THE ELABORATION OF EXTRACELLULAR CAPSULAR …

THE ELABORATION OF EXTRACELLULAR CAPSULAR POLYSACCHARIDE BY

KLEBSIELLA PNEUMONIAE AND ITS RELATIONSHIP TO VIRULENCE

By

PHILIP DOMENICO, B.A.

DISSERTATION

IN

MICROBIOLOGY

Presented to the Graduate Faculty of Texas Tech University Health Sciences Center

in Partial Fulfillment of the Requirements for

the Degree of

DOCTOR OF PHILOSOPHY

Approved

Acc'epted

December, 1983

f^C-'^'^ . ACKNOWLEDGEMENTS

I extend my gratitude to all who have made this dissertation

possible:

To my supervising professor, Dr. David C. Straus, who

encouraged me when I was frustrated and frustrated me when I was

encouraged.

To Dr. Dana L. Diedrich, who enriched my scientific repertoire

with his insight into microbial biochemistry and physiology, and

who burned the late-night candle with me on numerous occassions.

To Dr. Charles W. Garner for his guidance and his knowledge of

chemical phenomena.

To Dr. Rial D. Rolfe and Dr. David J. Hentges for their

assistance and constructive criticisms.

To Cathy Portnoy Duran who put up with my mess for two years.

To Linda Froelich, who inspired me and helped me through a

variety of difficulties during this period.

And finally to my parents who taught me the gift of

perseverence.

n

I-V"*.

TABLE OF CONTENTS

Page

ACKNOWLEDGEMENTS i i

LIST OF TABLES vi

LIST OF FIGURES x

LIST OF ABBREVIATIONS xiii

I. INTRODUCTION AND LITERATURE REVIEW 1

II. MATERIALS AND METHODS 11

Bacterial Strains 11

Media and Growth Conditions 11

Purification of the Extracellular Polysaccharides

of K_. pneumoniae 12

Preparation of Rabbit Anti-Type-Specific Antiserum.. 16

Rocket Immunoelectrophoresis 17

Opsonophagocytic Assay and Serum Sensitivity 18

Assay for Virulence of l<. pneumoniae in a Mouse

Model 20

Assay for the Production of Lobar Pneumonia in a

Rat Model 21

Assay for Characterization of Outer Membrane

Proteins of j<. pneumoniae 23

Electrodialysis of Extracellular

Polysaccharides from l<. pneumoniae 24

m

Page

Determination of Capsule Size of J<. pneumoniae 26

In Vitro Quantitation of Extracellular

Polysaccharides Produced by K. pneumoniae 26

Electron Microscopy 27

Gel Diffusion Method for Immunological Analysis 27

Saponification of K. pneumoniae Polysaccharides

and Quantitation of Fatty Acids 28

Hydrofluoric Acid Treatment 29

Statistical Analysis 29

III. RESULTS 30

Strain Variation and the Production of Apparent

Isogenic Sets 30

The Establishment of a Chronic Lobar Pneumonia by J<.

pneumoniae in a Rat Model 32

J<. pneumoniae Virulence in a Mouse Model 59


Polysaccharides Produced by J<. pneumoniae 61

Serum Sensitivities and Opsonophagocytic Assays

for J<. pneumoniae 75

Purification of the EPS of J<. pneumoniae 80

Effect of Purified Extracellular Products from K_.

pneumoniae on Virulence in a Mouse Model 126

Structural Studies on the EPS Produced by

J<. pneumoniae 146

Gel Immunodiffusion Studies for Identification and

Quantitation of ECPS Produced by K_. pneumoniae.. 186

iv

Page

Survey of the Outer Membrane Proteins of

J<. pneumoniae 194

IV. DISCUSSION 200

LITERATURE CITED 225

LIST OF TABLES

Page

1. Capsule size of K. pneumoniae 31

2. Establishment of a chronic Lobar Pneumonia in Rats 37

3. Effect of Dosage on the Ability of KPl to Produce

Pneumonia in Rats 51

4. Establishment of a Chronic Lobar Pneumonia in Rats

Emp 1 oyi ng KP1 -0 54

5. KPl-T in the Rat Lung Model 55

6. KP2-0 in the Rat Lung Model 57

7. KP2 2-70 in the Rat Lung Model 58

8. LDcn Values in Mice and ID^n Values in Rats for Strains

50 50

of J<. pneumoniae 60

9. ECPS Production by Strains of K_. pneumoniae Serotype 1 at

Various Intervals of Incubation 62

10. ECPS Production by Strains of j<. pneumoniae Serotype 2 at

Various Intervals of Incubation 68

11. Production of ELPS by K. pneumoniae Serotypes 1 and 2

after 48h of Culture in Defined Medium 71

12. Comparison of ECPS, ELPS, Capsule Size and Virulence of

J<. pneumoniae Serotypes 1 and 2 73

13. Correlations Between Polysaccharide Production and

Virulence in the Mouse Model 74

14. Serum Sensitivity of K. pneumoniae 76

15. Opsonophagocytic Assay 78

16. Effect of the Addition of EPS on the OPA 79

VI

Page

17. Elution Ionic Strength and Apparent Molecular Weights of

the ECPS from Various Strains of KPl and KP2 82

18. Elution Volumes for Dextran Calibration Standards on

Sepharose 2B (S-2B) 103

19. The Extracellular Products Found in Ethanol Fractionated

Supernatants of K_. pneumoniae 104

20. Comparison of the ECPS and ELPS Content in the Neutral

(N) and Acidic (A) Fractions from DEAE-Sephacel 106

21. The Extracellular Products Found in KPl and KP2 EPS after

Purification 107

22. Percent Yield Obtained from ECPS Purification for Two KPl

Strains 109

23. Pur i f i ca t ion of KPl-0 (EtOH) EPS by ED, cetavlon and Gel

Fi 1 t r a t i on I l l

24. Pur i f i ca t ion of KPl-0 (EtOH) EPS by ED, Cetavlon and Gel

F i l t r a t i o n : Percent Contamination wi th ELPS and

Protei n 115

25. Purification of KP2-0 EPS by ED and Cetavlon 121

26. Purification of KP2 2-70 EPS by ED and Cetavlon 123

27. Effect of KPl EPS on KPl-T Virulence in the Mouse Model.. 127

28. Probability Matrix Comparing the Virulence Enhancing

Potentials for all EPS Fractions Co-injected with

the KPl-T Strain in the Mouse Model 129

29. Effect of KP2 EPS on KPl-T Virulence in the Mouse Model.. 132

v n

-y^mmfia

Page

30. Effect of KPl or KP2 EPS on the Virulence of KP2-0 in

the Mouse Model 134

31. Effect of ED on the Virulence Enhancement of KPl-T by

KP2 2-70 EPS in the Mouse Model 137

32. Effect of ED on the Virulence Enhancement of KPl-T by

KP2-0 EPS in the Mouse Model 138

33. Effect of Saponification on the Virulence Enhancement

of KPl-T and KP2-0 by KPl-0 (N) EPS

in the Mouse Model 140

34. Virulence Enhancement of KPl-T in the Mouse Model:

Comparison to the Dosage of ELPS in KPl EPS

Sampl es 141

35. Virulence Enhancement of KPl-T in the Mouse Model:

Comparison to the Dosage of ELPS in KP2 EPS

Sampl es 142

36. Effect of an Al ternat ive Pur i f i ca t ion of KPl-0 EPS on

the Virulence Enhancement of KPl-T in the

Mouse Model 145

37. The pH D i f fe ren t ia l of the Anode and Cathode Chambers

During ED 161

38. Quantitation of Ions Retrieved from the Cathode and the

Anode Chambers During ED of KP2 2-70 (EtOH) EPS 163

39. Effect of ED on the Quantity of Divalent Cations Found

in KP2 EPS 164

vm

IV

Page

40. Effect of ED on the Quantity of Phosphate Found in the

KP2 EPS 165

41. Quantitation of Fatty Acid Methyl Ester (FAME) Released

from EPS after Saponification 181

42. Quantitation of FAME Released from the EPS of KPl-0

and KPl-T Obtained from Gel Filtration 182

43. Standard Curve for the Rocket Immunoelectrophoresis

(RIE) of KPl-0 HMW and KPl-0 LMW EPS 187

44. RIE of Standard Concentrations of KP2 2-70 ECPS and

Serum from an Infected Rat 189

45. Quantitation of KP2 2-70 EPS in the Serum of an

Infected Rat by Radial Immunodiffusion (RID) 193

IX

LIST OF FIGURES

Page

l a . Transmission Electron Micrograph (TEM) of KP2-0 and i t s

Capsule 34

36 lb. TEM of KP2-T and its Capsule

2. Photomicrocrographs of H&E Stained Lung Tissue Sections

During J<. pneumoniae Infection in Rats 39

2a. Normal Rat Lung Section 41

2b. Rat Lung Section at 24h Post-inoculation 44

2c. Rat Lung Section at 3 days Post-inoculation 46

2d. Rat Lung Section at 6 days Post-inoculation 48

2e. Rat Lung Section at 9 days Post-inoculation 50

3. Comparison of the Rate of Production of ECPS by KPl-0 and

KPl-T at Various Intervals of Incubation 64

4. TEM of KPl 2-70: Example of Capsule Sloughing

5. Comparison of the Rate of Production of ECPS by KP2-0

and KP2 2-70 at Various Intervals of Incubation.

66

6.

7.

8.

9.

10.

11.

12.

13.

Elution

Elution

Elution

Elution

Elution

^ Elution

Elution

Elution

Profi

Profi

Profi

Profi

Profi

Profi

Profi

Profi

le fo r KPl-T EPS on DEAE-Sephacel

le fo r KPl-0 (A) EPS on Sepharose 2B (S-2B)

le fo r KPl-0 (N) EPS on S-2B

le fo r KPl-T (A) EPS on S-2B

le fo r KPl-T (N) EPS on S-2B

le fo r KP2-0 (A) EPS on S-2B

le fo r KP2-0 (N) EPS on S-2B

le for KP2 2-70 (A) EPS on S-2B

70

84

87

89

91

93

95

97

99

Page

14. Elution Profile for Dextran Calibration Standards

on S-2B 101

15. Elut ion Pro f i le fo r KPl-0 F r I I on S-2B 114

16. Elut ion Pro f i le for KPl-0 Fr I I I on P-300 118

17. Elut ion Pro f i le fo r KP2-0 F r I I on S-2B 120

18. Effect of an Al ternat ive Pur i f i ca t ion on the Elut ion

Pro f i le for KP2 2-70 EPS on S-2B 125

19. Effect of ED on the Elut ion Pro f i le for KP2 2-70 EPS on

BGA-150m 148

20. Effect of ED on the RID pro f i les of KP2 2-70 EPS 151

21. Effect of ED on the Elution Pro f i le fo r KP2 2-70 LMW

EPS on BGA-150m 154

22. Effect of ED on the Elution Pro f i le for KP2-0 EPS

on S-2B 157

23. Effect of ED on the Elution Profile for KP2-0 EPS on

S-2B: Effect of a Small Sample Volume 159

24. Histogram of the Effect of ED on the [PO^"^] in the

KP2 2-70 EPS 167

25. Effect of Sodium Dodecyl Sulfate (SDS) on the Elut ion

Pro f i le fo r Electrodialyzed KP2-0 EPS on S-2B 171

26. Effect of ED on the Elut ion Pro f i le for KPl-0 EPS

on S-2B 173

27. Effect of Hydrofluoric acid (HF) on the Elut ion

Pro f i l e fo r KPl-0 EPS on S-2B 177

XT

Page

28. Effect of Saponification on the Elution Profi le for

KPl-0 EPS on S-2B 179

29. Elution Profile for KP2 2-70 (EtOH) EPS Obtained from

Growth in DMH on BGA-150m 185

30. Standard Curve for RIE of KP2 2-70 ECPS 192

31. Outer Membrane Protein Profiles for KPl Strains 196

32. Outer Membrane Protein Profiles for KP2 Strains 198

33. Hypothetical Model for the Electrophilic Associations

Between Strands of KP2 2-70 ECPS 219

x n

LIST OF ABBREVIATIONS

AB antibody/antiserum

BGA-150m BioGel A 150m gel filtration resin

BSA bovine serum albumin

Ca calcium

CET cetavlon

CFU colony forming units

CPS capsular polysaccharide

DEAE diethyl amino ethyl ion exchange resin

DHpO deionized water

DMH defined medium with Hepes buffer

DW defined medium with phosphate buffer

ECPS extracellular capsular polysaccharide(s)

ED electrodialysis

ELPS extracellular lipopolysaccharide(s)

EPS extracellular polysaccharide(s)

EtOH ethanol

FAME fatty acid methyl ester

Fr I fraction I

Fr II fraction II

Fr III fraction III

GLC gas liquid chromatography

H&E hematoxylin and eosin

HIAB heat-inactivated antibody

• • •

xm

HIRS heat-inactivated rabbit serum

HMW high molecular weight

ID^Q 50% infectious dose

IP intraperitoneal

KDO ketodeoxyoctanate

KPl Klebsiella pneumoniae serotype 1

KPl-0 J<. pneumoniae serotype 1 (ATCC 8047), opaque

variant

KPl-Or KPl-0 revertant

KPl-T l<. pneumoniae serotype 1 (ATCC 8047),

translucent variant

KPl-Tr KPl-T revertant

KPl 2-70 K.. pneumoniae serotype 1 CDC 2-70

KP2 K.. pneumoniae serotype 2

KP2-0 K.. pneumoniae serotype 2 (ATCC 29011), opaque

variant

KP2-T K,. pneumoniae serotype 2 (ATCC 29011),

translucent variant

J<. pneumoniae serotype 2 (CDC 2-70)

K. pneumoniae serotype 2 (ATCC 8052)

Limulus amoebocyte lysate

50% lethal dose

low molecular weight

lipopolysaccharide

milliamperage

minimal essential medium

xiv

KP2

KP2

LAL

LD50

LMW

LPS

MA

MEM

2-70

8052

1

Mg magnesium

MW molecular weight

ND none detected

NP not performed

OD optical density

OPA ppsonophagocytic assay

PBS phosphate-buffered saline

PMN polymorphonuclear neutrophil -3

PO. phosphate

PPT precipitate

RID radial immunodiffusion

RIE rocket immunoelectrophoresis

RS rabbit serum

S-2B Sepharose 2B gel filtration resin

SCD surface charge density

SDS sodium dodecyl sulfate

SUPE supernatant

TBC total bacterial count

TD transverse diameter(s)

TEM transmission electron micrograph

TSA Trypticase Soy Agar

TSB Trypticase Soy Broth V volts

WBC white blood cells

XV

CHAPTER I

INTRODUCTION AND LITERATURE REVIEW

The prolonged survival of chronically and critically ill

patients, due to the increased quality of medical care in this

country, has been paralleled by a striking increase in the

occurrence of gram-negative bacterial infections (65). This has

been especially true in the last three decades, when

hospital-acquired infections due to gram-negative bacilli have been

quite devastating (65,77). Hospital-acquired pneumonias caused by

these organisms have increased to where they now comprise nearly 50

percent of all nosocomial pneumonias (38).

One gram negative rod, Klebsiella pneumoniae, accounts for 25

to 43 per cent of gram-negative nosocomial pneumonias, thus making

it the most common agent in this disease process (77). Pneumonia

caused by K. pneumoniae is particularly dangerous, because once it

is established, it is difficult to control (38,65) and mortality

rates may reach or exceed 50 percent, even in treated cases (34,

41, 53). K.. pneumoniae pneumonias differ from most other

pneumonias in that lung tissue destruction seen in this disease

process is often extensive (70). Little is known about why such

extensive tissue necrosis is seen in this form of pneumonia.

The association between the ability of bacteria, such as J<.

pneumoniae and Streptococcus pneumoniae (serotype 3), to produce

large quantities of capsular polysaccharide (CPS) and to cause a

destructive lobar pneumonia is highly suggestive of a relationship

between these two parameters. Undoubtedly the rate of production

1

of CPS for these organisms is intimately associated with their

pathogenicity (28,45,55), and generally differentiates these

species from other medically important bacteria. Many theories

have been proposed as to how the CPS functions as a virulence

factor, most of which regard the CPS as an antiphagocytic substance

(45). Fukutome et al. (33) were able to show that j<. pneumoniae

could not be phagocytosed by polymorphonuclear leukocytes (PMN) nor

by alveolar macrophages in the presence of normal serum, unless

anti-capsular antibody (AB) was present. Escherichia coli strains

possessing CPS have also been shown to be resistant to phagocytosis

by PMN in normal serum, in contrast to £. coli strains without CPS

(78). These investigators presented evidence that the decreased

phagocytosis of encapsulated strains was caused by a low rate of

complement activation of the strains, as shown by the absence of

C3b or C3d fixation to the cell wall of the bacteria. Verbrugh et

al. (79) showed that encapsulation of several bacterial species

interfered with the process of C3 fixation in normal human serum.

Coonrod et al. (15) showed that systemic decomplementation of-rats

did not affect the severity of J<. pneumoniae pneumonia.

It is apparent, then, that the bactericidal and opsonic

properties of normal serum are ineffective against certain

encapsulated gram-negative organisms. The capsule is thought by

some to provide a cover for certain bacterial structures that are

known to be reactive with bactericidal and opsonic components found

in normal serum. For example, it was demonstrated that complement

component CI directly interacts with bacterial lipopolysaccharides

(LPS) and lipid A, independent of AB, while retaining its

esterolytically active form (52). Moreover, pure or soluble

polysaccharides are generally poor immunogens (31), and it may be

that the capsule also functions to make the exterior surface of

encapsulated bacteria comparatively unreactive, immunologically

speaking.

Although the presence of a certain amount of cell associated

capsule is regarded as necessary for the virulence of J<.

pneumoniae, the presence of additional cell wall-associated capsule

may not necessarily make the organism more virulent. For example,

Mizuta et al. (57) showed that, of 9 Klebsiella 01:K2 strains, 7

were highly virulent, whereas the other 2 strains were avirulent,

even though they were encapsulated to the same extent as the

virulent strains. One possible explanation for the role of the

capsule in pathogenicity has to do with the density rather than the

size of the capsule (20,82). It is possible that a more dense

capsular network could better inhibit nonspecific defense

mechanisms, such as C3 binding, from gaining access to the cell

wall of the bacillus. Recent studies by Wilkinson et al. (47,82),

and others (84), have suggested that the capsule of many bacteria

is readily penetrable by high molecular weight proteins, such as AB

and complement. Whether the same holds true for the virulent

strains of K.. pneumoniae has yet to be determined.

Another possible role for the capsule in virulence deals with

what can be referred to as the surface charge density (SCD). The

SCD can be construed as the net negative potential of the

polysaccharide that can interact with the environment. It may be

that the negative charge on the surface of phagocytic cells and the

negatively charged polysaccharide polymers of encapsulated bacteria

tend to repel one another, thereby explaining the antiphagocytic

nature of bacterial capsules. Perhaps because acidic

polysaccharides are able to avidly bind divalent cations (9, 16),

the CPS may create a microscopic zone around the organisms where

defense mechanisms dependent on the presence of these cations

(i.e., complement activation and initiation of phagocytosis) are

unable to function. In l<. pneumoniae, glucuronic acid is the

component accounting for most of the negative charge of the

capsule, and it is found in the repeating unit structure of the

polymer in most of the l<. pneumoniae serotypes (39,59). Appendix 1

shows the repeating unit structure found in the capsule

polysaccharides from K_. pneumoniae serotypes 1 and 2. Since most

serotypes possess uronic acids in more or less the same ratio (31

and 26 percent of the repeat unit weight for KPl and KP2,

respectively), this alone cannot explain the broad range of

virulence seen among, and especially within, the 72 serotypes of K.

pneumoniae. Virulent strains of Cryptococcus neoformans however,

have been shown to produce CPS having a greater uronic acid content

and a larger molecular size than relatively nonvirulent variants

within the same population (48). Additional negative potential may

be imparted to the Klebsiella capsule by covalently linked

non-carbohydrate substituents, such as pyruvate, acetate and

formate (35,76). These organic acids are detected in some, but

5

not all strains of a given serotype in various quantities. The KPl

ECPS has been shown to have an additional pyruvyl group linked to

its repeat unit structure (30) whereas only some KP2 ECPS have

these organic acids.

A third explanation for the role of the CPS in virulence is

related to the large quantities of these substances that are

produced and exuded into the medium by J<. pneumoniae (28). This

extracellular capsular polysaccharide (ECPS) may serve as an

antiphagocytic structure for K_. pneumoniae in a variety of ways.

The production of large amounts of ECPS provides an increasing

viscosity to solutions. It may be that the high producers of ECPS

are protected in vivo from intruding immune defense cells by

slowing the flow of particles in their iminediate environment. A

zoogleal mass, such as this could surround a microcolony of

bacteria, which may leave it relatively impenetrable to phagocytic

cells. Another virulence enhancing mechanism for the ECPS could be

to compete with the cell-associated capsular material for AB

produced against the capsular polysaccharide. Circulating

cell-free CPS in the blood of a patient infected with K. pneumonaie

could conceivably neutralize any previously or newly synthesized

antibodies before opsonization of the bacterium occurred (66).

Another explanation for the role of ECPS as a virulence factor

is its ability to paralyze the immune system against challenge with

the homologous (type-specific) bacterium or against heterologous

immunogens. Batshon et al. (2) induced immunological paralysis

against homologous challenge IP in mice with the ECPS isolated from

a KP2 strain. In the groups of animals receiving 2.5 yg of KP2

ECPS, there was a greater proportion of survivors than in groups

given 750 yg of ECPS, especially within the first 20 days after

ECPS administration. Protective AB were detected within 5 days in

the serum of mice given 2.5 yg, whereas in the serum of animals

given 750 yg of KP2 ECPS, such AB were not evident before 60 days.

Nakashima et al. (58) were able to show similar phenomena using the

ECPS from KPl strains. They found that anti-ECPS titers on day 11

post-administration were highest when mice were injected IP with 1

or 10 yg of KPl ECPS, but when 100 or 1000 yg of ECPS were

injected, virtually no AB titers were evident at this time period.

These same authors were able to show an increased response to

bovine serum albumin (BSA) when the mice were pre-injected with

these low ECPS doses, but a suppressed response compared to

controls occurred at the higher doses. Therefore, the effects of

KPl or KP2 ECPS on the immune system of mice appear to vary with

the amount introduced, with an adjuvant effect seen at low dosages,

and a suppression or tolerance phenomenon seen at higher dosages.

Pollack (66) demonstrated that the presence of detectable CPS in

the serum of human patients infected with J<. pneumoniae appeared to

correlate with the severity of infection, with persistence of

active foci (i.e., lung infections), and with a poorer prognosis

than in those patients who had no detectable circulating CPS. This

phenomenon was also observed in a study with S_. pneumoniae (19) and

group C meningococcemia (42) in humans, and for the type III group

B Streptococcus in a mouse model (20).

A series of publications by Yokochi et al. (85, 86)

demonstrated that minute quantities of KPl ECPS (0.05 mg/ml)

inhibits the maturation and functional capacity of macrophages.

They also showed that co-injecting a Salmonella strain IP in mice

with 200 yg of their KPl ECPS preparation markedly increased the

virulence of the Salmonella strain over controls without ECPS

treatment. Electron micrographs of peritoneal fluids showed that

the Salmonella were being phagocytosed in both the control and the

experimental groups, but the peritoneal macrophages were seen to be

killing and digesting the bacteria in the control group, whereas no

evidence for bactericidal or digestive processes were observed in

the treated group.

The production of CPS by J<. pneumoniae, both in the form of

cell-associated and soluble, cell-free (ECPS) polysaccharide, has

been studied extensively (20,21,26,27,46). Most of these studies,

however, did not relate the production of CPS or ECPS to virulence.

Early studies with the pneumococcus and Klebsiella demonstrated the

importance of the rate of capsule production in virulence. In a

classic study, MacLeod and Krauss (55) demonstrated the

relationship of virulence of pneumococcal strains for mice to the

quantity of CPS formed in vitro. Among several strains within

three different pneumococcal serotypes, they found that the

virulent strains formed more CPS than moderately virulent or

avirulent strains. Ehrenworth and Baer (28) reported a similar

phenomenon with a Klebsiella pneumoniae isolate. In both studies

the cell-associated CPS and the ECPS were measured, and virulence

8

was correlated with total CPS production. These investigators also

found a direct relationship between total CPS production and

virulence. Virulence was correlated with capsule size as

determined by packed-cell volume, and with soluble CPS as measured

by quantitative precipitin tests. One KP2 strain and three

variants of this strain were examined both for total CPS production

and virulence as determined by IP injections in mice. It was found

that the parent strain had nearly twice the packed-cell volume and

two times the antibody-precipi table soluble substance as did two of

the three variants at 3 h culture, yet no difference in virulence

was seen among these populations. Even at 24 h of incubation the

parent strain produced nearly twice the packed-cell volume and 1.5

times the ECPS as the two other variants with the same virulence

potential. Finally the remaining variant was shown to possess a

much smaller capsule and produced a somewhat smaller amount of

soluble antigen than the two other KP2 variants at 3 h of

incubation, but by 24 h of culture this strain increased production

of both capsular and soluble CPS to equal the total CPS production

of the two other variants. This last KP2 sub-strain was then shown

to be at least 5 log-jQ units less virulent in the mouse model than

all the other KP2 strains used. The authors concluded that the

slower rate of production of CPS by this latter KP2 variant was

slower than the other strains in the earlier stages of culture; it

was this slower rate which made it less virulent. Assuming that

the rather crude and outdated methodology used by these

investigators was approximating the actual total CPS production.

one is still not able to draw the conclusion that the rate of

production of CPS was the determining factor for virulence. Their

data would have been more convincing had they been able to show

differences in virulence between the parent strain, which produced

by far the most CPS at all intervals, and all the other variants

derived from this strain.

It was the intent of this proposal to further clarify the role

of the capsular substance; especially that of the ECPS, in the

pathogenicity of j<. pneumoniae. Moreover, careful consideration

was also given for other cell wall substances, such as LPS and

outer membrane proteins, for their possible functioning in these

virulence phenomena, and in their structural association with the

CPS of J<. pneumoniae. The present study addresses, to a

considerable extent, a number of structural issues heretofore

not reported in the literature, while attempting for the most part

to relate these structural phenomena to pathogenesis. One of the

inherent difficulties in working with the CPS of gram-negative

bacteria is the difficulty in obtaining CPS free from LPS. One

important question that is addressed in this study is an

examination of the nature of the association between extracellular

CPS and LPS and their relative roles as virulence factors.

Although Salmonella and £. coli LPS have been shown to possess a

wealth of biological activities, little is known of the LPS

produced by j<. pneumoniae. Reports have indicated that the LPS of

j<. pneumoniae is a more powerful adjuvant than the £. coli LPS (58,

10

87). The following, then, is an examination of the cell wall

associated structures of l<. pneumoniae and their relationship to

the virulence of the organism for mice.

CHAPTER II

MATERIALS AND METHODS

Bacterial Strains

Two strains of Klebsiella pneumoniae serotype 1 (KPl) and two

strains of l<. pneumoniae serotype 2 (KP2) were utilized in these

studies. The KPl strains are as follows: KPl ATCC 8047, lung

isolate, and KPl CDC 2-70 (Difco Labs, Detroit, MI). KPl ATCC 8047

was found to have two predominant variants seen on Trypticase Soy

Agar (TSA) (BBL, Division Beckton Dickinson Co., Cockeysville,

MD.), one opaque and the other translucent. These were isolated

and designated KPl-0 (opaque) and KPl-T (translucent). The KP2

serotypes employed in these studies were KP2 ATCC 29011, blood

isolate, and KP2 CDC 2-70 (Difco). KP2 ATCC 29011 was also found

to contain two variants, one opaque and one translucent. These

were subsequently isolated and designated KP2-0 and KP2-T. For the

fatty acid analyses the KP2 ATCC 8052 strain (bronchus isolate) was

utilized, but not included elsewhere in these studies. All strains

were maintained frozen in Trypticase Soy Broth (TSB) (BBL, Division

Becton, Dickinson and Co., Cockeysville, MD) with 20% glycerol at

-70°C. Prior to use, the stock cultures were added to liquid

medium (1 ml of stock for each 100 ml medium) and incubated at 37°C

in a shaking water bath (200 rpm) until the medium was slightly

turbid (ODggQ = 0.20).

Media and Growth Conditions

Several liquid media were used in the course of these studies.

11

12

TSB was prepared and sterilized according to manufacturers'

instructions and was mainly used for growing the organisms for

virulence studies and for use in the phagocytic and serum

sensitivity assays.

A chemically defined medium (DW) was prepared according to the

procedure of Duguid and Wilkinson (21) with one modification,

namely that zinc sulfate was added to a concentration of 0.0085 mM.

This medium was mainly used for growing the organisms for

quantitating and purifying the extracellular capsular

polysaccharide. In one study the phosphate buffer in this defined

medium was replaced with a 10 mM Hepes buffer, pH 7.3 (DMH).

For experimental procedures utilizing DW or TSB, all cultures

were grown at 37 C at 200 rpm in various volumes. Starter cultures

were grown to early logarithmic phase (ODccn = 0-2) in the

appropriate medium and 1 ml was inoculated per 100 ml of broth.

Growth was monitored by measuring the absorbance in a Bausch and

Lomb Spectronic 20 at 550 nm. Plate counts of bacteria on TSA were

routinely performed for each study utilizing DW medium. Cultures

were grown in TSB to early logarithmic phase (OD^^Q =0.2) and the

cells were harvested at this time. Cultures in DW medium were

grown for 48 hours before the cells were harvested unless otherwise

indicated. To terminate growth, cultures were immediately placed

in ice before centrifugation.

Purification of the Extracellular Polysaccharides

of K. pneumoniae

Cultures were grown in DW medium as described in Materials and

Methods, part B. Usually one liter of medium was used for each

13

organism. Two hundred ml samples were removed from the culture

after 18, 24, 36 and 48 h of growth. The total number of viable

organisms was monitored at each time period by examining colony

forming units (CFU) on TSA at appropriate dilutions. Samples were

centrifuged in a Beckman Model J2-21 refrigerated centrifuge

(Beckman Instruments, Inc., Palo Alto, CA) at 4°C at 17,700 x g for

60 min in a JA-10 rotor. The supernatant was collected and 2

volumes of cold ethanol (95%): methanol (19:1) were added to

precipitate the extracellular polysaccharides (EPS). These

solutions were left overnight stirring at 4°C or statically at

-20 C. The supernatant was then discarded or, if not clear, was

centrifuged for 30 min as above. Twenty ml of deionized water

(DH2O) was then added to the precipitate per 100 ml of original

culture volume which allowed the precipitate to go back into

solution. The crude EPS was then dialyzed with three changes

against 8 liters of DHpO at 4°C overnight. Samples were then

shell-frozen and lyophilized on a Virtis Freezemobile 6 lyophilizer

(Virtis Co., Inc., Gardiner, N.Y.).

Dried samples were dessicated over Drierite (W.A. Hammond

Drierite Co., Xenia, Ohio) overnight before being weighed out.

Various quantities of crude, dry EPS were weighed on a Mettler H51

balance (Mettler Instrument Corp., Hightstown, N.J.) and suspended

in an appropriate buffered solution. Early studies incorporated

DEAE-Sephacel ion exchange chromatography (2.5 x 25 cm column)

(Pharmacia Fine Chemicals AB, Uppsala, Sweden) as the next

purification step. The ethanol extracted, dried EPS was brought up

14

in 0.02 M (NH^)2C03 (Sigma Chemical Co., St. Louis, MO). The

DEAE-Sephacel column was equilibrated with the 0.02 M (NH4)2C03

solution and the sample (usually 10 ml) was placed on the column.

After allowing approximately 200 ml to elute from the column a

gradient was applied, which was normally from 0.02 M to 1.0 M

(NH^)2C02. The EPS elution profile was monitored by three

different methods: 1) a capillary precipitin reaction using rabbit

antiserum against whole, formalin killed cells of KPl or KP2; 2)

the anthrone assay (51) for total hexose and; 3) the uronic acid

assay of Blumenkrantz and Asboe-Hansen (8). Fractions were pooled

and dialyzed 3 times against 8L DHÔ at 4°C. The final

purification step was gel filtration on Sepharose 2B (S-2B) or

Sepharose 4B (S-4B), (Pharmacia), or BioGel A 150m (Bio-Rad Labs,

Richmond, CA) in 1 meter x 2.5 cm columns. Initial studies

utilized 0.5 M NaCl, both for the column buffer and for

resuspending samples. It was later determined that 0.01 M Tris, pH

12 (Sigma Chemical Co., St. Louis, MO) was ideal for the column

buffer and for bringing up samples of EPS to be placed on these

columns. Tris buffer (0.01 M, pH 12) was also used for subsequent

ion exchange chromatography. Various molecular weight (MW)

fractions were pooled separately and dialyzed against three 8L

changes of DH2O at 4°C while stirring, and lyophilized to dryness."

The various fractions were dessicated over CaSO. under a

vacuum, weighed and were then suspended in deionized water and

analyzed for their hexose and uronic acid content as well as for

the amount of protein present, as determined by the method of Lowry

15

(54), and for their lipopolysaccharide (LPS) content by the method

of Osborn et al. (60) using the LPS of Escherichia coli 055:B5

(Difco) as the standard. Phosphate content, as measured by Chen et

al. (13), as well as the calcium and magnesium content, assayed for

by atomic absorption spectrophotometry on a Perkin Elmer AAS model

303, were also determined.

In order to obtain a better purification procedure for the KPl

EPS, the following protocol was used. One hundred fifty mg of

ethanol extracted material from 48h supernatants of KPl-0 were

electrodialyzed at 2000 V, as described in Section J of Materials

and Methods. The electrodialyzed sample was then extracted with

10% cetavlon (docecyl trimethyl ammonium bromide, Sigma) to 1%

total cetavlon as described by Scott (72). The precipitate was

pelleted by centrifugation at 12,700 x g for 10 min and the

supernatant separated. Three ml of DH2O were then added to the

cetavlon fractionated precipitate (Fr I) and a 4M CaCl2 solution

was used to bring the solution to IM. This 4M CaCl^ solution was

also added to the cetavlon fractionated supernatant (Fr II) to

bring this solution to IM. Ninety-five percent ethanol was then

added to both Fr I and Fr II to 80% by volume and placed at -20°C

for 30 min. The precipitate from both fractions was centrifuged as

above and washed 2 times with 20 ml of 95% ethanol at -20°C for 30

min. The fractions were then resuspended in 10 ml of DH2O and

lyophilized. Both fractions were assayed for uronic acid and KDO

to quantitate LPS (60). Fr II was boiled for 5 min in 0.1% SDS and

placed on a Bio Gel P-300 gel filtration column equilibrated with

16

0.1 M ammonium acetate, pH 8.1, and 0.1% SDS. The fractions from

P-300 containing hexose were collected separately, dialyzed and

precipitated with 3 vol of acetone at -20°C for one hour. The

precipitates were collected by centrifugation as above and washed

once with 95% ethanol overnight at -20°C. Again the precipitates

were collected and resuspended in a small quantity of DHÔ and

lyophilized. All fractions were assayed for uronic acid and KDO as

before. A 20 mg dry weight sample of Fr I was placed on S-2B, the

fractions collected as above and tested for uronic acid and KDO.

Several of the KP2 EPS preparations were also partially purified in

a similar manner.

Preparation of Rabbit Anti-Type-Specific

Antiserum

Antisera were obtained against KPl-0, KPl-T, KP2-0 and KP2

2-70 by the procedure of Edmondson and Cooke (24). A fresh, early o

log phase (1 x 10 CFU) suspension of these organisms was killed in

10% formalin and inoculated separately into rabbits intravenously

once eêry three days for 13 days with increasing doses of the

organism, starting with 0.25 ml on the first day, 0.5 ml on the

fourth day, 1.0 ml on the seventh day, and 1.5 ml on the tenth and

thirteenth day. Rabbits were exsanguinated 5 days later by cardiac

puncture. Approximately 100 ml of rabbit serum was obtained from

each rabbit, and frozen at -70°C in 10 ml aliquots.

Type-specificity of the antiserum was tested in capillary

precipitin reactions against purified capsular material and in

double immunodiffusion assays (62).

17

Rocket Immunoelectrophoresis

A method similar to that described by Weeke (81) was used in

this procedure. Gels were prepared using 0.2% agarose (Sigma) in

0.075 M Gelman High Resolution Buffer (Tris-Barbital), pH 8.8. The

agarose was dissolved by heating to boiling and cooling to 50°C.

Anti-type 1 or anti-type 2 rabbit antiserum (2 ml) was mixed with

23.0 ml of agarose and poured onto a 4 x 6 in. sheet of Gel Bond

film (FMC Corp., Rockland, Maine). This volume gave a gel

thickness of approximately 2 mm.

Wells (4 mm) were punched in the agarose and 15 yl quantities

of EPS samples in DH2O were added. Standards included five 2-fold

serial dilutions (2 mg/ml-0.125 mg/ml in dry weight) of low

molecular-weight EPS from KPl-0. Samples were electrophoresed on a

Pharmacia FBG 3000 apparatus coupled to an Bio Rad Model 500/200

Power Supply at 5 V/cm in the same Gelman High Resolution Buffer

for 3 h. The gels were then washed 3 times in saline at 4°C and

once in DH2O at 4°C. The gels were then pressed lightly with

several layers of filter paper until dry. The gels were then

stained with Coomassie Blue and destained as described by Weeke

(81).

For the experiments on testing for antigenemia in infected

rats and mice, the sera of these animals were first electrodialyzed

as described in Materials and Methods, Section J. The height of

the rockets produced by the sera of these animals was then compared

to the standard curve and the levels of ECPS in the sera of these

animals was determined in yg/ml.

18

Opsonophagocytic Assay and Serum Sensitivity

The OPA was similar to the one described by Edwards et al.

(25). Bacteria were grown in 50 ml TSB to early log phase

(0DcrQ=0.2). Ten ml were removed and the organisms were pelleted

by centrifugation at 17,400 x g for 20 min on a Beckman J2-21

centrifuge using a JA-10 rotor. The pellet was resuspended in 10

ml sterile, cold PBS (FTA Hemagglutination Buffer, BBL), pH 7.2,

and centrifuged at 17,4000 x g for 20 min as above. The pellet was

then resuspended in 10 ml cold normal rabbit serum (Gibco

Laboratories, Grant Island, N.Y.) that was heat inactivated at 56 C

for 30 min (HIRS) and placed on ice. Plate counts were obtained in

duplicate on TSA at appropriate dilutions. The bacteria were

further diluted to achieve a ratio of bacteria to white blood cells

(WBC) of 3-4:1 in the opsonic reaction mixture.

The WBC suspension was prepared by drawing 10-12 ml of

peripheral venous blood from normal volunteers in a non-heparinized

plastic syringe. The blood was added to 50 ml Corning centrifuge

tubes (Corning Glass Works, Corning, N.Y.) containing 4.0 ml of 6%

Dextran in DH2O (Dextran MW 80,700, Sigma) and 3.0 ml citrate

solution (16 g citric acid and 59 g sodium citrate per liter;

sodium citrate from Fischer Scientific Co., Fairlawn, N.J.; citric

acid from Sargent-Welch Scientific Co., Skokie, IL.). This mixture

was incubated at 37°C for 45 min to sediment the erythrocytes. The

WBC-rich plasma supernatant fluid was removed and washed once in

minimal essential medium (MEM) (Gibco) plus 1% bovine serum albumin

(BSA) (Gibco). Lysis of erythrocytes was accomplished by adding 5

19

ml of 0.84% ammonium chloride for 20 min at room temperature. The

WBC were then washed twice in HIRS. The WBC were then resuspended

in 3 ml HIRS and placed on ice. A 1:10 dilution of the WBC

suspension was made in sterile, cold PBS for counting purposes

performed on a hemocytometer (American Optical Corp., Buffalo,

N.Y.). The cells were adjusted to yield 1 x 10^ WBC/ml with cold

HIRS. The OPA was performed in 1.5 ml polypropylene micro test

tubes (Bio-Rad). The reaction mixture contained a total volume of

0.4 ml, consisting of 0.1 ml WBC suspension, 0.1 ml bacterial

suspension, 0.1 ml of serum [as either HIRS, rabbit antiserum (AB)

or heat-inactivated rabbit serum (HIAB) against the homologous

strain being tested, or normal rabbit serum (NRS)], and 0.1 ml of

either PBS or a suspension of EPS from the same serotype in PBS.

Control tubes were included in each experiment that were lacking in

WBC, antibodies, complement, or EPS. For serum sensitivity assays,

0.1 ml of the bacterial suspension was inoculated into 0.9 ml

rabbit serum which contained no WBC. Tubes were incubated at 37°C

for 60 min on an Ames aliquot mixer (Miles Laboratories, Inc.,

Elkhart, IN). In the OPA samples (0.01 ml) were removed

post-incubation and added to 0.99 ml sterile, cold DH2O to lyse the

WBC. Additional dilutions were prepared in sterile, cold PBS and

0.1 ml of the appropriate dilutions were streaked on TSA. After

overnight incubation at 37°C, colonies were counted and the net

growth for both the OPA and serum sensitivity assays were

calculated as follows:

CFU 60 min Log 10

CFU 0 min

20

Assay for Virulence of K. pneumoniae

in a Mouse Model

A standard virulence assay was performed, adapted from the

procedure of Baltimore et al. (1). Bacteria were grown to an early

log phase (ODncn = 0-20) in TSB and prepared as previously

described (Section B). Ten-fold dilutions in sterile, cold PBS

were prepared and groups of 3, 4, or 5 mice (Swiss Webster, males,

20-25 g) (Laboratory Supply, Indianapolis, IN) injected intra

peritoneal ly (IP) with 1.0 ml of the appropriate dilution. Dead

mice were counted and removed from their cages at 24 h intervals

for 96 h. Virulence was expressed as the 50% lethal dose ( L D ^ Q ) ,

which was calculated by the method of Reed and Muench (68).

In the studies involving the effects of crude or partially

purified EPS on virulence, bacteria were grown and inoculated as

above, and various dilutions of EPS in PBS were injected IP

simultaneously in 0.1 ml volumes to all groups of mice. The amount

of EPS for all mice in each study was held constant while the

number of organisms differed ten-fold from group to group. A

control study was performed each time a virulence assay was set up.

Controls received the bacteria in the same manner as the

experimental and received 0.1 ml IP of sterile PBS simultaneously

in place of EPS. An EPS toxicity control was also performed by

injecting the EPS in 0.1 ml amounts IP, without administering

bacteria. The EPS controls employed EPS in concentrations from

twice to one-half that of the amount used in the virulence studies.

In these control studies, four mice were injected with each

two-fold dilution of the EPS solution.

21

Assay for the Production of Lobar Pneumonia

in a Rat Model

Male Sprague-Dawley rats (Laboratory Supply) weighing 200-250

g were used in these studies. The rats were housed in plastic

cages in groups of 4 and had access to commercial chow and water ad

libitum. Before inoculation, the rats were lightly anesthetized

with ether and the ventral cervical region was cleaned with a 95

per cent alcohol rub. A 1 cm medial longitudinal incision was made

in the animal to expose the trachea. The trachea was incised and

0.05 ml of a washed suspension of varying concentrations of

log-phase organisms in PBS was placed into the left diaphragmatic

lobe of the lung via a bead-tipped, curved inoculating needle. The

incisions healed rapidly without evidence of infection.

At various intervals after inoculation, groups of rats were

exsanguinated by cardiac puncture under ether, the thoracic cavity

opened and the lungs aseptically removed. The lungs were weighed

and then used for either bacterial quantitation or for histological

examination. Small samples of lung tissue were excised from

affected areas and fixed immediately in 10 per cent formalin.

Samples were cut at 4 ym and stained with hematoxylin and eosin

(H&E) on Brown and Haup stains. The sections were examined

microscopically and photomicrographs were made from representative

areas.

For bacterial quantitation, lungs were homogenized in 5 ml of

sterile PBS at 4°C using a Brinkman polytron homogenizer (Brinkman

Inst. Houston, TX). Serial ten-fold dilutions were made from the

homogenate and 0.1 ml of selected dilutions were plated out on TSA

22

and incubated overnight at 37°C. The following day, colonies were

counted and concentrations of J<. pneumoniae in lung specimens were

determined. Lung bacterial counts were calculated as the total

number of bacteria present in an entire lung specimen and were

reported as the total bacterial count (TBC) per set of lungs. For

convenience the TBC was expressed in logarithmic units to the base

ten (log^Q TBC).

To determine the fifty percent infectious dose (IDCQ) of

organisms used in these studies, a rat with a lung TBC value equal

to or exceeding 5'x 10 CFU (log^g TBC = 4.70) was considered

infected. Infected rats were counted at day 6 post-inoculation and

the loq-, IDrn was calculated by the method of Reed and Muench IU oU

(68). All rats succumbing to the infection before day 6 were

considered infected.

Chronicity studies were also performed with KPl ATCC 8047

before it was separated into its two subvariants, KPl-0 and KPl-T.

In these studies eight experimental (receiving 0.05 ml containing 5

x 10^ CFU of early log phase KPl transtracheal ly in PBS) and two

control rats (0.05 ml sterile PBS) were sacrificed on day 1, 3, 6,

and 9 post-inoculation. Lungs from 4 of the 8 experimental rats

and from one of the two control rats on each day of sacrifice were

processed for histology, while the remaining 4 experimental and one

control rat had their lungs removed for bacterial quantitation.

Four experimental and one control rat were also sacrificed on days

7, 14, 21 and 28 post-inoculation, having also received 5 x 10 CFU

23

of KPl, and all of these rats were processed for bacterial

quantitation.

Rat sera were obtained by allowing the blood to clot at 4°C

overnight, and the serum obtained by centrifugation at room

temperature at high speed in a clinical table-top centrifuge

(International Equipment Co., Needham Hts., MA). Lysozyme levels

in serum were measured by the lyso-plate method of Osserman and

Lawlor (61) using human urine lysozyme (Kallestad) or hen egg white

lysozyme (Difco) as the standard. Serum zinc determination was

performed by atomic absorption spectrophotometry on a Perkin-Elmer

model 2380 AAS using zinc chloride as the standard.

Assay for Characterization of Outer Membrane

Proteins of K. pneumoniae

The method for preparing the outer membranes for gel

electrophoresis was that of Diedrich et al. (18). One liter

cultures were grown in DW medium as already described. At 18, 24,

36, and 48 h intervals, 200 ml samples were taken and the cells

pelleted by centrifugation. The supernatant was decanted and the

cells were frozen at -70°C until further use. The frozen pellets

were broght up in 8-10 ml of Hepes buffer, pH 7.4 (Sigma) and

fractionated in a Franch Pressure Cell (Amico, Silver Spring, MD)

at 1.8-2.0 X 10^ pounds/square inch (PSI). The fractionated

material was then centrifuged at 3020 x g for 10 min in a JA-20

rotor (Beckman) to remove cell debris. The supernatant was then

placed in ultracentrifuge tubes (DuPont Instruments, Newtown, CN)

and centrifuged in an OTD-75 centrifuge (Sorvall, DuPont) at

24

50,000 X g for 45 min in a T-865 rotor (Sorvall). The supernatant

was decanted and the pellet resuspended in 10.8 ml of Hepes. A

volume of 0.12 ml of a 0.1 M MgCl2 solution was added, the tube

inverted several times and 1.0 ml 20% Triton X-100 in Hepes was

added and inverted several times again. This was allowed to stand

for 20 min at room temperature. The solution was again centrifuged

at 50,000 X g for 45 min, the supernatant decanted and the pellet

resuspended in DH2O to one-thousandth of the original culture

volume. The samples (10-30 yg protein) were then electrophoresed

at 30 V to 60 V in 12% polyacrylamide slab gels according to the

method of Pugsley et al. (67) and stained with Coomassie blue.

Electrodialysis of Extracellular

Polysaccharides from K. pneumoniae

This procedure was adapted from the method of Galanos et al.

(33). The electrodialysis apparatus was built from a 2 1/2 x 3 1/2

x 2 1/2 in. plastic tray with cover. Three chambers were assembled

inside the tray by two plastic slide frames held fast and made

leak-proof with silicone cement. Holes were drilled in the cover

over the two outside chambers and electrodes were fastened to the

cover. Dialysis tubing (Fisher Scientific Co., Pittsburgh, PA)

with a 10,000 molecular weight cut off was inserted within the two

slide frames. The chambers were filled with cold DH2O (resistivity

equal to 15-18 megohms) and the apparatus was placed on ice.

Initially, a 1000 V potential from an ISCO Model H92 Power Supply

(ISCO, Lincoln, Nebraska) was placed across the terminals of the

apparatus without sample to remove any contaminating ions. When

25

the ammeter approached 25 mA the power was shut off, the DH2O

decanted and refilled with cold DH2O. This was performed in the

absence of sample until no increase in mA was noted. The EPS

sample was placed within a benzoylated dialysis bag (Sigma) with a

molecular weight cutoff of either 1000 or 2000, and the bag was

placed in the center chamber of the electrodialysis apparatus.

Initially the voltmeter was set at 400 V and the ammeter was

allowed to increase up to 25 mA. At this time the power supply was

shut off and one half of the DH2O (25 ml of 50 ml) was collected

separately from both the cathode and the anode chambers. The

apparatus containing the sample was washed extensively in DHpO and

refilled with cold DH2O. The process was repeated at 400 V up to

25 mA and half the liquid in the anode and cathode chambers

collected and pooled with prior trials until the increase in

amperage at 400 V was negligible over a 30 min period. At this

time the dialysis bag containing the sample was opened and an

aliquot removed. The bag was then closed and placed back into the

electrodialysis apparatus.

The voltage was then increased to 1000 V and electrodialysis

continued. The liquid from the cathode and anode chambers was

pooled seperately from each other and separately from the preceding

runs at 400 V. Electrodialysis was complete when no further

increase in amperage was noted over a 30 min period. The sample

was removed and placed at 4°C. The cathode and anode pools were

lyophilized to dryness and brought back up in DH2O at approximately

one-tenth the original volume. From electrodialysis was obtained a

26

sequential set of electrodialysed samples, and the material from

both the cathode and the anode chambers from one electrodialysed

sample to the next. These solutions were then tested for hexose,

uronic acid, protein and LPS as described above, for phosphate,

according to the procedure of Chen et al. (13), and for calcium and

magnesium as determined on a Perkin-Elmer model 303 Atomic

Absorption Spectrophotemeter (Perkin-Elmer Corp., Norwalk., CN)

according to the Perkin-Elmer manual.

Determination of Capsule Size of K. pneumoniae

Capsule size of bacteria from DW medium was determined by the

method of Duguid (20) using India ink preparations. Capsule

production was expressed as the transverse diameter (TD), which is

a measurement of both the width of the bacillus and the width of

the capsule on either side of the bacillus. One hundred bacilli

were randomly selected under oil immersion, measured with an ocular

micrometer, and the average capsule size was calculated.


Polysaccharides Produced by K. pneumoniae

Bacteria were grown in a defined medium as described in

Section B of Materials and Methods. Samples (100 ml) of growing

cultures were taken at 18, 24, 36 and 48 h and the organisms

pelleted by centrifugation at 12,700 x g for 30 min. The

supernatant obtained was dialyzed 3 times in 8L cold DH2O overnight

while stirring, and assayed for uronic acid. Colony forming units

per ml culture at these time periods were also determined on TSA.

The production of ECPS for each organism was expressed as yg ECPS

27

ml cell (xlO ). ECPS was calculated by dividing the uronic

acid determinations by 0.3098 (for KPl) or by 0.2643 (for KP2),

which reflects the proportion of ECPS which is uronic acid (30,64).

The production of extracellular lipopolysaccharide (ELPS) for each

strain was also monitored at these time periods by the method of

Osborn et al. (60). The ELPS data were expressed as the yg ECPS

ml cell (x 10 ). ELPS data were further characterized in some

of the studies by the Limulus Amoebocyte Lysate (LAL) Assay

(Pyrotell Associates of Cape Cod, Inc., Woods Hole, Mass.) as

described by Levin (50). Both assays utilized the purified LPS

from Escherichia coli 055:B5 as the standard and the ELPS units are

expressed as yg of E.. coli LPS equivalents.

Electron Microscopy

Electromicroscopy for the visualization of the capsular

substances of J<. pneumoniae was performed according to the method

of Cassone and Garaci (12) on early log phase cultures grown in DW

medium. Preparations were observed and photographed with a Hitachi

H-600 Transmission Electron Microscope. These studies were kindly

performed by Dr. Jack Yee of the Department of Anatomy, Texas Tech

University Health Sciences Center, Lubbock, Texas.

Gel Diffusion Method for Immunological Analysis

The immunological characterization of the various fractions of

EPS from KPl and KP2 strains were performed by the method of

Ouchterlony (62). Ion agar (0.5 to 1.0% solutions) (Difco) in DHÔ

were brought to boiling to dissolve the agar and cooled to about

50°C. Twenty-five ml were then poured on 3 1 / 4 x 4 inch glass

28

plates and allowed to solidify. Holes of 2 mm in diameter and 0.5

cm apart in a circular fashion were made in the gel and 5 yl of EPS

at various concentrations were placed in the outside wells.

Another well was cut in the middle of the circular wells and 5 yl

of KPl or KP2 rabbit antiserum was added to this middle well. The

reaction took place at room temperature overnight in a humidifying

chamber. The gels were then washed three times in 200 ml

physiological saline at 4°C and once with 200 ml DH2O at 4°C. The

gels were then stained with 0.2% Coomassive blue in methanol,

acetic acid and DH2O (5:1:5, by volume), and then destained in the

same solution in the absence of stain.

Radial immunodiffusion studies were also performed to

•quantitate the ECPS found in the serum of infected rats, and to

follow the effect of electrodialysis on EPS samples. A 0.5%

agarose solution in DH2O was heated to boiling and allowed to cool

to 50°C. One to two ml of type-specific antiserum was then added

to 23 or 24 ml of the agarose slurry, and poured onto 3 1 / 4 x 4

inch glass plates or in 150 mm petri plates and allowed to

solidify. Hole were made in the gel (2 mm) and 5 to 10 yl of

sample was placed in the wells. The gel plates were incubated for

18-24 h at room temperature. Zone diameters of the precipitin

reaction within the gel were measured and compared to known

quantities of ECPS tested in the same manner.

Saponification of K. pneumoniae

Polysaccharides and Quantitation

of Fatty Acids

In some studies the EPS, that was partially purified up to the

29

by EPS after saponification. No attempt was made to identify the

various methyl ester fractions.

Hydrofluoric Acid Treatment

Various fractions of EPS were weighed out (20 mg) and placed

in 1 ml 60% hydrofluoric acid (HF) for 3 h at 0°C with occasional

mixing. NaOH (4M) was then added to pH 12 and the sample was

centrifuged at 12,700 x g for 10 min to remove debris. The

supernatanat fluid was collected and tested for serological

activity with the appropriate antiserum. The supernatant was then

placed on a gel filtration column to characterize the effect of HF

on the molecular weight fractions of EPS.

Statistical Analysis

All statistical analyses performed in these studies utilized

the student's t test for unpaired samples (73).

CHAPTER III

RESULTS

Strain Variation and the Production of

Apparent Isogenic Sets

India ink preparations of a number of strains of K,. pneumoniae

revealed that not all bacilli of the same strain possessed a

similar size capsule. Two predominant capsule sizes co-existed

within many of the strains. A closer inspection of isolated

colonies on TSA revealed that the two basic colony types within a

given strain corresponded to the capsule size differences seen

under India ink. In particular, for KPl ATCC 8047 and for KP2 ATCC

29011, there existed an opaque (0) and a translucent (T) colony

type. Thus the co-variants in the KPl population were labelled

KPl-0 and KPl-T, and those in KP2 were designated KP2-0 and KP2-T.

In both cases the opaque variant possessed the larger capsule.

KPl-0 possessed a capsule with an average transverse diameter (TD)

of 5.6 ym, while KPl-T exhibited a TD of 2.5 ym. The capsules of

KP2-0 and KP2-T had TD of 2.5 ym and 1.5 ym respectively. The data

for capsule sizes of all the strains used in this study are given

in Table 1. Biochemical and serological typing were performed on

all variants to confirm species and serotype. All strains of K,.

pneumoniae used in these studies had an API 20E (Analytab) code of

5215773 except for KPl 2-70 which coded out as 5005773.

It was much more difficult to obtain large and small capsule

variants for the KPl CDC 2-70. Under India ink KPl 2-70 had a TD

of 2.2 ym. A large encapsulated organism (TD=5.6 ym) was

occassionally seen under India ink, though these were rare. The

30

31

Table 1 . Capsule Size of K. pneumoniae

OiaMlim.^ Tp^(ym) Range(ym)^

KPl-0 5.6 3.9 - 8.6

KPl-T 2.5 2.2 - 3.0

KPl 2-70 2.2 1.8 - 2.5

KP2-0 2.5 1.9 - 3.0

KP2-T 1.5 1 . 4 - 1 . 6

KP2 2-70 3.0 2.5 - 5.0

Bacteria were grown in defined medium for 18h.

TD; Transverse diameter as measured under India ink, calculated as the average (mode) of 100 random determinations.

^The range in TD re f lec ts the smallest and largest TD seen in a s ingle preparat ion.

32

large capsuled variant was finally isolated by enriching the

population by passage through mice.

After several attempts, a sub-population of KP2 CDC 2-70

differing in capsule size could not be isolated. Under India ink

the average TD for KP2 2-70 was calculated to be 3.0 ym. However

the variance in capsule size in this strain ranged from 2.5 to 5.0

ym. A population rich in large encapsulated variants of KP2 2-70

was obtained by passage through mice. However, upon subculture,

the majority of large variants were no longer present in the

population. Furthermore there appeared to be an increase in the

presence of the larger encapsulated variants during the stationary

phase of growth, with an average TD of 2.75 ym at 24h growth, 3.0

ym at 36h growth and 3.3 ym at 48h. This is the only strain among

those utilized in these studies that had the propensity to change

its average capsular diameter at various stages of culture. Figure

la and lb show transmission electron micrographs (TEM) of KP2-0 and

KP2-T, respectively, as examples of the variance in the dimension

of capsule size within a given population.

The Establishment of a Chronic Lobar

Pneumonia by K. pneumoniae

in a Rat Model

Early studies employed KPl ATCC 8047 in chronicity studies, as

described in Materials and Methods. Results of lung bacterial

concentration are shown in Table 2. The total viable bacterial

count (TBC) is expressed as the log-iQ of the average total lung

bacterial concentration for the experimental animals in each group.

As can be seen, the log-j^TBC remained elevated (range of 4.11 to

33

34

36

37

Table 2. Establishment of Chronic KPl Pneumonia in Rats

Group Number of animals Day

dead/total Sacrificed

Log 10

TBC" (CFU)

(Range)

1

2

3

4

5

6

7

8

Control

0/4

0/4

0/4

0/4

0/4

0/4

1/4

2/4

0/8

1

3

6

9

7

14

21

28

3, 6, 7, 9,

14, 21, 28

6.51

7.02

6.32

6.91

9.47

6.20

3.84

3.19

ND^

(5.96-6.90)

(5.67-8.46)

(4.11-8.51)

(4.60-8.95)

(9.05-10.16)

(5.86-6.65)

(2.00-5.00)

(ND^-6.38)

^Rats were inoculated transtracheally with 5 x 10 CFU of KPl in 0.05 ml sterile PBS and sacrificed on the days indicated. Controls received 0.05 ml sterile PBS in the same manner. All surviving rats (except the four used for histological processing on days 1, 3, 6 and 9) were used for bacterial quantitation of lung tissue.

^TBC; viable bacterial count per whole lung expressed in log.|Q units from surviving rats.

^ND; none detected at 10' dilution of lung homogenate.

38

10.16) throughout the first fourteen days of the study, while no

KPl were detected in the lungs of the control animals. After day

14 it became difficult to obtain a statistically sound estimate of

the TBC in the rat lungs due to deaths occurring in the 21 and 28

day groups. Only one of the eight experimental rats in the 21 and

28 day groups cleared KPl from its lungs, while the four remaining

animals had log^^TBC of between 2.00 and 6.38. Mortality for the

entire population was 5 per cent (3/60), but for those rats that

were sacrificed on or after 21 days, 37.5% (3/8) of the animals

died before the time of sacrifice.

By day 2 post-infection virtually all experimental rats

appeared acutely ill. Mucous secretions exuded from their eyes and

most exhibited short and rapid breathing. As the infection

progressed, their coats became shabby and considerable weight loss

was obvious. Gross examination of the lungs showed involvement of

one or more lobes, often affecting the entire lobe in a typical

lobar distribution. The involvement was characteristically massive

and voluminous, presenting as dull, greyish regions that released

copious amounts of purulent exudate upon sectioning.

Histological examination also supported the establishment of a

lobar pneumonia in this rat lung model. Figure 2 (a-e) are

photomicrographs of H&E stained sections of rat lung tissue showing

the progressive development of a confluent pneumonia. Figure 2a

depicts normal lung tissue, representative of all control animals.

The integrity of the alveolar and bronchiolar structures can be

easily visualized. In marked contrast, lung tissue typical of 24h,

39

post-exposure rats (Fig. 2b) shows a phagocytic infiltrate

consisting primarily of polymorphonuclear leukocytes (PMN) filling

the alveolar spaces. By day 3 post-infection (Fig. 2c) a confluent

pneumonia had developed. The structural integrity of the

bronchiolar, columnar epithelium had been compromised, and signs of

necrosis and early abscess formation were evident. Large abscesses

and liquefication of structural walls were characteristic of

infection by day 6 (Fig. 2d). By day 9, foci of chronic abscess

formation were evident (Fig. 2e) with collagen fibers visibly

forming a wall to contain the abscess. This process of progressive

destruction of lung tissue continued up to day 28 when the study

was terminated.

The next set of experiments was performed to determine the 50

per cent infective dose (IDCQ) in the rat model for the KPl 8047

strain, before this strain was separated into its capsular o

variants. KPl was grown to a concentration of 1.55 x 10 CFU/ml in

TSB, washed twice and resuspended in cold PBS. Serial 10-fold

dilutions were then made in cold, sterile PBS and the organisms

were kept on ice until inoculation. Thirty rats were employed in

these studies and were divided into five groups of six animals

each. Group 1 received 0.05 ml of the undiluted KPl suspension

(7.76 X 10^) transtracheally into the left lower lobe of the lung.

Group 2 received the same volume of the first 10-fold dilution r 2

(7.76 X 10 CFU) and so on to group 5 which received 7.76 x 10 CFU

of KPl. All rats were sacrificed on the sixth day after KPl

administration. Table 3 summarizes the results obtained. Twenty

\

- -ijaja^ \

42

V9r

A ^

^

43

\

44

r \

"V^" . ^

45

' T' v ' " ^ '.r •^. ' % V- ^' •^•- ^

46

47

48

49

50

CO 4- i (T3

OC

M- .—^ - - ^ OJ E CD rs c: S- (O O) on

 ^ - -N E O ^>. CO CD > > =1.

_ l > — '

51

(0

O

O)

u : 3

-o o i-

o

O)

o OJ en ro CO O

Q

M-O

• a CJ CO

en o

OJ

c

CD

O

o V Q .

CO

O

i n

0 0 o

i n

I C>sJ

o •

o V Q .

CM

CO O

I CO

o V Q .

O

I

i n o o V

I— o

i n 00

I CO

o I

0 0

CQ

CO

OsJ

0 0

C\J 0 0 I—

cr> I I

o cvj

CM

cn

CO

o o

I

Q

0 0

0 0

l o

o o

I

cu

CO CO

CO I

Q

cr> i n

i n

• I - ( J

•—< C_)

o I—<D

CD CO

o o _ J Q

co CO >— CM O O

CTi O

0 0 CM CM

OJ Q

i n o

• c

4 -O

CO E O

: 3

O

^ CT>

c-^ cr CD ftJ

4 J 'CXL

cr> 0 0 CO

cr> 0 0 0 0

cr> 0 0

ô i n CO CM

oo CO CO

CD CD •r- C

(J OJ

co

J 3

-o - 

^ ^

00

CO

-

CO

r^

£5

-

CTi

^ i n

^

CO

»JD

^

i n

00

^ 1

i n

vo

i n

CM

1

a^

• o OJ

+-> ro

Z5 O

o

CM CM

I— ro J 3 Ci. ro 23

t— O

Ci3

CO O I — , — , —

vc "^ CM

CO

CM

VD

CM CO i n

o

o

o o

ro CU

u ro

+-) CO

c ro S -

+-> ro

CtL ro

•C3 C 4 -ro - i -

>>-a I— CU I— +-> ro O (J CU

•r- M-+J C CL- r -O) CO O) ro XJ

• o O

> o -o E S- CO «o OJ C (U 5 O i -

U CD 01 CD C_> I CU CQ

- > s : ^-co o -o s-CL E • ! -

ro 

ro x j s : <o QJ - r -U CU E

•r- x : o If- +J E •r- 3 S_ E CU (J • ! - E CO "O



CO CO CU o

I— - o +-> ro E CO - o

•r- ro CU 01

CD O > , E CU O - E

- E +J

>

rs

X i

(yO CQ D_

CU

cu • M CO

CO CU

CO

" O CU

x : CO ro

• a E fO

o II o i n i n

o

CU 0 1 ro

(J

i i -rO CD O

I X3

O +J

CQ 0 0

E

O S-cn CO ro

CU 4-> O E

+J O O

CU

ro

XJ CU

J 3 •r-

CJ CO CU

X J

CU 4-> ro E CU CD O E O

CD

O

E O

Q _ CO

U T3

X3

CM

O

4-J ro

X3 CU

4-> CJ CU

4-) CU

X5

CU E O E

Q

CU

CU

CO ro

CU E

E ro

CD E

• I —

>> O

CL E CU

o

ro

X3 E ro

ro E S -CU CO CO

o

CU S -1 3

XJ CU u o Q .

CU

r— 4->

XJ CU

E S -

o M -i . O) CL >^ ro lO oi ro CU XJ E S >> ro N XJ O E CO ro > ) - M

I CO

CO

O s-

- M E O CJ

o +J

XJ CU

ro CL E O

u ro +J ro

X )

CU

E N O CO

O

CO • I —

CO

> ro E ro

ro O

+-> CO

+ J ro 4-i CO

CD

52

per cent (6/30) of the rats died during the course of this

experiment with half of these belonging to group 1. Lung weight

was determined to document the marked increase in lung size in

infected rats. Serum lysozyme levels were also examined in this

study because these values have been shown to covary with the

extent of infection (4). As can be seen in Table 3, the serum

lysozyme levels of all groups of animals receiving KPl were

elevated with respect to control values, achieving significance at

the p < 0.01 level in two of the groups. Due to the marked

swelling during the infectious process, the weight of the lungs

increased up to more than three times that of normal. The average

lung weight for the infected rats in this study was 5.0 grams (3.1

grams above the control mean lung weight). Rats were considered

infected if they either succumbed to the KPl-induced pneumonia or

if a TBC of at least 5 x 10^ Oog^Q TBC=4.7) was found in the whole

lung of those rats harboring the organism. An ID^Q for KPl of 1.55

X 10 CFU was thus obtained.

In order to test the effect of the medium in which the

bacteria were grown as contributing to the virulence of the

organism, a defined medium was used and the effect of dosage of KPl

was repeated in the same manner as above. The ID^Q obtained using 5

the defined medium was found to be 2.22 x 10 CFU, which does not

differ significantly from the ID^Q value obtained using TSB.

Therefore, the comparative effect of growing l<. pneumoniae in two

different media on the pathogenicity of KPl in the rat lung model

appears to be negligible.

53

The remainder of the bacterial strains and their subvariants,

with the exception of the KP2-T and KPl CDC 2-70 strains were then

examined in the rat lung model. All organisms were grown in

defined medium and harvested as described in materials and methods.

Table 4 shows the results obtained when various concentrations of

KPl-0 were inoculated transtracheally into the lungs of normal

rats. With an initial inoculum of 5.0 x 10^ CFU all rats became

infected, three died, and the one remaining rat harbored a TBC of

1.43 X 10^ CFU at the time of sacrifice. All rats receiving 5.01 x

10^ CFU or 5.01 x 10^ CFU of KPl-0 also became infected. One rat

died in each of these two groups, while lung weight and serum

lysozyme were elevated. In groups 4 and 5, which received 5.01 x

10^ and 5.01 x 10^ CFU of KPl-0 respectively, three of the four

rats in each group were infected, one rat in each group died and

serum lysozyme as well as lung weight was elevated. Finally, a

dose of 5.01 X 10^ CFU of KPl-0 or less did not result in an

infection of any rats.

Table 5 shows the results obtained when various concentrations

of KPl-T were inoculated transtracehally into the lungs of healthy

rats. An initial inoculum of 7.07 x 10^ CFU of KPl-T resulted in

the death of three of the four rats in the first group. The one

remaining rat effectively cleared this massive inoculum of KPl-T

organisms placed in its lungs and showed no overt signs of

pathology. Only one of the four rats in the second group, which

received 7.07 x 10^ CFU of KPl-T, succumbed to the infection, while

the remaining three rats showed no signs of infection at the time

o

I

:^ CD E O

CO +J ro

ro

CU

CO

^ ^ CU E cn 13 E i- ro CU ex:

a> 0 0 ^ — E N E O ^ ^ CO e n >^ =1

X I C-J CQ

CD O

CD E ro

a:

a O r— b 13 CU E

Cl.

s-ro

X I o

—1

u •r-E O i -

x : CJ

ro

M-O

4-> E

ro LL. • - - CJ 4-> > — . f —

E E »—« rs

r— o r j <—o

CD o O E

_ j t—t

^ — V

CU en E

«4- ro O - ^ - Q : :

en—' +-> — JE CD CD

•r- E CU =3

1 —

X ) ro

4-> 01

UJ

XJ XJ CU

4-> (J CU

M-E

CO x : 4-> ro CU

Q

CU

X3

ro ro

o CD

WD

0 0 0 0

en

'-^ r^ -— ^-^ ,-^ ^-^ ,_^ •" : ^ ^ ^ f ^ LO CM O

^ O VO CO CO IJD CD o^ <— 00 00 vo S r^ A ' ' ' • I I CJ ^ l O , _ OQ Q^ Q2 £ > )JD i n C3^ r-^ c J C7 ^ r^ in in ko tjD Ln

l ^ CM «;d- r>^ O CM «X) • •

S S If ^ ^ ^ Ln 00 CT) r ^ r>>. ^ o « ^ vo

CM CM

I

C O

I— CO

CM ^ - s CM O ^ 0 0 CM

cr> . o I 0 0 >—

«x> I I ' d - CU CU

cr> z 2 -

»X) CM 0 0

0 0 cr> i n

CM

C O

CU Q

CU Q

o o o o o o o 1" r^ I— i ^ r^ r r^ CO i n ^ CO CM I — CD

CO CM ^ CO ^ ^

( ^ 0 0 «;r 0 0 CM CM'

J. ' ' ' ' ' cx) 1 ^ o CM o r> CO CO CM CM r-^ r-^

CM ^ f>0 C7^ cri I— I—

i >» i n r^ CO ^ CM CM

3 3 3 <— ^ o" o" ^ ^ ^ « ^ «;J- ^ «;j-

^ ^ <d- CO CO O O

<— CM CO i n CO

CU en

CU CO ro

x : CL

U

ro CD

o I

X I

o 4->

CO

XJ o

CU

X3 E ro CO

ro CU

l/J CQ Cl.

i n

CM 1

r^

>—

cr>

•""

o * • — '

•^ - ^ CT»

r— o S-+-> E O

o

. '—» ro

CU ->

o E

+J O o

M-* • — '

CM

CU r— X i ro

f—

E •n-

X I CU

X3 •r-S-o CO CU

XJ

CO <

ro

. "—^ X I

CU 4J o E

+J o o

M-^—' CM

CU f —

X2 ro

I—

E **" XJ CU

X I •r— %-u CO CU

XJ

CO <

X3

CU +-> ro

2 1

E • •p- c/)

CQ T3 Q_

crib

e

rile

01 CU CU +->

XJ u)

E E ZJ • ! -

• • "O CO CU CU E E

• r— XJ 4-> CU E CO

' T XJ CU CU

X I x : 01

E ro •r- S

E XJ S E o ro S-CT) ' -^

1 —

CO . ro o :s II

o o i n 1 LO

f— Q Q- o v : — -

CJ

. . . N

u

 o o

M-—'

•~~ CU —

X I ro

h -

E • 1 —

•C3 OJ

X ) •r— S-CJ CO CU

X3

CO <

XJ

ro E CU CD o E O

x :

lung

4 -o

E o

4 J :3

't—

X3

CM

O

4-> ro

X3 CU 4-i O CU

-l-> CU

X I

CU E o E

« W\

Q 2 ;

CU

• CM

CU

X3 ro

h -

E

XJ CU

XJ .,_ S-CJ CO CU

XJ

CO <c

4 -

54

4 -

"—^ E Z2 S-CU

CD CO

E ^ < — N E O --sv 01 CD

>^ n.

.—^ CU CD

c ro

Od

55

CU X3 O

cr E :3

_ j 4-) ro

CU x : +->

E

Q_

i n

 r»» C O 0 0

O 9 0 0 ^ r o c?^ • ^ t ^ un i n i n e n

C7 CO (v.. Ln CM CsJ C O

crj e n ^ un CT^ 0 ^ Ln vo CD CO CO i n i n CO

o C O CO

CU CD

CO

cri

I CU

Q

ro

4->

c E o

0 3 u . I — a c_j

CD O O E

O f— CU CM O

Q ^ .— C O

CU Q

i n C O

• CO

I CU

Q

0 0 CU Q

CU Q

CU

CU 4-) CO

i n o o E

>^

CU

X3 ro

o

CO E o

i n 0 0

e n CO

en CO

i n 0 0

i n 0 0

CO i n CO

i n 0 0 oo

CQ CM CL.

O M -

I O

ro

CU CO ro

. E CL

O

ro C D

o XJ •r— E O

4->

CO XJ

o . E +-> CU

X3 E ro CO

ro

s-OJ

4-J ro

CU +-> ro E CU en o

M-O

•M

en •r— CU

CU CD E

'-^ ro CDct:

CJ) E r3

r~s.

CM

0 0

-—

• CM

cr>

CM

o CM

CO

0 0

CM

0 0

^

^

CM

CM

1

' -

o

o CM

1 0 0

- -

cr>

en CM

1

—

en •— CM CM CM CM

XJ ^-^ CU 01

4-* J E U +J CU ro

4 - CU E Q

ro CL

o C5

CO o o o o o

•s*" "^t «:J- ^ ^ ^3- ^

< >— >— o o o o

o S-

4 J E

o C_)

I — CM CO i n CO

s-CU CO

x: 4-> • i ~

:5 >^

1 —

— ro CU

x: u ro s-

•i-> 01 E

ro s-+-> X J CU

4 J

ro 1 —

3 CJ

o E

• 1 —

CU S-CU

5 CO

+ J CyO ro CQ

CU Q_

ro

. * — X

X )

CU 4 J

o E

4-> O

o H-

>—' CM

CU p —

X J

ro 1— E

•r—

•X3 CU

X 3 • r ^

s-CJ CO CU

X 3

CO

< X 2

.,— X J CU

X J • r -S-CJ ' 0 1 tyO QJ C Q

X I ci -

E = rs ••"

• r -" O *^ CU ^

E £ ' p —

X I - - CU E CO

• 1 —

<4- X J CU CU

X J x : CO

E ro •-- 2 E XJ 2 E o ro s-C D ^ - - .

CM 01 . ro o 5 II

o 1— i n

1 i n f— Q CL O i ^

o

. — o CU

4-> O E

4-> O O

^-V ^

r—

CU r— X J

ro h-E

•r—

X J CU

J D •r— i -

o CO CU

X I

CO

< X J

o j e

CD E =3

C ( -

o E O

4 J 13

,__ .r— T 3

CM 1 O f~

4-> ro

X J CU

+ j CJ CU

+ j CU

X J

CU E o E

• €\

Q

CU

^ M -

CU 4-> o E

4-> o o

i f -

CM

CU

X J ro

1— E

'r—

X J OJ

X I * r -S-CJ CO CU

-a CO

<c M-

56

of sacrifice. One of four rats in group 3, which received 7.07 x

5 A 10 CFU of KPl-T, showed signs of infection (greater than 5 x 10

CFU per whole lung). Finally, all rats receiving 7.07 x 10^ CFU or

less of KPl-T (groups 4-6) did not become infected.

Table 6 shows the results obtained when various concentrations

of KP2-0 were inoculated transtracheally into the lungs of normal

rats. Doses of 7.2 x 10^ CFU did not result in infection in three

of the four animals in the first group. The fourth rat had a TBC 4

of 5.25 X 10 CFU with a lung weight of 3.2 grams and was

considered infected. No rats in group 2, which received 7.2 x 10

CFU of KP2-0 were considered infected using our criteria. One of 5

the four rats in group 3, which received 7.2 x 10 CFU of KP2-0 was

infected, but all other rats in this group and in the ensuing

groups (groups 4-6) had cleared K_. pneumoniae from their lungs.

Serum lysozyme was not significantly elevated in any group within

either the KPl-T or the KP2-0 study.

Table 7 shows the effect of various doses of KP2 2-70 4

inoculated into the lungs of rats. Doses of 6.5 x 10 CFU per rat

resulted in the infection of 6 of 8 animals at 7 days 3

post-inoculation. Group 2 received 6.5 x 10 CFU and showed 3 of 8

rats infected. Three animals from group 3 died and 1 of the

remaining 5 were infected. Group 4 rats which received 6.5 x 10

CFU revealed 5 of 8 animals having a TBC above the threshold for

infection. Finally group 5 rats, which received 6.5 x 10 CFU per

rat, showed 3 of the rats infected by day 7 post-inoculation. Both

lung weight and serum lysozyme were elevated in these rats, but

CU

57

4 -

'—^ E 3 S-CU

CU 0 1

E > ^ ^ Nl E O - v . CO CD > ^ 3 .

. — N

CU en E

ro cc

•

-69

•

CO

• CO CO

i n 1

CO

• i n

cr> en i n

0 0

CO

i n

CO

i n

r-..

CO

CM

•78.

r>~.

i n

CO

CO CO

"^

68.

1 CO

CO

i n

64.

CTl

69.

1

CM CO

r^

66.

o

70.

1 0 0

(59.

CO

65.

CU X J

o

cr E 13

ro cc: CU

XI 4 J

E

O i

CM CL.

C O

CU

X J

ro

X J C J CQ

CD O

CU CD E ro

CM

I

o CM

CO

ro

• i - C J

c : 3 cr o O E

CO

CO CO

CU CD

4 - E o •— ro

CDcc: 4-> "—-

x: O J CD

• r - E CU Z3

0 0 CO

CD CO

0 0 CO

0 0 0 0

CO CO

I CU

Q

cn i

CU Q

C O

CU Q

C O s ^

i I CU CU

Q Q

'— CM r>« cri o 1—

CO LO

CO CO 1 ^ I—

CO CO .— CM

CU

CO 0 0

CO 0 0

CO 0 0

CO 0 0

CO CO CO

CO 0 0 m

CQ CM D_

-o CU

-(-> u CU

^-E

CO - E

+-> ro CU

Q

CM

CO

CM

CD

CO

CM

r

o

f—

CO

«:J-

CD

cr>

CM

0 0

-

0 0

r—

1

en

CO

r -

CM

1

00

i n

CM i

cri

CM CM

ro CL

o i-

13

o o o o o o o

^ ^ ^ ^ ^ ^ *;

<— o >— o o o o '

o S-

4-) i n CO E

o C_3

I — CM CO

CU 4-> CO

in o

o I

CM ex.

CU

X J

ro

4-o CO E o

X J

o 4 -

I o

ro

CU CO

x: 4 J •r—

ro CU

. E CJ

ro 4-> CO E

ro 4 J

X 3 CU

+-> ro

o o E

CU i -cu

CO 4 J OO ro CQ

ro x : 4-> CD

ro

X i

CU

o E

4-) O O

CM

CU

ro

X 3 CU

X ) • ^ •

J -

o 00 CU

X J

01 <

X 3

CU CO

ro x: C2-

« — CU

X J

o E CO

• 1 —

x: + j

CD E

•r— CO r j

E CU CU CO

CO

ro ^

s-ro en o

'— X I • p ~

E O

4->

CO X J

o x: 4-> CU

s: • o E

ro CO

r—

ro • r -S-

cu • M

ro

• I - E

X J •

cuc>o X J CQ • I - Q_

s-CJ CU CO ' — O J - r -

CU

01 - M ro 0

E ^

CU OJ

E _E X J -M CU E CO

4 ^ X I CU CU

X I x : CO

E ro

E X I 2 E

o ro C D ^ - ^

O J 01

ro

O I

CM a . :^

u

o II o in en

Q

o

CU 4 J

o E

4-) O

o 4 -

C M

CU

ro

X J CU

X J

S-CJ 00 CU

X I

CO

< X J

CU 4->

ro E CU CD

o

en E 13

<4-

o E o

'r— 4-> 13

CM i O

4 J

ro X J CU

4-> o CU

+-> CU

X 3

CU E o

CU 4 J

o E

4-> O O

«+-

CM

CU

X J

ro

C 2

CU

X J CU

J D • t —

S -

CO CU

X J

00

58

CU X3 O

ro

Q i

CU

x : E

O I

CM

CM Q_

CU

 .

_ J

E 3 S-CU

CO

CU CD

XJ E CJ ro CQ or

CTi

o

CD QJ

E ro

cc CD-

2

-— 3 CJ ro «— -—^

O r - 3 ZD . — M CJ U -

C D - r - O CJ O E E — '

01 XJ

CU 4-> x : u -M en

CO i n

o 0 0

en cr>

• CTi

CTi O ^

CO

^— i n CO

CD

^-% " ^ r^

• CO en

1 CO

^

CO CO

CO i n

CO

r^

i n CO

• cr> o CM

CO

^—' ^ CO

en

^—.. i n

'â-»

o cr> 1 ^

(— •

CO r>»

i n CO

,— 00

o i n

CO

1

o^ Csj

' r^ o

^

„-^ CO CO

• r>.

1 CTi 0 0

i n LO

CO en

CM 1 ^

CO CO

• i n

t

CM

•?J-

^ O

i n

,_^ 0 0 0 0

t

r^

1 r^ 1 —

• CM CO

^ CNJ

CT\ CO

CM CM

• i n

1 1 ^ 0 0

CO ' — ^_ i n

« v ^

^_ E

en o

• o

E

o r^

1 CM

CM

CU

X J

ro >

M -O

00 E O

•r—

o 1 ^

• CO

^d-0 0

CO

cr. •

^

i n 0 0

CTi CO

• CM

0 0 cri

cri r>.

• CM

1 O cn

^ CO

• CM

1

0 0

CO

CO

CO

CM

CO

CM

CO

CM

O CM

CM

O 4 -

I O

ro • r -

cu 01

>^

CO 00 5 CO CO ^ fa

^ CO CM I— O QJ x : CJ ro s-+-> c/)

o o o o o ro — ' — ' — - — - — - S-

4-> CO CO CO 0 0 0 0

CO CO ^ i n CO CU 4 J ro

3 O

0 0 0 0 c n CO 0 0 o E •

• - - LT)

CU CL s-CU CU 2 <— 01 S--(-> QJ

f— c\j CO ^ i n ro 4-> Q: 01

ro

00 CQ Q_

CU

CU 4-> 01

01 CU

CO

X I CU x : CO ro

•o E ro

CM

• CM

CU r"" X J

ro t— E

•^-

• a CU

X ) ' p ™

s-u 01 CU

•o CO

< X J

o i n

Q o —

CU CD

ro x : CL

CJ •r—

E . E - l - >

•^-S-ro en o

•"" X J • n * E

O

+-' CQ Ul

I— E

" • -

O 4 J

E 2 O S-CD

01

ro 2

O r>H

1 CM

OvJ Q_ : ^

• X I

CU -> o E

4 J O o

^ -

•—

CU p —

X J

ro 1— E

• r -

X 3 CU

X J •n" s-u 01 O E

4 J O O

M -

CM

CU r ^ X J ro I—

E • r -

X J (U

X J ' r~ S-CJ CO CU

X J

01 <

<u

59

showed a gradual decline to near control values with increasing

10-fold dilutions of inoculum. Seven rats inoculated with 6.46 x 3

10 CFU KP2 2-70 (as in group 2) were sacrificed at 14 days

post-inoculation. Of these rats one died on day 3 and one had a

9 A TBC of 4.7 X 10 . Three rats had a TBC of above 1 x 10 but not

above the threshold of 5 x 10 . The remaining two rats had TBC of

3 3

1.6 X 10 and 3.6 x 10 . Therefore none of the rats had cleared

KP2 2-70 from their lungs though only 2 of 7 rats exhibited all the

signs of infection at day 14 post-inoculation.

The IDcQ in rats for each of the five J<. pneumoniae strains

employed in these studies can be seen in Table 8 along with the

LDcQ values for each strain in mice. The data are analysed in the

following section.

K. pneumoniae Virulence in a Mouse Model

All J<. pneumoniae serotype 1 and serotype 2 strains and their

variants were employed in standard virulence assays as described in

Materials and Methods. Result of these studies are compiled in

Table 8 together with the data obtained from studies of virulence

in the rat lung model. These data show that KPl-0, which exhibited

the largest capsule, was more virulent than its covariant, KPl-T,

by 4 or more log-jQ units. Similarly KP2-0 exhibited a larger

capsule than its covariant, KP2-T, and proved to be more virulent

in the mouse model. The TD of KPl-T was slightly larger than the

TD of KPl 2-70 and was more virulent than KPl 2-70 by more than 2

log-ip, units. Therefore, a direct correlation between two distinct

60

Table 8. LDHQ Values in Mice and ID^Q Values in Rats for Strains of l<. pneumoniae Serotypes 1 and 2.

Organism LD^Q (CFU)^ ID^Q (CFU)*^

KPl (mixed) 1.92 x 10 ^ 1.55 x 10^

KPl-0 4.99 X 10^^ 3.41 x 10^

KPl-T 6.03 X 10^^ 1.53 x 10^

KP2 2-70 1.00 X 10° 4.70 x 10^

KP2 (mixed) 4.29 x 10^ NP'

KP2-0 1.78 X 10^ >7.3 x 10^

KP2-T >6.2 X 10^ NP^

^Five groups of f i ve mice each were inoculated IP with ser ia l 10-fold d i l u t i ons of the appropriate K,. pneumoniae s t ra in in 1.0 ml of s t e r i l e PBS and observed for a 72 h period. LD^Q values were calculated by the method of Reed and Muench (65) ana represent at least two determinations for each organism.

^Rats were considered to be infected i f they succumbed to the i r pneumonia or i f the TBC was 5 x 10 CFU or greater.

^The IDrr. values for KPl-0 and KPl-T were shown to be s ign i f i can t l y d i f f e ren t (p<0.025).

°NP; not performed.

61

populations within a serotype regarding the relationship between

virulence and capsule size appears to exist.


Polysaccharides Produced by

K. pneumoniae

Bacteria were grown in a defined medium (DW) and harvested as

described in Materials and Methods. The ECPS production by KPl-0,

KPl-T and KPl 2-70 are summarized in Table 9. With regard to the

apparent isogenic covariants, the results indicate that KPl-0

produces ECPS to a much greater extent than KPl-T at all stages of

culture. -The expression of ECPS by both strains is linear with

respect to time over the entire period of analysis. The KPl-0

organism produced 0.37 yg ECPS ml" cell" (x 10 ) per hour of

culture, while the KPl-T strain produced 0.067 yg ECPS ml" cell"

(x 10^) per hour in this same period between 18 and 48h of growth

in DW. The overall ratio of the rate of production of ECPS between

KPl-0 and KPl-T is estimated to be 5.52:1. These data are

illustrated in Fig 3. The difference between the means of the

production of ECPS by these two bacteria is significant at p <

0.005. The production of ECPS by KPl 2-70 was intermediate between

that of KPl-0 and KPl-T and a rate of ECPS production of 0.15 yg

ECPS* ml"^ cell"^ x 10'^ per hour was characteristic of this

organism. Figure 4 shows a TEM of KPl 2-70, depicting both the

cell-associated and the extracellular capsular material.

62

Table 9. ECPS Production by Strains of 1<. pneumoniae Serotype 1 at Various Intervals of Incubation

Organisms Incubation Period (h) ECPS^ CFU/ml

5.50+3.68 1.17x10^

8.49+3.71 1.17x10^

10.75+3.70 5.64x10^

18.24+1.32 2.00x10^

0.67+0.03 3.7x10^

1.09+0.11 3.7x10^

1.63+0.03 2.67x10^

2.56+0.09 1.82x10'' p<0.001^

0.52+0.24 3.8x10^

2.12+1.55 4.0x10^

9.49+p.OO 3.0x10^

10.01+.2.48 1.1x10^ p<0.01^

p<0.05^

^Cultures were grown at 37°C in defined medium at 200 rpm.

ÊCPS; in yg ml"^ c e l l " ^ (xlO^) as quantitated by the method of Blumencrantz and Asboe-Hansen (7) .

^S ta t i s t i ca l analysis comparing ECPS ml"^ ce l l " ^ production of KPl-0 to KPl-T or to KPl 2-70 at 48 h incubation.

^Comparison of KPl-T to KPl 2-70 as in footnote C.

KPl-0

KPl-0

KPl-0

KPl-0

KPl-T

KPl-T

KPl-T

KPl-T

KPl 2-

KPl 2-

KPl 2-

KPl 2-

•70

•70

•70

•70

18

24

36

48

18

24

36

48

18

24

36

48

63

64

UJ

(Q.OI^)J|a3J^-Scd03 E'h'

65

66

67

Table 10 summarizes the ECPS production of the various KP2

serotypes. As can be seen, KP2-0 produces more ECPS than its

co-variant, KP2-T, which produces ECPS at near non-detectable

levels in the supernatant fluid at all intervals of incubation. A

comparison of ECPS production between KP2-0 and KP2 2-70 shows a

difference in kinetics, as illustrated in Figure 5. KP2-0 begins

to produce ECPS much earlier in culture than KP2 2-70, but by 36h,

KP2 2-70 has surpassed KP2-0 in total ECPS production and has

produced twice as much ECPS by 48 h. These differences are

reflected in the rate of ECPS production within this time period.

At up to 18h of culture, the rate of production for KP2 2-70 is

essentially negligible, but it increased significantly at

approximately this time period. Production is linear between 18

and 36 h for both strains, but the rates of production differ

markedly within this time frame, with KP2 2-70 producing 0.078 yg

ECPS ml"^ cell"^ (xlO"^) per hour and KP2-0 producing 0.027 yg ECPS

ml" cell" (x 10" ) per hour. Therefore, in this particular

medium, KP2 2-70 produces nearly three times as much of ECPS as

does KP2-0 within the linear phase of production.

Extracellular LPS (ELPS) production was also examined during

these time periods and the data are shown in Table 11. This table

shows the amounts of ELPS produced by all strains at 48 h of growth

in DW as measured by both the KDO and the LAL assays. Colony

forming units at the highest concentration during cultural growth

are also listed in Table 11. The ELPS is expressed as the yg ELPS

ml' cell' (x 10" ). It can be seen from this table that the

68

Table 10. ECPS Production by Strains of J<. pneumoniae Serotype 2 at Various Intervals of Incubation

Organisms Incubation Period (h) ECPS CFU/ml

KP2-0

KP2-0

KP2-0

KP2-0

18

24

36

48

0.48+0.20

0.59+0.11

0.90+0.001

1.30+0.13

9.22x10'

1.1x10^

9.2x10^

8.6x10^

8

KP2-T

KP2-T

KP2-T

KP2-T

18

24

36

48

0.002'

ND'

1.1x10-

0.003+0.002 9.9x10 8

1.1x10"

0.005+0.0002 1.0x10-

KP2 2-70

KP2 2-70

KP2 2-70

KP2 2-70

18

24

36

48

0.06+0.03

1.66+0.16

2.13+0.11

4.0x10^ p<0.20^

8 0.38+0.02 4.2x10 p<0.10

3.9x10^ p<0.025

3.9x10^ p<0.020

^Cul tures were grown a t 37°C i n def ined medium at 200 rpm.

ÊCPS; i n yg ml""" c e l l " ^ ( x l O ' ^ ) .

^ND; none de tec ted .

^ S t a t i s t i c a l ana lys is comparing ECPS ml"^ c e l l " product ion of KP2-0 and KP2 2-70 a t the same stage of incuba t ion .

Ônly one determinat ion made.

69

70

24 TIME(h)

71

_ Dtypes 1 and ; at 48h of Culture in Defined Medium

Table 11. Production of ELPS^ by J<. pneumoniae Serotypes 1 and 2

Organism

KPl-0

KPl-T

KPl 2-70

KP2-0

KP2-T

KP2 2-70

CFU/ml

1.17x10^

3.70x10^

4.00x10^

9.48x10^

1.05x10^

4.25x10^

(xlO"^)(KDO)^

1.43+0.04

0.38+0.14

0.54+0.03

0.03+p.OOl

0.01+0.002

0.07+0.007

yg ELPS mr^cell"^ yg ELPS ml'^ceir^

(xlO"^)(LAL)^

1.37 - 2.74

0.22 - 0.43

0.21 - 0.42

0.008-0.017

0.004-0.008

0.04 - 0.08

Êxt racel lu lar l ipopolysaccharide (N=2).

KDO; Thiobarbi tur ic acid assay for ketodeoxyoctanate (57).

^LAL; Limulus Amoebocyte Lysate assay. Values are presented as the range in which the quant i ty of ELPS in the samples is l im i ted (47).

72

KPl-0 organism produces the greatest quantity of ELPS per ml per

cell of all the strains, whereas the KP2 2-70 organism is the

highest ELPS producer among the KP2 strains.

Comparisons of ECPS, ELPS and capsule size at 48h growth for

all serotype 1 and 2 strains as well as virulence in both the rat

lung model and the mouse model can be seen in Table 12. First of

all, there exists a strong postive correlation (r=0.97) between the

production of ECPS and ELPS, per ml per cell. The two serotypes

differ in this regard in that the KPl strains produced

approximately 59 yg of ELPS for each mg of ECPS produced, while the

KP2 strains produced about 30 yg of ELPS per mg of ECPS. Therefore

the correlation between the production of ECPS and ELPS is

strongest within serotypes.

The relationship between ECPS and the capsule size of all

strains shows a direct positive correlation (r = 0.95), which is

even of greater magnitude when comparing strains within serotype

for both the KPl and KP2 strains. Therefore, in general, all of

these organisms seem to reflect their ability to produce

extracellular capsular polysaccharide by the size of their

capsules. A notable exception is the KPl 2-70 strain which seems

to have a TD less than that of KPl-T but produces more ECPS per

cell than KPl-T.

Finally in the comparison between the three parameters of

polysaccharide production (TD, ECPS, and ELPS per cell at 48h

incubation) and the virulence studies in the standard mouse

virulence model, an inverse correlation was noted. Table 13

73

Table 12. Comparison of ECPS, ELPS, Capsule Size and Virulence of J<. pneumoniae Serotypes 1 and 2

Organism

KPl-0

KPl-T

KPl 2-70

KP2-0

KP2-T

KP2 2-70

ECPS

18.24

2.56

3.73

1.30

0.005

2.13

ELPS^

1.43

0.38

0.54

0.03

0.01

0.07

TD^

5.6

2.5

2.2

2.5

1.5

3.0

LD ^

4.9x10^

5.34x10^

7.30x10^

1.78x10^

>6.2xl0^

1.0x10°

^^50'

3.41x10^

1.53X10'

NP^

>7.3xlO'^

NP

4.7x10^

a - 1 - 6 ECPS yg ml" cell"(xl0 ) in dialyzed supernatants of 48h growth at 37 C in defined medium.

h - 1 - 1 fi

ELPS yg ml" cell" (xlO ) in dialyzed supernatants of 48h growth at 37 C in defined medium.

^transverse diameter.

LDrni 50% lethal dose, obtained from IP injections in mice. bU

ÎDj-^; 50% infectious dose, obtained from transtracheal inoculations into the lungs of rats.

Not performed.

74

Table 13. Correlations Between Polysaccharide Production and Virulence in the Mouse Model

ECPS/cell^ ELPS/cell^ TD°

Serotype 1 ( L D ^ Q ) ^ -0.46 -0.37 -0.50

Serotype 2 ( L D ^ Q ) ^ -0.82 -0.72 -0.96

All strains (I-D^Q)^ -0.31 -0.15 -0.55

^Quantity of ECPS per cell in dialyzed supernatants at 48h growth in defined medium at 37 C.

Quantity of ELPS per cell as in footnote a.

^Transverse diameter of the capsule.

^The LDnn values obtained from 3 serotype 1 strains. bU

^The LDc^ values obtained from 3 serotype 2 strains. bU

^The LDj-f. values obtained from all 6 strains of type 1 and type 2 K_. pneumonVae.

75

represents the correlations obtained between the polysaccharide

parameters and the virulence data. The data from Table 13 show

that capsule size may be the best indicator of the ability of a

particular strain to be pathogenic, especially when the serotype is

unknown. However, the production of ECPS per cell correlates

nearly as well as does capsule size with these virulence

parameters. Finally the production of ELPS per cell gives the

least amount of information as to the virulence potential of a

given strain, though it still correlates with virulence. Again,

stronger correlations are found within serotype than in grouping

the two serotypes together, which is especially true for the KP2

strains where nearly perfect correlations were obtained.

Serum Sensitivities and Opsonophagocytic

Assays for K. pneumoniae

Of the six strains used in these studies, only one of the

strains (KP2-T) showed inhibition of growth in the presence of 90%

rabbit serum over a 60 min period. Data for all strains can be

seen in Table 14, which shows the change in the log-jQ CFU between

time 0 and 60 min of incubation in serum. The virulent KPl-0 and

KP2 2-70 strains seem to grow most favorably in 90% serum, but

their growth is not significantly different than those of the other

strains, with the exception of KP2-T.

The ability of these organisms to grow in the presence of

human leukocytes (WBC), type-specific antiserum (AB), or a

complement source (C) was then tested, as described in materials

76

Table 14. Serum Sens i t i v i t y of K_. pneumoniae

Strain^ ^Ô^^Q C F U ^

KPl-0 + 0.97

KPl-T + 0.54

KPl 2-70 + 0.50

KP2-0 + 0.42

KP2-T - 0.08

KP2 2-70 + 0.64

^Log phase organisms were washed 3 times and resuspended in PBS in various concentrations and added to 9 parts normal rabbi t serum.

^The change in the number of colony-forming units (CFU) in log,Q uni ts a f te r 60 min incubation in 90% normal rabbi t serum at 3/ C.

77

and methods under Opsonophagocytic Assays (OPA). Table 15

summarizes the results after normalizing the data for comparative

purposes. Normalization of the data was as follows: 1) Colony

counts in log-jQ CFU that were obtained for each strain at 0 min

were subtracted from the log-jQ CFU at 60 min incubation in the OPA.

The net change in the log^^ CFU (Alog,Q CFU) was thus obtained. 2)

Secondly the Alog.Q CFU for the control assay, containing only

heat-inactivated serum (without AB, WBC or C), was subtracted from

the Alog-jQ CFU of all other assays within a given strain. The

final value then shows the Alog-.^ CFU from 0 to 60 min compared to

the serum controls and allows for comparisons among the strains

tested. The results in Table 15 show that the most important

variables which affect the Alog,Q CFU is the presence of WBC and

type-specific AB in the assay mixture. The effect of a complement

source did not seem to affect the net log-jQ CFU in these assays.

Also it seems that WBC do not significantly alter the net

log-,p. CFU of any of the strains in question in the absence of AB.

One notable exception to this conclusion involves the KPl-T strain

and its ability to be more readily phagocytosed in the presence of

complement (C-E) than in the absence of complement (D-E), while in

the absence of AB. But also for the KPl-T strain, type specific

antiserum enhances the ability of WBC to ingest KPl-T moreso than

does complement, though this was not a significant difference.

Table 16 shows the results obtained when KPl 2-70 or KP2-0 was

tested in the OPA in the presence of EPS from either KPl or KP2,

containing both ECPS and ELPS. When the EPS from a type 1 strain

78

Table 15. Opsonophagocytic Assay^

(AB+WBC+C) (ABC+WBC) (WBC+C) (WBC)

Strain Control Control^ Control^ Control^

KPl-0 -2.08^ -1.49 -0.10 -0.12

KPl-T -0.27 -0.27 -0.16 +0.16

KPl 2-70 -0.39 -0.42 -0.02 -0.03

KP2-0 -0.19 -0.30 -0.09 -0.13

KP2-T -0.40 -0.22 -0.01 -0.07

KP2 2-70 -0.81 -0.79 +0.37 +0.05

Log-phase organisms were washed in PBS and various concentrations were placed in 3 parts normal rabbit serum containing combinations of the following components: 1) type specific antiserum (AB)-and 2) Human peripheral white blood cells (WBC). The serum was heat-inactivated in some of the trials to test for the effect of a complement (C) source. The assay was performed at 37 C for 60 min and the net CFU in log-iQ units was determined. See appendix 4 through 9.

The net log-,p. CFU obtained from the assay in which AB, WBC and C were present, minus the net log-jQ CFU obtained from the control assay where none of these components were present (heat-inactivated normal rat serum only).

^The net log-.^ CFU obtained from the assay in which AB and WBC but no C source were present, minus the net log-jQ CFU from the control assay.

^The net log-jp, CFU obtained from the assay in which WBC and C but no AB were firesent, minus the net log-jQ CFU obtained from the control assay.

^The net log-.^ CFU obtained from the assay in which WBC but no AB or C were fDresent, minus the net log^Q CFU obtained from the control assay.

79

Table 16. Effect of the Addition of EPS on the OPA^

Strain EPS Alog^^ CFU^

KPl 2-70 PBS 0.56

KPl^ 1.19 p<0.0l3

KP2^ 0.67

KP2-0 PBS 0.41

KPl^ 0.35

KP2^ 0.68 p<0.005^

The OPA contained the following components in equal volumes: Human peripheral white blood cells at 1 x 10 gper ml in normal rabbit serum; J<. pneumoniae strains at 1 x 10 to 1 x 10 CFU/ml in normal rabbit serum; Rabbit antiserum against the homologous serotype, and either KPl or KP2 EPS in PBS or PBS alone. A total serum concentration of 75% was used.

Change in CFU from 0 to 60 min incubation at 37°C in log units.

^KPl EPS from dialyzed supernatants of KPl-0 at 339 yg ECPS/ml and 20 yg ELPS/ml.

^KP2 EPS from the neutral fraction of KP2-0 at 850 yg ECPS/ml and 25 yg ELPS/ml.

^KPl EPS from the LMW fraction of KPl-0 at 331 yg ECPS/ml and 23 yg ELPS/ml.

^KP2 EPS from the ethanol extracted fraction of KP2-0 at 489 yg ECPS/ml and 16 yg ELPS/ml.

^Statistical analysis comparing the Alog.« CFU after treatment with homologous EPS compared to treatment witn either PBS or heterologous EPS.

80

was added to the KPl 2-70 OPA mixture at 339 yg ECPS/ml and 20 yg

ELPS/ml there was a significant difference in the Alog,Q CFU from

the same mixture without ECPS present. However, when KP2-0 EPS,

containing 850 yg ECPS and 25 yg ELPS per ml was added to the same

OPA, no significant difference was seen compared to the non-ECPS

control. In the reverse experiment the KP2-0 strain with the

addition of KP2 EPS at 489 yg ECPS/ml and 16 y ELPS/ml grew

significantly better than both the antiserum control (PBS treated)

or the KPl ECPS treated trials (331 yg ECPS and 23 yg ELPS per ml).

Therefore EPS from a type-specific strain allowed for enhanced

growth of KPl 2-70 over controls in the presence AB, whereas EPS

from a heterologous serotype did not.

Purification of the EPS of K. pneumoniae

Organisms were grown in the defined medium at 37 C for 48h

while shaking at 200 rpm for these studies. The purification

protocol, as described in Materials and Methods, involved ethanol

fractionation, DEAE-Sephacel ion exchange and gel filtration

chromatography. Fractionation with ethanol produced a hygroscopic,

white and fluffy material which was not easily resuspended in

hydrophilic solutions and was precipi table in non-polar solvents

(i.e., methanol, ethanol, chloroform, hexanes). When suspended in

DHpO these materials produced highly viscous solutions, especially

at a concentration of 2 mg dry weight or more per ml. These

solutions were characteristically opalescent and, for certain

preparations, a white precipitate was noted. Removal of the

precipitate by centrifugation caused as much as a 25% loss of ECPS

81

from KP2 preparations but no detectable loss of ECPS from KPl

preparations. The KP2 preparations in general, dissolved less

easily in DH2O than the KPl preparations, but seemed to dissolve

much better in alkaline solutions above a pH of 11. A Tris buffer

was then used (0.01 M Tris, pH 12) in many of the subsequent

purification steps, especially to avoid the cessation of a column

run due to aggregates of KP2 EPS clogging the column filters.

Ethanol fractionated EPS was resuspended in 0.01 M Tris, pH

12, or in another low ionic strength buffer [i.e., 0.02 M

(NH4)2C02], and placed on a DEAE-Sephacel column equilibrated with

the same buffer. The column was eluted with approximately 200 ml

of buffer before the salt gradient was applied. It was found that,

for all of the ethanol fractionated samples from the strains used

in this study, yielded a fraction of uronic acid containing

material eluting from the column at this time. This fraction was

labeled the neutral (N) EPS fraction. When a salt gradient was

applied [up to IM concentrations of either NaCl or (NH^)2C02],

a second uronic acid and hexose containing fraction eluted. An

example of an elution profile on DEAE Sephacel can be seen in

Figure 6 for the KPl-T ECPS. Table 17 shows the ionic strength of

the buffer at which two KPl and two KP2 EPS samples were eluted.

The acidic (A) fraction of EPS for the various samples eluted

between 0.2 and 0.4 M (NH^)2C03.

The acidic and the neutral fractions, after dialysis and

lyophilization, were resuspended in the appropriate column buffer

(O.OIM Tris, pH 12, or 0.5M NaCl were commonly used) and placed on

82

Table 17. Elution Ionic Strength and Apparent Molecular Weights of the ECPS from Various Strains of KPl and KP2

Elution

ECPS

KPl-O(N)^

KPl-O(A)^

KPl-T(N)

KPl-T(A)

KP2-0(N)

KP2-0(A)

KP2-270(N)

KP2 2-70(A)

HMW^

>3xl0^

>3xl0^

>,3xlO^

>^3xlO^

>3xl0^

>^3xlO^

>3xl0^

>3xl0^

%total

15

10

52

53

60

70

100

100

LMW^

8.9x10^

5.0x10^

9.4x10^

1.1x10^

2.3x10^

2.0x10^

—

_ — —

%total

85

90

48

46

40

30

--

_—

Ionic strength

0.02

0.42

0.02

0.21

0.02

0.23

0.02

0.17

^HMW; the high molecular weight or void volume fraction of EPS eluted from S-2B gel filtration.

^LMW; the low molecular weight fraction of EPS found within the, inclusion capabilities of the gel filtration column.

Êlution Ionic Strength; the salt gradient molarity of (NH^)2C03 at which the various EPS fractions eluted from DEAE-Sephacel.

*^(N); neutral fraction.

^(A); acidic fraction.

83

84

20 30 40 50 60 70 80

Fraction Numbtr 90

85

gel filtration (S-2B or BGA-150m) columns. Examples of elution

profiles for KPl-0 (A), KPl-0 (N), KPl-T (A), KPl-T (N), KP2-0 (A)

and KP2 2-70(A) EPS on S-2B can be seen in Figures 7 through 13.

For all strains utilized in these studies, the acidic and the

neutral fraction of EPS produced a high molecular weight (HMW),

hexose and uronic acid containing fraction at the void volume of

the gel filtration profile. For the KP2 strains the vast majority

of the polysaccharide material eluted in this HMW fraction, whereas

only 10 to 15 percent of the acidic or neutral EPS from KPl-0 and

52 to 53 percent from KPl-T eluted at this HMW fraction. The

remainder of the hexose and uronic acid containing material for all

strains was within the inclusion capabilities of the gel filtration

column, and was typically contained in one fraction. This fraction

was labelled the low molecular weight (LMW) fraction. For the

KPl-0 and the KPl-T strains this fraction accounted for 85-90% and

46-48% of all hexose and uronic acid containing material,

respectively. In contrast, it was the lesser of the two fractions

for the KP2 strains, with the KP2-0 strains producing a LMW

fraction which accounted for 30 to 40% of the polysaccharide in the

sample, and the KP2 2-70 strain apparently producing no LMW

material. It was found later that the KP2 2-70 strain does produce

a LMW component which appeared after ethanol extracted material was

placed on a BGA 150 m column equilibrated with 0.01 M Tris, pH 12,

but did not appear when 0.5 M NaCl was used with the acidic or the

neutral EPS. Table 17 displays the percentages of EPS that were in

86

87

Uronic Add/ ig /ml(o) o o o o o o o o o o o o

-r T

.a E

u o

« ^ ""—'—'—'—r

(•)|UJ/B7y980X9H

88

89

Uronic Acid / ig /ml (o)

(•) IUJ/BT/ OSOXOH

90

91

Uronic Acid/ig/mKo)

w — o a> l O CSJ —

(•) I U I / C T / 9S0X9H

92

93

Uronic Acid/ig/mI(o)

o Q o o o ^ (O (\j _

o o o O 0> OD

P o o o f «o «

( • ) | U J / 6 T / 9 « O X 9 H

94

95

10 15 20 25 30

FRACTION NUMBER

35 40

96

97

T

LU CO

o X UJ X

15 20 25

FRACTION NUMBER

40

98

99

Uronic Acid /ig/ml(o)

o o O O o - <D iO -t <Si

- I — I t . i

o <0

- o 'J-

«>

E 3

O o

o CVJ ro

O O fO

O cn CM

( • ) I U J / B T / 980X8H

100

101

250

200-

3 ' 50 UJ

o X Ul X

100-

60

FRACTION NUMBER

102

the HMW and the LMW peaks as well as the apparent molecular weights

calculated from dextran calibration standards (Fig. 14 and Table

18). The HMW fractions from all strains were outside the range of

the column and were calculated to be at least 3 x 10^ daltons in

molecular weight. The LMW fractions were, however, retained by the

column and gave molecular weights as seen in Table 17. For the KPl r c

LMW EPS the molecular weight range is between 5 x 10 and 1 x 10

daltons, while for the KP2-0 LMW EPS a molecular weight of around 2

X 10 daltons was estimated.

As purification proceeded, fractions containing hexose and

uronic acid were pooled, dialyzed and lyophilized. Dried materials

were then- brought up in 1 mg/ml or 2 mg/ml solutions in DHpO and

tested for their content of ECPS, ELPS and protein. Table 19 shows

the percentage of these components to the total dry weight of the

samples. It can be seen that nearly 50% of the KPl ethanol

extracted [KPl (EtOH) EPS] material was in the form of ECPS, while

approximately 5 to 7% of the dry weight was accounted for by ELPS,

and about 4 to 7% was protein. For the KP2 strains, 78 to 84% of

the dry weight of the ethanol extracted [KP2 (EtOH) EPS] material

was ECPS, 2.4 to 2.5% was ELPS and 1 to 1.7% was protein. Between

37 and 45% of the dry weight of KPl (EtOH) EPS and between 12 and

18% of the dry weight of KP2 (EtOH) EPS was unaccounted for in

these preparations by the methods used.

A second study of this type was then performed on the uronic

acid and hexose containing materials that were purified by DEAE-

103

Table 18. Elution Volumes for Dextran Calibration Standards on

S-2B

Dextran^ V ° K ^ av

5-40 X 10^ 159.8 0.0

2.0 X 10^. 250.0 0.27

5.0 X 10^ 379.8 0.67

2.3 X 10^ 396.7 0.72

1.7 X 10^ 413.6 0.76

8.1 X 10^ 415.5 0.77

9.4 X 10^ 453.08 0.89

^Dextran standards in average molecular weight (daltons)

V ; Elution volume in ml at peak of fraction.

''K ; determined by the equation av

V^ - V^ V^=159.8 ml e 0 0

âv

V. - V„ V =490 ml t o t

104

Table 19. The Extracellular Products Found in the Ethanol Fractionated Supernatants of K_. pneumoniae

Strain

KPl-0

KPl-T

KP2-0

KP2 2-70

ECPS^

52.2

42.6

78.2

84.1

ELPS^

6.5

5.2

2.5

2.4

Protein^

3.9

•7.0

1.0

1.7

Other^

37.4

45.2

18.3

11.8

Values of ECPS, ELPS and protein are expressed as the percentage of the dry weight of the sample. The ECPS was calculated from the uronic acid determinations which were performed by the method of Blumencrantz et al. (7). The ELPS was determined by the method of Osborn et al. (57). Protein was determined by the procedure of Lowry et al. (51).

Other; the percentage of unidentified components contributing to the dry weight of ethanol extracted materials.

105

Sephacel ion exchange chromatography. Table 20 shows the amounts

of ECPS and ELPS contained in 2 mg/ml solutions and the percent of

the total dry weight of samples from two KPl and one KP2 strains.

The EPS from KPl-0 exhibited the least amount of ECPS per mg of dry

weight sample, with the KPl-0 (A) EPS being 12.8% ECPS by weight

and KPl-O(N) EPS 20.2% by weight. The figures for KPl-T (A) and

KPl-T (N) were somewhat greater, as seen in Table 20. The KP2 2-70

(N) and KP2 2-70(A) EPS were the most highly purified ECPS samples

at this stage of purification (70 to 76% of the total dry weight).

The amount of ELPS contained in these samples was seen to correlate

directly with the amounts of ECPS present for both the KPl EPS

samples, and it was determined that 1.6 yg of ELPS was present for

eyery 10 yg of ECPS. For the KP2 2-70 EPS, which was a much more

highly purified preparation, 2.5 yg of ELPS was present for ewery

100 yg of ECPS. From 64 to 86% of the total dry weight for the KPl

EPS samples and between 22 and 28% of the KP2 2-70 EPS was left

unaccounted for by these measures.

The extracellular products found in KPl-0 and KP2 2-70 EPS

after gel filtration and the percentages of the total dry weight

that these products comprise are listed in Table 21 . It can be

seen that in 1 mg of the HMW fraction from KPl-0, about 30% was

ECPS, 8% was ELPS and about 9% was protein, and for the LMW

fraction 53% was ECPS, about 4% was ELPS and 3% was protein. Thus

for KPl-0 EPS the fraction most enriched with ECPS and containing

the least amount of ELPS and protein was the LMW fraction (with

106

Table 20. Comparison of the ECPS and ELPS content in the Neutral (N) and Acidic (A) Fractions from DEAE-Sephacel

Sample ECPS (N=2)

KPl-O(N)^ 403.1+45.1

KPl-O(A)^ 255.4+24.7

KPl-T(N) 608.3^51.8

KPl-T(A) 284.2+13.0

KP2 2-70(N) 1400.1+87.3

KP2 2-70(A) 1515.3+66.3

Quantity of component in yg contained in 2 mg of dried sample.

Percentage of component contained in the dried sample.

^(N); The neutral fraction of hexose-containing material obtained from DEAE Sephacel.

(A); The acidic fraction as in footnote c.

%total^

20.2

12.8

30.4

14.2

70.0

75.8

ELPS^ (N=2)

79.5+7.5

29.3+3.7

105.3+18.2

69.3+10.3

36.1+0.8

38.1+1.3

%total^

4.0

1.4

5.3

3.5

1.8

1.9

107

Table 21. The Extracellular Products Foynd in KPl and KP2 EPS After Purification

Strain

KPl-0

KPl-0

KP2 2-70

KP2 2-70

Fraction

HMW^

LMW^

HMW

LMW

ECPS'^

29.6

53.0

80.7

83.9

ELPS'^

7.9

3.7

2.8

1.6

n ^ . b Protein

8.7

3.0

4.9

3.4

Other^

53.8

40.3

11.6

11.1

^Purification procedures include ethanol extraction, DEAE-Sephacel and gel filtration chromatography.

^Values of ECPS, ELPS, and protein are expressed as percentage of the weight of the samples.

Ôther; the percentage of unidentified components contributing to the dry weight of the samples.

* HMW; the high molecular weight fraction.

^LMW; the low molecular weight fraction.

108

approximately 7 yg of ELPS and 6 yg of protein per 100 yg of ECPS).

The KPl-0 HMW EPS contained much more of these components

(approximately 27 yg ELPS and 29 yg protein per 100 yg ECPS). The

HMW and LMW fractions from KP2 2-70 contained far less quantities

of ELPS and protein, with the HMW fraction containing about 2 yg

ELPS and 6 yg of protein per 100 yg ECPS and the LMW fraction

containing 0.3 yg ELPS and 4 yg protein per 100 yg ECPS. The ECPS

accounted for the vast majority of the dry weight of KP2 2-70 ECPS

samples at this stage of purification (between 81 and 84%).

Finally, Table 22 shows the percent yield obtained for KPl-0

and KPl-T EPS at various stages of purification starting with a 200

ml culture of each strain grown in defined medium for 48h at 37 C.

The data reveal yields of between 24 and 80% throughout the

purification procedures. The final yield for KPl-0 ECPS was 22.5%

while the final yield for the KPl-T ECPS was 7.4%. Much of the

KPl-T ECPS was lost during purification on DEAE Sephacel, which was

also shown above not to further purify the ECPS from either the

KPl-0, KPl-T or the KP2 2-70 strain. Indeed, the eluted fractions

from ion exchange appear to be less pure than any of the fractions

before or after this purification step. These purification steps

apparently have not purified the KP2 ECPS beyond the level of

purification achieved by ethanol fractionation for the KPl ECPS.

Ethanol fractionation resulted in a product that is 43 to 52% ECPS

by weight, while no other subsequent fractions from the

purification schema, except for the LMW fraction from gel

109

Table 22. Percent Yield Obtained from ECPS Pur i f i ca t ion of KPl-0

Stage Preparation ygECPS/mg dry wt^ Total ECPS Yield(%)'

I 48h Dialyzed 79.3+10.2 108.2+2.0 100

Supernatants

II Ethanol 522.2+24.6 62.7+3.0 58

Fractionation

III DEAE 185.4^13.6 30.4+^2.2 28

Chromatography

IV BGA-150m

Chromatography HMW 287.0+38.3 2.4+0.3

IV BGA-150m 23

Chromatography LMW 471.8+96.4 21.9+4.5

The quantity of ECPS in the samples were determined from uronic acid measurements performed by the procedure of Blumencrantz et al. (8).

Yields were determined as the percent of ECPS at each stage of purification compared to the amount in the 48h dialyzed supernatant.

no

filtration, were of equal purity. The KPl LMW fraction retained

the highest purity of all the KPl fractions in that it contained

lower quantities of ELPS (7 yg per 100 yg ECPS) than the ethanol

extracted fraction (12 yg ELPS per 100 yg ECPS).

Due to the rather poor level of purity obtained for the ECPS

of the KPl strains by the procedures above, a different method was

utilized to purify the KPl ECPS from the KPl-0 strain. Ethanol

fractionated KPl-0 EPS was resuspended in DHÔ and subjected to

electrodialysis (ED) at 2000 V. All of the DHÔ in the cathode and

anode chambers were collected, lyophilized and referred to as

fraction I (Fr I). The electrodialyzed EPS was then fractionated

with cetavlon. The cetavlon precipitate, labelled Fraction II (Fr

II), and the cetavlon supernatant, labelled fraction III (Fr III)

were both washed with ethanol (3 times) and the product was tested

for its content of ECPS, ELPS and protein. The results of these

procedures are presented in Table 23. With 150 mg of starting

material, 130 mg total product were obtained gravimetrically after

these purification steps. Fr I was shown to contain no detectable

hexose or uronic acid and comprised at least 19 mg of material.

Therefore nearly 13% of the starting material was dialyzed free of

the EPS during ED. Fr II was found to comprise about 71% of the

weight of the starting material, and 77% of Fr II was found

chemically to be ECPS while 3% was ELPS, 3.5% was protein and 16%

was unaccounted for by these methods. Fr III, on the other hand,

comprised only 3% of the starting material. It was determined

Ill

Table 23. Pur i f i ca t ion of KPl-0 (EtOH) EPS by ED, Cetavlon and Gel F i l t r a t i o n

Sample

KPl-0 (EtOH)^

Fr 11^

Fr I I I ^

Fr I^

Fr I I , S-2B^

Fr I I , S-2B,ED

KPl-0 (EtOH)ED

mg dry wt

150.0

106.0

4.5

19.0

58.0

9 ^p

^ NP

ECPS

522.17+24.64

766.77+14.36

494.40+.56.15

ND" '

620.14j: 3.13

732.10+35.42

802.47^18.91

ELPS^

62.66+2.96

29.70+0.86^

212.73+2.96

ND

3 . 0 5 + 1 . 3 1

6 . 1 5 + 0 . 6 6

54.37+0.87

Protein

39.0+1.5

34.8+.7.6

NP^

ND

9 . 2 + 2 . 0

1 9 . 6 + 2 . 7

NP

Units are in yg per mg dry weight.

Êthanol extracted EPS.

^Fr I I ; the cetavlon prec ip i ta te from KPl-0 (EtOH) EPS washed x3 wi th EtOH.

^Fr I I I ; the supernatant from the cetavlon step washed x 3 with EtOH.

^Fr I ; the dialysable material col lected during ED (2000 molecular weight c u t o f f ) .

^Fr I I , S-2B; Fr I placed on Sepharose - 2B and the LMW component co l lec ted .

^Fr I I , S-2B, ED; Material in footnote d re-electrodialyzed.

•^KPl-O (EtOH) ED; KPl-0 (EtOH) EPS subjected to ED (separate experiment).

^NP; not performed.

"^ND; none detected.

112

chemically that 49.5% of Fr III consisted of ECPS while 21.3% was

ELPS, which left 30% of Fr III unaccounted for by these methods. A

portion of Fr II was then placed on a S-2B gel filtration column

and an elution profile was obtained as seen in Figure 15. This

figure illustrates that now the vast majority of the hexose

containing material elutes at a well defined LMW peak. Tube

fractions 25-45 were collected, dialyzed and lyophilized and

retested for ECPS, ELPS and protein content as seen under the label

Fr II, S-2B in Table 23. Nearly 90% of the ELPS has been removed

from this fraction, when comparing it to the Fr II starting

material, as well as 75% of the protein. However, the ECPS per mg

dry weight was nearly 15% less than that of Fr II. The Fr II, S-2B

material was then electrodialyzed to remove the salts acquired from

gel filtration. This final preparation, labeled Fr II, S-2B, ED in

Table 23, is now comparable to Fr II in its ECPS content (73%

compared to 77% respectively) and still shows comparatively low

levels of ELPS and protein. Finally, in a separate experiment

KPl-0 (EtOH) EPS was subjected to ED and then tested for ECPS and

ELPS content as shown in Table 23 under the label KPl-0 (EtOH) ED.

The ED step shows a 28% increase in the quantity of ECPS per mg dry

wt with comparable levels of ELPS, when compared to the KPl-0

(EtOH) EPS. Table 24 shows the amounts of ELPS and ECPS that were

contained in 100 yg ECPS from all of these purification steps.

Greater than 90% of the ELPS and between 65 and 80% of the protein

were removed from the KPl-0 (EtOH) EPS by these methods as seen in

113

114

80 '

70

- 6 0 ^ E ^ 50 r =k

U 4 0 o ^ 30 X

2 0

10

—

-

-

A A

10 20 30 40 50 60

FRACTION NUMBER

115

Table 24 Purification of KPl-0 (EtOH) EPS bv Fn cetavlon and S-2B: Percent of^LPs! Protean'''

and Other Materials

^^m}± ELPS/ECPS^ Protein/Frpc;a nû R ^ô^^^ri/tCPS Other material/ECPS^

KPl-0 (EtOH)^ 12.00 7.47

F^ n ^ 3.87 72.04

4-54 22.01

' ' ' ' ' " ^ 4 3 . 0 3 NPJ- .• d

^' I ' ND " Mnl

Fr I I , S-2B^ 0 . 1 6

Fr I I , S -2B , ED^ 0 . 8 4

NP

ND' i p J

T-48 5 9 . 2 8

2 - 6 8 3 3 . 0 8

a Aûn t of ELPS, protein or other materials in pg per 100 .g ECPS.

'•Âs In Table 23.

116

this table. Extracting the EPS with cetavlon alone without prior

ED gave a similar elution profile as was seen for the EtOH

extracted fractions and a similar level of purity (559.15 and

607.49 yg ECPS per mg dry weight and 20.89 and 52.05 yg ELPS per mg

dry weight for KPl-0 and KPl-T samples respectively).

Fr III (3.5 mg) was applied to a BioGel P-300 gel filtration

column (P-300), equilibrated with 0.1 M ammonium acetate, pH 8.1,

containing 0.1% SDS, after boiling for 5 min in the column buffer.

The elution profile in Figure 16 shows the hexose containing

fractions obtained. The Limulus Amoebocyte Lysate assay indicated

that the majority of the ELPS was located under the peak that

eluted at tube fractions 26-34. This latter pool was collected,

washed 3 times with EtOH and tested for its ECPS and ELPS content.

It was found that this pool was enriched with ELPS (45 yg of ELPS

per 100 yg of ECPS), more so than all other samples encountered.

The KP2-0 EPS was also subjected to the above purification

procedures (EtOH ED CET S-2B). No substantial difference was

seen in the quantitites of ECPS or ELPS between the KP2-0 (EtOH)

EPS and the EPS obtained by the above methods. However, the Fr III

material obtained from the supernatant after cetavlon extraction

was enriched with ELPS (38.1 yg ELPS per 100 y5 ECPS) compared to

the Fr II precipitate from cetavlon treatment (4.21 yg ELPS per 100

yg ECPS). Table 25 summarizes the data obtained from these studies

with the KP2-0 EPS. Figure 17 shows the elution profile of these

materials on S-2B. KP2 2-70 (A) EPS was also subjected to ED and

117

118

50

4 0 -

•g 3 0

C3>

3 U l CO

o X Ul X

20r

0 -

20 30 40 50 60

FRACTION NUMBER

119

120

FRACTION NUMBER

121

Table 25. Pur i f i ca t ion of KP2-0 EPS by ED and Cetavlon

Sample ELPS/ECPS^

KP2-0 (EtOH)^ 3.25+0.01

KP2-0 (EtOH)ED^ 3.20+0.27

KP2-0 (EtOH)ED, CET PPT^ 4.21+0.18

KP2-0 (EtOH)ED, CET SUPE^ 38.05+3.59

Âmourt of ELPS in yg found per 100 yg of ECPS (N=2).

^KP2-C (EtOH); Ethanol extracted KP2-0 supernatants from growth (48h) at 37 C in defined medium.

^KP2-0 (EtOH)ED; Electrodialysis of material in footnote b.

CET PPT; Ethanol washed precipitate from cetavlon extraction of material in footnote c.

^CET SUPE; Ethanol washed supernatant as in footnote d.

122

then fractionated with cetavlon, and both the cetavlon precipitate

(Fr II) and the cetavlon supernatant (Fr III) were ethanol

extracted. Table 26 shows the KP2 2-70 (A) EPS data utilizing

these purification methods. Figure 18 shows the elution profile

obtained for the Fr II material applied to the S-2B gel filtration

column. It can be seen in Figure 17 that these purification

procedures have produced a better separation of the two major

fractions of KP2-0 EPS (compare with Figures 11 and 22), but did

not dissociate the HMW form. Figure 18 shows that cetavlon

extraction did not further dissociate the HMW form of KP2 2-70 EPS

over that of ED alone, but, rather, tended to increase the

concentration of the HMW form with a decrease in the proportion of

lower molecular weight peaks. It can be seen in Figure 18 that

whenever ED was a part of the purification procedure, a more

prominent LMW fraction (tube fractions 22 to 44) can be seen,

regardless of cetavlon use in the purification protocol. Cetavlon

extraction alone does not reduce the proportion of HMW EPS to LMW

EPS, when compared to the profile for KP2 2-70(A) EPS, but the LMW

regions are markedly different from one another. Since cetavlon

extracts only acidic polysaccharides, the LMW peaks observed in

Figure 18 for the profile of KP2 2-70 (A) EPS, after tube fraction

36, are most likely composed of neutral polysaccharides that are

removed during cetavlon extraction. It is therefore likely that

only ED can effect a dissociation of HMW EPS, and that cetavlon

extraction serves to remove neutral polysaccharides, such as LPS,

from the EPS, which was found to be the case shown in Table 26.

123

Table 26. Purification of KP2 2-70 EPS by ED and Cetavlon

Sample^

KP2 2-70(A)

KP2 2-70(A), ED

KP2 2-70(A), ED CET PPT

KP2 2-70(A) CET PPT

KP2 2-70(A), ED, CET SUPE

KP2 2-70(A), CET SUPE

^KP2 2-70(A) EPS was subjected to ED (KP2 2-70(A), ED) at 2000 V and cetavlon extracted to give both the cetavlon precipitate (KP2 2-70(A), ED, CET PPT) and ethanol extracted material from the cetavlon supernatants (KP2 2-70 (A), ED, CET SUPE). KP2 2-70(A) EPS was also cetavlon extracted without ED (KP2 2-70, CET PPT) and ethanol extracted materials from the cetavlon supernatant were obtained (KP2 2-70(A), CET SUPE).

Amount of ELPS present in yg per 100 yg of ECPS.

Âmount of protein present in yg per 100 yg of ECPS.

^No ECPS detected.

^Not determined.

ELPS/ECPS^

3.67+0.19

3.26+0.23

3.42+0.29

2.89+0.12

46.15+0.11

d

Protein/ECPS^

2.51+0.09

0.04^0.03

0.07+0.04

ND^

2.38+1.50

d

124

125

6 0 ff

50

40

C7>

3 Ul 3 0 cn O X Ul X

20

10 -

• J — • -

FRACTION NUMBER

126

Effect of Purified Extracellular Products

from K. pneumoniae on Virulence

in a Mouse Model

The majority of the studies in this section were performed

with the moderately virulent KPl-T strain to test whether the

virulence of this strain could be enhanced by co-injection with

EPS at various stages of purity. These studies were performed

first by utilizing KPl EPS, as the material co-administered to mice

by IP injections, along with serial ten-fold dilutions of KPl-T.

At the dosages administered it was found that only the KPl-T (A)

and the LMW fractions of EPS from KPl-0 or KPl-T did not

significantly enhance the virulence of KPl-T over control values.

This was true for the LMW EPS fractions even at the high doses

(108-149 yg/mouse) administered. Table 27 displays the Alog-jQ LD^Q

per milligram of ECPS administered to these mice [A(log.|Q LDgQ)/mg

ECPS]. A wide range of differences are seen to exist among the

values presented in this table (from -1.00 to -40.00), and these

differences seem to correlate inversely with the degree of

purification. For example the A(log^Q LD^QJ/mg ECPS for the acid

or neutral ECPS from KPl-0 (-35.7 to -40.0) had at least three

times the virulence enhancing capability as did the same EPS

further purified by gel filtration (-3.89 to -11.78). These

differences reach significance at the levels shown in Table 28.

The KPl-0 HMW (N) EPS virulence enhancement values were also

significantly greater (p<0.01) than the KPl-0 LMW EPS values. Thus

127

Table 27. E f fec t of KPl EPS on KPl-T Vi ru lence in the Mouse Model

FCPS Adog^Q LD5Q)/mg ECPS a

KPl-0 (N)'^ -35.73 + 10.2

KPl-0 (A)^ -40.00 + 13.3

KPl-T (N)^ -9.12 + 0.99

KPl-T (A)^ -7.38 + 4.57

KPl-0 HMW (N)^ -11.78 + 0.52

KPl-0 LMW (N)^ -3.89 + 1.87

KPl-T HMW (N)^ -9.08 + 3.51

KPl-T LMW (N)^ -1.00 + 0.30

Â(log,Q LDr^)/mg ECPS; Change in the log.Q LD^Q compared to contrAV t r i a l s per mg ECPS co- in jected with tne KPl-T s t ra in ,

^Neutral EPS f rac t i on from DEAE-Sephacel.

Âcid EPS f rac t i on from DEAE-Sephacel.

*^High molecular weight f rac t ion from neutral EPS applied to Sepharose 2B gel f i l t r a t i o n .

^Low molecular weight f rac t ion as in footnote d.

CO

O

•4-J U (O i -

EPS

F

1 —

p —

for

A

CO

(O

+-> e f T 3 CU 1 —

-•-> CU

o -a Q - O

:E: CD E a j

•1— to CJ r j E O

E CJ LU x :

4-) CU

E - r -

cu 1 — E Z3 - 1 -i - rC3

the

1-T

S

CnCL E i k i

' f "

i - - E fO -(-> Q - ' i —

E ^ O c_) - o

cu X - M

• » - CJ S.- CJ

- M - r - , fTJ E

s : T -1

> o •M CD 'r—

• f —

KP

2-0

O l^>s

1 CM ^

CM >«-CL

KP

l-T

1 MW

K

Pl-

T HM

W

CD 1 I S

O 1

o

o- nr

, ^ <:C

1-T

(

I S

1— 1 (

C L <

^ •=c

l-0(

-7

r\

ex. c

' • — ^

•21

O 1

CL

0.0

25

1

" o c o - ^ ft

o

L O

0.00

\ 0.

05

LO

o

0.0

.02

o

.02

CZ5

• 3

D

D

1

0.0

5 .0

01

o 0.

01

0.10

LO CD> O

'O

.025

CD)

.02

CD

.02

CD

1 1

.70

o

0.0

25

o ft

o

0.00

1 1.

00

0.0

.02

o

.60

o

1 1

.02

o

.01

o

0.2

5

o CM • O

0.10

0

.70

0.2

5 .2

0

o

1 1 1

o CO

CD

CM CD

O

.02

o

0.00

5

, o o

ft

o

0.00

1 0

.25

0.01

1 1 1 1

LO

o CD

CM O

CD

CM O

O

.02

o

0.4

0

o LO

CD

0.0

5 0

.05

1 1 1

c:

CD

LO CM

CD

O

CD

005

o

.005

o

0.2

0

CM

o o

0.0

25

i 1

0.0

5

LO CM

CD

O

ft

o

o o

ft

o

CD

LO

o CD

0.6

0

o CM

o

1 1 1

30

*0

0.0

5 00

1

ft

o

o

CD

001

ft

O

O

O

005

CD

0.7

0

1

1 1

0.2

0 0.

02

0.05

00

1

CD

O CM

ft

O

o

CD

001

o

001

• o

1 1 1

o ft

o

0.6

0 0

.20

0.40

00

5 ft

o

LO CM

ft

o

-

025

ft

o

LO O

ft

o

025

o

0.5

0

o ft

o

0.00

5 0.

30

0.4

0

CM

o ft

o

o ft

o

o CM

ft

o

LO

o ft

o

LO

o •

o

the

on

s.

M - - r -

O +-> C/^ E

CD E UJ S-

CU CD +-> E CU

' ^ " O

o -Lr>—

o o r-^+-

A(l

og

used

CU t/) - E fO

rin

g

test

n3

om

com

p e

nt'

s t

i - TD M- =5

+-)

tain

ed

The

s o .

n3 CU CO 3 to

ro CD > CO

Q . 03

CU CU . E x : 4-> + J

CO E •M o E

rese

ho

se

C L 4 J CU t- o

+-> X

• 1 — CU S.- 4-J

• M fO fO E E - r -

QJ S-- E O

128

o S-

CL

QJ

OO CM

QJ

JD 03

^ CD

1

KP

l

<: o

1

KP

l

2 :

1— 1

KP

l

<c 1 —

1

KP

l

i r o

1

KP

T

_ j

o 1 •L

dX

3 :

f—

KP

l-

—J

h-

KP

l-

o 1

CM

CM Q L

•—^ < "«— O

1 CM CL ^^

o - - ^ r n: 1 o

CM +-> UJ

CvJ — -CL i

J:: CJ 03

UJ fO

E O

I / ) QJ

1 —

CL E a in

129

as purification proceeded for the KPl-0 EPS, virulence enhancement

significantly decreased.

There were no significant differences seen between the KPl-T

(A), KPl-T (N) and the KPl-T HMW EPS in their virulence enhancing

properties for the KPl-T strain. However, the KPl-T (N) and the

KPl-T HMW (N) ECPS were significantly greater virulence enhancers

for KPl-T than was the KPl-T LMW EPS (p values are shown in Table

28). The A(log^Q ^^^0^/^^ ^^P^ values for KPl-T (N), KPl-T (A) and

KPl-T HMW are all quite similar (-9.12, -7.38, and -9.08,

respectively) whereas the value for KPl-T LMW is -1.00. Thus for

both the KPl-0 and the KPl-T EPS, the LMW fraction showed

significantly less virulence enhancement for KPl-T in the mouse

model than the HMW fractions from either strain.

When comparing the KPl-0 EPS to the KPl-T EPS at various

stages of purity as to their virulence enhancing capabilities, it

was found that the KPl-0 (N) and the KPl-0 (A) EPS were

significantly more virulence enhancing than the same fractions of

KPl-T EPS, but the HMW (N) EPS fractions from either strain did not

differ significantly in this parameter (see Table 28). The LMW EPS

from KPl-0 (N) did, however, differ significantly (p<0.05) from the

KPl-T LMW EPS, even though both values were small compared to the

other values obtained for KPl EPS in Table 27. With the exception

of KPl-0 (N) and KPl-0 (A) EPS on one extreme, and KPl-0 LMW and

KPl-T LMW EPS on the other, the remaining four fractions of KPl EPS

show very similar A(log^Q LD^Q)/mg ECPS values, the mean of these

values being -10.58 ± 3.60.

130

Cetavlon extracted EPS from KPl-0 cultures [KPl-0 (CET) EPS]

was also tested in the mouse virulence assay, with both the KPl-T

and the KP2-0 s t r a i n , to see i f a d i f fe ren t pu r i f i ca t i on procedure

could a f fec t the virulence enhancement properties of KPl-0 EPS.

At 200 yg/mouse, the KPl-0 (CET) EPS was shown to enhance the

virulence of both the KPl-T and the KP2-0 s t ra in s i gn i f i can t l y over

control values. The A(log^Q ^^Q^^^ ECPS for KPl-0 (CET) EPS in

the virulence studies with the KPl-T s t ra in (-8.40 + 0.92) is not

s i g n i f i c a n t l y d i f f e ren t from the mean value (-10.58 ± 3.60)

obtained in the ea r l i e r studies in th is sect ion. Therefore the EPS

of KPl-0 retains i t s virulence enhancing properties whether an

ethanol-based or a cetavlon-based extract ion is u t i l i z e d .

The next series of studies were performed in order to

determine the a b i l i t y of the KP2 EPS to enhance the virulence of a

KPl organism. KP2 EPS, at various stages of p u r i f i c a t i o n , from

e i ther the v i ru len t KP2 2-70 or the moderately v i ru len t KP2-0

s t r a i n , were co-administered IP with the KPl-T s t ra in into mice,

and lOrr. values were obtained. Appendix 15 displays the dosages of

the various EPS f rac t ions along with the log-jg LDrg data and the

A(logiQ LDCQ) wi th respect to contro ls . S ta t i s t i ca l analysis of

these values showed that (1) administrat ion of KP2 2-70 ethanol

extracted [KP2 2-70 (EtOH)] EPS at 336.2 yg ECPS/mouse enhanced the

virulence of KPl-T s i gn i f i can t l y over control values (p<0.005); (2)

administ rat ion of KP2 2-70 (A) EPS into mice at a dosage of 454.1

yg ECPS/mouse, but not at 204.7 or 101.5 yg ECPS/mouse, enhanced

the virulence of KPl-T s i gn i f i can t l y over control values; and (3)

131

administration of KP2-0 (A) EPS at 428.6 yg ECPS/mouse did not

significantly enhance the virulence of KPl-T in the mouse model.

Table 29 shows the A(log^Q '-D5Q)/mg ECPS for each of the ECPS

fractions utilized in these studies. The KP2 2-70 (EtOH) fraction

is seen here to have produced the greatest increment of virulence

enhancement for the KPl-T strain in the mouse model (-5.38 +0.38),

followed by the KP2 2-70 (A) EPS at the higher doses (-3.56 + 0.48

and -4.21 + 1.18). The KP2-0 (A) EPS produced a relatively small

increment in virulence enhancement for KPl-T (-2.03 + 2.77) as well

as did the lowest dosage of KP2 2-70 EPS (-1.28 + 2.35). None of

these differences in virulence enhancement among the KP2 EPS

fractions were significantly different, however. Table 29 also

shows the A(log^Q '-' SO /' ^ ECPS obtained for the six KP2 2-70 (A)

EPS trials at three different doses. Four of the six values used

to calculate this mean also approximated the mean closely, whereas

two of the data points (one in the highest and one in the lowest

dosage) were unrelated to the mean obtained. Similarly, the two

data points used to obtain the mean A(logiQ LDrQ)/mg ECPS value for

KP2-0 (A) EPS were dissimilar. A high degree of variability was

thus seen in these experiments, as well as in the experiments

utilizing KPl EPS to enhance the KPl-T strain in the mouse model.

When comparing the ffect of KPl EPS to the effect of KP2 EPS

on the virulence of KPl-T in the mouse model, the following trends

were evident: (1) the HMW fractions or the fractions from

ion-exchange for the KPl EPS are between 2 and 10 times more

virulence enhancing for the KPl-T strain than are the KP2 EPS

132

Table 29. Effect of KP2 EPS on KPl-T Virulence in the Mouse Model

hPb

KP2 2-70 (EtOH)^

KP2 2-70 (A)^

KP2 2-70 (A)

KP2 2-70 (A)

KP2-0 (A)^

Âs in Table 27.

yg ECPS/mouse

336.2

101.5

204.7

454.1

428.6

Adog^Q LD3Q)/mg ECPS

-5.38 ± 0.38

-1.28 +2 .35^

-3.56 + 0.48^

-4.21 + 1.18^

-2.03 + 2.77

Ethanol extracted EPS from 48h cu l tura l supernatants.

Âcid EPS f rac t ion from DEAE-Sephacel

^The mean and standard deviat ion of a l l the KP2 2-70 (A) EPS t r i a l s (N=6) was -3.02 ± 1.83.

133

fractions; and (2) the LMW fractions of KPl EPS are equal in

potency to the KP2 EPS in enhancing the virulence of KPl-T. Table

28 shows the p values for the comparison of the means of the

A(log^Q '- So'/' S ECPS for KPl and KP2 EPS. The KPl-0 (N), KPl-0

(A) and the KPl-0 HMW EPS are all significantly greater virulence

enhancers than any one of the KP2 EPS fractions, while the KPl-T

(N) EPS fraction is significantly more potent as a virulence

enhancer of KPl-T than two of the three KP2 EPS fractions [KP2 2-70

(A) and KP2-0 (A), but not KP2 2-70 (EtOH) EPS]. The KPl-T HMW

EPS was a significantly greater virulence enhancer than the KP2

2-70 (A) EPS (p<0.02) but did not differ significantly from the

KP2-0 (A) or the KP2 2-70 (EtOH) EPS. There were no significant

differences seen between the KPl-T (A) or the KPl-0 LMW EPS when

compared to any of the KP2 EPS fractions in virulence enhancement

of KPl-T. Finally, the A(log^Q ^^SO^^^^ ^^^^ ^ ° ^ ^^^ ^^^''^ "" ^

EPS was significantly lower (p<0.005) than the value obtained for

the KP2 2-70 (EtOH) EPS, but not different significantly from the

other KP2 EPS fractions.

The reverse experiment was then performed, wherein a KP2

strain was utilized in standard mouse virulence assays, with the

additional co-administration of EPS from KPl or KP2, to see if EPS

from either serotype could enhance the virulence of the KP2 strain.

The moderately virulent KP2-0 strain was chosen for these trials,

and either KPl-0 (N) EPS at 40.3 or 200.0 yg ECPS per mouse were

co-administered. The A(log^Q ldc^Q)/mg ECPS are given in Table 30.

These data reveal that the virulence of the KP2-0 strain was

134

Table 30. E f fec t of KPl or KP2 EPS on the V i ru lence o f KP2-0 in the Mouse Model

EPS Adog^Q LDgQJ/mg ECPS^

KP2 2-70 (EtOH)^ -1.70 + 0 . 4 6

KPl-0 (N)^ - 3 . 9 7 + 0 . 0 0 (p<0.005)^

KPl-0 (CET)^ -5.08 + 0.88

Âs i n Table 27.

^KP2 2-70 (EtOH); Ethanol ex t rac ted EPS from 48h supernatants of KP2 2-70 grown in def ined medium a t 37 C.

^KPl-0 (N) ; Neutral EPS f r a c t i o n from DEAE sephacel.

*^KPl-0 (CET); Cetavlon ex t rac ted EPS

^ S t a t i s t i c a l ana lys is o f the means comparing the A(log•,f^ LDt-p,)/mg ECPS of KP2 2-70 (EtOH) EPS to t ha t o f KPl-0 (N) EPS.""

135

enhanced in the mouse model by co-injection of EPS from either the

KPl-0 or the KP2 2-70 strain. When cetavlon extracted KPl-0 EPS

was co-injected with the KP2-0 strain, a A(logiQ LDgQ)/mg ECPS

value of -5.08 j 0.88 (N=2) was obtained, which was not

significantly different than the value obtained for KPl-0 (N) EPS.

Both values, however, were significantly greater than that obtained

for the KP2 2-70 EPS with the KP2-0 strain. In general, however,

the extent of virulence enhancement was relatively small compared

to that manifested by the KPl-T strain with these same EPS

fractions. The KP2 2-70 (EtOH) EPS enhanced the virulence of KPl-T

with approximately three times the magnitude of that which it

enhanced the KP2-0 strain (-5.38 compared to -1.70, respectively)

per mg of ECPS. This difference is significant at the p<0.001

level. The same phenomenon was observed for the effect of KPl (N)

EPS on KP2-0 and KPl-T virulence, namely that the KPl-0 (N) EPS was

significantly more virulence enhancing (p<0.01) for KPl-T than for

KP2-0 (-35.7 compared to -3.97, respectively) per mg of ECPS.

Finally, when comparing the effect of KPl EPS versus the KP2 EPS in

the virulence enhancement of KP2-0, it was seen that the KPl-0 (N)

EPS was significantly more enhancing (p<0.005) than the KP2 2-70

(EtOH) EPS. It has thus been determined that the KPl EPS was

significantly more virulence enhancing for both serotype 1 and

serotype 2 K.. pneumoniae than the KP2 EPS in the mouse model.

Furthermore, the virulence of the KPl-T strain was affected

significantly more so by the co-administration of either KPl or KP2

EPS than was the KP2-0 strain.

136

The next set of experiments were performed to determine

whether the virulence enhancement properties of EPS could be

influenced by electrodialysis (ED) of the EPS as a purification

step before utilization in the mouse model. Both the KP2 2-70

(EtOH) and the KP2-0 (EtOH) EPS were electrodialyzed at lOOOV until

no further marked increases in mA were observed over a 30 minute

period. Both of these types of KP2 EPS were then injected

separately at various dosages with the KPl-T strain into mice.

Table 31 summarizes the results for KP2 2-70 (EtOH) EPS in terms of

the A(logiQ LDcgj/mg ECPS. It can be seen from this table that the

EPS before ED enhanced the virulence of KPl-T to a significantly

greater extent (p<0.005) than did the EPS after ED (-5.38 compared

to -3.01, respectively). Table 32 summarizes these results in

terms of the A(logiQ LDcQ)/mg ECPS. In contrast to the effect of

KP2 2-70 (EtOH) EPS, the KP2-0 (EtOH) EPS enhanced KPl-T virulence

significantly less (p<0.005) before ED. However, after subjecting

KP2-0 (EtOH) EPS to ED, its virulence enhancement capabilities

significantly increased (p<0.005) over that of the same EPS before

ED. There was no significant difference between the effect of

electrodialyzed KP2-0 (EtOH) EPS and the effect of electrodialyzed

KP2 2-70 (EtOH) EPS on KPl-T in the mouse model. There still

remained a significant difference (p<0.02) between the

nonelectrodialyzed KP2 2-70 (EtOH) and the electrodialyzed KP2-0

(EtOH) EPS as to their virulence enhancement capabilties. These

experiments then showed a difference in the effects of ED upon the

ECPS from one strain (KP2-0) having manifested an increase in

137

Table 31 . Effect of E lect rodia lys is (ED) on the Virulence Enhancement of KPl-T by KP2 2-70

EPS in the Mouse Model

KP2 2-70 EPS^ Adog^Q LD5Q)/mg ECPS*

Before ED -5.38 ± 0.66

Af ter ED^ -3.01 +0 .56 (p<0.005)

^KP2 2-70 EPS; obtained by ethanol extract ion of 48h supernatants of 37°C cul tures grown in defined medium.

^ As in Table 27.

Êlectrodialysis proceeded at lOOOV until no marked increase in mA occurred over a 30 min period.

^Statistical comparison of the means of the A(log.jQ LDgQ)/mg ECPS obtained before and after electrodialysis.

138

Table 32. Effect of ED on the Virulence Enhancement of KPl-T by KP2-0 EPS in the Mouse Model

KP2-0 EPS^ ^loSlO '-^50'^^^ ^^^^- (^^^'^^^

Before ED -1.15 + 0.36 (N=2)^

Af ter ED^ -3.81 +0 .14 (N=3) (p<0.005)^

^KP2-0 EPS; obtained by ethanol extract ion of 48h supernatants of 37 C cultures grown in defined medium.

Âs in Table 27.

^N equals the number of LD^Q t r i a l s performed.

^S ta t i s t i ca l analysis (Student's t tes t ) of the Alog^Q ' -^SQ/^^ ECPS for KP2-0 EPS before and a f te r e lec t rod ia lys is .

139

virulence enhancement, while in the other strain (KP2 2-70),

virulence enhancement of the ethanol extracted EPS decreased as the

result of ED. Moreover, the EPS from the less virulent KP2-0

strain was seen to be significantly less virulence enhancing for

KPl-T than was the EPS from the highly virulent KP2 2-70.

The next set of experiments were performed to determine the

effect of saponification, as a means to remove covalently linked

fatty acid esters from EPS, on the virulence enhancing properties

of EPS in the mouse model. In these trials the neutral EPS from

KPl-0 [KPl-0 (N)] was saponified before utilization in mouse

virulence tests. The A(logiQ LDrgj/mg ECPS is displayed in Table

33. Both the KPl-T and the KP2-0 strain were tested. It was found

that saponification of the KPl-0 (N) EPS.significantly decreased

its virulence enhancing properties both for the KPl-T and the KP2-0

strain in the mouse model without affecting the type specificity of

the ECPS. The A(log^Q LD5Q)/mg ECPS dropped from -35.7 +_ 10.20 to

-3.4 +_ 0.92 for KPl-T after saponification while, in tests

involving KP2-0, it dropped from -3.97 ± 0.00 to -0.70 J: 0.01.

These data suggest that certain ester linkages (presumably fatty

acids esterified to the ECPS or ELPS of these preparations, as

addressed in Results, Section H) were contributing to the virulence

enhancing capabilities of these materials. Tables 34 and 35 show

the A(log.Q L D ^ Q ) per yg of ELPS contained in a variety of the

polysaccharide samples of KPl and KPl EPS, respectively, utilized

in these virulence enhancement studies on the KPl-T strain. These

tables show that, when comparing the ELPS content of these

140

Table 33. Effect of Saponif ication^ on the Virulence Enhancement of KPl-T and KP2-0 by KPl-0 (N)

EPS in the Mouse Model

Stra in in jected EPS A(log^Q L^5o)/"^9 ^^^^^

KPl-T KPl-0 (N)^ -35 .70 + 10.2

KPl-0 (N) - 3 .40 + 0.92 (p<0 .05 )^

(saponified)

KP2-0 KPl-0 (N) -3.97 + 0.01

KPl-0 (N) -0 .70+ 0.01 (p<0.001)^

(saponified)

^KPl-0 (N) EPS (10 mg) was placed in 0.5 M NaOH overnight at room temperature and dialyzed 3 times against 8L DHpO while stirring.

Âs in Table 27.

^KPl-0 (N); the neutral fraction of ethanol fractionated EPS placed on DEAE-Sephacel at 200 yg/mouse.

^Statistical analysis (Student's t test) of the A(log.^ LDc.^)/mg ECPS for saponified and untreated EPS in the KPl-T stady.^^

^Statistical analysis as for the KP2-0 study as in footnote d.

141

Table 34. Virulence Enhancement of KPl-T in the Mouse Model: Comparison to the Dosage of ELPS in

KPl EPS Samples

Sample^

KPl-0 (N)

KPl-T (N)

KPl-0 (N)

KPl-0 (N)

HMW

LMW

yg ELPS/mouse

8.0

12.2

16.8

6.8

Adog^Q LD5Q)/yg ELPS^

-0.18 + 0.08 (N=3)

-0.09 + 0.01 (N=3)

-0.06 + 0.002 (N=2)

-0.07 + 0.04 (N=3)

Samples were resuspended in PBS at 2 mg dry weight per ml and 0.1 ml was co- in jected with the KPl-T s t ra in IP into mice.

b A(log,Q LDcQ)/yg ELPS; the change in the log^Q LD^Q per yg injected.

^N equals the number of LDrQ trials performed.

142

Table 35. Virulence Enhancement of KPl-T in the Mouse Model: Comparison to the Dosage of ELPS in KP2 EPS Samples

Sample^ ng ELPS/mouse A(log^Q L^5o'/^9 ^ ^ b

KP2 2-70 (EtOH) 9.44 -0.19 + 0.13

KP2 2-70 (A)

KP2 2-70 (A)

KP2 2-70 (A)

11.40

5.70

2.85

-0.17 + 0.07^

-0.13 +0 .06^

-0.04 + 0.17^

KP2-0 (A) 9.97 -0.09 +0 .12

Samples were resuspended in PBS at various concentrations and 0.1 ml was CO-injected wi th the KPl-T s t ra in IP into mice.

Âs in Table 33.

^The mean and standard deviation of all the KP2 2-70 (A) EPS trials (N=6) is -0.11 + 0.10.

143

preparations, no significant differences between different EPS

samples were seen as to their virulence enhancement capabilities

per yg of ELPS. The average A(log^Q "-^so'/^S ELPS for all of the

KPl EPS and KP2 EPS samples was calculated to be -0.10 ± 0.05 and

-0.12 +_ 0.06, respectively. These results argue strongly for a

central role of ELPS in the virulence enhancing properties of these

preparations.

To further clarify the roles of both ELPS and ECPS in

virulence enhancement, the KPl-0 (EtOH) EPS was subjected to an

alternative purification procedure involving electrodialysis (ED),

followed by cetavlon fractionation (CET) and gel filtration on

Sepharose -2B. This alternative procedure allowed for separation

of the ECPS from the ELPS in this sample, as seen in Table 24. The

Fr III sample was then re-extracted with cetavlon after ED at 2000V

to give a sample of ELPS, found in the cetavlon supernatant, that

was free of ECPS as demonstrated by the absence of uronic acid

activity in the preparation. This ELPS fraction was also shown to

precipitate with anti-type 1 antiserum in double diffusion studies

but did not exhibit identity with the ECPS from Fr II. It was

found that, at a dosage of 400 yg of ECPS (and 15.5 yg ELPS) per

mouse, the Fr II material enhanced the virulence of KPl-T

significantly over controls (p<0.01). However, after Fr I was

passed over S-2B and the majority of the ELPS removed, and then

subjected to ED to remove salts obtained from the column [KPl-0 Fr

II, S-2B (ED) EPS], the resultant material at 400 yg/mouse did not

enhance the virulence of KPl-T over controls. The Fr III material

144

at 9.4 yg ELPS/mouse and with no detectable ECPS decreased the LDrg

of KPl-T by 0.45 log,Q units but this was not significantly

different than control LD^Q values (p<0.10). Table 36 shows the

A(log.jQ LD^Q)/mg ECPS and the A(log^Q LD^gj/yg ELPS which resulted

from these studies using the alternative purification methods. The

points to be made here are that, 1) with further purification the

A(log-.Q LDrQ)/mg ECPS values decreased, 2) regardless of the extent

of purification the A(logiQ LD^QJ/yg ELPS remained essentially the

same and 3) a dosage of somewhere between 10 and 15 yg of ELPS per

mouse was sufficient to significantly enhance the virulence of

KPl-T over control values. In comparing the A(log-jQ LDcQ)/yg ELPS

values in this study with those from Table 34, it can be seen that

there were no significant differences in these values for all of

the polysaccharide samples tested in the mouse model. The mean

value for the A(logiQ LDrQ)/yg ELPS for KPl samples turns out to be

-0.086 j: 0.044 (N=7). Therefore, approximately 11.6 yg of ELPS

from KPl-0 or KPl-T should decrease the LDCQ of KPl-T in the mouse

model by one log-jQ unit. For the KP2 EPS the mean value for the

A(log,Q LD5Q)/yg ELPS was determined to be -0.12 + 0.06 (N=5).

Therefore 8.3 yg of KP2 ELPS were apparently needed to decrease the

LD -f value of KPl-T by one log-jQ unit. Due to the high variability

in these assays, no significant difference could be seen between

the effects of KPl and KP2 samples on KPl-T virulence enhancement,

with respect to the dose of ELPS.

145

Table 36. Effect of an Al ternat ive Pur i f i ca t ion of KPl-0 EPS on the Virulence Enhancement of KPl-T in the Mouse Model

Sample^ ^(ÔS^Q LDggJ/mg ECPS ^(^09lO ^^50 ' /^^ ^^^^^

KPl-0 Fr I I - 2 . 6 8 + 0 . 4 4 - 0 . 0 7 + 0 . 0 1

KPl-0 Fr I I - 0 . 7 4 + 1 . 1 4 - 0 . 0 8 + 0 . 1 3

S-2B (ED)

KPl-0 Fr I I I - ^ -0.05 + 0.00

^Samples were prepared from ED of the ethanol f ract ionated prec ip i ta ted from 48h supernatant f l u i d of KPl-0 cultures followed by cetavlon f rac t i ona t ion . The Fr I I sample is the cetavlon p rec ip i t a te . The Fr I I S-2B (ED) sample is material obtained from placing the Fr I I sample on S-2B, co l lec t ing the LMW f rac t ion and e lect rod ia lyz ing at 2000V. The Fr I I I sample is the ethanol extracted material obtained from the supernatants a f te r cetavlon ex t rac t ion .

As in Table 27: footnote a.

Âs in Table 34: footnote b.

No ECPS was detected in th is f rac t ion as determined by the uronic acid assay of Blumencrantz et a l . (7 ) .

146 Structural Studies on the EPS Produced

by K. pneumoniae

The primary focus of these studies was on the effects of ED on

the EPS from the KP2 strains. Ethanol extracted KP2 2-70 [KP2 2-70

(EtOH)] or KP2-0 [KP2-0 (EtOH)] EPS from 48h supernatants were

resuspended in column buffer (0.01 M Tris, pH 12) and applied to

either BGA-150 m or S-2B gel filtration columns equilibrated with

the same buffer. A portion of these samples were subjected to ED

for various time periods before application to the column. Figure

19 shows the effect of ED on the elution profile from BGA-150 m for

the KP2 2-70 (EtOH) EPS at 4 mg dry weight per ml. Before ED there

were two prominent hexose containing peaks, one eluting at the void

volume of the column (HMW) and one retained by the column (LMW)

with a peak of hexose activity at fraction 33-36. After subjecting

the KP2 2-70 (EtOH) EPS to ED at 400V (intermediate ED), an elution

profile on BGA-150m was obtained, denoted by the profile seen in

Figure 19. In this profile at least two thirds of the HMW fraction

from the profile before ED is now absent, and the LMW component has

increased in magnitude. The 400V ED product was then subjected to

ED at lOOOV and the resulting material was placed on BGA 150m. The

elution profile for this final ED product of KP2 2-70 (EtOH) EPS

can be seen in Figure 19. The HMW component in this final ED

profile is now virtually absent and the LMW peak is apparently more

refined (higher concentrations of hexose containing material in a

fewer number of fractions). It was found that approximately 10% of

the hexose activity was lost during ED and the majority of this

hexose activity was recovered in the fluids collected outside of

147

148

I I 1 1 1 1 — O «rt O lO O uo r o CM CM — —

( I O J / D T / ) 9S0X9H

149

the ED dialysis bag (in the cathode and anode fluids). Subsequent

studies were performed utilizing dialysis tubing with a 1000 or

2000 molecular weight cutoff rather than a 10,000 molecular weight

cutoff, so as to retain all of the hexose activity within the

dialysis tubing.

The KP2 2-70 (EtOH) EPS at 4 mg/ml was again subjected to ED,

this time at 400V, lOOOV and then at 2000V and subjected to RID in

agarose impregnated with antiserum prepared in rabbits against the

KP2 2-70 organism. Figure 20 shows the precipitin zones formed from

these antigen-antibody interactions in agarose at various stages of

ED. Well a in Figure 20 shows the KP2 2-70 (EtOH) EPS before ED.

In this figure a number of diffuse precipitin zones can be seen

within a rather broad and hazy background. When this material was

subjected to ED at 400V the RID profile seen in well b in Figure 20

was obtained. At this stage one predominant precipitin zone was

observed, with a diameter far less than that in well a. Apparently

a small portion of the EPS was also precipitating with antibody at

the periphery of the sample well. It was not certain, however,

whether this precipitate was due to antibody- antigen interactions.

Figure 20, well c, shows the RID profile for KP2 2-70 (EtOH) EPS

after ED at lOOOV, and is essentially the same profile as that seen

in well b. To be sure that ED was complete, the EPS was subjected

to 2000V, and the RID profile seen in Figure 20, well d, was

obtained. In this figure it can be seen that the zone diameter of

the precipitin ring has apparently increased over that of well c,

even though the concentration of ECPS placed in this well was

150

151

iîsissssas^

152

approximately the same as that in the other RID profiles.

From the data obtained from gel filtration and RID studies on

electrodialyzed products of KP2 2-70 (EtOH) EPS, it can be

concluded that ED affects the EPS in two different fashions: 1) HMW

components have apparently dissociated and have given rise to LMW

components and, 2) The LMW components have become more homogeneous

in molecular size as ED proceeded. To substantiate these

conclusions the KP2 2-70 LMW EPS from the BGA-150m profile of KP2

2-70 (EtOH) EPS electrodialyzed at lOOOV (Figure 19) was pooled,

collected, dialyzed and lyophilized to dryness. A comparatively

small sample (2mg dry weight) of the LMW EPS was reapplied to

BGA-150m before and after another round of ED at 1000 V. The

elution profile for the KP2 2-70 (EtOH) LMW EPS before and after ED

can be seen in Figure 21 . Before ED it can be seen that the

elution profile contains a predominant HMW peak that eluted at the

void volume (fractions 16-18), even though no HMW material was

carried over from the initial column run. Also at this low

concentration of EPS one can now see that the LMW fraction

apparently consisted of a number of distinct lower molecular weight

species that, with larger sample volumes, seemed to coalesce into

one broad fraction. These individual LMW fractions were somewhat

symnetrical and occured in the profile at a rather distinct

periodicity (peaks occured for every 60 ml of eluant on the

average). After ED an elution profile was obtained as seen in

Figure 21. The HMW component was again seen to diminish, while the

LMW components seem to have become somewhat less broad and more

refined toward the center of the LMW region. Moreover a new peak

153

154

8

o» 6 U l CO

o X Ul X i

30 40

FRACTION NUMBER

155

was seen near fraction 70.

The ethanol extracted 48h supernatant from the KP2-0 strains

[KP2-0 (EtOH) EPS] was then subjected to ED at up to 2000V to

compare the elution profiles of this ECPS on gel filtration before

and after ED. Figure 22 depicts the elution profile of the KP2-0

(EtOH) EPS before and after ED at 2000V on BGA-150m (6 mg dry

weight applied). Before ED the hexose and uronic acid activity was

limited largely to the void volume of the column, with some

evidence of a second fraction manifested as a shoulder of hexose or

uronic acid activity to the right of the void volume peak at around

fraction 24, and a third small fraction that peaked at fraction 42.

After ED a slight shift to the right is noted, with more hexose

activity found in later fractions (fractions 27-60) and a small,

though not well-defined peak at fraction 57. To demonstrate the

existence of two juxtaposed peaks of activity at or near the void

volume, a much smaller sample of KP2-0 (EtOH) EPS was applied to

the same column and the elution profile can be seen in Figure 23.

In Figure 23' the peak to the right of the void volume fraction was

somewhat more evident in the profiles before and after ED but in

contrast to the EPS from KP2 2-70, the HMW fraction was not

drastically diminished in magnitude after extensive ED.

The ability of ED to dissociate higher molecular weight KP2

2-70 EPS to that of lower molecular weight components argues for

the presence of electrophilic interactions between polysaccharide

strands. During ED it was found that as the mA increased across

the terminals, there was a concommitant increase in pH at the

cathode and a decrease in pH at the anode. The surge in mA during

156

157

(iuj/6rf)aiov DiNOdn

cr Ul

m

<

Li-

- O

-

o CM

1

o o

1

o GO

1

o CD

1

o 1

o CM

(|ai/Brr)3S0X3H

158

159

CP

3 U l CO

o X Ul X

FRACTION NUMBER

160

ED tended to produce high temperatures, so it was necessary to

change the DHÔ in all chambers when a current of approximately 25

mA was reached. The column labeled "Wash" in Table 37 refers to

these DHÔ changes as ED proceeded, the washes being numbered

progressively. The pH of the DH2O was measured, as well as was the

pH in the cathode and the anode, once this 25 mA current was

achieved, for each wash. The change in the pH in the cathode or

the anode was calculated by subtracting the pH of DHpO from the

resulting pH in either chamber. The ratio of the change in the pH

in the cathode to the change in pH in the anode (ApH cathode/ApH

anode) was then determined, and these results are listed in Table

37. From this table it can be seen that the highest ratio between

the pH changes at the cathode and the anode occured within the

first three washes, where over 2 pH unit changes have occurred in

the cathode for ewery one pH unit change in the anode. Washes 4 to

10 produced a ratio of 1.76, and washes 11-19 produced a 1.60 pH

change ratio, which were both significantly less in magnitude than

the ratio obtained in the first three washes, but were not

significantly different from each other. Therefore the very early

stages of KP2 2-70 (EtOH) EPS electrodialysis showed a greater

surge of cations to the cathode than the latter stages of ED in

relation to the surge of anions to the anode. In a total of 19

washes the overall ApH cathode/ApH anode value was determined to be

1.75.

During ED of KP2 2-70 (EtOH) EPS, the DH2O from both the anode

and the cathode chambers were pooled separately, lyophilized and

resuspended in concentrated volumes to test for certain divalent

161

Table 37. The pH Differential of the Anode and Cathode Chambers During Electrodialysis

ApH Cathode/ApH Anode^

2.16 + 0.15

1.76 + 0.06

1.60 +. 0.13

1.75 +0.22 (N=19)

The number of changes of DHpO in the electrodialysis chambers as electrodialysis proceeded.

ApH Cathode/ApH Anode; the ratio of the pH changes in the cathode chamber and the anode chamber with respect to the pH of dHÔ (pH = 6.10). ^

Wash^

1 - 3

4 - 10

11 - 19

Total

162

cations [calcium (Ca"*" ) and magnesium (Mg"^^)] in the cathode wash

and phosphate (P0^~^), which is the major anion in the defined

medium, in the anode wash. Table 38 shows the results of the

quantitation of these ions in the first three washes during ED. It

was found that the major divalent cation dissociated from the KP2

2-70 EPS during the first 3 washes of ED was Mg" ^ (1648.8 yM),

+2 while 829.0 yM of Ca was determined from the same cathode sample. _3

No PO. was found in the cathode sample but 779.3 yM was found in

the corresponding anode sample after 3 ED washes. Comparatively

+2 +2 small amounts of Mg and Ca were found in the anode also (22 and

51 yM, respectively). These data indicate that approximately 3.2

moles of these divalent cations are retrieved at the cathode for _3

ewery 1 mole of PO^ retrieved at the anode.

+2 +2 Table 39 shows the quantities of Mg and Ca found in yg per

mg of ECPS for both the KP2 2-70 (EtOH) and the KP2-0 (EtOH) EPS

before and after ED at 400V. In this table it can be seen that the

vast majority of Mg is lost from either EPS fraction after ED at

+2 400V, while approximately two-thirds of the Ca is lost during the

_3 same interval. Table 40 shows the effect of ED on the PO^

concentration for both KP2 2-70 (EtOH) and KP2-0 (EtOH) EPS.

Nearly 50% of the PO." was found to be extractable from both KP2

EPS fractions after ED at 400V. Further ED at lOOOV removed 18%

more PO.'^ from the KP2 2-70 (EtOH) EPS and 32% more PO^'^ from the

KP2-0 (EtOH) EPS. Figure 24 shows a histogram of the effect of ED

on the concentration of PO "' [P04"^] in mM, both in the sample

(open bars) and in the anode (stippled bars) for KP2-70 (EtOH) EPS

at various intervals of ED. The cross-hatched bars reflect the

163

"°'^^ 22.0+0.3 51.0+18.4 779.3+2.80

Cathode 1648.8+14.7 829.0+7.1 ND"

ND; none detected.

164

Table 39. Effect of ED on the Quantity of Divalent Cations Found in KP2 EPS

EPS

KP2-70 (EtOH) Before ED

After ED^

KP2-0 (EtOH) Before ED

After ED^

yg Mg'*" /mg ECPS

8.44+1.07 (N=4)^

0.10+0.05 (N=3)

8.50+2.76 (N=4)

0.54+0.37 (N=4)

yg Ca'^^/mq ECPS

1.70+0.15 (N=2)

0.62+0.09 (N=3)

1.14+0.05 (N=2)

0.38+0.00 (N=2)

ED proceeded at 400V until no marked surge in mA was noted over a 30 min period.

'N equals the number of determinations made from two separate preparations.

165

Table 40. Effect of ED on the Quantity of Phosphate Found in the KP2 EPS

EPi yg P04"^/mq ECPS (N=2)^

KP2 2-70 (EtOH) Before ED 36.81 + 1.31

400V ED^ 17.83 + 2.41

lOOOV ED^ 11.05 + 1.64

KP2-0 (EtOH) Before ED 2 9 . 2 8 ^ 2 . 8 2

400V ED 15.69 +_ 0.58

lOOOV ED 6.24 + 0.13

N equals the number of phosphate determinations made on the same sample.

400V ED; Electrodia lys is at 400V un t i l no marked surge in mA was noted over a 30 min period.

lOOOV ED; Electrodia lys is at lOOOV as in footnote b.

166

167

( V ) (l"^/îJLi)Sd03 i n ^ fO CVJ —

I I i I I

O

GO

1 r I "1—r ro OJ

(TV)(i^^)*'0<d

168

concentration of ECPS in mg/ml. The [PO^"^] in the sample before

ED (position A) was calculated to be 55.5 ymoles, but after ED at

400V (position B) the [PO^"^] in the sample diminished to 30.9

ymoles, while the [PO^"^] recovered in the anode equalled 19.9

ymoles (92% recovery of [PO^"^] from position A to position B).

After ED at lOOOV (position C), 15.4 ymoles of [PO^"^] remained in

the sample, while 9.1 ymoles were recovered in the anode (a

recovery of 79% from position B to position C). Finally, the ECPS

concentration in the sample remained essentially the same

throughout ED as seen in this figure. The results from this study

and the above data show that the EPS from two strains of KP2

contained appreciable amounts of both cations and anions even after

extensive dialysis, and the majority of the cations and anions

measured were extractable by the ED procedures, as shown both in

the concentrations of ions retained by the sample and the ions

recovered at the ED terminals.

The major difference between the two KP2 EPS was that the KP2

2-70 EPS was seen to readily dissociate into a LMW component after

ED while the majority of the KP2-0 EPS stayed in the HMW form even

after extensive ED. A study was then undertaken to see whether the

KP2-0 EPS could be dissociated into a lower molecular weight form

by the addition of sodium dodecyl sulfate (SDS). The KP2 (EtOH)

EPS was first electrodialyzed at 2000V to rid of all possible

hydrophilic associations of EPS due to inorganic cations and

anions. The electrodialyzed EPS was then subjected to boiling in

2% SDS for 5 min to disrupt noncovalent, hydrophobic interactions.

169

and applied to a Sepharose 2B column equilibrated with 0.01 M Tris,

pH 12, and 0.1% SDS. The elution profile obtained is displayed in

Figure 25, and fractions were monitored for both hexose and uronic

acid activity. Both the hexose and uronic acid profiles exhibited

two major fractions on S-2B; the HMW void volume fraction and a

second fraction immediately to the right of the HMW fraction, as

was seen in the profiles for KP2-0 EPS after ED alone (Figure 22).

A third and a fourth minor fraction can also be seen in this

profile and correspond proportionally to the same minor fractions

seen after ED alone (Figure 22). Although the two major fractions

produced on gel filtration after ED in the presence of SDS seem to

have been more clearly separated from one another than in the

profile obtained from ED alone, the differences are minor at best.

Thus, with the methods used in these studies to (1) remove all

dialysable ions that may contribute to the aggregation of the KP2-0

EPS and (2) to disrupt any noncovalent, hydrophobic interactions

between KP2-0 EPS polymers, it was found that no major shift to a

LMW form of ECPS could be achieved, although a slight shift to a

still rather high molecular weight fraction was evident.

The EPS from KPl-0 48h ethanol extracted supernatants was also

subjected to ED at 2000V to determine the effect of ED on the S-2B

elution profile of this material, which is shown in Figure 26.

Again, a relatively small quantity (1 mg total ECPS) was placed on

the column to avoid the coalescence of adjacent peaks (see Figure 7

for comparison). In both the profile before ED and after ED, a

number of symmetrical and rather evenly spaced peaks can be seen.

170

171

9 0^

7 0 -

E 50 C7»

3 Ul 8 30H X UJ X

20 30 40

FRACTION NUMBER

172

173

cr UJ 00

Z

< cr.

(J> 00 r^ c^ in

|iJU/&^3S0X3H

CM —

174

which do not necessarily correspond to the fractions seen in

earlier profiles (see Figure 7). Both profiles in Figure 26 have

retained a similar quantity of the HMW component of KPl-0 EPS,

though there appears to be a shift to the right of 4 fractions for

the ED profile. Most of the other fractions coincide as to

fraction number when comparing the two profiles in the LMW region,

except for the materials eluting at the far right of the profile

(between tube fractions 55 and 72). There were, however,

quantitative differences between many of the coincidental fractions

in the two profiles. For example, the fractions which eluted at

tube fractions 23, 27 and 34 for the ED profile were of greater

magnitude with regard to hexose than were the corresponding

fractions seen in the profile before ED, whereas the peaks of

hexose at tube fractions 38 and 43 were much larger in the profile

before ED than in the corresponding fractions after ED. There was

then an apparent shift of hexose from later to earlier fractions in

the LMW region as a result of ED, as well as an apparent shift of

the HMW fraction to a higher elution volume. There were also two

prominent peaks of hexose in the material after ED at tube

fractions 53 and 65 that were in contrast to the material before

ED.

Two different methods were utilized to ascertain the

contribution of hydrophobic interactions in the KPl-0 EPS, both of

which were based on the assumption that lipid groups are covalently

linked to either the ECPS or to the contaminating ELPS and that

these hydrophobic groups may tend toward micellar formation in

175

aqueous solutions (see Results, Section G). The first of these

methods was hydrolysis of KPl-0 EPS with 60% hydrofluoric acid (HF)

at 0 C, which has been shown to liberate acylglycerols and

diacylglycerols from the capsular polysaccharides of the Group B

meningococci, but leave the glycosidic linkages intact (37).

Figure 27 shows the elution profile on S-2B of KPl-0 EPS obtained

before and after HF .treatment, utilizing the EPS obtained from

dialyzed 48h supernatants. The profile after HF treatment is seen

here to be a more homogenous preparation of LMW EPS, with much of

the higher molecular weight hexose activity absent and a more

refined singular LMW fraction. The second method used to define

these putative hydrophobic interactions was that of treatment of

the KPl-0 EPS with 0.5 M NaOH (saponification). Figure 28 shows

the effect of saponification of KPl-0 (N) ECPS on the elution

profile obtained on S-2B. The profile obtained before

saponification contained a number of prominent peaks. After

saponification the vast majority of the hexose activity was found

to elute between tube fractions 32 and 50. Thus with HF treatment,

or with NaOH treatment of KPl-0 EPS, the fractions in the HMW

region were not as prominent and the LMW fraction apparently became

more homogeneous. It was also found that ED of the KPl-0 EPS,

followed by cetavlon fractionation, produces essentially the same

profile, with the LMW peak predominant (Figure 15). However, there

is a difference in the elution volume between the LMW fractions

from HF and NaOH treatment on the one hand and the LMW fraction in

Figure 15.

176

177

130

FRACTION NUMBER

178

179

FRACTION NUMBER

180

Further studies on the saponif icat ion of KPl-0 EPS, as well as

on various f ract ions of KPl and KP2 EPS from other sources,

revealed that f a t t y acids (FA) were being released from a l l of

these f rac t i ons , even a f te r the EPS was pre-extracted with organic

solvents to remove any noncovalently attached l i p i d s . The presence

of covalently l inked f a t t y acids was demonstrated by the i r release

from the EPS treated with 0.5 M NaOH, and the i r conversion to

v o l a t i l e methyl esters to be detected in gas l i qu id chromatography

(GLC). A known concentration of bacterial methyl ester standard

was run along with j<. pneumoniae EPS to quanti tate the to ta l amount

of f a t t y acid methyl ester (FAME) released by a known quantity of

ECPS (yg FAME/100 yg ECPS). Table 41 shows the results obtained

fo r the ethanol extracted ECPS from 5 strains of K pneumoniae. I t

can be seen from th is table that the KPl-0 variant produced ethanol

extractable material that contained s ign i f i can t l y more FA than did

i t s covar iant, the KPl-T s t r a i n . Among the KP2 (EtOH) EPS

f ract ions tested, the KP2 2-70 EPS and the KP2-0 EPS did not d i f f e r

s i gn i f i can t l y in the i r content of FA. However KP2 8052 EPS had

s i gn i f i can t l y less FA than did the EPS from KP2 2-70.

A s imi lar study was then performed on the HMW and LMW

fract ions from gel f i l t r a t i o n of both the KPl-0 and KPl-T EPS.

Various amounts of these f ract ions were saponi f ied, methylated and

assayed fo r the i r to ta l quanti ty of FAME on gas- l iqu id

chromatography. Table 42 shows the ra t ios of FAME to ECPS and to

ELPS in yg per lOOyg. The HMW samples from ei ther the KPl-0 or

181

Table 41. Quantitation of Fatty Acid Methyl Ester (FAME) Released from EPS after Saponification

FAME/ECPS^

6.13 + 1.45 (N=3)^

2.39 ± 1.61 (N=3) (p < 0.05)^

5.05 t 0.16 (N=3)

4.33 + 1.51 (N=2)

1.35 ± 0.54 (N=3) (p < 0.001)^

EPS; All fractions of EPS from the 5 strains tested were ethanol extracted from 48h supernatants.

FAME/ECPS: The amount of fatty acid methyl ester in yg found in 100 yg of ECPS as quantitated by the procedure of Blumencrantz et al. (7).

^N equal the number of trials performed, each from a different preparation of EPS.

Statistical analysis (Student's t test) between the mean values obtained for KPl-0 and KPl-T EPS.

^Statistical analysis (Student's t test) between the mean values obtained for KP2 2-70 and KP2 8052 EPS.

EPS^

KPl-

KPl-

KP2

KP2-

KP2

•0

•1

2-

-0

70

8052

182

Table 42. Quantitat ion of FAME Released from the EPS of KPl-0 and KPl-T Obtained from Gel F i l t r a t i o n on S-2B

Sample^ yg FAME/100 yg ECPS yg FAME/100 yg ELPS^

KPl-0 HMW 5.49 + 2.91 (N=4)^ 13.11 + 6.95 (N=4)

KPl-0 LMW 0 . 2 7 + 0 . 0 8 (N=4 ) 4.92 + 1.40 (N=4)

KPl-T HMW 3 . 7 7 + 1 . 3 3 (N=3) 8.91 +3.14 (N=3)

KPl-T LMW 0.18 + 0.02 (N=3) 6.25 + 0.54 (N=3)

^The HMW and LMW fract ions from KPl-0 and KPl-T obtained from gel f i l t r a t i o n on S-2B.

The number of yg of f a t t y acid methyl ester obtained per 100 yg of ECPS.

The number of yg of f a t t y acid methyl ester obtained per 100 yg of ELPS.

The number of determinations of FAME performed on two separate preparations.

183

KPl-T strain was seen to contain significantly more yg FAME than

the LMW samples per lOOyg ECPS (p < 0.02 and p < 0.005,

respectively). Essentially 20 to 25 times the amount of FAME per

lOOyg ECPS in the LMW fractions was found in the HMW fractions. In

contrast, the differences in the amount of FAME determined per

lOOyg of ELPS in the HMW versus the LMW fractions of these two

organisms was not nearly so marked. There were no significant

differences seen in the yg FAME/1 OOyg ELPS between any of the

samples, though there was a trend toward higher values being found

in the HMW component. These data suggest that the FAME content of

these samples was primarily related to the amount of ELPS present

in the fractions, rather than to the amount of ECPS.

The last structural study in this section involved growing the

KP2 2-70 strain in the same defined medium, with the exception that _o

the PO. buffer was eliminated and replaced by a 10 mM Hepes 3

buffer. The final molarity of PO. in the new medium (DMH) was

one-thousandth that of the old medium. Figure 29 shows the elution

profile of KP2 2-70 (EtOH) EPS obtained from growth in DMH on

BGA-150m. It can be seen in this profile that a HMW, void volume

fraction was present (tube fraction 18) even under conditions of

low [PO.''^] in the medium. The LMW fractions differed somewhat

from that seen for this EPS grown in the phosphate-buffered defined

medium (see Figure 19) in that there were two prominent fractions

(at tube fraction 24 and 33) rather than one LMW fraction. Thus, _3

even with the [PO. ] at 3 orders of magnitude less in

concentration in the defined medium, the HMW component was

produced.

184

185

FRACTION NUMBER

186

Gel Immunodiffusion Studies for Identification

and Quantitation of ECPS Produced

by K. pneumoniae

Ouchterlony double diffusion studies were performed in gels

for the various EPS fractions obtained from ion exchange for a

number of KPl and KP2 strains. 1 mg of lyophilized samples in one

ml DH2O were prepared for these studies. One precipitin line

appeared for each of the EPS fractions and each was shown to be

immunologically identical to one another. Therefore, the acid and

the neutral fractions were identical serologically within a given

strain and identical between strains of the same serotype. The

same patterns were seen for the acid and neutral EPS isolated from

three different strains of KP2.

When various fractions of KPl EPS were tested by RIE using

type-specific antiserum (AB), all of the fractions tested, except

for the HMW fractions, appeared to contain a mobile component that

reacted with AB with a peak as far as 2 to 3 or more cm above the

well, whereas the fractions, containing only HMW EPS at the same

ECPS concentration show some difficulties in entering the gel, and

show distortion of the materials that did enter. This was

especially true for the KPl HMW fraction, where it was seen that

only a small portion of EPS had entered the gel. Table 43 shows

the calibration of the standard curve obtained in RIE for 1:2

dilutions of both KPl HMW and KPl-O LMW EPS. The standard curve

for the LMW EPS produced a straight line with a correlation

coefficient(r) of 0.94, while the curve for the HMW fractions did

187

Table 43. Standard Curve for the RIE^ of KPl-0 HMW and KPl-0 LMW EPS

Sample

KPl-0 HMW

KPl-0 LMW

ECPS(yg/ml)

628.

314.

157,

78.

1081.

540.

270,

135,

.5

,3

.1

.6

,3

.6

.3

.2

Rocket

1.

1.

1.

0,

4.

3.

3.

2.

Hei

,10

.05

.12

.74

,25

,80

,00

.45

ght (cm)

Êlectrophoresis took place at 4-6V/cm for 3h in 0.2% agarose impregnated with anti-KPl antiserum. Ten yl of sample were added to each wel l .

188

not assume a straight line function (r=0.57). Moreover the

majority of the HMW EPS precipitated by AB at about 2mm above the

well in all four dilutions, and this EPS was believed not to have

entered the gel, but was rather a distortion of the sample well.

It was found that the ECPS present in the serum of rats

infected- with KPl-0 was in the HMW form, or of even a higher

molecular weight form, since an immunological precipitate was seen

around the upper periphery of the sample well after RIE, but no

ECPS appeared to enter the gel. After a number of unsuccessful

manipulations, the quantitation of KPl EPS in rat sera was

abandoned. However, RIE was used successfully to quantitate the

ECPS in rat sera infected with the KP2 2-70 strain. Out of the six

sera studied, only two were found to contain measurable quantities

of ECPS, and one of these sera were selected to perform

quantitative RIE. This serum sample was isolated from a rat which

was found to have nearly 1 x 10 bacteria per total lung

homogenate; the highest level of organisms found for all the rats

in this particular study. The other rat serum that had measurable 9

ECPS came from a rat that had approximately 5 x 10 bacteria per

total lung homogenate. The rats with the next 3 highest

concentrations of bacteria in their lungs (between 1.2 and 2.4 x

10 CFU per total lung homogenate) were found not to have

measurable ECPS in their sera by RIE.

Table 44 shows the RIE rocket heights obtained on the serum

from the rat that had the highest CFU titer of the KP2 2-70 strain

189

a Table 44. RIE of Standard Concentrations of KP2 2-70

ECPS and Serum from an Infected Rat

Sample

KP2 2-70 (A)^

KP2 2-70 HMW

ECPS (yq/ml)

756.9

378.5

189.2

1614.8

807.4

403.7

201.9

Rocket Height

2.25

1.55

0.45

3.30

2.27

2.00

0.70

Rat Serum 820" 2.50

Êlectrophoresis took place at 4-6V/cm for 3 h in 0.2% agarose impregnated with anti-KPl antiserum. Ten yl samples were used,

The acid f rac t ion of KP2 2-70 EPS from ion exchange chromatography.

^The HMW f rac t i on of KP2 2-70 EPS from gel f i l t r a t i o n chromatography.

Calculated from the standard curve.

190

in its lungs and the rocket heights produced from 2-fold dilutions

of KP2 2-70 (A) EPS, as well and the rocket heights obtained from

2-fold dilutions of KP2 2-70 HMW EPS. In Figure 30 a graph was

drawn to show the standard curve produced by the rocket heights in

cm against the quantity of ECPS that produced these rockets. Five

of the seven data points obtained conformed to a straight line,

which was then used to calculate the amount of ECPS in the rat

serum. The dotted lines in the graph depict the rocket height (2.5

cm) of the unknown sample and the calculated yg of ECPS. Thus this

rat serum was found to contain approximately 820 yg ECPS per ml.

Radial immunodiffusion was also performed to quantitate the

ECPS found in the sera of these rats infected with KP2 2-70. Table

45 shows the zone diameters of the serological precipitin reaction

obtained both for the known KP2 2-70 ECPS standards and for the

sera from the same rat as used above. By these methods the

concentration of ECPS in this rat serum was calculated to be 496.8

yg/ml. So by the two methods used in these studies, the level of

KP2 2-70 ECPS in the serum of a rat infected with high levels of

KP2 2-70 in its lungs, was found to be between 500 and 800 yg/ml.

However, in four of the five remaining sera obtained from animals

less moribund from the infection, no ECPS was measurable. There

was however, an apparent ring of precipitation around the periphery

of the wells and within the gel well itself for most of these sera.

Whether or not these precipitation reactions were immunological was

not readily demonstrable and, therefore, not determined.

191

192

0.5 1.0 1.5

Peak Height (cm)

2.0 2.5

193

Table 45. Quantitation of KP2 2-70 ECPS in the Serum of Infected Rats by Radial Immunodiffusion

Sample^

KP2 2-70 (A)"

yg ECPS

1914.6

957.3

478.65

239.33

Zone diameter

9.15 + 0.07 (r

8.15 + 0.49

7.00 + 0.14

6.30 + 0.42

Rat Serum 496.8 6.95 j 0.76 (N=4)^

^10 yl of sample were placed in the gel wells and incubated overnight for 24h.

Measurement of the immunologic precipitin zone around the wells at 24h in mm (N=2).

^The acid fraction of KP2 2-70 EPS.

Correlation coefficient for the standard curve.

^N equals the number of determinations made.

194

Survey of the Outer Membrane Proteins

of K. pneumoniae

The Triton insoluble outer membrane proteins (OMP) from

strains of both serotypes 1 and 2 were prepared and examined by SDS

polyacrylamide gel electrophoresis (SDS-PAGE) in gels containing

8.0 M urea. Figure 31 shows the Coomassie blue-stained OMP from

the KPl strains. Lanes A and B in Figure 35 are the OMP profiles

for KPl-0 and KPl-T, respectively. Lanes C and D are OMP profiles

from revertants in the KPl-O and KPl-T population, respectively

[Lane C is from a KPl-T like strain isolated from the KPl-0

population (KPl-Or), and Lane D is from a KPl-0 like strain

isolated from the KPl-T population (KPl-Tr)]. Lanes E and F are

OMP profiles from a large and a small capsular variant from the KPl

2-70 population, respectively. No significant differences could be

detected among the profiles of the six strains examined. Lane G

shows, for reference, the OMP profile of i- coli strain CS138,

which had been induced for the LamB protein (3). There are two

major protein bands from the profiles of the six KPl strains that

are apparently analogous to two major OMP of E_. coli namely the

OmpC and the OmpA proteins.

Figure 32 illustrates a similar study on the OMP from three

KP2 strains. Lanes A and D are profiles from KP2-0, with lane D

containing twice the volume of sample as Lane A. Lanes B and E are

OMP profiles from the KP2-T strain, and Lanes C and E are OMP

profiles from the KP2 2-70 strain, again with double the volume of

sample in the latter profiles. The OMP composition was

195

196

A B C D E F G

Lam B a & OmpC

OmpF ^ ^ H ^ O m p A

197

B rm •»•-•

D

i^,v •• MiM

198

4i# •-.îf'

j^sii-m-

.f-<^'^

Sft?".

199

fundamentally the same for these three strains also, and these

strains appear to possess the £. coli OmpC and OmpA protein

analogues, as did the KPl strains. A minor difference in the

profiles was observed in the high molecular weight region of the

KP2-T strain compared to that of the KP2-0 and KP2 2-70 strain,

namely that a distortion of the p.rotein bands was observed. This

distortion is best visualized in Lane E in Figure 32. It is known

that this high molecular weight region of the gel typically

contains the largest mass of LPS and these types of distortions are

usually attributed to contamination with LPS. It is then likely

that the KP2-T sample contained more LPS than did the other

samples.

CHAPTER IV

DISCUSSION

Classically the histologic and pathologic features of

Klebsiella pneumonia in humans include a massive, confluent lobar

consolidation consisting primarily of PMN, a voluminous edema, and

abscess formation with massive cavitation. Unfortunately, at least

for the development of animal models, there exists a wide

diversity of clinical manifestations of this disease process.

Several classifications have been proposed which distinguish

between an acute and chronic pneumonia pattern (29,34,43,70), a

primary versus secondary (suprainfecting) pneumonia (34,43,70), and

endogenous or epidemic sources of the organism (56,65). The

progression of the disease and its prognosis are primarily related

to age and predisposing variables (77). Since most Klebsiella

lobar pneumonias are seen in debilitated, middle-aged males (43,

65,77), it is difficult to establish a model that closely

approximates the human condition, especially when the large numbers

of predisposing factors are considered. The rats used in the

present study were healthy, young males more likely to effectively

combat the experimental infection than the aged and debilitated

human patient. This may explain the relatively low mortality and

the greater chronicity of infection seen in this rat model.

Nevertheless, in the present study, we were able to establish a rat

model which displayed the classical symptoms for J<. pneumoniae

pneumonia found in humans.

200

201

Berendt et al. (4) produced a bronchopneumonia in rats using

intranasal inoculations of 5 x 10^ l<. pneumoniae (strain A-D). The

authors eventually abandoned their rat model, concluding that the

squirrel monkey provided a more satisfactory experimental model for

lobar pneumonia. This non-human primate model allows measurements

of clinical signs that a rodent model does not afford, such as

fever, respiratory rate, and throat cultures (5). However, the

squirrel monkey pneumonia pattern mimics only the acute form of the

disease. The rapidity with which death occurred and the low

frequency of abscess formation severely restricted the utility of

this.model. Moreover, in their discussion (4), the authors noted

the economic, practical and statistical advantages of using a

rodent model to study this type of infectious process.

Sale and Wood (71) reported the production of a lobar

pneumonia in rats. They described a highly acute infection, with

the majority of their rats succumbing to their pneumonia by day 3

post-exposure. Again, their model simulated, at best, the acute

pattern of Klebsiella pneumonia in humans. The high mortality

encountered by these researchers was most likely due to the

administration of mucin into the lungs as an adjuvant, a procedure

that other authors felt could have "profound effects" on the

experimental animal (4).

Establishing the threshold of infection in the rat model at a

4 TBC of 5 X 10 was based on several meaningful observations.

First, it was at this approximate titer when morphological changes

in the lungs were seen. Berendt et al. (4) reported that lysozyme

202

levels did not become elevated until the number of bacteria in the

lungs reached four to five logs. Serum lysozyme is a convenient

assay for determining the extent of infection because it reflects

the appearance, frequency and severity of pyogranulomatous lesions

(11). Second, with only a few exceptions, the present study showed

an all or none response to the j<. pneumoniae challenge, with rat

lungs containing either well over this 5 x 10^ CFU threshold or

well below it. Third, changes in lung weight also supported a TBC 4

of 5 X 10 CFU as the threshold of infection. None of the rats

with a lung content of under 5 x 10 bacteria had any marked

elevation in lung weight.

In the present study we were able to produce a chronic lobar

pneumonia in rats without the aid of adjuvants. This model allows

for the colonization and infection of the rat lung by j<. pneumoniae

for at least 28 days. Previous attempts at establishing an

experimental paradigm for Klebsiella respiratory infections have

disregarded the infectivity or virulence of the bacterium. In this

report, two serotypes and variants within these serotypes were

examined for their ability to produce a chronic lobar pneumonia in

rats, and a comparision was made to their virulence in mouse

lethality tests. The results show how important it is to know the

pathogenic nature of the organism before model construction is

possible.

A relationship of both practical and economical significance

developed from the comparison of the ability of a given strain to

infect the lungs of rats and to be virulent in the mouse model. A

203

positive correlation of 0.95 was obtained by comparing the four

strains in which both a quantifiable LD^Q and ID^Q were determined.

In essence it is possible to gain insight into the ability of a

strain of l<. pneumoniae to produce an infection in the lungs of

rats by performing the simpler and cheaper mouse virulence assay.

If this relationship holds true for other strains of K,. pneumoniae,

performing a mouse virulence test, obtaining and LD^Q and then

adding 1.7 log-jQ units to this LD^Q calculation should approximate

the IDrQ for that particular strain.

It was seen that in 3 of the 4 strains tested, a large and a

small encapsulated variant could be isolated by various means. Two

of these strains revealed prominent variants without manipulation

while the third strain required passage through mice to-enrich for

a large encapsulated variant. These variations in capsule size

within a defined population have been described elsewhere for J<.

pneumoniae (26, 46), though little is known about the reasons for

the existence of these variations. The smaller capsule size has

been suggested (46) as the more stable structure, and it may be

that the large capsule (or what could be an unregulated synthesis

of CPS) is a mutation occurring at low frequency. Passage through

mice, or infection in general, gives a selective advantage to the

large encapsulated form and thereby enriches its presence in the

population. Since most of the strains in this study were isolated

from human disease, it is not surprising that the large

encapsulated variant was seen to be predominant in 2 of the 3

strains in question, both of which were lung isolates. The KPl

204

ATCC 8047 strain manifested two variants with respect to capsule

size, one which had a TD of 5.6 ym and the other a TD of 2.5 ym.

The KP2 ATCC 29011 strain exhibited one variant with a TD of 2.5 ym

and the other with a TD of 1.5 ym. The KP2 strain then produced a

large encapsulated variant which had the same TD as the small

variant in the KPl strain. The KPl CDC 2-70 strain, which was the

remaining strain shown to produce two capsular variants, had a TD

of 2.2 ym. It was seen, however, that a large encapsulated variant

appeared under India ink preparations at a frequency of

-4 approximately 10 , which had a TD as large as the KPl-0 strain

(5.6 ym on the average). After mouse passage, many of these large

encapsulated variants were isolated from the KPl 2-70 population.

It was seen then that the KPl strains in these studies generally

were capable of producing much larger capsules than the KP2

strains.

These variations in capsulze size within a given bacterial

population were considered to be isogenic variations for a number

of reasons. First of all, these variations were seen to arise

commonly from a single colony. Secondly, the biotypes of these

variants, determined on API-20E strips, were found to be identical,

as well as was their OMP profiles as seen in SDS PAGE gels. If

these variants in capsule size were really a mixture of

contaminating j<. pneumoniae, one would expect to see several

serotypes represented, since there are 72 known serotypes of this

organism. On the contrary, all variants possessed the same

serotype within a given population, and the ECPS isolated from

205

these variants exhibited immunological identity in gel

immunodiffusion. Therefore, it is likely these variants arose from

a single strain and are isogenic with respect to capsule

production.

The KP2 2-70 strain was not seen to produce two separate and

distinct variants in capsule size even after extensive

manipulation, which included both mouse passage and low speed

differential centrifugation. Both manipulations originally were

able to separate large and small encapsulated variants, but this

distinction was again lost on subsequent plating. An increase in

capsule size for this strain was also seen during culture,

exhibiting a TD of 2.5 ym at 18h growth and a TD of 3.3 ym at 48h

growth in defined medium. It was also noted that the KP2 2-70

strain produced rather small (1 mm), nonmucoid colonies at 18h on

TSA, and contained organisms possessing a rather homogenous TD of

2.5 ym. On the other hand, plating these same organisms on

nutrient agar produced a relatively large (3-4 mm) and mucoid

colony type at 18h, with organisms possessing an assortment of TD

ranging from 2.5 to 5.0 ym. This phenomenon was not seen for any

of the other strains in these studies. The KP2 2-70 strain then

has the peculiar ability to regulate its capsule diameter with

changing environmental influences. This is further substantiated

by the changing kinetics regarding the production of ECPS by this

organism in defined medium. At up to 18h of culture, comparatively

small quantities of ECPS were produced, even though the stationary

phase of growth had begun approximately 10 hours earlier. After

206

18h the rate of production of ECPS increased tremendously. It is

possible that there were certain components in the medium that

regulated the production of capsule and, that by 18h of culture,

these components were at low enough concentrations to permit

deregulation of CPS production. This is also supported by the fact

that nutrient agar is of a much lower ionic strength than TSA and

seems to allow for this deregulation of CPS production much

earlier in culture. By 48h the colonies of KP2 2-70 on TSA became

mucoid and large, and exhibited the large assortment of capsular

types seen on nutrient agar at 18h.

Although the identity of these regulating components for CPS

production by KP2 2-70 have not been determined, it is believed

that metal ions, such as magnesium and calcium, may play a role.

It has been demonstrated that these ions are important for the

stability and organization of the outer membranes of gram negative

bacteria (16). Divalent cations are thought to play a role both in

reduction of change repulsion between highly anionic polysaccharide

molecules and are thought to bridge adjacent LPS molecules and to

link LPS with membrane proteins (16). Chelating agents have been

shown to effect the release of up to 50% of the LPS from the whole

cells (49). The CPS of J<. pneumoniae has been shown in this study

to be complexed with divalent cations. When these metal ions

attain a low concentration in the medium, much of the CPS (and LPS)

may begin to loosen from the outer membrane and exude into the

growth medium. The loss of these polymers from the cell may

trigger a constitutive response of CPS and LPS production to

207

replace these lost surface polymers. The end result manifests as a

larger complex of CPS surrounding the bacterium, coupled with a

higher rate of escape of CPS into the medium.

Although the KP2 2-70 strain seems to be regulated and,

therefore, conservative in CPS production, other strains seem to

regulate CPS production by producing low levels of CPS and are not

inducible during the same conditions that affect the KP2 2-70

strain. Alternatively these other strains may undergo low level

mutations for high capsule production, which are then selected for

under the right conditions, such as during infection. Both the

variations in phenotypic expression of the KP2 2-70 strain and the

genotypic mutational variations of other K_. pneumoniae for capsule

production may then serve the same requirement for these cells,

that is the need to produce extraordinary quantities of both

capsule and slime (ECPS) under certain conditions.

It is plausible that this high rate of CPS production plays

some functional role for K_. pneumoniae, otherwise the metabolic

expense would seem too costly. In light of the evidence that much

of this CPS is found in the extracellular milieu, it is suggested

that the ECPS also functions for the benefit of the organism in

some manner. In a more static environment, as opposed to cultures

shaking at high speeds, the ECPS may exist as a large extension of

the bacterium and may surround huge numbers of bacteria in the form

of microcolonies. These static conditions are probably found in

the lungs of infected animals. If the capsule is envisioned as a

large chelating complex for divalent cations, these large

208

extensions of the capsule (microcolonies) may not only provide

better protection against phagocytosis, but may also serve as an

ion escalator for certain cations needed for growth. Since it has

been shown by Fukutome et al. (32) and others (31, 45), that the

primary host defense against I<. pneumoniae involves the production

of opsonic AB specific to the capsular type, it is reasonable to

view these putative microcolonies of J<. pneumoniae as being more

resistant to phagocytosis than an individual bacterium, even in the

presence of AB to the capsule. These arguments then consider

freely soluble ECPS to be, in part, an artifact of rotating

cultures, and call for a distinction in capsular organization

between the in vivo and in vitro situation. It has been shown by

Pollack (66) and in the present study, however, that circulating

cell free CPS can be demonstrated in the sera of infected animals,

especially in cases of severe infection.

In the OPA studies performed, we were able to support the

hypothesis that PMN will not phagocytose l<. pneumoniae in the

absence of AB. In view of the theories posited in the Introduction

as to the possible functional roles of the ECPS in virulence, the

OPA studies have apparently shed some light as to which of these

functions may apply. It was shown .that the KPl 2-70 strain was

able to thrive significantly better in the OPA in the presence of

AB to the type 1 capsule, when homologous KPl EPS was present

rather than when twice the quantity of heterologous KP2 EPS was

present. Although the KP2 EPS at these dosages imparted much more

viscosity to the OPA, it did not help the KPl 2-70 organism survive

209

any better than control trials, which contained no EPS. In the

reverse experiment KP2 EPS allowed the KP2-0 organism to grow

significantly above that of either KPl EPS treated or control

assays. Therefore, the notion that EPS influences the ability of

PMN to phagocytose K_. pneumoniae due to its viscosity is not

supported in these studies. There also appears not to be any

immediate biological effect of the EPS on the WBC. However, the

type-specific, homologous EPS allowed these organisms to survive

the effects of 90% serum, WBC and AB to the capsule significantly

better than both controls and heterologous EPS treated trials.

Therefore, it is most likely an AB neutralization phenomenon taking

place. It was also found in the OPA that the large encapsulated

KPl-0 variant is not better protected from phagocytosis under these

conditions than is the small-capsuled KPl-T variant. Whether the

same level of protection holds true in vivo is an entirely

different question.

It was determined in these studies that the production of

ECPS, ELPS and the size of the capsule were all positively related

to the ability of an organism to infect the lungs of rats or to

kill mice. This relationship was greatest for the cell associated

capsule size parameter, but was also seen to correlate highly with

ECPS and ELPS production. These correlations were, however, much

greater when examined within, rather than among serotypes. It was

also apparent that even within a given serotype, a strict

correlation between total polysaccharide production and virulence

does not hold true, nor does the positive relationship always hold

210

between the capsule size and ECPS production. A primary example of

deviations from the rule involves the KPl 2-70 strain, which

produces more ECPS per cell in defined medium than does the KPl-T

strain, but harbors a smaller capsule and is at least two orders of

magnitude less virulent in the mouse model. Although KPl 2-70 was

not seen to be sensitive to 90% serum and not subject to in vitro

phagocytosis by PMN in the absence of AB, it may be that the

antiphagocytic properties of the relatively small capsule of this

strain are more readily overcome by host defenses in vivo. Yet

this organism was seen to produce more ECPS than the large

encapsulated KPl-T strain. It may be that it is not possible to

compare unrelated strains, regardless of a common serotype, as to

their capsule to ECPS ratio. A strict relationship between these

two parameters of CPS production may only apply within isogenic

pairs, and may be reflective of other surface components (LPS, OMP)

that are intimately involved in the stabilizing forces of the outer

membrane of K_. pneumoniae. It was seen, however, that there were

no remarkable differences in the OMP profiles of these organisms.

This strict capsule to ECPS ratio was observed, however, for the 4

variants obtained from one KP2 strain in a study by Ehrenworth and

Baer (28). Duguid and Wilkinson (21) also demonstrated that, with

varying culture conditions, the capsular and slime (EPS)

polysaccharide generally increased or decreased together, though

not in strict proportion. Thus a general relationship does exist

within serotype as to capsule and ECPS production but only in the

isogenic situation does it appear to strictly apply.

211

The capsule size and the quantity of ECPS produced by a given

strain are considered to be manifestations of the same phenomenon.

It appears to be the actual rate of total CPS production which

distinguishes a small from a large capsular variant, and a high

from a low ECPS producer. • The capsule is viewed as a dynamic

intermediate between CPS production and release into the growth

medium. A high CPS producer may harbor a large cap-sule due to the

rate of CPS production being faster than the rate of release into

the medium.

If the association between the capsule size and ECPS

production holds true, then the question still remains as to

whether the capsule or the ECPS or both are responsible for the

virulence of j<. pneumoniae. It is well established that the

presence of a capsule is essential for this organism to be

pathogenic (28, 32, 45). A relationship between the rate of

production of CPS and the degree of pathogencity has been shown

with J<. pneumoniae (28) as well as with Streptococcus pneumoniae

(55, 74). These studies suggest that the antiphagocytic properties

of the capsule increase as the cell-associated capsular material

increases in volume. There also exist acapsular variants of l<.

pneumoniae that produce copious amounts of ECPS but are avirulent

(28) as was also seen with the KP2 8052 strain utilized in this

study (L-D-Q > 7.3 x 10 ). Although it is generally true that when

a large capsule is being produced, large quantities of ECPS are

being released, the reverse phenomenon is not always true, and high

ECPS producers exist which are acapsular. These latter variants

212

These lines of evidence support the role of the capsule as a

necessary structure for virulence, but do not rule out the ECPS as

a virulence factor when a capsule is present. It was found in the

present study that as low as 25.5 yg of KPl-0 (A) ECPS could

enhance the virulence of KPl-T significantly over control values as

could approximately 400 yg of KP2 2-70 ECPS. It was also seen that

in one of the rats infected with the KP2 2-70 strain, between 500

and 800 yg of ECPS were detected per ml of serum. These values are

also in line with the amount of KP2 EPS used by Batshon et al. (2)

to suppress the immune response to this antigen. KPl EPS was shown

also to be present, but not quantifiable, in the serum of rats

infected with KPl-0 or KPl-T, since the EPS was presumably in a

high molecular weight form and was not mobile in RIE or RID assays

even in 0.2% agarose. The KPl EPS was also shown to suppress AB

production to the capsule at a dosage of between 10 and 100 yg per

mouse (58). These values are also in line with the virulence

enhancement data obtained in the present study. It is therefore

conceivable that the capsular and slime polysaccharide work

together to enhance the pathogenicity of j<. pneumoniae; the capsule

providing protection against nonspecific immune defense postures of

the host, and the ECPS possibly enhancing the antiphagocytic

potential of the organism by neutralizing AB, by suppressing the

antibody response via a tolerance phenomenon, and by affecting the

proper functioning of the macrophage. The effect shown by Yokochi

et al. (85, 86) on the macrophage could also explain the tolerance

phenomenon in that the poorly functioning macrophage may not be

213

processing the ECPS to present it to B-lymphocytes for AB formation

to occur. Unfortunately due to the contaminating LPS in the above

preparations, the notion that the LPS may be responsible for these

effects can not be ruled out. Macrophages have been shown to be

particularly sensitive in vivo to LPS administration (81). In

fact, from the results on virulence enhancement described in the

present work, the LPS may be the virulence component in the studies

from these other investigators.

In regard to the associations that hold the CPS to the cell

wall, it is believed that the LPS is intimately involved for a

number of reasons. First it was seen that both the ECPS and ELPS

are released together in a consistent ratio among all strains

within a given serotype. This ratio of ELPS to ECPS was constant

within serotype and the KPl strains were found to release nearly

twice the quantity of ELPS to ECPS as did the KP2 strains. These

stable proportions of ELPS to ECPS within serotype may then be a

reflection of the actual surface composition of these polymers.

Electron micrographs of these bacteria support this idea, since the

extracellular capsular materials were seen to slough off of these

bacteria in large masses (Fig. 5). If the KPl strains actually

have twice the quantity of LPS on their surface as the KP2 strains

this may explain why they were seen to produce much larger

capsules, since it is possible that the CPS is anchored to the

outer membrane in association with LPS. An alternative explanation

for the larger capsules seen among the KPl isolates has to do with

the extra negative charge per repeat unit of the CPS polymer

imparted by the pyruvyl linkage (30). This extra anionic component

214

may increase the adhesiveness of the CPS to itself, thereby

allowing for greater aggregation.

Secondly, the ECPS and ELPS were seen to retain their

association and to co-purify together, and were shown in gel

filtration and in gel immunoelectrophoresis studies to exist in a

highly aggregated form. In gel filtration both the KPl and the KP2

EPS aggregates eluted at the void volume of the column, so no

molecular weight differences could be discerned by these methods.

However, in the RIE and RID studies, the HMW aggregates from the

KPl strains were noticeably larger than the HMW EPS from the KP2

strains. The HMW aggregates from the KPl strains also contained at

least twice the quantity of ELPS as did the HMW KP2 EPS. It is

believed that the ELPS functions to hold together these aggregates

and is responsible for the larger size of the KPl HMW EPS.

Finally, it was shown that the EPS from the acapsular KP2 8052

strain contained significantly less fatty acids than the KP2 2-70

EPS. As the quantity of FA was considered to be related to the

ELPS content of the samples tested, it is apparent then that the

KP2 8052 EPS contains significantly less ELPS than the KP2 2-70 EPS

or, alternatively, that the KP2 8052 ELPS contains significantly

less FA per polymer than does the KP2 2-70 ELPS. The latter

conclusion is supported by the fact that the KDO content in yg per

mg of ECPS was quite similar. In any case the KP2 8052 strain

apparently has one fourth to one-fifth the FA content in its

extracellular polysaccharides as do the polysaccharides of the

encapsulated KP2 strains. If the extracellular polysaccharides are

215

truly reflective of the surface composition on the cell wall, it is

suggestive that acapsular, mucoid strains are the result of

anchored LPS polymers on their surfaces. In summary, LPS may be

responsible for holding the apsule together on the cell surface due

to the intimate association seen between CPS and LPS polymers, due

to the constant ratios seen within serotype in the extracellular

material, and in light of the data obtained from an acapsular

variant.

The ECPS:ELPS complex is felt to be held together not only by

ionic interactions in the form of salt bridges between anionic

components of the polymers, but also by hydrophobic interactions

imparted by the lipid A of the LPS and possibly by a lipid terminus

on the CPS polymer, as was shown for £. coli and meningococcal ECPS

(37). These aggregates were seen to dissociate for the KP2 2-70

EPS after ED only, which argues that the predominant association

between these polymers is that of salt bridging. However, for the

KPl-0 EPS the aggregates did not dissociate after ED alone and were

found to require an additional cetavlon extraction step or

treatment with either 0.5 M NaOH or 60% HF at 0°C, both of which

are known to remove covalently linked fatty acid esters. Cetavlon

is a cationic detergent and is known to separate LPS from the CPS

of K_. pneumoniae (72). Cetavlon alone, however, was not able to

disrupt the KPl-0 EPS aggregates as seen on gel filtration.

Therefore it is believed that both hydrophilic and hydrophobic

interactions are dominant in the HMW EPS of KPl-O. The KP2-0

strain, on the other hand, produced a HMW aggregate that was not

216

seen to dissociate to any appreciable extent after both ED and

cetavlon extraction or after both ED and boiling in SDS. The

possibility of covalent linkages in the KP2-0 EPS is suggested. It

has been demonstrated, for at least the KPl-0 and the KP2 2-70

EPS, that ED serves as a powerful tool for the purification of

these polymers.

The LMW forms of EPS from either the KPl or KP2 strains were

shown to contain 50% or less ELPS than their respective HMW forms.

We do not believe that the ECPS and ELPS in the LMW regions of gel

filtration are linked together in any regular fashion, since these

LMW samples exhibit a great deal of variability in their ELPS

content. It was possible to remove virtually all ELPS from the LMW

EPS of KPl-0 by ED and cetavlon extraction. The LMW ECPS and ELPS

are probably not efficiently separable on the gel filtration

resins that were used in the earlier studies (S-2B and BGA-150m).

The utilization of a P-300 column was shown to produce better

separation of ECPS and ELPS (Fig. 16).

The OMP profiles of K.. pneumoniae were unremarkable as to

their differences between high and low CPS producers. This is in

contrast to was seen in the OMP profiles of different isolates of

£. coli, where marked variation from strain to strain is the rule

(44). Also, in E_. col i, there was demonstrated a new OMP that was

associated only with encapsulated organisms (63). The OMP profiles

of K_. pneumoniae did not show these differences, but did reveal an

interesting phenomenon that distinguished high from low ECPS

producers, especially in the KP2 group. A greater distortion of

217

OMP bands in the HMW region of the gels were seen in the OMP

profiles of low ECPS producers as compared to high producers. This

distortion is usually attributable to high LPS levels in the

Triton-insoluble membrane preparations. If, for instance, KP2-T

actually has much more LPS on its surface, it may be that LPS

replaces the CPS in low producers or nonproducers of CPS.

Therefore CPS and LPS may be freely exchanged for one another on

the cell surface in order to fill any gaps. If the CPS contains a

lipid terminus, it may utilize the same space in the outer membrane

as does LPS for anchorage. Yet if the lipid terminus is absent in

J<. pneumoniae, it may be anchored through electrophilic (and

possibly covalent) interactions with LPS. Electrodialysis of whole

organisms at 2000V did not reveal any differences in the capsule

size of the KPl-0 strain and allowed for the release of less than

1% of the ECPS that was obtained in the cultural supernatant.

Therefore a hydrophobic linkage is more than likely present.

In light of the evidence obtained on the quality and quantity

of undialyzable inorganic ions attached to the KP2 2-70 EPS, a

hypothetical structure was drawn to explain these associations, as

can be seen in Figure 33. Two ECPS strands are here held together

by magnesium phosphate bridges, thus forming a highly stable

complex through structural complementarity. The negatively charged

uronic acid moiety in the repeating tetrasaccharide of the ECPS

polymer may orient at precise angles from the next uronic acid in

the same chain, thereby allowing for three dimensional

electrophilic interactions. Uronic acids lining up in the same

218

219

0

I-MG-O-P-O-MG

0 6'

0

•MG-0-P-O-MG l l

OUTER

ME.MBRANE

220

plane may form a ladder-like structure with another juxtaposed ECPS

strand, due to these putative divalent cation phosphate bridges.

In such an arrangement, a free negative charge resonates between

the two unoccupied oxygens in the phosphate ion. This structure

may then function as an ion escalator to transport positive ions

from the outer periphery of the capsule to the outer membrane,

where a convenient ion acceptor or porin may be located. The

capsule may then serve at least two important functions for j<.

pneumoniae; one being as an antiphagocytic structure, and the other

for the provision of metal ions from the external milieu.

The major problem that existed in the present study and in

many of the studies cited on ECPS as a virulence factor involved

the extent of purification of the ECPS and, in particular, the

amount of LPS contained in the EPS fractions. Batshon (2) worked

with 3 different preparations of KP2 EPS and showed that two of

these preparations were lethal for a significant proportion of mice

injected with a dose of 250 yg. The third preparation, which had

similar values of hexuronic acid as the two others, appeared to be

completely nontoxic even in doses of 2500 yg. The authors

concluded that there was a possibility of endotoxin contamination

in two of the 3 preparations. It has been shown in the present

study that doses of KP2 2-70 (A) up to 800 yg per mouse were not

lethal, although the mice became quite ill and listless for a 24 to

48 h period at these doses. Much of the early work performed by

Kato's group (58) spoke of the powerful adjuvanticity of the CPS of

K. pneumoniae. These publications were then followed by a focus on

221

the strong adjuvanticity of the LPS of l<. pneumoniae (87) without

providing the reader many clues as to what role CPS actually

played. It was most likely that it was the LPS contamination in

the ECPS preparations that produced these biological effects. It

has long been known that LPS is a virulence factor in gram-negative

bacteria (6), and has been shown to cause a transient leukopenia,

(6, 14, 81) in the host. The macrophage is particularly sensitive

to LPS, which was seen to both inhibit macrophage migration and to

cause severe morphologic damage to these cells (40, 81). A

significant reduction in the number of mature macrophages in the

peritoneum of mice was demonstrated after a dosage intravenously of

0.1 to 20 yg of Salmonella LPS (81). The response to LPS is also

known to be biphasic, with a transient leukopenia in the first 3h

after administration followed by an augmentation of phagocytic

activity by 48h. Furthermore, approximately 10 yg of E . coli LPS

was shown to increase the infectivity of pathogenic staphylococci

in rabbit skin, with an absence of leukocytic infiltration into the

focal area of infection (14). Therefore, there is evidence in the

literature to support the correlation of ELPS with virulence

enhancement seen in the present study.

It has been seen here that without appropriate precautions,

the ELPS will co-purify with the ECPS. Gotschlich et al. (37) were

able to separate meningococcal LPS and CPS by the use of

zwitterionic detergents. The method used in the present study to

remove ELPS was that of electrodialysing the ECPS in order to

222

remove the large quantities of small molecular weight ions which

contaminate these preparations. The ECPS is in many ways a magnet

for both positive and negative ions in that it produces salt

bridges presumably between adjacent uronic acid (or pyruvyl)

groups. Not only is the ED step removing these ions, but it is

also allowing for the dissociation of ECPS and ELPS with the

subsequent cetavlon extraction. ELPS is most likely associated

with the ECPS hydrophilically through the phosphate groups on LPS

covalently linked to lipid A. A divalent cation could provide a «3

salt bridge between the PO- on the LPS and a negatively charged

group on the ECPS. There is still the possibility of the bond

between ELPS and ECPS being covalent, and labile under acidic

conditions with a rise in temperature above 60°C, as can happen in

ED without a proper cooling apparatus to keep the temperature down.

Whether the bond is covalent or noncovalent, this procedure has

allowed for separation of the two polysaccharide components to a

far greater degree than without ED, at least for the KPl-0 ECPS.

Having finally separated the KPl-0 ECPS from the ELPS and

protein, virulence studies were then performed using these

preparations. It was found that our purest ECPS preparation

(containing <1% ELPS) in doses of up to 400 yg per mouse, did not

significantly enhance the virulence of KPl-T over control values.

Again, with further purification of the ECPS, the A(logiQ LD^Q)/mg

ECPS decreased accordingly. Both protein and ELPS, as well as

other contaminating materials (i.e., salts), were removed from the

ECPS during these purification steps. However, it was found that

223

the ability of these materials to decrease the LD^Q of KPl-T was

directly correlated to the amounts of ELPS present in the samples,

with approximately 11.6 yg of KPl ELPS or 8.3 yg of KP2 ELPS needed

to decrease the LD^Q of KPl-T by one log^Q unit, regardless of the

quantity of protein or ECPS present. Therefore, the ECPS of K,.

pneumoniae is believed not to have significant virulence

enhancement properties by itself, but the ELPS appears to be a

powerful virulence enhancer at low dosages. In light of the fact

that LPS from j<. pneumoniae has been shown to possess much more

powerful adjuvant activity than E_. coli LPS (58, 87), there is good

reason to believe that the ELPS of j<. pneumoniae acts as an

important virulence factor during infection. This is also

supported by the apparent loss of virulence enhancement seen after

mild saponification of the polysaccharide samples, which has been

shown to decrease the biologic activity of endotoxin (69).

In conclusion, the polysaccharides of K_. pneumoniae have shown

to be highly complex structures that aid the organism in a number

of ways during pathogenesis. The capsule plays an antiphagocytic

role, as does the ECPS, at least in AB neutralization, while the

ELPS seems to be responsible for certain biological effects that

decrease the host's resistance to infection. The ECPS was seen

also to act as a magnet for a variety of inorganic ions. This avid

attraction for ions may serve to provide necessary nutrients in the

highly competitive host environment and may also affect complement

activation and phagocytosis, both of which need divalent cations

for proper functioning. It was also seen that KPl-0 ELPS alone did

224

not significantly enhance the virulence of KPl-T at 9.4 yg/mouse

(Appendix 21) whereas the KPl-0 (N) ECPS significantly enhanced

KPl-T virulence while containing only 8.0 yg ELPS per mouse dosage.

It is suggested that the ECPS may augment the biologic activity of

ELPS in these virulence studies. This may in fact be why the 1<.

pneumoniae LPS is considered to be a more potent biologic effector

molecule than the E. coli CPS.

LITERATURE CITED

1. Baltimore, R.S., D.L. Kasper, and J. Vecchitto. 1979. Mouse

protection test for group B streptococcus type III. J.

Infect. Dis. 140:81-88.

2. Batshon, B.A., H. Baer, and M.F. Schaffer. 1963. Immunologic

paralysis in mice by Klebsiella pneumoniae type 2

polysaccharide. J. Immunol. 90^:121-126.

3. Beher, M.G., C A . Schnaitman, and A.P. Pugsley. 1980. Major

heat-modifiable outer membrane protein in Gram-negative

bacteria: Comparison with the OmpA protein of Escherichia

col i. J. Bacterid. 143:906-913.

4. Berendt, R.F., G.G. Long, F.B. Abeles, P.G. Canonico, M.R.

Elwell, and M.C. Powanda. 1977. Pathogenesis of respiratory

Klebsiella pneumoniae infection in rats: Bacteriological and

histological findings and metabolic alterations. Infect.

Immun. J_5:586-593.

5. Berendt, R.F., G.L. Knutsen, and M.C. Powanda. 1978.

Nonhuman primate model for the study of respiratory Klebsiella

pneumoniae infection. Infect. Immun. 22:275-281.

6. Biozzi, G., B. Benacerraf, and B.N. Halpern. 1955. The

effect of Salmonella typhi and its endotoxin on the phagocytic

activity of the reticuloendothelial system in mice. Brit. J.

Exp. Path. 36:226-235.

7. Bligh, E.G. and W.J. Dyer. 1959. A rapid method of total

lipid extraction and purification. Can. J. Biochem. Physiol.

37.:911-917.

225

226

8. Blumenkrantz, N. and G. Asboe-Hansen. 1973. New method for

quantitative determination of uronic acids. Anal. Biochem.

5^:484-489.

9. Brown, M.J. and J.N. Lester. 1982. Role of bacterial

extracellular polymers in metal uptake in pure bacterial

culture and activated sludge. Water Res. 16^:1539-1548.

10. Bukantz, S.C, P.F. DeGara, and J.G.M. Bullowa. 1942.

Capsular polysaccharide in the blood of patients with

pneumococcal pneumonia. Arch. Int. Med. 69^:191-212.

11. Canonico, P.G., M.C. Powanda, G.L. Cockerell, and J.B. Moe.

1975. Relationship of serum 6-glucuronidase and lysozyme to

pathogenesis of tularemia in immune and nonimmune rats.

Infect. Inmun. J_2:42-47.

12. Cassone, A. and E. Garaci. 1977. The capsular network of

Klebsiella pneumoniae. Can. J. Microbiol. 23:684-689.

13. Chen, P.S., T.Y. Toribara, and H. Warner. 1956. Microdeter-

mination of phosphorus. Analyt. Chem. 28:1756-1758.

14. Conti, C.R., L.E. Cluff, and E.P. Scheder. 1961. Studies on

the pathogenesis of Staphylococcal infection. IV. The effect

of bacterial endotoxin. J. Exp. Med. 113:845-859.

15. Coonrod, J.D. 1981. Pulmonary opsonins in Klebsiella

pneumoniae pneumonia in rats. Infect. Immun. 33:533-539.

16. Coughlin, R.T., S. Tonsager, and E.J. McGroarty. 1983. Quan

titation of metal cations bound to membranes and extracted

lipopolysaccharide of Escherichia coli. Biochem. 22:2002-2007.

227

17. Coughlin, R.T., A. Haug, and E.T. McGroarty. 1983. Physical

properties of defined lipopolysaccharide salts. Biochem.

22_:2007-2013.

18. Diedrich, D.L., A.O. Sunmers, and C A . Schnaitman. 1977.

Outer membrane proteins of Escherichia coli. V. Evidence that

protein 1 and bacteriophage-directed protein 2 are different

polypeptides. J. Bacterid. 131:598-607.

19. Dochez, A.R. and O.T. Avery. 1917. The elaboration of

specific soluble substance by pneumococcus during growth. J.

Exper. Med. 26^:487-493.

20. Duguid, J.P. 1951. The demonstration of bacterial capsules

and slime. J. Path. Bact. 63^:673-685.

21. Duguid, J.P. and J.F. Wilkinson. 1953. The influence of

cultural conditions on polysaccharide production by Aerobacter

aerogenes. J. Gen. Microbiol. 9^:174-189.

22. Durham, D.L., S.J. Mattingly, T.I. Doran, T.W. Milligan, and

D.C Straus. 1981. Correlation between the production of

extracellular substances by type III group B streptococcal

strains and virulence in a mouse model. Infect. Immun.

34.:448-454.

23. Dutton, G.G.S., K.L. Mackie, and A.V. Savage. 1980.

Preparation of oligosaccharides by bacteriophage degradation

of polysaccharides from Klebsiella serotypes K21 and K32.

Cabohydrate Res. 8^:161-170.

24. Edmondson, A.S. and E.M. Cooke. 1979. The production of

antisera to the Klebsiella capsular antigens. J. Appl.

Bacterid. 46:579-584.

228

25. Edwards, M.S., A. Nicholson-Weller, C J . Baker, and D.L.

Kasper. 1980. The role of specific antibody in alternative

complement pathway-mediated opsonophagocytosis of type III

Group B streptococcus. J. Exp. Med. 151:1275-1287.

26. Edwards, P.R. and M.A. Fife. 1952. Capsule types of

Klebsiella. J. Infect. Dis. 91:92-104.

27. Edwards, P.R. and M.A. Fife. 1955. Studies on the

Klebsiella-Aerobacter group of bacteria. J. Bacteriol.

7^:382-390.

28. Ehrenworth, L. and H. Baer. 1956. The pathogenicity of

Klebsiella pneumoniae for mice: The relationship to the

quantity and rate of production of type-specific capsular

polysaccharide. J. Bacteriol. 72:713-717.

29. Erasmus, L.D. 1956. Friedlander bacillus infection of the

lung. Quart. J. Med. 25^:507-521.

30. Erbing, C , L. Kenne, B. Lindberg, and J. Lonngren. 1976.

Structural studies of the capsular polysaccharide from

Klebsiella type 1. Carbohyd. Res. 5£:115-120.

31. Fournier, J.M., C Jolivet-Reynaud, M.M. Riottot, and H.

Jouin. 1981. Murine immunoprotective activity of Klebsiella

pneumoniae cell surface preparations: Comparative study with

ribosomal preparations. Infect. Immun. 32:420-426.

32. Fukutome, T., M. Mitsuyama, K. Takeya, and K. Nomoto. 1980.

Importance of antiserum and phagocytic cells in the protection

of mice against infection by Klebsiella pneumoniae. J. Gen.

Microbiol. 119:225-229.

229

33. Galanos, C and 0. Luderitz. 1975. Electrodialysis of

lipopolysaccharides and their conversion to uniform salt

forms. Eur. J. Biochem. 54:603-610.

34. Gill, R.J. 1951. Treatment of Friedlander's pneumonia.

Amer. J. Med. Sci. 221:5-9.

35. Gormus, B.J., R.W. Wheat, and J.F. Porter. 1971. Occurrence

of pyruvic acid in capsular polysaccharides from various

Klebsiella species. J. Bacteriol. 107:150-154.

36. Gormus, B.J. and R.W. Wheat. 1971. Polysaccharides of type 6

Klebsiella. J. Bacteriol. 108:1304-1309.

37. Gotschlich, E.C, B.A. Eraser, 0. Nishimura, J.B. Robbins, and

T.Y. Liu. 1981. Lipid on capsular polysaccharides of gram

negative bacteria. J. Biol. Chem. 256:8915-8921.

38. Graybill, J.R., L.W. Marshall, P. Charache, C K . Wallace, and

V.B. Melvin. 1973. Nosocomial pneumonia: A continuing major

problem. Amer. Rev. Resp. Dis. 108:1130-1140.

39. Heidelberger, M., W.F. Dudman, and W. Nimmich. 1970.

Immunochemical relationships of certain capsular

polysaccharides of Klebsiella, pneumococci, and Rhizobia. J.

Immunol. 104:1321-1328.

40. Heilman, D.H. 1965. The selective toxicity of endotoxin for

phagocytic cells of the reticuloendothelial system. Int.

Arch. Allergy. 26^:63-79.

41. Hoffman, N.R. and F.S. Preston. 1968. Friedlander's

pneumonia. Dis. Chest. 53:481-486.

230

42. Hoffman, T.A. and E.A. Edwards. 1972. Group-specific

polysaccharide antigen and humoral antibody response in

disease due to Neisseria meningitidis. J. Infect. Dis.

126:636-644.

43. Jackson, G.G. 1980. Host factors in gram-negative bacillary

pneumonia, p. 87-105. Ui_ J. P. Thys, J. Klastersky and E.

Yourassowsky (ed.). Aerobic gram-negative bonchopneumonias.

Pergamon Press, Brussels.

44. Jann, B. and K. Jann. 1980. SDS polyacrylamide gel

electrophoresis patterns of the outer membrane proteins from

E.- coli strains of different pathogenic origin. FEMS

Microbiol Letters. 7_:19-22.

45. Julianelle, L.A. 1926. A biological classification of

Encapsulatus pneumoniae (Friedlander's bacillus). J. Exper.

Med. 4i: 113-128.

46. Kaufmann, F. 1949. On the serology of the Klebsiella group.

Acta. Path. et. Micro. Scand. 26_:381-406.

47. King, B.F. and B.J. Wilkinson. 1981. Binding of human

immunoglobulin G to protein A in encapsulated Staphylococcus

aureus. Infect. Immun. 33:666-672.

48. Kozel, T.R. and J. Cazin, Jr. 1971. Nonencapsulated variant

of Cryptococcus neoformans. Infect. Immun. 3_(2) :287-294.

49. Leive, L. 1968. Studies on the permeability change produced

in coliform bacteria by ethylenediamine tetraacetate. J.

Biol. Chem. 243:2373-2380.

231

50. Levin. J. 1968. Clottable protein in Limulus: Its

localization and kinetics of its coagulation by endotoxin.

Thromb. Diath. Haemorrph. 21:186-197.

51. Loewus, F.A. 1952. Improvement in anthrone method for

determination of carbohydrates. Anal. Chem. 21:219-224.

52. Loos, M., B. Wellek, R. Thesen, and W. Opferkuch. 1978.

Antibody-independent interaction of the first component of

complement with gram-negative bacteria. Infect. Immun.

22.:5-9.

53. Lorian, V. and B. Topf. 1972. Microbiology of nosocomial

infections. Arch. Intern. Med. 130:104-110.

54. Lowry, O.H., N.J. Rosebrough, A.L. Farr, and R.J. Randall.

1951. Protein measurement with the Folin phenol reagent. J.

Biol. Chem. 193:265-275.

55. MacLeod, C M . and M.R. Krauss. 1950. Relation of virulence

of pneumococcal strains for mice to the quantity of capsular

polysaccharide formed in vitro. J. Exper. Med. 92:1-9.

56. Martin, W.J., P.K.W. Yu, and J.A. Washington II. 1971.

Epidemiologic significance of Klebsiella pneumoniae. Mayo

Clin. Proc. 46.:785-793.

57. Mizuta, K., M. Ohta, M. Mori, T. Hasegawa, I. Nakashima, and

N. Kato. 1983. Virulence for mice of Klebsiella strains

belonging to the 01 group: Relationship to their capsular (K)

types. Infect. Immun. 40:56-61.

232

58. Nakashima, I . , T. Kobayashi, and N. Kato. 1971. Al terat ions

in the antibody response to bovine serum albumin by capsular

polysaccharide of Klebsiel la pneumoniae. J . Immunol.

107:1112-1121.

59. Nimmich, W. 1968. Zur Isol ierung and qua l i ta t iven

Bausteinanalyse der K-Antigene von Klebsiel len. Z. Med.

Microb io l , u. Immunol. 154:117-131.

60. Osborn, M.J. 1963. Studies on the gram-negative ce l l w a l l ,

I . Evidence for the role of 2-keto-3-deoxyoctonate in the

l ipopolysaccharide of Salmonella typhimurium. Proc. Nat l .

Acad. Sc i . (U.S.A.). 50.:499-506.

61. Osserman, E.F. and D.P. Lawlor. 1966. Serum and urinary

lysozyme (muramidase) in monocytic and monomyelocytic

leukemia. J . Exper. Med. 124:921-952.

62. Ouchterlony, 0. 1968. Dif fusion in gel methods for

immunological analysis, p. 30. ln_ P. Kallos and B. Waksman

(ed . ) . Handbook of immunodiffusion and immunoelectrophoresis.

Ann Arbor Sc ien t i f i c Publ icat ions, Inc . , Ann Arbor, Mich.

63. Paakkanen, J . , E.C. Gotschl ich, and H. Makela. 1979. Protein

K: a new major outer membrane protein found in encapsulated

Escherichia c o l i . J . Bacter io l . 139:835-841.

64. Park, S.H., J . Eriksen, and S.D. Henriksen. 1967. Structure

of the capsular polysaccharide of Klebsiel la pneumoniae type 2

(B). Acta. Path. e t . m i c r o b i d . Scandinav. 69^:431-436.

65. Pierce, A.K. and J.P. Sanford. 1974. Aerobic gram-negative

bac i l l a ry pneumonias. Amer. Rev. Resp. Dis. 110:647-658.

233

66. Pollack, M. 1976. Significance of circulating capsular

antigen in Klebisella infections. Infect. Immun. 1_3:1543-1548.

67. Pugsley, A.P. and C A . Schnaitman. 1978. Identification of

three genes controlling production of new outer membrane pore

proteins in Escherichia coli K-12. J. Bacteriol.

135:1118-1129.

68. Reed, L.J. and H. Muench. 1938. A simple method of

estimating fifty per cent endpoints. Amer. J. Hyg.

2^:493-497.

69. Ribi, E., R.L. Anacker, K. Fukushi, W.T. Haskins, M. Landy,

and K.C Milner. 1964. Relationship of chemical composition

to biological activity. Iji Bacterial Endotoxins. M. Landy

and W. Braun (eds.). Rutgers, New Brunswick, N.J. pp.

427-436.

70. Ritvo, M. and F. Martin. 1949. The clinical and roentgen

manifestations of pneumonia due to Bacillus mucosus capsulatus

(primary Friedlander pneumonia). Amer. J. Roent. 62:211-222.

71. Sale, L. and W.B. Wood. 1947. Studies on the mechanism of

recovery in pneumonia due to Friedlander's bacillus. J.

Exper. Med. 86.:239-248.

72. Scott, J.E. 1965. Fractionation by precipitation with

quarternary ammonium salts. In Methods in Carbohydrate

Chemistry. R.L. Whistler (ed.). Academic Press, N.Y. pp.

38-44.

73. Snedecor, G.W. and W.G. Cochran. 1967. Statistical Methods,

pp. 102-106. The Iowa State University Press, Ames, Iowa.

234

74. Stark, O.K. 1959. Mechanisms of hypersensitivity, p. 519.

J.H. Shaffer, G.A. LoGrippo and M.W. Chase (ed.). Little,

Brown and Company, Boston.

75. Stein, R.A., V. Slawson, and J.F. Mead. 1976. Gas-liquid

chromatography of fatty acids and derivatives. Jji Lipid

Chromatographic Analysis, G.V. Marinetti (ed^.), 2nd ed. New

York, Dekker. pp 874-887.

76. Sutherland, I.W. 1970. Formate, a new component of bacterial

exopdysaccharides. Nature 228:280.

77. Thys, J.P., A. Decoster, and J. Klastersky. 1980. Clinical

manifestations of aerobic gram-negative bronchopneumonias. Iri

Aerobic gram negative bronchopneumonias. J.P. Thys, J.

Klastersky and E. Yourassowsky (ed.), Pergamon Press,

Brussells. pp 53-85.

78. Van Dijk, W . C , H.A. Verbrugh, M.E. van der Tol, R. Peters,

and J. Verhoef. 1979. Role of Escherichia coli K capsular

antigens during complement activation, C3 fixation, and

opsonization. Infect. Immun. 25:603-609.

79. Verbrugh, H.A., W.C. van Dijk, M.E. van Erne, R. Peters, P.K.

Peterson, and J. Verhoef. 1979. Quantitation of the third

component of human complement attached to the surface of

opsonized bacteria: Opsonin-deficient sera and phagocytosis-

resistant strains. Infect. Immun. 26:808-812.

80. Weeke, B. 1972. A manual of quantitative immunoelectro

phoresis methods and applications. 2lL N.A. Axelson, J. Kroll,

and B. Weeke (eds.). Scand. J. Immunol. 2_(supp. nol. 1).

Universitets Forlaget, Oslo, pp 15-46.

235

81. Wiener, E., A. Beck, and M. Shilo. 1965. Effect of bacterial

LPS on mouse peritoneal leukocytes. Lab. Invest. 11:475-487.

82. Wilkinson, B.J., S.P. Sisson, Y. Kim, and P.K. Peterson.

1979. Localization of the third component of complement on

the cell wall of encapsulated Staphylococcus aureus M:

Implications for the mechanism of resistance to phagocytosis.

Infect. Immun. 26^:1159-1163.

83. Wilkinson, J.F., J.P. Duguid, and P.N. Edmunds. 1954. The

distribution of polysaccharide production in Aerobacter and

Escherichia strains and its relation to antigenic character.

J. Gen. Microbiol. lL:59-72.

84. Winkelstein, J.A., A.S. Abramovitz, and A. Tomasz. 1980.

Activation of C3 via the alternative complement pathway

results in fixation of C3b to the pneumococcal cell wall. J.

Immunol. 124:2502-2506.

85. Yokochi, T., I. Nakashima, and N. Kato. 1977. Effect of

capsular polysaccharide of Klebsiella pneumoniae on the

differentiation and functional capacity of macrophages

cultured in vitro. Microbiol. Immunol. 21:601-610.

86. Yokochi, T., I. Nakashima, and N. Kato. 1979. Further

studies on generation of macrophages in in vitro cultures of

mouse spleen cells and its inhibition by the capsular

polysaccharide of Klebsiella pneumoniae. Microbiol. Immunol.

23:487-499.

236

87. Yokochi, T., I. Nakaskima, F. Nagase, N. Kato, M. Ohta, Y.

Fujii, K. Mizoguchi, K. Isobe, and M. Saito. 1982. Further

studies on the polysaccharide of J<. pneumoniae possessing

strong adjuvanticity. III. Augmentation of the antibody

response to subcutaneously injected sheep red blood cells by

the adjuvant polysaccharide. Microbiol. Immunol. 26:843-852.

APPENDICES

Page

1. CPS Structural Repeat Unit for K_. pneumoniae

Serotypes 1 and 2 239

2. ECPS Production 240

3. Serum Sensitivity 241

4. OPA for KPl-0 242

5. OPA for KPl-T 243

6. OPA for KPl CDC 2-70 244

7. OPA for KP2-0 245

8. OPA for KP2-T 246

9. OPA for KP2 2-70 247

10. Effect of Addition of EPS to the OPA 248

11. The Extracellular Products Found in the Ethanol

Fractionated Supernatants of K_. pneumoniae 249

12. The Extracellular Products Found in KPl and KP2 EPS

After Purification 250

13. Effect of KPl EPS on KPl-T Virulence in the Mouse Model.. 251

14. Effect of Cetavalon Extracted EPS on KPl-T and KP2-0

Vi rul ence in the Mouse Model 252

15. Effect of KP2 EPS on KPl-T Virulence in the Mouse Model.. 253

16. Effect of KPl or KP2 EPS on the Virulence of KP2-0


17. Effect of Electrodialysis (ED) on the Virulence

Enhancement of KPl-T by KP2 2-70 EPS


237

238

Page


Enhancement of KPl-T by KP2-0 EPS

i n the Mouse Model 256


of KPl-T by KPl-0 (N) EPS in the Mouse Model 257


of KP2-0 by KPl-0 (N) EPS in the Mouse Model 258

21. Effect of an Alternative Purification of KPl-0 EPS

on the Virulence Enhancement of KPl-T


22. Quantitation of FAME Released from the EPS of KPl-0

and KPl-T Obtained from Gel Filtration 260

239

1. CPS Structural Repeat Unit for K. pneumoniae

Serotypes 1 and 2

Serotype 1^

-4)-e-D-GlcAp-(l^)-a-L-Fuc (1.3)-6-D-Glc ( U / \ H p 2 3 V / c

/ \ H3C C02H

Serotype 2

a-D-GlcA P

j^

^3)-a-D-Glc -(l->4)-3-D-Man (l->4)-e-D-Glc (1-

a Erbing et al., 1976 (30).

^Park et al., 1967 (64).

240

2. ECPS Production

yg/ml^ yg/ml yg/ml X + S.D.

KPl-0 18h 82.00 96.58 15.78 64.79+43.06 24h 110.94 135.67 51.32 99 .31^43 .36 36h 126.84 168.43 81.96 125.74+43.25 48h 204.78 231.18 204.16 213.37^15.42

KPl-T 18h 23.95 25.63 24.79+ 1.19 24h 37.38 43.25 40.32 + 4.15 36h 59.52 61.23 60.38 + 1.21 48h 92.12 97.00 94.56 + 3.45

KPl 2-70 18h 10.65 22.34 29.24 20.74 + 9.40 24h 153.71 33.44 66.79 84 .65+62.09 36h - - - - 379.60 379.60 48h 293.87 489.96 416.91 400.25+99.10

KP2-0 18h 309.50 576.60 443.05 +188.87 24h 571.59 720.85 646.22+105.54 36h 906.24 907.78 907.01 + 1.09 48h 1035.68 1205.49 1120.59 +120.07

KP2-T 18h 1.89 24h 2.35 4.81 36h 48h 4.46 4.69

KP2 2-70 18h 31.14 16.04 24h 164.66 154.71 36h 602.19 691.11 48h 861.28 796.37

^yg/ml of ECPS calculated from uronic acid data on dialyzed supernatants.

1.89 3.58 +

4.58 +_

23.59 + 159.69 + 646.65 + 828.83 +

1.74

0.16

10.68 7.04

62.88 45.90

241

3. Serum Sensitivity

Strain

KPl-0

KPl-T

KPl 2-70

KP2-0

KP2-T

KP2 2-70

log^QCFU^

0 min

5.49 + 0.15

8.30 + 0.04

8.21 + 0.01

8.62 + 0.07

8.43 + 0.06

8.67 + 0.03

log^QCFU^

60 min

6.46 + 0.15

8.84 + 0.09

8.71 + 0.01

9.05 1 0.08

8.35 + 0.18

9.31 + 0.05

^Log phage organisms were washed 3 times and resuspended in PBS in

various concentrations, and added to 9 parts normal rabbit serum.

" The number of colony forming units (CFU) in log.|Q units at 0 time

(N=3).

^The CFU after 60 min incubation in 90% normal rabbit serum (N=2).

242

4. OPA for KPl-0

Time 0 Log^Q CFU = 5.28 + 0 . 1 4 (N=4)

60 min

A)

B)

c)

D)

E)

AB^

+

+

-

-

-

c^

+

-

+

-

-

WBC^

+

+

+

+

-

log^Q CFU (N=2)

4.15 1 0.16

4.74 + 0.06

6.13 + 0.02

6.11 + 0.39

6.13 + 0.12

Alog^QCFU^

-1.13

-0.54

+0.85

+0.83

+0.95

ÂB, Type specific antibody.

C, complement source (normal rabbit serum).

^WBC, white blood cells (human peripheral leukocytes).

Net growth from 0 to 60 min in log-jQ units.

243

5. OPA for KPl-T

Time 0 Log^Q CFU = 6.16 10.16 (N=2)

As in Appendix 4.

As in Appendix 4.

60 min

AB^ C^

A) + +

B) +

c) - +

D) -

E) -

Âs in Appendix 4.

As in Appendix 4.

WBC^

+

+

+

+

-

log^QCFU (N=4)

6.79 + 0.18

6.79 + 0.05

6.90 1 0.07

7.22 + 0.28

7.06 + 0.24

Alog^Q CFU^

+0.63

+0.63

+0.74

. +1.06

+0.90

244

6. OPA for KPl CDC 2-70

Time 0 Log^Q CFU = 6.75 1 0 . 0 9 (N=4)

Time 60 min

AB*

A) +

B) +

0 -D) -

E) -

c"

+

-

+

-

WBC

+

+

+

+

^

60 min

log^QCFU (N=2)

7.34 + 0.15

7.31 + 0.01

7.71 + 0.08

7-70 ± 0 . 0 1

7.73 + 0.18

Alog^QCFU^

+ 0.59

+ 0.56

+ 0.96

+ 0.95

+ 0.98

Âs in Appendix 4.

As in Appendix 4.

Âs in Appendix 4.

Âs in Appendix 4.

245

OPA for KP2-0

Time 0 Log^QCFU = 7.15 + 0.21 (N=2)

60 min

A)

B)

0 D)

E)

*As

"AS

=As

''AS

Âs

AB* c"

+ +

+

+

-

-

in Appendix 4.

in Appendix 4.

in Appendix 4.

in Appendix 4.

in Appendix 4.

WBC^

+

+

+

+

-

log^QCFU (N=2)

7.99 1 0.02

7.87 + 0.12

8.08 1 0.12

8.04 1 0,06

8.17 + 0.04

Alog^QCFU

+0.83

+0.72

+0.93

+0.89

+1.02

246

8. OPA for KP2-T

0 min 60 min

A)

B)

C)

D)

E)

AB^

+

+

-

-

-

c"

+

-

+

-

-

WBC^

+

+

+

+

-

log^QCFU (N=2)

7.66 + 0.07

7.26 + 0.09

6.96 + 0.24

7.19 + 0.08

7.00 + 0.24

log^QCFU (N=2)

7.50 1 0.25

7.28 1 0.19

7.19 1 0.27

7.36 + 0.18

7.24 + 0.18

Alog^QCFU

-0.16

+0.02

+0.23

+0.17

+0.24

Âs in Appendix 4.

As in Appendix 4.

Âs in Appendix 4.

As in Appendix 4.

247

9. OPA for KP2 2-70

Time 0 Lo9in

AB*

A) +

B) +

0 -D) -

El -

CFU =

c"

+

-

+

-

.

6.57 1 0.02

WBC^

+

+

+

+

^

(N=2)

60 min

log^QCFU (N=2)

6.70 + 0.07

6.72 + 0.04

7.88 + 0.02

7.56 + 0.14

7.51 + 0.21

Alog^QCFU^

+ 0.13

+ 0.15

+ 1.31

+ 0.99

+ 0.94

Âs in Appendix 4.

As in Appendix 4.

Âs in Appendix 4.

As in Appendix 4.

248

10. Effect of Addition of EPS to the OPA

Time 0 Log^QCFU = 6.75 +0.09 (N=4) for KPl 2-70

Time 0 Log^QCFU = 6.33 +0.15 (N=4) for KP2-0

Time 60 min

60 min

Strain AB^ C* WBC^ EPS log^^CFU (N=2) Alog^QCFU^

KPl 2-70 + - + - 7.31 + 0.01 0.56

+ - + KPl^ 7.94 + 0.07 1.19

+ - + KP2^ 7.42 + 0.11 0.67

KP2-0 + - + - 6.74 + 0.12 0.412

+ - + KPl^ 6.68 1 0 . 1 6 0.350

+ - + KP2' 7.01 + 0.12 0.675

a As i

b

n Appendix 4.

n Appendix 4.

n Appendix 4.

n Appendix 4.

n Table 16: footnote c.

n Table 16: footnote d.

Âs in Table 16: footnote e.

' As in Table 16: footnote f.

As i

Âs i

Âs i

Âs

Âs

249

11. The Extracellular Products Found in the Ethanol

Fractionated Supernatants of J<. pneumoniae

Strain

KPl-0

KPl-T

KP2-0

KP2 2- 70

ECPS'' (N=3)

522.2124.6

425.6115.8

782.4110.0

840.6+11.5

ELPS'' (N=2)

64.514:5

52.111.2

25.412.1

23.6+0.12

Protein' ' (N=2)

39.211.5

70.115.3

9.610.6

16.9+0.2

^Values for ECPS, ELPS and protein expressed in yg per mg dry

weight.

^N equals the number of determinations performed on separate samples.

250

12. The Extracellular Products Found in KPl and KP2 EPS

After Purification^

Strain Fraction ECPS^ ELPS^ Protein^

KPl-0 HMW 296.2131.1(N=3) 79.4ll6.3(N=3) 86.612.1(N=2)

KPl-0 LMW 529.9118.1(N=3) 37.41 5.8(N=3) 30.0l0.2(N=2)

KP2 2-70 HMW 807.4ll6.7(N=3) 28.ll 0.0(N=2) 48.7l0.7(N=2)

KP2 2-70 LMW 838.7ll0.4(N=3) 15.7l 0.9(N=2) 33.7l0.2(N=2)

^Purification procedures included ethanol extraction, DEAE-Sephacel

and gel filtration chromatography on S-2B.

^Values of ECPS, ELPS and protein are expressed in yg per mg dry

weight.

http://28.ll

251

13. E f f e c t o f KPl EPS on KPl-T V i r u l ence

i n the Mouse Model

EPS yg ECPS/mouse log^Q LD^Q ^ ^ ° 9 I O ^^50^

KPl-0 (N)^ 40.3 3.3410.64(N=3)^ -1 .4410.41(p<00.001)^

KPl-0 (A )^ 25.5 3.7510.56(N=3) - 1 . 02 l 0 .34 (p<0 .01 )

KP l -T (N)^ 121.7 3 .67 l0 .12(N=3) - 1 . l l l 0 . 1 2 ( p < 0 . 0 0 5 )

KPl -O(A)^ 56.9 4.3610.26(N=3) -0 .4210.26(p<0.02)

KPl-0 HMW(N)^ 62.8 4.0410.41(N=3) - 0 . 74 l 0 .03 (p<0 .05 )

KPl -0 LMW(N)^ 108.0 4.3610.20(N=4) -0 .4210.20(p<0.10)

KPl-T HMW(N)^ 96.9 3.90l0.52N=2) -0 .8810.34(p<0.05)

KPl-T LMW(N)^ 149.0 4.63l0.37(N=3) -0.15l0.04(p<0.80)

Control PBS 4.78l0.44(N-10)

Âlog.Q LD^Q; Change in the log^Q LD^Q over PBS control values.

^Neutral EPS fraction from DEAE-Sephacel.

Âcid EPS fraction from DEAE-Sephacel.

^High molecular weight fraction from neutral EPS applied to

Sepharose 2B gel filtration.

^Low molecular weight fraction as in footnote d.

'N equals the number of LDnn trials performed.

^Statistical analysis (Student's t test) between experimental (EPS

treated) and control (PBS treated) trials.

252

14. Effect of Cetavlon Extracted EPS on KPl-T

and KP2-0 Virulence in the Mouse Model

Strain

Injected

KPl-T

KP2-0

EPS

KPl-O(CET)^

Control

KPl-O(CET)^

Control

yg ECPS/mouse log-jQ LD^Q ^"'oQio ^^50

200.0

200.0

PBS

200.0

200.0

PBS

3.23 -1.55

2.97 -1.81(p<0.0001)^

4.7810.44(N=10)

4.87 -1.14

5.12 -0.89(p<0.05)^

6.01+0.21(N=2)^

Âlog-|Q LDCQ; Change in the log^Q LD^Q compared to control values.

^KPl-0 (CET); cetavlon fractionated supernatant of cultures of

KPl-0 grown in defined medium for 48h at 37 C

^Statistical analysis (Student's t test) of the comparison of the

means of the log^Q LD^Q values for experimental (ECPS) and control

(PBS) trials.

^N equals the number of LD^Q trials performed.

15. Effect of KP2 EPS on KPl-T

Virulence in the Mouse Model

253

EPS

KP2 2-70 EtOH^

KP2 2-70 (A)^

KP2 2-70 (A)

KP2 2-70 (A)

KP2-0 (A)^

ug ECPS/mouse

336.2

101.5

204.7

454.1

428.6

PBS

logô LDgQ (N=2)^ Alog^Q LD 50

2.97 1 1.27

4.67 1 0.49

4.07 1 0.35

2.87 1 0.77

3.91 + 1.19

-1.81(p<0.005)'

-0.11(p<0.80)

-0.71(p<0.01)

-1.91(p<0.001)

-0.87(p<0.10)

4.78 + 0.44 (N=10)

Âlog,Q LDCQ; change in the log-jQ LD^Q over PBS control values.

Êthanol extracted EPS from 48h cultural supernatants.

Âcid EPS fraction from DEAE-Sephacel.

^Statistical analysis (Student's t test) between experimental (EPS

treated) and control (PBS treated) trials.

^N equals the number of determinations performed from two separate

trials.

254

16. Effect of KPl or KP2 EPS on the Virulence

of KP2-0 in the Mouse Model

EPS yg ECPS/mouse log.|Q LD^Q ^^°9IO ^^50

KP2 2-70 (EtOH)^

KPl-0 {Hf

432

467

467

536

40.

40.

3

3

5.42

5.16

5.42

4.80

5.86

5.86

-0.60

-0.86

-0.60

-1.22

-0.16

-0.16

Control PBS 6.0210.22 (N=2)

Âlog,Q LDCQ; change in the log^Q LD^Q over PBS control values.

' KP2 2-70 (EtOH); ethanol extracted EPS from 48h supernatants of

KP2 2-70 grown in defined medium at 37 C

^KPl-0 (N); neutral EPS fraction from DEAE-Sephacel.

255

17. Effect of Electrodialysis (ED) on the Virulence Enhancement

of KPl-T by KP2 2-70 EPS in the Mouse Model

KP2 2-70 EPS^ yg ECPS/mouse LD^Q log^Q LD^Q ^^^9^Q LD^Q"^

Before ED

After ED^

470

470

470

380

380

409

432

1.17x10^

1.17x10^

4.07x10^

4.11x10^

2.56x10"^

7.30x10'^

2.75x10^

2.07

2.07

2.61

3.61

3.41

3.86

3.44

-2.71

-2.71

-2.17

-1.17

-1.37

-0.92

-1.34

Control PBS 6.03x10^ 4.78l0.44(N-10)^

^KP2 2-70 EPS obtained by ethanol extraction of 48h supernatants of

cultures grown in defined medium at 37 C

Âlog^Q LD^Q; the change in the log^Q LD^Q from the PBS control

data.

Êlectrodialysis proceeded at lOOOV until no marked increase in mA

occurred over a 30 min period.

Âs in Table 31.

256


Enhancement of KPl-T by KP2-0 EPS in the Mouse Model

KP2-0 EPS^ yg ECPS/mouse LD^Q log^Q LD^Q ^^^9-^Q LD^Q^

Before ED 441 1.45x10^ 4.16 -0.62

470 2.31x10^ 4.36 -0.42

After ED^ 392 2.08x10^ 3.32 -1.46

392 2.08x10^ 3.32 -1.46

421 1.30x10^ 3.11 -1.67

Control PBS 6.03x10^ 4.78 (N=10)^

^KP2-0 EPS; obtained by ethanol extraction of 48h supernatants of

cultures grown in defined medium at 37 C

Âlog^Q LD^Q; the change in the log^Q LD^Q from the PBS control

data.

Êlectrodialysis proceeded at lOOOV until no marked increase in mA

occurred over a 30 min period.

^N refers to the number of LD^Q trials performed.

257


of KPl-T by KPl-0 (N) EPS in the Mouse Model

Strain

KPl-T

EPS

KPl-O(N)^

KPl-O(N)^

(Saponified)

KPl-O(N)^

(Saponified)

yg ECPS/mouse

40.3

200.0

200.0

Control PBS

log^Q LD^Q Alog^Q LD^Q

4.23

3.97

3.34 + 0.64 -1.28 + 0.41

-0.55

-0.81

4.78 + 0.44 (N=10)

Âlog-|Q LDCQ; the change in the log.|Q LD^Q from PBS controls.

^KPl-O(N); the neutral fraction of ethanol fractionated EPS placed

on DEAE-Sephacel (N=3).

^KPl-O(N) (Saponified); neutral EPS as in footnote b placed in 0.5

M NaOH overnight at room temperature and dialyzed 3 times against

8LDH2O in the cold.

258


of KP2-0 by KPl-0 (N) EPS in the Mouse Model

Strain EPS yg ECPS/mouse log^Q LD^Q ^^^^-[Q LD^Q^

KP2-0 KPl-O(N)'^

KPl-O(N)

KPl-O(N)^

(Saponified)

KPl-O(N)^

(Saponified)

Control

40.3

40.3

200.0

200.0

PBS

5.86 -0.16

5.86 -0.16

5.87 -0.15

5.87 -0.15

6.02+0.22(N=2) d

Âlog,Q LDCQ; the change in the log^Q LD^Q from PBS controls.

^ KPl-O(N); the neutral fraction of ethanol fractionated EPS placed

on DEAE-Sephacel.

^KPl-O(N) (Saponified); neutral EPS as in footnote b placed in 0.5

M NaOH overnight at room temperature and dialyzed 3 times against

BLDHÔ in the cold.

^N equals the number of LD^Q trials performed.

259

21. Effect of an Al ternat ive Pur i f i ca t ion of

KPl-0 EPS on the Virulence Enhancement

of KPl-T in The Mouse Model

Sample^

KPl-0 Fr II

KPl-0 Fr II

KPl-0 Fr II

S-2B (ED)

KPl-0 Fr II

S-2B (ED)

KPl-0 Fr II

S-2B (ED)

KPl-0 Fr III

KPl-0 Fr III

ECPS/mouse°

400.6

400.6

402.7

201.3

100.7

ND^

ND

ELPS/mouse^

15.5

15.5

3.4

1.7

0.9

9.4

9.4

^^910^^50

3.58

3.83

4.58

4.83

4.58

4.33

4.33

Alog^Q LD5Q

-1.20

-0.95

-0.20

+0.05

-0.20

-0.45

-0.45

Control PBS PBS 4.7810.44 (N=10)^

Âs i n Table 23.

^The quantity of ECPS in yg co-administered with serial dilutions

of the KPl-T strain IP into mice.

^The quantity of ELPS in yg as in footnote b.

^The change in the log^Q LD^Q with respect to controls.

^ND; none detected.

^N equals the number of separate trials performed.

260

22. Quantitation of FAME Released from the EPS of

KPl-0 and KPl-T Obtained from Gel F i l t r a t i o n

Sample FAME ECPS ELPS^

KPl-0 HMW 24.22112.84 (N=4)^ 440.76137.97 184.7312.36

KPl-0 LMW 34.0419.72 (N=4) 12769.951596.59 692.56114.18

KPl-T HMW 59.84121.11 (N=3) 1587.14155.22 671.6813.55

KPl-T LMW 20.1211.73 (N=3) 11516.921867.26 321.71114.18

^The HMW and LMW fract ions from both KPl-0 and KPl-T EPS obtained

from gel f i l t r a t i o n .

The quanti ty of f a t t y acid metyl esters in yg obtained from

saponi f icat ion of EPS.

^The quant i ty of ECPS in yg (N=3).

* The quant i ty of ELPS in yg (N=2).

^N equals the number of determinations performed on two separate

preparations.

Documents

THE ELABORATION OF EXTRACELLULAR CAPSULAR …