 2 pi SDF (89 pg protein) and 18 pi buffer A was added to the well.

Fibrinolytic activity was observed by the method of Yang and Ru (1997). A fibrin plate

was prepared by adding 75 mg agarose to 10 ml 0.02M Tris buffer, pH 8.0, containing 85

mg NaCl and boiling for 5 min. When the temperature declined to 58°C, 3 ml of a

fibrinogen solution (150 mg Sigma, F-8630 in 3 ml of the above buffer) and thrombin

(100 pi containing 200 units, Sigma T-4648) were added, and the mixture was poured

into a diffusion plate and allowed to solidify for 10 min. Samples of 2 pi SDF (89 pg

protein) made to 20 pi with buffer A were placed in the wells. The plates were incubated

11

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. overnight and the diameter of the ring of digestion of elastin and fibrin was measured in mm.

All protein concentrations were estimated by the method of Lowry et al. (1951)

using bovine serum albumin as a standard.

1.2.4. The effect of EDTA on spider proteases.

To observe whether SDF proteases are metalloproteases, 100 pi SDF (4.45 mg

protein) was made to 3.0 ml with 25 mM Tris buffer, pH 8.1, containing 20 mM EDTA

and set at room temperature for 1 h. SDF was then dialyzed against the same buffer

without EDTA for an hour. 100 pi of SDF (4.45 mg protein) made to 3.0 ml with 25 mM

Tris, pH 8.1, and containing 2 mM CaCl2 served as the control. One hundred microliter

aliquots of EDTA-treated enzyme were made to 1 ml with increasing quantities of 2.0

mM ZnS04 and incubated for 30 min at room temperature. One ml of 1% Hammersten

casein (Schwarz/Mann) was then added and all tubes were incubated for 15 min at 37°C.

The reactions were stopped by the addition of 3 ml 10% TCA. After 30 min the samples

were filtered through Whatman No.l filter paper and the optical density of the eluant was

read at 275 nm (Kunitz, 1947).

1.2.5. Inhibition of spider digestive proteases by spider hemolymph

To determine the quantity of hemolymph necessary to inhibit the SDF proteases,

90 mg of agarose and 90 mg Carnation dry milk were added to 9 ml of buffer A. The

mixture was boiled for 5 min and poured into a diffusion plate. Wells were cut when the

gel solidified. Two microliters of SDF (89 pg protein) mixed with 18 pi buffer A was

added to one well and 2 pi SDF (89 pg protein) mixed with 18 piA. aurantia

12

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. hemolymph was added to a second well. The plate was incubated overnight at room

temperature. .

1.2.6. The effect o f spider digestive fluid on the skin o f rabbits

To observe if SDF from A. aurantia is capable of causing necrotic lesions samples

were injected into the skin of rabbits. Because it is the best documented example of

spider-induced necrotic lesions on rabbit skin (Atkins et al., 1958; Smith and Micks,

1968; Geren et al., 1976; Barbara et al., 1992, 1996; Gomez et al., 1999; Elston et al.,

2000), the venom of L. reclusa served as a positive control. A. aurantia venom has not

been examined in the skin of rabbits. Three adult female New Zealand white rabbits

were injected in a shaved area on the back with the following solutions: site I: 200 pi

PBS, pH 7.4; site II: SDF (20 pg protein) in 200 pi PBS; site III:L. reclusa venom gland

homogenate (10 pg protein) in 200 pi PBS; site IV:A. aurantia hemolymph (100 pg

protein) in 200 pi PBS; site V: SDF (20 pg protein) +A. aurantia hemolymph (100 pg

protein) in 200 pi PBS; and site VI:L. reclusa venom gland homogenate (10 pg protein)

+ A. aurantia hemolymph (100 pg protein) in 200 pi PBS. All preparations were filtered

through a 0.2 micron filter prior to injection. The injections were administered

intradermally with a 26 gauge needle approximately 6-8 cm apart. The rabbits were

observed hourly for the first six hours and then every six hours for 72 hours. The lesions

were photographed with a digital camera and scored according to the intracutaneous test

described in the USP 24-NF19 (2000). The experiments were carried out in compliance

with the Public Health Service Policy, the Guide for the Care and Use of Laboratory

Animals (1996), and University Policy (see Appendix).

13

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The rabbits were euthanized 72 hours after injection. Tissues samples including

skin and subcutis were taken from all injection sites and fixed for 48 hours in 10% neutral

buffered formalin, dehydrated in an ethanol series, cleared in Histoclear II (National

Diagnostics), then embedded in Paraplast Plus (Fisher Scientific) using a Citadel 2000

(Shandon) carousel style automatic tissue processor. Sections were cut at 6 pm, mounted

on glass slides, and stained with Harris’ hematoxylin and eosin (Luna, 1979) and

examined under light microscopy.

1.3. Results

There are at least four major proteases in the SDF. They have apparent molecular

weights of 120, 38, 30, and 12 kDa when observed by zymography (Fig. 1.1). In addition

to their ability to cleave casein, the SDF proteases are capable of cleaving native collagen

and gelatin (Fig. 1.2), fibrinogen and fibronectin (Fig. 1.3), as well as elastin and fibrin

(Fig. 1.4). The exposure of 100 pi SDF to 20 mM EDTA for 1 h reduced the proteolytic

activity to about 35% of that of the control. The activity was largely restored by

exposure to 1.4 mM ZnS04, above which the activity was reduced (Fig. 1.5).

The spider proteases in 2 pi SDF were effectively blocked by 18 pi hemolymph

(Fig. 1.6a). The hemolymph of A. aurantia also inhibited proteases from the venom

gland homogenates of L. reclusa (Fig. 1.6b).

The effect of SDF on New Zealand white rabbit skin was examined by

intradermal injection. The effectiveness of inhibition by spider hemolymph was observed

by mixing a pre-established volume of hemolymph with each of the fluids immediately

prior to injection. The injection of SDF (20 pg protein) produced a well defined oval

area of redness approximately 1.0 x 0.7 cm in diameter and 1.0 mm in height (Fig. 1.7,

J 14

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 1.1. Proteases in the digestive fluid (SDF) of A. aurantia using casein as a substrate.

Spider digestive fluid (0.5 pi) from 3 different Argiope aurantia were electrophoresed on 4-16% zymogram (blue casein) gels under non-reducing conditions. The gels were renatured, developed, then stained according to the manufacturer's recommendations.

15

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ;.-iA («&$*» la# M ikm?’h i® ;||i® lfP|; J i t UpP Jfif§! lo u if' <1 2417 l'J ? iy t

1} :•

■«m, .■■!'►. :

1 2 3-4 5 67 MPa

im# m 4 m s # u 4fJ 3fi4 347 ifj .13.1 §j

Figure 1.2. Effect of A. aurantia digestive fluid on (a) Type III collagen and (b) gelatin

Nine microliters of spider digestive fluid was added to a solution containing 3 mg collagen or 4 pi of spider digestive fluid was added to a solution containing 1.2 mg of gelatin. The mixtures were incubated for 23 h at room temperature (20 ° C). Samples (20 gl) were removed at: 0 time (lane 2), 30 min (lane 3), 1 h (lane 4), 2 h (lane 5), 4 h (lane 6), and 23 h (lane 7). Lane 1 in each case was a control (con.) containing substrate and buffer only. The fibrinogen and fibronectin samples were electrophoresed on 4-20% SDS-PAGE gradient gels under reducing conditions. The gels were stained with Coomassie blue to visualize substrate fragments.

16

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. MDa 172.6 1IJW 79.6 61.3 f l H B I 49,0

con, 0 0.5 1

Figure 1.3. Effect of A. aurantia digestive fluid on (a) fibrinogen and (b) fibronectin.

Two microliters of spider digestive fluid was added to a solution containing 0.5 mg fibrinogen or fibronectin in 200 pi 25 mM Tris buffer, pH 8.1. The mixtures were incubated for 23 h at 37°C. Samples (20 pi) were removed at: 0 time (lane 2), 30 min (lane 3), 1 h (lane 4), 2 h (lane 5), 4 h (lane 6), 8 h (lane 7), and 23 h (lane 8). Lane 1 in each case was a control (con.) containing substrate and buffer only. The samples were added to an equal volume of sample buffer and electrophoresed on 10% SDS-PAGE gels under non-reducing and reducing conditions respectively. The gels were stained with Coomassie blue to visualize substrate fragments.

17

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 1.4. Effect of A. aurantia digestive fluid on (a) fibrin and (b) elastin.

Fibrin and elastin gels were prepared with Tris buffer, pH 8.1, as described in the text. Two microliters of A. aurantia digestive fluid made to 20 pi with the same buffer, then added to the wells. The plates were incubated overnight.

18

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ZiiKoaaaiiifiMcii. (mM)

Figure 1.5. The effect of EDTA on A. aurantia digestive fluid (SDF) proteases.

One hundred microliters SDF was made to 3 ml with 25 mM Tris buffer, pH 8.1, containing 20 mM EDTA incubated for 1 hour, then dialyzed to remove the chelator. One hundred microliter aliquots of treated SDF were incubated with increasing concentrations of ZnS04 then assayed for protease activity. The results were compared to the protease activity of spider digestive fluid in the above buffer, but lacking EDTA and containing 2 mM CaCl2.

19

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. SDF + Hemolymph

^ m m w i

/ aurantsa Hemolymph L teclusa VGH

Figure 1.6. Inhibition of (a) A. aurantia digestive fluid (SDF) proteases and (b) L. reclusa venom proteases by A. aurantia hemolymph as shown by diffusion in agarose/casein gels.

(a) Two microliters (89 pg protein)A. aurantia digestive fluid (SDF) mixed with 18 pi Tris buffer (left well), and 2 pi SDF mixed with 18 pi A. aurantia hemolymph (right well), (b) Twenty five microliters A. aurantia hemolymph (left well) and 10 piL. reclusa venom gland homogenate (VGH) made to 25 pi with PBS (right well). Note the inhibition on the left side of the right well of b.

20

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 1.7. Comparison of the effects of A. aurantia digestive fluid and L. reclusa venom on the skin of the rabbit 18 hours post-injection.

New Zealand White rabbits were injected intradermally at the numbered sites shown on the dorsum with the following solutions made to 200 pi with PBS. II. A. aurantia digestive fluid (20 pg protein), III. brown recluse venom (10 pg protein), V. A. aurantia digestive fluid (20 pg protein) plus A. aurantia hemolymph (100 pg protein), and VI. brown recluse venom (10 pg protein) plus A. aurantia hemolymph (100 pg protein). Sites I and IV were saline and hemolymph controls respectively, and are not shown.

21

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. site Et). It was most intense after 6 h, but stabilized thereafter, decreasing in color

intensity after 24 h. Mixing A. aurantia hemolymph (100 pg protein) with the SDF

almost completely negated these effects (Fig. 1.7, site Y).

Histological sections of skin and subcutis from site I (PBS) showed no observable

changes and served as a negative control. Although SDF is rich in connective tissue

degrading proteases, sections of tissue from site II (SDF) showed only minimal reaction

characterized by low numbers of heterophilic leukocytes (Fig. 1.8). Tissues from site IV

(A. aurantia hemolymph) showed little or no observable reaction and were comparable to

site I. There was negligible to minimal reaction in the tissues from site V, injected with a

mixture of A. aurantia hemolymph and SDF. The microscopic appearance of the

injection sites was consistent in all three rabbits.

1.4. Discussion

While a number of spider species have been listed as potentially causing necrotic

lesions (Russell and Gertsch, 1983), the agent responsible for these lesions remains a

subject of debate. There is strong evidence that the venom enzyme, sphingomyelinase

D, is the causative agent (Kurpiewski et al., 1981) in the bite of the L. reclusa. However,

Atkinson and Wright (1991, 1992) have presented evidence that spider digestive

collagenases rather than venom enzymes are the cause. I have tested their hypothesis

with the digestive fluid from A. aurantia.

The digestive fluid of A. aurantia contains casein cleaving proteases ranging in

molecular weights from 12 to 140 kDa. The most evident of these had apparent

molecular weights of 12, 30, 38, and 120kDa (Fig. 1.1). The proteases proved capable of

22

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 1.8. A histological comparison of the effects of intradermal injection of Argiope aurantia digestive fluid (II a and b; from site II, Fig. 1.7) and Loxosceles reclusa venom gland homogenate (III a and b; from site III, Fig. 1.7), 72 hours post-injection. Note edema and heavy infiltration of heterophils (HI a) and degeneration of connective tissue and panniculus muscle (arrow, site III b).

23

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. cleaving native collagen (Fig. 1.2), fibrinogen and fibronectin (Fig. 1.3), as well as

elastin and fibrin (Fig.4), Feitosa et al. (1998) have reported that the venom of L.

intermedia has proteolytic activity against gelatin, fibrinogen, and fibronectin, but is

unable to cleave types I and Et collagen. This is consistent with the observations of

Kaiser and Raab (1967) who reported that the venom of both Phoneutriafera and Lycosa

erythrognatha can readily digest gelatin, but neither had the ability to cleave native

collagen. By contrast, A. aurantia digestive proteases can not only cleave all of these

substrates, but native collagen as well (Fig. 1.2). The presence of eollagenases in spider

digestive fluid is in agreement with the report of collagenase activity in spider midgut

extracts (Atkinson and Wright, 1992).

Most of the spider digestive proteases appear to be metalloproteases. SDF

protease activity is greatly inhibited by EDTA, but not serine, cysteine or aspartate

protease inhibitors (Tillinghast and Kavanagh, 1977: Tugmon and Tillinghast, 1995).

The inhibition by EDTA is relieved by the addition of zinc ions (Fig. 1.5). Spider venom

proteases also appear to be metalloproteases. They are inhibited by EDTA (Jong et al.,

1979; Feitosa et al., 1998), and not by serine protease inhibitors (Mebs, 1972). Both

spider digestive proteases (Tillinghast and Kavanagh, 1977; Mommsen, 1978) and spider

venom proteases (Mebs, 1972; Jong et al., 1979; Veiga et al., 2000) have alkaline pH

optima.

Whereas the preceding observations suggest a strong similarity between digestive

fluid proteases and those of venom, there is one important difference: neither Kaiser and

Raab (1967), Wright et al. (1973), nor Feitosa et al. (1998) could find evidence of

eollagenases in spider venom. The digestive fluid of A. aurantia, by contrast, has

24

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. collagenase activity. Perret (1977) has observed that spider venom collected by

electrostimulation can be contaminated by digestive fluids. My observations suggest that

spider venom collected in this manner might first be screened for eollagenases as a

precaution against possible contamination from digestive fluid.

In spite of the fact that SDF has the ability to cleave a variety of connective tissue

proteins, this fluid produced only a minimal effect in the skin of rabbits (Fig. 1.7, Site II).

By contrast, brown recluse venom caused all the classical features associated with its bite

(Fig. 1.7, site III). These observations are consistent with those of Smith and Micks

(1968) who found that homogenates of the abdomens of L. reclusa do not evoke necrotic

lesions in the skin of rabbits. In addition, Geren et al. (1973) reported that brown recluse

spiders often regurgitate when biting, however, when the regurgitated material was tested

for toxicity, none was found. Taken together, these observations suggest that the

digestive proteases of spiders have no primary role in necrotic lesion formation.

The role Of venom proteases in spider bite necrosis remains uncertain.

Preliminary studies in the laboratory suggest that proteases are not a major component of

this fluid, and Kuhn-Nentwig et al. (2000) have estimated that the venom glands have

only one-ten thousandth of the protease activity of the digestive fluid. However, if

proteases had a primary role in spider bite necrosis, one would have expected that the

effects of the brown recluse venom in the skin of rabbits would have been reduced by the

simultaneous injection of hemolymph as this fluid neutralized brown recluse venom

proteases in vitro (Fig. 1.6a). However, it did nothing to suppress the effect of the brown

recluse venom (Fig. 1.7, site VI). At this time the most convincing agent associated with

spider bite necrosis appears to be sphingomyelinase D. The isolated agents appear to

25

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. cause both necrosis and hemolysis (Kurpiewski et al., 1981; Tambourgi et al., 1998) and

antibodies to these enzymes effectively prevent necrosis (Rees et al., 1984). At this time

there is no corresponding set of evidence in support of a primary role for venom or

digestive fluid proteases in necrosis. It is conceivable, however, that the proteases could

enhance the effect of venom in spider bites.

26

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 2

SURVEY FOR POTENTIALLY NECROTIZING SPIDER VENOMS, WITH SPECIAL

EMPHASIS ON CHEIRACANTHIUM MILDEI

Abstract

It has proven difficult to identify those spiders which cause necrotic lesions. In an

effort to design a simple, inexpensive screening method for identifying spiders with

necrotizing venoms, I have examined the venom gland homogenates of a variety of spider

species for their ability to cause red blood cell lysis. Those venoms which were positive

were further examined for the presence of sphingomyelinase D, and their ability to evoke

necrotic lesions in the skin of rabbits. Sphingomyelinase D is known to be the causative

agent of necrosis and red blood cell (RBC) lysis in the venom of the

(.Loxosceles reclusa), and my assumption was that this would be the same agent in other

spider venoms as well. This did not prove to be the case. Of 45 species examined, only

the venom of L. reclusa and Cheiracanthium mildei lysed sheep red blood cells. Unlike

L. reclusa venom, however, C. mildei venom did not possess sphingomyelinase D nor did

it cause necrotic lesions in the skin of rabbits. In an effort to determine the hemolytic

agent in C. mildei venom, I found evidence for a phospholipase A2.

27

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2.1. Introduction

It has proven difficult to identify which spider species produce necrotic lesions.

Only rarely does a person seeking medical attention for a spider bite have the specimen in

their possession. Moreover, if a culprit is not present, it is nearly impossible to

distinguish spider bites from those produced by other (Russell and Gertsch,

1983) or even some disease states (Vetter, 2000). It is also difficult for arachnologists to

link particular spider species directly with clinically significant bites. First, it is

necessary to inspect the area where suspected spider bites occurred to confirm that the

spider exists in sufficient numbers as to suggest a probable cause. In addition, it is

necessary to demonstrate that the injection of the venom will produce the appropriate

lesions (Spielman and Levi, 1970; Vest, 1987). And, whereas reviews listing spider

species that potentially cause lesions have been assembled (Russell and Gertsch, 1967;

White et al., 1995; Vetter and Vischer, 1998; Sams et al., 2002), the lists continue to be

modified. The continuing lack of consensus has led White (1999) to remark that "a

concerted effort to determine which species can cause necrotic arachnidism should

become a prime focus for research in this field".

It is now well established that Loxosceles spp. are capable of causing severe

necrotic lesions (Macchiavello, 1937; Atkins et al., 1958). In a systematic search for the

necrotizing agent, Forrester and coworkers (1978) discovered sphingomyelinase D, a

hemolytic constituent in the venom of L. reclusa. One microgram of the isolated enzyme

was later shown to cause necrotic lesions in rabbits (Kurpiewski et al., 1981). Antibodies

to the purified enzyme are able to prevent necrosis (Rees et al, 1984; Gomez et al., 1999).

28

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. These observations have since been extended to South American Loxosceles species

(Barbara et al., 1994: Tambourgi et al, 1998; Guilherme et al., 2001).

In an effort to develop a simple, inexpensive screening method to identify those

spiders which are capable of causing necrotic lesions, I have examined the venom of a

variety of spiders for the ability to cause sheep red blood cell hemolysis. Those which

were positive were further screened for sphingomyelinase D and dermonecrotic activity.

My survey was based upon the assumption that those spider venoms which caused

necrosis would also possess sphingomyelinase D and, therefore, cause sheep rbc lysis. I

recognize that this assumption is not completely valid: there are antimicrobial peptides in

the venom of some spiders that are also hemolytic (Yan and Adams, 1998; Corzo et al.,

2002; Kuhn-Nentwig et al., 2002a,b). By using sheep red blood cell suspensions and by

keeping the venom extracts in high dilution I hoped to avoid any antimicrobial peptide

influence with the assay.

2.2. Materials and Methods

2.2.1. Reagents

Defibrinated sheep blood was obtained from Northeast Laboratory (Waterville,

ME), co-trinitrophenylaminolauric acid (TNPAL)-sphingomyelin (T 1014),

sphingomyelinase D from Staphylococcus aureus (S 8633), snake venom from Crotalus

atrox (V 7000), and ethylene glycol-bis(2-aminoethylether)-N,N,N’,N’-tetraacetic acid

(EGTA, E 3889) were purchased from Sigma (St. Louis, MI). Silica Gel 60 TLC plates

(5721-7) were purchased from EM Sciences (Cincinnati, OH). Purified

phosphatidylcholine (PC), phosphatidylethanolamine (PE), phosphatidylserine (PS),

phosphatidylinositol (PI), and sphingomyelin (SPM) were acquired as a

29

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. phosphoglycerides kit (#1053) from Matreya, Inc. (State College, PA). BODIPY-labeled

phosphatidylcholines 2-(4,4-difluoro-5,7-dimethyl-4-bora-3a,4a- diaza-s-indacene-3-

pqntanoyl)-1 -hexadecanoyl- sn-glycero-3-phosphocholine (D-3803) and 2-decanoyl-l-

(0-(l l-(4,4-difluoro- 5,7-dimethyl-4-bora-3a,4a-diaza-s- indacene-3-

propionyl)amino)undecyl)-sn- glycero-3-phosphocholine) (D-3771) were purchased from

Molecular Probes (Eugene, OR).

2.2.2. Preparation of venom gland homogenates

All spiders were collected between June 2001 and November 2002 in the vicinity

of Durham, New Hampshire, unless otherwise noted. The spiders were fasted for at least

three days post-capture to ensure venom glands were not depleted. Spiders were

narcotized by carbon dioxide, and venom glands were removed by micro-dissection. The

glands were either used immediately or frozen at -20° C for future use.

2.2.3. Hemolysis assay

Hemolytic activity was determined using the spectrophotometric method of

Forrester et al. (1978). Individual venom glands were homogenized in 50 pi of a 20 mM

Tris buffered saline with 8 mM CaCl2, pH 7.4 (TBSC; after Babcock et al., 1981). The

entire venom gland homogenate (VGH) was added to 1.0 ml suspension of 2% sheep red

blood cells in a TBSC suspension and incubated at 37° C for 1.5 hours with gentle

agitation. The suspension was then diluted with 3.0 ml TBSC and centrifuged at

approximately 760 X g for 5 minutes. The optical densities of the supernatants were

measured at 540 nm with a Spectronic 601 spectrophotometer (Milton-Roy). Within

each assay, a 1 ml aliquot of sheep red blood cells was incubated as stated above, then

30

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. lysed with 3 nils distilled water, and used as the standard for 100% hemolysis. Percent

hemolysis was determined against total hemolysis.

2.2.4. Sphingomyelinase Assay

Sphingomyelinase D activity was determined using the spectrophotometric

method of Gatt et al. (1978) as modified by Young and Pincus (2001) using (TNPAL)-

sphingomyelin as substrates. Each venom gland was homogenized in 30 pi 20 mM

HEPES with 8 mM CaCl2 pH 7.1 then added to 170 pi solution of 0.3 mM TNPAL-

sphingomyelin containing 1.5 mM Triton X-100 in the above buffer. All samples were

incubated at 37° C for 2 hours. The positive control substituted 1.0 unit of

sphingomyelinase D from Staphylococcus aureus in place of spider VGH. Reactions

were stopped by the addition of 750 pi of isopropanol: heptane: 5M sulfuric acid (40:10:1

v/v). The mixtures were extracted with 450 pi heptane, followed by 400 pi distilled H 20.

All samples were vortexed then centrifuged for ten minutes. The upper heptane phase

was removed and its absorbance was read at 330 nm.

2.2.5. Phospholipase Assays

Defibrinated sheep red blood cell ghosts were prepared according to Dodge et al.

(1963). Individual venom glands from spiders were homogenized in 50 pi of 20 mM

HEPES buffer pH 7.1 with 8 mM CaCl2. The VGH was then added to 500 pi of a 17%

RBC ghost suspension in the same buffer, and incubated for 3 hrs at 37° C. Eight pg

protein of snake venom from Crotalus atrox were incubated with the same amount of the

RBC ghost suspension and used as a positive control. Membrane phospholipids and

cleavage products were extracted with chloroform: methanol (2:1 v/v) and separated on

Silica Gel 60 TLC using Solvent B (chloroform: methanol: 2-propanol:0.25% potassium

31

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. chloride: triethylamine (30:9:25:6:18 v/v)) of Touchstone et al. (1980). Results were

visualized by staining with a 3% cupric acetate, 8% phosphoric acid solution then heating

at 180° C for 10 minutes, or by using a 1% Rhodamine 6G stain (for quantitative use).

Phospholipase A activity was determined by the method of Farber et al. (1999).

BODIPY®-FL labeled phosphatidylcholine (Molecular Probes, WA) was used to

determine whether the enzyme cleaved the fatty acid from the sn-l (D-3771) or sn-2 (D-

3803) position. Cleavage products were separated by TLC methods using solvent B of

Touchstone et al. (1980) and visualized with a 488 nm laser on a Bio-Rad Molecular

Imager FX Pro (Bio-Rad).

2.2.6. The effect of C. mildei venom gland homogenate on the skin o f rabbits

To observe if venom from C. mildei is capable of causing necrotic lesions VGH

was injected into the skin of rabbits. The experiments were carried out in compliance

with the Public Health Service Policy, the Guide for the Care and Use of Laboratory

Animals (1996), and University Policy (see Appendix). Three adult female New Zealand

white (NZW) rabbits were injected in a shaved area on the back with the following

solutions: site I: 200 pi PBS, pH 7.4; site II: 15 pg of C. mildei VGH protein in 200 pi

PBS. All preparations were filtered through a 0.2 micron filter prior to injection. The

injections were administered intradermally with a 26 gauge needle approximately 10-12

cm apart. The rabbits were observed hourly for the first six hours and then every twelve

hours for the next 138 hrs. The injection sites were photographed with a digital camera

and scored according to the intracutaneous test described in the USP 24-NF19 (2000).

Inasmuch as no necrosis was observed, the same rabbits were administered a 5-fold

increase in concentration of C. mildei VGH protein seven days later. Site I again served

32

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. as the control, being 200 jil PBS. Site II was moved to the opposite side of the freshly

shaved rabbit. The injectate contained 79 pg of C. mildei VGH protein (approximately

4.4 venom glands) in 200 pi PBS. The rabbits were observed hourly for the first six

hours and then every twelve hours for 48 hours. Rabbits were euthanized and skin

samples were taken as described in the methods of Foradori et al. (2001).

2.2.7. The calcium dependency of the hemolytic factor in C. mildei venom

To observe whether calcium was essential to the hemolytic factor, individual

venom glands were homogenized in 1.0 ml 2% sheep RBC in 20 mM Tris buffered

saline, pH 7.4 (300 mOsm) with or without 8mM CaCl2. Hemolysis was evaluated as

described above (part 2.3). The comparison was repeated three times and the data was

evaluated using a t-test.

2.3. Results

2.3.1. Hemolysis assay

Cheiracanthium mildei was the only VGH in this survey found to cause sheep red

blood cell hemolysis (Table 2.1), other than Loxosceles reclusa, which was used as the

positive control. C. mildei females caused an average of 65.4% hemolysis; males 38.9%

hemolysis. L. reclusa females caused 37.5% hemolysis. The VGH of all other spiders

surveyed showed little (< 0.5%) or no ability to cause hemolysis.

2.3.2. Sphingomyelinase Assay

Sphingomyelinase D activity was not observed in female C. mildei VGH using

the modified method of Gatt (1978). L. reclusa VGH homogenate was rich in this

enzyme. See also figure 2.1 where L. reclusa VGH completely removed sphingomyelin,

but C. mildei VGH had no effect on sphingomyelin concentration.

33

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Family Species Sheep RBG Hemolysis

Loxoscelidae Loxosceles reclusa (3 )1 ++

Miturgidae Cheiracanthium mildei female (9) ++ male (8) +

Gnaphosidae Herpyllus ecclesiastices (2) Micaria longipes (2) - Unidentified Gnaphosidae (6)

Dtsderidae Dysderia crocata (1) "

Salticidae Phidippus audax <8) Phidippus clarus (2) Platycryptus undatus (4) Salticus scenicus (1) Unidentified Salticidae (5)

Tetragnathidae Tetragnatha elongata (6) Nephila madigascari (3) Nephila clavipes (4 )f

Araneidae Argiope aurantia (4) Argiope trifasciata (6) - Larinioides cornutus (10) p Neoscona arabesca (5) Araneus cavaticus (4) Araneus diadematus (3) Araneus marmoreus (4) Araneus trifolium (8) Gastercantha cancriformis (8 )r Micrathena sagittata (4 )8 Micrathena gracilis (3 )8 - Metazyga sp. (5) Verrucosa sp. (3) f - unidentified Areneidae (9)

Table 2.1. Survey of the venom of selected spider species for their ability to cause hemolysis. One venom gland was homogenized from each spider and assayed for its ability to cause hemolysis in 1.0 ml of 2% sheep RBC suspension. Number of spiders tested is indicated in parentheses. All spiders tested were female unless noted otherwise. All spiders were collected in New Hampshire unless otherwise noted (f = Florida; g = Georgia; k=Kansas; n = North Carolina; p = Pennsylvania; t = Texas; v = Virginia).

34

Reproduced with permission of the copyright owner. Fudher reproduction prohibited without permission. Family Species Sheep RBC Hemolysis

Agelenidae Agelenopsis aperta (4 )v - Unidentified Agelenopsis (30+) -

Filistatidae Kukulcania hibemalis (6 )f -■

Scytodidae Scytod.es thoracica (2) -

Pholcidae Pholcus phalangoides (4) -

Mimetidae Mimetus sp. (5) -

Theridiidae Steatoda borealis (7) Theridionfrondeum (3) - Achaearanea tepidariorum (8) - Latrodectus geometricus (4) r - Latrodectus maetans (3) -

Pisanridae Pisaurina mira (7) Dolomedes tenebrosus (3) - Dolomedes triron (3) Dolomedes vittatus (3) “ Unidentified Pisauridae

Thomisidae Misumena vatia (3) . Misumena asperatus (1) - Tibellus sp (5) “ Unidentified Thomisidae Yellow crab spider (7)

Oxyopidae Oxyopes salticus salticus (3 )k - Peucetia viridans (2 )f "

Linyphiinae Frontinella sp (5) -

Lycosidae Gladicosa gulosa (1) - Unidentified Lycosidae (3) -

Table 2.1. Continued.

35

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2.3.3. Phospholipase assay

The analysis of cleavage products from RBC ghosts exposed to venom gland

homogenates of C. mildei revealed the release of free fatty acids signaling the presence of

a phospholipase (Figure 2.1). Phosphatidylcholine (PC), phosphatidylethanolamine (PE),

and phosphatidylserine (PS) all released a free fatty acid and lysophospholipid when

exposed to VGH of C. mildei (Fig. 2.2) suggesting the presence of a Phospholipase A.

Phosphatidylcholine fluorescently labeled on the sn-1 position released fluorescent

labeled lysophosphatidylcholine; phosphatidylcholine fluorescently labeled on the sn-2

(D-3803) position released fluorescently labeled free fatty acid upon exposure to C.

mildei venom. These observations suggest the presence of a phospholipase A2 (Fig. 2.3).

2.3.4. The effect of C. mildei venom gland homogenate on the skin o f rabbits

Histological sections of skin and subcutis of NZW rabbits injected intradermally

with C. mildei VGH showed a mild to moderate reaction in comparison with the lesions

which I observed earlier in rabbits injected with Loxosceles reclusa VGH (Foradori et al.,

2001). Microscopic examination of the 5-fold increases of C. mildei VGH protein

injection site revealed a well demarcated and focal reaction with edema and mixed

inflammatory cell infiltrates (Fig. 2.4). In some instances myofibril necrosis was

observed (Fig. 2.4). A 6mm diameter red flare appeared within hours of the C. mildei

VGH injection, but slowly disappeared within 24 hours (Fig. 2.5). PBS was used as a

control and produced no observable changes in the skin and subcutis. The microscopic

appearance of the injection sites was consistent in all three rabbits.

36

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. f a ■ : IK

■vIKv^

PE 9 ^ m m m

PI PS LP

SM • •

RBC ghost RBC ghost RBC ghost RBC ghost RBC ghost blank + ::^Z; R 't+:\/ \/;.G:Y. v : r C. mildei P.andax L. geometriais L.reclusa VGH VG-H VGH VGH

Figure 2.1. The effects of spider venom gland homogenates (VGH) on RBC ghosts.

Extracts of RBC ghosts treated with VGH were chromatographed in the one­ dimensional system of Touchstone et al. (1980). The main components were phosphatidylethanolamine (PE), phosphatidylserine (PS), and phosphatidylinositol (PI), and sphingomyelin (SM). RBC ghosts treated with Cheiracanthium mildei VGH had a decreased density of PC, PE, and PS, and the appearance of free fatty acids (FA), as well as lysoforms (LP) of the previously mentioned phospholipids. Note the complete disappearance of SM withLoxosceles reclusa VGH. (Spiders in order are: C. mildei, Phidippus audax, Latrodectus geometricus, and L. reclusa)

37

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ffa *

t

PE i PS PC • • m y ' W"- m : m

* # * * , ■v: -i PE+ PE+ PE + PS + PS+ PS + PC + PC + PC + Cjnidei Cjnildei CMrax Cmiciei Cjnildei C.atrox C.mildei Cjnildei Catrcsc 0 mm 120 min 120 min. Omin. 120 min. 120 mini Omin. 120 min. 120 mm.

Figure 2.2. The effect of Cheiracanthium mildei venom gland homogenate (VGH) on individual purified phospholipids: Phosphatidylethanolamine (PE), phosphatidylserine (PS), and phosphatidylinositol (PC). Snake venom {Crotalus atrox) was used as a positive control.

38

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. /FA

/PC /LPC

W

Figure 2.3. The effect of Cheiracanthium mildei venom gland homogenate (YGH) on phosphatidylcholine (/PC) fluorescently labeled on the sn-l (D-3771) or sn-2 (D-3803) position. The release of fluorescent fatty acid (/FA) from D-3803 and fluorescent lysophosphatidylcholine (/LPC) from D-3771 indicate the presence of phospholipase A2 in the VGH.

39

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 2.4. The effect of Cheiracanthium mildei venom gland homogenate (VGH) on the skin of New Zealand white rabbits. Approximately 79pg protein of C. mildei VGH was injected intradermally. Picture taken 48 hours post-injection. Note: unlabelled dark spots are permanent marker used to mark coordinates of injection sites.

40

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 2.5. Histology of the skin of New Zealand white rabbit 48 hours after intradermal injection of 79pg protein Cheiracanthium mildei venom gland homogenate. A. View of the dermis and epidermis reveals edema and mixed inflammatory cell infiltrates (25x); B. Hair follicle showing heavy heterophil infiltration (100X); C. The panniculus (p) is beginning to lose banding, signaling degeneration and mild necrosis (200x)

41

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2.3.5. The calcium dependency o f the hemolytic factor in C. mildei venom gland

homogenate

The absence of calcium significantly reduced the hemolytic capacity of the

enzyme by almost 3-fold (p > 0.001) as seen in figure 2.6.

2.4. Discussion

Venom survey

The VGH of a variety of spiders has been surveyed for their ability to cause

hemolysis of sheep erythrocytes as a potentially simple and inexpensive way to screen for

necrotizing spider venoms. This survey was based upon the assumption that such

venoms would contain sphingomyelinase D, the necrotizing component in the venom of

the brown recluse (Kurpiewski et al, 1981), which is also known to cause red RBC lysis

(Forrester et al., 1978). Of the 45 spider species examined, representing 20 spider

families, only the VGH of Cheiracanthium mildei and L. reclusa were observed to cause

hemolysis (Table 2.1). The results accord with those of Croucamp and Veale (1999),

who observed hemolytic activity in the venom of Cheiracanthium furculatum. However,

in contrast to L. reclusa, C. mildei VGH did not possess sphingomyelinase D.

There are hemolytic factors in the venom of some spiders other than

sphingomyelinase D. Lycosids (Yan and Adams, 1998) and Oxyopids (Corzo et al.,

2002) have amphipathic peptides that are both antimicrobial and hemolytic. Their

effectiveness on red blood cells, however, seems to depend upon the phosphatidylcholine

(PC) content of the membranes. The PC content of sheep erythrocytes is less than 1%

(See Fig. 2.1) and higher concentrations of the amphipathic peptides are required for lysis

than is the case for the guinea pig and pig (Corzo et al., 2002). One venom gland in 1.0

42

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 80

7 0 -

6 0 -

20 -

TBS + Ca TBS w/o Ca

Figure 2.6. Calcium dependency of Cheiracanthium mildei venom gland homogenate hemolytic factor. Homogenates of individual venom glands were incubated with 1.0 ml 2% sheep RBC in 20 mM Tris buffered saline, pH 7.4 (300 mOsm) with or without 8 mM CaCl2. (n=3; error bars = SEM)

43

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ml of 2% sheep red blood cell suspension was used for an assay, it is likely that the

amphipathic peptides were too dilute to observe an effect with those species (Table 2.1).

Cupiennius salei (Family Ctenidae) venom has also been shown to contain antimicrobial

peptides (Kuhn-Nentwig et al., 2002a,b), so it may be that antimicrobial peptides are

universal components in the venom of spiders. Their likely function is that of disrupting

ion and voltage gradients across the membranes of excitable cells (Yan and Adams,

1998).

Cohen and Quistad (1998) observed that the venom of Phidippus audax contained

a potent cytolytic factor that disrupted cells in culture. However, neither lysis of

erythrocytes nor cleavage of products when the VGH of P. audax was incubated with

RBC ghosts were observed (Fig. 2.1).

Hemolytic factor in the venom o f Cheiracanthium mildei

To establish the factor present in the VGH of C. mildei responsible for RBC lysis,

filtered VGHs were incubated with sheep RBC ghosts. Thin layer chromatography of

incubation extracts revealed a decreased density of PE, PS, and PI as well an appearance

of fatty acids and lysoforms of the lipids, suggesting the presence of a phospholipase

(Fig. 2.1). Incubation of C. mildei VGH with commercially available PE, PS, and PC

confirmed the cleavage of these compounds (Fig. 2.2). The reduction in activity of the

hemolytic agent by the removal of calcium (Fig. 2.6) also indicates the presence of

phospholipase A2(PLA2), as Ca2+ is required for optimal activity (Hanahan, 1997). To

identify the site of cleavage of the phospholipids, VGH was incubated with PC

fluorescently labeled on the sn-l (D-3771) or sn-2 position (D-3803). Whereas

fluorescently labeled fatty acid was released from D-3803, fluorescently labeled

44 '

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. lysophospholipid was released from D-3771 (Fig. 2.3) suggesting the presence of a

phospholipase A2 in the venom gland homogenates ofC. mildei.

PLA2 has been observed in the venom of the spider Eresus cinnaberinus (= niger)

and has been shown to produce an irreversible blockage of frog neuromuscular junctions

as well as anticoagulant activity (Usmanov and Nuritova, 1994). The enzyme is also

present in the venom of most snakes (Tu, 1977), and like the snake venom PLA2, C.

mildei venom gland homogenate PLA2 has a calcium requirement (Fig. 2.6).

The venom of C. mildei has been observed to cause necrotic lesions in the skin of

guinea pigs (Spielman and Levi, 1970) and hamsters (Hughes, 1981), and

Cheiracanthium lawrencei venom has been shown to cause necrosis in the skin of rabbits

(Newlands et al., 1980). We, however, observed only a mild erythema on the skin of

rabbits following the injection of the equivalent of 4 glands of protein (Fig. 2.4). Whereas

Spielman and Levi (1970), Hughes (1981), and Newlands et al. (1980) allowed spiders to

bite their prey, I injected the rabbits intradermally with VGH. The difference in the

technique may account for the differences in the results obtained. And whereas Newlands

et al. (1980) observed ulcerating lesions in humans resulting from the envenomation of C.

lawrencei, most observations suggest that the effects of Cheiracanthium spider bites in

humans are mild. Spielman and Levi (1970) observed only a mild reddening of the skin

of a human volunteer resulting from C. mildei spider bite. Minton (1972) and Krinsky

(1987) also observed only redness, swelling, and itching in humans resulting from the

bite of C. mildei. Similar observations have been made for the venom of C. inclusum

(Furman and Reeves, 1957; Gorham and Rheney, 1968; Leech and Brown, 1994), C.

punctorium (Maretic, 1982), and C. japonicum (Owri, 1978).

45

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The foregoing experiments support the theory that severely necrotizing spider

venoms are rare. Of all the spider venoms examined, only two produced hemolysis, and

of these, only the brown recluse produced necrotic lesions in the skin of New Zealand

white rabbits. Although the venom of the local yellow , Cheiracanthium

mildei, produced hemolysis, its effect upon the skin of New Zealand white rabbits at

normal levels was mild. My results suggest that the widespread fear of spider bites, for

the most part, is unwarranted.

46

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 3

FRACTIONATION OF SPIDER DIGESTIVE FLUID AND PARTIAL

CHARACTERIZATION OF THE PROTEASES AND PROTEASE INHIBITORS

ABSTRACT

Digestive fluid proteases and serine protease inhibitors play an important role in

the life of the spider. The proteases allow for external digestion of prey, as well serving

as a tool to recycle silk. Through a combination of ion exchange chromatography, size t exclusion chromatography, ultracentrifugation, and electrophoresis, I have tried to isolate

several digestive fluid proteases and serine protease inhibitors in milligram quantities

from the digestive fluid of Argiope aurantia. The proteases, however, eluted

anomalously during the size exclusion chromatography and migrated anomalously during

ultracentrifugation trials. I have hypothesized that proteases can undergo a reversible

aggregation to account for this behavior. I have also examined the possibility that spider

digestive enzymes are related to crayfish gastric proteases. While my evidence supports

this in part, I found major differences between the two groups of enzymes. I found

evidence for 10 or more serine protease inhibitors in Argiope aurantia digestive fluid.

3.1. Introduction

The digestive proteases of orb weaving spiders play an important role not only in

the consumption of prey, but also in the recycling of their orb web (Tillinghast and

Kavanagh, 1977; Townley and Tillinghast, 1988) and the escape of spiderlings from the

47

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. egg sac (Foradori et al., 2002). Little is known about the biochemical characteristics of

spider digestive enzymes which must function during the initial stages of digestion

without the aid of a controlled environment (e.g., stomach of mammals). The

evolutionary relationship of spider digestive proteases to those of other animals has not

been established either.

Bertkau (1885), Schlottke (1936), and Pickford (1942) were among the first to

study the digestive hydrolases of spiders. A more complete inventory was provided by

Mommsen (1978 a,b,c) who fractionated the digestive fluid by polyacrylamide gel

electrophoresis. Mommsen (1978a) identified one protease capable of cleaving natural

proteins and several enzymes capable of cleaving synthetic substrates typically employed

for studying the functioning of mammalian proteases. Tillinghast and Kavanagh (1977)

brought two spider digestive proteases to apparent homogeneity. They had masses of

approximately lOkDa and were metalloproteases with alkaline pH optima. One of these,

Argiope protease B, proved capable of cleaving major ampullate silk (Kavanagh and

Tillinghast, 1983). More recently, serine protease inhibitors have also been demonstrated

in the digestive fluid of A. aurantia (Tugmon and Tillinghast, 1995).

Proteases have also been reported in a number of spider venoms (reviewed by

Perret, 1977). In one of these reports, Mebs (1972) isolated three proteases with

molecular weights of approximately 11 kDa. They had pH optima of 8.5, did not cleave

trypsin substrates, and were equally unaffected by serine protease inhibitors. These

characteristics are essentially identical to proteases later described in spider digestive

fluid (Kavanagh and Tillinghast, 1977). Mebs also noted that the spider proteases were

48

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. similar in molecular weight and catalytic properties to a protease isolated from crayfish

gastric fluid (Pleiderer et al., 1967).

Previous efforts to isolate individual spider digestive proteases resulted in a very

low yield (Kavanagh and Tillinghast, 1983). In this study, I have attempted to develop a

more effective protocol for fractionating digestive fluid and purifying its proteases and

serine protease inhibitors for biochemical characterization. In these experiments, I have

used the spider digestive fluid amylases and N-acetylglucosaminidase as internal

indicators of the effectiveness of my methods. Both are major enzymes in the spider’s

digestive fluid and have established molecular weights (Mommsen, 1978). I have also

examined the possibility that spider digestive proteases are related to crayfish digestive

proteases

3.2. Materials and Methods

3.2.1. Reagents

Casein substrate was Carnation Dry Milk or Hammersten casein (Schwarz-Mann,

Orangeburg, NJ). Tris (15504-020) was purchased from GIBCO-BRL. All dialysis was

performed using 3500MWCO Slide-A-lyzer dialysis cassettes (Pierce). Size exclusion

columns were calibrated with various combinations of alcohol dehydrogenase (Sigma,

A8656, MW 150kDa), bovine serum albumin (A8531, MW 66 kDa), carbonic anhydrase

(C 7025, MW 29kDa), cytochrome C (C7150, MW 12.4 kDa). All other reagents were

products of the Sigma Chemical Company, St. Louis MO.

3.2.2. Collection of biological fluids.

Spider digestive fluid (SDF) was collected from Argiope aurantia by holding the

in one hand and putting a micropipette to its mouthparts. Under these

49

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. circumstances A. aurantia freely released digestive fluid (SDF) which was collected and

frozen until used. The SDF was then centrifuged for 2 min at 14,000 rpm in an

Eppendorf centrifuge 5415 to remove particulate matter and then dialyzed overnight

against the buffer in which it would be fractionated.

Crayfish gastric fluid (CGF) was obtained by inserting a 5 ml syringe into the

stomach of an anesthetized Astacus sp. and withdrawing fluid. Hemolymph was obtained

from both animals by clipping a leg and drawing fluid into micropipettes.

3.2.3. Isolation o f SDF Proteases

SDF was dialyzed against 25 mM Tris, pH 8.1 containing 2 mM CaCl2 (buffer

A). A 1.5 ml sample was injected into a 5 ml Q econo-cartridge (Bio-Rad) and was

eluted with a 0 to 0.25 M NaCl gradient in the same buffer. The fractions were

monitored at OD 280nm for protein. Fractions were pooled and assayed for protein,

proteases, amylase, N-acetylglucosaminidase, and protease inhibitors by methods

described below.

The first ten fractions from anion-exchange chromatography were pooled and

dialyzed against 50 mM Tris buffer, pH 7.6 and the dialysate was applied to a column

(1.4x100 cm) of Sephacryl S-100 HR pre-equilibrated with the same buffer. The sample

was eluted at a flow rate of 12.8 ml per hr and 3.2 ml fractions were collected. The

fractions were monitored at 280nm for protein. Samples were assayed for amylase

activity and proteases. The proteases were Visualized by zymography (3.2.3 below).

Fractions 48-54 from the Sephacryl S-100HR column were pooled and concentrated by

lyophilization, dialyzed against 20 mM HEPES buffer (H7006), pH 7.5 and applied to a

5ml High S econo-cartridge (Bio-Rad) pre-equilibrated with the same buffer. The

50

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. proteins were eluted with a gradient of 0 to 0.25M NaCl in the same buffer. Samples

were assayed for protease activity. Active fractions were pooled, lyophilized, and

aliquots electrophoresed to observe the proteases. Selected bands were electroblotted for

sequencing.

^ 3.2.4. Fractionation of whole SDF

In other studies, whole SDF was fractionated on Sephacryl S-100 HR column

(1.4x100cm). The SDF was first dialyzed against 50 mM Tris buffer, pH 7.6, with 100

mM KC1 and eluted with the same buffer at a flow rate of 13.6 ml per hour with 3.4 ml

fractions being collected per tube. Fractions 26-32 were further fractionated on a column

(1.5X75 cm) of Sephadex G-75 using a 50 mM Tris buffer, pH 7.6 containing 100 mM

KC1. Proteins were eluted at a flow rate of 20 min/tube and 3.0 ml collected per tube.

3.2.5. Isolation of SDF serine protease inhibitors

One ml of SDF was applied to a Q-Sepharose anion-exchange column (1.4x20.3

cm) and eluted with 100 ml 20 mM Bis-Tris (B7535), pH 6.8, followed by 400 ml of a 0

to 500 mM NaCl gradient in the same buffer. This was followed by an additional 100 ml

of the buffer containing 500 mM NaCl. The flow rate was 1.3 ml /min and 3.4 ml

fractions were collected. The eluate was monitored continuously at 280 nm. Peak

fractions and regions between were pooled separately, lyophilized to reduce volume and

de-salted with Sephadex PD-10 columns. The high molecular weight fractions were

reduced in volume in a Savant speed-vac concentrator, boiled for 4 min, and centrifuged

at 14,000 rpm for 5 min to remove inactivated proteases. The fractions were reduced to

250 pi in the Savant concentrator and 20pl volumes were used to locate protease inhibitor

activity by radial diffusion. Pooled fractions containing the highest inhibitor activity

51

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. were examined by native and denaturing PAGE with sample aliquots of 7.5 to 15 pi per

lane.

3.2.6. N-terminal sequencing

Proteins were transferred from SDS-PAGE gels to polyvinylidene difluoride

(PVDF) membranes at 30 V (constant) for 45 min to 1 hr using the Nu-PAGE system

(Invitrogen). The PVDF membrane was stained with 50% methanol, 10% acetic acid and

0.1% Coomassie blue solution briefly to visualize proteins of interest, then destained with

50% methanol (Matsiduria, 1993). Selected bands were cut out and sent to the University

of Texas Medical Branch Protein Chemistry Laboratory for sequencing.

3.2.7. Comparison of spider and crayfish proteases

Electrophoresis-horizontal; 1% argarose (A6013) gels were prepared with barbital

buffer, pH 8.6 (B6632). Samples of SDF or CGF were electrophoresed at 30 mA

(constant) for 1 hr. To visualize protease activity, selected lanes were removed and

placed on gels prepared with 1% agarose (A6013) and 1% casein (Carnation dry milk) in

buffer A. Other lanes were stained for protease inhibitor activity as described below.

Hemolymph protease inhibitor studies (radial diffusion): Gels composed of 1%

argarose (A6013) and 1% casein were prepared with buffer A. Mixtures were prepared

containing 5pi SDF or CGF and 15 pi spider hemolymph buffer A or 15 pi buffer A.

These were vortexed and transferred to wells cut into the gels. The gels were incubated

overnight and the diameter of digestion was recorded.

3.2.8. Sucrose Density Centrifugation

SDF was centrifuged at 3000 rpm for 3 min, then 15pl aliquots were mixed with

180 pi buffer (50 mM Tris-HCl, pH 7.5) and carefully applied to a 5 ml linear gradient

52

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (5-30% w/v) of sucrose in the same buffer. The samples were centrifuged at 4°C for 21

hr at 160,000 x g in a Beckman model L8-70 ultracentrifuge using a SW65 rotor. Spun

tubes were placed in a holder and the bottom of the tube was punctured with a needle and

0.25 ml fractions collected. The fractions were assayed for proteases, protease inhibitors,

amylase, and N-acetylglucosaminidase.

3.2.9. Electrophoresis

Native PAGE: The high pH discontinuous system described in Hames and

Rickwood (1981) was used with 8.5% resolving gels containing AcrylAide (EMC

BioProducts , Rockland, ME) as the cross-linker.

SDS-PAGE: Small format, Novex Tris Glycine gels (Invitrogen) of different

percentages were employed and run at 125 V (constant) for 2 hrs. For large format

electrophoresis, the discontinuous system of Laemmli (1970) was used with 13% gels

resolving gels, again containing AcrylAide, and samples were boiled for 1 min after

addition of sample buffer.

Zymography: 4-16% Novex Zymogram gels (Invitrogen; EC 6415) were used to

visualize proteases according to the manufacturer’s instructions.

SDF serine protease inhibitors were visualized in gels by the method of Broadway

(1993). Briefly, following SDS-PAGE, SDS was removed from the gels by washing the

gels sequentially in 40 mM Tris HC1 (T8529), pH 9.0 for 15 min, distilled water for 5

min, the same Tris buffer for 25 min and distilled water for 10 min at 37°C. The gels

were then covered with a Trypsin, (TPCK treated; T8642) solution (0.04 mg/ml PBS, pH

7.4) for 10 min at 37°C. The solution was then decanted and the residual trypsin allowed

to penetrate the gel for 30 min at 37°C. The gel was then covered with tow solutions

53

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. mixed just prior to staining. The first was 50 mg of the tryptic substrate N-Acetyl-DL-

phenylalanine (Tnaphthyl ester (A7512) dissolved in 20 ml N ,N-dimethylformamide

(D9654); the other waslOO mg O-dianisidine, tetrazotized (D3502) dissolved in 180 ml

PBS, pH 7.4. The stain was decanted and the gels stored in 2% acetic acid.

Elution: Proteins were eluted from agarose and polyacrylamide gel slices in

volatile NH4HC03 buffer using Bio-Rad Model 422 electro-eluter. Electroelution was at

a constant 8 mA per tube for 6 hr. The apparatus was placed on ice to prevent

overheating.

3.2.10. Enzyme Assays

Protease activity was assayed by radial diffusion agarose/casein gels as described

by Schumacher and Schill (1972) or the method of Kunitz (1947) using a 1%

Hammersten casein solution as the substrate. SDF serine protease inhibitors were

assayed using radial diffusion in argarose/casein gels as described by Tugmon and

Tillinghast (1995). Amylase was assayed according to the procedure of Bernfeld (1951)

using dinitrosalycylic acid as the color reagent. Maltose was used to construct a standard

curve. N-acetylglucosaminidase was assayed by the method of Peters et al. (1998) using

4-Nitrophenyl N-acetyl-(TD-glucosaminide as substrate.

3.3. Results

3.3.1. Isolation o f Proteases

The fractionation of SDF by anion-exchange chromatography is presented in

Figure 1. The bulk of the proteases eluted prior to the application of the salt gradient

(Fig. 3.1 A). The amylase also appeared with unbound material. Protease inhibitors and

N-acetylglucosaminidase were both released early in the salt gradient (Fig. 3. IB).

54

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A

3.0 r 0.50 ■OD 280 2.5 ■[NaCl] - 0.40

2.0 0.30 COo Q 1.5 O 0.20 <3 1.0 O Z

0.0 0.00 o in o in o in o in co co it Fraction Number B

1 0 0 -i El% amylase D % protein ■% protease 80 - P % inhibitor 70 - E0% glucosaminidase

60 - c O4) <0 CL

O o O lO o o CM CO CO Tj- lO in

Pooled Fraction Fig. 3.1. Fractionation of SDF by anion exchange chromatography

A. SDF (0.5ml) dialyzed against 0.25 mM Tris buffer, pH 8.1 containing 5 mM CaCl2. A 1.5 ml sample was applied to a 5 ml high Q econo-cartridge (Bio-Rad) pre- equilibrated with the same buffer. The sample was eluted with a gradient of 0 to 0.25 M NaCl. B. Pooled fractions were assayed for protease, amylase, N-acetylglucosaminidase, protein and protease inhibitor. The specific activities of the enzymes were determined and the percent f the total calculated for each pooled fraction.

55

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. When the first 10 tubes from anion-exchange chromatography were pooled and applied to

Sephacryl S-100 HR amylase appeared in fraction 36 whereas pl2 and pl5 eluted prior to

it and p27 and p35 eluted later. Protease pl2 appeared in all fractions (Fig. 3.2). To

further purify p27, SDF fractions 48-54 were pooled and passed through a High S Bio-

Rad econo-column at pH 7.5 and eluted with a Salt gradient. A doublet appeared in

fractions 26-40 with apparent molecular weights of 24 and 28 kDa. These two isolated

proteases were electroblotted and sent out for sequencing (Fig. 3.3).

3.3.2. Whole SDF fractionation on Sephacryl S-100 HR and Sephadex G-75

When whole SDF was passed through Sephacryl S-100 HR column, N-

acetylglucosaminidase (MW108kDa; Mommsen, 1980) appeared in fractions 21-25;

proteases pl5, p20 and p i 10 appeared in fractions 26-32; proteases p27 and p35 appeared

in fractions 33-38 along with amylase (MW approx. 60kDa; Mommsen, 1978d).

Protease pl2 appeared in all fractions including all of those after fraction 38. When

fractions 26-32 were further subject to size exclusion chromatography on Sephadex G-75,

pl5 and pllO emerged following p20. pl2 appeared in all fractions (Fig 3.4)

3.3.3. SDF serine protease inhibitors

The bulk of the protease inhibitors eluted in fractions 56-91 during anion-

exchange chromatography. Ten serine protease inhibitor bands were seen when pooled

fractions 56-67 and 68-91 were examined by 8.0% PAGE. These were designated A, B,

C, D, E, E’, F, G and G’ with E ’ and G’, appearing in fractions 68-91 and not in 56-67.

Likewise A and B occurred in 56-91 but not in 68-91 (Fig. 3.5). Individual inhibitors C,

D, E, F, and G were eluted from unstained parallel lanes of gels containing fractions 56-

67 and re-electrophoresed on 8% PAGE gel (Fig. 3.5). Three of these were sufficiently

56

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 0.6 25 27 33 35 37 38 40 42 45 Fraction # 0 .4

0.2

i i i i i 1"I I i I i TTTT'f I 'I "1 l"'T'"T I I M TTTTTTT i "i i m i i r~ oocvi

Fig. 3.2. Fractionation of fractions 1-10 on Sephacryl S-100 HR column.

The first ten fractions from anion exchange chromatography (Fig. 3.1) were pooled, lyophilized to reduce volume, dialyzed against 0.25 mM Tris buffer, pH 8.1 and applied to a 1.4x100 cm column of Sephacryl S-100 HR pre-equilibrated with the same buffer. The sample was eluted at a flow rate of 12.8 ml per hr and 2.8 ml fractions were collected. Selected fractions were examined for protease activity by zymography (inset) using 20pl aliquots. Note that pl2 occurs in every fraction.

57

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. kDa fraction

20.5

Fig. 3.3. Fractionation of fractions 48-54 on cation exchange chromatography

Fractions 48-54 from the S-100 HR column (Fig. 3.2) were pooled, lyophilized to 3.0 ml, then dialyzed overnight against 25 mM HEPES, pH 7.5. The dialysate was applied to 5ml High S econo-cartridge (Bio-Rad) pre-equilibrated with the same buffer. The sample was eluted at a flow rate of 2.0ml per minute with a gradient of 0 to 0.25 M NaCl. Fractions 30-36 containing high protease activity were subject to 10% SDS-PAGE and stained for protein.

58

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. kDa

178 0.8 111

0.7

0.6

0.5

0.4 m

0.3 32 37 0.2 fraction #

0.1

0 j | — CD co CD CO lO N- CO h - CD T- T— CM CM CM CM CO CO CO CO 't f

Fraction #

Figure 3.4. Fractionation of fractions 26-32 on a Sephadex G-75 column.

Whole SDF was fractionated on Sephacryl S-100 HR column using a 50 mM Tris buffer, pH 7.6 containing 100 mM KC1. Intense protease activity was found in fractions 26-32, which included pl5, p20, and pi 10. The pooled fractions were placed on a Sephadex G- 75 column, with the above mentioned buffer. Selected fractions were examined for proteases activity by zymography (inset) using 20 pi aliquots. Note that p 15 and pi 10 emerged ahead of p20 and that a majority of the fractions contain pl2.

59

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 58-67 68-91 C D E F G

I I I

Fig. 3.5. Separation of SDF protease inhibitors by natural PAGE.

SDF was fractionated by anion-exchange chromatography and those fractions containing serine protease inhibitors were separated by natural electrophoresis on a large format 8% PAGE gel. Ten serine protease inhibitors are seen in the two samples of pooled fractions (Gel I). C, D, E, F, and G were electroeluted individually from parallel lanes of Gel I and electrophoresed on a second gel (Gel II).

60

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. concentrated (C, E, F) to again elute from parallel unstained lanes and re-electrophorese

on a 13% SDS-PAGE gel. All inhibitors now appeared as doublets (Fig. 3.6). When 2-

mercaptoethanol was included during electrophoresis no inhibitors were observed.

3.3.4. N-terminal amino acid sequences o f proteases

A pattern emerged from the N-terminal amino acid sequences I observed for

several proteases (Table 3.1). The N-terminal amino acid sequence pattern for the first

20 amino acids was as follows: NAIPGQHYR-PGAKVPY-I-. This generic amino acid

sequence, as well as each of the individual sequences were subjected to a protein

homology and similarity search using FAST A (Pearson and Lipman, 1988; Pearson,

1990). All of the sequences had a 52.94% identity (52.94% ungapped) to chitinase.

Protease SDF 25 (S) A 20-25 was unlike any of the other protease, and when subjected to

a FAST A search, had 61.111% identity (61.111% ungapped) to an astacin precursor and

a 40.0% identity (40.0% ungapped) to Astacin family metalloendopeptidase, both of

which are found in crayfish digestive fluid.

3.3.5. Comparison o f spider and crayfish digestive proteases

When SDF and CGF were co-electrophoresed at pH 8.6, the crayfish proteases

moved rapidly toward the anode, whereas the spider proteases migrated toward the

cathode (Fig 3.7A). Radial diffusion studies showed that whereas spider hemolymph

readily inhibits SDF proteases, the spider’s hemolymph had only a little effect upon

crayfish protease (Fig 3.7B). The converse experiment showed that whereas crayfish

hemolymph readily inhibited crayfish proteases their hemolymph had little effect upon

spider digestive proteases.

3.3.6. Ultracentrifugation of SDF

61

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. -a tm ttMM C E F

Fig. 3.6. Separation of SDF protease inhibitors by SDS-PAGE.

C, E and F were eluted from parallel lanes in gel II (Fig 3.5). These were subject to denaturing electrophoresis using a 13% SDS-PAGE gel. Note that all three eluted inhibitors now appear as doublets. No bands appeared if 2-mercaptoethanol was included in the gel

62

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. amino acid residue sample 10 SDF 25 (S) A 20-25 Kd E E SDF 25 (S)B 20-25 Kd SDF 08 01-10 15 Kd SDF 08 Q l-10 19 Kd X X SDF 2003 Series A 15KD SDF 2003 Series A 23KD SDF 2003 Series A 27KD SDF 2003 Series A 28KD SDG-15-15 W ! F I ! 1 SDG-15-15.2 SDG-15-36 STB-15-12

Table 3.1. N-terminal amino acid sequence of isolated protease from the digestive fluid of Argiope aurantia, i

Digestive fluid enzymes were isolated by ion exchange chromatography and SDS- PAGE, blotted to PVDF for determination of N-terminal amino acid sequence; All of the proteins sequences possess the same core sequence (except for SDF 25 (S) A 20-25 Kd): NAIPGQHYR-PGAKVPY-I-. Queries using a FAST A protein homology and similarity search (Pearson and Lipman,1988) revealed ho significant matches. (“X” denotes no amino acid given, shaded amino acids are identical)

63

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A

'CGF+"SH OOF

Fig. 3.7. Comparison of spider and crayfish digestive proteases

A. 20 pi samples of spider digestive fluid (SDF) and crayfish gastric fluid (CGF) were electrophoresed in a 1% agarose gel, pH 8.6 for 1 hr at 20 mA (constant). Selected lanes were removed and placed on gels composed of 1% casein and 1% agarose,'pH 8.1 to visualize proteases. Note that the bulk of the spider proteases move slowly toward the cathode, whereas the crayfish proteases move rapidly toward the anode.

B. Five pi samples of SDF or CGF was mixed with 15 pi spider hemolymph or 15 pi buffer A (25 mM Tris, pH 8.1 containing 2 mM CaCl2). These were placed in wells of gels composed of 1% agarose and 1% casein, pH 8.1 and allowed to incubate overnight.

64

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. To observe if the proteases clustered into higher aggregates, SDF was subjected to

sucrose density gradient centrifugation at 160,000 x g for 21 hr, Whereas the spider N-

acetylglucosaminidase, amylase and protease inhibitors separated from each other as

rather narrow peaks in accordance with their masses, the proteases formed a single broad

peak with protease activity appearing in every fraction (Fig. 8). In part this broad peak

might be explained by the widely differing masses of the individual proteases (Fig. 9b),

however, proteases pi 10, p32, p27 and pl2 were still seen together in every fraction (Fig.

9a).

3.4. Discussion

This study was initiated with the goal of isolating spider digestive proteases in

milligram quantities for biochemical characterization. Although the digestive fluid of A.

aurantia contains approximately 35mg protein per ml (Tillinghast and Kavanagh, 1977),

and at least six major proteases (pi2, pl5, p27, p35, and pi 10), I encountered more

difficulty than anticipated. In no chromatographic separation did a protease, other than

p i2, ever elute as a single protein species and pl2 eluted with almost every fraction. My

experience was much the same as Mebs (1972) in his attempt to purify spider venom

proteases. In passing venom through a series of Sephadex gels, Mebs observed

anomalous elution times such as to suggest that the proteases were aggregating and

disassociating. He isolated proteases from three very different chromatographic fractions

and yet all had masses of approximately 1 lkDa, identical catalytic properties, did not

cleave trypsin substrates, and were equally unaffected by serine protease inhibitors. He

concluded that they were the same enzyme and that the reversible aggregation of the

proteases accounted for their differing elution rates. The data for spider gut proteases are

65

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1.600 ■0 “ —=>pratease activity OD 275 60 -A— amylase activity OD 540 1.400 ~e—-N-acetyl glucosaminidase OD 410 •B—-percent inhibition 50 1.200

S ' 1 .0 0 0 40

0.600 3 0

0.600

20

0.400

0 .2 0 0

0.000 1 2 3 4 5 6 10 1 1 12 13 1 4 1516 17 18 Fraction #

Figure 3.8. Fractionation of SDF by ultracentrifugation

15 |il SDF was applied to 5 ml of a 5-30% sucrose density gradient and centrifuged for 21 hr at 160,000 x g (SW65 rotor) at 4°C. Fractions were assayed for the distribution of amylase, proteases, N-acetylglucosaminidase and protease inhibitors.

66

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. fractions 2 4 6 8 10 12 14 16. 18

kDa iPllilgi 182 9 113.7 80 9 63 8 HIS 37 4 26 0 ilplSlP 149 3BMItey _ — IlilliS MHM pBBliMIp I fractions- 1-5 6-9 10-13 14-18

Figure 3.9. Whole SDF fractionation by sucrose density gradient ultracentrifugation examined by (A) zymography and (B) SDS-PAGE.

A. Five pi of the even fractions from SDF ultracentrifugation were examined for protease activity by 4-16% SDS-PAGE zymography. Note the persistence of pl2, p27, p32, and pllO throughout the fractions.

B. Fractions from SDF ultracentrifugation were pooled into groups of four or five. Twenty pi of each pooled fraction were examined for protein content by 4- 20% SDS-PAGE.

67

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. consistent with Mebs (1972) interpretation and I believe that it explains why only low

molecular weight proteases have been purified to date and why the yield has been so

small (Tillinghast and Kavanagh, 1977, Kavanagh and Tillinghast, 1983). It has long

been known that related proteins such as hemocyanin undergo reversible association and

dissociation depending on protein concentration, pH and the ionic environment

(Pedersen, 1950). Future efforts to isolate individual proteases may find it useful to first

identify conditions of pH, temperature and ionic environment such as to reduce the

association of the proteases more fully prior to column chromatography. High pH, for

example, has proven a very effective way of dissociating spider hemocyanin monomers

(Sugita and Sekiguchi, 1975).

It is possible that the reversible association of the spider proteases may have a

functional importance in the digestion of insect prey. Having two or more proteases with

differing cleavage specificities in close proximity may help to more rapidly inactivate the

proteases of the prey during extra-oral digestion. Or it may be that differing catalytic

rates occur in the associated and dissociated forms as Hipps and Nelson (1974) suggested

for cockroach digestive esterases. And, whereas I did not find conclusive evidence for

aggregates by sucrose density gradient centrifugation, this may be explained by the high

dilution (15pl SDF in 5ml sucrose) used in these experiments. However, the appearance

of the p32, p27 and pl2 in all fractions (Fig. 6) does suggest that these proteases

maintained an affinity for one another. The SDF would not become so dilute when the

spider consumes its prey. It is often easy, for example, to obtain 100 pi or more of the

SDF from a single spider depending on the state of hydration of the animal. If this

volume were injected into an insect of equal size the dilution would only be about 1:2,

68

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and this would likely favor the equilibrium towards the association of the proteases. At

first, I thought that the serine protease inhibitors would be linked with the hypothetical

proteases aggregates. This does not seem to be the case. During ultracentrifugation

experiments the inhibitors separated discretely from the bulk of the proteases (Fig. 6).

Mebs (1972) noted a similarity of the masses and enzymatic properties of the

isolated spider venom proteases compared to those of the crayfishAstacus fluviatilis. I

examined the possibility that the spider and crayfish digestive proteases are closely

related by determining the N-terminal sequences of isolated spider proteases. I found,

however, only one set of data indicating a close relationship (SDF 25 (S) A 20-25 Kd;

Table 3.1). There must however be major differences as: 1) the spider hemolymph

exerted only a slight inhibition of crayfish gastric proteases and likewise the crayfish

hemolymph exerted only a slight inhibition of spider digestive prbteases (Fig.5b) and

whereas the bulk of the spider digestive proteases appear to be cationic at pH 8.6, the

major crayfish proteases are anionic (Fig.5a). My observations are consistent with

Zwilling and Neurath (1981) who noted the low isoelectric point of the LMW proteases

of the crayfish gastric fluid.

I found 10 serine protease inhibitors when SDF was fractionated by natural PAGE

(Fig 3.4). They had masses ranging from 14 to 43 kDa. When three of these were

isolated and subject to SDS-PAGE, each separated into two bands. The additional bands

may indicate that the pairs of inhibitors had co-migrated with natural PAGE or possibly

that the inhibitors are heterodimers. In any case no inhibitors were seen when 2-

mercaptoethanol was used during electrophoresis suggesting that disulfide bonds are

essential to their structure and function. The SDF serine protease inhibitors are

69

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. reminiscent of those found in nematodes (Zang and Maizel, 2001). They are also of low

molecular weight, and have disulfide bonds which are thought to render them heat

resistant. Presumably the inhibitors protect the nematodes from proteolytic degradation

by the host’s digestive enzymes (Zang and Maizels, 2001). The serine protease inhibitors

of the spider digestive fluid are also of relatively low molecular weight and are heat

stable. They may serve to protect the spider digestive enzymes during the digestion of

prey.

Note: For the following reasons I believe that Mebs (1972) inadvertently described

spider digestive proteases rather than venom proteases. It seems likely to me that his

venom sample was contaminated by SDF. 1) He acknowledged that the venom was a gift

to him; he did not collect it. 2) Very little protease activity has ever been observed in any

spider venom which has been collected carefully so as to avoid contamination by SDF

(see Perret, 1977 and Kuhn-Nentwig et al., 1994 for reviews of this problem). 3) The

proteases he described are essentially identical in their biochemical features to those

described for SDF proteases (Kavanagh and Tillinghast, 1983).

The results of this study demonstrate that SDF is a complex fluid containing both

proteases and protease inhibitors. The proteases appear to self-assemble into clusters,

making their purification difficult. A greater knowledge of the protein-protein interaction

should help to develop a more effective means of purification. My results suggest that

SDF proteases are members of the astacin family of metalloproteases, and this accords

with earlier observations of the ion requirements of spider digestive enzymes. My

research has also revealed at least ten serine protease inhibitors in the SDF, whose

function remains unknown.

70

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF REFERENCES

Atkins, J.A., Wingo, C.W., Sodeman, W.A., Flynn, J.E., 1958. Necrotic arachnidism. Am. J. Trop. Med. Hyg. 7, 165-184.

Atkinson, R.K., Wright, L.G., 1991. Studies on the necrotic actions of the venoms of several Australian spiders. Comp. Biochem. Physiol. 98C, 441-444.

Atkinson, R.K., Wright, L.G., 1992. The involvement of collagenases in the necrosis induced by the bites of some spiders. Comp. Biochem. Physiol. 102C, 125-128.

Babcock, J.L., Civello, D.J., Geren, C.R., 1981. Purification and characterization of a toxin from brown recluse spider (Loxosceles reclusa) venom gland extracts. Toxicon. 19(5), 677-689.

Barbara, K.C., Cardosa, J.L.C., Eickstedt, V.R.D., Mota, I., 1992. Dermonecrotic and lethal components of Loxosceles gaucho spider venom. Toxicon. 30, 331-338.

Barbara, K.C., Eickstedt, V.R.D., Mota, I, 1994. Antigenic cross-reactivity of venoms from medically important Loxosceles (Araneae) species in Brazil. Toxicon. 32(1), 113-120.

Barbara, K.C., Sousa, M.V., Morhy, L., Eickstedt, V.R.D., Mota, I., 1996. Compared chemical properties of dermonecrotic and lethal toxins from spiders of the Loxosceles (Araneae). J. Prot. Chem. 15(4), 337-343.

Bettini, S., Brignoli, P.M., 1978. Review of the spider families, with notes on the lesser- known poisonous forms. In: Bettini, S. (Ed.), Handbook of Experimental Pharmacology, vol.48. Venoms. Springer-Verlag, New York, pp. 101- 120.

Bemfeld, P., 1951. Enzymes of starch degradation and synthesis. In: Nord, F.F. (Ed.) Advances in Enzymology, vol. 7. Interscience Publishers, New York N.Y. pp. 379.

Betkau, P., 1885. Uber den Verdauungsapparat der Spinnen. Arch mikrosk. Anat. EntwMech., 24, 398-451.

B inford, G., 2001. An analysis of geographic and intersexual chemical variation in venoms of the spider Tegenaria agrestis (Agelenidae). Toxicon. 39(7), 955-968.

71

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Broadway, R. M., 1993. Purification and partial characterization of trypsin/chymotrypsin inhibitors from cabbage foliage. Phytochemistry. 33(1), 21-27.

Cohen, E., Quistad, G. B., 1998. Cytotoxic effects of arthropod venoms on various cultured cells. Toxicon. 36(2), 353-358.

Corzo, G., Villegas, E., Gomez-Lagunas, F., Possani, L.D., Belokoneva, O.S., Nakajima, T., 2002. Oxyopinins, large amphipathic peptides isolated from the venom of the wolf spider Oxyopes kitabensis with cytolytic properties and positive insecticidal cooperativity with spider neurotoxins. J. Biol. Chem. 277(26), 23627-37.

Couch, T.L., Benton, A.W., 1971. Comparison of gut and hemolymph enzymes and proteins of several spider species. Ann. Ent. Soc. Am. 64, 409-413.

Croucamp, W., Veale, R., 1999. The venom of the sac spiderCheiracanthium furculatum Karsch, 1879. Toxicon. 37(10), 1335.

Dodge, J.T., Mitchell, C., Hanahan, D.J., 1963. The preparation and chemical characteristics of hemoglobin-free ghosts of human erythrocytes. Arch. Bioch. Biophys. 100, 119-130.

Elston, D.M., Eggers, J.S., Schmidt, W.E., Storrow, A.B., Doe, R.H., McGlasson, D., Fischer, J.R., 2000. Histologicaal findings after brown recluse spider envenomation. Am. J. Dermatopathol. 22, 242-246.

Farber, S.A., Olson, E.S., Clark, J.D., Halpem, M.E., 1999. Characterization of Ca2+- dependent phospholipase A2 activity during zebrafish embryogenesis. I. Biol. Chem. 274(27), 19338-19346.

Feitosa, L. W., Gremski, W., Veiga, S.S., Elias, M.C.Q.B., Graner, E., Mangili, O.C., Brentani, R.R., 1998. Detection and characterization of metalloproteinases with gelatinolytic, fibronectinolytic and fibrinogenolytic activities in brown spider (Loxosceles intermedia) venom. Toxicon 36, 1039-1051.

Foelix, R.F., 1996. Biology of Spiders. 2nd Ed. Oxford University Press, NY.

Foradori, M.J., Keil, L.M., Wells, R.E., Diem, M., Tillinghast, E.K., 2001. An examination of the potential role of spider digestive proteases as a causative factor in spider bite necrosis. Comp. Biochem. Physiol. 130C(2), 209-218.

Foradori, M.J., Kovoor, J., Moon, M.J., Tillinghast, E.K., 2002. Relation between the outer cover of the egg case of Argiope aurantia (Araneae: Araneidae) and the emergence of its spiderlings. J. Morphol. 252(2), 218-26

72

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ;

Forrester, L.J., Barrett, J.T., Campbell, B.J., 1978. Red blood cell lysis induced by the venom of the brown recluse spider: the role of sphingomyelinase D. Arch. Biochem. Biophys. 187, 355-365.

Furman, D.P., Reeves, W.C., 1957. Toxic bite of a spider, Cheiracanthium inclusum Hentz. Calif. Med. 87,114.

Futrell, J.M., 1992. Loxoscelism. Am. J. Med. Sci. 304, 261-267.

Gatt, S., Dinur, T., Barenholz, Y., 1978. A spectrophotometric method for determination of sphingomyelinase. Biochim. Biophys. Acta. 530(3), 503-507.

Geren, C.R., Chan, T.K., Ward, B.C., Howell, D.E., Pinkston, K., Odell, G.V., 1973. Composition and properties of extract of fiddleback (Loxosceles reclusa) spider venom apparatus. Toxicon. 11, 471-479.

Geren, C.R., Chan, T.K., Ward, B.C., Howell, D.E., Odell, G.V., 1976. Isolation and characterization of toxins from brown recluse spider venom (Loxosceles reclusa). Arch. Biochem. Biophys. 174, 90-99.

Gomez, H.F., Miller, M.J., Desai, A., Warren, J.S., 1999. Loxosceles spider venom induces the production of ot and (3 chemokines: Implication for the pathogenesis of dermonecrotic arachnidism. Inflammation. 23, 207-215.

Guilherme P., Fernandes, I., Barbara, K.C., 2001. Neutralization of dermonecrotic and lethal activities and differences among 32-35 kDa toxins of medically important Loxosceles spider venoms in Brazil revealed by monoclonal antibodies. Toxicon. 39(9), 1333-1342.

Hames, B.D., Rickwood, D., 1981. Gel electrophoresis of proteins: a practical approach IRL Press, Washington, D.C.

Hanahan, D.J., 1997. A Guide to Phospholipid Chemistry. Oxford University Press. New York.

Herrera, M.V., Odell, G.V., 1988. Characteristics of the venoms and digestive secretions of Aphonopelma seemanni and Sphaerobothria hojfmanni (Araneae: Theraphosidae) of Costa Rica. Rev. Biol. Trap. 36(2A), 241-246

Hipps, P.P., Nelson, D.R., 1974. Esterases from the midgut and gastric caecum of the American cockroach, Periplaneta americana (L.). Isolation and characterization. Biochim. Biophys. Acta. 341, 421-436.

Houghton, D.E.F., Ledsham, N., Pearson, R.G., Poliak, P.M., 1967. Esterase and phosphatase activity in the opisthosomal region of spiders. Nature. 213, 290-291.

73

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Hughes, S.E., 1981. Necrotic arachnidism induced by the envenomation of Chiracanthium mildei Koch (Araneae: Clubionidae). University of Massachusetts. Master’s thesis

Jong, Y-S., Norment, B.R., Heitz, J.R., 1979. Separation and characterization of venom components in Loxosceles reclusa - II. Protease enzyme activity. Toxicon. 17, 529-537.

Joo, H.S., Park, G.C., Cho, W.R., Tak, E., Paik, S.R., Chang, C.S., 2002. Purification and characterization of a prothrombin-activating protease from Nephila clavata. Toxicon. 40(3), 289-296.

Kaiser, E., Raab, W., 1967. Collagenolytic activity of snake and spider venoms. Toxicon. 4, 251-255.

Kavanagh, E.J.,Tillinghast, E.K., 1983. The alkaline proteases of Argiope-H Fractionation of protease activity and isolation of a silk fibroin digesting protease. Comp. Biochem. Physiol. 24B, 365-372.

Kuhn-Nentwig, L., Schaller, J., Nentwig, W., 1994. Purification of toxic peptides and the amino acid sequence of CSTX-1 from the multicomponent venom of Cupiennius salei (Araneae:Ctenidae). Toxicon. 32(3), 287-302.

Kuhn-Nentwig, L., Schaller, J., Kampfer, U., Imboden, H., Malli, H., Nentwig, W., 2000. A lysine rich C-terminal tail is directly involved in the toxicity of CSTX-1, a neurotoxic peptide from the venom of the spider Cupiennius salei. Arch. Insect Biochem. Phys. 44, 101-111.

Kuhn-Nentwig L, Dathe M, Walz A, Schaller J, Nentwig W., 2002a. Cupiennin Id*: the cytolytic activity depends on the hydrophobic N-terminus and is modulated by the polar C-terminus. FEBS Lett. 527(1-3), 193-8.

Kuhn-Nentwig L, Muller J, Schaller J, Walz A, Dathe M, Nentwig W., 2002b. Cupiennin 1, a new family of highly basic antimicrobial peptides in the venom of the spider Cupiennius salei (Ctenidae). J. Biol. Chem. 277(13), 11208-16.

Kunitz, M., 1947. Crystalline soybean trypsin inhibitor II. General proteases. J. Gen. Phys. 30, 291-302.

Kurpiewski, G., Forrester, L.J., Barrett, J.T., Campbell, B.J., 1981. Platelet aggregation and sphingomyelinase D activity of a purified toxin from the venom of Loxosceles reclusa. Biochim. Biophys. Acta. 678, 467-476.

Krinsky, W.L., 1987. Envenomation by the sac spider Chiracanthium mildei. Cutis. 40(2), 127-129.

74

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Laemmli, U.K., 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature. 227, 680-685.

Leech, R., Brown, T.M., 1994. The first Alberta record for Cheiracanthium inclusum (Hentz) (Araneida:Clubionidae), with observation on a human envenomation reaction. The Canadian Entomologist. 126(1), 187.

Luna, L.G., 1979. Manual of Histological Staining of the Armed Forces Institute of Pathology, Third Edition. McGraw-Hill Book Company, New York.

Lowry, O.H., Rosebrough, N.J., Farr, A.L., Randall, R.J., 1951. Protein measurements with Folin phenol reagent. J. Biol. Chem. 193, 265-275.

Macchiavello, A., 1937. La Loxosceles laeta, causa del aracnoidisma cutaneo manoha gangrenosa, de Chile. Rev. Chelena. de His. Nat. 41, 11-19.

, Maretic, Z., 1982. Some clinical and epidemiological problems of venom poisoning today. Toxicon. 20(1), 345-348.

Matsudaira, P.T., 1989. A Practical Guide to Protein and Peptide Purification for Microsequencing. Academic Press, N.Y.

Mebs, D., 1970. Proteolytische Aktivitaet eines Vogelspinnengiftes (Pamphobeteus roseus, Dipluridae). Naturwissenschaften. 57, 308.

Mebs, D., 1972. Proteolytic activity of a spider venom. In; DeVries, A., Kochva, E. (Eds.) Toxins of Animal and Plant Origin, vol. II. Gordon and Breach, New York, pp. 493-496

Minton, S.A., 1972. Poisonous spiders of Indiana and a report of a envenomation by Chiracanthium mildei. J. Indiana. State Med. Assoc. 65(5), 425-6.

Mommsen, T.P., 1979a. Digestive enzymes of a spider (Tegenaria atrica Koch) - 1. General remarks, digestion of proteins. Comp. Biochem. Physiol. 60A, 365-370.

Mommsen, T.P., 1979b. Digestive enzymes of a spider (Tegenaria atrica Koch) - II. Carbohydrases. Comp. Biochem. Physiol. 60A, 371-375.

Mommsen, T.P., 1979c. Digestive enzymes of a spider (Tegenaria atrica Koch) - III. Esterases, Phosphatases, nucleases. Comp. Biochem Physiol. 60A, 377-382.

Mommsen, T.P., 1978d. Comparison of digestive a-amylase from two species of spiders (Tegenaria atrica) and Cupiennius salei). J. comp. Physiol. 127 B, 355-361.

Mommsen, T.P., 1980. Chitinase and b-N-acetylglucosaminidase from the digestive fluid of the spider Cupiennius salei. Biochim. Biophys. Acta. 612, 361-372

75

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Newlands ,G., Martindale, C., Berson, S.D., Rippey, J.J., 1980. Cutaneous necrosis caused by the envenomation of Chiracanthium spiders. S. Afr. Med. J. 57(5), 171-173.

Owri, M., 1978. Envenomation of Chiracanthium japonicum and the properties of the spider venom [proceedings]. Jpn. J. Med. Sci. Biol. 31(2), 200-203.

Pedersen K.O., 1950 Size relationship among similar proteins. Association and dissociation reactions of protein units. Cold Spring Harbor Symposia on Quantitative Biology XIV The Biological Laboratory. Cold Spring Harbor, L.I. New York.

Pearson, W.R., Lipman, D.J., 1988. Improved Tools for Biological Sequence Comparison. PNAS. 85, 2444- 2448.

Pearson, W.R., 1990. Rapid and Sensitive Sequence Comparison with FASTP and FAST A. Methods in Enzymology. 183, 63-98.

Perret, B.A., 1977. Proteolytic activity of tarantula venoms due to contamination with saliva. Toxicon. 15, 505-510.

Peters, G., Saborowski, R., Mentlein, R., Buchholz, F., 1998. Isoforms of an N-acetyl- b-D-glucosaminidase fro the antarctic krill, Euphausia superba: purification and antibody production. Comp. Biochem. Physiol. 120B, 743-751.

Pfleiderer, G.R., Zwilling, R., Sonneborn, H.H., 1967. Eine Protease vom Molekulargewicht 11,000 und eine trypsinahnliche Fraktion aus Astacus fluviatus Fabr. Hoppe-Seyler’s Z. Physiol. Chem. 348, 1319-1331.

Phillips, J., Dresden, M.H., 1973. A collagenase in extracts of the invertebrate Bipalium kewense. Biochem. J. 133, 329-334.

Pickford, G.E. 1942. Studies on the digestive enzymes of spiders. Trans. Conn. Acad. Arts. Sci. 35, 33-72.

Pincus, G.L., Winkel, K.D., Hawdon, G.M., Sutherland, S.K., 1999. Acute and recurrent skin ulceration after spider envenomation. Med. J. Aust. 171, 99-102. I Plateau, F., 1877. Recherches sur la structure de l’appareil digestif et sur le phenomenes de la digestion chez les araneides dipneumones. Bull. Acad. R. Belgique. 44, 129

Pollack, P.M., 1966. Esterase activity in the arachnida. Nature. 211, 546-547.

76

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Rees, R.S., Nanney, L.B., Yates, R.A., King, L.E., 1984. Interaction of brown recluse venom on cell membranes: the inciting mechanism? J. Invest. Derm. 83, 270- 275.

Rekow, M.A., Civello, D.J., Geren, C.R., 1983. Enzymatic and hemolytic properties of brown recluse spider (Loxosceles reclusa) toxin and extracts of venom apparatus, cephalothorax and abdomen. Toxicon. 21(3), 441-444.

Russell, F.E., Gertsch, W.J., 1983. For those who treat spider or suspected spider envenomations. Toxicon. 21, 337-339.

Schlottke, E., 1936. Uber die Verdauungsfermente der Vogelspinnen. Sber. Abh. Naturf. Ges Rostock 3. 6, 89-105.

Schumacher, G. F. B., Schill W. B., 1972. Radial; diffusion in gel for micro determination of enzymes. II Plasminogen activator, elastase and nonspecific proteases. Anal. Biochem. 48, 9-26.

Smith, C.W., Micks, D.W., 1968. A comparative study of the venom and other components of three species of Loxosceles. Am. J. Trop. Med. Hyg. 17, 651- 656.

Spielman, A., Levi, H.W., 1970. Probable envenomation by Chiracanthium mildei', a spider found in houses. Am. J. Trop. Med. Hyg. 19, 729-732.

Stocker, W., Zwilling, R., 1995. Astacins, In: Barrett, A J. (Ed.), Methods in Enzymology, vol 248. Proteolytic Enzymes: Aspartic and Metallo Peptidases. Academic Press, New York, pp. 305 - 325.

Sugita, H. , Sekiguchi, K. 1975. Heterogeneity of the minimal functional unit of hemocyanins from the spider {Argiope bruennichii), the scorpion (Heterometrus sp.), and the horseshoe crab {Tachypleus tridentatus). J. Biochem. 78, 713-718.

Tambourgi, D.V., Magnoli, F.C., van den Berg, C.W., Morgan, B.P., deAraujo, P.S., Alves, E.W., Da Silva W.D., 1998. Sphingomyelinases in the venom of the spider Loxosceles intermedia are responsible for both dermonecrosis and complement-dependent hemolysis. Biochem. Biophys. Res. Comm. 251, 366- 373.

Tillinghast, E.K., E.J. Kavanagh., 1977. The alkaline proteases of Argiope and their possible role in web digestion. J. Exp. Zool. 202, 213-222.

Townley, M.A., Tillinghast, E.K., 1988. Orb web recycling in Araneus cavaticus (Araneae, Araneidae) with and emphasis on the adhesive spiral component GAB amide. J. Arachnol. 16, 303-319.

77

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Touchstone, J.C., Chen, J.C., Beaver, K.M., 1980. Improved separation of phospholipids in thin layer chromatography. Lipids. 15(1), 61-62.

Tu. A.T., 1977. Venoms: chemistry and molecular biology. New York : Wiley.

Tugmon, C.R., Tillinghast, E.K., 1995. Proteases and protease inhibitors of the spider Argiope aurantia (Araneae, Araneidae). Naturwissenschaften. 82(4), 195-197 ( Usmanov P.B, Nuritova, F.A., 1994. The anticoagulant action of phospholipase A from Eresus niger spider venom. Toxicon. 32(5), 625-628

USP 24-NF 19., 2000. The United States Pharmacopeia - The National Formulary. National Publishing, Philadelphia, PA. pp. 1834-1835.

Veiga, S.S., da Silveira, R.B., Dreyfuss, J.L., Haoach, J., Pereira, A.M., Mangili O.C., Gremski, W., 2000. Identification of high molecular weight serine-proteases in Loxosceles intermedia venom. Toxicon. 38, 825-839.

Vest, D.K., 1987a. Necrotic arachnidism in the northwest United States and its probable relationship to Tegenaria agrestis (Walckenaer) spiders. Toxicon. 25(2), 175- 184,

Vest, D.K., 1987b. Envenomation by Tegenaria agrestis (Walckenaer) spiders in rabbits. Toxicon. 25(2), 221-224.

Vetter, R.S., 2000. Myth: Idiopathic wounds are often due to brown recluse or other spider envenomations throughout the United States. West. J. Med. 173, 357-358.

Vetter, R'.S, Visscher, P.K., 1998. Envenomations and stings of medically important venomous arthropods. Int. J. Dermatol. 37(7), 481-496.

White, J.A., 1999. Necrotising arachnidism: Does the white-tailed spider deserve its bad name? Med. J. Aust. 171, 98.

White, J.A., Cardosa, J.C., Fan, H.W., 1995. Clinical Toxicology of Spider Envenomations. In: Meier, J., White, J.A. (Eds.) Handbook of Clinical Toxicology of Animal Venoms and Poisons. CRC Press, Boca Raton, pp. 261- 329.

White, J.A., Hirst, D., Hender, E., 1989. 36 cases of envenomations by spiders, including the white-tailed spider, Lampona cylindrata. Med J Aust. 150(7), 401-403.

Wolz, R.L., Bond, J.S., 1995. Meprins A and B. In: Barrett, A.J. (Ed.), Methods in Enzymology, vol 248. Proteolytic Enzymes: Aspartic and Metallo Peptidases. Academic Press, New York, pp. 325 - 345.

78

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Wright, R.P., Elgert, K.D., Campbell, B.J., Barrett, J.T., 1973. Hyaluronidase and esterase activities of the venom of the poisonous brown recluse spider. Arch. Biochem. Biophys. 159, 415-426.

Yan, L., Adams, M.E., 1998. Lycotoxins, antimicrobial peptides from venom of the wolf spider Lycosa carolinensis. J. Biol. Chem. 273(4), 2059-66. ; Yang, J.S., Ru, B.G., 1997. Purification and characterization of an SDS activated fibrinolytic enzyme from Eisenia foetida. Comp. Biochem. Physiol, 118B, 623- 631

Zang, X., and Maizels, R.M., 2001. Serine proteinase inhibitors from nematodes and the arms race between host and pathogen. Trends in Biochemical Science. 26, 191- 197.

Zwilling, R., Neurath, H., 1981. Invertebrate Proteases In: Colowick, P.S.; Kaplan, N.O., (Eds) Methods in Enzymology, vol. 80. Academic Press, New York. pp. 633-664.

79

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX

80

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. JjU §*», NIYERSITY y OF-JNIEW T v i ■v? & t JrlAMPSHIRE 1 . jT a ~sl *&*yyr*sr~%, Tg^T Office o tSponsored Research

mm mum « if c g t*y FIRST MAME {Stew*

man a pp'l da te a g a i n

OWJMMR® IACBC # . s m s A»aiss {if «ppMM«; « f iK W L E m m

RR©«€T TITLE t e s t e in tabbiis and the |te ® fe ta ^ « te B of ifcg t e h » i ^ |p s ^ r pretease M t e i s lt e t aM f m hamoiymptt

4 1 ««#«} j m M M W a f l w t t wrecords must meted# yettrM CiJC P roiom i# « flrlwtabm e.

The Imtitatioiial Animal Care and Use Committee has reviewed and approved the protocol submitted for this study under Category 2 on Page 3 Of the “Application far Review of Animal Use or Instmaion ProtoeolH - the research potentially involves minor tb a f c to m pats, discomfort or dlsttess which w il l* treated with appropriate anesthetics/analgesics.

Approval is pantedfor a period ofte e yeas fpa»the approval date above. Continued approval thma^aoat the te e yearperiod is contingent upon com plete ofannual tspoas on the use of animals. At the end o f the te e year approvalperiod you may submit a new application and request for extension to continue te fta p e i. -fepeste for extension most be filed prior to the expiration of tire original approval.

For the Institutional Animal Cate aid Use Commi ttee,

Kara L. BMy, MJLA, Manager, Regulatory Compliance

81

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. U n iv er sity o f N e w H a m p sh ir e Gffies «f Spoil Wed .Research SetvfajeBuading 51 College Road Durham, Nev. Hampshire 03S24-35e5 w y m a m F A X '

IAS7 NAME RRSTNAME

0EPT NEXT REVIEW DATE

OFF-CAMPUS lACUC f ADDRESS (If applicable) REVIEW LEVEL I

DATE OF NOTICE

PROJECT i t spiders wife necrotizing venoms TITUs

The instiHrfopai Animat Care and Use Committee {iACUC} has reviewed and approved your request for a time extension fcr this protocol. Approval is granted until the "Next Review Date" indicated above. You will be asked to subih't s orcjcct roport witn regard to tie mvoivument cr animals before that date. If your project Is still active, you may apply fot exteusmfi of IACUC approval w ougb ihis olnce.

The appropriate use and care of animats In your project is an ongoing process for which you hold primary responsibility. Changes in your protocol must be submitted to the IACUC tor review and approval prior to their implementation. If you nave questions or concerns about your project or this approval, please feel free to contact the Regulatory Compliance Office at 862-2003 or 882-3536.

Please note; Use of anraais in research and instruction is approved contingent upon participation in fee UNH Occupational Health Program for persons handling animate. Participation is mandatory ior all principal investigators and tneir affiliated paraon'wl, employees of fee University and students alike. A Medical History Questionnaire accompanies tnis approval, ploasa copy and oistnoute to sll listed project staff who have not completed thisfcrrr. already. Completed questionnaires should be sent to Dr. Giadi Porsche, UNH Health Services.

Please refer to fee IACUC # above in ail correspondence related to this project. The IACUC wishes you success wife your research.

Foofee IACUC, f t i

Jbhn A, Utvaitfe, Ph.D. Chair

cc: File . .. grig apfl ti»8ip#ea

82

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.