Retrospective Theses and Dissertations Iowa State University Capstones, Theses andDissertations

1992

The role of tumor necrosis factor and secondarybacteria in the pathogenesis of swine dysenteryStuart Kent NibbelinkIowa State University

Follow this and additional works at: https://lib.dr.iastate.edu/rtd

Part of the Animal Sciences Commons, and the Veterinary Medicine Commons

This Dissertation is brought to you for free and open access by the Iowa State University Capstones, Theses and Dissertations at Iowa State UniversityDigital Repository. It has been accepted for inclusion in Retrospective Theses and Dissertations by an authorized administrator of Iowa State UniversityDigital Repository. For more information, please contact digirep@iastate.edu.

Recommended CitationNibbelink, Stuart Kent, "The role of tumor necrosis factor and secondary bacteria in the pathogenesis of swine dysentery " (1992).Retrospective Theses and Dissertations. 10336.https://lib.dr.iastate.edu/rtd/10336

INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI films the text directly fi-om the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.

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

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

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

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

University Microfilms international A Bell & Howell Information Company

300 North Zeeb Road. Ann Arbor. Ml 48106-1346 USA 313/761-4700 800/521-0600

Order Number 9234889

The role of tumor necrosis factor and secondary bacteria in the pathogenesis of swine dysentery

Nibbelink, Stuart Kent, Ph.D.

Iowa State University, 1992

U M I 300 N. ZeebRd. Ann Arbor, MI 48106

The role of tumor necrosis factor and secondary bacteria

in the pathogenesis of swine dysentery

by

Stuart Kent Nibbelink

A Dissertation Submitted to the

Graduate Faculty in Partial Fulfillment of the

Requirements for the Degree of

DOCTOR OF PHILOSOPHY

Department: Microbiology, Inununology and Preventive Medicine

Major: Veterinary Microbiology

Approved:

In pharge of Major Work

For Ae Major Progra

For the Graduate College

Iowa State University Ames, Iowa

1992

Signature was redacted for privacy.

Signature was redacted for privacy.

Signature was redacted for privacy.

Signature was redacted for privacy.

i i

TABLE OF CONTENTS

Page

GENERAL INTRODUCTION 1

EXPLANATION OF DISSERTATION FORMAT 3

LITERATURE REVIEW 4

PAPER I. SUSCEPTIBILITY OF INBRED MOUSE STRAINS TO INFECTION WITH Serpulina (Treponema) hyodysenteriae 19

ABSTRACT 20

INTRODUCTION 21

MATERIALS AND METHODS 23

RESULTS 26

DISCUSSION 37

REFERENCES 42

PAPER n. AN ENHANCED MURINE MODEL FOR STUDIES ON THE PATHOGENESIS OF Serpulina ("Treponema) hyodysenteriae 47

ABSTRACT 48

INTRODUCTION 49

MATERIALS AND METHODS 51

RESULTS 53

DISCUSSION 60

REFERENCES 62

PAPER m. PATHOGENESIS OF Serpulina hyodysenteriae: IN VIVO INDUCTION OF TUMOR NECROSIS FACTOR BY A SERPULINAL

i i i BUTANOLAVATER EXTRACT (ENDOTOXIN) 65

INTRODUCTION 66

RESULTS 69

DISCUSSION 79

MATERIALS AND METHODS 83

REFERENCES 86

PAPER IV. ABROGATION OF CECAL LESIONS, BUT NOT COLONIZATION, IN CONVENTIONAL MICE EXPERIMENTALLY CHALLENGED WITH Serpulina hyodysenteriae AND MAINTAINED ON ORAL ANTIBIOTICS 89

ABSTRACT 90

INTRODUCTION 91

MATERIALS AND METHODS 94

RESULTS 97

DISCUSSION 104

REFERENCES 106

PAPER V. EXPERIMENTAL INFECTION OF GNOTOBIOTIC MICE WITH Serpulina hyodysenteriae OR S. innocens, IN THE PRESENCE OR ABSENCE OF Bacteroides vulgatus 110

ABSTRACT 111

INTRODUCTION 112

MATERIALS AND METHODS 115

RESULTS 117

DISCUSSION 127

i V

REFERENCES 131

PAPER VI. EXPERIMENTAL INFECTION OF SEVERE COMBINED IMMUNODEFICIENT (SCID) MICE WITH Serpulina hyodysenteriae 134

ABSTRACT 135

INTRODUCTION 136

MATERIALS AND METHODS 138

RESULTS 142

DISCUSSION 154

REFERENCES 157

GENERAL SUMMARY 160

BIBLIOGRAPHY 162

ACKNOWLEDGMENTS 172

1

GENERAL INTRODUCTION

Swine dysentery, as a disease, has been described and the etiologic

agent identified, however, the disease pathogenesis and involvement of host

response(s) have not been characterized. An inflammatory component is

an integral facet of lesion development in the large bowel, but this

inflammatory component has not been characterized as being either

protective or detrimental to the host. The bacterium, Serpulina

hyodysenteriae (previously genus Treponema), fulfills Koch's postulates in

the induction of the disease in conventional animals; the spirochetes may be

isolated from clinical cases of disease, propagated in pure culture, and

cause disease in naive pigs upon challenge. In general, gnotobiotic pigs and

mice require the presence of a secondary bacteria for the development of

lesions following infection with S. hyodysenteriae. However, it has been

reported that development of lesions in gnotobiotic pigs may occur with

monoassociation with S. hyodysenteriae.

Research in disease pathogenesis is being performed in a mouse

model, but inconsistency exists in results between mouse strains. The

overall emphasis of this work is to further the understanding of this

enteric, inflammatory disease process, predominantly using the mouse

model. The work presented in this dissertation will address three general

goals: 1) to characterize inbred mouse strains as to susceptibility to

infection and to describe an enhanced model of susceptibility; 2) to

elucidate pathogenic mechanisms by characterizing germfree mouse

responses to infection with S. hyodysenteriae, and 3) to determine if tumor

2

necrosis factor, a putative factor in disease pathogenesis, is produced in

vivo in response to Serpulina endotoxin.

3

EXPLANATION OF DISSERTATION FORMAT

The alternate format has been used for this dissertation, which

includes six manuscripts. Four manuscripts, Papers 1,2,4, and 5, are

written in the American Society for Microbiology style (Infection and

Immunity). The remaining two manuscripts are written in the style of

Microbial Pathogenesis (Paper 3) and Laboratory Animal Science (Paper

m.

A literature review precedes the first manuscript, with a general

summary following the last manuscript. References for the literature

review and the general summary are listed in a literature cited section

which follows the general summary. Literature cited in the manuscripts

are cited at the end of each manuscript.

4

LITERATURE REVIEW

Swine dysentery has been estimated to cost producers 100 milUon

dollars annually (60). This cost encompasses mortality, as well as

decreased rate of gain and the routine incorporation of antibiotics in

rations to suppress dysentery outbreaks.

Swine dysentery was originally described by Whiting in 1921 (92).

The disease, also known as bloody diarrhea or necrotic enteritis, was

postulated to be an infectious entity by tracing outbreaks to stockyards,

with new stock arrival leading to clinical disease on the whole farai (90,

91). In addition, the feeding of feces, stomach tissue, or large intestine

from pigs clinically affected with swine dysentery also reproduced the

disease in naive pigs (91). Initial trials were performed to determine if

Bacillus suipestifer was the causitive agent (90). The cause of swine

dysentery was still unknown in 1944, when Doyle (16) described a Vibrio

species associated with the disease. The Vibrio was described in two

forms; one form was "coarse", and the other "fine". Pure cultures of this

bacteria fed to susceptible swine induced a diarrheic condition; however,

the clinical signs were less severe than those induced by feeding diseased

pig colon to susceptible swine (16). Doyle subsequently described the

feeding of a pure culture of this Vibrio to swine and inducing clinical

swine dysentery (34). Among other bacteria and protozoans observed in

dysenteric lesions (thus being possible causal candidates) were Salmonella

spp., paratyphoid organisms, and "protozoal" spirochetes and balantidia

5

(17). In 1968, Terpstra (86) demonstrated that convalescent sera from

pigs affected with swine dysentery recognized spirochetes in colonic

sections prepared from pigs clinically experiencing dysentery. Glock also

observed spirochetes in tissue sections of clinically affected pigs (22).

Concurrently, Taylor and Alexander (82) and Harris et al. (26)

described the pure culture cultivation of a spirochete {Treponema

hyodysenteriae) and the reproduction of swine dysentery by administration

of the organism to pigs (26). In 1991, the name Treponema

hyodysenteriae was changed to Serpula hyodysenteriae based on the lack of

homology of 16s ribosomal RNA between Treponema hyodysenteriae and

the type species of the genus, Treponema pallidum (79). Subsequently, the

genus Serpula was changed to Serpulina, as Serpula was a genus previously

assigned to a wood rot fungus (77).

S. hyodysenteriae was originally described as an aerotolerant

anaerobe (24); Stanton et al. (78), however, has shown the bacteria to

consume substrate amounts of oxygen and that growth of the organism is

enhanced by 1% oxygen in the growth media. Methods used to isolate the

organism include filtration (25) and growth on blood agar containing

various selective antibiotics (35, 51, 76). Propogation of the organism may

be accomplished by growth on solid media as well as in broth culture (48).

One of the earliest histopathological findings in swine dysentery is

the expulsion of mucigen from goblet cells in the basilar crypts of the large

intestine (93), concurrent with acute coagulative necrosis of the mucosal

epithelium and superficial lamina propria (31). The areas of coagulative

necrosis coincide with high levels of colonizing spirochetes (31).

6

Transmission electron microscropy shows S. hyodysenteriae within

necrotic mucosal epithelial cells, but tissue invasion is not necessary for

lesion development (1). Pohlenz et al. (69), working with gnotobiotic pigs,

observed S. hyodysenteriae in goblet cells in high numbers as early as 2

days post infection. By 4 days post infection, S. hyodysenteriae had

invaded enterocytes adjacent to goblet cells. Subsequent to infection, the

character of the mucus appeared changed; the mucus was more

homogeneous, less electron dense, and less sulphated (69). Mild mucosal

epithelial erosion was observed, but little inflammatory response was noted

in these pigs (69). It was hypothesized that S. hyodysenteriae has a

predilection for mucus, colonizes goblet cells, and causes the

hypersecretion of mucus and hyperplasia of immature basilar crypt cells

(69).

Following the inoculation of conventional pigs with S.

hyodysenteriae, one of the first findings noted by scanning electron

microscopy is a roughened, corrugated appearance of the cecal and colonic

mucosa (83). Albassam et al. (1) described initial lesions of superficial

vascular congestion in the cecum/colon and edema of the lamina propria.

Lesions progress to intracellular separation of epithelial cells in the crypt

shoulders (1), and a severe coagulative necrosis of the superficial mucosa

(31). Fibrin thrombi are observed within superficial lamina propria

vessels, and eroded surfaces are covered with exudate of mucus,

erythrocytes, fibrin, and bacteria (31). Under the eroded surface was

found an area of polymorphonuclear leukocyte infiltrate, and the lamina

7

propria and submucosa are infiltrated with predominantly mononuclear

leukocytes (31).

Spirochetes have been observed between colonic epithelial cells (83)

and free within the cytoplasm of enterocytes (1, 83). However, the actual

invasion of the enterocytes by the spirochete is not believed to be necessary

to cause disease (1, 45, 84). Conflicting views exist as to the need for

specific attachment by receptor:ligand interaction of the spirochete and

enterocyte to cause disease (9,45).

One of the putative virulence factors of S. hyodysenteriae is the

production of a beta-hemolysin (46, 47, 50, 52, 59, 70-73). Picard et al.

(68) initially described the increase of beta-hemolysin production by

spirochetes grown in broth supplemented with sodium ribonucleate.

Knoop (50) reported enhanced production of hemolysin in bacteria-free

filtrates by supplementation with yeast core RNA. The beta-hemolysin is

oxygen stable, heat labile, and is stable through a wide range of pH values

(71). The hemolysin is deduced as being a protein associated with lipids,

or a lipo-protein molecule (71). Hemolysin production by S.

hyodysenteriae is maximal at late-logarithmic phase (52). Evidence that

hemolysin synthesis is stimulated by sodium RNA and is secreted by viable

spirochetes was described by Lemcke and Burrows (52) when cell free

supematants of centrifuged washings or ultrasonically dismpted spirochetes

did not produce/contain hemolysin (52). Molecular mass of the hemolysin

was originally reported as 74 kDa (71), but more recently has been

reported at 26.9 kDa (59) to 19 kDa (47).

8

The hemolysin has recently been described as having action on

epithelial cells and lymphocytes (53), but most work testing the hemolysin's

mechanism of lysis has focussed on red cells. Lytic mechanisms on rabbit

erythrocytes did not require divalent cations and did not appear to be

lipolytic or proteolytic (71). Fixation of the beta-hemolysin to the

erythrocyte is temperature independent, but the hemolysis of the red cell is

temperature dependent (70). Lytic action appears to be associated with a

swelling (increase in intracellular volume) of the target red cells (70). The

pre-lytic swelling would suggest a colloid-osmotic lysis, similar to

streptolysin S; however, lysis by S. hyodysenteriae hemolysin is only partly

blocked by sucrose, with some hemoglobin release observed before or in

the absence of red cell swelling (70). This would indicate an alternate

mechanism of red cell lysis for the hemolysin. Saheb and Lafleur (70)

postulated that lysis without swelling would be due to a change in red cell

permeability caused by a digestion or disruption of some membrane

component.

Early work comparing the nucleic acid-bound hemolysin of S.

hyodysenteriae and nonpathogenic S. innocens showed that while both

hemolysins shared similar molecular masses, stability to oxygen, and heat

lability, the hemolysins were different on the basis of phospholipid

sensitivity, hemolytic spectra, and specific activity on rabbit erythrocytes

(72). While both S. hyodysenteriae and S innocens hemolysins were

efficient in lysing rabbit erythrocytes, 5. innocens hemolysin was 40-60%

less efficient in lysing bovine, sheep, horse, or pig erythrocytes (72).

Although most investigation deals with the effect of the hemolysin on red

9

cells, the hemolysin can be cytotoxic for various cells in vitw, fibroblasts

and porcine lymphocytes appear to be relatively sensitive to S.

hyodysenteriae beta-hemolysin killing (46). In addition, injection of high

levels of S. hyodysenteriae hemolysin (100,000 hemolytic units) into

colonic loops established in germfree pigs caused severe epithelial damage

to the cells in the intercrypt zone (53).

Much research has been performed to identify resistance to swine

dysentery in convalescent pigs. Pigs multiply exposed to S. hyodysenteriae

have been shown to become resistant to disease (65). Joens et al. (38)

demonstrated that in naive pigs, a high incidence of diarrhea and mortality

was observed after challenge with S. hyodysenteriae strain B204. After

convalescence (varying from 4 to 17 weeks post exposure), significant

protection against clinical signs as well as mortality was observed after

experimental re-challenge (38). Serum responses in these pigs, measured

by microtitration agglutination (39), indicated an anti-5. hyodysenteriae

response lasting up to 8 weeks post inoculation (38). It was also

hypothesized that, since shedding of S. hyodysenteriae by asymptomatic

pigs can occur as long as 89 days post infection (PI), a low level of colonic

colonization of S. hyodysenteriae could induce a protective response in the

convalescent pig (38). Fisher and Olander (21), in a similar study,

described the shedding (as determined by fecal culture) of S.

hyodysenteriae to 83 days post challenge, with the ability to transmit the

disease to sentinel animals after 57 days of convalescence. Pigs

experimentally infected with S. hyodysenteriae, both untreated and

virginiamycin-treated, exhibited elevated agglutinin titers out to 150 days

10

after challenge (21). These data suggest some means of immunologic

protection developed in convalescent animals, but that this protected state

might also coexist with a low level of S. hyodysenteriae colonization in the

pig.

Early attempts to induce protection against disease in naive pigs by

parenteral or oral immunization with inactivated or attenuated S.

hyodysenteriae showed either no protection (30) or a low level of

protection (23, 29), not thought to be of practical value in a field setting

(38). A bacterin has been described (67) which consists of a strain B204

bacterin in adjuvant which is administered parenterally. Protection studies

with this type of bacterin offer conflicting results (66, 67), with Olsen et

al. (66) reporting increased death and longer periods of clinical signs in

parenterally immunized pigs than non-vaccinated pigs experimentally

challenged with S. hyodysenteriae.

Subsequent work showed that convalescent pigs mount a humoral and

mucosal antibody response to S. hyodysenteriae (36), which also cross

reacts with many antigens of S. innocens, a non-pathogenic porcine

spirochete. Joens and Marquez (36) and Wannemuehler et al. (87)

analyzed and characterized the outer membrane components of S.

hyodysenteriae by electrophoresis and immunoblotting. Protein antigens

were recognized with molecular masses ranging from greater than 100 kDa

to 14 kDa (36) with many protein bands common to both S. hyodysenteriae

and S. innocens. Chatfield et al. (12) showed similar protein profiles of S.

hyodysenteriae by electrophoresis, and showed, through inmiune electron

microscopy, that the major antigen(s) recognized were associated with the

11

cell envelope. A wide band at 16 kDa was recognized by convalescent pig

serum which was species specific for S. hyodysenteriae (36). This 14-19

kDa antigen (87) was shown to be endotoxin- or lipopolysaccharide (LPS)-

like in character, as reactivity to immunoblot was only removed by

absorbing the convalescent pig serum with S. hyodysenteriae endotoxin

(butanol/water extract) (87). Treatment of the cell components with meta-

periodate eliminated the 16-19 kDa band, but treatment with proteinase K

did not (87). These results show that this antigen, which is species specific

for S. hyodysenteriae (36), is probably composed of carbohydrate and/or

lipid moieties (87).

Animal models used for in vivo investigation of S. hyodysenteriae

pathogenesis include chicks (81), guinea pigs (41), and mice (37). Joens

and Glock (37) first described the use of CF-1 mice as an infectious disease

model. Mice, after challenge with S. hyodysenteriae B234, showed mucoid

feces as a clinical sign (37). At necropsy, gross pathological changes in

mice included hyperemia of the cecal and colonic mucosa, catarrhal

inflammation of the cecal mucosa, and a varying incidence of cecal

petechial hemorrhage (37). Microscopically, the murine ceca showed

mucosal edema, hyperemia, dilitation of crypt areas, and a multifocal

epithelial erosions (37).

One of the first applications of the murine model was to differentiate

S. hyodysenteriae from S. innocens by virtue of S. hyodysenteriae

enteropathogenicity in CF-1 mice (42). While S. hyodysenteriae induced

grossly visible and microscopic lesions in murine ceca and large intestine,

S. innocens caused no lesions. In addition, S. hyodysenteriae could be

12

readily reisolated from infected mice ceca at necropsy (12-15 days), but S.

innocens was only shed for 7 to 10 days (42). While CF-1 mice were

originally described as a model (37), inbred mice are also susceptible to

development of lesions (61), however, susceptibilities do vary among

different inbred strains (61). Congenitally athymic (nude) mice also

appear susceptible to development of lesions after S. hyodysenteriae

challenge (80).

The murine model was subsequently used to investigate a potential

virulence attribute of S. hyodysenteriae. Nuessen et al. (64) described the

C3HeB/FeJ mouse strain, which possesses a functional Ips gene and

responds to the administration of LPS, as being susceptible to the

development of lesions after challenge with S. hyodysenteriae. In contrast,

C3H/HeJ mice, which are hyporesponsive to bacterial LPS, appeared to be

refractory to the development of lesions after S. hyodysenteriae challenge

(64). Nibbelink and Wannemuehler (61) described the decreased

susceptibility of C3H/HeJ mice to lesion development (when compared to

LPS responsive murine strains, such as C3Heb/FeJ, C3H/HeSn, and

C3H/HeN), but LPS hyporesponsive mice (C3H/HeJ) did develop lesions

when challenged with doses greater than 10^ bacteria. These two reports

showed the importance of the response to bacterial LPS in the development

of cecitis after S. hyodysenteriae infection, but also showed that lesion

development is multifactorial, and not influenced solely by response against

LPS.

After the description of the etiologic agent in 1972, a logical

extension of this work would be to investigate whether S. hyodysenteriae

13

could induce clinical swine dysentery in gnotobiotic pigs. Meyer et al. (56)

initially challenged germfree swine (5-20 days old) with pure cultures of S.

hyodysenteriae B78, with Vibrio (Campylobacter^ coli, or a combination

of the both. Germfree pigs of all three challenge groups did not show

clinical signs of disease but all agents readily colonized the pig gut. Only

minced dysenteric gut scrapings, fed to the germfree swine, would induce

clinical dysentery (which was attributed to Salmonella infantis

contamination) (56). Since S. hyodysenteriae, in pure culture fed to

conventional pigs, will induce disease, the inability of the same organism in

germfree swine to induce lesions was unexpected. Meyer et al. (56)

theorized that S. hyodysenteriae was an opportunist or conditional pathogen

that required another, possibly synergistic, organism to effect a disease

state (56). Subsequent work (10, 57) showed that S. hyodysenteriae in

combination with Escherichia coli or Lactobacillus or Campylobacter coli

or Clostridium also did not induce clinical dysentery in germfree swine.

Germfree pigs inoculated with S. hyodysenteriae and 4 anaerobes (not

characterized; bacteroides or fusobacteria) (58) did develop clinical disease

and lesions typical of swine dysentery. This work suggested an interaction

between certain anaerobic bacteria and S. hyodysenteriae to cause clinical

dysentery (58).

Alexander et al. (2) described the finding of high numbers of

Bacteroides vulgatus and Fusobacterium necrophorum in the colonic

mucosa of conventional pigs in early lesions of swine dysentery. Harris et

al. (27) demonstrated that gnotobiotic pigs, challenged with Bacteroides

vulgatus and S. hyodysenteriae did develop clinical disease and lesions of

14

swine dysentery. Pigs inoculated with S. hyodysenteriae alone or with B.

vulgatus alone did not show disease. This work demonstrated that a bi-

associated gnotobiotic pig would develop lesions. The role of the

secondary bacteria (in this case, B. vulgatus) was hypothesized to facilitate

colonization by the spirochete of the gnotobiotic pig; 4 of 6 pigs

monoassociated with S. hyodysenteriae were culture negative for the

spirochete (27). Gnotobiotic pigs exposed and colonized either with

Fusobacterium necrophorum, three strains of Bacteroides vulgatus, a

Clostridium species, or Listeria denitrificans and then secondarily

challenged with S. hyodysenteriae B204 showed colonic lesions of swine

dysentery with excretion of large amounts of mucus and some fibrin clots

(88). Clinical signs were observed 4 to 20 days after secondary challenge

with B204 (clinical signs were similar regardless of the species/type of

'secondary' bacteria S. hyodysenteriae was co-infected with) (88) which

indicated that S. hyodysenteriae was the etiologic agent of dysentery.

While interpreting this data (88), the authors hypothesized that: 1) the

colonizing numbers of S. hyodysenteriae associated with the mucosa was of

more importance than the identity of the 'other' bacterium; and 2) a

synergism exists between the two bacteria such that S. hyodysenteriae may

colonize/replicate to higher numbers on/in the mucosa. The synergistic

factor(s) of the anaerobic bacteria has not been identified. Histologically,

gnotobiotic pigs challenged with S. hyodysenteriae, B. vulgatus, and F.

necrophorum show B. vulgatus and F. necrophorum colonization of the

colonic lumenal surface, whereas S. hyodysenteriae invaded the crypts of

the pig colon (43). This work supported the assertation that synergistic

15

bacteria enable S. hyodysenteriae to colonize the crypts of the mucosa (43).

While most reports recognize a need for a secondary bacteria to be present

in a gnotobiotic animal for S. hyodysenteriae to cause lesions and disease

(27, 40, 43, 56, 58, 88), it was reported (89) that S. hyodysenteriae alone

caused disease in gnotobiotic pigs. The reason(s) for the expression of

pathogenicity of S. hyodysenteriae in these monoassociated gnotobiotic pigs

is unknown, but might have been due to the diet or subtle changes in the

growth media for the inocula (89).

In a very similar protocol, Joens et al. (40) showed that gnotobiotic

mice, colonized with Bacteroides vulgatus alone or S. hyodysenteriae B204

alone, do not develop gross cecal lesions when examined 9 days PL

Gnotobiotic mice challenged with B204 and B. vulgatus, however, showed

gross cecal lesions beginning at 3 days PI. Recent work, comparing effects

of colonization by S. hyodysenteriae B204 alone or S. innocens B256 alone

(62) showed that, despite colonization by both the pathogenic and the non­

pathogenic spirochete, no grossly evident lesions were observed out to 115

days PI. Cecal sections of S. hyodysenteriae-mitcitd mice, however,

showed invasion by the spirochete into the lamina propria, whereas

invasion was not noted in mice infected with S. innocens (62). This work

was in agreement with previous observations (40), in that S. hyodysenteriae

could colonize monoassociated gnotobiotic mouse ceca without synergistic

bacteria present. A hypothetical reason for the need of synergistic bacteria

to colonize gnotobiotic pigs with S. hyodysenteriae may be a function of

the disease progression: spirochetes are very viscotactic, motile in mucus

strands, which, during early lesion development, would enable S.

16

hyodysenteriae to colonize the crypts. Whether or not the hyper-secretion

of mucus is needed for lesion development remains to be answered.

The germfree model (specifically, the mouse) does show some innate

immunological differences from conventional animals. Germfree mice

have been described as having smaller secondary lymphoid organs than

normal animals. Secondary lymphoid organ development of germfree

animals is especially retarded if the lymphoid organs are in an area

nomially colonized by a bacterial flora, such as Peyer's patch areas and

lymphoid aggregates in the mouse intestinal tract (4). Germfree animals

also have low gamma globulin levels in the serum (75). Germfree mice,

immunized with specific antigen, do develop a comparable specific

antibody response with respect to conventional mice, but a longer delay

before serum antibody appears is evident (5). Macrophage numbers and

capability appear similar between germfree and conventional mice in

spleens and lymph nodes (3). Phagocytic indices for clearance of carbon

are the same between germfree and conventional mice, however, activity of

macrophages (higher killing of Listeria) in response to microorganisms are

increased or higher in conventional animals (15, 54). Also, indirectly

focusing on the germfree macrophage, germfree mice are five-fold more

resistant to endotoxin-induced lethality than are conventional mice (49),

and germfree mouse peritoneal exudate cells (PEC)(peritoneal

macrophages) produce less TNF when stimulated with LPS than

conventional mouse PECs ( M. J. Wannemuehler, unpublished data).

McGhee et al. (55) postulated that lipid A of bacterial LPS derived from

17

normal gram-negative microflora induces the production of a T cell

population which regulates B cell responses to bacterial LPS.

Bacterial LPS, injected in conventional animals in vivo, causes

neutrophil accumulation in tissue (32). Since LPS exhibits no chemotactic

properties for neutrophils in in vitro systems, endogenous factors secreted

by the host in response to the LPS probably would cause the neutrophil

accumulation (74). One of the cytokines that bacterial endotoxin/LPS

induces the production of is tumor necrosis factor (TNF) (11).

TNF-alpha (cachectin) was first described as an endotoxin-stimulated

protein which mediated hemorrhagic necrosis of tumors in mice (11).

Macrophages and monocytes, classically, are the major source of TNF (7,

33). Tumor necrosis factor is a polypeptide cytokine which is involved in

the induction of a pro-inflammatory state; subsequent proteins/factors

released include interleukin-1 (IL-1), prostaglandin E2, up-regulation of

MHC class I antigens (6,13,14,18), intercellular adhesion molecule-1

(ICAM), and endothelial leukocyte adhesion molecule-1 (ELAM-1) (3).

Also of potential interest in the pathogenesis of swine dysentery is the

involvement in many inflammatory diseases of platelet activating factor

(PAF), which may, with TNF, mutually upregulate each cytokine's

production (19). Platelet activating factor is synthesized by a wide variety

of cells, including basophils, neutrophils, platelets, and endothelial cells in

response to inflammatory stimuli. Platelet activating factor has been

implicated as being increased in tissues of patients undergoing ulcerative

colitis (20), and experimental ischemic bowel necrosis is resultant from

administration of synthetic PAF to rats (28).

18

One of the hallmark characteristics of bacterial endotoxin is the

ability to elicit the Shwartzman reaction, a local dermal

necrotic/hemorrhagic reaction attributed to induction of inflammatory

mediators, namely, IL-1, gamma interferon, and TNF (8). Nordstoga (63)

has described the similar histologic appearance of fibrinonecrotic colitis in

swine and the dermal Shwartzman reaction. If this similarity is true, then a

probable role for TNF in the pathogenesis of swine dysentery exists.

19

PAPER 1.

SUSCEPTIBILITY OF INBRED MOUSE STRAINS TO

INFECTION WITH Serpulina (Treponema) hyodysenteriae

20

ABSTRACT

Several inbred strains of mice were infected with Serpulina

(Treponema) hyodysenteriae strain B204 to determine susceptibility to

infection. Challenge doses of 10^ or 10^ spirochetes induced cecal lesions

in C3H/HeJ and other C3H strains of mice. However, more than a 2 log

difference existed between the dose required to induce lesions in 50%

(ID50) of the infected C3H/HeJ (8.3 X 10^) or C3H/HeN (5 X lO^) mice.

C3H/HeJ mice lack a splenocyte mitogenic response to Escherichia coli

lipopolysaccharide, but exhibited a mitogenic response comparable to other

C3H strains of mice when stimulated with S. hyodysenteriae endotoxin

(butanol/water extract). Different inbred strains exhibited differing

susceptibilities to infection, with the C3H/HeN strain being the most

susceptible based on colonization and development of macroscopic cecal

lesions. The ity gene had no apparent effect on susceptibility of mice

challenged with S. hyodysenteriae. The involvement of H2 haplotype with

susceptibihty is unclear but the mice bearing H-2K were more susceptible

than mice with H-2b, H-2d, or H-2^1 haplotypes. These data support the

hypothesis that the host's responsiveness to LPS influences the susceptibility

to infection with S. hyodysenteriae. However, susceptibility differences

between inbred mice exist independent of the Ips locus, suggesting other

inherent differences between mouse strains which affect susceptibility to

infection by S. hyodysenteriae.

21

INTRODUCTION

Swine dysentery is a mucohemorrhagic diarrheal disease of pigs

caused by Serpulina (Treponema) hyodysenteriae (7,33). Affected pigs

become gaunt and dehydrated due to non-reabsorbtion of fluids and colonic

ion transport abnormalities (2). Affected cecal and colonic mucosa exhibit

mucohemorrhagic exudation and fibrinonecrotic membrane formation

(10). Histologic lesions include mucosal epithelial erosions, inflammatory

cell infiltration of the lamina propria, and coagulative necrosis of the

superficial mucosa (1,10). Potential virulence determinants of S.

hyodysenteriae include an endotoxin-like moiety (21,22) and a 6-hemolysin

(18,31). In vivo models (other than pigs) used to examine the virulence of

S. hyodysenteriae have included guinea pigs (13), mice (12, 14, 20, 35),

and chicks (36). Because of availability of reagents and the ability to

adoptively transfer cells, we have utilized inbred mice as a model to study

the pathogenesis of S. hyodysenteriae.

Ceca from mice infected with S. hyodysenteriae exhibit

histopathological lesions similar to those of pigs with swine dysentery (12).

The endotoxin of S. hyodysenteriae is biologically active (5,6,22) and was

suggested as a virulence determinant by work which showed that LPS

hyporesponsive C3H/HeJ mice did not develop macroscopic cecal lesions

following S. hyodysenteriae infection, while LPS responsive C3HeB/FeJ

mice did develop cecitis (21). The host responsiveness to

Enterobacteriaceae endotoxin has been shown to be important in resistance

22

to mucosal infections with Escherichia coli (37). The Ips gene has been

shown to control the host's sensitivity to lipid A, B cell mitogenesis, and

the induction of tumor necrosis factor (3,30,40). Another murine genetic

loci, which determines susceptibility to gram negative bacterial infections,

has been identified as ity. (26,27). The role of the ity gene has been shown

to control the proliferation of Salmonella typhimurium within the spleen

and liver (26,27). In contrast to a previous report (21), we have shown

that C3H/HeJ mice developed mucosal lesions following S. hyodysenteriae

infection which are comparable to those observed in other LPS responsive

C3H strains of mice. The present work was undertaken to define murine

susceptibility to S. hyodysenteriae infection in terms of: 1) the influence of

Ips and ity genes; 2) influence of murine H2 haplotype; and 3) to

determine a 50% infective dose (ID50) for different inbred murine strains.

23

MATERIALS AND METHODS

Mice. C3H/HeSn (Ipsn/lpsn, ity Q, C3HeB/FeJ (Ips^Hpsn, ity 0,

DBA/1 J (Ips^llps^), and BALB/cByJ {Ips^llps^, ity -î) mice were purchased

from Jackson Laboratories, Bar Harbor, Maine. C57BL/6J (Ips^llps^, ity

mice were kindly provided by Dr. C. J. Warner, Department of

Biochemistry, Iowa State University. C3H/HeN Qps^llps^, ity 0 and

C3H/HeJ {Ips^/lps^, ity mice were obtained from breeding colonies

which were originally procured from Harlan Sprague Dawley,

Indianapohs, Indiana, and Jackson Laboratories, respectively. Mice were

maintained on Mouse Lab Chow #5010 (Purina Mills, Inc., St. Louis,

Missouri) under controlled conditions in the Laboratory Animal Resource

Facility of the College of Veterinary Medicine, Iowa State University.

Mice were 6 to 10 weeks of age at the time of experimentation.

Bacterial Strains. S. hyodysenteriae strain B204 (serotype 2) with less

than 20 in vitro passages was grown anaerobically at 37°C in Trypticase Soy

Broth (BBL Microbiological Systems, Cockeysville, Maryland) supplemented

with 5% horse serum (HyClone Laboratories, Logan, Utah), 0.5% yeast extract

(BBL), and 1% VPI salt solutions as previously described (38). Highly motile

(motility > 90%) log phase cultures were obtained by inoculating warm (37°C)

broth with a 33% inoculum from an overnight culture and incubating for

approximately 5 h. Bacteria were enumerated using a Petroff-Hauser counting

chamber and diluted with warm complete media prior to inoculation of mice.

24

Mitogenic Response Determination. C3H/HeSn, C3H/HeJ,

C3FeB/HeJ, and C3H/HeN splenocytes (5 x lO^/well) were stimulated with

0.2, 1.0, 5.0, or 25.0 p-g/ml of E. coli LPS or 1.0, 5.0 or 25.0 |Lig/ml of S.

hyodysenteriae endotoxin (5) as previously described (39).

Infection Procedure. Mice were administered 2 doses of bacteria

24 h apart by gastric intubation. Feed was removed 6 h prior to the first

challenge and withheld for 36 h. The standard infective challenge utilized

to determine susceptibility of various mouse strains was 10^ bacteria in 1

ml culture broth. In dose titration studies, spirochetes were diluted to the

appropriate concentration and delivered in 1 ml of culture medium.

Necropsy Procedure. Mice were sacrificed at 10 or 15 days post­

infection (PI). Criteria used to determine the presence of disease were 1)

macroscopic cecal lesions, 2) re-isolation of S. hyodysenteriae from cecal

contents, and 3) levels of S. hyodysenteriae colony forming units (CPUs)

per gram of cecal tissue. Macroscopic cecal lesions were scored as follows;

cecal organ atrophy and excess intraluminal mucus, 3; atrophy and excess

intraluminal mucus localized in the cecal apex only, 2; excess cecal mucus

with no evidence of atrophy, 1; and no gross lesions (NGL), 0. Cecal

atrophy was determined by comparison of affected ceca with control ceca.

Cecal apices were fixed in neutral buffered formalin, sectioned, and stained

with hematoxylin-eosin.

Isolation of 5. hyodysenteriae. Cecal contents were streaked

onto citrated-ovine blood agar plates containing antibiotics as previously

described (16). Plates were incubated anaerobically at 37°C for 96 h.

25

Strongly B-hemolytic colonies were confirmed to be large spirochetes by

dark field microscopy.

Plate Count Technique. Ceca were aseptically removed and

weighed in sterile WhirlPak bags (Baxter Co., Minneapolis, Minnesota),

suspended 1:100 (w/v) in phosphate-buffered saline (NaCl 136.9 mM,

Na2HP04 8.1 mM, KH2PO4 1.5 mM, KCl 2.7 mM, pH 7.2), and

homogenized in a Stomacher Laboratory Blender (A. J. Seward Co.,

London, U. K.). This suspension was directly examined by dark field

microscopy for the presence of large spirochetes. Appropriate dilutions of

the homogenate were added to 5 ml molten (45°C) Trypticase Soy agar

tubes supplemented with 5% ovine blood and antibiotics, vortexed, and

poured into 60 mm Petri dishes (#25010, Coming Glass Works, Coming,

N. Y.). Petri plates were incubated anaerobically at 37°C for 96 h. The

zones of B-hemolysis were enumerated and the number of S.

hyodysenteriae colony forming units per gram of cecal tissue was

calculated.

Statistical Evaluation. Student's t-distribution (32) was used to

determine significance. The number of spirochetes required to produce

disease in 50% of the infected mice, infectious dose-50 (ID50), was

estimated by the method of Reed and Muench (28).

26

RESULTS

Evaluation of Susceptibility of C3H Mouse Strains. The

macroscopic cecal scores of C3H mice challenged with 1 x 10^ 5.

hyodysenteriae are represented in Figure 1, and corresponding colonization

values are expressed in Figure 2.

g 10 d PI

0 15dPI

C3H/H0N C3H/HeJ G3Heb/FeJ C3H/HeSn

Figure 1. Susceptibilities of C3H strains of mice following challenge with IxlO? S. hyodysenteriae strain B204. Cecal scores range from 3 (severe lesions) to 0 (no macroscopic lesions). Results are expressed as the mean ± S.E.M. At 10 days PI (solid bars), n= 4, 4,5, and 5 for C3H/HeN, C3H/HeJ, C3HeB/FeJ, and C3H/HeSn, respectively. At 15 days PI (hatched bars), n= 5, 5, 5, and 4, for C3H/HeN, C3H/HeJ, C3HeB/^eJ, and C3li'HeSn, respectively.

27

C3H/HeN C3H/HeJ C3HeB/FeJ

^ CSH/HeSn

10dPI 15dPI

Figure 2. Level of cecal colonization of C3H strains as estimated by recovered S. hyodysenteriae colony forming units (CPUs). Mice were challenged with IxlO? bacteria as described in Materials and Methods. Results are expressed as the logio of the mean number of CPUs per gram of cecal tissue.

28

0 C3H/HeJ • C3H/HeN

Group mean cecal scores

Figure 3. Ability of C3H/HeN and C3H/HeJ mice to resolve cecal lesions following infection with Serpulina hyodysenteriae B204. Mice were infected with 1x10^ S. hyodysenteriae as described in Materials and Methods. Cecal scores range from 3 (severe lesions) to 0 (no macroscopic lesions). Results are expressed as the mean ± S.E.M. Percentages indicate proportion of each respective group culture positive for S. hyodysenteriae. n=5 for all groups. *, no C3H/HeJ mice exhibited macroscopic lesions at 60 days PI.

29

AU mice, C3H/HeN, C3H/HeJ, C3H/HeSn, and C3HeB/FeJ,

necropsied at 10 days PI were culture positive for S. hyodysenteriae, and

susceptibility to disease was similar. At 15 days PI, C3H/HeN mice

appeared most susceptible (as evidenced by greatest development or

persistence of macroscopic lesions) but all strains exhibited signs of

infection by S. hyodysenteriae (Fig. 1). At 15 days PI, the lesion scores

appeared more severe in C3H/HeN mice than the other C3H strains (Fig.

1), but the numbers of S. hyodysenteriae recovered were similar for all

strains examined (Fig. 2). To determine whether the lower lesion scores

associated with the C3H/HeJ at 15 days PI were associated with enhanced

resolution of the lesions, C3H/HeJ and C3H/HeN mice were sacrificed at

various times over a 70 day period (Fig. 3). Following infection, S.

hyodysenteriae was isolated from both C3H/HeJ and C3H/HeN mice at all

times points (Fig. 3), and both strains of mice presented cecal lesions

except the C3H/HeJ mice at 60 days PI.

Mitogenesis. As expected, C3H/HeJ splenocytes were

hyporesponsive to E. coli LPS-induced mitogenesis (Fig. 4a) in comparison

to spleen cells from other C3H strains of mice. However, S.

hyodysenteriae endotoxin (butanol/water extract, 14) induced a mitogenic

response from C3H/HeJ spleen cells similar to that of spleen cells from

LPS-responsive mice (Fig. 4b).

30

50

40

30

20

10

0

£ co//LPS (ng/ml)

50H

40.

c o

3 E 55

30.

C3H/HeSn ^ CSHeB/FeJ .{J CSH/HeJ

C3H/HeN 20.

S. /j/ooysenfer/ae endotoxin (ng/ml)

Figure 4. Splenocyte mitogenic response to E. coli LPS and S. hyodysenteriae endotoxin. Splenocytes from two non-infected mice of each inbred strain were pooled and stimulated with either E. coli LPS (panel A) or S. hyodysenteriae endotoxin (panel B). Results are expressed as the stimulation index (cpm of stimulated cells/cpm of non-stimulated cells). Mouse strains are identified for both panels by the inset in panel B.

31

This indicated that C3H/HeJ mice were able to mitogenically respond

to spirochetal cell wall components.

Titration of Infective Dose for C3H/HeN, C3H/HeJ,

BALB/cByJ, C57BL/6J, and DBA/IJ. Mice were challenged with

increasing doses of S. hyodysenteriae strain B204 ranging from 1 x 10^ to

1 X 108 organisms. Mice were sacrificed at 10 or 15 days PI and

evaluated for lesion development. Cecal scores are represented in Figure

5.

3

C3H/HeN C3H/HeJ

2

1

0

Challenge dose of Serpulina hyodysenteriae

Figure 5. Susceptibilities of C3H/HeN and C3H/HeJ mice to titrated doses of S. hyodysenteriae challenge based on the determination of cecal scores. Mice were challenged orally on two consecutive days with the indicated dose of S. hyodysenteriae B204. Results are expressed as the mean of the observed cecal scores. For the C3H/HeN mice sacrificed at 10 days PI n= 14,13,19,20, and 7 for 108,10^, 10^, 10^, and 10^ challenge doses, respectively. At 10 days PI, n=7 for all C3H/HeJ groups. *, p<0.05.

32

At the infective dose of 1 x 10^, both of the C3H strains developed

similar macroscopic cecal lesions. As the challenge dose was decreased,

fewer of the C3H/HeJ mice developed macroscopic lesions. The

differences in the cecal scores between C3H/HeJ and C3H/HeN mice (Fig.

5) were significantly different (p< 0.05) at the challenge dose of 1 x 10^.

The proportion of each group of mice that was culture positive for

S. hyodysenteriae when sacrificed is shown in Table 1.

Table 1. Percentage of mice culture positive 10 days after challenge with various dosages of S. hyodysenteriae ^

Infective Dose^ Strain 108 107 106 105 104 103 102

C3H/HeN 93(14) 85(13) 100(19) 100(20) 93(14) 86(7) 57(7)

C3H/HeJ 86(7) 71(7) 64(14) 38(13) 13(8) N.D.C N.D.

C57BL/6J 29(7) 71(7) 0(6) 0(6) N.D. N.D. N.D.

DBA/IJ 100(6) 86(7) 14(7) 29(7) N.D. N.D. N.D.

BALB/cByJ 0(8) 0(8) 0(8) 0(8) N.D. N.D. N.D.

^The percentage of mice in each challenge group which were culture positive for S. hyodysenteriae strain B204. Parenthetical numbers indicate group size. ^Mice were challenged with titrated doses of S. hyodysenteriae strain B204 as described in Materials and Methods. CNot determined, N.D.

33

The percentage of C3H/HeN, C3H/HeJ, and DBA/1 J mice which

were culture positive for S. hyodysenteriae were similar (i.e., 71 to 100%)

for infective doses of 1 x 10^ or 1 x 10^. However, cecal samples from

DBA/1 J mice produced fewer positive cultures than the C3H mice at doses

of 1 X 106 and 1 x 10^ S. hyodysenteriae . At a challenge dose of 1 x 10^

organisms, C3H/HeJ mice developed less severe macroscopic lesions than

C3H/HeN mice (p < 0.05, Fig. 5), and the recovery of S. hyodysenteriae

from cecal samples was lower in comparison to C3H/HeN mice (Fig. 6).

C3H/HeN C3H/HeJ

III II ° o oi

10 10 10 10 10 10 10 S. hyodysenteriae chaWenge inoculum

Figure 6. Level of cecal colonization of C3H/HeN and C3H/HeJ mice as estimated by the recovery of S. hyodysenteriae. Results are expressed as the logic of the mean number of colony forming units (CPUs) per gram of cecal tissue. *p<0.05.

34

In addition, examination of mice at 15 days PI demonstrated that

C3H/HeJ mice developed significantly less (p < 0.05) disease than similarily

infected C3H/HeN mice following challenge with 1 x 10^ or fewer

organisms (data not shown). Cecal lesions were noted in C3H/HeN mice

following infection witii as few as 100 spirochetes; however, 1 x 104

organisms failed to induce macroscopic lesions in C3H/HeJ mice sacrificed

at 15 days PI (data not shown). C57BL/6J mice were culture negative at 1

X 106 and 1 x 10^, and BALB/cByJ mice were culture negative at all

challenge doses examined (Table 1).

The data in Figures 6 and 7a indicates the CPUs of S.

hyodysenteriae recovered from the ceca of infected mice. The recovery of

S. hyodysenteriae from C3H/HeJ mice following infection with 1 x 10^ to

1 X 104 organisms was significantly different (p < 0.05) from that

obtained from similarly infected C3H/HeN mice (Pig. 6). In comparison to

the C3H/HeN mice, the CPUs of S. hyodysenteriae recovered from DBA/IJ

and C57BL/6J were at least 100- to 1000-fold less at all challenge doses

(Fig. 7a). In addition, the presence of cecal lesions in C57BL/6J or

DBA/IJ mice were much less frequent than those observed in C3H mice

(Fig. 7b). BALB/cByJ mice failed to develop lesions irregardless of the

number of spirochetes in the challenge inoculum (data not shown).

Using a cecal score of 3 as evidence of murine susceptibility, an

ID50 dose of S. hyodysenteriae was calculated to be 5 x 10^ for C3H/HeN

mice, and 8.3 x 10^ for C3H/HeJ mice. Other inbred murine strains

35

(BALB/cByJ, DBA/IJ, C57BL/6J) exhibited low susceptibility to the

development of cecal lesions and this prohibited calculating ID50 doses.

36

8

•C I 4

i î g-c/i 2 oi O)

0

3

B

C57BL/6J DBA/1J 2

9- 1

0

10 10 10 10

S. hyodysenteriaechaWenge inoculum

Figure 7. Susceptibilities of DBA/1 J and C57BL/6J mice to titrated doses of S. hyodysenteriae challenge. Panel A: Colony forming units (CPUs) of S. hyodysenteriae recovered 10 days post infection (DPI) Results are expressed as the loglO of die mean number of CPUs per gram of cecal tissue. Panel B; Results are expressed as the mean cecal score observed for mice sacrificed 10 dPI. Six to eight mice were used for each dose for each strain of mice. Mouse strains are identified for both panels by the inset in panel B.

37

DISCUSSION

The difference in the susceptibility between C3H/HeN and C3H/HeJ

mice following infection with S. hyodysenteriae suggests a role for

endotoxin or the host's responsiveness to endotoxin in the development of

cecal lesions. In agreement with previous work (21), we observed

differences in susceptibility of the LPS hyporesponsive C3H/HeJ (8,40)

mice to challenge with S. hyodysenteriae in comparison to LPS-responsive

mice. Lesion development in C3H/HeJ mice was similar to that observed

for C3H/HeN mice when the mice were challenged with doses > 1 x 10^

spirochetes (Fig. 5). C3H/HeJ mice were less susceptible to infection at 1 x

104 and 1 x 10^ challenge than were C3H/HeN mice, which were

susceptible to lesion development at a challenge dose of 100 organisms

(data not shown). In addition, susceptibility differences to infection by S.

hyodysenteriae existed between the other murine strains tested, independent

of Ips genotype.

The reason(s) for the increased susceptibility of the C3H/HeN mice

in comparison to C3H/HeJ mice to infection with S. hyodysenteriae is

unknown. Previous work shows that C3H/HeN mice are less susceptible to

infection with Salmonella typhimurium (23,24,29) dnà Klebsiella

pneumoniae (25) than the LPS hyporesponsive C3H/HeJ mice. The

increase in susceptibility of the C3H/HeJ mice to Salmonella infections has

been attributed to the inability to activate macrophages via the lipid A

moiety of bacterial LPS (30). Conversely, C3H/HeJ mice have been shown

to be less susceptible to infection with Neisseria gonorrhoeae, to develop a

38

greater peritoneal leukocyte response, and to clear gonococci from the

circulation faster than C3H/HeN mice (34). These observations indicate

that the host's responses play an active role in elimination of the pathogen

but that an elevated response may result in a prolonged infection or

development of more severe lesions.

Most pathogenicity studies comparing lps<^llps^ and Ips^/lps^ mice

have involved invasive microorganisms which were administered either

intravenously or intraperitoneally (23-25, 29). Recently, Svanborg-Eden

et al. (37) evaluated both local and disseminated infection in C3H and

C57BL mice, describing the effect of the Ips and iîy genes on susceptibility

to disease at a mucosal surface. Disseminated infection (i.e., intravenous S.

typhimurium challenge) was shown to be influenced by Ips and ity genes,

with the Ips gene influence being greater in determining susceptibility.

Local infection (i.e., intravesicular E. coli challenge) was not influenced by

the ity genotype, but mice with the lps<^/lpsd genotype did exhibit

impaired clearance (37). The effector mechanism(s) involved in clearance

of the mucosal infections appeared to require polymorphonuclear leukocyte

infiltration and acute inflammation (37). This mucosal inflammatory

response was absent in the C3H/HeJ mouse (37). Treatment of C3H/HeN

mice with indomethacin resulted in the development of mucosal infection

and subsequent disease, indicating the importance of the host's

inflammatory response as a means of mucosal bacterial resistance (19,37).

The lesions associated with swine dysentery are not induced by a massive

tissue invasion of microorganisms (1,10). It is possible that the continued

presence of S. hyodysenteriae in the intestinal lumen will stimulate a

39

persistant inflammatory response which would result in the presence of

chronic lesions. This was suggested by the fact that the intestinal lesions

were not resolved within 70 days following infection of either C3H/HeJ or

C3H/HeN mice (Fig. 3). Indeed, a threshold level of colonizing S.

hyodysenteriae appears necessary to cause macroscopic lesion development

in mice; approximately 1 X 10^ or greater spirochetes per gram of cecal

tissue was necessary before severe cecitis (cecal score of 3) was observed

(personal observation). C3H/HeJ mice may require a higher threshold

stimulus (i.e., number of CFU/gram cecum) to initiate pathogenesis, thus,

the higher ID50 in comparison to C3H/HeN mice. A higher stimulus

threshold might also explain why C3H/HeJ mice have a lower incidence of

cecal lesions (i.e., 60 days PI, Fig. 3) than similarly infected C3H/HeN

mice, even though the frequency of S. hyodysenteriae re-isolation is

comparable (Fig. 3).

Previous work failed to detect a neutrophil infiltration into the cecal

lamina propria of C3H/HeJ mice following challenge with S.

hyodysenteriae (21). Our results indicate that C3H/HeJ mice, when

challenged with >1 x 10^ S. hyodysenteriae, exhibit similar neutrophil

accumulation in the lamina propria/submucosa as the C3H/HeN.

Since mucosal inflammation caused by E. coli LPS appears to be

triggered at the mucosal epithelial surface (19), it is possible that the

pathogenesis of swine dysentery also involves interaction of LPS/endotoxin

with the colonic epithelium. Serpulina hyodysenteriae has been shown to

closely associate with porcine mucin in vitro and porcine colonic mucosal

tissue in vivo (15).

40

This research suggests a multifaceted pathogenesis, whereby the

degree of host responsiveness to LPS (whether it be spirochetal endotoxin

or LPS from the indigenous flora) corresponds to the relative susceptibility

to disease. However, alternate pathways for the induction of inflammation

at the mucosal level might occur in C3H/HeJ mice, such as the induction of

IL-6 resulting from tissue trauma (4). Also, C3H/HeJ macrophages have

been shown to be activated if "primed" first with interferon gamma and

then exposed to bacterial endotoxin (butanol/water extract) (9). It must be

remembered that C3H/HeJ mice are LPS "hyporesponders", and may,

therefore, be capable of a limited LPS/endotoxin-induced inflammatory

response. The ability of C3H/HeJ mice to respond to bacterial endotoxin

has been demonstrated (30, Fig. 4). However, preliminary observations

indicate that C3H/HeJ (in comparison to C3H/HeN) peritoneal exudate cells

produce less tumor necrosis factor and interleukin-1 after stimulation with

S. hyodysenteriae endotoxin in vitro (M. J. Wannemuehler et al.,

manuscript in preparation). This observation may suggest that the mucosal

inflammatory response(s) induced by the spirochetal cell components in the

C3H/HeJ mice is less severe than in C3H/HeN mice.

Others have observed differences in murine strain susceptibility

upon infection with S. hyodysenteriae (35), theorizing that differences in

the intestinal microflora would affect susceptibility or colonization.

Presently, we have demonstrated that manipulation of the intestinal

microflora (by oral antibiotic treatment or diet changes) does affect

susceptibility of mice to infection with S. hyodysenteriae (work in

progress). However, all mice used during these studies were maintained

41

under similar environmental conditions, feed, and care, thereby

minimizing differences in gut microflora (11). We have found the mouse

model to be optimized by challenging with S. hyodysenteriae in log phase,

using only highly motile (> 90% motile by dark-field microscopy)

cultures. We have not determined the difference between challenging with

log vs. late-log or stationary cultures, but Lee and Falkow (17) have shown

that Salmonella pathogenicity is greatly increased in log phase cultures as

opposed to stationary cultures.

These results suggest the importance of host responsiveness to LPS in

the development of an intestinal inflammatory disease. However, the

response to LPS alone is not the only determinant in the development of the

intestinal lesions, as C3H/HeJ mice are susceptible to lesion development at

doses in excess of 1 X 10^ bacteria. This may reflect the hyporesponsive

nature of C3H/HeJ mice to LPS and indicate that a sufficient challenge

inocula must excede a "threshold" for lesion development. Involvement of

H-2 haplotype on susceptibility to S. hyodysenteriae is unclear at this time;

however, all C3H strains (H-2k) were more susceptible than BALB/cByJ

(H-2d), DBA/IJ (H-2q), or C57BL/6J (H-2b).

42

REFERENCES

1. Albassam, M. A., H. J. Olander, H. L. Thacker, and J. J. Turek. 1985. Ultrastructural characterization of colonic lesions in pigs inoculated with Treponema hyodysenteriae. Can. J. Comp. Med. 49:384-390.

2. Argenzio, R. A. 1985. Pathophysiology of swine dysentery, p. 3-19. In C. J. Pfeiffer (ed.). Animal Models for Intestinal Disease. CRC Press Inc., Boca Raton, Florida.

3. Beutler, B., N. Krochin, I. W. Milsark, C. Luedke, and A. Cerami. 1986. Control of cachectin (tumor necrosis factor) synthesis; mechanisms of endotoxin resistance. Science 232:977-980.

4. De Man, P., C. Van Kooten, L. Aarden, I. Engberg, H. Linder, and C. Svanborg Eden. 1989. Interleukin-6 induced at mucosal surfaces by gram negative bacterial infection. Infect. Immun. 57:3383-3388.

5. Greer, J. M., and M. J. Wannemuehler . 1989. Pathogenesis of Treponema hyodysenteriae: induction of interleukin-1 and tumor necrosis factor by a treponemal butanol/water extract (endotoxin). Microbiol. Pathogen. 7:279-288.

6. Greer, J. M., and M. J. Wannemuehler. 1989. Comparison of the biological responses induced by lipopolysaccharide and endotoxin of Treponema hyodysenteriae and Treponema innocens. Infect. Immun. 57:717-723.

7. Harris, D. L., R. D. Glock, C. R. Christensen, and J. M. Kinyon. 1972. Swine dysentery-1. Inoculation of pigs with rreponma hyodysenteriae (new species) and reproduction of the disease. Vet. Med. Sm. Animal Clin. 67:61-68.

8. Heppner, G., and D. W. Weiss. 1965. High susceptibility of strain A mice to endotoxin and endotoxin-red blood cell mixtures. J. Bacteriol. 90:696-703.

43

9. Hogan, M. M., and S. N. Vogel. 1987. Lipid-A associated proteins provide an alternate "second signal" in the activation of recombinant interferon-gamma-primed, C3H/HeJ macrophages to a fully tumoricidal state. J. Immunol. 139:3697-3702.

10. Hughes, R., H. J. Olander, and C. B. Williams. 1975. Swine dysentery: pathogenicity of Treponema hyodysenteriae. Am. J. Vet. Res. 36:971-977.

11. Itoh, K., T. Mitsuoka, K. Sudo, and K. Suzuki. 1983. Comparisons of fecal flora of mice based upon different strains and different housing conditions. Zeit. Versuchstierk. 25:135-146.

12. Joens, L. A., and R. D. Clock . 1979. Experimental infection in mice with Treponema hyodysenteriae. Infect. Lnmun. 25:757-760.

13. Joens, L. A., J. C. Songer, D. L. Harris, and R. D. Clock. 1978. Experimental infection with Treponema hyodysenteriae in guinea pigs. Infect. Immun. 22:132-135.

14. Joens, L. A., R. D. Clock, and J. M. Kinyon. 1980. Differentiation of Treponema hyodysenteriae from T. innocens by enteropathogenicity testing in the CFl mouse. Vet. Rec. 107:527-529.

15. Kennedy, M. J., D. K. Rosnick , R. C. Ulrich , and R. J. Yancey. 1988. Association of Treponema hyodysenteriae with porcine intestinal mucosa. J. Gen. Microbiol. 134:1565-1576.

16. Kunkle, R. A., and J. M. Kinyon. 1988. Improved selective medium for the isolation of Treponema hyodysenteriae. J. Clin. Microbiol. 26:2357-2360.

17. Lee, C. A., and S. Falkow. 1990. The ability of Salmonella to enter mammalian cells is affected by bacterial growth state. Proc. Natl. Acad. Sci. USA 87:4304-4308.

18. Lemcke, R. M., and M. R. Burrows. 1982. Studies on a haemolysin produced by Treponema hyodysenteriae. J. Med. Microbiol. 15:205-214.

44

19. Linder, H., I. Mattsby-Baltzer, and C. Svanborg Eden. 1988. Natural resistance to urinary tract infection determined by endotoxin-induced inflammation. FEMS Microbiol. Lett. 49:219-222.

20. Nibbelink, S. K., and M. J. Wannemuehler. 1990. Effect of Treponema hyodysenteriae infection on mucosal mast cells and T cells in the murine cecum. Infect. Immun. 58: 88-92.

21. Nuessen, M. E., L. A. Joens, and R. D. Glock. 1983. Involvement of lipopolysaccharide in the pathogenicity of Treponema hyodysenteriae. J. Immunol. 131:997-999.

22. Nuessen, M. E., J. R. Birmingham, and L. A. Joens. 1982. Biological activity of a lipopolysaccharide extracted from Treponema hyodysenteriae. Infect. Immun. 37:138-142.

23. O'Brien, A. D., D. L. Rosenstreich, I. Scher, G. H. Campbell, R. P. MacDermott, and S. B. Formal. 1980. Genetic control of susceptibility to Salmonella typhimurium in mice. Role of the LPS gene. J. Immunol. 124:20-24

24. O'Brien, A. D., E. S. Metcalf, and D. L. Rosenstreich. 1982. Defect in macrophage function confers Salmonella typhimurium susceptibility on C3H/HeJ mice. Cell. Immunol. 67:325-333.

25. Parant, M., F. Parant, and L. Chedid, 1977. Inheritance of lipopolysaccharide-enhanced nonspecific resistance to infection and susceptibility to endotoxic shock shock in LPS low responder mice. Infect. Immun. 16:432-438.

26. Plant, J., and A. A. Glynn. 1976. Genetics of resistance to infection with Salmonella typhimurium in mice. L Infect. Dis. 133:72-78.

27. Plant, J., and A. A. Glynn. 1979. Locating salmonella resistance gene on mouse chromosome 1. Clin. Exp. Immunol. 37:1-6.

28. Reed, L. J., and H. Muench. 1938. A simple method of estimating fifty per cent endpoints. Am. J. Hygiene 27:493-497.

45

29. Robson, H. G., and S. I. Vas . 1972. Resistance of inbred mice to Salmonella typhimurium. J. Infect. Dis. 126:378-386.

30. Rosenstreich, D. L. 1985. Genetic control of endotoxin response: C3H/HeJ mice, p. 82-122. In L. J. Berry (ed.), Handbook of Endotoxin: Cellular Biology of Endotoxin, vol. 3. Elsevier, New York.

31. Saheb, S. A., L. Massicotte, and B. Picard. 1980. Purification and characterization of Treponema hyodysenteriae hemolysin. Biochimie 62:779-785.

32. Snedecor, G. W., and W. G. Cochran. 1967. Statistical Methods, p.59-62. 6th Ed. Iowa State University Press, Ames, lA.

33. Stanton, T. B., N. S. Jensen, T. A. Casey, L. A. Tordoff, F. E. Dewhirst, and B. J. Paster. 1991. Reclassification of Treponema hyodysenteriae and Treponema innocens in a new genus, Serpula gen. nov., as Serpula hyodysenteriae comb. nov. and Serpula innocens comb. nov. Lit. J. Systematic Bacteriol. 41:50-58.

34. Streeter, P. R., and L. B. Corbeil. 1981. Gonococcal infection in endotoxin-resistant and endotoxin-susceptible mice. Infect. Immun. 32:105-110.

35. Suenaga, I., and T. Yamazaki. 1984. Experimental Treponema hyodysenteriae infection of mice. Zbl. Bakt. Hyg. A257:348-356.

36. Sueyoshi, M., Y. Adachi, and S. Shoya. 1987. Enteropathogenicity of Treponema hyodysenteriae in young chicks. Zbl. Bakt. Hyg. A266:469-477.

37. Svanborg-Eden, C., R. Shahin, and D. Briles. 1988. Host resistance to mucosal gram negative infection. Susceptibility of lipopolysaccharide nonresponder mice. J. Immunol. 140:3180-3185.

46

38. Waimemuehler, M. J., R. D. Hubbard, and J. M. Greer. 1988. Characterization of the major outer membrane antigens of Treponema hyodysenteriae. Infect. Immun. 56:3032-3039.

39. Wannemuehler, M. J., S. M. Michalek, E. J. Jirillo, S. I. Williamson, M. Hirasawa, and J. R. McGhee. 1984. LPS regulation of the immune response; Bacteroides endotoxin induces mitogenic, polyclonal, and antibody responses in classical LPS responsive but not C3H/HeJ mice. J. Immunol. 133:299-305.

40. Watson, J., R. Riblet, and B. A. Taylor. 1977. The response of recombinant inbred strains of mice to bacterial lipopolysaccharides. J. Immunol. 118:2088-2093.

47

PAPER n.

AN ENHANCED MURINE MODEL FOR STUDIES ON THE

PATHOGENESIS OF Serpulina (Treponemal hyodysenteriae

48

ABSTRACT

A defined diet was utilized to increase the susceptibility of mice to

Serpulina hyodysenteriae. BALB/cByJ, C3H/HeN, and C3H/HeJ mice, when

fed the diet 7 to 14 days prior to and throughout the challenge period,

consistently showed higher incidences of disease than mice maintained on

normal rodent chow. The use of this diet will increase consistency of in

vivo studies following infection with S. hyodysenteriae in the mouse

model.

49

INTRODUCTION

Swine dysentery is a muco-hemorrhagic diarrheal disease of growing

swine, affecting the cecum and colon (5,8). Mortality is primarily a result

of metabolic acidosis and a reduction of water resorbtion in the colon (1).

Although the etiologic agent, Serpulina (Treponema) hyodysenteriae, has

been described (5), the mechanism(s) of pathogenesis is unknown.

Potential virulence determinants of S. hyodysenteriae are an LPS or LPS-

like moiety (2-4,19,20) and a 6-hemolysin (14,21). A murine model was

initially described by Joens and Glock (10), and characterized as a disease

exhibiting catarrhal typhlitis after infection with S. hyodysenteriae.

Histopathological changes in the mouse cecum are similar to those observed

in swine (mucosal edema, hyperemia, crypt dilatation, and epithelial

erosion) (10); however, clinical signs in the mouse are diminished when

compared to swine. The mouse model has been used extensively in

pathogenesis research of S. hyodysenteriae, (7,9-12,16-19,22) but the

inconsistency of disease in the mouse model often makes interpretation of

results difficult, especially involving certain inbred strains of mice (18).

Differences in susceptibility to development of disease have been described

as a result of genetic differences (18), suppliers of mice (22), and the Ips

gene locus of C3H mice (18,19). Previous research described enhanced

susceptibility with oral antibiotic administration (16, M. I. Kennedy, K. R.

Goodenough, C. P. Cornell, C. W. Ford, and R. J. Yancey, Ir., Abstr.

Proc. 90th Am. Soc. Microbiol. 1990, p. 72) or with oral doses of specific

50

anaerobic bacteria (7). In the present study, we describe the use of a diet

which renders resistant mice (e.g., BALB/cByJ) (18) more susceptible to

disease, as well as increasing susceptibility of other inbred (e.g., C3H)

mouse strains.

51

MATERIALS AND METHODS

BALB/cByJ and C3H/HeJ mice were originally procured from

Jackson Laboratories, Bar Harbor, ME. C3H/HeN mice were originally

obtained from Harlan Sprague Dawley, Indianapolis, IN. All mice for this

work were obtained from breeding colonies maintained in the Laboratory

Animal Resource Facility of the College of Veterinary Medicine, Iowa

State University, Ames. The inbred mouse strains used in this study were

selected to compare the effect of the diet on a "resistant" and "susceptible"

mouse strain; it was previously shown (18) that BALB/cByJ mice were

refractory, whereas C3H strains were susceptible, to development of cecitis

after S. hyodysenteriae challenge. All breeding colonies and principals of

these trials were maintained on Mouse Lab Chow #5010 (nutritionally

complete)(Purina Mills, Inc., St. Louis, Mo.) until experimentation. For

the sake of convenience. Mouse Lab Chow will be referred to herein as

"normal diet". Mice were age-matched within each trial and were 3-5

months old at time of experimentation. S. hyodysenteriae strain B204 was

cultivated in broth as previously described (18). Some mice were fed

Teklad Diet TD 85420, a nutritionally complete defined diet for 7 to 14

days prior to and throughout the challenge period. Mice were challenged

on two consecutive days by intragastric intubation of various doses of S.

hyodysenteriae B204 in 1 ml of broth as previously described (18). Mice

were necropsied at 10 to 20 days post-challenge and ceca visualized for

development of grossly evident lesions. Cecal lesions were scored as

52

previously described (18); briefly, a cecum with no grossly evident lesions

was scored a 0; those ceca with excess intraluminal mucus, a 1; ceca

exhibiting excess intraluminal mucus and organ atrophy in the cecal apex

only, a 2; and excess cecal intraluminal mucus and organ atrophy, a 3. To

determine colonization, cecal swabs were cultured on trypticase soy blood

agar plates containing antibiotics (13,18). In some trials the number of S.

hyodysenteriae colony forming units (CPU) per gram cecal weight were

enumerated by serial dilution and pour-plating on selective media (18).

53

RESULTS

Experiment 1 (Table 1) illustrates the susceptibility of C3H/HeN

mice to lesion development from S. hyodysenteriae when maintained on

Teklad diet TD 85420. Mice fed the Teklad diet showed increased group

mean cecal scores (greater severity of cecal lesions) (3.0, 2.6, and 3.0 for

challenge doses 10^, 10^, and 10^, respectively) when compared to

previously published data of C3H/HeN mice similarly challenged on normal

diet (2.1, 1.8, and 1.3 for challenge doses 10^, 10^, and 10"^, respectively)

(18).

To further this observation, we investigated whether the Teklad diet

could make a more resistant mouse strain, C3H/HeJ, more susceptible to

development of lesions. Experiment 2 (Table 1) shows that at challenge

doses of 108,10^, and 10^ spirochetes, the C3H/HeJ mice developed more

consistent lesions (3.0, 3.0, and 2.6 vs. 2.1, 1.2, and 1.4) and higher levels

of CFU/g-cecum values (8.72-8.89 vs. 8.00-6.51) than previously described

for C3H/HeJ mice maintained on normal rodent chow, respectively (18).

In addition, in a challenge dose titration study (Exp 3, Table 1), the

C3H/HeJ mice maintained on the Teklad diet during experimentation were

sensitive to development of cecal lesions even when challenged with as few

as 100 organisms. This is in contrast to previous results (18) which

indicated that the ID50 (infectious-dose which yielded a lesion score of 3

for 50% of challenged mice) for C3H/HeJ mice was 8.3 x 10^ S.

hyodysenteriae B204 when maintained on normal diet.

54

Table 1. Susceptibilities of mice maintained on Teklad diet TD85420 to challenge with S. hyodysenteriae B204

Group Mean Log 10 Exp. Strain Protocol^ Dose^ Cecal Score^ Culture^ CFU/g^

I C3H/HeN 7d 20dPI I06 3.0 dLO (5) 100% 7.95 105 2.6 + 0.4 (5) 100% 8.29 104 3.0 + 0 (5) 100% 8.08

2 C3H/HeJ 14d lOdPI 108 3.0 ±0 (4) 100% 8.72 107 3.0 ±_0 (5) 100% 8.88 106 2,6 + 0.4 (5) 100% 8.89

3 C3H/HeJ 14d lOdPI 105 3.0 ±0 (7) 100% N.D. 104 1.8 + 0.5 (5) 100% N.D. 103 1.8 + 0.7 (5) 100% N.D. 102 2.7 ± 0.3 (6) 100% N.D.

IjMice were maintained on Teklad diet 85420 7 or 14 days prior to and throughout the challenge period with S. hyodysenteriae strain B204. In each experiment, all mice were sacrificed at either 10 or 20 days post­infection (dPI).

^Mice were challenged per os with the indicated dose of log-phase S. hyodysenteriae B204 once a day on 2 consecutive days.

^Group mean cecal scores are based on grossly observable lesions ranging from 0 (no lesions present) to 3 (severe cecal lesions) +/- standard error. Parenthetical numbers indicate group size.

^Percent of individuals in group in which S. hyodysenteriae was re-isolated from cecal contents.

^Estimation of S. hyodysenteriae colonization density by enumerating S. hyodysenteriae colony forming units (CPU) per gram of cecum. Individual values were determined and are expressed as group mean values. N.D.=not determined.

55

The BALB/cByJ inbred mouse strain has been decribed as

refractory to development of lesions after challenge with S. hyodysenteriae

(18). Figure 1 illustrates that BALB/cByJ mice, fed the Teklad diet during

experimentation, become susceptible to lesion development after challenge

with S. hyodysenteriae.

I 8 M

Î H)

percent culture positive

normal normal Teklad Teklad

10®^ 10^ 10*^ 10^ Challenge Dose

Figure 1. Effect of diet on BALB/cByJ susceptibility to challenge with S. hyodysenteriae B204. Mice were maintained on normal feed (Purina Mouse Chow 5010) or Teklad diet TD85420 10 days prior to and throughout the challenge period, were challenged per OS on 2 consecutive days with either 10^ or 10^ bacteria, and were sacrificed 15 days post-challenge. Cecal scores represent mean values ranging from 0 (no grossly evident lesions) to 3 (severe organ atrophy with intraluminal mucus). Percent culture positive values indicate the percentage of individual ceca in each treatment group in which S. hyodysenteriae was re-isolated. The standard error of the mean is indicated. n=group size.

56

BALB/cByJ mice maintained on normal mouse chow diet exhibited

no gross lesions at challenge doses of 10^ or 10^ spirochetes. BALB/cByJ

mice fed the Teklad diet, challenged with 10^ organisms, were much more

sensitive to development of lesions than their experimental counterparts fed

normal diet (Fig. 2). Similarly, C3H/HeJ mice, challenged with 10^

organisms, developed more severe lesions following S. hyodysenteriae

challenge when fed the Teklad diet (Fig. 2).

Since the BALB/cByJ mice fed the Teklad diet appeared more

susceptible to S. hyodysenteriae challenge (Fig. 1 and 2), we titrated the

infective challenge of S. hyodysenteriae for BALB/cByJ mice fed the

Teklad diet (Fig. 3). BALB/cByJ mice developed lesions at all challenge

doses, however, the susceptibility to challenge was less at the lower doses

than that exhibited by the C3H/HeJ mice (Table 1).

57

Percent Culture Positive

1 CJ CO

1 I

BALB/cByJ C3H/HeJ

normal Teklad normal Teklad

Diet

10» 10^ 10^ 10^

Infective Challenge

Figure 2. Effect of diet on BALB/cByJ and C3H/HeJ susceptibility to challenge with S. hyodysenteriae B204. Mice were maintained on normal feed (Purina Mouse Chow 5010) or Teklad diet TD85420 for 7 days prior to and throughout the challenge period, were challenged per os on 2 consecutive days with either 10^ (BALB/cByJ) or 10^ (C3H/HeJ) bacteria, and were sacrificed 10 days post-challenge. Cecal scores represent mean values ranging from 0 (no grossly evident lesions) to 3 (severe organ atrophy with intraluminal mucus). Percent culture positive values indicate the percentage of individual ceca in each treatment group in which S. hyodysenteriae was re-isolated. The standard error of the mean is indicated. n=group size.

58

Percent Culture Positive 100% 100% 83% 100% 67% n=6

l(f 10^ 10^ 10^

Challenge Dose

Figure 3. Susceptibility of BALB/cByJ mice to decreasing challenge doses of S. hyodysenteriae B204. Mice were maintained on Teklad diet TD85420 for 7 days prior to and throughout the challenge period, were challenged per os on 2 consecutive days, and were sacrificed 10 days post-challenge. Cecal scores represent mean values ranging from 0 (no grossly evident lesions) to 3 (severe organ atrophy with intraluminal mucus). Percent culture positive values indicate the percentage of individual ceca in each treatment group in which S. hyodysenteriae was re-isolated. The standard error of the mean is indicated. n=group size.

BALB/cByJ mice (n=4), maintained on Teklad TD85420 for 20 days

were injected intraperitoneally with 150 |xg S. hyodysenteriae whole cell

lysate on day 10. BALB/cByJ mice maintained on normal diet (n=4) were

also similarly immunized. All mice in both groups were sacrificed 10 days

following immunization. Anti-5. hyodysenteriae immunoglobulin

59

responses were not significantly different between the two groups of mice

as determined by ELISA (data not shown). The mice maintained on Teklad

TD85420, however, did show an approximate 4 log increase in gram-

positive organisms recovered from the cecum as estimated by spread plate

counts on phenylethanol agar, and an approximate 2 log increase in gram-

negative organisms as estimated by spread plate counts on EMB agar of

cecal contents. All culture plates were incubated at 37°C aerobically (data

not shown). To determine whether a factor(s) in Teklad TD85420

increased susceptibility of mice to challenge, C3H/HeSnJ and C3H/HeJ

mice, maintained on normal diet, were fed water containing 0.75 M

dextrose ad libitum 1 days prior to and throughout a challenge period with

S. hyodysenteriae B204. At 12 days postinfection, no increase in the

susceptibility to induction of cecal lesions was observed when compared to

mice drinking water not supplemented with dextrose (data not shown).

This would suggest that the inclusion of the major component of the

TD85420 diet (i.e., 63% dextrose) in the water of mice maintained on a

normal diet is insufficient to enhance susceptibility to disease following

infection with S. hyodysenteriae.

60

DISCUSSION

The reason for the increase in susceptibihty to S. hyodysenteriae is

unclear, but the Teklad diet appears to facilitate cecal colonization of S.

hyodysenteriae more readily and possibly to a higher density than a normal

rodent diet. Possible explanations for this hypothesis include alterations of

the cecal microflora, changes in cecal volatile fatty acid levels, or substrate

availability differences in the cecal contents. Cecal microflora differences

have been postulated to play a major role in the susceptibihty of mice to

challenge with S. hyodysenteriae (7,16,22, M. J. Kennedy, K. R.

Goodenough, C. P. Cornell, C. W. Ford, and R. J. Yancey, Jr., Abstr.

Proc. 90th Am. Soc. Microbiol. 1990, p. 72). Previous work with

gnotobiotic pigs (6,15,23) characterized the necessity for a secondary

bacteria to be present in conjunction with S. hyodysenteriae for lesions to

develop. For the mouse model, Joens et al. (11) have shown that S.

hyodysenteriae and Bacteroides vulgatus will induce cecal lesions in

gnotobiotic mice; however, S. hyodysenteriae alone or B. vulgatus alone

would not induce gross lesions. In conventional mice, Suenaga and

Yamazaki (22) reported differing susceptibilities to S. hyodysenteriae of

mice obtained from different suppliers, theorizing that differing gut flora

of mice derived from different breeding colonies (9) was affecting lesion

development. Subsequently, Hayashi et al. (7) described Bacteroides

uniformis as a cecal flora component which facilitates infection of Ta:CF-

1 mice by S. hyodysenteriae. Kennedy et al. (M. J. Kennedy, K. R.

Goodenough, C. P. Cornell, C. W. Ford, and R. J. Yancey, Jr., Abstr.

61

Proc. 90th Am. Soc. Microbiol. 1990, p. 72) described a mouse model of

S. hyodysenteriae infection which was made more reproducible by oral

pretreatment of mice with spectinomycin. Others have also described

increased susceptibility to S. hyodysenteriae infection after oral antibiotic

treatment (16). These results, taken with the present work, indicate the

importance of the character of the resident intestinal microflora in the

development of lesions after challenge with S. hyodysenteriae.

Quite possibly, with shifts in the intestinal microflora, the cecal

levels of volatile fatty acids may also change, which have been shown to

inhibit S. hyodysenteriae growth in vitro (M. J. Wannemuehler, M. H.

Crump, R. Hubbard, and S. K. Nibbelink, Abstr. Proc. 66th Conf.

Research Workers in Animal Dis., 1985, p.25). The Teklad diet TD 85420

contains a high percentage composition of dextrose and egg white solids.

Different levels of microbial substrates could shift the predominant

microbial flora, thus effecting increased susceptibility of the mouse to S.

hyodysenteriae colonization and subsequent lesion development.

This diet, when administered to mice before S. hyodysenteriae

challenge, causes mice to be more susceptible to lesion development. The

use of this diet will result in a more consistent model and will permit the

utilization of genetically diverse strains of inbred mice for the

determination of the pathogenesis of disease caused by S. hyodysenteriae.

62

REFERENCES

1. Argenzio, R. A. 1985. Pathophysiology of swine dysentery, p. 3-19. M C. J. Pfeiffer (ed.), Animal models for intestinal disease. CRC Press Lie., Boca Raton, Fla.

2. Greer, J. M., and M. J. Wannemuehler. 1989. Comparison of the biological responses induced by lipopolysaccharide and endotoxin of Treponema hyodysenteriae and Treponema innocens. Infect. Immun. 57:717-723.

3. Greer, J. M., and M. J. Wannemuehler. 1989. Pathogenesis of Treponema hyodysenteriae: induction of interleukin-1 and tumor necrosis factor by a treponemal butanol/water extract (endotoxin). Microb. Pathog. 7:279-288.

4. Halter, M. R., and L. A. Joens. 1988. Lipooligosaccharides from Treponema hyodysenteriae and Treponema innocens. Infect. Immun. 56:3152-3156.

5. Harris, D. L., R. D. Clock, C. R. Christensen, and J. M. Kinyon. 1972. Swine Dysentery-1. Inoculation of pigs with Treponema hyodysenteriae (new species) and reproduction of the disease. Vet. Med. Small Anim. Clin. 67:61-68.

6. Harris, D. L., T. J. L. Alexander, S. C. Whipp, I. M. Robinson, R. D. Clock, and P. J. Matthews. 1978. Swine dysentery: studies of gnotobiotic pigs inoculated with Treponema hyodysenteriae, Bacteroides vulgatus, and Fusobacterium necrophorum. J. Am. Vet. Med. Assoc. 172:468-471.

7. Hayashi, T., I. Suenaga, T. Komeda, and T. Yamazaki. 1990. Role of Bacteroides uniformis in susceptibility of Ta:CF#l mice to infection by Treponema hyodysenteriae. Zentralbl. Bakteriol. 274:118-125.

63

8. Hughes, R., H. J. Olander, and C. B. Wilhams. 1975. S wine dysentery: pathogenicity of Treponema hyodysenteriae. Am. J. Vet. Res. 36:971-977.

9. Itoh, K., T. Mitsuoka, K. Sudo, and K. Suzuki. 1983. Comparison of fecal flora of mice based upon different strains and different housing conditions. Z. Versuchstierkd. 25:135-146.

10. Joens, L. A., and R. D. Glock. 1979. Experimental infection in mice with Treponema hyodysenteriae. Infect. Immun. 25:757-760.

11. Joens, L. A., I. M. Robinson, R. D. Glock, and P. J. Matthews. 1981. Production of lesions in gnotobiotic mice by inoculation with Treponema hyodysenteriae. Infect. Immun. 31:504-506.

12. Joens, L. A., R. D. Glock, and J. M. Kinyon. 1980. Differentiation of Treponema hyodysenteriae from T. innocens by enteropathogenicity testing in the CFl mouse. Vet. Rec. 107:527-529.

13. Kunkle, R. A., and J. M. Kinyon. 1988. Improved selective medium for the isolation of Treponema hyodysenteriae. J. Clin. Microbiol. 26:2357-2360.

14. Lemcke, R. M., and M. R. Burrows. 1982. Studies on a haemolysin produced by Treponema hyodysenteriae. J. Med. Microbiol. 15:205-214.

15. Meyer, R. C., J. Simon, and C. S. Byerly. 1975. The etiology of swine dysentery. DI. The role of selected gram-negative obligate anaerobes. Vet. Pathol. 12:46-54.

16. Nibbelink, S. K. 1989. M. S. thesis. Iowa State University, Ames, Iowa. pp. 105-115.

17. Nibbelink, S. K., and M. J. Wannemuehler. 1990. Effect of Treponema hyodysenteriae infection on mucosal mast cells and T cells in the murine cecum. Infect. Immun. 58:88-92.

64

18. Nibbelink, S. K., and M. J. Wannemuehler. 1991. Susceptibility of inbred mouse strains to infection with Serpula {Treponema) hyodysenteriae. Infect. Immun. 59:3111-3118.

19. Nuessen, M. E., L. A. Joens, and R. D. Glock. 1983. Involvement of lipopolysaccharide in the pathogenicity of Treponema hyodysenteriae. J. Immunol. 131:997-999.

20. Nuessen, M., E., J. R. Birmingham, and L. A. Joens. 1982. Biological activity of a lipopolysaccharide extracted from Treponema hyodysenteriae. Infect. Inrniun. 37:138-142.

21. Saheb, S. A., L. Massicotte, and B. Picard. 1980. Purification and characterization of Treponema hyodysenteriae hemolysin. Biochimie 62:779-785.

22. Suenaga, L, and T. Yamazaki. 1984. Experimental Treponema hyodysenteriae infection of mice. Zentralbl. Bakteriol. Mikrobiol. Hyg. Reihe A 257:348-346.

23. Whipp, S. C., I. M. Robinson, D. L. Harris, R. D. Glock, P. J. Matthews, and T. J. L. Alexander. 1979. Pathogenic synergism between Treponema hyodysenteriae and other selected anaerobes in gnotobiotic pigs. Infect. Immun. 26:1042-1047.

65

PAPER m.

PATHOGENESIS OF Serpulina hyodysenteriae: IN VIVO INDUCTION

OF TUMOR NECROSIS FACTOR BY A SERPULINAL

BUTANOL/WATER EXTRACT (ENDOTOXIN)

66

INTRODUCTION

Serpulina (Treponemal hyodysenteriae is the causative agent of

swine dysentery, a muco-hemorrhagic colitis of growing swine (1).

Morbidity of the disease is high in naive herds, with mortality resulting

from dehydration due to loss of resorbtive fimction of the colon (2). A

mouse model of the disease has been described (3), and mice exhibit a

catarrhal cecitis after oral challenge with the spirochete.

Serpulina hyodysenteriae is described as a strongly 6-hemolytic (1),

anaerobic, gram negative spirochete which can consume oxygen (4). A

non-pathogenic porcine spirochete, Serpulina (Treponema) innocens, also

exists as a component of the normal colonic flora and is distinguished from

S. hyodysenteriae by weak 6-hemolysis on blood agar (1). Although the

pathogenesis of the disease is unknown, putative virulence determinants of

S. hyodysenteriae may include multiple fi-hemolysin(s) (5,6,26) and a

lipopolysaccharide (LPS) or LPS-like moiety (7-9). A phenol/water

extract of S. hyodysenteriae has been used to serotype isolates of S.

hyodysenteriae (10) and has been shown to possess in vitro biologic

activities, including toxicity for mouse peritoneal macrophages,

enhancement of C3 and IgG-Fc mediated phagocytosis, mitogenicity for

murine splenocytes, and stimulation of chemotaxis of porcine leukocytes

(7). Greer and Wannemuehler (11) have shown that S. hyodysenteriae

endotoxin (butanol/water extract) posesses more biologic activity than LPS

(phenol/water) and that the S. hyodysenteriae butanol/water preparation

67

contains comparable endotoxin activities as Escherichia colt LPS as assayed

by the Limulus amebocyte assay. In addition, S. hyodysenteriae endotoxin

did induce production of tumor necrosis factor (TNF) and interleukin-1

(IL-1) activity by murine peritoneal exudate cells in vitro (9). Although S.

hyodysenteriae endotoxin does exhibit in vitro biological activities, it does

not induce dermal Shwartzman reactions or pyrogenic responses in rabbits

(11).

In vivo, S. hyodysenteriae endotoxin has been implicated in the

pathogenesis of the disease, as C3H/HeJ inbred mice, which possess a defect

in the Ips gene and are hyporesponsive to LPS, are more resistant to S.

hyodysenteriae challenge than are Ips responsive C3HeB/FeJ or C3H/HeN

inbred mice (8,12). Following challenge with Salmonella spp., LPS

hyporesponsive C3H/HeJ mice are more susceptible to disease and mortality

than are LPS responsive C3H mice (13). The apparent discrepancy

between LPS response and the above two diseases may be explained by the

mucosal nature of S. hyodysenteriae infection. The pathogenesis of S.

hyodysenteriae is strictly confined to the mucosa of the pig (14) (as well as

the mouse (3)), and does not involve a systemic component, as does

Salmonella. Fibrino-necrotic colitis in swine has been likened to a dermal

Shwartzman reaction (15), and if so, tumor necrosis factor in concert with

interleukin-1 (IL-1) and gamma interferon (IFN-gamma) would be

instrumental in the development of mucosal lesions associated with S.

hyodysenteriae (16). In this situation, LPS-hyporesponsive C3H/HeJ mice

would be more refractory to development of lesions after challenge with S.

hyodysenteriae.

68

The current study was designed to determine whether or not the

administration of S. hyodysenteriae LPS or endotoxin in vivo would induce

detectable levels of TNF activity in the semm of swine or mice. The

relative levels of TNF activity detected and the dose of stimulant required

was compared with the levels of TNF activity produced by S. typhimurium

or E. coli LPS. These studies will further investigate the potential role of

the spirochetal endotoxin in the induction of colonic mucosal inflammatory

responses in swine dysentery.

69

RESULTS

Stimulation of serum TNF activity in mice after intra-peritoneal

administration of serpulinal preparations

After intraperitoneal (i.p.) injection of E. coli or S. typhimurium

LPS, mouse sera exhibited increased TNF activity beginning at 30 minutes

post-administration and continuing to 120 minutes or longer (Fig. 1). At

the high dose of E. coli LPS (10 jig/mouse) TNF activity in the serum

remained elevated (compared to control values) to the termination of the

experiment (360 min.). The lower dose of E. coli LPS (1 |Xg/mouse)

elicited comparably increased serum TNF activity from 30 to 120 min.,

after which TNF activity dropped to control levels. Both S. typhimurium

LPS doses (10 and 1 |ig/mouse) caused peak serum TNF concentrations at

90 minutes, and then decreased to comparable control values at later time

points.

Serpulina hyodysenteriae endotoxin (butanol/water) and LPS

(phenol/water) preparations also elicited increased serum TNF activity in mice

(Fig. 2). The doses of S. hyodysenteriae endotoxin (10 and 100 |ig/mouse)

caused a peak of serum TNF at 90 minutes, after which serum TNF levels fell to

control values. S. hyodysenteriae LPS showed less ability to increase semm TNF

in mice, supporting the finding of less bioactivity of S. hyodysenteriae LPS when

compared to S. hyodysenteriae endotoxin (11). Figure 3 shows serum TNF in

mice after administration of endotoxin from S. innocens. It was previously

70

reported that S, innocens endotoxin had in vitro bioactivity (11), and Fig. 3

shows that, at doses of 10 and 100 |i.g/mouse, S. innocens endotoxin can cause a

very transient peak (30-60 min.) of serum TNF activity.

100

ô 80 -X o

>> o

c 0) u 0) a

60

^ 40

20 -

—O— Control # Ecoli 1ng •— E coli 10fig

Salm. lug Salm. lOng

—1— 30

—T— 60

—r-90

T T I

120 180 360

Time (minutes)

Figure 1. Levels of serum tumor necrosis factor activity in mice after intraperitoneal injection of E. coli and S. typhimurium LPS at various doses. Data is presented as the mean percentage cytotoxicity of L929 cells from individual sera (1:50 dilution) collected at the indicated time point as described in Materials and Methods. Group sizes: for control mice, n=8, 7, 7, 8, 7, 7 for time points 30, 60, 90, 120, 180, and 360, respectively; all other data points, n=3 for each time point.

71

100

'3 X o

O

c u u 0) a

80

60

40

20

0 —r-30

—1— 60

—o— Control • B204P/W10fig Q B204P/W100ng •0-- B204 Bu/WIOng

B204 Bu/W lOOng

90 I

120 180 360

Time (minutes)

Figure 2. Levels of serum tumor necrosis factor activity in mice after intraperitoneal injection of S. hyodysenteriae B204 LPS (P/W) or endotoxin (Bu/W) at various doses. Data is presented as the mean percentage cytotoxicity of L929 cells from individual sera (1:50 dilution) collected at the indicated time point as described in Materials and Methods. Group sizes; for control mice, n=8, 7, 7, 8, 7, 7 for time points 30, 60, 90, 120, 180, and 360, respectively; all mice receiving B204 Bu/W doses, n=6 for each time point; all mice receiving B204 P/W doses, n=3 for each time point.

72

100 1

Control B256Bu/W10|ig B256 Bu/W 100ng

0 30 60 90 120 180 360

Time (minutes)

Figure 3. Levels of serum tumor necrosis factor activity in mice after intraperitoneal injection of S. innocens B256 endotoxin at various doses. Data is presented as the mean percentage cytotoxicity of L929 cells from individual sera (1:50 dilution) collected at the indicated time point as described in Materials and Methods. Group sizes: for control mice, n=8, 7, 7, 8, 7, 7 for time points 30, 60, 90, 120, 180, and 360, respectively; all mice receiving B256 Bu/W doses, n=3 for each time point.

73

'3 x o

O

c 0) u 0) a

100 -

90 -

8 0 -

70

60 -

5 0 -

40

3 0 -

2 0 -

10

0 H

-10

—o— Control •— E. coli 20fig/kg D E coli 10O^g/kg

Salm. 20^g/kg Salm. lOOng/kg

0 30 60 —r-90 120

Time (minutes)

I 180 3 6 0

Figure 4. Levels of serum tumor necrosis factor activity in pigs after intravenous injection of E. coli and Salmonella LPS at various doses. Data is presented as the mean percentage cytotoxicity of L929 cells from individual sera collected at the indicated time point as described in Materials and Methods. Group sizes: control pigs, n=8 for each time point; E. coli 20 |ig/kg, n=3 for each time point; E. coli 100 |ig/kg, n=6 for each time point; Salmonella LPS (both doses), n=3 for each time point.

74

50 -I

Control B2Q4 P/W 500ng/kg B204 Bu/W 10O^g/kg B204 Bu/W 500ng/kg

> 40 -

1 3 0 -o >. O 20 -

c 0) u 10 -

0) Q.

30 6 0 9 0 1 2 0 1 8 0 3 6 0 0

Time (minutes)

Figure 5. Levels of serum tumor necrosis factor activity in pigs after intravenous injection of S. hyodysenteriae B204 LPS (P/W) and endotoxin (Bu/W) at various doses. Data is presented as the mean percentage cytotoxicity of L929 cells from individual sera collected at the indicated time point as described in Materials and Methods. Group sizes: control pigs, n=8 for each time point; B204 P/W 500 |ig/kg, n=3 for each time point; B204 Bu/W at both doses, n=6 for each time point.

75

50

> 40 -

o o > O

u x

30 -

•Ct Control 8256 Bu/W 10Ong/kg B256 Bu/W 500fig/kg

2 0 -

c 0) u 10 -

o a. 0 -

-10

0 3 0 6 0 9 0 1 2 0 1 8 0 3 6 0

Time (minutes)

Figure 6. Levels of serum tumor necrosis factor activity in pigs after intravenous injection of S. innocens B256 endotoxin (Bu/W) at various doses. Data is presented as the mean percentage cytotoxicity of L929 cells from individual sera collected at the indicated time point as described in Materials and Methods. Group sizes: control pigs, n=8 for each time point; B256 Bu/W at both doses, n=3 for each time point.

76

Stimulation of serum TNF activity in pigs after administration of

serpulinal preparations

Pigs which received either E. coli or S.typhimurium LPS at doses of either

20 or 100 |ig/kg responded by elevating serum TNF activity, peaking at

30-120 min. (Fig. 4). Serum levels of TNF in pigs which received the E.

coli or S. typhimurium LPS preparations decreased to control levels at the

termination of the experiment (360 min.). In contrast, pigs injected with S.

hyodysenteriae endotoxin (100 and 500 |Lig/kg) or LPS (500 jXg/kg)

developed little or no increases in serum TNF activity (Fig. 5). Pigs

injected with S. innocens endotoxin (100 or 500 |ig/kg) also did not exhibit

elevated levels of serum TNF activity (Fig. 6). However, pigs receiving

the high doses of S. hyodysenteriae, S. typhimurium, and E. coli

LPS/endotoxin preparations did exhibit clinical signs of endotoxemia

(vomiting, depression, diarrhea) within the collection period; 2 pigs died

in the S. typhimurium LPS group (high dose). Interestingly, pigs

receiving S. hyodysenteriae endotoxin also exhibited clinical signs (i.e.,

vomiting, depression, diarrhea) but did not show elevated serum TNF

activity (Fig. 5).

To further characterize the serum factor (in mice and pigs injected

with stimulants) that exhibited cytotoxicity for L929 cells, selected animal

sera was supplemented with either anti-human- or anti-murine-TNF

antibody. Reduction in L929 cytotoxicity in mice was observed in samples

supplemented with anti-mouse-TNF antibody (Table 1). In addition, pig

sera supplemented with either anti-mouse-TNF or anti-human-TNF

77

antibody also decreased L929 cytotoxicity (Table 1). These results further

the conclusion that in vivo stimulation with endotoxin/LPS preparations in

mice and pigs leads to increases in serum TNF.

78

Table 1. TNF-like activity in serum after treatment of serum samples with

anti-tumor necrosis factor antibody

Animal^ Control^ Anti-Murine TNF^ Anti-Human TNF^

207 91* 0 81 208 76 0 0 609 93 45 89

372 79 0 0 374 73 24 0 373 63 1 0

^ animal numbers 207, 208, and 609 were mice treated with 100|a.g S. hyodysenteriae B204 endotoxin (#207 and 208) and 10|ig S. typhimurium LPS (#609) at 90 minutes post-injection. Animal numbers 372, 374, and 373 were pigs treated with 100 fig/kg S. typhimurium LPS (# 372 and 374) and 100|ig/kg E. coli LPS (#373) at 90 minutes post-injection.

b sera from animals injected with stimulants and assayed for TNF-like activity by an L929 assay as described in Materials and Methods,

c sera from animals injected as described above but supplemented with 15 |j,l/ml anti-murine TNF.

^ sera from animals injected as described above but supplemented with 25 Hl/ml anti-human TNF.

* Tumor necrosis factor-like activity is quantitated as percent cytotoxicity (mean of sample values performed in duplicate).

79

DISCUSSION

Swine dysentery manifests as a cecal and colonic mucosal

inflammmatory disease. The disease is localized to the mucosal surface and

lamina propria (17); Serpulina do not invade deeper than the mucosa or

become systemic. Some of the earliest histopathological changes noted in

the pathogenesis of swine dysentery include microthrombi formation and

inflammatory cell infiltration in lamina propria tissues (17). Tumor

necrosis factor has been shown to have many pro-inflammatory activities,

including pro-coagulant activity and toxicity for vascular endothelium

(18,19). As inflammatory mediators (i.e., TNF) have been shown to be

produced in vitro in response to S. hyodysenteriae endotoxin (9), this work

addressed the production of TNF in both mice and pigs after administration

of various serpulinal endotoxin or LPS preparations.

The mouse model, which has been used in research on S.

hyodysenteriae (3,8,12), responded to parenteral administration of

serpulinal endotoxin preparations by the production of serum TNF activity.

TNF production was elicited by both S. hyodysenteriae and S. innocens

endotoxin preparations (Fig. 2 & 3). Serum TNF activity was less in mice

receiving S. hyodysenteriae LPS than mice receiving S. hyodysenteriae

endotoxin (Fig. 2), which agrees with published in vitro data showing S.

hyodysenteriae LPS preparations possessing less bioactivity than endotoxin

preparations (11). While S. hyodysenteriae and S. innocens endotoxin

preparations did stimulate serum TNF activity in mice, serpulinal

80

endotoxin was at least 10 times less potent than the LPS preparations of E.

coli and S. typhimurium on a weight basis for comparable activity. The

observation of less in vivo potency of serpulinal endotoxin preparations to

elicit serum TNF activity also agrees with previous research in which LPS

and endotoxin from S. hyodysenteriae was 100- to 500-fold less toxic for

mice than E. coli LPS (11).

Serum from pigs contained less TNF activity (in the L929 bioassay)

after injection of endotoxin/LPS than did the mice. Murine sera were

assayed for TNF activity using a 1:50 serum dilution; porcine sera were

diluted 1:10 (Figs. 1-3 and 4-6, respectively).

We have tested the sensitivities of both L929 and WEHI 164 cells

(another TNF-sensitive fibroblast cell line), and have found that both cell

lines exhibit comparable sensitivities to porcine TNF (unpublished data).

Pigs appeared to constituitively produce less serum TNF activity than mice

(control lines, Figs. 1-3 and 4-6). Mice in some high dose groups showed

clinical signs of endotoxemia (depression, diarrhea) during the collection

period, however, no mortality resulted. Pigs, however, showed clinical

signs of endotoxemia (diarrhea, vomiting, and depression) including 2

mortalities during the first 6 hours (high dose Salmonella LPS) and 1

mortality after the collection period (high dose E. coli LPS). These

findings insinuate some form of in vivo response to the administered

endotoxin or LPS, whether increases in serum TNF were evident (i.e., S.

typhimurium or E. coli LPS) or not {S. hyodysenteriae preparations).

Pigs did exhibit a serum TNF activity increase after administration

of E. coli and Salmonella LPS (Fig. 4). Response to S. hyodysenteriae

81

endotoxin or LPS was confined to a small peak in serum TNF at 30 min.

post-administration (Fig. 5). The low serum TNF activity in pigs injected

with serpulinal endotoxin might reflect a low bioactivity of S.

hyodysenteriae endotoxin in porcines. Paired serum samples (pre-infection

and during various stages of clinical swine dysentery) of pigs

experimentally infected with S. hyodysenteriae tend to show increased

levels of serum TNF activity with disease (data not presented). In this

situation, the stimulus for the increased serum TNF activity might result

from the spirochete, or from endotoxin released from normal flora in the

lumen. Gnotobiotic pigs or mice will not develop clinical gross lesions

indicative of swine dysentery when mono-associated with S. hyodysenteriae

alone (20-22). However, when specific secondary bacteria are bi-

associated with S. hyodysenteriae, clinical disease and gross lesions are

observed (20,23,24). In gnotobiotic mice mono-associated with S.

hyodysenteriae, spirochetes were observed invading the cecal lamina

propria, but no inflammatory changes were observed (25). This may

indicate that bacterial cell wall components released from secondary

bacteria are important in the induction of an inflammatory response

associated with enteritis.

The current study shows that, in the mouse, S. hyodysenteriae and S.

innocens endotoxin can elicit increases in serum TNF activity. A slight

increase in serum TNF was observed in pigs treated with 5. hyodysenteriae

extracts, and was less than that induced by E. coli or S.typhimurium LPS.

Although S. hyodysenteriae endotoxin did cause clinical signs of

endotoxemia, only small amounts of serum TNF activity could be detected

82

in pigs. Taken in toto, these results show that serpulinal endotoxins are

biologically active in mice in vivo by inducing increases in serum TNF

activity, but are probably not the sole agent(s) inducing the inflammatory

responses associated with swine dysentery.

83

MATERIALS AND METHODS

Animals. CD-I mice were obtained from Jackson Laboratory (Bar

Harbor, ME). Mice were maintained on acidified water and Mouse Lab

Chow #5010 (Purina Mills, St. Louis, MO) in the Laboratory Animal

Resources Facility , College of Veterinary Medicine, Iowa State University,

Ames, Iowa. Mice were age-matched and averaged 30 grams/mouse. Two

trials were performed with mice. Conventional, outbred growing pigs

were obtained through Laboratory Animal Resources, Iowa State

University. All pigs were culture negative for S. hyodysenteriae by rectal

swabs. Two trials were performed with pigs, with each trial having

individual pig weights within one standard deviation of the group mean.

Stimulants. Lipopolysaccharide from Escherichia coli K235 and

Salmonella typhimurium (both phenol purified) was obtained from Sigma

Chemical Co., St. Louis, MO. Endotoxin from Serpulina hyodysenteriae

B204 or S. innocens B256 was prepared by a butanol/water extraction

procedure from washed broth-grown cells (11). Lipopolysaccharide from

S. hyodysenteriae was extracted from washed broth cultures of strain B204

by a modified hot phenol/water method (11). Solutions of endotoxins or

LPS s were dissolved in pyrogen-free saline immediately before

experimentation from lyophilized stock. Prior to use, all preparations

were heated for 10 minutes at 100°C.

Protocol. All mice were injected at time zero i.p. with the

appropriate stimulant suspended in pyrogen-free saline (1 ml). Control

84

mice received an i.p. injection of pyrogen-free saline alone (1 ml).

Beginning at 30 minutes post-injection, blood samples were obtained by

retro-orbital puncture. All mice were bled twice; i.e., the mice bled at 30

minutes were also bled at 120 minutes, the mice bled at 60 minutes were

also bled at 180 minutes, and mice bled at 90 minutes were also bled at 360

minutes. Sera was collected by centrifugation and stored at -70°C until

assayed for TNF activity. For pigs, pre-inject serum samples were

obtained, and then pigs were injected with stimulants (2.5 ml) via anterior

vena cava puncture. Blood samples were collected at each time point (0-

360 min) from each pig by retro-orbital puncture. Serum was harvested

by centrifugation and stored at -70°C until assayed for TNF activity.

Tumor necrosis factor assay. Murine L929 cells were grown in

minimal essential medium (MEM) supplemented with 3.7% NaHCOg, 25

units/ml penicillin, 25 |ig/ml streptomycin, 2mM L-glutamine, and 10%

horse serum (Hyclone, Logan, Utah). Cells were detached, washed in

complete media, and resuspended in complete medium to 5 x 10^ cells/ml.

One tenth ml of cells was added to each well of a 96 well microtiter plate

and incubated overnight at 37°C in 5% C02. One tenth ml of serum

diluted (1:10 for swine, 1:50 for mice) in complete medium plus

actinomycin D (3.0 p-g/ml) was added to wells of L929 monolayers for a

1:1 dilution. The assay was terminated at 18 h after addition of serum

dilutions by adding 50 |il of formalin/well. After 15 min, non-adherent

cells were removed from the plates by washing extensively in saline

solution. Residual monolayers were stained with a 0.2% crystal violet-20%

formalin solution for 20 minutes. Plates were washed with tap water five

85

times to remove excess stain. Crystal violet was solubilized with 50%

ethanol (100 |il/well) and quantitated by measuring absorbance at 595 nm

on an ELIS A reader (model EL301, Biotek Instruments, Winooski, VT).

Sera dilutions were tested in triplicate wells and the results are expressed as

the mean percent cytotoxicity for each group.

Anti-tumor necrosis factor antibodies. Rabbit anti-murine and rabbit

anti-human polyclonal antibodies were obtained from Genzyme, Boston,

MA.

86

REFERENCES

1. Harris DL, Glock RD. Swine dysentery. M: Léman AE, Glock RD, Mengeling WL, Penny RHC, Scholl E, Straw B, eds. Diseases of swine. 5th ed. Ames, lowa: lowa State University Press, 1981; 432-44,

2. Argenzio RA. Pathophysiology of swine dysentery. M: Pfeiffer CJ, ed. Animal models for intestinal disease. Boca Raton, Florida: CRC Press, Inc, 1985; 3-19.

3. Joens LA, Glock RD. Experimental infection in mice with Treponema hyodysenteriae. Infect Immun 1979; 25:757-69.

4. Stanton TB. Glucose metabolism and NADH recycling by Treponema hyodysenteriae, the agent of swine dysentery. Appl Environmental Microbiol 1989; 55:2365-71.

5. Lemcke RM, Burrows MR. Studies on a haemolysin produced by Treponema hyodysenteriae. J Med Microbiol 1982; 15:205-14.

6. Saheb SA, Massicotte L, Picard B. Purification and characterization of Treponema hyodysenteriae hemolysin. Biochimie 1980; 62:779-85.

7. Nuessen ME, Birmingham JR, Joens LA. Biological activity of a lipopolysaccharide extracted from Treponema hyodysenteriae. Infect Immun 1982; 37:138-142.

8. Nuessen ME, Joens LA, Glock RD. Involvement of lipopolysaccharide in the pathogenicity of Treponema hyodysenteriae. J Immunol 1983; 131:997-99.

9. Greer JM, Wannemuehler MJ. Pathogenesis of Treponema hyodysenteriae: induction of interleukin-1 and tumor necrosis factor by a treponemal butanol/water extract (endotoxin). Microbial Pathogenesis 1989; 7:279-88.

87

10. Baum DH, Joens LA. Serotypes of beta-hemolytic Treponema hyodysenteriae. Infect Immun 1979; 25:792-6.

11. Greer JM, Waimemuehler MJ. Comparison of the biological responses induced by lipopolysaccharide and endotoxin of Treponema hyodysenteriae or Treponema innocens. Infect Immun 1989; 57:717-23.

12. Nibbelink SK, Wannemuehler MJ. Susceptibility of inbred mouse strains to infection with Serpula {Treponema) hyodysenteriae. Infect Immun 1991; 59:3111-18.

13. Rosenstreich DL. Genetic control of endotoxin response: C3H/HeJ mice. In: Berry LJ (ed). Handbook of endotoxin, vol. 3. New York: Elsevier, 1985; 82-122.

14. Hughes R, Olander HJ, Williams CB. Swine dysentery: pathogenicity of Treponema hyodysenteriae. Am J Vet Res 1975; 36:971-77.

15. Nordstoga K. Fibrinous colitis in swine, a manifestation of Shwartzman reaction? Vet Rec 1973; June 30:698.

16. Billiau A. Gamma-interferon: the match that lights the fire? Immunol Today 1988; 9:37-40.

17. Albassam MA, Olander HJ, Thacker HL, Turek JJ. Ultrastructural characterization of colonic lesions in pigs inoculated with Treponema hyodysenteriae. Can J Comp Med 1985; 49:384-90.

18. Beutler B, Cerami A. Cachectin: more than a tumor necrosis factor. New Engl J Med 1987; 316:379-85.

19. Beutler BA, Milsark IW, Cerami A. Cachectin/tumor necrosis factor: production, distribution and metabolic fate in vivo. J Immunol 1985; 135:3972-7

20. Joens LA, Robinson IM, Glock RD, Matthews PJ. Production of lesions in gnotobiotic mice by inoculation with Treponema hyodysenteriae. Lifect Immun 1981; 31:504-506.

88

21. Meyer RC, Simon J, Byerly CS. The etiology of swine dysentery, I. Oral inoculation of germ-free swine with Treponema hyodysenteriae and Vibrio coli. Vet Path 1974; 11:515-26.

22. Brandenburg AC, Miniats OP, Geissinger HD, Ewert E. Swine dysentery: inoculation of gnotobiotic pigs with Treponema hyodysenteriae and Vibrio coli and a Peptostreptococcus. Can J Comp Med 1977; 41:294-301.

23. Harris DL, Alexander TJL, Whipp SC, Robinson IM, Glock RD, Matthews PJ. Swine dysentery: studies of gnotobiotic pigs inoculated with Treponema hyodysenteriae, Bacteroides vulgatus, and Fusobacterium necrophorum. J Am Vet Med Assoc 1978; 172:468-71.

24. Whipp SC, Robinson IM, Harris DL, Glock RD, Matthews PJ, Alexander TJL. Pathogenic synergism between Treponema hyodysenteriae and other selected anaerobes in gnotobiotic pigs. Infect Lnmun 1979; 26:1042-47.

25. Nibbelink SK, Morfitt DC, Wannemuehler MJ. Comparative effects of Treponema hyodysenteriae and Treponema innocens infection in germfree mice. 1990; In: Proceedings, 71st Ann Mtg Conf Research Workers in Animal Disease, Chicago, IL, p 14.

26. ter Huume, A. A. H. M., M. van Houten, S. Muir, J. G. Kusters, B. A. M. van der Zeijst, and W. Gaastra. Inactivation of a Serpula (Treponema) hyodysenteriae hemolysin gene by homologous recombination: importance of this hemolysin in pathogenesis of S. hyodysenteriae in mice. F. E. M. S. Microbiol. Letters 1992; 92:109-114.

89

PAPER IV.

ABROGATION OF CECAL LESIONS, BUT NOT

COLONIZATION, IN CONVENTIONAL MICE EXPERIMENTALLY

CHALLENGED WITH Serpulina hyodysenteriae AND MAINTAINED

ON ORAL ANTIBIOTICS

90

ABSTRACT

CSH/HeSnJ and BALB/cByJ mice were infected per os with

Serpulina (Treponema) hyodysenteriae. Mice maintained on a mixture of

antibiotics (spiramycin, spectinomycin, vancomycin, colistin, and

rifampicin) in the drinking water showed no signs of gross cecal lesion

development at 5, 10, or 15 days post infection. Conversely, mice

maintained on normal water (no antibiotics) developed gross cecal lesions

as early as 5 days post-infection. Despite the differences in development of

cecal lesions, levels of S. hyodysenteriae~co\ony forming units recovered

from ceca were comparable for both treatment groups. This observation

was similar to that observed in gnotobiotic mice or pigs infected with S.

hyodysenteriae', that is, monoassociation of gnotobiotic pigs or mice with S.

hyodysenteriae does not induce gross lesions or clinical disease, whereas

biassociation of S. hyodysenteriae plus one of other defined anaerobic

bacteria will cause lesion formation and clinical disease. The present work

demonstrated that conventional animals, depleted of gut flora components

by oral antibiotic treatment, do not develop cecitis following S.

hyodysenteriae challenge and colonization. It appears that secondary

bacteria perform a dynamic function in the development of S.

hyodysenteriae-mduced cecitis, and that the absence of lesions in

monoassociated S. hyodysenteriae-mÎQcted gnotobiotic mice is not

artifactually due to the immunologic immaturity of a gnotobiotic animal.

91

INTRODUCTION

Serpulina hyodysenteriae, the causative agent of swine dysentery

(10), is a gram-negative facultative anaerobic spirochete, consuming

substrate levels of oxygen (30). The dysenteric condition in growing pigs

is characterized as a fulminating muco-hemorrhagic diarrhea (9),

frequently leading to death by dehydration and acidosis (2).

Mucosal invasion of the cecum or large bowel does not appear

necessary for lesion development (14); however, large numbers of

spirochetes are found in distended goblet cells in crypt areas (1). The need

for tissue attachment by the spirochete to cause lesions is not clear, and a

cytotoxin has been postulated to be involved in the pathogenesis of the

disease (18). Virulence factors of the bacteria include biologically active

endotoxin (7, 8, 25, 26) and a secreted beta-hemolysin which has in vitro

cytocidal properties (17, 27, 31).

A mouse model has been described for the disease (12), with

histopathological changes being similar to those observed in swine.

Whereas pigs develop a mucohemorrhagic diarrhea with lesion

involvement of the cecum and large bowel, the mouse develops a catarrhal

cecitis with slight diarrhea (12, 22).

Using germfree pigs, it was shown that monoassociation with S.

hyodysenteriae would not cause clinical disease or development of typical

swine dysentery lesions (5, 21). Infection of germfree pigs with S.

hyodysenteriae plus certain anaerobes was demonstrated to induce disease

and gut lesions (11, 20, 32). It was hypothesized that secondary anaerobic

92

bacteria in some fashion allowed the spirochete to fully express virulence

factors or to colonize to a higher density (32). There is evidence which

supports the ability of S. hyodysenteriae alone to induce lesions in

gnotobiotic pigs (33), as well as evidence that secondary bacteria are

required for the induction of lesions and disease in gnotobiotic pigs (11,

20, 32).

Using germfree mice (13) challenged with S. hyodysenteriae, it was

shown that gross lesions (cecitis) only developed in mice infected with S.

hyodysenteriae and Bacteroides vulgatus. Gnotobiotic mice monoassociated

with 5. hyodysenteriae were observed to have spirochetes within cecal

lamina propria by 10 days post infection (DPI) out to 115 DPI, with

minimal inflammatory response observed (24). We investigated the

hypothesis that the inability of a monoassociated gnotobiotic animal

infected with S. hyodysenteriae to develop lesions is due to an immature

inflammatory response or an inability to react with S. hyodysenteriae

endotoxin or LPS, which has been shown to have decreased bioactivity

when compared to E. coH or Salmonella LPS (7). It has been shown that

germfree mice exposed to gram-negative bacteria are more susceptible to

lethal effects of endotoxin (28). If the above hypothesis were true (i.e., if

the absence of lesions ma.S. hyodysenteriae-monoâssociatQd animal is

artifactual due to inherent physiological constraints), conventional, flora-

depleted mice would develop lesions after S. hyodysenteriae challege. The

present work shows that conventional, flora-depleted mice do not develop

gross cecal lesions after S. hyodysenteriae challenge, and suggest an active

93

role for secondary bacteria in disease pathogenesis not restricted to the

gnotobiotic model.

94

MATERIALS AND METHODS

Mice BALB/cByJ and C3H/HeSnJ mice were originally procured

from Jackson Laboratories, Bar Harbor, ME., and maintained as breeding

colonies at the Laboratory Animal Resource Facility, College of Veterinary

Medicine, Iowa State University. Mice were maintained on autoclaved

Mouse Lab Chow #5010 (Purina Mills, St. Louis, MO.) until time of

experimentation, when diet was changed to a defined but nutritionally

complete diet, Teklad TD 85420 (Teklad diets, Madison, WI). Mice were

approximately 12 weeks of age at the start of experimentation.

Antibiotic solutions Drinking water for oral antibiotic treatment

was supplemented with the following five antibiotics originally described in

selective media for S. hyodysenteriae (16) (Sigma Chemical Company, St.

Louis, MO): Colistin, 6.2 |ig/ml. Vancomycin, 6 |lg/ml. Spiramycin, 15.2

Hg/ml, Rifampicin, 12.5 |J.g/ml, and Spectinomycin, 200 |J.g/ml. Antibiotic

drinking water was changed every 2-3 days during the trial.

Bacteria S. hyodysenteriae B204 was cultivated in trpticase soy

broth supplemented with yeast extract, VPI salts, cysteine, and equine

serum as described previously (22). Highly motile spirochete cultures,

enumerated by Petroff-Hauser counting chamber, were utilized for

challenge inocula.

Challenge procedure All mice were fed Teklad 85420 diet

beginning 6 days prior to and throughout the challenge period to increase

susceptibility of the mice to S hyodysenteriae challenge (23). Antibiotic-

treated mouse groups recieved antibiotic water beginning at 6 days prior to

95

and throughout the challenge period. Mice were challenged once a day for

two consecutive days with 1 x 10^ log phase S. hyodysenteriae B204 (less

than 20 in vitro passages) by intragastric intubation (22). Concurrent with

the challenge period, the mice were fasted for 36 hours.

Treatment groups Mice were predominately assigned to either an

oral antibiotic treated group (antibiotics in the water) or to a control

infected group (normal water). Mice in these groups were necropsied at 5,

10, and 15 DPI. Groups of mice which received oral antibiotics but were

not infected (non-challenged treatment controls) were necropsied at 10 DPI

to ascertain cecal changes induced by antibiotics alone. In addition, mice

which were challenged and received antibiotics through day 15 PI had their

antibiotic water replaced with normal water for 15 additional days to

determine if the removal of antibiotics would facilitate lesion development.

Necropsy procedure Mice in treatment groups were sacrificed at

5, 10, 15, and 30 DPI. Ceca were examined for gross lesions and scored as

described previously (22): 0, no grossly evident lesions; 1, excess

intraluminal mucus within the cecal lumen; 2, excess intraluminal mucus

and organ atrophy in the apex; and 3, excess intraluminal mucus and severe

organ atrophy. Cecal tips were fixed in neutral buffered formalin for

histologic exam, and ceca and contents were individually weighed in

Whirlpak bags and blended in a Stomacher laboratory blender in a 1:100

w:v dilution in PBS. S. hyodysenteriae colony forming units (CFU) were

quantitated in a pour plate assay (22) using selective antibiotics for S.

hyodysenteriae (16).

96

Estimation of cecal microflora shifts Two treatment groups

(one receiving antibiotic water, one receiving normal water) of non-

infected mice were utilized to ascertain shifts in cecal microflora.

Antibiotic water was provided to the appropriate group for 6 days prior to

necropsy. Both groups of mice received Teklad TD 85420 diet during the

experimentation period. At necropsy, individual ceca were removed,

weighed, homogenized, and diluted in sterile PBS. Spread plate counts of

gram-negative bacteria were estimated by utilizing EMB agar (Difco,

Detroit, MI), and gram-positive bacterial counts were estimated by

phenylethanol agar spread plate counts, with both media being incubated at

37 °C in an aerobic incubator. Total anaerobic bacteria were estimated by

spread plate analysis on trypticase soy agar (BBL, Cockeysville, MD)

under anaerobic conditions.

Statistics Group mean cecal scores, indicative of severity of

lesions, were determined significantly different between treatment groups

at p<0.05 by Student's t-distribution.

97

RESULTS

Gross lesions All mice (both BALB/cByJ and C3H/HeSnJ)

receiving normal water showed severe cecitis at all sacrifice time points

(Figures 1 and 2). In contrast, challenged mice maintained on antibiotics in

water showed minimal (Fig. 1, 5 dPI) or no development of grossly

evident cecal lesions (Figures 1 and 2).

Figure 1. Development of macroscopic cecal lesions in BALB/cByJ mice maintained on normal water (control) or water supplemented with antibiotics (oral antibiotics; spectinomycin, spiramycin, rifampicin, colistin, and vancomycin). Data is presented as group mean cecal scores, with 0 being no gross lesions observed, progressing to 3, severe cecal lesions. Group sizes: 30 days post infection (dPI) control, n=5; all other groups, n=6. *,p<0.05 between treatment groups at the time point shown, t. mice which were administered antibiotics and challenged had antibiotic water replaced by normal water at day 15 PI and necropsied 15 days later.

4/ *

• oral antibiotics 0 control

5dPI 10dPI 15 dPI 30dPI

98

• oral antibiotics 0 control

30dPI

Figure 2. Development of macroscopic cecal lesions in C3H/HeSnJ mice maintained on normal water (control) or water supplemented with antibiotics (oral antibiotics; spectinomycin, spiramycin, rifampicin, colistin, and vancomycin). Data is presented as group mean cecal scores, with 0 being no gross lesions observed, progressing to 3, severe cecal lesions. Group sizes: all groups n=6. *, p<0.05 between treatment groups at the time point shown, f, mice which were administered antibiotics and challenged had antibiotic water replaced by normal water at day 15 PI and necropsied 15 days later.

The gross appearance of ceca from mice maintained on oral

antibiotics and challenged with S. hyodysenteriae was different than that of

a normal, non-infected mouse; the antibiotic treated mouse ceca were more

flaccid and filled with more fluid cecal contents than a normal non-infected

mouse. Control, non-infected, antibiotic-treated mice were necropsied at

99

10 dPI, and the ceca in these mice also appeared flaccid and cecal contents

were more fluid than normal. This finding suggested that the change in

cecal character of the antibiotic-treated mice was due to the oral antibiotics,

not due to the presence of S. hyodysenteriae.

Tables 1 and 2 illustrate the colony forming units (CPU) of

spirochetes present within the ceca of treatment groups. The data is

represented in two forms: the number of S. hyodysenteriae CPU per gram

of cecum (Table 1) and the total number of spirochete CPU recovered

from individual mouse ceca (Table 2). Mice which developed lesions

(received normal water) had higher values of CPU/g cecum; however,

these same mice had less total numbers of spirochetes in their ceca than did

the antibiotic-treated mice. Mice which developed severe lesions also had

atrophied ceca, approximately 25% or less the weight of an antibiotic-

treated mouse ceca; this atrophy would effectively increase CPU/g cecum

values for diseased animals relative to non-diseased animals (mice receiving

antibiotic water). At 15 dPI, enumerating S. hyodysenteriae CPU became

difficult, as oral antibiotics presumeably selected for antibiotic resistant

cecal flora which began to over-grow the selective media used for the pour

plates.

Histopathological findings Differences between mice infected

with S. hyodysenteriae but receiving oral antibiotics (no lesions) and mice

infected and receiving normal water (lesions) were noted. Mice with

lesions exhibited increased crypt depth and higher incidences of mixed

(inflammatory and lymphocytic/plasmacytic) infiltrates in the lamina

propria. Warthin-Starry silver stains showed spirochetes present in all

100

BALB/cBvJ C3H/HeSnJ

5dPI 9.1 ±2.4 40±4.0§ 4.6 ± 1.2 37±14§

lOdPI 3.6 + 0.7 19±7§ 0.7 ±0.4 6.0 ± 3.7§

15dPI N.D.t 2.2±1.1§ N.D. 49±15§

30dPI* 77±19§ 55±21§ 160±40§ 22±9§

oral control oral control

antibiotics antibiotics

Table 1. Levels of recovered 5. hyodysenteriae colony forming units (CPU) expressed as mean CFU/g cecum of mice maintained on oral antibiotics (spiramycin, colistin, rifampicin, spectinomycin, and vancomycin) or normal water (control). Ceca were individually homogenized, diluted, and soirochetes enumerated by a pour-plate assay as described in Materials and Methods. All values are represented as group mean (x 10" 6) ± standard error of the mean. *, mice previously on oral antibiotics and infected had normal water returned to them at 15 dPI and sacrificed at 30 dPI. tN.D., not determined. §, groups which developed macroscopic lesions.

101

BALB/cByJ C3H/HeSnJ

5dPI 2.4 ±0.5 0.9±0.1§ 1.2 ±0.5 0.4±0.1§

lOdPI 1.1 ±0.3 0.4±0.1§ 0.1 ±0.1 .01±.01§

15 dPI N.D.t .05±.01§ N.D. 0.7±0.2§

30dPI* 3.7±1.1§ 2.2±0.8§ 1.7±0.5§ 0.4±.13§

oral control

antibiotics

oral control

antibiotics

Table 2. Levels of recovered S. hyodysenteriae colony forming units (CPU) expressed as mean total S, hyodysenteriae CPU in individual ceca of mice maintained on oral antibiotics (spiramycin, spectinomycin, rifampicin, vancomycin, and spectinomycin) or normal water (control). Ceca were individually homogenized, diluted, and spirochetes enumerated by a pour-plate assay as described in Materials and Methods. All values are represented as group mean (x 10"^) ± standard error of the mean. *, mice previously on oral antibiotics and infected had normal water returned to them at 15 dPI and sacrificed at 30 dPI. tN.D., not determined. §, groups which developed macroscopic lesions.

102

infected cecal sections, with ceca showing gross lesions having relatively

more spirochetes within the crypts than the lumen, and mice not developing

gross lesions having relatively more spirochetes within the cecal lumen

than the crypts. Mice treated with oral antibiotics (and not developing

gross lesions) did have spirochetes in the deep crypts. A hypothesis why S.

hyodysenteriae was in relatively higher number in the crypts of diseased

ceca is that S. hyodysenteriae is very viscotactic, and will reach high

numbers in the base of crypts due to the overproduction of and the

spirochete's ability to be motile in mucus in the diseased cecum.

Bactériologie characterization of antibiotic treatment

Individual cecal pour plate results for gram negative, gram positive, and

total anaerobic bacterial CFU/g of cecum are pooled by treatment group

and illustrated in Figure 3. Six day treatment with the oral antibiotic

combination in the water decreased gram negative, gram positive, and total

anaerobic bacterial counts approximately 5 log units/g cecum when

compared to counts observed in control mice receiving normal water.

103

/

• Gram negative B3 Gram positive ^ Anaerobe

Control Antibiotics

Figure 3. Estimation of cecal microflora shifts after treatment with oral antibiotics (spiramycin, spectinomycin, vancomycin, rifampicin, and colistin). C3H/HeSnJ mice treated with oral antibiotics (see Materials and Methods) (n=6) were administered antibiotics in the water for 6 days prior to sacrifice. C3H/HeSnJ mice receiving normal water (Control) (n=4) were sacrificed at the same time point. Individual ceca were homogenized, diluted, and gram-negative, gram-positive, and anaerobic bacteria CFU were counted by pour plate analysis on EMB, phenylethanol, and trypticase soy agar plates, respectively, in the appropriate incubation atmosphere. Results are expressed as the group mean logic CFU/g cecum value. Standard errors are indicated. All antibiotic treated group values were significantly lower than control groups (p<0.05).

104

DISCUSSION

Germfree mice, as a model, are dissimilar in some immunological

aspects from conventional mice (3, 4, 15, 29). Sera from germfree mice

are usually low in gamma globulin concentration (when compared to

conventional mice), but immunoglobulins are present (29). When

germfree rats are conventionalized, rapid increases in alpha globulins

(presumed acute phase proteins) and beta globulins are observed. Gamma

globulins, however, remain at low levels for two weeks (3, 4). Germfree

mice develop significant antibody responses to bacterial lipopolysaccharide;

however, peak titers occur at a later time point than those attained by

conventional mice (15). Although macrophages from germfree mice

possess similar phagocytic indices in comparison to macrophages from

conventional mice (6), conventional macrophages are more active and

better able to kill Listeria sp., both in vivo and in vitro than macrophages

from germfree mice (3, 19). Germfree mice also have less developed

peripheral lymphoid organs, especially noticeable in Peyer's patches and

mesenteric lymph nodes (3). With immunologic capability differences

described between germfree and conventional mice, especially involving

the gut, this work addressed the validity of the S. hyodysenteriae-

monoassociated germfree animal. In other words, does the S.

hyodysenteriae-monosLSSOcisited animal fail to develop lesions simply

because of anatomic/iirununologic constraints inherent to the germfree

model?

105

The present work demonstrates that conventional mice, maintained

on drinking water containing antibiotics and demonstrating cecal

microflora shifts, will be colonized by S. hyodysenteriae but will not

develop lesions. Even though mice which were on oral antibiotics did not

develop lesions, levels of spirochetes present in the cecum appeared

comparable with levels determined in groups on normal water but which

developed lesions (Tables 1 and 2). Alternatively, it is possible that this

combination of antibiotics affected the expression of virulence attributes by

S. hyodysenteriae and thus reduced the potential for development of

lesions.

The absence of lesions in conventional mice treated with oral

antibiotics, along with previous research (13, 32), suggests that secondary

bacteria are dynamically involved in the progression of lesions. This need

for a secondary bacteria to aid in development of lesions is now suggested

not to be only applicable in the gnotobiotic system. Whether the secondary

bacteria (which, in this work, is inhibited by the oral antibiotics) contibutes

a vital pathologic factor(s) or enables S. hyodysenteriae to express

unknown virulence factors is not clear at this time. Further research into

the role of secondary bacteria in disease progression, by supplying or

replacing secondary bacteria fractions (i.e., LPS or inactive cellular

components), or supplying intermediate inflammatogenic substances (i.e.,

platelet-activating factor) in this antibiotic-treated system will further

address disease pathogenesis of swine dysentery.

106

REFERENCES

1. Albassam, M. A., H. J. Olander, H. L. Thacker, and J. J. Turek. 1985. Ultrastructural characterization of colonic lesions in pigs inoculated with Treponema hyodysenteriae. Can. J. Comp. Med. 49:384-390.

2. Argenzio, R. A. 1985. Pathophysiology of swine dysentery, p. 3-19. In; C. J. Pfeiffer (ed.). Animal models for intestinal disease. CRC Press Inc., Boca Raton, Fla.

3. Bealmear, P. M., 0. A. Holterman, and E. A. Mirand. 1984. Influence of the microflora on the immune response, p.335-384 In: Coates, M. E., and B. E. Gustafsson (eds.). Laboratory animal handbooks 9, the germ-free animal in biomedical research. London, Laboratory Animals LTD.

4. Bolish, E., C. E. Yale, and R. Hong. 1973. Influence of a defined flora on the serum proteins of gnotobiotic rats, p 485-492 jta: Germfree research: biologic effect of gnotobiotic environments. J. B. Heneghan (ed.). New York: Academic Press.

5. Brandenburg, A. C., O. P. Miniats, H. D. Geissinger, and E. Ewert. 1977. Swine dysentery: inoculation of gnotobiotic pigs with Treponema hyodysenteriae and Vibrio coli and a Peptostreptococcus. Can. J. Comp. Med. 41:294-301.

6. Doll, J. P. 1962. Rate of carbon clearance in three strains of germfree mice. Am. J. Physiol. 203:291-295.

7. Greer, J. M., and M. J. Wannemuehler. 1989. Comparison of the biological responses induced by lipopolysaccharide and endotoxin of Treponema hyodysenteriae and Treponema innocens. Infect. Immun. 57:717-723.

8. Greer, J. M., and M. J. Wannemuehler. 1989. Pathogenesis of Treponema hyodysenteriae: induction of interleukin-l and tumor necrosis factor by a treponemal butanol/water extract (endotoxin). Microb. Pathog. 7:279-288.

107

9. Harris, D. L., and R. D. Glock. 1986. Swine dysentery and spirochetal diseases, p. 494-507. In: A. D. Leman, B. Straw, R. D. Glock, et. al., (eds.), Diseases of swine, 6th ed., Iowa State University Press, Ames, Iowa.

10. Harris, D. L., R. D. Glock, C. R. Christensen, and J. M. Kinyon. 1972. Swine dysentery-I. Inoculation of pigs with Treponema hyodysenteriae (new species) and reproduction of the disease. Vet. Med. Small Animal Clin. 67:61-68.

11. Harris, D. L., T, J. L. Alexander, S. C. Whipp, I. M. Robinson, R. D. Glock, and P. J. Matthews. 1978. Swine dysentery: studies of gnotobiotic pigs inoculated with Treponema hyodysenteriae, Bacteroides vulgatus, and Fusobacterium necrophorum. J. Am. Vet. Med. Assoc. 172:468-471.

12. Joens, L. A., and R. D. Glock. 1979. Experimental infection in mice with Treponema hyodysenteriae. Infect. Lnmun. 25:757-760.

13. Joens, L. A., I. M. Robinson, R. D. Glock, and P. J. Matthews. 1981. Production of lesions in gnotobiotic mice by inoculation with Treponema hyodysenteriae. Infect. Lnmun. 31:504-506.

14. Kennedy M. J., D. K. Rosnick, R. G. Ulrich, and R. J. Yancey. 1988. Association of Treponema hyodysenteriae with porcine intestinal mucosa. J. Gen. Microbiol. 134:1565-1576.

15. Kiyono, H., J. R. Mcghee, and S. M. Michalek. 1980. Lipopolysaccharide regulation of the immune response: comparison of responses to LPS in germfree, Escherichia co//-monoassociated, and conventional mice. J. Immunol. 124:36-41.

16. Kunkle, R. A., and J. M. Kinyon. 1988. Improved selective medium for the isolation of Treponema hyodysenteriae. J. Clin. Microbiol. 26:2357-2360.

17. Lemcke, R. M., and M. R. Burrows. 1982. Studies on a hemolysin produced by Treponema hyodysenteriae. J. Med. Microbiol. 15:205-214.

108

18. Lysons, R. J., K. A. Kent, A. P. Bland, R. Sellwood, W. F. Robinson, and A. J. Frost. 1991. A cytotoxic haemolysin from Treponema hyodysenteriae-z. probable virulence determinant in swine dysentery. J. Med. Microbiol. 34:97-102.

19. MacDonald, T. T., S. G. Simkins, G. L. Spitalny, and P. B. Carter. 1978. Macrophage activation in the germfree mouse. 16th Ann. Mtg. Assoc. Gnotobiotics, Madison, WI., p. 10.

20. Meyer, R. C., J. Simon and C. S. Byerly. 1975. The etiology of swine dysentery, m. The role of selected gram-negative obligate anaerobes. Vet. Path. 12:46-54.

21. Meyer, R. C., J. Simon, and C. S. Byerly. 1974. The etiology of swine dysentery. I. Oral inoculation of germfree swine with Treponema hyodysenteriae and Vibrio coli. Vet. Path. 11:515-526.

22. Nibbelink, S. K., and M. J. Wannemuehler. 1991. Susceptibility of inbred mouse strains to infection with Serpula {Treponema) hyodysenteriae. Infect. Lnmun. 59:3111-3118.

23. Nibbelink, S. K., and M. J. Wannemuehler. 1992. An enhanced murine model for studies of Serpulina {Treponema) hyodysenteriae pathogenesis. Infect. Immun. 60: (in press).

24. Nibbelink, S. K., D. C. Morfitt, and M. J. Wannemuehler. 1990. Comparative effects of Treponema hyodysenteriae and Treponema mnoce/i5 infection in germfree mice. p.l3 Di: Proceedings, 71st Ann. Mtg. Conf. Res. Workers in Animal Dis., Chicago, IL.

25. Nuessen M. E., J. R. Birmingham, and L. A. Joens. 1982. Biological activity of a lipopolysaccharide extracted from Treponema hyodysenteriae. Infect. Lnmun. 37:138-142.

26. Nuessen, M. E., L. A. Joens, and R. D. Glock. 1983. Involvement of lipopolysaccharide in the pathogenicity of Treponema hyodysenteriae. J. Dnmunol. 131:997-999.

109

27. Saheb, S. A., L. Massicotte, and B. Picard. 1980. Purification and characterization of Treponema hyodysenteriae hemolysin. Biochimie 62:779-785.

28. Schaedler, R. W., and R. J. Dubos. 1961. The susceptibiUty of mice to bacterial endotoxins. J. Exp. Med. 113:559-570.

29. Sell, S., and J. L. Fahey. 1964. Relationship between gamma globulin metabolism and low serum gamma globulin in germfree mice. J. Immunol. 93:81-87.

30. Stanton, T. B. 1989. Glucose metabolism and NADH recycling by Treponema hyodysenteriae, the agent of swine dysentery. Appl. Environ. Microbiol. 55:2365-2371.

31. ter Huume, A. A. H. M., M. van Houten, S. Muir, J. G. Kusters, B. A. M. van der Zeijst, and W. Gaastra. 1992. Inactivation of a Serpula (Treponema) hyodysenteriae hemolysin gene by homologous recombination: importance of this hemolysin in pathogenesis of S. hyodysenteriae in mice. F. E. M. S. Microbiol. Letters 92:109-114.

32. Whipp, S. C., I. M. Robinson, D. L. Harris, R. D. Glock, P. J. Matthews, and T. J. L. Alexander. 1979. Pathogenic synergism between Treponema hyodysenteriae and other selected anaerobes in gnotobiotic pigs. Infect. Immun. 26:1042-1047.

33. Whipp, S. C., J. Pohlenz, D. L. Harris, I. M. Robinson, R. D. Glock, and R. Kunkle. 1982. Pathogenicity of Treponema hyodysenteriae in uncontaminated gnotobiotic pigs. p. 27 k: 7th Int. Pig Vet. Soc. Congress, Mexico City, Mexico.

110

PAPER V.

EXPERIMENTAL INFECTION OF GNOTOBIOTIC MICE WITH

Serpulina hyodysenteriae OR S. innocens, EST THE PRESENCE OR

ABSENCE OF Bacteroides vulgatus.

I l l

ABSTRACT

Germfree CD-I mice were orally infected with Serpulina

hyodysenteriae or S. innocens, with or without Bacteroides vulgatus

biassociation to determine pathogenic mechanisms of S. hyodysenteriae.

Mice monoassociated with S. hyodysenteriae B204 showed no development

of macroscopic cecitis, with minor microscopic changes noted in cecal

epithelia. Mice monoassociated with S. innocens B256 did not develop

macroscopic lesions. The major difference noted between the two groups

was that the pathogenic spirochete, S. hyodysenteriae B204, appeared to

invade the cecal lamina propria, whereas S. innocens B256 did not.

Biassociation of S. hyodysenteriae-mÎQcXtà. gnotobiotic mice with

Bacteroides vulgatus resulted in the development of macroscopically

evident lesions; biassociation of S. innocens infected mice with B. vulgatus

did not induce macroscopic lesions. Sera obtained from gnotobiotic groups

indicated an early antibody response against spirochetal whole cell antigen,

this antibody response was cross reactive between both species of the

spirochete. These studies confirm the finding of S. innocens B256 being

nonpathogenic, and they also describe an invasive capability of S.

hyodysenteriae B204 which S. innocens B256 does not possess.

112

INTRODUCTION

Swine dysentery, a mucohemorrhagic diarrheal disease of growing

pigs, was shown to be caused by an aerotolerant (20) gram negative

spirochete, S. hyodysenteriae (5). When conventional pigs were challenged

with pure cultures of S. hyodysenteriae, clinical swine dysentery was

observed in naive pigs, with S. hyodysenteriae being shed in the feces (5).

Virulence determinants of the spirochete are postulated to be a beta

hemolysin (10, 11, 18) and a lipopolysaccharide (LPS) or LPS-like moiety

(3, 4, 16, 17). Although virulence attributes are proposed, the initial

pathogenesis of the disease is not known and the character of a protective

response in pigs has not been well described.

Early research into disease pathogenesis, by infecting germfree pigs

with pure cultures of S. hyodysenteriae, proved difficult as

monoassociation with the spirochete did not lead to disease or lesion

development (1, 13). Various attempts to cause disease in gnotobiotic pigs

were unsuccessful unless other bacteria were concomittantly administered

with S. hyodysenteriae in gnotobiotic pigs (6, 14, 23). Certain bacteria,

such as Lactobacillus, Escherichia coli, or Vibrio coli could not, in concert

with S. hyodysenteriae in gnotobiotic pigs, produce lesions (13). Other

bacteria, however, such as Bacteroides vulgatus, Fusobacterium

necrophorum, and Clostridium sp., could, in association with S.

hyodysenteriae, induce the development of lesions in gnotobiotic pigs (6,

23). A report (24) has described development of lesions in S.

hyodysenteriae-m.onodi.ssoQ.\d.tQà gnotobiotic pigs; however, the factor(s)

113

necessary to allow S. hyodysenteriae to cause disease in this report have not

been identified (24).

These data suggested a multi-factorial disease process, with S.

hyodysenteriae being the causative agent. Possible hypotheses explaining

the need for a secondary bacteria for S. hyodysenteriae to cause disease in

gnotobiotic animals include: 1)5. hyodysenteriae is a "conditional"

pathogen, requiring other bacteria to fully express virulence (14), or 2)

"synergistic" secondary bacteria allow or cause S. hyodysenteriae to

colonize more readily and to a higher density (23).

A mouse model exists for the study of swine dysentery (7) which is

characterized by the development of a catarrhal cecitis which, in some

cases, also involves areas of the proximal colon. Mice infected with S.

hyodysenteriae and which develop catarrhal cecitis do not appear to clear

the organism or resolve the lesion even to 70 days post infection (dPI)

(15). Although clinical signs in infected mice are less severe than those of

pigs (15), histopathological changes are similar (7).

The mouse has been used to identify enteropathogenic porcine

spirochetes (9), thus classifying a spirochete as S. hyodysenteriae or S.

innocens, a nonpathogenic porcine spirochete which may be cultured from

clinically normal pigs. The gnotobiotic mouse has also been used in S.

hyodysenteriae research (8). Joens et al. (8) reported similar findings in

gnotobiotic CF-1 mice infected with S. hyodysenteriae as that previously

reported for gnotobiotic pigs (6, 23); monoassociation with S.

hyodysenteriae in gnotobiotic mice caused minor histopathologic changes

but no grossly evident lesions, whereas biassociation with Bacteroides

114

vulgatus did cause grossly evident lesions. As with gnotobiotic pigs,

monoassociation with B. vulgatus as a control did not cause histopathologic

or gross lesions (8).

This work attempts to elucidate pathogenic mechanisms of S.

hyodysenteriae by infecting gnotobiotic mice with either S. hyodysenteriae

B204 (serotype 2) or with S. innocens B256 for extended periods and

compare histopathologic findings to determine differences between the two

groups. Previous work with germfree mice challenged with S.

hyodysenteriae (8) was terminated at 9 days post-challenge; longer

incubation periods were used in this study to ascertain possible later lesion

development or histopathological changes. In addition, sera was tested to

determine which antigens of the spirochetes were recognized by the animal.

115

MATERIALS AND METHODS

Mice Germfree CD-I mice were utilized in these trials. Mice were

maintained in flexible polyethylene gnotobiotic isolators (Standard Safety

Equipment, Palatine, IL) using standard germfree protocols. Germfree

status was ascertained by periodically streaking soiled bedding or cecal

content swabs on both aerobic and anaerobic blood agar plates. Mice

received sterilized Mouse Lab Chow #5010 (Purina Mills, St. Louis, MO)

and sterilized water ad libitum.

Bacteria S. hyodysenteriae strain B204 (serotype 2) was used as the

pathogenic spirochete in this study. S. innocens B256 was used as the non­

pathogenic spirochete. Both species were propogated anaerobically at 37°C

in trypticase soy broth (BBL Microbiology Systems, Cockeysville, Md)

supplemented with 5% equine serum (Hyclone Laboratories, Logan, Utah),

0.5% yeast extract (BBL), and 1% VPI salt solutions as described (22).

Bacteria were enumerated by Petroff-Houser chamber counts. A

Bacteroides vulgatus isolate, obtained from laboratory culture collections,

was grown anaerobically in the above media omitting equine serum.

Infection Protocol Mice were intragastrically intubated once with

1 X 108 log phase broth cultures of S. hyodysenteriae or S. innocens. In

monoassociated gnotobiotic groups, this time point will be referred to as 0

days post-infection (DPI). In mice which were biassociated with either of

the spirochetes plus B. vulgatus, mice were intragastrically intubated at -10

DPI with either of the spirochetes, and secondarily challenged with 1 x 10^

B. vulgatus at day 0 PL

116

Necropsy Procedure Mice were sacrificed at various time points

PI and ceca visualized to assess development of grossly evident cecal

lesions. Specifically, in monoassociated studies, S. hyodysenteriae infected

mice were sacrificed at DPI 5, 10, 15, 30, 45, 70, and 115. Mice

monoassociated with S. innocens were sacrificed at DPI 5, 10, 15, 30, 45,

70, and 90. Mice biassociated with either of the spirochetes plus B.

vulgatus were sacrificed at DPI -1, 5, 10, 15, and 30. Blood was collected

via retro-orbital puncture to assay sera for anti-spirochete antibodies. Ceca

were fixed in Kamovsky's fixative for histologic examination.

Serology Sera was pooled from treatment groups sacrificed at

specific DPI. Specific antibody responses of the sera samples against S.

hyodysenteriae whole cell lysate (WCL), endotoxin (butanol/water extract),

and lipopolysaccharide (LPS) (phenol/water extraction), and to S. innocens

WCL, endotoxin, and LPS were performed by ELISA. In biassociated

groups, sera samples also were evaluated as to their anti-5. vulgatus WCL

and endotoxin antibody content.

ELISA Procedure Bacterial antigens (50 ng/well) were coated on

polyvinyl 96 well microtiter plates in 0.1 M NaHC03 buffer (pH 9.6).

Plates were washed using Tween-saline (0.85% NaCl, 0.025% Tween 20)

and blocked with 2% skim milk. Diluted sera (1:50 in Tween-saline) was

incubated 4 h at 37° C. Alkaline phosphatase-labeled conjugate was added

afteer washing and allowed to incubate approximately 1 h at 37 C. Plates

were washed, substrate was added, and plates were incubated

approximately 20 minutes, and ELISA optical densities were determined at

405 nm.

117

RESULTS

Monoassociated studies Mice monoassociated with either S.

hyodysenteriae or S. innocens had not developed macroscopic lesions by

any of the necropsy time points. Mice challenged with S. hyodysenteriae

were culture positive at all time points. Mice challenged with S. innocens

were culture negative at 5 dPI and were re-challenged with 10^ S innocens

B256 on day 10 PI. Subsequently, mice were culture positive for S.

innocens, but not heavily colonized by S. innocens until 30 dPI, as

subjectively evaluated by darkfield microscopy of cecal contents.

Histopathologically, minor changes could be observed in the cecal

epithelium of S. hyodysenteriae infected mice, such as crypt epithelial cell

hyperplasia and increased crypt depth. These subtle changes were also

observed in the S. innocens infected mouse tissues, but to a lesser degree.

Serpulina hyodysenteriae was observed between and within cecal epithelial

cells, and in the lamina propria, beginning at approximately 10-15 dPI.

This invasion through the mucosal epithelium was not observed at any time

point in the S. innocens infected mice, hi spite of spirochetes within the

cecal lamina propria in S. hyodysenteriae infected mice, minimal or no

inflammatory reactions were noted.

Specific serum antibodies against S. hyodysenteriae antigens are

shown in Figures 1 and 2. Figure 1 shows antibody responses against B204

(S. hyodysenteriae) whole cell lysate; antigen recognition is evident,

however, cross reactivity to this antigen by heterologous challenge groups

is shown. When S. hyodysenteriae B204 endotoxin is used as an ELISA

118

antigen, a more specific serum response is noted (Fig. 2). Similarly, serum

responses against S. innocens B256 WCL show cross reactivity (Fig. 3);

anti- S. innocens endotoxin (Fig. 4) show specific serum responses. Both

S. hyodysenteriae and S. innocens LPS (phenol/water extraction) were also

used as ELISA antigens; however, sera ELISA values were negligible (data

not shown).

Figure 1. Responses of pooled sera by S. hyodysenteriae-monoa.ssocia.tQd

(B204 Infected) and S. innocens-monoassociated (B256 Infected) and control (Non-infected) mice to S. hyodysenteriae B204 WCL antigen as detected by ELISA (see Materials and Methods). Mice infected with S. hyodysenteriae B204 were necropsied at 5, 10, 15, 30, 45, 70, and 115 dPI. Mice infected with S. innocens were necropsied at 5, 10, 15, 30, 45, and 90 dPL Control mice were sacrificed at days 5, 30, 45, 70, and 90. Group sizes: B204 infected; 5-15 dPI, each group n=3, for 30 and 45 dPI, each group n=4; for 70 dPI, n=5, and for 115 dPI, n=6. For B256 infected; 5-45 dPI, each group n=3, for 70 and 90 dPI, each group n=2.

3 y

B204 Infected 8256 Infected Non-infected

5 1 0 1 5 3 0 4 5 7 0 9 0 1 1 5

Days Post-infection

119

m o

(0 c 0) Û (0 o

a. O

2 -

1

10 15 30 45 70

Days Post-infection

9 0 1 1 5

• B204 Infected 0 B256 Infected • Non-infected

Figure 2. Responses of pooled sera by S. hy0dysenteriae-m0n0SLSS0cia.tQd (B204 Infected) and S. mnocew^-monoassociated (B256 Infected) and control (Non-infected) mice to S. hyodysenteriae B204 endotoxin as detected by ELIS A (see Materials and Methods). Sacrifice time points and group numbers are listed in Fig. 1.

120

in o

OT C 0) Q

« o Q. O

1 0 15 30 45 70

Days Post-infection

1 1 5

• B204 Infected 0 B256 Infected • Non-infected

Figure 3. Responses of pooled sera by S. hyodysenteriae-monosissocidLted (B204 Infected) and S. mnocen^-monoassociated (B256 Infected) and control (Non-infected) mice to S. innocens B256 WCL as detected by ELISA (see Materials and Methods). Sacrifice time points and group numbers are listed in Fig. 1.

121

B204 Infected B256 Infected Non-infected

5 1 0 1 5 3 0 4 5 7 0 9 0 1 1 5

Days Post-infection

Figure 4. Responses of pooled sera by S. hyodysenteriae-monoâssociatQà (B204 Infected) and S. /nnocen^-monoassociated (B256 Infected) and control (Non-infected) mice to S. innocens B256 endotoxin as detected by ELISA (see Materials and Methods). Sacrifice time points and group numbers are listed in Fig. 1.

Biassociated studies Serpulina hyodysenteriae-iiâQcttà. (primary

challenge) gnotobiotic mice biassociated with Bacteroides vulgatus showed

gross lesions consisting of thickened mucosa, serosal edema, and increased

mucus within the cecal lumen beginning at 5 DPI. In addition, ceca were

somewhat smaller in size than ceca of other groups not showing gross

lesions. The character of the gross lesions remained similar at 5, 10, 15,

and 30 DPI. S. innocens-ivdtoXtd gnotobiotic mice biassociated with 5.

vulgatus did not exhibit gross cecal lesions at 5,10,15, or 30 dPI. Mice

in o

c 0) O (0 o a O

1 -

0 ^

122

which were monoassociated with B. vulgatus did not show gross lesions at

5, 10, 15, or 30 DPI.

Histologically, S. hyodysenteriaelB. vulgatus-kdect&d gnotobiotic

mice showed increased crypt depth and crypt epithelial cell hyperplasia

when compared to S. innocensIB vulgatus, B. VM/gamj'-monoassociated, or

non-infected control germfree ceca. As observed in the Serpulina spp.

monoassociated trial, S. hyodysenteriaelB. vulgatus biassociated gnotobiotic

mouse tissues had invading spirochetes present within lamina propria at

days 10,15, and 30 PL Although S. innocens was observed in cecal crypts,

no S. innocens were observed in lamina propria, even in this biassociated

system.

Specific serum antibody responses of biassociated mouse groups

were similar to those observed for monoassociated mice infected with

either S. hyodysenteriae or S. innocens; cross reactivity was observed for

WCL antigens of both spirochete species, with endotoxin responses being

more specific (data not shown). Anti-5. vulgatus responses of all

biassociated groups are represented in Figures 5 and 6, showing

recognition of antigens of the secondary bacteria in all treatment groups.

Figure 5. Levels of serum antibody produced by S.

hyodysenteriae/Bacteroides vulgatus biassociated (B204+B.V.)

and S. innocens/Bacteroides vulgatus biassociated (B256+B.V.)

and Bacteroides vulgatus monoassociated mice to B. vulgatus

WCL antigen as detected by ELIS A (see Materials and

Methods). Group sizes: For B204 + B.v., n=3, 3, 3, 8 for dPI

5, 10, 15, and 30. For B256 + B. v., n=3, 3, 3, and 4; For B.

vulgatus monoassociated groups, n= 2, 3, 2, and 3.

ELISA Optical Density

% y ç» ça Ç) Ç) Ç) ça *A ers Va "tù \j\ "<3^

%

\ '

o o o o o o o ^ hû CO 4^ en ày

Figure 6. Levels of serum antibody produced by S.

hyodysenteriaefBacteroides vulgatus biassociated (B204+B,v.)

and S. innocenslBacteroides vulgatus biassociated (B256+B.V.)

and Bacteroides vulgatus monoassociated mice to B. vulgatus

endotoxin as detected by ELISA (see Materials and Methods).

Group sizes: For B204 + B.v., n=3, 3, 3, 8 for dPI 5, 10, 15,

and 30. For B256 + B. v., n=3, 3, 3, and 4; For B. vulgatus

monoassociated groups, n= 2, 3, 2, and 3.

ELISA Optical Density

127

DISCUSSION

This work attempted to elucidate pathogenic mechanisms of S.

hyodysenteriae by comparing gnotobiotic mouse lesion formation between

mice infected with the pathogenic S. hyodysenteriae and the nonpathogenic

S. innocens. In monoassociation trials, it was found that S. hyodysenteriae

infection did not cause gross cecal lesions, even at protracted incubation

times (115 DPI). The observation was made, however, that only S.

hyodysenteriae-mS.QCie,à. mice showed spirochetes invading the cecal

epithelium and into the lamina propria in contrast to S. innocens-inî&cXQd.

mice. In spite of large numbers of spirochetes in the lamina propria,

minimal cellular reaction (inflammation) or damage was noted. It should

be emphasized that the actual need for spirochetal invasion to cause disease

has not been described to this point, and has been discounted as essential in

conventional pigs (2).

A virulence determinant of S. hyodysenteriae has been postulated to

be an endotoxin moiety which has in vitro biological activity (3, 4, 16).

Additionally, influence on susceptibility to S. hyodysenteriae-m&MCQà

cecitis has been shown to be influenced by murine Igs phenotype (15,17).

Germfree mice, as a model, are inherently refractory to LPS-induced

lethality (19), and germfree mouse peritoneal macrophages are refractory

to LPS/endotoxia stimulation and secretion of tumor necrosis factor, an

inflammatory cytokine (M. J. Wannemuehler, unpublished data). These

observations might partially explain the lack of lesion development in

monoassociated gnotobiotic mice infected with S. hyodysenteriae. Another

128

postulated virulence determinant is a beta-hemolysin/cytotoxin (21), which

has been shown to be cytotoxic for gnotobiotic pig colonic loop epithelial

cells (11). Interestingly, in S. hyodysenteriae monoassociated gnotobiotic

mice, minimal histopathological changes were noted, even upon invasion of

the cecal lamina propria. Since previous work (11) has shown S.

hyodysenteriae hemolysin to have direct cytotoxic activity against germfree

colonic epithelial cells in vivo, the absence of cytotoxic effects in

monoassociated mouse ceca suggests either: 1) monoassociation with

germfree animals by S. hyodysenteriae will not cause the spirochete to

produce hemolysin in vivo; 2) S. hyodysenteriae, monoassociated, cannot

produce enough hemolysin in vivo to cause tissue damage; or 3) Secondary

bacteria provide some sort of stimulus for S. hyodysenteriae to produce

hemolysin in vivo.

Pathogenic synergism was reported between S. hyodysenteriae and

other anaerobic or facultative anaerobic bacteria in gnotobiotic pigs (23).

The authors noted that biassociation of gnotobiotic pigs with S.

hyodysenteriae plus F.necrophorum, three strains of B.vulgatus, a

Clostridium species, or Listeria denitrificans led to clinical dysentery.

These observations allowed for the hypothesis that the presence of other

bacterial organisms was needed for S. hyodysenteriae to express

pathogenicity, and that this contribution by a secondary bacteria is

relatively non-specific (23). It was also theorized that secondary bacteria

could facilitate colonization of gnotobiotic mucosa by the spirochete (23).

Comparing lesion development of atrophic rhinitis in conventional and

gnotobiotic pigs, it was similarly hypothesized that slight damage to the

129

nasal mucosa caused by commensal bacteria causes leakage of exudate

which increases colonization by Pasteurella multocida (12). In the present

studies, using monoassociated gnotobiotic mice, however, high numbers of

spirochetes were observed on wet mount preparations of cecal contents, in

cecal crypts, and in lamina propria of the tissues, without biassociation with

other bacteria. These findings tend to discount the hypothesis of secondary

bacteria needed to allow S. hyodysenteriae to colonize to high density.

Lysons et al. injected germfree pig ileal and colonic gut loops with

high levels of sterile S. hyodysenteriae P18A hemolysin and observed

severe epithelial damage to villus tip epithelium in the ileal loops, and to

the intercryptal epithelia in the colonic loops (11). These findings, if

applicable to the mouse model, would suggest that, since S. hyodysenteriae

monoassociated gnotobiotic animals do not develop lesions, the hemolysin

presumeably produced by S. hyodysenteriae in monoassociated gnotobiotic

mice is either not active or is not being produced. No studies have been

implemented which directly show that S. hyodysenteriae produces beta

hemolysin in vivo. Additional possible hypotheses for the inability of S.

hyodysenteriae to cause gross cecitis in monoassociated gnotobiotic mice

would be a lack of "reactive" endotoxin--^, hyodysenteriae LPS/endotoxin

has been described to be less potent in inducing biological responses than E.

coli or S. typhimurium LPS (3)~or, possible unknown virulence factors

which are induced in vivo (possibly through the influence of the secondary

bacteria) which have not been detected as of yet.

The biassociated gnotobiotic mouse studies are consistent with

previous reports (8) in that the S. hyodysenteriaelB. vulgatus gnotobiotic

130

system yields development of gross cecal lesions. S. innocens, which does

not cause disease, does not appear to colonize to as high of density as does

S. hyodysenteriae in disease conditions (personal observations). Our

research demonstrates that in the S. innocens/B. vulgatus biassociated

gnotobiotic mouse, "artificially" colonized heavily by S. innocens, no gross

cecitis is observed.

Serological responses of gnotobiotic mice infected with either

spirochete indicate relative cross reactivity to whole cell antigens of both

bacteria. More species-specific immune responses were noted at latter time

points. No influence of histopathological changes were noted due to serum

immunoblobulin responses. These findings support those previously

reported in gnotobiotic pigs and mice, and further support the observations

that secondary bacteria perform a vital and dynamic function in the

pathogenesis induced by S. hyodysenteriae.

131

REFERENCES

1. Brandenburg, A. C., O. P. Miniats, H. D. Geissinger, and E. Ewert. 1977. S wine dysentery; inoculation of gnotobiotic pigs with Treponema hyodysenteriae and Vibrio coli and a Peptostreptococcus. Can. J. Comp. Med. 41:294-301.

2. Glock, R. D., and D. L. Harris. 1972. Swine dysentery II. Characterization of lesions in pigs inoculated with Treponema hyodysenteriae in pure and mixed culture. Vet. Med. Small Animal Clin. 67:65-68.

3. Greer, J. M., and M. J. Wannemuehler. 1989. Comparison of the biological responses induced by lipopolysaccharide and endotoxin of Treponema hyodysenteriae and Treponema innocens. Infect. Immun. 57:717-723.

4. Greer, J. M., and M. J. Wannemuehler. 1989. Pathogenesis of Treponema hyodysenteriae: induction of interleukin-1 and tumor necrosis factor by a treponemal butanol/water extract (endotoxin). Microb. Pathog. 7:279-288.

5. Harris, D. L., R. D. Glock, C. R. Christensen, and J. M. Kinyon. 1972. Swine dysentery. I. Inoculation of pigs with Treponema hyodysenteriae (new species) and reproduction of the disease. Vet. Med. Small Animal Clin. 67:61-68.

6. Harris, D. L., T. J. L. Alexander, S. C. Whipp, I, M. Robinson, R. D. Glock, and P. J. Matthews. 1978. Swine dysentery: studies of gnotobiotic pigs inoculated with Treponema hyodysenteriae, Bacteroides vulgatus, and Fusobacterium necrophorum. J. Am. Vet. Med. Assoc. 172:468-471.

7. Joens, L. A., and R. D. Glock. 1979, Experimental infection in mice with Treponema hyodysenteriae. Infect. Immun. 25:757-760.

8. Joens, L. A., I. M. Robinson, R. D. Glock, and P. J. Matthews. 1981. Production of lesions in gnotobiotic mice by inoculation with Treponema hyodysenteriae. Infect. Immun. 31:504-506.

132

9. Joens, L. A., R. D. Glock, and J. M. Kinyon. 1980. Differentiation of Treponema hyodysenteriae from Treponema innocens by enteropathogenicity testing in the CF-1 mouse. Vet. Rec. 107:527-529.

10. Lemcke, R. M., and M. R. Burrows. 1982. Studies on a haemolysin produced by Treponema hyodysenteriae. J. Med. Microbiol. 15:204-214.

11. Lysons, R. J., K. A. Kent, A. P. Bland, R. Sellwood, W. F. Robinson, and A. J. Frost. 1991. A cytotoxic haemolysin from Treponema hyodysenteriae-di probable virulence determinant in swine dysentery. J. Med. Microbiol. 34:97-102.

12. Martineau, G. P., B. Martineau-Doize, and A. Broes. 1985. Atrophic rhinitis caused by Pasteurella multocida: some factors influencing pathogenicity in gnotobiotic and conventional piglets. Zbl. Vet. Med. B 32:583-592.

13. Meyer, R. C., J. Simon, and C. S. Byerly. 1974. The etiology of swine dysentery. H. Effect of a known microbial flora, weaning and diet on disease production in gnotobiotic and conventional swine. Vet. Pathol. 11:527-534.

14. Meyer, R. C., J. Simon, and C. S. Byerly. 1975. The etiology of swine dysentery. HI. The role of selected gram negative obligate anaerobes. Vet. Pathol. 12:46-54.

15. Nibbelink, S. K. and M. J. Wannemuehler. 1991. Susceptibility of inbred mouse strains to infection with Serpula (Treponema) hyodysenteriae. Infect. Immun. 59:3111-3118.

16. Nuessen, M. E., J. R. Birmingham, and L. A. Joens. 1982. Biological activity of a lipopolysaccharide extracted from Treponema hyodysenteriae. Infect. Immun. 37:138-142.

17. Nuessen, M. E., L. A. Joens, and R. D. Glock. 1983. Involvement of lipopolysaccharide in the pathogenicity of Treponema hyodysenteriae. J. Immunol. 131:997-999.

133

18. Saheb, S. A., L. Massicotte, and B. Picard. 1980. Purification and characterization of Treponema hyodysenteriae hemolysin. Biochimie 62:779-785.

19. Schaedler, R. W., and R. J. Dubos. 1961. The susceptibility of mice to bacterial endotoxins. J. Exp. Med. 113:559-570.

20. Stanton, T. B. 1989. Glucose metabolism and NADH recycling by Treponema hyodysenteriae, the agent of swine dysentery. Appl. Environm. Microbiol. 55:2365-2371.

21. ter Huume, A. A. H. M., M. van Houten, S. Muir, J. G. Kusters, B. A. M. van der Zeijst, and W. Gaastra. 1992. Inactivation of a Serpula {Treponema) hyodysenteriae hemolysin gene by homologous recombination: importance of this hemolysin in pathogenesis of S. hyodysenteriae in mice. F. E. M. S. Microbiol. Letters 92:109-114.

22. Wannemuehler, M. J., R. D. Hubbard, and J, M. Greer. 1988. Characterization of the major outer membrane antigens of Treponema hyodysenteriae. Infect. Immun. 56:3032-3039.

23. Whipp, S. C., I. M. Robinson, D. L. Harris, R. D. Glock, P. J. Mattihews, and T. J. L. Alexander. 1979. Pathogenic synergism between Treponema hyodysenteriae and other selected anaerobes in gnotobiotic pigs. Infect. Immun. 26:1042-1047.

24. Whipp, S. C., J. Pohlenz, D. L. Harris, I. M. Robinson, R. D. Glock, and R. Kunkle. 1982. Pathogenicity of Treponema hyodysenteriae in uncontaminated gnotobiotic pigs. p. 27 In: 7th Int. Pig Vet. Soc. Congress, Mexico City, Mexico.

134

PAPER VI.

EXPERIMENTAL INFECTION OF SEVERE COMBINED

IMMUNODEFICIENT (SCID) MICE WITH Serpulina hyodysenteriae

135

ABSTRACT

Serpulina hyodysenteriae, the causative agent of swine dysentery,

also causes a catarrhal cecitis in mice. The importance of a specific

immune response in the disease is not well characterized; specific anti-

spirochetal serum antibody response in the pig have been postulated to

protect, or, conversely, to exacerbate lesion development. We have orally

challenged severe combined immunodeficiency (SCID) mice (derived from

C3H/HeSnJ lineage) with S. hyodysenteriae strain B204 to characterize

development of gross and microscopic cecal lesions in mice incapable of

developing an immune response. In addition, the ability of macrophages

recovered from a population of resident peritoneal exudate cells (PEC) to

produce interleukin-1 and/or tumor necrosis factor when stimulated with

Escherichia coli LPS or S. hyodysenteriae endotoxin in vitro was

examined. Following infection, SCID mice developed gross cecal lesions

similar to those seen in C3H/HeN mice. Microscopically, SCID mice

infected with S. hyodysenteriae showed increased neutrophil infiltration of

the cecal lamina propria in comparison to infected C3H/HeN mice. Ability

of peritoneal macrophages from SCID mice to produce interleukin-1 and

tumor necrosis factor in response to E. coli or S. hyodysenteriae endotoxin

was similar to the amounts produced by macrophages from normal C3H

mice.

136

INTRODUCTION

Serpulina hyodysenteriae, a gram negative porcine spirochete, is the

causal agent of swine dysentery, a muco-hemorrhagic diarrheal disease of

swine (9). Virulence factors of the bacteria have been postulated to be a

beta-hemolysin (13, 21) and a lipopolysaccharide (LPS) or LPS-like moiety

(19, 20). A murine model of the disease has been described (10), and has

been used to elucidate pathogenic mechanisms of the bacteria. Differences

in susceptibility to lesion development after challenge with 5.

hyodysenteriae have been shown to be influenced by Ips genotype (17, 19),

inbred strain of mouse (17), suppliers of mice (23), and diet (18). Effect

of oral antibiotics on susceptibility has also been investigated (15).

Gnotobiotic mice, when mono-associated with S. hyodysenteriae, do

not develop gross lesions or develop disease (11). Gnotobiotic pigs, when

monoassociated with S. hyodysenteriae, do not develop lesions or clinical

dysentery (5, 14); however, in one system it was reported that S.

hyodysenteriae was pathogenic in uncontaminated gnotobiotic pigs (25).

When gnotobiotic mice or swine are bi-associated with S. hyodysenteriae

plus Bacteroides vulgatus, gross lesions and clinical dysentery (in

gnotobiotic pigs) are observed (8,11). Several bacterial species can fulfill

the role of the "secondary bacteria"(26); however, the contribution of the

secondary bacteria towards disease production is unknown.

Suenaga and Yamazaki (22) have shown lesion development in nude

mice, and Nibbelink and Wannemuehler (16) have shown lesion

development in mast cell-deficient W/W^ mice after S. hyodysenteriae

137

challenge. To date, the authors are unaware of any mutant mouse strain,

deficient in some cell or response, which will fail to develop lesions if

colonized heavily by S. hyodysenteriae (approximately 10^ spirochetes per

gram of cecal tissue). SCID mice possess a genetic defect, wherein the

lymphocytic cell precursors cannot join V-D-J coding regions, thus

severely decreasing functional T and B lymphocyte populations (2).

Natural killer cell (NK) activity in SCID mice is present (24), but

circulating antibody is not present.

Levels of anti-5. hyodysenteriae antibody have been implicated in

protection against the disease, however, our laboratory has considered the

possibility of an immune response being detrimental to the host, i.e.,

exacerbating or instigating mucosal lesions after S. hyodysenteriae

challenge. To this end, we experimentally challenged SCID mice with S.

hyodysenteriae to eliminate the variable of antibody formation or cell

mediated immunity on cecal lesion development.

138

MATERIALS AND METHODS

Mice C3HSmn.C-scid mice, derived from C3H/HeSnJ lineage

rips^/lps^^). were obtained from Jackson Labs, Bar Harbor, ME. C3H/HeN

miceHps^/lps^^). originally obtained from Harlan-Sprague Dawley,

Madison, WI, were maintained in breeding colonies at the Laboratory

Animal Resources Facility, College of Veterinary Medicine, Iowa State

University, Ames, Iowa. SCID mice were housed in a Thoren Laminar

Flow Unit and maintained under sterile conditions with sterile water

supplemented with Septra (Phoenix Pharmaceuticals, St. Joseph, MO, final

concentration in drinking water, 200 mg sulfamethoxazole and 40 mg

trimethoprim/L). All mice were fed autoclaved Mouse Lab Chow #5010

(Purina Mills, St. Louis, MO). Septra was withdrawn from the water 6

days prior to challenge with S. hyodysenteriae. Mice were 8-12 weeks of

age at the time of experimentation. All mice were housed and cared for

under the guidelines of the Committee on Animal Care, Iowa State

University.

Bacteria Mice were challenged with various doses of log phase

broth cultures of S. hyodysenteriae strain B204 (18-20 passages in vitro) as

previously described (17). Briefly, mice were intragastrically intubated

once a day for two days concurrent with a 30 hour fast. Varying doses

(10^-10^ spirochetes) of challenge inocula were delivered in a 1 mi

volume.

Necropsy Procedure Mice were necropsied 15 days post infection

(dPI) and ceca scored for gross lesions as previously described (17): 0; no

139

grossly evident lesions present, 1; excess intraluminal mucus evident, 2;

excess intraluminal mucus and cecal apex atrophy, and 3; excess

intraluminal mucus and complete organ atrophy. Ceca were fixed in

neutral buffered formalin for histopathological examination. Cecal

contents were swabbed on blood agar plates containing selective antibiotics

(12) for S. hyodysenteriae as previously described (17) to reisolate the

spirochete.

Collection of peritoneal macrophages Resident peritoneal

macrophages of all infected and control mice were harvested by

intraperitoneal lavage with 6 ml pyrogen-free phosphate buffered saline

containing 10 units/ml heparin and 1% fetal bovine serum (FBS, JR

Scientific, Irvine, CA). Cells were washed twice and suspended in MEM

media containing 3.7% NaHCOs, 25 units/ml penicillin, 25 |ig/ml

streptomycin (P/S), 2mM L-glutamine (L-glu), and 1% FBS. Cells were

plated at the density of 0.25 x 10^ cells/well (0.5 ml) in a 24-well plate.

Macrophages were allowed to adhere for 12 hours, washed three times

with PF PBS, one time with media (without FBS) and stimulants were

added to the macrophage cultures. Stimulants used were E. coli K235 TCA

extract (endotoxin) (Sigma Chemical Company, St. Louis, MO) and S.

hyodysenteriae B204 butanol/water extract (endotoxin) in various dilutions.

After 24 hours, supematants of stimulated peritoneal macrophages were

harvested, centrifiiged, and stored at -70 C until testing for IL-1 and TNF

activity.

TNF assay Murine L929 cells were suspended to a density of 5 x

105 cells/ml in MEM supplemented with 3.7% NaHCOs, 25 units/ml

140

penicillin, 25 |Xg/ml streptomycin (P/S), 2mM L-glutamine (L-glu), and

10% horse serum (Hyclone, Logan, Utah). One tenth ml of the cell

suspension was added to each well of a 96-well microtiter plate and

incubated overnight in a 5% C02 atmosphere. The supernatant fluids from

stimulated macrophages were diluted in MEM containing actinomycin D (3

|Xg/ml, final concentration), and 0.1 ml of diluted macrophage supernatant

was added to each microtiter well. Dilutions were performed in triplicate.

The assay was terminated at 18 hours after addition of supematants. Dead

(non-adherent) L929 cells were removed by multiple plate washes in PBS.

Residual monolayers of L929 cells were stained with 0.2% crystal violet-

2% fomialin solution for 15 minutes. Plates were extensively washed with

tap water to remove excess stain. Crystal violet was solubilized in 50%

ethanol (0.1 ml/microtiter well) and quantitated by measuring the

absorbance at 595 nm on an ELISA reader (model EL301, Biotek

Instruments, Winooski, VT). Supematants were tested in triplicate and the

results expressed as the percent cytotoxicity.

Interleukin-1 assay D10.G4.1 cells were harvested and

centrifuged over Histopaque 1077 (Sigma, St. Louis, MO) to remove

nonviable cells. The resulting cells were washed three times in RPMI1640

containing P/S, L-glu, and 10 mM HEPES, and suspended to 5 x 10^

cells/ml in RPMI containing P/S, L-glu,10 mM HEPES, and 5% FBS.

One-tenth ml of the cell suspension was added to each well of a 96-well

microtiter plate. To each well was added 100 |il of supernatant fluid

dilution from stimulated macrophage cultures. The D10.G4.1 cells were

incubated 72h and 0.5 jiC of 3H-labeled thymidine (Amersham, Arlington

141

Heights, IL) was added during the last 8 h. Samples were harvested onto

filter paper discs with an automatic sample harvester (Bellco, Vineland,

NJ) and counted using liquid scintillation.

142

RESULTS

Development of gross cecal lesions SCID mice, when

compared to C3H/HeN mice, were as susceptible to development of

catarrhal cecitis after S. hyodysenteriae challenge as were conventional

C3H/HeN mice. Figure 1 illustrates that at doses of 1 x 10^ and 1 x 10^

spirochetes, 100% of the SCID mice showed severe lesion development. In

addition, all challenged mice were culture positive for S. hyodysenteriae.

With lower challenge doses, decreasing incidences of lesions were

observed, with the SCID mice showing equal or greater susceptibility to

lesion development as C3H/HeN mice challenged with 1 x 10^ spirochetes

(Fig. 2).

While mean cecal lesion scores were not significantly different at

p<.05 (Student's t-distribution) between SCID and C3H/HeN challenged

with 1 X 103 spirochetes, the SCID mice were 100% culture positive for S.

hyodysenteriae (Fig. 2) while C3H/HeN mice were only 20% culture

positive. Individual sera were tested post-necropsy for serum antibodies by

capture ELISA; SCID mice did not show circulating antibody (data not

shown).

143

i M S « u

1 a 3

2 e

2 -

1 -

-7-.

1 z

8 • C3H/HeN, 10

a SCID, 10^

m SCID, 10^

0 SCID, non-infected

n=6 n=4 n=4 n=2

Figure 1. Development of gross cecal lesions in SCID and C3H/HeN mice c h a l l e n g e d w i t h e i t h e r 1 x 1 0 ^ o r 1 x 1 0 ^ B 2 0 4 S . hyodysenteriae. Mice were necropsied at 15 dPI as described in Materials and Methods. Cecal scores range from 0 (no grossly evident lesions) to 3 (severe cecal lesions). Bars indicate standard error of the mean. All challenged mice were culture positive for S. hyodysenteriae. n indicates group size.

144

1 IZ3 S O) u

I a

I O

• C3H/HeN, 10

El SCID.lO^

0 SCID.IO^

S SCID, noninfected

n=5 n=4 n=4 n=2

Figure 2. Development of gross cecal lesions in SCK) and C3H/HeN control infected mice challenged with either 10^ or 10^ B204 S. hyodysenteriae and necropsied at 15 dPI as described in Materials and Methods. Cecal scores range from 0 (no grossly evident lesions) to 3 (severe cecal lesions). Bars indicate standard error of the mean. Percentages above bars represent mice in the group which were culture positive for S. hyodysenteriae. n indicates group size.

145

Histopathologic findings In both C3H/HeN and SCID mice, cecal

crypt elongation and goblet cell hyperplasia was observed. In addition,

some ceca from SCID mice exhibited necrosis and dismption of the crypt

shoulder epithelium. The most noteable difference observed between the

infected SCID and C3H/HeN mice cecal tissue was the higher numbers of

neutrophils present in SCID lamina propria, especially in areas subjacent to

the base of the crypts.

Ability of SCID and C3H/HeN mouse cells to produce IL-1

and TNF in vitro PEC from SCID mice, when stimulated with E. colt

or S. hyodysenteriae endotoxin, produced comparable amounts of IL-1

activity when compared to that produced by C3H/HeN macrophages

(Figures 3 and 4). Differences in IL-1 activity production were not

observed between cells from infected or non-infected (control) SCID mice

groups. Similarly, C3H/HeN and SCID peritoneal macrophages produced

comparable amounts of TNF activity when stimulated with bacterial

endotoxin (Fig. 5 and 6). Again, no differences in ability to produce TNF

activity was noted between infected and non-infected SCID mice.

Figure 3. Production of IL-1 activity by SCID and C3H/HeN macrophages

in vitro. IL-1 activity from pooled macrophage supematants

(1:20 dilution) from C3H/HeN mice, challenged with 1 x 10^

spirochetes, and from SCID mice challenged with either 1 x 10^

or 1 X 107 spirochetes, is represented as stimulation indicies as

described in Materials and Methods. Stimulants used were

various dilutions of E. coli or S. hyodysenteriae endotoxin,

notated in |ig. SCID control mice were non-infected. Group

sizes are the same as listed in Fig. 1.

stimulation Index

Figure 4, Production of IL-1 activity by SCID and C3H/HeN macrophages

in vitro. IL-1 activity from pooled macrophage supematants

(1:20 dilution) from C3H/HeN mice, challenged with 1 x 103

spirochetes, and from SCID mice challenged with either 1 x 103

or 1 X 102 spirochetes, is represented as stimulation indicies as

described in Materials and Methods. Stimulants used were

various dilutions of E. coli or S. hyodysenteriae endotoxin,

notated in |ig. SCID control mice were non-infected. Group

sizes are the same as listed in Fig. 2.

stimulation Index

t—' vo

Figure 5, Production of TNF activity by C3H/HeN and SCID peritoneal

macrophages in vitro. Supematants (1:10) from pooled

peritoneal macrophages stimulated with (jig) E. coli or S.

hyodysenteriae endotoxin were tested for L929 cell cytotoxicity

as described in Materials and Methods from C3H/HeN mice

infected with 1 x 10^ or from SCID mice infected with either 1

X 108 or 1 X \Çp spirochetes. SCID control mice were non-

infected, Group sizes are listed in Fig. 1.

Percent Cytotoxicity

—k. VO OA —A CD VO O

Figure 6. Production of TNF activity by C3H/HeN and SCID peritoneal

macrophages in vitro. Supematants (1:10) from pooled

peritoneal macrophages stimulated with (jig) E. coli or S.

hyodysenteriae endotoxin were tested for L929 cell cytotoxicity

as described in Materials and Methods from C3H/HeN mice

infected with 1 x 103 or from SCID mice infected with either 1

X 103 or 1 X 102 spirochetes. SCID control mice were non-

infected. Group sizes are listed in Fig. 2.

Percent cytotoxicity

' J \ » A L \iim\>ui\»ni\»n\inn\mi\u>

o o o

I X

t 0

Ln U)

O — ^ - t ^ O C O - P ^ C / l O i ^ O O C O - ^ o o o o o o o o c ^ o o

154

DISCUSSION

The SCID syndrome has been reported in humans, Arabian horses,

and mice (4). A homozygous autosomal recessive mutation on

chromosome 16 results in a severe dysfunction of T and B lymphocytes

(2). Lymphoid organs are very small in SCID animals, and Peyer's patches

in the intestinal tract are nonexistant or rare (4, 6). The SCID mutation

does not affect nonlymphoid lineages; mast cells, NK cells, and myeloid

cells are present and respond to bacterial stimuli (4). Macrophage

activation in vivo does occur in the absence of T and B cells in SCID mice

(7). The reason for the lack of functional T and B cells in the SCID mouse

appears to be a defective recombinase system, where SCID lymphocytes

cannot join coding regions of V, D, and J segments of the T cell receptor

or of antibody molecules (2).

While SCID mice lack functional T or B lymphocytes, CB-17 SCID

mice, which partially resist Listeria monocytogenes infection, still

markedly increase la antigen expression on various cell types after L.

monocytogenes challenge (7). It was later implicated that, in the same

mouse system, the observed resistance to L. monocytogenes was dependent

on NK cell gamma interferon secretion (24). Activation of these CB-17

SCID NK cells appeared to require two signals: TNF alpha, and a second

signal, of presumed bacterial origin and possibly processed by

mononuclear phagocytes (24).

155

The present experiment demonstrates that S. hyodysenteriae causes

characteristic lesion development in vivo regardless of the immune

competence of the host. Lesion development was observed at comparable

frequencies or at heightened frequencies when compared to C3H/HeN

infected controls. Gross lesions appeared similar; however, SCID tissues

exhibited greater numbers of neutrophils within cecal lamina propria.

The virulence factors of S. hyodysenteriae have been postulated to be

its LPS (19, 20) and a beta-hemolysin (13, 21). With development of

lesions in mice, these moieties would affect cecal mucosal epithelial cells,

macrophages, NK cells, intra-epithelial lymphocytes, and stromal tissue. S.

hyodysenteriae infection in the mouse or pig is confined to the mucosa (1);

the organism does not invade past the lamina propria, therefore a local host

response is important in determining the character of the lesion. This

work has also shown S. hyodysenteriae to be only a mucosal pathogen in

the SCID mouse as determined by histological analysis.

In vitro, production of IL-1 and TNF by SCID peritoneal

macrophages were comparable with those from C3H/HeN mice, showing

that SCID macrophages could be stimulated to release cytokines after

exposure to S. hyodysenteriae endotoxin. This finding is in agreement with

previous work, where SCID splenic macrophages secreted TNF in response

to killed Listeria monocytogenes cells (3). Indeed, cytokine release in vivo

in the S. hyodysenteriae-mfected SCID mouse could be inferred by the

relatively larger neutrophil infiltration in SCID mice (when compared to

C3H/HeN) cecal lamina propria during infection. With grossly evident

cecal lesions in S. hyodysenteriae-chaWenged mice being similar between

156

C3H/HeN and SCK) mice, the role of functional T and B lymphocytes

being involved in lesion production appears minimal. From these data it

might be postulated that the cell type most important in the induction of

lesions following challenge with S. hyodysenteriae might be the

macrophage, which, through some form of stimulation by the spirochete,

would secrete inflammatogenic cytokines and increase local infiltration by

neutrophils. Further work utilizing mice with the SCK) mutation, and

passive transfer of serum, cells, or both and S. hyodysenteriae challenge

might elucidate protective mechanisms and protective epitopes involved in

the disease.

157

REFERENCES

1. Albassam, M. A., H. J. Olander, H. L. Thacker, and J. J. Turek. 1985. Uitrastmctural characterization of colonic lesions in pigs inoculated with Treponema hyodysenteriae. Can. J. Comp. Med. 49:384-390.

2. Ansell, J. D., and G. J. Bancroft. 1989. The biology of the SCID mutation. Immunology Today 10:322-325.

3. Bancroft, G. J., K. C. F. Sheehan, R. D. Schreiber, and E. R. Unanue. 1989. Tumor necrosis factor is involved in the T cell-independent pathway of macrophage activation in scid mice. J. Immunol. 143:127-130.

4. Bosma, M. J. 1989. The scid mouse: a model for severe combined immune deficiency, p. 1-11. hi B. Wu and J. Zheng (eds.). Immune deficient animals in experimental medicine. 6th Int. Workshop on Immune-Deficient Animals, Beijing, 1988. Karger, Basel.

5. Brandenburg, A. C., Miniats, O. P., Geissinger, H. D., and E. Ewert. 1976. Swine dysentery: inoculation of gnotobiotic pigs with Treponema hyodysenteriae and Vibrio coli and a Peptrostreptococcus. Can. J. Comp. Med. 41:294-301.

6. Custer, R. P., Bosma, G. C., and M. J. Bosma. 1985. Severe combined immunodeficiency (SCID) in the mouse. Am. J. Pathol. 120:464-477.

7. Deschryver-Kecskemeti, K., Bancroft, G. J., Bosma, G.l C., Bosma, M. J., and E. R. Unanue. 1988. Pathology of Listeria infection in murine severe combined immunodeficiency. A study by immunohistochemistry and electron microscopy. Lab. Invest. 58:698-705.

8. Harris, D. L., Alexander, T. J. L., Whipp, S. C., Robinson, I. M., Glock, R. D., and P. J. Matthews. 1978. Swine dysentery: studies of gnotobiotic pigs inoculated with Treponema hyodysenteriae.

158

Bacteroides vulgatus, and Fusobacterium necrophorum. J. Am. Vet, Med. Assoc. 172:468-471.

9. Harris, D.L., and R. D. Glock. 1986. Swine dysentery and spirochetal diseases, p. 494-507. M A. D. Leman, B. Straw, R. D. Glock, et al. (eds.). Diseases of swine, 6th ed., Iowa State University Press, Ames, Iowa.

10. Joens, L. A., and R. D. Glock. 1979. Experimental infection in mice with Treponema hyodysenteriae. Infect. Immun. 25: 757-760.

11. Joens, L. A., Robinson, I. M., Glock, R. D., and P. J. Matthews. 1981. Production of lesions in gnotobiotic mice by inoculation with Treponema hyodysenteriae. Infect. Immun. 31:504-506.

12. Kunkle, R. A., and J. M. Kinyon. 1988. Improved selective medium for the isolation of Treponema hyodysenteriae. J. Clin. Microbiol. 26:2357-2360.

13. Lemcke, R. M., and M. R. Burrows. 1982. Studies on a haemolysin produced by Treponema hyodysenteriae. J. Med. Microbiol. 15:205-214.

14. Meyer, R. C., Simon, J., and C. S. Byerly. 1974. The etiology of swine dysentery. I. Oral inoculation of germ-free swine with Treponema hyodysenteriae and Vibrio coli. Vet. Pathol. 11:515-526.

15. Mysore J. V., Duhamel, G. E., and M. R. Mathiesen. 1992. Morphologic analysis of enteric lesions in conventional and streptomycin-treated inbred C3H/HeN mice infected with Serpulina (Treponema) hyodysenteriae. Lab. Animal Science 42:7-12.

16. Nibbelink, S. K., and M. J. Wannemuehler. 1990. Effect of Treponema hyodysenteriae infection on mucosal mast cells and T cells in the murine cecum. Infect. Immun. 58:88-92.

17. Nibbelink, S. K., and M. J. Wannemuehler. 1991. Susceptibility of inbred mouse strains to infection with Serpula {Treponema) hyodysenteriae. Meet. Immun. 59:3111-3118.

159

18. Nibbelink, S. K., and M. J. Waimemuehler. 1992. An enhanced murine model for studies on the pathogenesis of Serpulina hyodysenteriae. Infect. Immun. 60:(in press).

19. Nuessen M. E., L. A. Joens, and R. D. Glock. 1983. Involvement of lipopolysaccharide in the pathogenicity of Treponema hyodysenteriae. J. Immunol. 131:997-999.

20. Nuessen, M. E., J. R. Birmingham, and L. A. Joens. 1982. Biological activity of a lipopolysaccharide extracted from Treponema hyodysenteriae. Infect. Immun. 37:138-142.

21. Saheb S. A., L. Massicotte, and B. Picard. 1980. Purification and characterization of Treponema hyodysenteriae hemolysin. Biochimie 62:779-785.

22. Suenaga, I., and T. Yamazaki. 1983. Experimental infection with Treponema hyodysenteriae in nude mice. Zbl. Bakt. Hyg. A. 254:528-533.

23. Suenaga, I., and T. Yamazaki. 1984. Experimental Treponema hyodysenteriae infection in mice. Zentralbl. Bakteriol. Mikrobiol. Hyg. Reihe A 257:348-356.

24. Wherry, J. C., Schreiber, R. D., and E. R. Unanue. 1991. Regulation of gamma interferon production by natural killer cells in SCID mice: roles of tumor necrosis factor and bacterial stimuli. Infect. Immun. 59:1709-1715.

25. Whipp, S. C., J. Pohlenz, D. L. Harris, I. M. Robinson, R. D. Glock, and R. Kunkle. 1982. Pathogenicity of Treponema hyodysenteriae in uncontaminated gnotobiotic pigs. p. 27 M: 7th Int. Pig Vet. Soc. Congress, Mexico City, Mexico.

26. Whipp, S. C., Robinson, I. M., Harris, D. L., Glock, R. D., Matthews, P. J., and T. J. L. Alexander. 1979. Pathogenic synergism between Treponema hyodysenteriae and other selected anaerobes in gnotobiotic pigs. Infect. Immun. 26:1042-1047.

160

GENERAL SUMMARY

The data presented in this dissertation have underscored the need to

fully understand disease pathogenesis in order to formulate efficacious

vaccines. Firstly, since many laboratories are using the mouse model in S.

hyodysenteriae research, this work has defined susceptibilities of various

inbred strains of mice, which, if needed, could be used in adoptive cellular

transfer techniques. The Ips gene in mice did affect susceptibility to

challenge with S. hyodysenteriae, as reported previously. As described in

paper 1, the hypo-responsive Ips^/lpsd phenotype did not totally abrogate

lesion development following S. hyodysenteriae challenge. A multi­

factorial pathogenic process, involving a host response to bacterial

endotoxin, among other instigating factors, is suggested for S.

hyodysenteriae.

Although S. hyodysenteriae endotoxin did not elicit large increases in

serum tumor necrosis factor activity in vivo in pigs, other work in our

laboratory suggests an increase in serum TNF activity in dysenteric pigs.

Since the disease has an inflammatory component, and epithelial

architecture is disrupted during the disease, it would be expected that after

the disease is initiated large amounts of bacterial cell wall components,

including LPS, would be in contact with colonic lamina propria. Tliis

"inappropriate" or "exaggerated" exposure to LPS may be the initiating

factor in relation to the increases in semm TNF activity (as well as

increasing ICAM and ELAM adhesion molecules for neutrophil

161

extravasation) in diseased pigs. The actual need for increases in TNF

locally (within the affected mucosa) to initiate disease was not determined

and would be a subject of future research efforts. Specifically, future

projects might address the increase in tissue TNF mRNA early in the

infection.

This work has also addressed the importance of secondary bacteria in

the disease pathogenesis, which effected not only gross lesions in

gnotobiotic animals, but also seemed to affect susceptibility of conventional

animals through dietary changes or ingestion of oral antibiotics. Secondary

bacteria may facilitate production of active hemolysin in vivo by S.

hyodysenteriae; Lysons et al. (53) have shown that gnotobiotic pig colonic

mucosa is susceptible to damage by spirochetal hemolysin preparations.

Since monoassociation of gnotobiotic animals frequently does not induce

mucosal damage, the question of the presence of active hemolysin in S.

hyodysenteriae-monoQiSSodiàtQà animals is raised. Additionally, the

contribution by the secondary bacteria of various cellular components

might also act in synergy with S. hyodysenteriae to cause disease in bi-

associated gnotobiotic animals.

The actual contribution of the secondary bacteria to disease

pathogenesis is unclear, but it is hypothesized that the germfree mouse

model in S. hyodysenteriae research is not artifactual; conventional mice,

treated with oral antibiotics and exhibiting decreased levels of cecal

microflora, are also incapable of developing lesions after S. hyodysenteriae

challenge.

162

BIBLIOGRAPHY

1. Albassam, M. A., H. J. Olander, H. L. Thacker, and J. J. Turek. 1985. Ultrastmctural characterization of colonic lesions in pigs inoculated with Treponema hyodysenteriae. Can. J. Comp. Med. 49:384-390.

2. Alexander, T. J. L., P. D. Wellstead, and M. J. Hudson. 1976. Studies of bacteria other than Treponema hyodysenteriae which may contribute to the lesion of swine dysentery, p. LI In: Proceedings, 4th Cong. Int. Pig Vet. Soc.

3. Bauer, H., R. B. Horowitz, S. M. Levenson, and H. Popper. 1963. The response of the lymphatic tissue to the microbial flora. Studies on germfree mice. Am. J. Pathol. 42:471-484.

4. Bealmear, P. M., O. A. Holterman, and E. A. Mirand. Influence of the microflora on the immune response, p 335-384, In: Coates, M. E., and B. E. Gustaffson (eds.). The germ-free animal in biomedical research, London, Laboratory Animals, LTD, 1984.

5. Beabner, P. M., and R. Wilson. 1967. Antibody responses of neonatally thymectomized germfree mice to Salmonella typhimurium H antigen. Exp. Hematol. 12:37-40.

6. Beutler, B., I. W. Milsark, and A. Cerami. 1987. Cachectin/tumor necrosis factor: production, distribution, and metabolic fate in vivo. J. Immunol. 135:3972-3977.

7. Beutler, B., J. Mahoney, N. LeTrang, P. Pekala, and A. Cerami. 1985. Purification of cachectin, a lipoprotein lipase-suppressing hormone secreted by endotoxin-induced RAW 264.7 cells. J. Exp. Med. 161:984-995.

8. Billiau, A. 1988. Gamma interferon: the match that lights the fire? Immunol. Today 9:37-40.

163

9. Bowden, C. A., L, A. Joens, and L. K. Kelly. 1989. Characterization of the attachment of Treponema hyodysenteriae to Henle intestinal epithelial cells in vitro. Am. J. Vet. Res. 50:1481-1485.

10. Brandenburg, A. C., O. P. Miniats, H. D. Geissingwer, and E. Ewert. 1977. Swine dysentery: inoculation of gnotobiotic pigs with Treponema hyodysenteriae and Vibrio coli and a Peptostreptococcus. Can. J. Comp. Med. 41:294-301.

11. Carswell, E. A., L. J. Old, R. L. Kassel, S. Greer, N. Fiore, and B. Williamson. 1975. An endotoxin-induced serum factor that causes necrosis of tumors. Proc. Natl. Acad. Sci. USA 72:3666-3670.

12. Chatfield, S. N., D. S. Femie, J. Beesley, C. Penn, and G. Dougan. 1988. Characterization of the cell envelope of Treponema hyodysenteriae. FEMS Microbiol. Letters 55:303-308.

13. Collins, T., L. A. LaPierre, W. Fiers, J. L. Strominger, and J. S. Pober. 1986. Recombinant human tumor necrosis factor increases mRNA levels and surface expression of HLA-A, B antigens in vascular endothelial cells and dermal fibroblasts in vitro. Proc. Natl. Acad Sci. USA 83:446-450.

14. Dinarello, C. A. 1984. Interleukin 1. Rev. Infect. Dis. 6:51-95.

15. Doll, J. P. 1962. Rate of carbon clearance in three strains of germfree mice. Am. J. Physiol. 203:291-295.

16. Doyle, L. P. 1944. A Vibrio associated with swine dysentery. Am. J. Vet. Res. 5:3-5.

17. Doyle, L. P. 1945. Swine dysentery. J. Am. Vet. Med. Assoc. 106:26-28.

18. Drayer, J. M., B. Beutler, and A. Cerami. 1985. Cachectin/tumor necrosis factor stimulates collagenase and prostaglandin E2 production by human synovial cells and dermal fibroblasts. J. Exp. Med. 162:2163-2168.

164

19. Dubois, C., E. Bissonette, and M. Rola-Pleszczynski. 1989. Platelet activating factor (PAF) enhances tumor necrosis factor produced by alveolar macrophages. J. Immunol. 143:964-970.

20. Eliakim, R., F. Karmeli, E. Razin, and D. Rachmilewitz. 1988. Role of platelet activating factor in ulcerative colitis. Gastroenterology 95:1167-1172.

21. Fisher, L. F. and H. J. Olander. 1981. Shedding of Treponema hyodysenteriae, transmission of disease, and agglutinin response of pigs convalescent from swine dysentery. Am. J. Vet. Res. 42:450-455.

22. Glock, R. D. 1971. Studies on the etiology, hematology, and pathology of swine dysentery. Ph.D. dissertation, Iowa State University.

23. Glock, R. D., K. J. Schwartz, and D. L. Harris. 1978. Parenteral immunization of pigs against infection with Treponema hyodysenteriae. Am. J. Vet. Res. 39:639-642.

24. Harris, D. L. and R. D. Glock. 1986. Swine dysentery and spirochetal diseases, p. 494-507. In A. D. Leman, B. Straw, R. D. Glock, W. L. Mengeling, R. H. C. Penny, and E. Scholl (ed.), Diseases of swine, 6th ed. Iowa State University Press, Ames.

25. Harris, D. L., J. M. Kinyon, M. T. Mullin, and R. D. Glock. 1972. Isolation and propogation of spirochetes from the colon of swine dysentery affected pigs. Can. J. Comp. Med. 36:74-76.

26. Harris, D. L., R. D. Glock, C. R. Christensen, and J. M. Kinyon. 1972. Swine dysentery-I. Inoculation of pigs with Treponema hyodysenteriae (new species) and reproduction of the disease. Vet. Med. Small Animal Clin. 67:61-68.

27. Harris, D. L., T. J. L. Alexander, S. C. Whipp, I. M. Robinson, R. D. Glock, and P. J. Matthews. 1978. Swine dysentery: studies of gnotobiotic pigs inoculated with Treponema hyodysenteriae, Bacteroides vulgatus, and Fusobacterium necrophorum. J. Am. Vet. Med. Assoc. 172:468-471.

165

28. Hsueh, W., F. Gonzalez-Crussi, and J. L. Arroyave. 1986. Platelet-activating factor-induced ischemic bowel necrosis. Am. J. Pathol. 122:231-240.

29. Hudson, M. J., T. J. L. Alexander, and P.J. Lysons. 1976. Swine dysentery: protection of pigs by attenuated Treponema hyodysenteriae. Res. Vet. Sci. 21:366-367.

30. Hudson, M. J., T. J. L. Alexander, and R. J. Lysons. 1974. Swine dysentery: failure of an attenuated strain of spirochete, given orally, to protect pigs against subsequent challenge. Br. Vet.J. 130:37-41.

31. Hughes, R., H. J. Olander, and C. B. Williams. 1975. Swine dysentery: pathogenicity of Treponema hyodysenteriae. Am. J. Vet. Res. 36:971-977.

32. Issekutz, A. C., and S. Bhimji. 1982. Role for endotoxin in the leukocyte infiltration accompanying Escherichia coli inflammation. Infect. Immun. 36:558-566.

33. Jaattela, M. 1991. Biologic activities and mechanisms of action of tumor necrosis factor-alpha/cachectin. Lab. Invest. 64:724-742.

34. James, H. D. and L. P. Doyle. 1947. Further studies with a Vibrio as the etiologic agent of swine dysentery. J. Am .Vet. Med. Assoc. 111:417.

35. Jenkinson, S. R. and C. R. Wingar. 1981. Selective medium for the isolation of Treponema hyodysenteriae. Vet. Rec. 109:384-385.

36. Joens, L. A. and R. B. Marquez. 1986. Molecular characterization of proteins from porcine spirochetes. Infect. Immun. 54:893-896.

37. Joens, L. A. and R. D. Glock. 1979. Experimental infection in mice with Treponema hyodysenteriae. Inf. Immun. 25:757-760.

38. Joens, L. A., D. L. Harris, and D. H. Baum. 1979. Immunity to swine dysentery in recovered pigs. Am. J. Vet. Res. 40:1352-1354.

166

39. Joens, L. A., D. L. Harris, and J. M. Kinyon. 1978. Micro titration agglutination for detection of Treponema hyodysenteriae antibody. J. Clin. Microbiol. 8:293-298.

40. Joens, L. A., I. M. Robinson, R. D. Clock, and P. J. Matthews. Production of lesions in gnotobiotic mice by inoculation with Treponema hyodysenteriae. Infect. Lnmun. 31:504-506.

41. Joens, L. A., J. C. Songer, D. L. Harris, and R. D. Clock. 1978. Experimental infection with Treponema hyodysenteriae in guinea pigs. Inf. Immun. 22:132-135.

42. Joens, L. A., R. D. Clock, and J. M. Kinyon. 1980. Differentiation of Treponema hyodysenteriae from Treponema innocens by enteropathogenicity testing in the CFl mouse. Vet. Rec. 107:527-529.

43. Joens, L. A., R. D. Clock, S. C. Whipp, I. M. Robinson, and D. L. Harris. 1981. of Treponema hyodysenteriae dûoâ. synergistic anaerobic bacteria in colonic lesions of gnotobiotic pigs. Vet. Microbiol. 6:69-77.

44. Kennedy, C. A., and A. C. Strafuss. 1976. Scanning electron microscopy of the lesions of swine dysentery. Am. J. Vet. Res. 37:395-401.

45. Kennedy, M. J., D. K. Rosnick, R. C. Ulrich, and R. J. Yancey. 1988. Association of Treponema hyodysenteriae with porcine intestinal mucosa. J. Cen. Microbiol. 134:1565-1576.

46. Kent, K. A., and R. M. Lemcke. 1984. Purification and cytotoxic activity of a haemolysin produced by Treponema hyodysenteriae. p. 785, Abstracts of the hitemational Pig Veterinary Society, Ghent, Belgium.

47. Kent, K, A., R. M. Lemcke, and R. J. Lysons. 1988. Production, purification and molecular weight determination of the haemolysin of Treponema hyodysenteriae. J. Med. Microbiol. 27:215-224.

167

48. Kinyon, J. M. and D. L. Harris. 1974. Growth of Treponema hyodysenteriae in liquid medium. Vet. Rec. 95:219-220.

49. Kiyono, H., J. R. McGhee, and S. M. Michalek. 1980. Lipopolysaccharide regulation of the immune response: comparison of responses to LPS in germfree, Escherichia co//-monoassociated, and conventional mice. J. Immunol. 124:36-41.

50. Knoop, F. C. 1981. Investigation of a hemolysin produced by enteropathogenic Treponema hyodysenteriae. Infect. Immun. 31:193-198.

51. Kunkle, R. A. and J. M. Kinyon. 1988. Improved selective medium for the isolation of Treponema hyodysenteriae. J. Clin. Microbiol. 26:2357-2360.

52. Lemcke, R. M. and M. R. Burrows. 1982. Studies on a haemolysin produced by Treponema hyodysenteriae. J .Med. Microbiol. 15:205-214.

53. Lysons, R. J., K. A. Kent, A. P. Bland, R. Sellwood, W. F. Robinson, and A. J. Frost. 1991. A cytotoxic haemolysin from Treponema hyodysenteriae-dL probable virulence determinant in swine dysentery. J. Med. Microbiol. 34:97-102.

54. MacDonald, T. T., S. G. Simkins, G. L. Spitalny, and P. B. Carter. 1978. Macrophage activation in the germfree mouse. 16th Ann. Mtg. of the Assoc. for Gnotobiotics, Madison, WI, p. 10.

55. McGhee, J. R., H. Kiyono, S. M. Michalek, J. L. Babb, D. L. Rosenstreich, and S. E. Mergenhagen. 1980. Lipopolysaccharide (LPS) regulation of the immune response: T lymphocytes from normal mice suppress mitogenic and immunogenic responses to LPS. J. Immunol. 124:1603-1607.

56. Meyer, R. C., J. Simon, and C. S. Byerly. 1974. The etiology of swine dysentery I. Oral inoculation of germ-free swine with Treponema hyodysenteriae and Vibrio coli. Vet. Path. 11:515-526.

168

57. Meyer, R. C., J. Simon, and C. S. Byerly. 1974. The etiology of swine dysentery E. Effect of a known microbial flora, weaning and diet on disease production in gnotobiotic and conventional swine. Vet. Path. 11:527-534.

58. Meyer, R. C., J. Simon, and C. S. Byerly. 1975. The etiology of swine dysentery HI. The role of selected gram negative abligate anaerobes. Vet. Path. 12:46-54.

59. Muir, S., M. D. H. Koopman, S. J. Libby, L. A. Joens, F. Heffron, and J. G. Kusters. 1992. Cloning and expression of a Serpula (Treponema) hyodysenteriae hemolysin gene. Infect. Immun. 60:529-535.

60. National Animal Health Monitoring System. 1986. USDA-APHIS ands Iowa High Technology Council. Dr. W. J. Owen, College of Veterinary Medicine, Iowa State University, Ames, Iowa.

61. Nibbelink, S. K. and M. J. Wannemuehler. 1991. Susceptibility of inbred mouse strains to infection with Serpula (Treponema) hyodysenteriae. Inf. Immun. 59:3111-3118.

62. Nibbelink, S. K., D. C. Morfitt, and M. J. Wannemuehler. 1990. Comparative effects of Treponema hyodysenteriae and Treponema innocens infection in germfree mice. p. 13 In: 71st Ann. Mtg. Conf. Res. Workers in Animal Disease, Chicago, IL.

63. Nordstoga, K. 1973. Fibrinous colitis in swine, a manifestation of Shwartzman reaction? Vet. Rec. June 30:698.

64. Nuessen, M. E., L. A. Joens, and R. D. Clock. Involvement of lipopolysaccharide in the pathogenicity of Treponema hyodysenteriae. J. Immunol. 131:997-999.

65. Olson, L. D. 1974. Clinical and pathological observations on the experimental passage of swine dysentery. Can. J. Comp. Med. 38:7-13.

66. Olson, L. D. and K. I. Dayalu. 1988. Exacerbation of swine dysentery in swine vaccinated with inactivated adjuvanted

169

Treponema hyodysenteriae. p. 122, In: Proceedings of the 10th Congress of the Litemational Pig Veterinary Society, Rio De Janeiro, Brazil.

67. Parizek, R., R. Stewart, K. Brown, and D. Blevins. 1985. Protection against swine dysentery with an inactivated Treponema hyodysenteriae bacterin. Vet. Medicine July 1985.

68. Picard, B., L. Maaicotte, and S. A. Saheb. 1979. Effect of sodium ribonucleate on the growth and the hemolytic activity of Treponema hyodysenteriae. Experientia 35:484-486.

69. Pohlenz, J. F., S. C. Whipp, I. M. Robinson, and J. A. Fagerland. 1984. A hypothesis on the pathogenesis of Treponema hyodysenteriae induced diarrhea in pigs. p. 178 In: 8th Int. Pig. Vet. Soc. Congress, Ghent, Belgium.

70. Saheb, S. A. and L. Lafleur. 1980. Characteristics of the interaction of a treponemal hemolysin with rabbit erythrocytes. Biochimie 62:787-793.

71. Saheb, S. A., L. Massicotte, and B. Picard. 1980. Purification and characterization of Treponema hyodysenteriae hemolysin. Biochimie 62:779-785.

72. Saheb, S. A., N. Daigneault-Sylvestre, and B. Picard. 1981. Comparative study of the hemolysins of Treponema hyodysenteriae and Treponema innocens. Current Microbiol. 5:87-90.

73. Saheb, S. A., R. Lallier, L. Massicotte, L. Lafleur, and S. Lemieux. 1981. Biological activity of Treponema hyodysenteriae hemolysin. Current Microbiol. 5:91-94.

74. Sayers T. J., T. A. Wiltrout, C. A. Bull, A. C. Denn, A. M. Pilaro, and B. Lokesh. 1988. Effect of cytokines on polymorphonuclear neutrophil infiltration in the mouse. J. Immunol. 141:1670-1677.

75. Sell, S., and J. L. Fahey. 1964. Relationship between gamma globulin and low serum gamma globulin in germfree mice. J. Immunol. 93:81-87.

170

76. Songer, J. G., J. M. Kinyon, and D. L. Harris. 1976. Selective medium for isolation of Treponema hyodysenteriae. J. Clin. Microbiol. 4:57-60.

77. Stanton, T. B. 1992. Proposal to change the genus designation Serpula to Serpulina gen. nov. containing the species Serpulina hyodysenteriae comb. nov. and Serpulina innocens comb. nov. Int. J. Syst. Bacteriol. 42:189-90.

78. Stanton, T. B., and D. F. Lebo. 1988. Treponema hyodysenteriae growth under various culture conditions. Vet. Microbiol. 18:177-190.

79. Stanton, T. B., N. S. Jensen, T. A. Casey, L. A. Tordoff, F. E. Dewhirst, and B. J. Paster. 1991. Reclassification of Treponema hyodysenteriae and Treponema innocens in a new genus, Serpula gen. nov., as Serpula hyodysenteriae comb. nov. and Serpula innocens comb. nov. Int. J. Syst. Bact. 41:50-58.

80. Suenaga, I. and T. Yamazaki. 1983. Experimental infection with Treponema hyodysenteriae in nude mice. Zbl. Bakt. Hyg. I Abt .Orig. A 254:528-533.

81. Sueyoshi, M. and Y. Adachi. 1990. Diarrhea induced by Treponema hyodysenteriae: a young chick cecal model for swine dysentery. Infect. Immun. 58:3348-3362.

82. Taylor, D. J and T. J. L. Alexander. 1971. The production of dysentery in swine by feeding cultures containing a spirochaete. Br. Vet. J. 127:viii-xi.

83. Taylor, D. J. and W. F. Blakemore. 1971. Spirochaetal invasion of the colonic epithelium in swine dysentery. Res. Vet. Sci. 12:177-179.

84. Teige, J., T. Landsverk, A. Lund, and H. J. Larsen. 1981. Swine dysentery: a scanning electron microscopy investigation. Acta. Vet. Scand. 22:218-225.

171

85. ter Huume, A. A. H. M., M. van Houten, S. Muir, J. G. Kusters, B. A. M. van der Zeijst, and W. Gaastra. 1992. Inactivation of a Serpula (Treponema) hyodysenteriae hemolysin gene by homologous recombination: importance of this hemolysin in pathogenesis of S. hyodysenteriae in mice. F.E.M.S. Microbiol. Letters 92:109-114.

86. Terpstra, J. I., J. P. Akkermans, and H. Ouwerkirk. 1968. Investigations into the etiology of vibrionic dysentery (Doyle) in pigs. Neth. J. Vet. Sci. 1:5-13.

87. Wannemuehler, M. J., R. D. Hubbard, and J. M. Greer. 1988. Characterization of the major outer membrane antigens of Treponema hyodysenteriae. Infect. Immun. 56:3032-3039.

88. Whipp, S. C., I. M. Robinson, D. L. Harris, R. D. Glock, P. J. Matthews, and T. J. L. Alexander. 1979. Pathogenic synergism between Treponema hyodysenteriae and other selected anaerobes in gnotobiotic pigs. Infect. Immun. 26:1042-1047.

89. Whipp, S. C., J. Pohlenz, D. L. Harris, I. M. Robinson, R. D. Glock, and R. Kunkle. 1982. Pathogenicity of Treponema hyodysenteriae in uncontaminated gnotobiotic pigs. p. 27 hi: 7th Int. Pig Vet. Soc. Congress, Mexico City, Mexico.

90. Whiting, R. A. 1924. Swine dysentery. J. Am. Vet. Med. Assoc. 17:600-610.

91. Whiting, R. A. 1928. Swine dysentery. J. Am. Vet. Med. Assoc. 28:721-728.

92. Whiting, R. A., L. P. Doyle, and R. S. Spray. 1921. Swine dysentery. Purdue University Agric. Exper. Station Bui. #257.

93. Wilcock, B. P., and H. J. Olander. 1979. Studies on the pathogenesis of swine dysentery, I. Characterization of the lesions in colons and colonic segments inoculated with pure cultures of colonic content containing Treponema hyodysenteriae. Vet. Pathol. 16:450-465.

172

ACKNOWLEDGEMENTS

Many people have helped and provided me with input in the

planning, execution, and preparation of this dissertation. Firstly, I would

like to thank my major professor, Mike Wannemuehler, for the direction

and support I have received over the years in his laboratory. I have

appreciated your scientific curiosity and knowledge, as well as providing

me the independence to follow my ideas and hypotheses. I would like to

thank my committee, Drs. Shannon Whipp, Prem Paul, F. Chris Minion,

and John Andrews for advice and help in preparing manuscripts and this

dissertation.

Technical help, provided by L. Vandermark and the late T. J. Bryant

was invaluable. Special thanks is extended to Drs. Dave Morfitt and Dave

Hutto for histopathologic interpretation of tissue sections and to J. Galvin

for help in collecting samples. I would also acknowledge and thank the

Comparative Gastroenterology Society for financial support of portions of

this work. To my fellow graduate students, who had the dubious honor of

being associated with me for varying lengths of time, I wish you

"successful" futures.

Lastly, I would thank my wife Sharon, and two daughters, Andrea

and Abby, for support and toleration during this tenure in graduate school.

Your faith in me made the endeavor worthwhile.