INFORMATION TO USERS

The most advanced technology has been used to photograph and reproduce this manuscript from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.

The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.

In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand corner and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book.

Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.

University Microfilms International A Bell & Howell Information Company 300 North Zeeb Road, Ann Arbor, Ml 48106-1346 USA 313/761-4700 800/521-0600 Order Number 00S1079

The survival ofStaphylococcus aureus in abscesses from streptozotocin-induced diabetic mice

Harvey, Kevin Michael, Ph.D. The Ohio State University, 1990

Copyright ©1990 by Harvey, Kevin Michael. All rights reserved.

UMI 300N.ZeebRd. Ann Arbor, MI 48106 THE SURVIVAL OF STAPHYLOCOCCUS AUREUS IN ABSCESSES FROM

STREPTOZOTOCIN-INDUCED DIABETIC MICE

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the

Degree Doctor of Philosophy in the Graduate School of The

Ohio State University

By

Kevin Michael Harvey, B.A.

*******

The Ohio State University

1990

Dissertation Committee: Approved by:

Frank A. Kapral, Ph.D.

Bernard U. Bowman, Jr., Ph.D.

Thomas F. DeMaria, Ph.D.

Abramo C. Ottolenghi, Ph.D. , Adviser Jepartment of Medical Norman L. Somerson, Ph.D. Microbiology and Immunology Copyright by Kevin Michael Harvey 1990 DEDICATION

To my wife, Betsy, whose love and faith provided the encouragement to reach this point. To my sons, Christopher and Brendan, for making the work worth the effort, and to my family especially my parents, for their love and support. ACKNOWLEDGEMENTS

I would like to thank my advisor, Dr. Frank A. Kapral, for his patience and guidance. My sincerest thanks to my committee members, Dr. Norman L. Somerson, Dr. Abramo C. Ottolenghi, Dr.Thomas F. DeMaria, and Dr. Bernard U. Bowman, for their suggestions and comments. I would also like to thank Judy Hart and Shelley Smith for their technical assistance and friendship. My gratitude is extended to the support staff in the "kitchen" and office for all their help. I also wish to thank all my colleagues who started their research with me especially, EAF, ALL, JLP, XZ, and GL, for their friendship. VITA

May 17, 1958...... Born - Pittsburgh, Pennsylvania

1980...... B. A. cum laude. University of North Texas, Denton, Texas

1980-198 1...... Junior Chemist, Kolmar Laboratories, Inc., Denton, Texas

1981-198 4...... Chemist, Sasco Cosmetics, Inc., Carrollton, Texas

1984-Present...... Graduate Research Associate, Department of Medical Microbiology and Immunology, The Ohio State University, Columbus, Ohio

PUBLICATIONS

Harvey, K. M., J. Hart, and F. A. Kapral. Survival of Staphylococcus aureus in intraperitoneal abscesses from diabetic mice. In: Jelijaszewicz, J. (ed) The Staphylococci. Zbl. Bakt. Suppl. 15. Gustav Fischer Verlag, Stuttgart (In Press).

FIELDS OF STUDY

Major Field: Medical Microbiology and Immunology

Major Area of Study: Host-Parasite Interactions

iv TABLE OF CONTENTS

DEDICATION ...... ii

ACKNOWLEDGEMENTS...... iii

VITA ...... iv

TABLE OF CONTENTS...... V

LIST OF T A B L E S ...... vii

LIST OF FIGURES...... ix

LIST OF P L A T E S ...... xi

CHAPTER

I. INTRODUCTION ...... 1

II. MATERIALS & METHODS...... 15 M i c e ...... 15 Staphylococcal strains and animal inoculation...... 15 Determination of blood glucose ...... 16 Generation of diabetic m i c e ...... 17 Evaluation of staphylococcal survival within a b s c e s s e s ...... 18 Bactericidal assay ...... 18 Histological sections ...... 19 Lipid e x t r a c t i o n ...... 20 Fractionation of lipids ...... 21 Collection of individual lipids ...... 22 Esterification of free fatty acids and gas chromatographic analysis ...... 23 Trimethylsilyl ether derivatives of monoglycerides and gas chromatographic analysis...... 24 Determination of glucose concentration in intraperitoneal abscesses ...... 26 Chemicals and reagents ...... 27 III. RESULTS...... 28 Generation of diabetic m i c e ...... 28 Survival of staphylococci in abscesses within diabetic mice ...... 29

vi Histology of intraperitoneal staphylococcal a b s c e s s e s ...... 32 Analysis of lipids in intraperitoneal abscesses...... 33 Glucose concentration within abscesses . . . 35

IV. DISCUSSION ...... 38

BIBLIOGRAPHY ...... 88 LIST OF TABLES

Table Page

1. Average blood glucose (mg/dl ± SD) before and after inoculation with streptozotocin (STZ) or citrate (CIT) ...... 54

2. Percent positive mice made diabetic with streptozotocin and controls receiving citrate...... 55

3. Average blood glucose (mg/dl ± SD) before and after infection with S. aureus 18Z ...... 57

4. Total lipids extracted from S. aureus 18Z intraperitoneal abscesses at 5 d a y s .... 73

5. Total lipids extracted from 10 day S. aureus 18Z intraperitoneal abscesses ...... 74

6. Total triglycerides extracted from 5 day S,. aureus 18Z intraperitoneal abscesses .... 75

7. Total triglycerides extracted from 10 day S. aureus 18Z intraperitoneal abscesses . . . 76

8. Composition of free fatty acid fraction extracted from 5 day S. aureus 18Z intraperitoneal abscesses ...... 77

9. Composition of free fatty acids extracted from 10 day S. aureus 18Z intraperitoneal abscesses...... 78

10. Free fatty acids extracted from 5 day S. aureus 18Z intraperitoneal abscesses . . . 79

11. Free fatty acids extracted from 10 day S. aureus 18Z intraperitoneal abscesses . 80

12. Total amount of monoglycerides extracted from 5 day S. aureus 18Z intraperitoneal abscesses...... 81

vii 13. Total monoglycerides extracted from 10 day S. aureus 18Z intraperitoneal abscesses . . . 82

14. Total unsaturated and saturated monoglycerides extracted from S. aureus 18Z intraperitoneal abscesses at 5 and 10 days . 83

15. Sensitivity of Staphylococcus aureus strains TG and 303 grown in different media to oleic acid and 1- and 2-monoolein...... 84

16. Evaluation of lipemic effect of abscess homogenate on Glucostix (Ames) ...... 87

*

viii LIST OF FIGURES

Figure Page

1. Fasting blood glucose concentrations of mice before and after treatment with streptozotocin or citrate ...... 56

2. The survival of Staphylococcus aureus strain 18Z in the abscesses from control and diabetic m i c e ...... 64

3. Survival curves of s. aureus strain PG114 in abscesses of control and diabetic mice . . . 65

4. The survival of S. aureus strain P78 in intraperitoneal abscesses from control and diabetic animals ...... 66

5. The survival of S. aureus strain P78-22 in the abscesses from control and diabetic animals...... 67

6. Bactericidal activity of the total lipid pool extracted from the abscesses of control and diabetic mice 5 days post infection with S. aureus 1 8 Z ...... 68

7. Bactericidal activity of the total lipid pool extracted from 5 day old abscesses of control and diabetic mice infected with S. aureus 1 8 Z ...... 69

8. Bactericidal activity of the total lipid pool extracted from the abscesses of control and diabetic animals 10 days post infection with S. aureus 1 8 Z ...... 70

9. The bactericidal activity of total lipids extracted from abscesses of control or diabetic mice at 10 days post infection with S. aureus strain 1 8 Z ...... 71

10. Fasting blood glucose concentrations of mice infected with S. aureus strain P78-22 .... 85

ix 11. The glucose concentration of S. aureus strain P78-22 abscesses in control and diabetic m i c e ...... 86

x LIST OF PLATES

Plate Page

I. Ixnmunocytochemical staining of pancreas islets from a normal mouse using the indirect immunoperoxidase technique to detect insulin 58

II. Immunocytochemical staining of pancreas islets from a normal mouse using the indirect immunoperoxidase technique to detect glucagon ...... 59

III. Immunocytochemical staining of pancreas islets from a normal mouse using the indirect immunoperoxidase technique to detect somatostatin ...... 60

IV. Immunocytochemical staining of pancreas islets from a diabetic mouse using the indirect immunoperoxidase technique to detect insulin ...... 61

V. Immunocytochemical staining of pancreas islets from a diabetic mouse using the indirect immunoperoxidase technique to detect glucagon ...... 62

VI. Immunocytochemical staining of pancreas islets from a diabetic mouse using the indirect immunoperoxidase technique to detect somatostatin ...... 63

VII. Ten day abscess sections stained with Oil Red 0 to detect lipid ...... 72

xi CHAPTER I

INTRODUCTION

In a lecture to the German Surgical Society in April

1880, and again a year later to the British Medical

Association, Sir Alexander Ogston presented convincing evidence to resolve the ongoing controversy about the role of micrococci in pyemia and septicemia. Ogston : ' differentiated between Billroth's streptococci or "chain micrococci" and the "grouped micrococci" he called staphylococci (1). He showed that when either pus containing staphylococci or organisms cultured from pus were inoculated subcutaneously into mice, disease symptoms similar to those in man were observed. There was no infection if the pus was treated with phenol or heated prior to infecting the animals (2). Ogston's publications offered foresight in future staphylococcal research. He made references to the ability of the organism to elicit toxins from sites of infection and noted that in older lesions the number of organisms declined, suggesting a relationship between the host and organism.

1 In the 110 years following Sir Ogston's observations, research has persistently provided more information about pyogenic infections caused by staphylococci. These organisms continue to have an impact on morbidity and mortality even today. One of the clinically significant, and most often studied, species of staphylococci is

Staphylococcus aureus. £!. aureus has remained a potent pathogen by acquiring antimicrobial resistance, even with the development of antistaphylococcal antibiotics over the last 40 years. £i. aureus is responsible for a spectrum of disease ranging from intoxication to local superficial or deep tissue infections. Staphylococcal infections are characterized by abscess formation. There is collection of pus, local tissue necrosis and the area is quickly surrounded by a thick-walled fibrin capsule (1). A variety of toxins can be elaborated from these abscesses. These virulence factors of the staphylococci are responsible for the symptoms associated with the different disease states.

Although much is known about S. aureus infections many gaps remain in our understanding the pathogenesis of staphylococcal infections. '

In order to study this host-pathogen relationship an acceptable model of a localized lesion is necessary. Many researchers have tried experimental models that include: incisions (3,4); damaging tissues prior to infection (5); introducing staphylococci impregnated on foreign material (6,7,8,9,10,11); and large inocula (12). Although these conditions result in staphylococcal infection, the naturally occurring conditions whereby a small number of cocci multiply and cause a disease, is not duplicated.

Kapral and others have described an appropriate model that produces a typical staphylococcal abscess (13). Mice can be infected intraperitoneally with a large inoculum of non-encapsulated staphylococci that possess clumping factor and a typical staphylococcal lesion is generated. These abscesses are easily removed from the peritoneal cavity and can be freed of extraneous host tissues. The abscesses are then available for a detailed study of the host-organism interactions.

The first step in the formation of an abscess in the peritoneal cavity is the production of a clump of organisms.

Since clumping is a requisite first step for organism survival and abscesses formation, a large dose of organism, usually 109, must be introduced into the mouse peritoneal cavity. When a dose of 109 or more non-encapsulated cocci possessing clumping factor (bound coagulase) is administered into the peritoneal cavity of mouse the clumping factor interacts with the fibrinogen present in the peritoneal fluid and a clump of organisms is formed (13). If less than

109 non-encapsulated cocci are inoculated the bacteria remain dispersed and are quickly phagocytized. If 10* or more encapsulated cocci are used to infect a mouse, the clumping factor is not exposed and can not interact with the fibrinogen to form a clump of organisms. However, the capsule protects the cocci from phagocytosis and the organism is able to elaborate enough toxins to kill the host

(13).

Soon after a clump is formed, neutrophils arrive at the site, but phagocytosis is minimal because only the cocci at the periphery of the clump are exposed to the neutrophils

(14). Before the leukocytes completely surround the clump, toxins (in particular alpha toxin) produced by the staphylococci are released into the peritoneal cavity. If enough toxins are elaborated into the host before the leukocytes enclose the clump, the host will die. Normally with the dose of organisms required to generate a clump, enough toxins can be produced to kill the animal. However, by prior growth of the organisms in the absence of carbon dioxide, subsequent toxin production can be sufficiently retarded so that an inoculum large enough to allow clumping to occur can be administered without killing the animal

(13).

Five to six hours post infection the clumps of organisms together with the leukocytes consolidate to form 1 or 2 large clusters that contain almost all of the initial inoculum. Fresh neutrophils continue to arrive and adhere to the periphery of the clusters and the lesion continues to enlarge. Deposition of a connective tissue capsule around the cluster begins at 24 hours post infection (14) and continues until the 4th day when the capsule has become vascularized and a typical staphylococcal abscess is formed.

The abscess consists of, beginning at the interior, a core of tightly packed cocci, a broad layer of acellular debris, a region of disintegrating leukocytes, a zone of intact leukocytes and finally the vascularized connective tissue capsule at the periphery of the abscess.

Previous work by Dye and others has shown that within the abscess the cocci exhibit one of three basic survival patterns (14,15). With some strains destruction begins early and continues progressively. Conversely other strains persist at initial population levels for long periods of time. With still other strains the population within abscesses first declines but then increases to almost original inoculum levels. This rebound phenomenon can occur between five and 15 days post infection. After the rebound the number of cocci decreases continuously over time.

Efforts were made to determine how th& host eliminates the staphylococci in the abscesses. Phagocytosis was not a good explanation because there are no intact leukocytes in the region of viable staphylococci and where the leukocytes are located there are no staphylococci. When abscesses were homogenized, bactericidal activity was found in the homogenates (16). The bactericidal material was able to kill certain strains of S. aureus more readily than other strains. This phenomenon was referred to as "differential

activity".

Furthermore there was a correlation between a strain's

sensitivity to the bactericidal substance and its ability to

survive in an abscess. Strains that are easily cleared from

abscesses are very sensitive to the substance, whereas,

strains that persist in lesions are more resistant. The

strains exhibiting the delayed clearance of survival have an

intermediate sensitivity (15).

The bactericidal material was differentiated from the

myeloperoxidase system of leukocytes by its resistance to

heat, catalase and pH extremes (16). The substance was

distinguished from cationic proteins by its lack of

solubility in weak acid, its insensitivity to iron and its

increased activity in 1 M sodium chloride (16). The material was heat stable and was found in the insoluble

fraction of the abscess homogenate. These properties

suggested that the material was lipid and this was verified

by demonstrating that the bactericidal activity was

recovered in the lipid fraction extracted from abscess

homogenates.

With the realization that the bactericidal substance was a lipid, further studies were done to examine where

lipids amass in an abscess (17). Frozen sections of abscess

taken at various stages of abscess development were stained with Oil Red 0 to study lipid accumulation in the lesions. Lipids were not detected until 12 hours after infection with the cocci. At that time lipids were seen at the periphery of the leukocytic layer. As abscess formation continued, and lipid concentrations increased, the largest extracellular lipid droplets were always observed in the area just beneath the connective tissue capsule. Smaller droplets were dispersed throughout the abscess even towards the core of the lesion where the viable staphylococci are located. One week after infection, the distribution of lipids remains unchanged with the greatest concentration of lipids just beneath the connective tissue and lesser amounts dispersed throughout the abscess (17).

Concomitant studies testing the abscess homogenate for bactericidal activity over time showed no measurable activity prior to 12 hours post infection. The amount of activity increased rapidly soon after, with peak levels being observed between seven and ten days post infection.

After the tenth day the bactericidal activity began to decrease (17).

Lipids were extracted from abscesses recovered at different time intervals after abscess formation and the amount of lipid was measured gravimetrically. Peak levels of total lipid were attained by day seven. When the lipids were then fractionated into phospholipids, glycolipids and neutral lipids classes, it was found that the neutral lipids were responsible for most of the increase seen in the amount of total lipids. Only smaller increases in phospholipids and glycolipids were found (17).

Further fractionation of the neutral lipids revealed that the bactericidal activity was present in the pool of long chain unsaturated free fatty acids (18) and the pool of monoglycerides (19). Individual fatty acids in the free fatty acid pool act equally well on different strains of S. aureus. and thus do not exhibit differential activity.

However, the monoglycerides, particularly the 2- monoglycerides, exhibit the differential activity associated with the original abscess homogenate. The 1-monoglycerides, although bactericidal, exhibit only slight differential activity.

The neutral lipid portion of abscess homogenates consists of triglycerides, cholesterol, free fatty acids, diglycerides and monoglycerides. Over 99% of the free fatty acids were identified by gas chromatography (17). The pool consists mostly of the long chain free fatty acids oleic, linoleic, palmitic and smaller amounts of palmitioleic, myristic, stearic and arachidonic acids. Although the unsaturated free fatty acids are bactericidal, none of these, either alone or in combination, possesses the differential activity seen with the monoglyceride pool (17).

When the monoglycerides were hydrolyzed, and the resultant acyl groups analyzed by gas chromatography, the fatty acids moieties were found to be the conventional C16-C„ acids found

in the free fatty acid pool present in the abscesses (18).

The staphylococci are capable of inactivating all the aforementioned bactericidal lipids in the abscess by producing an enzyme, fatty acid modifying enzyme (FAME)

(19, 20). FAME can esterify the free fatty acids and the

fatty acid moieties of the monoglycerides, with cholesterol

from cell membrane, thus rendering the lipids ineffective

(19, 20). The triglycerides which comprise a majority of the neutral lipid fraction, can inhibit the activity of FAME

(20). The cocci in turn, can elaborate lipase which degrades the triglycerides. This elaborate system of measures and countermeasures has provided a new insight into the host-organism interactions.

The most common staphylococcal infection is an abscess.

Normally, the host eventually eliminates the organisms from such lesions. However, some individuals demonstrate an inability to control the organism such that proliferation can continue and serious local or systemic disease may occur. Problems in the individual's host defense systems frequently exist in these cases. Some of these problems include: injury to normal skin such as a burn or wound, prior viral infection, defects with the leukocytes, humoral immunity deficiencies, the presence of a foreign body like a catheter or suture, or the presence of underlying illnesses 10 such as alcoholism, coronary artery disease, cystic fibrosis, or diabetes mellitus.

In addition to being susceptible to staphylococcal infections, individuals with diabetes mellitus have an altered lipid metabolism. Models studying staphylococcal infections in diabetic animals have been used in the past.

None of these studies, however, has examined the interaction of the host's bactericidal lipids produced in a lesion and the staphylococci. Information about differences in the aforementioned bactericidal lipids might be determined by comparing a staphylococcal infection in a mouse model of diabetes mellitus to normal mice.

There are two forms of diabetes mellitus, type I or insulin dependent (IDDM) and type II or non-insulin dependent (NIDDM). Recent studies have shown that there is a strong correlation between IDDM and certain HLA types.

Genetic susceptibility to IDDM is positively correlated with a neutral residue at position 57 of the HLA-DQB chain (21).

Although more common than type I, type II diabetes is still far from being understood. Genetic mechanisms, however, do appear to be important in the disease (22).

Both forms are characterized by insulin deficiency in the body. The principal role of insulin is to control the transport of glucose, the body's main energy source, from the bloodstream into target cells. With IDDM the pancreatic beta cells, which manufacture insulin, are 11 destroyed thus resulting in a deficiency of circulating insulin. A relative insensitivity of cells to insulin rather than a lack of insulin production, characterizes

NIDDM. Target cell resistance occurs in both forms of diabetes but is a central feature of type II diabetes (23).

The metabolic abnormalities that result from this insulin deficiency include a decreased utilization of glucose, which because it cannot enter cells, leads to an increase in the blood glucose concentration. The resultant hyperglycemia leads to the excretion of large volumes of urine which in turn leads to an intense thirst. There is also a marked increase in fat mobilization for energy production which causes abnormal lipid metabolism and deposition of fatty acid intermediates (ketone bodies) in the blood plasma. In addition, there is depletion of proteins in the body.

The acute symptoms of NIDDM are usually milder than those of IDDM, however, individuals with either form of diabetes are subject to the severe complications of the disease (24). These complications include kidney damage, retina destruction, atherosclerosis leading to heart disease and stroke, cataracts, neurological dysfunction, and a predisposition to infection from fungi, such as Candida albicans and bacteria, such as Staphylococcus aureus.

It has long been observed that individuals with diabetes mellitus are more susceptible to staphylococcal infections than are normal individuals. Much speculation has been given to this correlation. Early observations suggested that high blood sugar levels favored the growth of the cocci (1). Mice were challenged intraperitoneally with various S. aureus strains suspended in mucin. When the challenge was followed by intraperitoneal inoculation of glucose, the minimal lethal dose of the staphylococci was reduced (25). These conditions, however, are artificial.

When rabbits were given a high dose of glucose following challenge with staphylococci, an increase in the extent of experimental skin lesions was observed. However, similar results were observed when the rabbits were given sodium chloride of the same tonicity, thereby indicating that a specific effect of glucose was not the chief factor involved

(26). No convincing evidence supporting the theory that increased glucose levels, per se. are responsible for an increased susceptibility to staphylococcal infections has been made (1).

Another hypothesis proposed suggested that the presence of keto acids protect the cocci within polymorphonuclear leukocytes from the bactericidal effects of lactic acid

(27). The effects of keto acids on the staphylococci were only evaluated in vitro and ,/ere not tested in an actual abscess.

More recently, defects in the diabetic's cellular or humoral immunity have been examined as possible explanations 13 for this increased susceptibility. Production of antibody to a staphylococcal toxin in diabetic children is slower and the levels are not as high compared to normal controls (28).

Although phagocytosis and chemotaxis of leukocytes are impaired in diabetics in the presence of high glucose levels

(29, 30), this is more of an osmotic effect and can be demonstrated with a variety of sugars (31).

Other researchers have studied staphylococcal infections using experimental models of diabetes mellitus.

Dunn and colleagues first described the diabetogenic effects of alloxan in 1943 (32). In animals made diabetic with alloxan and infected with staphylococci, some workers have found that there was an increased bacteriemia in mice (33,

34) or that the toxic features of local skin lesions in rabbits were increased (35). Other workers, however, reported that alloxan diabetes had no influence on the susceptibility of mice to infection with staphylococci (36).

Only a small dose of cocci (5 x 107 org/ml) was used to intraperitoneally infect the mice, however, and the mice were able to easily handled the organisms.

Currently the antibiotic streptozotocin is the preferred drug for generating a model for diabetes mellitus.

The diabetogenic property of streptozotocin was first reported in 1963 (37). The compound is usually administered either in a large bolus (38) or in multiple subdiabetogenic doses over consecutive days (39). When streptozotocin is given in small doses it induces an autoimmune response that

involves T-cells and results in a hyperglycemic state (40).

A large dose of the drug results in a rapid necrosis of the

beta cells (41). It is believed that the glucose moiety of

streptozotocin has an affinity for islets of the pancreas

and acts as a carrier for the N-nitroso-N-methyl urea group

into the target cell (42, 43). Once in the cell, the toxic

damage to the cell is thought to occur by the generation of

highly reactive CH3* ions which alkylate DNA bases at various positions (43, 44). The DNA repair process stimulates

nuclear poly(ADP-ribose) synthetase which leads to increased

NAD degradation, thereby depleting the intracellular NAD.

The cell is unable to scavenge oxygen free radicals which

are generated during normal oxidative metabolism, which in turn leads to cell lysis (45). This then results in a deficiency of circulating insulin and a diabetic state is

induced.

The purpose of this study was to evaluate the

survival of different S. aureus strains within the abscesses of streptozotocin-induced diabetic mice. An evaluation of any changes in the diabetic host's production of lipids in these lesions compared to normal mice was also done. CHAPTER II

MATERIALS & METHODS

MICE

Female Swiss ICR mice, weighing approximately 21-24 grams and obtained from Harlan/Sprague-Dawley, were used throughout.

STAPHYLOCOCCAL STRAINS AND ANIMAL INOCULATION

Staphylococcus aureus strains 18Z, P78, and PG114 have been described previously (46, 47). The first two strains produce both the alpha and delta hemolysins whereas strain

PG114 produces only the delta hemolysin. These strains, together with the non-hemolytic variant of P78, P78-22 (15), were used to generate intraperitoneal abscesses. Strains

18Z and PG114 were selected because they exhibit a delayed pattern of clearance whereas strains P78 and P78-22 were selected because they are readily eliminated from intraperitoneal abscesses. All four strains produce the clumping factor which is required to generate abscesses in the peritoneal cavity.

Staphylococcus aureus strains used to generate abscesses (18Z, PG114, P78, and P78-22) were grown in trypticase yeast extract broth (TYB) and in the absence of

15 C02 in order to retard subsequent in vivo alpha toxin production, thus enabling animals to survive the large inocula necessary for initial clump formation. The cultures were continuously shaken at 37°C for 18-24 hours and harvested by centrifugation. The cocci were washed three times with a diluent consisting of sterile saline containing

8% trypticase soy broth (TSB) (w/w) (BBL, Cockeysville,

MD.). Washed cells were resuspended in the same diluent to give a final concentration of 1010 - 10" organisms/ml.

Aliquots were sealed in ampoules and stored at -70°C until used. The concentration of organisms in the suspensions were verified by plate counts done on the suspension at the time of preparation and again at the time of use.

In preparation for inoculation into mice, the suspensions were diluted in saline-8% TSB at a concentration of 5 x 109/ml. Of this, 0.2 ml aliquots were inoculated into the peritoneal cavity of the animals.

DETERMINATION OF BLOOD GLUCOSE

Mice were randomly separated into either experimental or control groups. The animals were color coded by staining their fur with crystal violet, basic fuschin, or picric acid in 95% ethanol. Various color combinations were used on five distinct areas of a mouse (head, back, tail, right or left side) thus enabling each mouse to be followed throughout an experiment. Mice were fasted 18-20 hours prior to obtaining blood for glucose measurements. Water was provided ad libitum.

The blood sugar baseline for each animal was established by collecting three blood samples over a one week period.

Samples consisting of a drop of blood from a tail snip were measured using Glucostix test strips (Ames Division, Miles

Laboratories, Inc. Elkhart, In.) and a Glucometer II reflectance meter (Ames Division, Miles Laboratories, Inc.

Elkhart, In.). A quality control check of the Glucometer II and Glucostix was done each time the meter was used. An acceptable range using a test paddle to check the Glucometer

II was 128-162 mg/dl, and our readings were 133 + 4 mg/dl.

An acceptable range using a glucose test solution to check the Glucostix was 88-111 mg/dl, and our readings were 100 +

7 mg/dl.

GENERATION OF DIABETIC MICE

Streptozotocin (Sigma Chemical Co., St. Louis, Mo.) was dissolved in 0.1 M citrate buffer (pH 3.8-4.2) to prepare a stock solution (6.5 mg/ml) and appropriate aliquots were introduced intraperitoneally into the mice. To induce a diabetic state, mice were given a dose of streptozotocin

(130 mg/kg) each day for two consecutive days. Control mice received the same amount of citrate diluent alone. Mice were considered diabetic when their fasting blood glucose level was 200 mg/dl or greater. Each animal was weighed each time a blood sample was taken. I

18

EVALUATION OF STAPHYLOCOCCAL SURVIVAL WITHIN ABSCESSES

Control animals and mice determined to be diabetic were inoculated intraperitoneally with 109 staphylococci one week after the mice had received the second dose of either streptozotocin or diluent. At each sample time over a 20 day period, 4-9 animals were randomly selected from each group and sacrificed by cervical dislocation. Abscesses from each animal were removed and placed in a glass tissue grinder containing 1 ml saline-8% TSB. The abscesses were then homogenized using a motor driven teflon pestle. An aliquot of the resultant abscess homogenate was serially diluted and plated on trypticase soy agar (TSA) to enumerate the staphylococcal population. The remaining abscess homogenate was sealed in an ampoule and stored at -70°C for future use. BACTERICIDAL ASSAY

S. aureus strains TG and 303 (15, 48) were used as the indicator strains in the bactericidal assay. Strain TG is very resistant to the bactericidal monoglycerides whereas strain 303 is very sensitive; thus the endpoints in the assay can reveal the presence of differential activity.

Both strains are equally sensitive to the unsaturated fatty acids.

S. aureus TG and 303, the bactericidal assay indicator strains, were grown in TSB under air with constant agitation. After incubation at 37°C for 18-24 hours, 19

cultures were centrifuged and the pelleted cocci were washed three times in saline-8% TSB. After the final washing,

cells were resuspended to a final concentration of 10s

cocci/ml. Aliquots were sealed in ampoules and stored at -

70°C until needed. Plate counts were done on the

suspensions at the time of preparation and when the

suspensions were used.

Lipids extracted from abscesses, derived from either diabetic or control mice, were dissolved in 95% ethanol to give a concentration of 1 mg/ml. Starting with a 1:10 dilution additional serial two-fold dilutions of the lipids were made in diluent consisting of 2 M NaCl with 2mM EDTA.

Suspensions of the test strains were diluted to 103 cocci/ml

in the same diluent and 1 ml aliquots of the suspension were added to 1 ml portions of each lipid dilution. Several 1 ml aliquots of the bacterial suspension added to 1 ml amounts of the diluent alone served as controls. The mixtures were

incubated for 1 hour at 37°C in a shaking water bath, after which time 0.5 ml aliquots were removed from each tube for plate counts to determine the number of viable cocci remaining. A bactericidal unit (LD50) was defined as that amount of lipid which destroyed 50% of the staphylococci under the test conditions and end-points were determined by probit analysis (49). All tests were performed in duplicate. HISTOLOGICAL SECTIONS 20

Groups of 2-3 control or diabetic animals inoculated

with staphylococci were randomly selected at various

intervals over a 22 day time period. Abscesses from a

single animal were removed and placed in neutral-buffered

formalin. Abscesses were embedded in O.C.T. compound

(Lab-tek, Naperville, IL.) and frozen sections (6 /xm) were

cut from different regions within the abscesses. The

sections were stained with Oil Red 0 to demonstrate lipid.

For every section stained with Oil Red 0 an adjacent section was stained with hematoxylin-eosin.

The pancreas of each of 4-5 control and diabetic mice were removed and fixed in neutral-buffered formalin. The tissues were embedded in paraffin and sectioned and adjacent

sections were stained with immunocytochemical stains to reveal insulin, glucagon, and somatostatin (50). To demonstrate beta cells the islets were stained with guinea pig anti-insulin primary antiserum (1:4000). Alpha cells were demonstrated using rabbit anti-glucagon primary antiserum (1:4). Rabbit anti-somatostatin primary antiserum

(1:1500) was used to demonstrate delta cells in the islets.

Sections which were stained with all the reagents except the primary antiserum were used as a negative control. LIPID EXTRACTION

Abscesses were produced in diabetic or control mice and representative individuals were sacrificed at 5 or 10 days after infection. Abscesses from 3-14 animals were pooled and homogenized in saline-8% TSB. The lipids were extracted

from the abscess homogenate using the Dole procedure (51).

For each ml of abscess homogenate, 5 ml of a mixture of

isopropanol-heptane-2N sulfuric (40:10:1 v/v/v) were added

and the preparation shaken vigorously. After 10 minutes, 3 ml of heptane and 2 ml of distilled water for each ml of

abscess homogenate were added, and the phases were allowed to separate. The upper heptane layer was collected in a

lipid-clean round bottom flask and dried on a rotary evaporator (Rotavapor-R; Buchi; Switzerland) until a final volume of 5 ml remained in the flask. The sample was then transferred to a lipid-clean vial and dried under nitrogen until a constant weight was obtained. The lipids were sealed in ampoules under nitrogen and stored at -70°C until needed. FRACTIONATION OF LIPIDS

The total lipid fractions extracted from each abscess homogenate pool were dissolved in 1 ml hexane and added to a

1 x 16 cm column containing Florisil (60-100 mesh Fisher

Scientific, Pittsburgh, PA.) impregnated with 10% boric acid

(52). The various neutral lipids were sequentially eluted from the columns as follows: the triglycerides, methyl esters, and cholesterol were eluted with 200 ml hexane:ethyl ether (75:25 v/v); diglycerides were eluted with 100 ml hexane:ethyl ether (50:50 v/v); monoglycerides were eluted with 250 ml ethyl ether:methanol (96:4 v/v); and the free 22 fatty acids were eluted with 250 ml ethyl ether:glacial acetic acid (98:2 v/v).

Purity of the different fractions was verified by thin layer chromatography (TLC) of 50-100 nq aliquots on prewashed Silica Gel G plates (Analtech, Newark, Delaware) using either petroleum ether:ethyl ether:glacial acetic acid

(15:85:1 v/v/v) or hexane:ethyl ether:glacial acetic acid

(90:10:1 v/v/v) as the solvent systems (52). Similar aliquots of known lipid standards were chromatographed on the same plate. Individual lipids were visualized either by exposure to iodine vapors or by charring after spraying the dried plates with a 50% aqueous sulfuric acid containing

0.6% potassium dichromate and heating for one hour at 180°C.

COLLECTION OF INDIVIDUAL LIPIDS

Column fractions containing either triglycerides, free fatty acids, or monoglycerides were further purified by TLC.

Lipid samples were dissolved in a minimum amount of hexane

(about 0.5 ml or less) and were loaded onto a prescored tapered Silica Gel G plates which had been previously washed in chloroform:methanol (2:1 v/v), dried, and then activated by heating for one hour at 100°C. A 20 ixq aliquot was spotted in the lane on each edge of the plate and the remainder was loaded in the lanes on the middle portion of the plate. To identify the particular lipid desired a 20 jug 23

sample of known lipid standard was spotted in the adjacent

lanes at both edges of the plate.

For the isolation of monoglycerides the plates were

chromatographed in petroleum ether:ethyl ether:glacial

acetic acid (15:85:1 v/v/v). For the recovery of either

triglycerides or free fatty acids chromatography in hexane:

ethyl ether:glacial acetic acid (90:10:1 v/v/v) was used

(52). When the solvent front was 3-4 cm from the top of the

plate, the scored lanes of the plate, containing the lipid

standard and the lipid fraction sample, were snapped off

from the rest of the plate and were developed and visualized

by charring. The two sides were realigned with the rest of

the plate and the region of the gel containing the desired

lipid was marked. The silica gel from this area was scraped

and collected in a lipid clean screw cap tube. The lipid

was extracted from the gel by washing three times with ethyl

ether:methanol (10:1 v/v), centrifuging the extract each

time at 2000 rpm for 10 minutes. The pooled extracts were

filtered through glass wool, collected in a 500 ml round

bottom flask, and the solvent was dried on a rotavap. When

about 5 ml of the solvent remained, the preparation was

transferred to a preweighed lipid-free vial and dried under

nitrogen to a constant weight.

ESTERIFICATION OF FREE FATTY ACIDS AND GAS CHROMATOGRAPHIC

ANALYSIS The free fatty acid fraction was dissolved in hexane at

a concentration of 1 ml/mg. A 50 jul aliquot of sample was

added to 1 ml BF3-methanol and incubated for 5 minutes at

60°C. The fatty acid methyl esters were extracted with two

1 ml portions of hexane. The esters were recovered by drying at 40°C under nitrogen. The esters were then

dissolved in 50 n 1 of hexane and about 1 jul portions were

injected into a gas chromatograph (Model GC 5800, Hewlett-

Packard Co., Avondale, Pa.) fitted with a 25 m x .22 cm i.d.

column with CP-Sil-5-CB (Chrompack, Inc., Bridgewater, New

Jersey). The injector heater was set at 250°C, the oven heater at 185°C, and the detector heater was set at 300°C.

Helium was the carrier gas at a flow rate of 0.5 ml/min. To determine the composition of the methyl esters, the

retention times of the methyl ester samples were compared to the retention times of known fatty acid methyl ester

standards run either singly or in mixtures. To determine the amount of fatty acid methyl esters the area percent of the peak for the methyl ester was compared to the area percent of a peak of a known amount of an internal standard

(pentadecanoic acid).

TRIMETHYLSILYL ETHER DERIVATIVES OF MONOGLYCERIDES AND GAS

CHROMATOGRAPHIC ANALYSIS

The trimethylsilyl (TMS) ether derivatives of 1- and 2- mono glycerides were prepared to permit analysis by gas chromatography (53). Briefly, up to 10 mg portions of dried monoglyceride sample were added to 350 /xl pyridine (Aldrich

Chemical Co., Inc., Milwaukee, Wi.), 150 /x 1 hexamethyldisilazone (Aldrich Chemical Co., Inc., Milwaukee,

Wi.), and 60 /xl of trichloromethylsilane (Aldrich Chemical

Co., Inc., Milwaukee, Wi.). The mixture was allowed to sit for 15 minutes at room temperature. The mixture was dried under nitrogen and the TMS ethers were extracted with 5 ml of HPLC grade hexane and 1 ml of HPLC grade water. The ethers were dried over sodium sulfate and the top hexane layer was recovered and dried under nitrogen. The TMS ethers samples were then resuspended in 50 /xl of HPLC grade hexane. Aliquots of about 1 fig/n 1 were injected into a gas chromatograph fitted with a25mx.22cm column with CP-

SIL-5-CB (Chrompack, Inc.). The injector and oven temperature was 250°C and the detector temperature was

300°C. The carrier gas was helium at a flow rate of 80 ml/min. Known 1- and 2-monoglyceride standards were silylated using the same procedure and chromatographed to establish retention times in order to permit identification and quanitifcation of the monolgycerides in the abscesses.

By comparing the area percent of the peak of the monoglyceride samples to the area percent of the peak of an internal standard (pentadecanoin) it was possible to determine the amount of monoglyceride. 26

DETERMINATION OF GLUCOSE CONCENTRATION IN INTRAPERITONEAL

ABSCESSES

After inoculation with staphylococci, groups of 4-10

diabetic or control mice were randomly selected and

sacrificed at various times over a 21 day period. Abscesses

from a single animal were removed and weighed. Using a pycnometer the specific gravity and volume of the abscesses were determined. Subsequently the abscesses were homogenized as previously described. The concentration of

glucose of the abscess homogenate was measured with the

Glucometer II reflectance meter using Glucostix test strips.

The readings were then adjusted for the dilution factor to determine the actual glucose concentration in the original

abscess.

In order to evaluate a possible lipemic effect on the test strips, studies were done to see whether the lipids in abscess homogenates could affect the accuracy of glucose measurements. Since the test strips function correctly only

in the presence of blood, 15 ml of human blood were drawn, the glucose levels measured, and then the blood was allowed to glycolyze overnight at 37°C with agitation. The blood glucose concentration was measured the next day to verify that there was no residual glucose in the blood. Next, a known amount of glucose was added to the blood and a 1 ml aliquot of the glucose solution was added to 1 ml of abscess homogenate. The glucose concentration of the resultant 27 mixture was measured to determine whether abscess homogenates interfered with the readings.

CHEMICALS AND REAGENTS

Oleic acid, 1-monoolein, triolein, and cholesterol were purchased from Sigma Chemical, Co., St. Louis, Mo. All 1- monoglycerides used as standards for the gas chromatography analysis of the monoglyceride fractions, as well as 1,2- diolein and 1,3-diolein, were obtained from Nu-Chek Prep,

Elysian, Mn. The fatty acid methyl esters were purchased individually from Nu-Chek Prep or as mixtures from Matreya,

Inc., Pleasant Gap, Pa.

All solvents were reagent or HPLC grade. Solvents used for thin layer chromatography were distilled prior to use.

Before use, all glassware was cleaned with chloroform or acetone. CHAPTER III

RESULTS

GENERATION OF DIABETIC MICE

Preliminary studies were done to determine the amount

of streptozotocin necessary to induce the diabetic state.

Groups of 3-4 mice received either one or two doses of

streptozotocin ranging from 70 mg/kg to 150 mg/kg body weight intraperitoneally. All mice receiving two doses of

130 mg/kg given on consecutive days, became hyperglycemic

and this level of drug had a low mortality rate (33%).

Since larger doses did not give a greater response, it was decided to use this streptozotocin dosage regimen in all

future studies.

When female Swiss ICR mice were given a cumulative

intraperitoneal dose of streptozotocin (260 mg/kg body weight), in 0.1 M citrate (pH 3.8-4.2), the animals that responded to the drug had fasting blood glucose levels of greater than 200 mg/dl consistently, as shown in Table 1.

Not all the mice receiving the streptozotocin became diabetic; on average 70% of the mice became diabetic (range

of 49-96%) (Table 2). A diabetic mouse was defined as a mouse having a fasting blood glucose level of 200 mg/dl or

28 29 greater. Once hyperglycemic, the mice remained diabetic

(Figure 1) even after infection with S. aureus (Table 3).

Control animals, which had received equal amounts of the citrate diluent alone, maintained fasting blood sugar levels within the established range for normal fasting blood glucose values as shown in Figure 1 and Table 3.

The pancreas was removed from 4-5 control and diabetic mice one week after receiving either streptozotocin or diluent. Sections were stained for insulin to verify that the beta cells of the diabetics had been destroyed.

Adjacent tissue sections were stained separately with immunoperoxidase stains to detect glucagon (alpha cells) and somatostatin (delta cells). In the pancreas of control mice numerous insulin containing cells were detected in the islets. These same islets also contained cells which stained positive for glucagon and somatostatin (Plates I-III).

However, the islets from the pancreas of diabetic mice either contained no insulin producing cells or had some cells which exhibited only minor staining for insulin (Plate

IV). These same islets did stain positive for both glucagon

(Plate V) and somatostatin (Plate VI).

SURVIVAL OF STAPHYLOCOCCI IN ABSCESSES WITHIN DIABETIC MICE

In the human the normal host can usually control the proliferation of staphylococci in the tissues. However, certain conditions, such as diabetes mellitus, may predispose an individual to infection by various 30 microorganisms, particularly staphylococci. To better understand how staphylococci survive in abscesses within diabetics, the diabetic mouse model was used.

Groups of diabetic and control mice received an intraperitoneal dose of 109 staphylococci one week after receiving either the streptozotocin or citrate diluent treatment. From each group four to nine randomly selected animals were sacrificed at various sample times over a 20 day time period. The abscesses from each mouse were collected, homogenized, and the number of staphylococci in them were enumerated by plate count. Four strains were evaluated; S. aureus 18Z and PG114 were strains that normally exhibited the delayed pattern of elimination whereas strains P78 and P78-22 usually manifested a pattern of immediate elimination (14).

S. aureus 18Z exhibited the typical rebound phenomenon during the first five days, but at this time no significant differences in survival were detected in abscesses from diabetics or controls (Figure 2). However, at ten days post infection the number of cocci in the abscesses from control animals began to decline, but the number of organisms in the abscesses from diabetic animals did not begin to decrease in number until 15 days after infection.

In the case of S. aureus PG114, (Figure 3) there was an initial drop in the staphylococcal population in abscesses from both groups. Thereafter, the number of cocci in the 31

abscesses from control mice rebounded until the tenth day

post infection. Subsequently, the number of organisms

declined continuously. In diabetic mice not only did the

population of staphylococci in abscesses rebound sooner, but

thereafter the organisms persisted in the abscesses

throughout the remainder of the twenty-one day observation

period.

S. aureus P78 and its variant P78-22 (Figures 4 and 5) were both eliminated from abscesses in control mice. In

diabetic animals the population of P78 persisted essentially

unchanged for twenty days, whereas P78-22 cocci were

eliminated, albeit at a much slower rate than that seen in

control mice.

Abscesses that were produced by S. aureus strain 18Z were recovered from the peritoneal cavities of both control

or diabetic mice at 5 and 10 days after infection. The

abscesses of 12-14 control or diabetic mice were pooled and homogenized. Lipids were extracted from the resultant

abscess homogenate using the Dole procedure (51). The total

lipids extracted were then assayed for bactericidal

activity. Lipids collected from abscesses of control or

diabetic mice five days after infection had no bactericidal

activity against either of the two S. aureus indicator

strains, 303 (Figure 6) or TG (Figure 7). By the tenth day post infection, a detectable amount of bactericidal activity

against S. aureus strains 303 (Figure 8) and TG (Figure 9) 32

was observed in the lipid pool extracted from the abscesses

of control animals whereas the lipids extracted from the

abscesses of diabetic mice still had no detectable

bactericidal activity. HISTOLOGY OF INTRAPERITONEAL STAPHYLOCOCCAL ABSCESSES

The organization of various regions within

intraperitoneal abscesses generated by S. aureus has been

described previously (14, 15). Abscesses produced by S.

aureus strain P78-22 were harvested from the peritoneal

cavity of control and diabetic mice at various times after

infection. The abscesses were fixed in formalin and frozen

sections were stained with Oil Red 0 for lipid or with hematoxylin-eosin. Histologically the abscesses from the

two groups of animals were found to be similar, revealing,

from the center to the periphery, the core of cocci, the region of acellular debris, degenerating leukocytes,intact

leukocytes, and the connective tissue capsule.

After 10 days, the clumps of cocci at the core of the

abscesses were more difficult to locate in abscesses from control animals, whereas the cocci were readily found in the core of the abscesses from diabetics as late as 22 days post

infection. This difference was not unexpected since there was about a two log difference in the number of cocci present in the abscesses from these two groups (Figure 5).

In sections stained with Oil Red 0, stainable lipids were found to accumulate in greatest amounts just beneath the connective tissue capsule in both diabetic and control

abscesses (Plate VII). Compared to the diabetic, there

appeared to be more stainable lipid amassed in this area in

the sections of abscesses from control mice until the

seventh day. After seven days there did not appear to be

any differences in the amount of stainable lipid present in

the sections of abscesses from the control or diabetic mice.

Since this is the region where intact leukocytes are always

found, it probably represents the location of the most

recently arriving cells. Although smaller droplets were

detected in the deeper regions of the abscesses from the

control animals by the seventh day, lipid droplets did not

reach the core of lesions in diabetics until day 14.

ANALYSIS OF LIPIDS IN INTRAPERITONEAL ABSCESSES

To determine if the difference in the survival patterns

of the staphylococci in the abscesses from diabetics and

control mice was due to any alteration in the bactericidal

lipids, lipids were extracted from abscess homogenates.

Groups of 3-14 control and diabetic mice were sacrificed at

5 and 10 days post infection with S. aureus strain 18Z. The

abscesses were collected from the peritoneal cavity of the

animals, pooled, and homogenized in saline with 8% TSB.

Total lipids were extracted from the resultant abscesses homogenates following the procedure of Dole (51).

When total lipids were extracted from 5 day old

abscesses, more total lipid was detected in the lesions from 34 control mice than from the abscesses of diabetic animals

(Table 4). After ten days there were no differences in the amount of lipid extracted from the abscesses of the two groups (Table 5). The neutral lipids were then fractionated on a Florisil column (impregnated with 10% boric acid) and the triglyceride, free fatty acid, and monoglyceride fractions were collected.

The amounts of triglycerides recovered were measured gravimetrically. Five days after infection, the triglyceride concentration from the abscesses of control mice was greater than the triglyceride concentration in the abscesses of diabetic animals (Table 6). By ten days, however, the amount of triglyceride extracted from the lesions of both the controls and diabetics was approximately the same (Table 7).

The free fatty acids collected were methylated using

BFj-methanol and the methyl esters were analyzed by gas chromatography. The fraction was found to consist of the conventional C14-C18 saturated and unsaturated fatty acids plus two unidentified fatty acids (Tables 8 and 9). The unsaturated free fatty acids are more effective antibacterial agents than are the saturated free fatty acids. At five days there were more unsaturated and saturated free fatty acids recovered from the abscesses of diabetic mice than in the abscesses of control mice (Table

10). However, by the tenth day, the amount of fatty acids, 35 both unsaturated and saturated, recovered from the abscesses of control animals had doubled whereas the amount of free fatty acids had only slightly increased in the abscesses from diabetic mice (Table 11).

The monoglyceride fractions were silylated (53) and the TMS ethers were then analyzed by gas chromatography.

Peaks corresponding to monoacylglycerols with fatty acid moietities of C15, C16, C18, c 1#:1, and C1#:2 were detected. At five days, there was variation in the amount of monoglycerides/mouse extracted from the abscesses of control and diabetic mice (Table 12) and in the amount of saturated and unsaturated monoglycerides (Table 14) between the two runs. By ten days greater amounts of monoglycerides (per mouse or per abscess) were found in abscesses from diabetic animals than in abscesses from control mice as shown in

Tables 13 and 14.

GLUCOSE CONCENTRATIONS WITHIN ABSCESSES

While testing the sensitivity of the indicator strains,

S. aureus TG and 303, to representative bactericidal lipids it was observed that the medium used to culture the organisms affected the strain's sensitivity. When grown in trypticase soy broth S. aureus TG and 303 were less sensitive to 1- and 2-monoolein than when they were grown in trypticase yeast extract broth. However, their sensitivity to oleic acid was not affected by the growth medium (Table 15). Since trypticase yeast extract broth is identical to

trypticase soy broth except that the former lacks glucose,

it seemed reasonable to determine whether growth in the

presence of glucose affects the sensitivity of Sj. aureus to

bactericidal lipids. A chemically defined medium

consisting of amino acids, salts, and vitamins was prepared with and without 0.25% glucose and was used to grow the

organisms. When S. aureus TG and 303 were grown in the

amino acid broth (AAB) alone, the cocci were more sensitive

to 1- and 2-monoolein than when they were grown in AAB

containing glucose. As previously observed, the two strains

were just as sensitive to oleic acid regardless of the

medium in which they were grown (Table 15). The minimum

glucose concentration that altered the staphylococci1s

sensitivity to the monoglycerides was 0.25% (250 mg/dl).

This effect was phenotypic in nature since once the glucose

had been consumed in the media, the cocci were found to

again become sensitive to the monooleins.

Previous studies have shown that the intraperitoneal

abscesses are vascularized by the fourth day post infection

(14). Since the blood glucose concentration in diabetics

(greater than 200 mg/dl) is sufficient to affect the

sensitivity of staphylococci to the monoglycerides, it became desirable to determine the glucose levels within

abscesses in diabetic mice, and to evaluate whether this 37 phenomenon might contribute to the increased survival of the cocci in abscesses.

Abscesses were generated in diabetic and control mice by inoculating 10’ P78-22 IP one week after treatment with streptozotocin or citrate. Abscesses were recovered from 4 to 10 animals sacrificed at 3, 7, 10, 14, and 21 days post infection. The glucose levels in abscesses were determined as described in Materials and Methods. While there were significant differences in the simultaneous blood glucose concentrations among the two groups of animals (Figure 10), there were no such differences in the glucose concentrations within abscesses (Figure 11). In neither group were the glucose levels sufficient to have altered the organisms sensitivity to monoglycerides.

Since lipids present in the abscess homogenates might interfere with the reflectance meter's measurement of the glucose concentration, it was necessary to evaluate this possibility. This was done by adding a known amount of glucose to the abscess homogenate and assaying the mixture.

However, in order to obtain reliable readings with the

Glucometer the mixtures had to be prepared in the presence of blood. Blood drawn was allowed to glycolyze by shaking at 37°C overnight. A known amount of glucose was added to the blood diluent and 1 ml of this mixture was added to 1 ml of abscess homogenate. When measured on the reflectance meter no lipemic effect was detected (Table 16). CHAPTER IV

DISCUSSION

Individuals with diabetes mellitus are more prone to certain opportunistic infections. These infections can be caused by bacteria, fungi, or parasites. Morbidity and mortality rates are high among diabetics with these infections despite the availability of antimicrobial or other types of therapy. Among the types of infections associated with diabetics are urinary tract infections, foot infections, mucormycosis, necrotizing pneumonia, malignant external otitis, Gram-negative or staphylococcal septicemia, hepatobiliary infections, and necrotizing soft tissue infections (54). The pathogenic mechanisms of these infections are not well understood, but speculation has been made about the underlying causes of these diseases.

Vascular insufficiency as a result of diabetic microangiopathy is thought to play a role in predisposing diabetics to malignant external otitis and emphysematous cholecystitis (55, 56, 57, 58). Mixed aerobic and anaerobic organisms responsible for necrotizing soft-tissue infections apparently favor the conditions created by vascular insufficiency and tissue hypoxia (54). In experimental

38 mucormycosis, ketosis impairs the inflammatory response,

permitting tissue invasion by the fungi (59, 60, 61, 62).

The combination of diabetic neuropathology, vascular

insufficiency, and high glucose concentrations, which can

impair phagocytosis of the polymorphonuclear leukocytes

probably explains the predisposition of diabetics to urinary tract infections (54, 63, 64, 65). To study the host- parasite interactions of these infections, researchers have used a variety of mechanisms to induce a hyperglycemic

state, including the antibiotic streptozotocin.

Streptozotocin-induced diabetic mice have been shown to have an increased sensitivity to bacteria, fungi, and parasites. The 50% lethal dose was significantly lowered in diabetic animals for some strains of type II Group B

streptococcus (66) and pneumococcus (31). The increased

susceptibility to type II Group B streptococcus was due to a

failure of the diabetic mice to clear the organisms with an effective complement system and type II-Group B

streptococcus specific antibodies. High concentrations of glucose were suspected as being responsible for rats being more susceptible to pneumococcal infection. Streptozotocin-

induced diabetic mice also show a decreased resistance to

Pseudomonas aeruginosa infections (67). An impaired antibody function, abnormalities in phagocytes, and a disturbed vasculature, caused by the diabetic state were cited as possible mechanisms for this increased sensitivity to the pseudomonad infection. Cell mediated immunity is impaired in diabetic mice infected with Listeria monocytogenes and the mice are unable to limit the

intracellular multiplication of the bacteria (68).

Opportunistic infections in the diabetic have been caused by both Candida albicans and C. trooicalis (69). Tanowitz and others have found that hyperglycemia significantly increased parasitemia and mortality rates in diabetic mice compared to control mice infected with Trypanosoma cruzi (70). Mahmoud and others have reported that diabetic mice infected with

Schistosoma mansoni exhibited a milder response to the parasite in the acute phase of the infection compared to controls due to a decreased granulomatous response of the diabetic host. However, during the chronic phase, there was exacerbation of hepatosplenic disease only in the streptozotocin-induced diabetic mice, apparently the result of sinusoidal obstruction caused by a pronounced megalocytosis (71).

Defects in the cellular and humoral responses of streptozotocin-induced diabetic animals are reportedly responsible for the increased susceptibility of the diabetic animals to certain types of infections. One defense mechanism available to the host that has not been well studied, however, is the bactericidal lipids. One of the associated complications of diabetes mellitus is an altered lipid metabolism. Changes in the production of bactericidal 41

lipids in a diabetic host might favor infections by specific

organisms.

Certain lipids present in tissues of normal and

infected animals have been shown to exhibit antimicrobial

activity. It is well established that fatty acids on the

surface of normal skin acts as a protective agent against

colonization by certain microorganisms. Kochan and

coworkers have found that fatty acids thought to be

hydrolyzed by lipases from the phospholipids or lipoproteins

of macrophage cell membranes are antimycobacterial in

infected animals (72, 73). Free fatty acids in the lungs

have also been shown to be lethal for pneumococci,

streptococci, and species of Bacillus (74). Campbell and

coworkers have suggested that the relative proportion of

linoleic and oleic acid present in the respiratory tract may

control the colonization of staphylococci in individuals with cystic fibrosis or essential fatty acid defects (75).

The effectiveness of certain lipids as antimicrobial

agents have been known for many years. Research has shown

that cell membranes of Gram-positive bacteria are more

susceptible to the action of free fatty acids than Gram-

negative organisms, probably because the Gram-negative

organism are protected by the lipopolysaccharide layer of

the outer membrane (76). The bactericidal effects of fatty

acids include the inhibition of amino acid uptake and oxygen

consumption (77), the induction of protein and enzyme 42

leakage (77, 78), lysis of protoplasts (78), and the

uncoupling of electron transport with a decrease in ATP

(79). The bactericidal activity of free fatty acids is

dependent on chain length, degree of unsaturation, and

configuration (80, 81). When fatty acids are esterified to

glycerol a more active derivative is formed (82, 83, 84).

Kapral and coworkers have described a novel defense

mechanism operative in the staphylococcal abscess where

bactericidal lipids produced by the host in the lesions are

responsible for the destruction of the cocci (15). These

lipids consist of a pool of long chained unsaturated free

fatty acids (LCFA) and monoglycerides, which contain the

same fatty acids as the acyl group; both of these pools are

bactericidal. The LCFA act equally on different strains of

S. aureus whereas the monoglycerides, particularly the 2- monoglycerides, exhibit a differential activity unique to

the staphylococcal abscess.

The staphylococci can produce an enzyme FAME, that is

capable of esterifying the LCFAs and the fatty acid moieties

of the monoglycerides thus rendering the lipids ineffective

as antibacterial agents. In this context, the triglycerides

are important because they are capable of inhibiting FAME

and also because they are thought to be one source of the

free fatty acids and monoglycerides.

The bactericidal free fatty acids and monoglycerides are believed not only to be produced by the hydrolysis of 43 the triglycerides by the staphylococcal lipase, these bactericidal lipids may also be the result of hydrolysis of the triglycerides by host lipases. There is also evidence to suggest that there is a second source of monoglycerides.

When abscess homogenate was treated with calcium ionophore or inositol triphosphate there was an increase in the amount of monoglyceride fraction along with an increase in the differential activity (19).

If this delicate balance of host-organism interactions is tipped in favor of the cocci, the organism is capable of setting up a site of infection. The most common type of infection caused by S. aureus is an abscess. If the organism can survive in these lesions the cocci can produce various toxins. The specific disease which would be manifest would depend on the toxins produced by the organism, the location of the cocci in the body, and the ability of the host to control the staphylococci. Such conditions can exist in the presence of tissue damage, invasive procedures, or underlying illness, such as diabetes mellitus. The present study has examined this model of staphylococcal lesions in streptozotocin-induced diabetic mice.

An average of 70% of the female Swiss ICR mice inoculated with two consecutive doses of streptozotocin (130 mg/kg) became hyperglycemic. While the beta cells were preferentially destroyed in the pancreas of the experimental 44

animals, the beta cells were still present in the islets of

control mice as shown by an indirect peroxidase stain. When

the pancreas was stained using primary antiserum to glucagon

and somatostatin the islets of the pancreas of both diabetic

and control animals stained positive indicating functional

alpha and delta cells.

The amount of cumulative streptozotocin administered to

an animal, to induce hyperglycemia, is dependent on several

factors. These factors include the species of animal, the

strain of animal, and the sex of the animal. Agarwal has reported the sensitivity of an animal to streptozotocin may be established as rat>mouse>dog>guinea pig (42). Cats have been reported to be resistant to the diabetogenic effects of multiple subdiabetogenic doses of streptozotocin (85).

Various strains of mice have been shown to have distinct sensitivities to streptozotocin (86, 87, 88). Kim and

Steinberg found that several strains of mice differed in their response to the effects to streptozotocin, and also that the same strain of mouse from separate sources differed markedly in their sensitivity to two regimens of treatment

(89). The effect of streptozotocin has been shown to be enhanced in male mice by androgen (90, 91) and male cats are more sensitive to streptozotocin than are female cats (85).

Even though the number of mice that became diabetic varied from 49-96% after treatment with streptozotocin, a sufficient number of animals were available to examine the 45

host-pathogen interactions in staphylococcal lesions. Four

distinct Staphylococcus aureus strains were better able to

survive within abscesses in diabetic animals than they were

in such lesions in control mice.

When inoculated into the peritoneal cavity of control

mice the four strains of staphylococci exhibited the same

pattern of survival previously described by Dye and others

(14, 15). S. aureus strains 18Z and PG114 displayed a

pattern of delayed clearance whereas strains P78 and P78-22

were quickly eliminated from the abscesses. These same

strains in diabetic mice either persisted the entire 20 day

observation period (strains PG114 and P78) or were cleared

from the abscesses, but at a slower rate than cocci in

normal mice (strains 18Z and P78-22).

The effect of glucose on S. aureus has been reported in

other studies. Mailman observed that when grown in the

presence of fermentable sugars, including glucose, S. aureus was more resistant to quaternary ammonium compounds than were cocci grown in media without the fermentable sugars

(92). Hayashi and others have reported that when S. aureus

is grown in the presence of glucose there is a change in the phospholipid make up of the cell membrane (93). More even- number long chain fatty acids are found in the membrane of

cocci grown in the presence of glucose, whereas the membrane

of organisms grown in the absence of glucose contain more

anteiso odd-number long chain fatty acids. This might be an 46

indication that the membrane is more rigid in the

staphylococci grown in media containing glucose.

In the present studies the staphylococci were less

sensitive to 1- and 2-monoglycerides when grown in the

presence of glucose. This effect appeared to be phenotypic

in nature since the staphylococci became sensitive to the

bactericidal action of the monoglycerides once the glucose

in the media was completely utilized. This phenomenon was

thought to be important because the glucose concentration that alters the organism's sensitivity to the bactericidal

lipids (250 mg/dl) is well within the range of the blood glucose concentration of diabetic mice (greater than 200 mg/dl). Since the abscesses are vascularized, if glucose gained access to the core of the abscess the cocci could be affected by the blood glucose, enabling the organism to have

an increased resistance to the bactericidal activity of the monoglycerides and persist.

In spite of obvious differences in blood glucose

levels, the glucose concentration in abscesses from diabetic animals was no different than that in abscesses from, control mice. Furthermore, in both instances the level of glucose measured was too low to have altered the organism's resistance to the bactericidal monoglycerides present in the abscesses.

When lipids were extracted from five and 10 day old

intraperitoneal abscesses produced by S^. aureus strain 18Z, 47 from both groups of mice, and assayed for bactericidal activity against the two S. aureus indicator strains, TG and

303, no bactericidal activity was detected before ten days.

The total lipid pool extracted from the abscesses of control mice possessed a small amount of bactericidal activity whereas the total lipids extracted from abscesses of diabetic mice did not after 10 days.

Efforts were made to quantify the total lipid and the free fatty acid, triglyceride, and monoglyceride fractions from abscesses of the two groups of mice at five and ten days post infection with S. aureus strain 18Z. At five days, the amount of total lipid (per mouse) was greater in abscesses from control than in abscesses from diabetic animals. This difference was attributed to larger amounts of triglycerides in abscesses from control mice. Whereas the abscesses from diabetic mice had larger amounts of free fatty acids than those from control mice, there was no consistent difference in the amount of monoglycerides extracted from the abscesses of the two groups of animals.

By ten days post infection, the amount of total lipid recovered from the abscesses of diabetic mice was approximately equal to that extracted from control animals.

While the amount of monoglycerides present in the abscesses of diabetic mice was greater than the monoglyceride pool from the abscesses of controls, the differences alone could not account for the increase in total lipids. The 48 triglyceride fraction recovered from the abscesses of control animals was still greater than the amount of these lipids in abscesses of diabetic mice.

This difference in the composition of the lipids extracted from abscesses of control animals may give those mice an advantage over the diabetic mice in eliminating the staphylococci in lesions. Since the abscesses of control mice contain more triglycerides than abscesses of diabetic mice, the control animals may be better able to neutralize the effects of FAME. Therefore, more free fatty acid and monoglycerides would be available to destroy staphylococci in abscesses of control animals.

Not all fatty acids are bactericidal, in particular the long-chain unsaturated free fatty acids are decidedly more bactericidal than their saturated counterparts. For this reason, each of fatty acids in the fatty acid pools was quantified. Two of the free fatty acids present in the extracted lipids from both control and diabetic mice were not identified. The retention times suggest that one is probably a Cie and the second a C20 fatty acid. The identification of these particular free fatty acids and their possible role in the host-organism relationship is unclear. While the abscesses from diabetic mice had larger amounts of both unsaturated and saturated free fatty acids at five days than did abscesses from controls, these differences were no longer apparent by the 10th day. At 49

that time the amount of both saturated and unsaturated free

fatty acids had nearly doubled in the abscesses from control

mice, whereas the amounts had increased only slightly in the

ones from diabetic mice. At both sample times the percent

composition of the free fatty acid pool was approximately

the same. Although the abscesses from the two groups of

animals did possess bactericidal free fatty acids, recall

that the monoglycerides, more specifically the 2-

monoglycerides, seem to be primarily responsible for

eliminating the staphylococci within the abscesses.

We did not distinguish between 1- and 2-monoglycerides in

samples extracted from abscesses and analyzed by gas

chromatography. Although the abscesses of diabetic animals

had more monoglycerides at ten days than did those from

control mice, we do not know what proportion were in the more stable isomeric form, 1-monoglyceride. While the 1- monoglycerides are bactericidal, they do not express the

same degree of differential bactericidal activity as the 2- monoglycerides. A difference in the type of monoglyceride present in the abscesses of diabetics could allow the four

strains of staphylococci tested to persist longer than when present in abscesses of control mice. The amount of each

isomer in abscesses from control and diabetic mice remains to be determined.

Perhaps the presence of a greater concentration of triglycerides, free fatty acids, and possibly 2- monoglycerides after 10 days, is not the only advantage the normal mice have over the diabetic animals. Normal mice may be able to mobilize these lipids within abscesses sooner than the diabetic animals. This possibility is supported by the observation made on sections of abscesses stained with

Oil Red 0. Whereas the largest amounts of lipid accumulated just beneath the connective tissue capsule, of abscesses in both control and diabetic mice, more "stainable lipid" was detected in this area earlier in control animals than in diabetic mice.

Not only might the lipids accumulate slower in the diabetic mice, but the distribution of the lipids throughout the abscess might also differ between the two groups of animals. When sections of abscesses from control and diabetic mice were stained with Oil Red 0 to detect lipid, differences in the dissemination of the lipid droplets was observed. Numerous small lipid droplets were seen in the deeper regions of the abscess, even in the core where the cocci are located, in the abscesses of control mice at seven days. Such droplets were not seen in the same vicinity in the abscesses of diabetic animals until the 14th day. While the distribution of the different kinds of lipid is not known, the survival studies suggest that the bactericidal lipids reach the staphylococci sooner in abscesses obtained from control mice than from diabetic animals. 51

The source of the lipids is not known, however, but the macrophages remain the prime candidate as the cell origin of the host's bactericidal lipids in the abscesses. Macrophage

accumulate a detectable amount of lipid bodies in their cytoplasm at sites of infection, inflammation, and neoplastic processes (94, 95). Preliminary findings in our

laboratory show that macrophages recovered from peritoneal washings of mice after infection with staphylococci, have

large lipid droplets in the cytoplasm upon staining with

Nile Red.

While examining the immune response of streptozotocin-

induced diabetic animals, Saiki and coworkers found that B- cell functions remain intact, whereas the T-cell function and phagocytic activity of macrophages in diabetic animals are depressed (96). Jones and coworkers reported that the macrophages of streptozotocin-induced diabetic rats have a decreased sensitivity to macrophage inhibitory factor. This may preclude the cells from accumulating at the site of an infection (97). Any detrimental effects of the streptozotocin itself on the immune response is a possibility. Mahmoud and coworkers found that the T-cell activity returns to normal once the mice were treated with insulin, indicating the effects are due to the metabolic disturbances rather than the direct toxic effects of the drug (71). When we counted the leukocytes in peritoneal washings, no differences were detected in the inflammatory 52 response to intraperitoneal inoculation with staphylococci between control and the streptozotocin-induced diabetic mice

(data not shown).

In summary, the staphylococci were more readily cleared from intraperitoneal abscesses from control mice than from diabetic mice over the 20 day observation period. When the bactericidal lipids were quantitated, no differences were detected in the amount of free fatty acids present in the abscesses from the two groups of animals at either 5 or 10 days. The amount of monoglycerides present in the abscesses from control or diabetic mice varied until 10 days, after which time the diabetic mice clearly had the greater amount of monoglycerides. The staphylococci1s ability to persist in the abscesses from diabetic mice does not appear due to a diabetic's inability to produce bactericidal lipids, however, the amount of 1- and 2-monoglycerides present in the abscesses from the two groups of animals is not known.

There were more triglycerides recovered from abscesses from control mice than from diabetic animals. Since the triglycerides neutralize the activity of FAME, the staphylococcal enzyme may have a better opportunity to inactiviate the bactericidal free fatty acids and monoglycerides present in abscesses from diabetic mice.

This may allow the staphylococci to persist in abscesses from diabetic mice. More stainable lipid was detected earlier and more lipid droplets were observed sooner throughout the abscesses from control mice than from diabetic animals. Therefore, the difference in the ability of the diabetic to clear the staphylococci may be due to an inability to mobilize the bactericidal lipids effectively. If diabetics have a defect in the macrophage's ability to mobilize and accumulate the bactericidal lipids in the abscesses, it remains to be determined whether insulin treatment restores normal function in these animals. 54

TABLE 1.

AVERAGE BLOOD GLUCOSE (mg/dl ± SD) BEFORE AND AFTER INOCULATION WITH STREPTOZOTOCIN (STZ)' OR CITRATE (CIT)

BEFORE AFTER” BEFORE AFTER GROUP STZ STZ CIT CIT

1 115 (± 21) 325 (± 86) 128 (+ 21) 127 (±18) n=60 n=50 n=50 n=39

2 82 (± 18) 281 (± 49) 84 (± 20) 115 (± 15) n=50 n=29 n=30 n=20

3 109 (+ 31) 304 (± 58) 97 (± 33) 66 (± 7) n=35 n=14 n=30 n=15

4 69 (± 12) 288 (+ 50) 63 (± 12) 63 (± 11) n=60 n=42 n=40 n=40

Streptozotocin given as mg/kg body weight IP in 0.1 M citrate (pH 3.8-4.2).

One week after inoculation. 55

TABLE 2.

PERCENT POSITIVE MICE MADE DIABETIC WITH STREPTOZOTOCIN AND CONTROLS RECEIVING CITRATE

CONTROLS______DIABETICS BLOOD' BLOOD' GROUP %POSITIVE GLUCOSE______%POSITIVE GLUCOSE

1 0 (0/39) 127 (±18) 93 (50/54) 321 (±77)

2 0 (0/20) 115 (±15) 60 (29/46) 281 (±49)

3 0 (0/30) 88 (±15) 62 (18/29) 287 (±45) 4 67 (10/15) 256 (±46)

5 0 (0/40) 63 (±11) 71 (42/59) 288 (±50)

6 0 (0/30) 55 (±11) 49 (33/67) 256 (±40) 7 0 (0/40) 124 (±18) 64 (34/53) 257 (±37) 8 0 (0/37) 116 (±28) 81 (35/43) 280 (±45)

9 0 (0/39) 110 (±16) 89 (47/53) 314 (±50)

10 0 (0/38) 124 (±21) 76 (39/51) 288 (±50)

11 0 (0/38) 103 (±16) 97 (34/35) 359 (±44)

12 0 (0/37) 84 (±13) 78 (42/54) 264 (±49) 13 0 (0/32) 101 (±12) 62 (45/72) 279 (±64)

14 0 (0/39) 73 (±18) 54 (30/56) 266 (±69)

15 52 (48/93) 292 (±74)

MEAN 0 99 (±24) 70 286 (±28)

One week after treatment with streptozotocin (260 mg/kg) or citrate. 56

4 0 0 -

DIABETIC ■o ^a 300 E

ut OT O O a a 200 Q o o _4 m CONTROL

100

STZ or CIT

10 20. 30 DAYS

FIGURE 1. Fasting blood glucose concentrations of mice before and after treatment with streptozotocin or citrate. Each time point represents the arithmetic mean of the average fasting blood glucose concentration of 4-9 control and diabetic mice. The animals received either a cumulative dose of 260 mg streptozotocin (STZ)/kg body weight in 0.1 M citrate (pH 3.8-4.2) or an equal amount of citrate (CIT) alone at day 0. The vertical bars represent the standard error of the mean. 57

TABLE 3.

AVERAGE BLOOD GLUCOSE (mg/dl ± SD) BEFORE AND AFTER INFECTION WITH S. AUREUS 18Z

BLOOD GLUCOSE______CONTROLS______DIABETICS

BEFORE INFECTION 63 (+ 11) 288 (± 50)

DAYS POST INFECTION 1 70 (± 12) 287 (± 89)

3 74 (+ 17) 329 (+ 76)

5 56 (± 32) 349 (± 66)

10 77 (± 38) 305 (± 96)

15 66 (± 19) 359 (± 52) Plate I. Immunocytochemical staining of pancreas islets from a normal mouse using the indirect immunoperoxidase technique to detect insulin. The panels represent a 3 /im section of pancreas (A. x 25.2) and islet (B. x 160) stained with guinea pig anti-insulin primary antiserum. The J3 cells make up the "core" of the islet. Plate II. Immunocytochemical staining of pancreas islets from a normal mouse using the indirect immunoperoxidase technique to detect glucagon. The panels represent a 3 fim section of pancreas (A. x 25.2) and islet (B. x 160) stained with rabbit anti-glucagon primary antiserum. The a cells are present around the periphery of the islet. 60

Plate III. Immunocytochemical staining of pancreas islets from a normal mouse using the indirect immunoperoxidase technique to detect somatostatin. The panels represent a 3 /Ltm section of pancreas (A. x 25.2) and islet (B. x 160) stained with rabbit anti-somatostatin primary antiserum. The S cells are located around the periphery of the islet. Plate IV. Immunocytochemical staining of pancreas islets from a diabetic mouse using the indirect immunoperoxidase technique to detect insulin. The panels represent a 3 p section of pancreas (A. x 25.2) and islet (B. x 160) stained with guinea pig anti-insulin primary antiserum. The 6 cells are noticeably lacking in the islet. Plate V. Immunocytochemical staining of pancreas islets from a diabetic mouse using the indirect immunoperoxidase technique to detect glucagon. The panels represent a 3 /im section of pancreas (A. x 25.2) and islet (B. x 160) stained with rabbit anti-glucagon primary antiserum. The a cells are present around the periphery of the islet. Plate VI. Immunocytochemical staining of pancreas islets from a diabetic mouse using the indirect immunoperoxidase technique to detect somatostatin. The panels represent a 3 jttm section of pancreas (A. x 25.2) and islet (B. x 160) stained with rabbit anti-somatostatin primary antiserum. The S cells are located at the periphery of the islet. 64

10

DIABETIC

CONTROL

10

S. aureus 18Z 10

5 10 15 20 DAYS

Figure 2. The survival of Staphylococcus aureus strain 18Z in the abscesses from control and diabetic mice. Each point represents the geometric mean of 4-5 animals. The vertical bars represent the standard error of the mean. 65

DIABETIC

10

Ul W 10 O s o o o 10 CONTROL

S. aureusPG114 10

10 20 DAYS

Figure 3. Survival curves of S. aureus strain PG114 in abscesses of control and diabetic mice. Each time point represents the geometric mean of 4-5 animals. The vertical bars represent the standard error of the mean. 66

10

DIABETIC

LU . _ CD 10 D Os

o o O _ o 1 0 7-

C0NTR0L S. aureus P78 10

10 15 20 DAYS

Figure 4. The survival of S. aureus strain P78 in intraperitoneal abscesses from control and diabetic animals. Each point represents the geometric mean of 4-5 mice. The vertical bars represent the standard error of the mean for that time point. 67

10 DIABETIC

III 0) 3 Os o o o

10 CONTROL

S. aureus P78-22

15 20 DAYS

Figure 5. The survival of S. aureus strain P78-22 in the abscesses from control and diabetic animals. Each time point represents the geometric mean of 4-5 animals. The vertical bars represent the standard error of the mean. % SURVIVAL 100 40 20 60 80 days post infection with S. aureus 18Z. The indicator The 18Z.S. aureusinfection with post days iue6 Bceiia ciiyo h oa lipidpool total ofactivity the Bactericidal 6.Figure strain is303.strain 5 mice anddiabetic controlof abscesses from the extracted DY H OA PO v 303 TOTAL vs POOL AH DAY 5 DILUTION 10 CONTROL DIABETIC 10

68

% SURVIVAL 100 mice infected with S. aureus 18Z. The indicator strain is strain indicator The 18Z.S. aureus infected with mice extracted from 5 day old abscesses of control and diabetic and ofcontrol abscesses old 5from day extracted lipid pool total ofthe activity Bactericidal 7.Figure S . aureus TG. aureus S. - 0 8 - 0 6 - 0 4 20 - - DY H TOTAL AH TG DAY LIPID POOL vs 5 DILUTION CONTROL DIABETIC

TT 10

69

70

10 DAY AH TOTAL LIPID POOL vs 303

100-

8 0 -

£ 6 0 H > u D 4 0 - V) V

20- DIABETIC CONTROL

I I -2 -3 -4 10 10 10

DILUTION

Figure 8. Bactericidal activity of the total lipid pool extracted from the abscesses of control and diabetic animals 10 days post infection with S.. aureus 18Z. The indicator strain is aureus 303. iue9 Tebceiia ciiyo oa lipids oftotal activity bactericidal The 9. Figure asps neto ihS aru tan1Z Indicator 18Z.strain S. aureus with infection post days strain is S. aureus TG. aureusS. is strain 10 at mice ordiabetic controlof abscesses from extracted SURVIVAL lOO— — 0 4 20 — 0 8 — 0 6 - 0 1 A A TTL II PO v TG vs POOL LIPID TOTAL AH DAY " T " 10 -2 DILUTION " T " 10 -3 1 CONTROL DIABETIC

10 -----

71

72 Plate VII. Ten day abscess sections stained with Oil Red O to detect lipid. The panels represent a 6 /xm section of abscess from a control (A) and diabetic (B) mice, x 160. The connective tissue capsule is at the left of the panel with the extreme outer region of the leukocytic layer just beneath the capsule. Lipid droplets appear dark black. 73

TABLE 4.

TOTAL LIPIDS EXTRACTED FROM S. AUREUS 18Z INTRAPERITONEAL ABSCESSES AT 5 DAYS.

RUN 1 RUN 2

CONT DIAB CONT DIAB

NUMBER OF MICE 8 14 8 3

TOTAL LIPID (mg/mouse) 5.3 1.9 1.5 0.77

TOTAL LIPID/ABSCESS (mg/mg) 0.58 0.35 0.16 0.038 74

TABLE 5.

TOTAL LIPIDS EXTRACTED FROM 10 DAY S. AUREUS 18Z INTRAPERITONEAL ABSCESSES

RUN 1 RUN 2

CONT DIAB CONT DIAB

NUMBER OF MICE 8 8 10

TOTAL LIPID (mg/mouse) 1.0 1.3 1.6 2.4

TOTAL LIPID/ABSCESS 0.07 0.06 0.12 0.11 (mg/mg) 75

TABLE 6.

TOTAL TRIGLYCERIDES EXTRACTED FROM 5 DAY S. AUREUS 18Z INTRAPERITONEAL ABSCESSES

RUN 1 RUN 2

CONT DIAB CONT DIAB

NUMBER OF MICE 8 14 8 3

TOTAL TRIGLYCERIDE 2.5 0.77 0.78 0.3 (mg/mouse)

TOTAL TRIGLYCERIDE/ABSCESS 280.0 140.0 85.0 15.0 (jug/mg) 76

TABLE 7.

TOTAL TRIGLYCERIDES EXTRACTED FROM 10 DAY S. AUREUS 18Z INTRAPERITONEAL ABSCESSES

RUN 1 RUN 2

CONT DIAB CONT DIAB

NUMBER OF MICE 8 8 10 5

TOTAL TRIGLYCERIDE 0.32 0.4 0.5 0.64 (mg/mouse)

TOTAL TRIGLYCERIDE/ABSCESS 22.0 19.0 38.0 30.0 (Atg/mg) 77

TABLE 8.

COMPOSITION OF FREE FATTY ACID FRACTION EXTRACTED FROM 5 DAY S. AUREUS 18Z INTRAPERITONEAL ABSCESSES

RETENTION RUN 1 RUN 2 AVERAGE

TIMES CONT DIAB CONT DIAB CONT DIAB

MYRISTIC 3.85 2.5 1.4 2.5 3.3 2.5 2.4 PALMITOLEIC 6.00 5.0 2.8 2.5 <3.3 3.8 2.8 PALMITIC 6.40 40.0 26.0 26.0 50.0 33.0 38.0 UNKNOWN #1 9.60 6.2 3.6 5.0 17.0 5.6 10.0 LINOLEIC 10.47 19.0 16.0 8.8 20.0 14.0 18.0 OLEIC 10.80 32.0 24.0 16.0 -3-7.0 24.0 30.0 LINOLENIC 10.96 11.0 9.3 12.0 37.0 12.0 23.0 STEARIC 11.93 15.0 14.0 16.0 47.0 16.0 30.0 UNKNOWN #2 38.58 12.0 7.2 11.0 27.0 12.0 17.0

Quantities are in ngs/mouse. Retention times are in minutes. 78

TABLE 9.

COMPOSITION OF FREE FATTY ACIDS EXTRACTED FROM 10 DAY S. AUREUS 18Z INTRAPERITONEAL ABSCESSES

RETENTION RUN 1 RUN 2 AVERAGE

TIMES CONT DIAB CONT DIAB CONT DIAB

MYRISTIC 3.85 19.0 20.0 4.0 2.0 11.0 11.0 PALMITOLEIC 6.00 6.2 5.0 4.0 6.0 5.1 5.5 PALMITIC 6.40 56.0 44.0 66.0 56.0 61.0 50.0 UNKNOWN #1 9.60 6.2 6.2 6.0 8.0 6.1 7.1 LINOLEIC 10.47 34.0 28.0 35.0 34.0 34.0 31.0 OLEIC 10.80 51.0 36.0 34.0 44.0 42.0 40.0 LINOLENIC 10.96 19.0 15.0 19.0 20.0 19.0 18.0 STEARIC 11.93 26.0 22.0 32.0 34.0 29.0 28.0 UNKNOWN #2 38.58 11.0 15.0 10.0 16.0 10.0 16.0

Quantities are expressed as ngs/mouse. Retention times are expressed as minutes. 79

TABLE 10.

FREE FATTY ACIDS EXTRACTED FROM 5 DAY S. AUREUS 18Z INTRAPERITONEAL ABSCESSES

RUN 1 RUN 2 AVERAGE

CONT DIAB CONT DIAB CONT DIAB

UNSATURATED 67.0 52.0 39.0 94.0 54.0 73.0 FFA (47%)’ (50%) (40%) (39%) (44%) (43%)

SATURATED 58.0 41.0 44.0 100.0 52.0 70.0 FFA (41%) (40%) (44%) (42%) (42%) (41%)

UNKNOWN 18.0 11.0 16.0 44.0 18.0 27.0 FFA (12%) (10%) (16%) (18%) (14%) (16%)

Amounts of free fatty acids (FFA) are expressed as ngs/mouse.

Percent total of free fatty acid recovered. TABLE 11.

FREE FATTY ACIDS (FFA) EXTRACTED FROM 10 DAY S. AUREUS 18Z INTRAPERITONEAL ABSCESSES

RUN 1 RUN 2 AVERAGE

CONT DIAB CONT DIAB CONT DIAB

UNSATURATED 110.0 84.0 92.0 104.0 101.0 94.0 FFA (48%)‘ (44%) (44%) (47%) (46%) (46%)

SATURATED 101.0 86.0 102.0 92.0 100.0 89.0 FFA (44%) (45%) (48%) (42%) (46%) (43%)

UNKNOWN 17.0 21.0 16.0 24.0 16.0 22.0 FFA (8%) (11%) (8%) (11%) (8%) (11%)

Amount of free fatty acids (FFA) expressed as ngs/mouse.

Percent total free fatty acid recovered. 81

TABLE 12.

TOTAL AMOUNT OF MONOGLYCERIDES EXTRACTED FROM 5 DAY S. AUREUS 18Z INTRAPERITONEAL ABSCESSES

RUN 1 RUN ______CONT DIAB______CONT DIAB

NUMBER OF MICE 8 14 8 3

TOTAL MONOGLYCERIDES 40.0 18.0 51.0 220.0 (ng/mouse)

TOTAL MONOGLYCERIDES/ 4.0 3.3 6.0 11.0 ABSCESS (ng/mg) Table 13.

TOTAL MONOGLYCERIDES EXTRACTED FROM 10 DAY S. AUREUS 18Z INTRAPERITONEAL ABSCESSES

RUN 1 RUN 2 ______CONT___DIAB_____CONT DIAB

NUMBER OF MICE 8 8 10 5

TOTAL MONOLGYCERIDES 111.0 579.0 27.0 356.0 (ng/mouse)

TOTAL MONOGLYCERIDES/ 7.5 27.0 2.0 17.0 ABSCESS (ng/mg) TABLE 14.

TOTAL UNSATURATED AND SATURATED MONOGLYCERIDES EXTRACTED FROM S. AUREUS 18Z INTRAPERITONEAL ABSCESSES AT 5 AND 10 DAYS

RUN 1 RUN 2

CONT DIAB CONT DIAB

5 DAYS:

UNSATURATED 24.0 5.7 21.0 100.0 MONOLGYCERIDES (60%)’ (30%) (41%) (45%)

SATURATED 16.0 13.0 30.0 120.0 MONOGLYCERIDES (40%) (70%) (59%) (55%)

10 DAYS:

UNSATURATED 69.0 392.0 8.0 250.0 MONOGLYCERIDES (62%) (68%) (30%) (70%)

SATURATED 42.0 186.0 19.0 106.0 MONOGLYCERIDES (38%) (32%) (70%) (30%)

Quantities expressed as ngs/mouse.

Percent total of monoglycerides recovered. 84

TABLE 15.

SENSITIVITY OF STAPHYLOCOCCUS AUREUS STRAINS TG AND 303 GROWN IN DIFFERENT MEDIA TO OLEIC ACID AND 1- AND 2-MONOOLEIN

LD.n/ma LIPID*

GROWTH OLEIC 1- 2 - ORGANISM MEDIUM ACID MONOOLEIN MONOOLEIN

S. aureus TG TYB 2.3 X 103 7.0 x 102 9.0 x 102 TSB* 2.4 X 103 <4.0 x 10' <4.0 X 101

S. aureus 303 TYB 4.2 X 103 2.0 X 103 1.0 X 102

TSB 3.3 X 103 <4.0 X 10’ <4.0 X 101

S. aureus TG AAB >5.1 X 103 8.5 X 102 7.5 x 102

AAB + 3.1 X 103 <4.0 x 10' <4.0 X 10' GLU'"

S. aureus 303 AAB 2.3 x 103 1.4 x 103 1.1 x 103

AAB 1.6 X 103 <4.0 X 10' <4.0 X 10' GLU

As determined by the method of probit analysis of the dose response curve.

Trypticase soy broth contains 0.25% glucose trypticase yeast extract broth (TYB) contains no glucose.

Amino acid broth (AAB) with 0.25% glucose. BLOOD GLUCOSE (mg/dll 200 300 100 400 represent the standard error of the of errormean. standard the represent diabetic animals infected with S. aureus strain P78-22 and P78-22strain S. aureusinfected with animals diabetic of the blood glucose concentration of 4-10 control or control of4-10 concentration glucose blood ofthe Figure 10. Fasting blood glucose concentrations ofmice concentrations bloodglucose Fasting 10. Figure sacrificed over a 22 day period. The vertical bars Thevertical 22period.aday over sacrificed infected with S. aureus strain P78-22. The arithmetic mean arithmetic The P78-22.strain S. aureus infected with — - - - % 5 I V " T" 10 DAYS T 15 DIABETIC CONTROL 20 I 4

85

86

80-

» DIABETIC - 60 - III CO O o a

CO 2 40- o CO a < CONTROL

20-

l o DAYS

Figure 11. The glucose concentration of S. aureus strain P78-22 abscesses in control and diabetic mice. The arithmetic mean of abscesses glucose concentration of 4-10 control and diabetic mice infected with S. aureus strain P78-22 and sacrificed over a 22 day time course. The vertical bars represent the standard error of the mean. TABLE 16.

EVALUATION OF LIPEMIC EFFECT OF ABSCESS HOMOGENATE ON GLUCOSTIX (AMES)

GLUCOSE’: GLUCOSE: MIXTURE THEORETICAL SOLUTION ABSCESS READING” AVERAGE______DIFFERENCE

189 mg/dl 68 mg/dl 110 mg/dl 128 mg/dl 18

185 115 110 150 40

179 87 106 133 27

190 61 104 126 22

208 60 140 134 6

112 31 72 75 3 107 35 80 71 9

97 53 75 74 1

Glucose solutions were prepared using glycolyzed blood.

Readings for a mixture of 1 ml abscess homogenate and 1 ml of glucose solution. BIBLIOGRAPHY

1. Elek, S. D. Staphylococcus pvoaenes and its Relation to Disease. 1st. ed. Edinburgh and London: E & S Livingstone Ltd., 1959.

2. Cohen, J. 0. The Staphylococci. 1st ed. New York: John Wiley & Sons, Inc. 1972.

3. Moller, A. and B. Rydberg. Influence of cationic detergent on development of infection in experimental wounds contaminated with staphylococci. Acta Chiur. Scand. 135:459-465 (1969).

4. Winn, H. R., M. Mendes, P. Moore, C. Wheeler, G. Rodeheaver. Production of experimental brain abscess in the rat. J. Neurosurg. 51:685-690 (1979).

5. Anderson, J.C. The effect of teat damage on the incidence of experimental mastitis in the mouse. Res. Vet. Science. 13:390-402 (1972). 6. Agarwal, D. S. Subcutaneous staphylococcal infection in mice. I. The role of cotton dust enhancing infection. Brit. J. Exper. Path. 48:436-449 (1967).

7. Hays, R. C. and G. L. Mandall. p02, pH, and redox potential of experimental abscesses. Proc. Soc. Exper. Biol. Med. 142:29-30 (1974).

8. Isenberg, H. D., S. L. Wiener, G. A. Isenberg, J. Sampson-Scherer, M. Urivetzky and J. I. Berkman. Rat polyvinyl sponge model for the study of infections: initial investigations. Infec. Immun. 14:483-495 (1976).

9. James, C. R. and C. J. MacLeod. Induction of staphylococcal infections in mice with small inocula introduced on sutures. Brit. J. Exper. Path. 42:266-267 (1961).

10. Noble, W. C. The production of subcutaneous staphylococcal skin lesions in mice. Brit. J. Exper. Path. 46:254-262 (1965).

88 89

11. Taubler, J. H. and F. A. Kapral. Staphylococcal population changes in experimentally infected mice: infection with suture-absorbed and unabsorbed organisms grown in vitro and in vivo. J. Infect. Dis. 116:257- 262 (1966).

12. Kapral, F. A. Staphylococcus aureus: some host- parasite interactions. Annals of the New York Academy of Sciences 236:267-276 (1974).

13. Kapral, F. A. Clumping of s^. aureus in the peritoneal cavity of mice. Jour, of Bact. 92: 1188- 1195 (1966).

14. Kapral, F. A., J. R Goodwin and E. S. Dye. Formation of intraperitoneal abscesses. Infect. Immun. 30:204-211 (1980).

15. Dye, E. S. and F. A. Kapral. Survival of S. aureus in intraperitoneal abscesses. Jour. Med. Micro. 14: 185-194 (1981).

16. Dye, E. S. and F. A. Kapral. Partial characterization of a bactericidal system in staphylococcal abscesses. Infect. Immun. 30:198-203 (1980).

17. Shryock, T. R., E. S. Dye, and F. A. Kapral. Accumulation of bactericidal lipid within staphylococcal abscesses. (Manuscript in preparation).

18. Dye, E. S. and F. A. Kapral. Characterization of a bactericidal lipid developing within staphylococcal abscesses. Infect. Immun. 32: 98-104 (1981).

19. Engler, H. D. and F. A. Kapral. The production of a bactericidal monoglyceride in murine abscesses that are generated by Staphylococcus aureus. (Manuscript in preparation).

20. Kapral, F. A. and J. E. Mortensen. The inactivation of bactericidal fatty acids by an enzyme of jLj. aureus. In: Kabara, J. (ed) The Pharmacological Effects Of Lipids II. The American Oil Chemists Society, Champaign p. 103-109 (1985).

21. Erlich, H. A. HLA and insulin-dependent diabetes mellitus. Nature. 337:415 (1989).

22. Pyke, D.A. Introduction. Brit. Med. Bull. 45:1-3 (1989). 90

23. Kahn, C. R. and B. Goldstein. Molecular defects in insulin action. Science. 245:13 (1989).

24. Chenault Maurer, A. The therapy of Diabetes. Amer. Scient. 67:422-431 (1979).

25. Parfentjev, I. A. and S. W. Collins, Jr. The effects of the injection of glucose upon experimental infection in mice. Jour. Immun. 38.:137-141 (1940).

26. Pillsbury, D. M. and G. V. Kulcher. The relation of experimental skin infection to carbohydrate metabolism. Amer. Jour. Med. Sci. 190: 169-1178 (1935).

27. Dubos, R. J. Effect of ketone bodies and other metabolites on the survival and multiplication of staphylococci and tubercle bacilli. Jour. Exper. Med. 98: 145-155 (1953). 28. Bates, G. and C. Weiss. Delayed development of antibody to staphylococcus toxin in diabetic children. Amer. J. Dis. Child. 62:346-351 (1941).

29. Nolan, C. M., H. N. Beatty and J. D. Bagdade. Further characterization of the impaired bactericidal function of granulocytes in patients with poorly controlled diabetes. Diabetes. 27:889-894 (1978).

30. Mowat, M. E. and J. Baum. Chemotaxis of polymorphonuclear leukocytes from patients with diabetes mellitus. New Eng. J. Med. 284:621-627 (1971).

31. Drachman, R. H., R. K. Root, and W. B. Wood, Jr. Studies on the effect of experimental nonketotic diabetes mellitus on antibacterial defense. Jour. Exper. Med. 124:227-240 (1966).

32. Dunn, J. S., H. L. Sheehan, N. G. B. McLetchie. Necrosis of islets of Langerhans produced experimentally. Lancet. 1:484-487 (1943).

33. Bobr,J. The effect of alloxan diabetes on experimental staphylococcal infection in mice. Jour. Path. Bact. 89:749-752 (1965).

34. Krizek, T. J. and J. H. Davis. Experimental staphylococcal infection in the diabetic. Pancreas. 5:394-395 (1964). 91

35. Fiorio, C. and G. Garrone. Diabete e infezioni- ricerche spermentali.- Nota III. Diabete da allossana e tossina stafilococcica (potere dermonecrotico e portere letale). G. Batt. Immun. 40:49-70 (1948).

36. Cohn, Z. A. Determinants of infection in the peritoneal cavity II Factors influencing the fate of Staphylococcus aureus in the mouse. Yale Jour. Biol. Med. 25:29-47 (1962).

37. Rakieten, N., M. L. Rakieten, and M. V. Nadkarni. Studies on the diabetogenic action of streptozotocin (NSC-37917). Cancer Chem. Rep. 29:91-97 (1963).

38. Rerup, C. and F. Tarding. Streptozotocin- and alloxan- diabetes in mice. Eur. J. Pharmocol. 7:89-96 (1969).

39. Like, A. A. and A. A. Rossini. Streptozotocin-induced pancreatic insulitis: New model of diabetes mellitus. Science. 193:415-415 (1976).

40. Buschard, K. The thymus-dependent immune system in the pathogenesis of type I (insulin-dependent) diabetes mellitus. Dan. Med. Bull. 32:139-151 (1985).

41. Johanssan, E. B. and H. Tjalve. Studies on the tissue deposition and fate of [MC]streptozotocin with special reference to the pancreatic islets. Acta Endocrin. (Copenh) 89:339-351 (1978).

42. Agarwal, M. K. Streptozotocin: mechanism of action. FEBS Letters. (1980).

43. Beppu, H., K. Maruta, T. Kurnwer, and H. Kolb. Diabetogenic action of streptozotocin: essential role of membrane permeability. Acta Endocr. (Copenh). 114:90-95 (1987).

44. Yamamoto, H., Y. Uchigata, and H. Okamoto. Streptozotocin and alloxan induce DNA strand breaks and poly(ADP-ribose) synthetase in pancreatic islets. Nature. 294: 284-286 (1981). 45. Kolb, H. Mouse models of insulin dependent diabetes: low-dose streptozotocin-induced diabetes and nonobese diabetic (NOD) mice. Diabetes/Metabolism Rev. 2: 751- 778 (1987).

46. Kapral, F. A. and M. Gh. Shayegani. Intracellular survival of staphylococci. J. Exper. Med. 110:123-138 (1959). 92

47. Kapral, F. A. and M. M. Miller. Product of Staphylococcus aureus responsible for the scalded-skin syndrome. Infect. Immun. 4:541-545 (1971).

48. Melish, M. E. and L. A. Glasgow. The staphylococcal scalded-skin syndrome. New. Eng. J. Med. 282:1114- 1119 (1970).

49. Finney, D. J. Probit analysis, 3rd ed., Cambridge University Press, Cambridge (1966).

50. Sharma, H. M., E. M. Kauffman, and C. M. Conrad. An improved immunoperoxidase technique using horseradish peroxidase avidin D. Lab. Med. 20:109-112 (1989).

51. Dole, V. P. and H. Meinertz. Microdetermination of long-chain fatty acids in plasma and tissues. Jour. Biol. Chem. 235:2595-2599.

52. Kates, M. Techniques of lipidology. Isolation, analysis and identification of lipids. 2nd Rev Ed. Amsterdam: Elsevir Science Publishers. 1986. pp 196- 198.

53. Kuksis, A. Handbook of lipid research I. Fatty acids and glycerides. 1st Ed. New York: Plenum Press. 1978. p. 129.

54. Wheat, J. L. Infection and diabetes mellitus. Diabetes Care. 3:187-197 (1980).

55. Chandler, J. R. Malignant external otitis: further considerations. Ann. Otol. 86:417-428 (1977).

56. Zaky, D. A., D. W. Bentley, K. Lowy, R. F. Betts, and R. H. Douglas. Malignant external otitis: a severe form of otitis in diabetic patients. Amer. J. Med. 61:298-302 (1976).

57. Cohn, A. M. Progressive necrotizing otitis. Arch. Otolaryncol. 99:136-139 (1974).

58. Mentzer, R. M., Jr., G. T. Golden, J. G. Chandler, and J. S. Horsley, III. A comparative appraisal of emphysematous cholecystitis. Amer. J. Surg. 129:10-15 (1975).

59. Sheldon, W. H. and H. Bauer. The development of the acute inflammatory response to experimental cutaneous mucormycosis in normal and diabetic rabbits. J. Exper. Med. 110:845-858 (1959). 93

60. Bauer, H., J. F. Flanagan, and W. H. Sheldon. Experimental cerebral mucormycosis in rabbits with alloxan diabetes. Yale J. Biol. Med. 28:29-36 (1955).

61. Bauer, H., J. F. Flanagan, and W. H. Sheldon. The effects of metabolic alteration on experimental Rhizopus oryzae. Yale J. Biol. Med. 29:23-32 (1956).

62. Gale, G. R. Gas exchange in Rhizopus oryzae. J. Infect. Dis. 106:149-153 (I960).

63. Ellenberg, M. Diabetic neuropathy: clinical aspects. Metabolism 25:1627-1655 (1976).

64. Frumodt-Moller, C. Diabetic cystopathy I-IV. Dan. Med. Bull. 23:267-278, 279-286, 287-290, 291-294 (1976).

65. Cherenew, J. and A. Braude. Depression of phagocytosis by solutes in concentrations found in the kidney and urine. J. Clin. Invest. 41:1945-1953 (1962).

66. Edwards, M. S. and P. A. Fuselier. Enhanced susceptibility of mice with streptozotocin-induced diabetes to type II group B streptococcal infection. Infect. Immun. 39:580-585 (1983).

67. Kitara, K., T. Ishibashi, Y. Harada, M. Takamoto, and K. Tanaka. Reduced resistance to Pseudomonas septicaemia in diabetic mice. Clin. Exp. Immun. 43:590-598 (1981).

68. Rodman, H. M., M. Olszewski, D. Little, and T. Butler. Defective cell mediated immunity to infection in diabetes. Diabetes. 26(Suppl. 1):369 (1977).

69. Fromtling, R. A., G. K. Abruzzo, and D. M. Giltinan. Candida tropicalis infection in normal, diabetic and neutropenic mice. Jour. Clin. Micro. 25:1416-1420 (1987).

70. Tanowitz, H. B., B. Amole, D. Hewlett, and M. Wittner. Trypanosoma cruzi infection in diabetic mice. Trans. Royal. Soc. Trop. Med. Hyg. 82:90-93 (1988).

71. Mahmoud , A. A. F., A. W. Cheever, and K. S. Warren. Streptozotocin-induced diabetes mellitus and the host-parasite relation in murine schistosomiasis mansoni. Jour. Infec. Dis. 131:634-642 (1975). 94

72. Kochan, I. and C. A. Golden. Immunological nature of antimycobacterial phenomenon in macrophages. Infect. Immun. 9:249-254 (1974).

73. Hemsworth, G. R. and I. Kochan. Secretion of antimycobacterial fatty acids by normal and activated macrophage. Infect. Immun. 19:170-177 (1978).

74. Coonrod, J. D., R. L. Lester, and L. C. Hsu. Characterization of antibacterial factors of rat alveolar lining material. J. Clin. Invest. 74:1269- 1279 (1984).

75. Campbell, I. M., D. N. Crozier, A. B. Pawagi, and I. A. Buivids. In vitro response of Staphylococcus aureus from cystic fibrosis patients to combinations of linoleic and oleic acids added to nutrient medium. J. Clin. Micro. 18:408-415 (1983).

76. Sheu, C. W. and E. Freese. Lipopolysaccharide layer protection of Gram-negative bacteria inhibition by long-chain fatty acids. J. Bact. 115:869-875 (1973).

77. Galbraith, H. and T. B. Miller. Effect of long chain fatty acids on bactericidal respiration and amino acid uptake. J. Appl. Bact. 365:659-675 (1973).

78. Galbraith, H. and T. B. Miller. Physiochemical effects of long chain fatty acids on bacterial cells and their protoplasts. J. Appl. Bact. 36:647-658 (1973).

79. Coles, R.S. and H. C. Lichstein. The inhibition of malic enzyme of Lactobacillus arabinosus 17-5 by oleic acid I. Observations on the reaction. Arch. Biochim. Biophys. 103:186-190 (1963).

80. Sheu, C. W. and E. Freese. Effects of fatty acids on growth and envelope proteins of Bacillus subtilis. J. Bact. 111:516-524 (1972).

81. Butcher, G. W., G. King, and K. G. H. Dyke. Sensitivity of Staphylococcus aureus to unsaturated fatty acids. J. Gen. Micro. 94:290-296 (1976).

82. Kanai, K. and E. Kondo. Antibacterial and cytotoxic aspects of long chain fatty acids as cell surface events: Selected topics (A review). Jap. J. Med. Sci. Biol. 32:135-174 (1974). 95

83. Kabara, J. J., D. M. Swieczkowski, A. J. Conley, and J. P. Truant. Fatty acids and derivatives as antimicrobial agents. Micro. Agents and Chem. 2:23-28 (1972).

84. Kabara, J. J. and R. Vrable. Antimicrobial lipids: natural and synthetic fatty acids and monoglycerides. Lipids. 12:753-759 (1977).

85. Jergan, A. D., E. L. Nguyen, J.L. Sally, and V. M. Wong. Resistance of cats to induction of diabetes mellitus by administration of multiple doses of streptozotocin. FASEB Jour. 3,:A927. (1989).

86. Kiesel, U. and H. Kolb. Low-dose streptozotocin- induced autoimmune diabetes is under the genetic control of the major histocompatibility complex in mice. Diabetologia. 23:69-71 (1982).

87. Kromann, H., M.Christy, J. Egeberg, A. Lernmark, and J. Nerup. Absence of h-2 genetic influence on streptozotocin-induced diabetes in mice. Diabetologia. 23:114-118 (1982).

88. Wolf, J., F. Lilly, and S.-I. Shin. The influence of genetic background on the susceptibility of inbred mice to streptozotocin-induced diabetes. Diabetes. 33:567-571 (1983).

89. Kim, Y. T. and C. Steinberg. Immunological studies on the induction of diabetes in experimental animals. Cellular basis for the induction of diabetes by streptozotocin. Diabetes. 33.:771-777 (1984).

90. Skett, P. Sex-dependent effect of streptozotocin- induced diabetes mellitus on hepatic steroid metabolism in the rat. Acta Endocr. 111:217-221 (1986).

91. Paik, S.-G., M. A. Michelis, Y. T. Kim, and S. Shin. Induction of insulin-dependent diabetes by streptozotocin. Inhibition by estrogens and potentiation by androgens. Diabetes. 31:724-729 (1982).

92. Mailman, W. L., A. Leavitt, and D. A. Joslyn. Effect of medium on resistance of Staphylococcus aureus to quaternary ammonium compounds. Proc. Soc. Exp. Biol. Med. 68:401-403 (1948). 96

93. Hayashi, H., H. Hamada, Y. Hirari, R. Kariyama, I. Koujima, and Y. Kanemasa. Effect of glucose and oxygen on the structure of the plasma membrane of Staphylococcus aureus. Acta Med. Okayama 33.:379-387 (1979).

94 Dvorak, A. M. Lipid bodies: cytoplasmic organelles important to arachidonate metabolism in macrophages and mast cells. J. Immun. 131:2965-2976 (1983).

95. Schlayer, S. J. and M. S. Meltzer. Macrophage activation for tumor cytotoxicity: analysis of cellular lipid and fatty acid content during lymphokine activation. J. Reticul. Soc. 29:227-240 (1980).

96. Saiki, 0., S. Negoro, I. Tsuyuguchi, and Y. Yamamura. Depressed immunological defense mechanisms in mice experimentally induced diabetes. Infect. Immun. 28:127-131 (1980).

97. Jones, C. A., M. F. Seifert, and P. K. Dixit. Macrophage migration inhibition in experimental diabetes (41434). Proc. Soc. Exper. Biol. Med. 170:298-304 (1982).