Iowa State University Capstones, Theses and Retrospective Theses and Dissertations Dissertations

1996 The ap thogenesis of respiratory infection of lambs by mycoplasmas Mamadou Niang Iowa State University

Follow this and additional works at: https://lib.dr.iastate.edu/rtd Part of the Microbiology Commons, and the Veterinary Pathology and Pathobiology Commons

Recommended Citation Niang, Mamadou, "The ap thogenesis of respiratory infection of lambs by mycoplasmas" (1996). Retrospective Theses and Dissertations. 12265. https://lib.dr.iastate.edu/rtd/12265

This Dissertation is brought to you for free and open access by the Iowa State University Capstones, Theses and Dissertations at Iowa State University Digital Repository. It has been accepted for inclusion in Retrospective Theses and Dissertations by an authorized administrator of Iowa State University Digital Repository. For more information, please contact [email protected]. INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter &ce, 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 comer 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. ISgher 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. UMI A Bell & Howell Information Coiiq)aiQr 300 North Zeeb Road, Ann Aibor MI 48106-1346 USA 313/761-4700 800/521-0600

The pathogenesis of respiratory infection of Iambs by mycoplasmas

by

Mamadou Niang

A dissertation submitted to the graduate faculty

in partial fulfillment of the requirements for the degree of

DOCTOR OF PHILOSOPHY

Major: Veterinary Microbiology

Major Professor: Merlin L. Kaeberle

Iowa State University

Ames, Iowa

1996 XJMI Number; 9737778

UMI Microform 9737778 Copyright 1997, by UMI Company. All rights reserved.

This microform edition is protected against unauthorized copying under Title 17, United States Code.

UMI 300 North Zeeb Road Ann Arbor, MI 48103 u

Graduate College Iowa State University

This is to certify that the Doctoral dissertation of

Mamadou Niang has met the dissertation requirements of Iowa State University

Signature was redacted for privacy. Major Professor

Signature was redacted for privacy.

For the Major Program

Signature was redacted for privacy.

For the Graduate College m

TABLE OF CONTENTS

CHAPTER 1. GENERAL INTRODUCTION

CHAPTER 2. LITERATURE REVIEW

CHAPTERS. DEMONSTRATION OF A CAPSULE ON MYCOPLASMA OVIPNEUMONIAE 39

CHAPTER 4. CYTOPATmC EFFECTS OF MYCOPLASMA OVIPNEUMONIAE FIELD ISOLATES IN OVINE TRACHEAL ORGAN CULTURES 79

CHAPTER 5. IMMUNE RESPONSE IN LAMBS NATURALLY INFECTED WITH MYCOPLASMA SPECIES 114

CHAPTER 6. OCCURRENCE OF AUTOANTIBODIES TO CILIA IN LAMBS WITH A "COUGHING SYNDROME" 135

CHAPTER 7. GENERAL CONCLUSIONS 173

APPENDDC A. ASSOCIATED OBSERVATIONS: EXPRESSION OF MHC CLASS I AND n ANTIGENS ON SHEEP ALVEOLAR MACROPHAGES 180

APPENDIX B. ASSOCIATED OBSERVATIONS: TYPE I HYPERSENSITIVITY IN LAMBS WITH COUGHING SYNDROME 191

LITERATURE CITED 197

ACKNOWLEDGMENTS 215 1

CHAPTER 1. GENERAL INTRODUCTION

The etiology of respiratory diseases in sheep is considered to be very complex and

has not been clearly defined. There is an apparent association of several microorganisms

including Pasteurella sp.. Mycoplasma sp., chlamydia, viruses, nematode parasites, fungal agents and occasionally other bacteria such as Escherichia call, Corynebacterium sp..

Staphylococcus sp., and Neisseria sp. with the disease (Alley, 1975; Davies, 1985).

Establishment of Koch's postulates has been difficult due to the complex etiology and poorly defined lesions.

Surveys conducted in many countries have indicated that, other than

Pasteurella sp.. Mycoplasma ovipneumoniae is the bacterium most commonly isolated from the ovine respiratory tract (Bakke, 1982; Pfeffer et al., 1983; Brogden et al., 1988; Malone et al., 1988). However, due to inconsistent results in experimental reproduction of the disease, its role as an important etiological factor in respiratory disease in sheep is not well accepted. In contrast, there is general acceptance that Pasteurella sp. are contributory to pneumonia in sheep, although they may have limited ability to induce pneumonia unless predisposing factors are present. Several viruses, chlamydia and M. ovipneumoniae may serve as predisposing factors.

During the past several years, an ovine respiratory disease that we have termed the

"coughing syndrome" has been observed in the mid-west states of the United States of

America. The terminology stems from primary clinical signs which include a severe 2

paroxysmal cough. The disease is chronic in young lambs with morbidity and severity

variable among flocks.

While there is an obvious need for information on the role of viruses in this clinical

syndrome, the evidence available does not favor viruses as being important in the etiology

of this disease. Such evidence includes absence of serological responses in paired serum

samples to several common sheep viruses and absence of histopathological lesions

consistent with viral infections (unpublished observations). On the other hand, evidence for

prevalence and association of mycoplasmas, in particular M. ovipneianoniae, with

outbreaks of this syndrome have been suggested by stodies conducted in this laboratory

(Kaeberle and Eness, 1988; Eness et al., 1990; Niang, 1992; Khan, 1993). Unfortunately,

however, we have not been able to routinely reproduce the clinical syndrome by exposure

of lambs to live M. ovipneianoniae or M. arginini organisms although infection and

histopathological lung lesions can be consistently demonstrated, in particular with M.

ovipneianoniae. This does not necessarily rule out a causal relationship since establishment of cause and effect, or fulfillment of Koch's postulates, has been difficult with several

mycoplasma species, even with those proven to be highly virulent such as M. mycoides subsp. mycoides (Gourlay and Howard, 1982; Stalheim, 1983; Cassell et al., 1985).

Microorganisms, including mycoplasmas, differ in their capacity to cause disease.

Some are highly virulent and can readily produce disease when they infect; others with moderate virulence will produce disease by progressively deteriorating conditions of the

host defense mechanisms, allowing opportunistic proliferation of secondary 3 microorganisms. Most pathogenic mycoplasmas, including those that are now being recognized as primary etiological agents in several respiratory diseases such as enzootic pneumonia in calves and swine, fall into the latter category. The etiological significance of such mycoplasmas may only be expressed fully in association with other organisms

(Gourlay and Howard, 1978; Gourlay and Howard, 1982)

Apart from the fiilfillment of Koch's postulates, evaluation of the pathogenic potential of an organism may be of value in determining its etiological significance

(Gourlay and Howard, 1978; Gourlay and Howard, 1982). Such evidence can include the ability of the microorganism to produce certain factors that may be of pathogenic potential such as a toxin, capsule, hydrogen peroxide, attachment molecules, ability to cause damage to any host tissue (either in vivo or in vitro), or ability to manipulate or to affect the elements of the host immune system.

In relation to the aforementioned concepts, we have attempted in the present investigation to determine the significance of mycoplasma species, in particular

M. ovipneumoniae, isolated from lambs with the coughing syndrome. The approach used was to determine specific activities or factors that may contribute to the disease producing potential of M. ovipneumoniae.

Characteristics of the disease and histopathologic lesions in the lungs of affected lambs suggested that immunopathological mechanisms contributed the disease process.

Immune mechanisms have been implicated in mycoplasmal infection in humans and several other domestic animals (Cassell et al., 1985). One approach was to smdy the 4 microorganism for mechanisms of colonization (adherence to respiratory epithelimn) and effects on epithelial cells. This led to demonstration of a capsule on the microorganism and possible role of this capsule in adherence. Also, this adherence was associated with ciliostatic activity. A second general approach was to demonstrate aberrant immune responses in association with infection of lambs by mycoplasmas, and potential role of these responses in the pathogenesis of this disease. First, a serological smdy was conducted in order to gain a better understanding of the nature of immime response against these mycoplasmas. Secondly, an allergic type of hypersensitivity to M. ovipneumoniae in these lambs was demonstrated. Thirdly, an autoimmune response associated with mycoplasmal infection in these lambs was noted and led to a study on the effects of mycoplasmas on

MHC molecule expression by sheep alveolar macrophages.

Dissertation organization. This dissertation was written using the alternate format.

It contains a general introduction, review of literature, four manuscripts, general summary and two appendices. References cited for each manuscript and each appendix follow the corresponding manuscript and the appendix, while references cited in the general introduction, literature review, and general conclusions are listed following the appendixes.

Manuscripts were written in the style of the appropriate journal to which submission was intended. The doctoral candidate, Mamadou Niang, is the senior author and principal investigator for each of the manuscripts, with the advice of the major professor. Dr. Merlin

L. Kaeberle, and other co-authors of the manuscripts. 5

CHAPTER 2. LITERATURE REVIEW

Ovine pneumonia is one of tlie most common infectious diseases of sheep

worldwide. Surveys conducted in many countries have shown that M. ovipneumoniae is probably the most conmionly isolated microbial species from the respiratory tract of pneumonic sheep (Bakke, 1982; Pfeffer et al., 1983; Brogden et al., 1988; Malone et al.,

1988; Khan, 1993). The agent has tropism for the respiratory tract where it colonizes, like many other respiratory mycoplasmas, the ciliated epithelial cells. There is, however, considerable controversy concerning the experimental reproduction of the disease with this mycoplasma and fulfillment of Koch's postulates, as with several proven highly virulent mycoplasmas, has been difficult to achieve. Still, the organism is regarded by several investigators (Nostvold and Bakke, 1982; Bakke and Nostvold, 1982; Rosendal, 1994) to be a confirmed pathogen for the sheep respiratory tract.

The mechanism of pathogenesis of mycoplasma diseases is complex and can be divided into a number of interrelated events including attachment, toxin production, and suppression of the host defense mechanisms. Also, in addition to eliciting specific immune responses, mycoplasmas have the ability of inducing immune responses that are detrimental to the host (Cassell et al., 1985). However, there is very litde information available regarding these aberrant immtme responses although they may play an important role in the pathogenesis of several diseases caused by the organisms. Therefore, this literature review 6 will address the pathogenesis of mycoplasmal infections, in general. Prior to that, a brief review of respiratory mycoplasmas in sheep is presented.

Respiratory Mycoplasmas in Sheep

The etiology of sheep respiratory diseases is extremely complex and relates to synergistic effects of several factors including environment and microorganisms. A wide variety of microorganisms have been isolated from the respiratory tract of pneumonic sheep. These include Pasteurella sp.. Mycoplasma sp., chlamydia, viruses, nematode parasites, ftmgal agents, and occasionally other bacteria such as £. coli, Corynebaaerium sp.. Staphylococcus sp.. Neisseria sp. (Alley, 1975; Davies, 1985; Malone, 1988).

Historically, the isolation of Mycoplasma sp. from the respiratory tract of sheep was first reported by Greig (1955). However, a potential role for mycoplasmas in ovine pneumonia was initially suggested by Hamdy et al. (1959) who consistently were able to reproduce the disease by stressing lambs inoculated intratrachealy with a mycoplasma agent and Pasteurella multocida. Meanwhile, there were reports of isolation of mycoplasma from the respiratory tract of pneumonic sheep in many countries (Jones et al., 1976). Also, some investigations conducted in Australia associated Mycoplasma sp. with many cases of pneumonia appearing in lambs between the age of 5-10 weeks (St. George et al., 1971;

Sullivan et al., 1973). Clinical signs reported were coughing, sneezing, a mucoid nasal discharge and reduced weight gain. While the morbidity was high the mortality was low. 7

The lungs from all animals that died or were sacrificed had pnemnonic lesions that were characterized as proliferative interstitial pneumonia. At the same time, a cytopathogenic agent was cultivated in bovine testicular cell culture and pneumonia similar to the natural disease was produced in lambs by intratracheal inoculation of the cell culture grown organisms. Further studies by Cannichael et al. (1972) confirmed the nature of the agent and they named it M. ovipneumoniae.

Since then, there have been several reports of isolation of M. oynpnewnoniae from different parts of the world, mostly in connection with respiratory disease in sheep (Bakke,

1982; Pfeffer et al., 1983; Brogden et al., 1988; Malone et al., 1988; Khan, 1993). Other

Mycoplasma sp. including M. arginini, M. conjunctivae, Acholeplasma laidlawii and ureaplasmas have also been isolated from the respiratory tract of both healthy and pneumonic sheep (Brogden et al., 1988; Davies, 1985). However, these mycoplasmas are not considered to play a major role in the disease process (Foggie and Angus, 1972;

Davies, 1985).

Attempts to produce pneumonia in sheep with pure cultures of M. ovipneumoniae have yielded conflicting results. In some early smdies conducted in Australia, nearly all animals developed severe pneumonia (St. George et al., 1971; Sullivan et al., 1973). On the other hand, only 25 to 50% of the animals developed mild pneumonia in smdies conducted in Scotland (Foggie et al., 1976; Gilmour et al., 1979), and less than 20% showed lesions in experiments conducted in New Zealand (Alley and Clarke, 1977; Davies et al., 1981). Attempts to recover M. ovipneumoniae from the pneimionic lungs in all these 8

experiments usually were unsuccessfiil and, even when M. ovipneumoniae was present,

bacterial numbers in both healthy and pneumonic lungs were similar. These observations

and theiact that M. ovipneumoniae could also be isolated from the respiratory tract of

normal sheep raised the question of whether the organism is part of normal flora of sheep.

However, attempts to produce disease in lambs with a homogenate of pneumonic limg containing M. ovipneumoniae alone or in combination with P. haemofytica yielded proliferative pneumonia in most animals (Jones et al., 1982a). These findings led to the hypothesis that the challenge organisms in the early smdies might have been attenuated during isolation and cloning on artificial medium or that strains varied in virulence (lonas et al., 1985; Jones et al., 1982a; Jones et al., 1982b). In this regard, it is interesting to note that the early Australian smdies were conducted utilizing tissue culture grown organisms

(Sullivan et al., 1973) while the other investigators utilized artificial mediimi grown organisms. Another factor contributing to variability in the challenge studies could be the nature of the experimental animals. Naturally reared lambs were more likely to develop pneumonia than were specific-pathogen free animals (Jones et al., 1982a). It was also suggested that isolation of M. ovipneumoniae from normal sheep should not be an indication that the organism is part of the normal flora of sheep since only a low percentage of normal sheep carried the organism and many of these normal positive sheep were negative when retested (lonas, 1985). This transitory carriage of M. ovipneumoniae in normal sheep was probably important and would provide an obvious source from which young lambs could be naturally infected (lonas al., 1985). Moreover, reported 9

experimentation indicated presence of M. ovipneumoniae in large numbers in pneumonic

sheep and absence or presence in low numbers in normal controls, suggesting a direct or

indirect role in ovine pneumonia (Alley et al., 1975; Bakke, 1982; Buddie et al., 1984).

Pathogenesis of Mycoplasmal Infection

Assessment of Mycoplasma Pathogenicity

Most mycoplasma diseases are multifactorial in which a variety of host, bacterial and environmental factors determine the final outcome of infection. Fulfillment of Koch's postulates, the usual accepted criterion for proof of etiological significance, has proved to be difficult to achieve even with M. mycoides, the most pathogenic of mycoplasmas

(Gourlay and Howard, 1982, Stalheim, 1983; Cassell et al., 1985). The reasons for these difficulties are poorly understood. However, it is becoming more and more clear that some of the assumptions implicit in Koch's postulates are limited. For instance, complex requirements for in vitro growth of mycoplasmas renders fulfillment of Koch's postulates difficult to achieve. This delayed for many years the acceptance of M. mycoides subsp. mycoides as the etiological agent of contagious bovine pleuropneumonia (Gourlay and

Howard, 1982). Mycobacterium leprae, the causative agent of leprosy, has not yet been cultured in laboratory medium (Salyers and Whitt, 1994). Other difficulties in achieving the criteria of Koch's postulates during the experimental production of disease include loss of virulence, strain variation and susceptibility of the experimental animal In addition. 10 mycoplasmas, in general, are not among the most highly pathogenic microorganisms. Also, because they primarily have predilection for mucosal surfaces and generally establish persistent superficial infections, their effects are less immediate and more subtle.

Moreover, many of the diseases in which mycoplasmas are associated are diseases of complex etiology. Their effects may be heavily influenced by the presence of other bacteria, such as Pasteurella sp. which may be conmion inhabitants of mucosal surfaces of clinically normal animals (Gourlay and Howard, 1978).

Apart firom animal models, various organ culture model systems have been extensively used to smdy the pathogenicity of mycoplasmas. These models have provided valuable information complementing that obtained from natural and experimental infections by mycoplasmas. In these model systems, subjective measurement of ciliary activity, light and electron microscopic observations, measurements of kinetics of oxygen uptake, ATP content, calmodulin levels, and dehydrogenase enzyme measurements have been used to assess explant epithelial cell damage (Gabridge, 1986; Bermudez et al., 1992). These systems appear to be easy to control and manipulate, and to accurately mimic the natural host tissue configuration (Gabridge, 1984).

The differentiated epithelium lining mammalian mucosal surfaces comprises perhaps the single most important first line of defense against infectious agents including mycoplasmas. Mucus is made up mainly of acid glycoproteins and is thick, gel-like, viscous, and sticky. Mucosal ciliated cells contain cilia which beat in a coordinated manner, moving the mucus blanket upward toward the pharynx. These properties combine to protect the underlying mucosal surfaces against foreign invaders (Alder, 1986).

Alteration of normal functions of the mucociliary blanket can lead to a deficiency of the host defense mechanisms and can be an important factor in the disease process. The pathogenesis of several mycoplasma respiratory pathogens seems to relate, at least in part, directly to loss of mucociliary clearance mechanisms. This deficiency may contribute to the establishment of secondary bacterial infections in the lung and its relevance in vivo has been proven beyond doubt (Araake, 1982; Gourlay and Howard, 1982; Almeida and

Rosenbusch, 1994). Numerous mycoplasma respiratory pathogens including

M. hyopneumoniae G^ebey and Ross, 1994), M. pneumoniae (Gabridge et al., 1974),

M. gallisepticum (Levishon et al., 1986), M. dispar (Thomas et al., 1987), M. mycoides subsp. capri (Cherry and Taylor-Robison, 1970), M. pulmonis (Stadtlander et al., 1991), and M. ovipneumoniae (Jones et al., 1985) have been studied using tracheal organ cultures and, in all these studies, progressive ciliostasis and epithelial exfoliation were reported.

Another in vitro model of mycoplasma infection is a monolayer cell culnire of ciliated epithelial cells (Gabridge, 1986). These ciliated cells in monolayer format can be infected with mycoplasmas and provide a unique perspective to examine the membrane- membrane interactions between host and parasites (Gabridge, 1986). Moreover, monolayers of cell types other than the natural target cells such as lung fibroblasts have proven to be very useful for investigating mycoplasma-host cell interactions. The cytopathic effects of several mycoplasma species have been studied using this model. Also, 12

cell monolayers of Imig €n)roblasts have been used to study the adherence mechanism of

several respiratory mycoplasmas (Razin, 1985).

However, proper control of certain fectors is necessary when developing an in vitro

organ culture system which accurately reflects pathogenic mechanisms operative in vivo.

Culture conditions including variations of CO^ concentration in culture mediimi, type and

content of serum, nature of the host cells including species and tissue of origin, and stams

and passage history of the mycoplasma strain are factors that need to be controlled when

assessing the pathogenicity of mycoplasmas in organ cultures (Debey, 1992).

Alternatively, evaluation for evidence of certain components that may be of

pathogenic potential such as a toxin, capsule, attachment molecules, or ability to

manipulate or to affect the host immune defense systems may also be important in determining the etiological significance of an organism (Gourlay and Howard, 1978;

Gourlay and Howard, 1982).

Mechanism of Attachment to Host Tissues

The success of surface parasitism by microorganisms such as mycoplasmas that colonize the mucosal surfaces requires an adequate attachment mechanism to resist and circumvent the flushing and cleaning mechanisms that protect these surfaces and to acquire

nutrients from host cells. Those organisms that can effectively overcome these first lines of defense have a better chance of colonization and increased opportunity for induction of pathogenic changes. Hence, attachment of mycoplasmas to host cells is crucial for their 13 survival within a host (Razin, 1978) and is also a critical step in the initiation of tissue colonization and subsequent induction of cellular damage (Baseman et al., 1982).

Mycoplasmas have been demonstrated to attach to inert surfaces as well as to a variety of cells or tissues including tracheal or fallopian-tube epithelial cells, erythrocytes, lymphocytes, phagocytic cells and various other cultured mammalian cells (Collier, 1979;

Razin, 1985). The consequences of attachment of mycoplasmas are often injury to the host cells through biochemical alterations such as decreases in ATP, cyclic AMP, carbohydrate utilization, amino acid transport and oxidative metabolism (Cassell et al., 1985). However, adhesion per se should not be an indicator of pathogenicity since both pathogenic and non­ pathogenic mycoplasma species have the ability to attach to host cells, although loss of ability to attach may lead to avirulence (Gourlay and Howard, 1982).

The general ability of mycoplasmas to attach to host cells has been studied most extensively with the respiratory pathogens. The most extensively studied is

M. pneumoniae, the causative agent of atypical pneumonia in man. M. pneumoniae is known to attach to host cells by means of a specialized tip-structure which was described by Collier and Clyde Jr, (1969), and Collier and Baseman (1973). This specialized structure of M. pneumoniae is mostly concentrated at one site on the microorganism but also can be present on other areas of the membrane (Collier, 1979). In further studies it was demonstrated that trypsin treatment of M. pneumoniae could result in the loss of a 170 kDa protein designated as PI. Removal of this PI protein correlated with loss of ability of the microorganism to attach to eucaryotic cells, while regeneration of the PI protein restored this activity, suggesting that this PI protein is associated with the adherence

mechanism of the mycoplasma (Hu et al., 1977; Gorski and Bredt, 1977). Further support

for the importance of the PI protein in the adherence mechanism of M. pneumoniae has

been provided by morphologic and immunologic smdies. Monoclonal antibody against PI

reacted with cultured mycoplasmas, and was found to be localized around the tip structure

with a ferritin-labelled secondary antibody (Hu et al., 1982). Monoclonal antibodies against

the PI blocked the adherence of virulent organisms to hamster tracheal organ cultures

(Collier et al., 1983).

However, the significance of the PI protein in vivo is not clear since avirulent strains of M. pneumoniae possess the PI protein but are unable to attach (Collier et al.,

1985). In addition, immunization of guinea pigs with purified PI protein did not protect these animals against virulent M. pneumoniae, even though there was a strong humoral

response to the protein (Jacobs et al., 1988a; Jacobs et al., 1988b). Moreover, in organ cultures, Gabridge (1982) found that M. pneumoniae organisms that are located on nonciliated cells lie parallel to the plasma membrane of these cells while those on ciliated cells are oriented vertically. This finding suggested that the specialized tip-structure may not be mandatory for attachment.

Strong evidence exists to suggest that expression of PI protein alone is not sufficient for adherence of M. pneumoniae to host cells. For instance, experiments with adherent and nonadherent mutants of M. pneumoniae indicated involvement of a P30 protein in the adherence mechanism of the mycoplasma (Krause and Baseman, 1982; 15

Baseman et al., 1987). Like PI, this P30 protein is also clustered at the tip-structure of virulent Af. pneumoniae and antibodies against the protein blocked adherence of the mycoplasma to erythrocytes (Baseman et al., 1987).

Another respiratory mycoplasma with specialized structures possibly involved in attachment is M. gallisepticum which attaches to the tracheal epithelium of chickens by means of its specialized adherence organelle or bleb (Uppal and Chu, 1977). The existence of cross reactions between epitopes of PI protein of Af. pneumoniae and a protein of similar size in M. gallisepticum was noted, but adherence-inhibiting monoclonal antibodies to PI did not react with that protein.

There are many other respiratory mycoplasmas such as M. dispar,

M. hyopneumoniae and Af. pulmonis that attach to a wide variety of host cells but apparently do not have any specialized attachment organelles. Adhesion to host cells in these mycoplasmas is mediated by a general interaction of the mycoplasma membrane with the host, implying a random distribution of adhesin molecules over the whole mycoplasmal surface. As will be discussed later in this review, the capsular material of these organisms has been suggested as a mediator in the attachment process to host cells (Rosendal, 1993;

Minion and Rosenbusch, 1993). Several reports of morphological studies have suggested participation of mycoplasma capsules in the attachment process of mycoplasma species to host cells (Green and Hanson, 1973; Robertson and Smook, 1976; Wilson and Collier,

1976; Taylor-Robmson et al., 1981; Tajima and Yagihashi, 1982; Rosenbusch, 1992), but the nature of this binding process is still unclear. 16

Adhesins expressed on the surface of bacteria including mycoplasmas are usually

protein in nature. However, there are a few instances in which noiqjrotein molecules such as carbohydrate capsules can be used by the bacteria as adhesins. The capsule of many

bacteria such as Pasteurella sp., Bacteroides sp. and Pseudomonas aeruginosa have been proposed as mediators of their adhesion to host cells (Marcus and Baker, 1985; Ramphal and Pier, 1985; Inzana, 1990; Pijoan and Trigo, 1990; Baselga et al., 1994). Moreover, the involvement of lipopolysaccharide in the adhesion of Actinobacillus pleuropneumoniae, a swine respiratory pathogen, to porcine epithelial cells has been reported (Belanger et al.,

1990 ). Binding of a streptococcus to human fibronectin via lipoteichoic acid has also been described (Stanislawski et al., 1985).

Characteristics of the mycoplasma cell surface determine the capacity for adhesion

(Gourlay and Howard, 1982) and adherence is probably a multifactorial event involving both nonspecific (hydrophobic interactions) and specific Qigand-receptor interactions) mechanisms. The chemical nature of receptors on animal cells for microbial pathogens including mycoplasmas are usually glycoconjugates. They can be either glycoproteins or glycolipids, in which the carbohydrate moieties serve as receptor epitopes. The receptors for several mycoplasmas including M. pneumonia, M. gallisepticum, M. genitalium,

M. synoviae and M. bovis on host cells are neuraminidase sensitive, suggesting that glycoproteins containing sialic acid residues serve as receptors for these mycoplasmas

(Razin, 1985). However, sialylated glycoconjugates are not the sole receptors on host cells for mycoplasmas. Another common type of receptor identified for mycoplasmas is sulfated glycolipid which is a common constituent of cell membranes. Mycoplasmas that are

reported to utilize sulfated glycolipids on host cells as receptors include M. pneumoniae,

M. hominis, M. genitalium, M. salivarium, M. orale, M. hyorhinis, M. arginini,

M. pulmonis and Ureaplasma ureafyticum (Razin, 1985). More recently, Zhang et al.

(1994) studied the mechanism of adherence of M. hyopneumoniae to swine ciliated cells

and reported that sulfated glycolipids were the major native receptors for the organism on

these cells. The aforementioned mechanisms of adherence can be considered as specific

interactions between microbial surface molecules (adhesins) and complementary sugar

moieties (receptors) on host cells.

Other less specific but irreversible interactions such as hydrophobic interactions can

also play a role in the adherence mechanism of microorganisms to host cells. High surface

hydrophobicity was correlated with increased adherence properties in E. coli, in

Streptococcus mutans, and in Salmonella typhimurium (Zielinski, 1991). The cell surface

hydrophobicity of mycoplasmas has not been as well documented as for other bacteria.

However, in a recent study, Zielinski and Ross (1992) presented evidence that the cell

surface of M. hyopneumoniae is hydrophobic and may play in part in the adherence process of the mycoplasma to swine ciliated cells. Previously, Minion et al. (1984) proposed a

multiphasic model of attachment for M. pulmonis which includes an initial recognition step

mediated by long-range hydrophobic forces and followed by ligand-receptor specific

interactions. These interactions also have been suggested in the adherence of several other

pathogenic mycoplasmas to host cells (Kahane, 1984). The involvement of hydrophobic 18

interactions in the adherence process of several bacteria other than mycoplasmas has also

been documented (Zielinsid, 1991).

In regard to M. ovipneumoniae, morphologic studies have revealed its ability to

attach to a variety of host cells including phagocytic cells (Al-Kaisi and Alley, 1983) and

ovine ciliated cells (Jones et al., 1985) with profound effects on these cells. However, no

information on mechanisms of adherence to host cells has been reported.

Mechanism of Host Cell Injury

Ciliostasis is a common result of infection of tracheal organ cultures with

respiratory mycoplasmas including M. hyopneumoniae (Debey and Ross, 1994),

M. gallisepticum (Levishon et al., 1986), Af. pneumoniae (Gabridge et al., 1974),

M. pulmonis (Stadtlander et al., 1991), Af. dispar (Thomas et al., 1987), M. mycoides subsp. capri (Cherry and Taylor-Robinson, 1970) and M. ovipneumoniae (Jones et al.,

1985).

There are reports that the basis for these cytopathic effects might be oxidative attack by hydrogen peroxide generated by many of the mycoplasmas. Cherry and Taylor-

Robinson (1970) reported that the cytopathic effects induced by M. mycoides subsp. capri in chicken tracheal organ cultures were related to hydrogen peroxide produced by the organism and that the addition of exogenous catalase could delay these cytopathic effects.

Cytopathic effects on porcine and rat tracheal organ cultures by M. mobile 163 K was thought to be mediated by hydrogen peroxide generated by the organism (Stadtlander and 19

Kirchhoff, 1988). M. pneumoniae is known to actively generate superoxide and hydrogen peroxide, and these oxidants have been incriminated in cytopathological changes observed in a variety of M. pneumoniae-infectsd human cells including ciliated epithelial cells

(Almagor et al., 1983; Almagor et al., 1984; Kahane, 1984). The role of hydrogen peroxide in the pathogenicity of M. mycoides subsp. mycoides was suggested by Nicholas and Bashiruddin (1995). Furthermore, M. equigenitalium produced damage and cytotoxicity to equine uterine tube epithelial cells; the effects could be reversed by addition of catalase (Bermudez et al., 1990; Rosendal, 1994), suggesting a role for hydrogen peroxide. Also, hydrogen peroxide produced by M. equigenitalium has been proposed as possible mediator of membrane leakage of calmodulin from infected equine uterine mbe epithelial cells (reported by Debey, 1992). Additionally, there are reports which indicated that both high and low concentrations of exogenous hydrogen peroxide alone disrupted airway epithelial cells and inhibited ciliary activity in sheep (Kobayashi et al., 1992) as well as in guinea pig (Khan et al., 1990) and rat (Burman and Martin, 1986) tracheal organ cultures. Hydrogen peroxide was the most damaging oxygen radical when lung fibroblasts were exposed to superoxide, hydrogen peroxide, hydroxyi radical or singlet oxygen (Simon et al., 1981). Moreover, hydrogen peroxide was demonstrated to be responsible for lysis of erythrocytes in vitro and also could potentially damage tissue of the respiratory tract

(Clyde, 1963).

Conversely, there are pathogenic mycoplasmas, like M. hyosynoviae, that do not produce hydrogen peroxide or nonpathogenic mycoplasmas, like A. ladlawii, that produce it (Razin, 1978). Also, comparative studies on in vivo effects of virulent and homologous

attenuated strains of M. pneumoniae failed to show a consistent relationship between

pathogenicity and peroxide production (Lipman and Clyde Jr, 1969). However, hydrogen

peroxide produced by M. pneumoniae and other mycoplasmas is rapidly destroyed by host

catalase activity. Therefore, for the hydrogen peroxide to exert its toxic effect, either host

catalase must be depleted or mycoplasmas must adhere close enough to the host cell sinface

to maintain a toxic, steady state concentration of hydrogen peroxide sufficient to cause direct damage to the cell membrane (Cohen and Somerson, 1969). Cohen and Somerson

(1969) also reported the presence of peroxidase-like activity in M. pneumoniae and suggested that this may account for a lack of accumulation of hydrogen peroxide in growth media. Kahane (1984) and Almagor et al. (1984) presented evidence that the effectiveness of hydrogen peroxide may depend on both the intimate contact of the organism with the

host cell membrane and production of superoxide anions which inactivate host cell catalase; this could permit the oxidative action of hydrogen peroxide produced by both the organism and the host cells. Furthermore, catalase deficient mice were more likely to develop pneumonia after infection with M. pulmonis than normal mice, providing circumstantial evidence that in vivo production of hydrogen peroxide by the organism contributes to its virulence (Brennan and Finsten, 1969).

Some mycoplasmas cause cytopathic effects in cell cultures. This may be due to

metabolic competition for essential nutrients. Arginine-requiring mycoplasmas may deplete arginine to such an extent that cell damage results. Alternatively, ammonia or urea, die end products of arginine-requiring mycoplasmas and ureaplasma, respectively, as well as acid

production by glucose fermenting mycoplasmas may cause cell damage (Gourlay and

Howard, 1982).

Very little work on mycoplasmal toxins has been reported in the literature. A neurotoxin produced by M. neurolyticwn, the causative agent of rolling disease in mice and rats, has been described since the work of Sabin (1938). Neurotoxic properties of three other mycoplasmas species, including M. arthritidis, M. pulmonis and M. gallisepticum, were described by Thomas (1969) who demonstrated that the organisms must be alive to produce toxin and larger doses were necessary to produce toxicity. A toxin capable of eliciting an inflammatory response in the skin of guinea pigs and the mammary gland of cows was produced by M. bovis (Geary et al., 1981); however, the relevance to the cytopathic effects of this organism for organ cultures could not be demonstrated (Thomas et al., 1987). There are studies which suggest that proteins associated with the mycoplasma cell membrane are the toxic factors responsible for the cell damage caused by the mycoplasmas. Both M. pneumoniae live organisms and membrane-associated preparations induced ciliostasis in organ culmres of adult hamster trachea (Gabridge et al., 1974).

Similarly, whole cells and purified membranes of M. hyopneumoniae were reported to be cytotoxic for porcine lung fibroblasts (Geary and Walczak, 1983). In both cases, membrane associated proteins were implicated in these cytopathic effects. In contrast to these reports, however, Hu et al. (1976) were unable to induce ciliostasis or necrosis in tracheal organ cultures treated with M. pneumoniae membranes. 22

Other potential mediators of toxicity in mycoplasma infections are capsular galactan, a phenol-water extract of M. mycoides subsp. mycoides, and a lipoglycan extracted from several acholeplasma species. Evidence that these products may play a pathological role similar to the endotoxin of gram negative bacteria has been suggested by several authors, although biochemically they are different from classical lipopolysaccharides (Gourlay and Howard, 1982; Rosendal, 1993; Nicholas and

Bashiruddin, 1995).

Numerous enzymatic activities have been described and suggested to play a role in mycoplasma pathogenicity. Proteases specific for IgA have been found in U. urealyticum

(Robertson et al., 1984; Kapatais-Zoumbos et al., 1985). Liquefaction of gelatin by cultures of M. mycoides subsp. mycoides and subsp. capri was described by Freundt

(1958). Liquefaction of coagulated serum by M. capricolum and M. mycoides subsp. mycoides has also been described (Gabridge et al., 1985). Increased collagenase activity of certain M. arthritidis strains have been reported to have some role in the development of arthritis in mice (Kluve et al., 1981). Phospholipase activities associated with membranes of M. hominis, M. mycoides and A. laidlawii have been observed to be responsible for lysis of red blood cells from various sources (Tryon and Baseman, 1992).

Mycoplasma Capsules

Mycoplasmas, unlike bacteria, lack a cell wall and cell-wall precursors. However, similar to many bacteria, capsules have been described in several mycoplasma species. The common techniques used for detection of bacterial capsules, based on light

microscopy combined with India ink to enhance contrast, have not been useful with mycoplasmas because of their small size and complexity of the media or animal tissue environment in which capsule is being expressed (Rosenbusch and Minion, 1992).

Demonstration of mycoplasma capsules has been achieved mainly by using positively charged electron dense compounds such as ruthenium red or polycationic ferritin which react with substances that are negatively charged including acidic sugars, phospholipids and fatty acids (Luft, 1971). Another technique that has been widely used for demonstration of mycoplasma capsules is lectin reactivity. Since lectins are known to exhibit carbohydrate specificity, this technique can also provide some basic information on the sugar composition of the capsule (Bayer, 1990; Minion and Rosenbusch, 1993).

The presence of a capsule has been reported on several mycoplasma species that are proven respiratory pathogens including M. pneumonia, M. mycoides subsp. mycoides,

M. dispar, M. hyopneumoniae, M. gallisepticum, M. pulmonis. Among the ureaplasmas,

U. urealyticum has been reported to express a capsule. Capsules have not been reported among the acholeplasmas, anaeroplasmas, or asteroleplasmas, but lipoglycans have been found in the members of these species (Razin, 1978; Rosenbusch and Minion, 1992;

Minion and Rosenbusch, 1993). In contrast to capsules, lipoglycans are considered integral to the mycoplasma membrane. Moreover, lipoglycans do not provide significant staining with rutheniimi red and have lower molecular weight than capsules (Rosenbusch and

Minion, 1992). Little is known concerning the effects of enviroiunental factors on capsule production and the nature of the interaction occurring between eucaryotic cells and the mycoplasmas. Some mycoplasma species seem to synthesize their capsule constimtively, while others sjoithesize the capsule in a considerable amount only during in vivo growth or when co-cultured in vitro with eucaryotic cells. M. dispar produces a higher percentage of encapsulated cells and thicker capsule when grown in vivo or co-cultured with fetal bovine limg cells than when grown in vitro (Almeida and Rosenbusch, 1991), implying that capsule synthesis was induced in this mycoplasma. Similarly, synthesis of capsular material in M. penetrans has been shown to be a regulated process (NeyroUes et al., 1996).

Increases in the capsular thickness have also been observed to occur with co-cultured M. hyopneumoniae (Tajima and Yagihashi, 1982), M. pulmonis (Taylor-Robinson et al., 1981) and M. gallisepticum (Tajima et al., 1982; Yagihashi et al., 1987).

Several techniques have been used for capsule extraction and purification. Hot phenol extraction procedures followed by ion-exchange chromatography have been used extensively (Rosenbusch and Minion, 1992). However, capsular material that has been isolated by the use of hot phenol extraction procedures may contain lipoglycans; these are integral membrane components and therefore can be a source of contamination. Capsular material can also be extracted by prolonged exposure of the whole mycoplasma cells to buffered saline at 37°C (Almeida et al., 1992). The capsular material of M. hyopneumoniae was also extracted by boiling the organisms (Young et al., 1996). These observations may relate to the supposition that capsule may be loosely attached to the mycoplasma 25 membrane. There are also reports of releasing carbohydrates from mycoplasma cells by mild sonication (Minion and Rosenbusch, 1993). The ability of lectins to exhibit carbohydrate specificity has also been explored as a component of capsular marprial purification protocols. Alemeida and Rosenbusch (1991) purified the capsule of M. dispar based on its specific reactivity to Ricinus communis lectin. The capsule of U. urelyticum which binds to concanavalin A lectin (Whitescarver et al., 1975; Robertson and Smook,

1976) has also been purified in this way for further smdy.

Information regarding the chemical composition of mycoplasma capsules is lacking and is mostly derived from indirect evidence provided by electron microscopic observations of preparations stained with positively charged electron dense compoimds such as ruthenium red or polycationic ferritin, and also the ability of specific lectins to react with different mycoplasmas. The results obtained from such studies suggest that mycoplasma capsules are composed at least in part of acidic sugars (Rosenbusch and Minion, 1992).

Few mycoplasma capsules have been chemically analyzed. The capsular galactan of

M. mycoides subsp. mycoides is reported to be composed of polymers of galactose which are linked in a beta (1-6) configuration (Rosenbusch and Minion, 1992). The composition of a polysaccharide isolated from mycoplasma CF-38 strain), die etiological agent of contagious caprine pleuropneumonia in goats, was determined to be of equal quantities of glucose, galactose, mannose, fiicose, glucosamine and galactosamine (Rurangirwa et al.,

1987). The chemical composition of the capsular polysaccharide of M. dispar was a polymer of galactose and galacturonic acid (Bansal, 1995). More recently, the sugar 26 composition of the capsular material of M. penetrans was determined by gas liquid chromatography and was found to be mannose:glucose:N-acetylglucosamine:N- acetylgalactosamine in the proportion of 1:6:1:2 (NeyroUes et al., 1996).

Pathogenic roles of mycoplasma capsules. Capsules have long been associated with virulence properties of bacteria including mycoplasmas. Since they are composed of

99% water, these highly hydrated polysaccharide capsules may protect the organisms from desiccation thereby enhancing their survival and facilitate their spread from one host to another (Dudman, 1977; Whitfield, 1988). This observation is quite interesting at least for the mycoplasma respiratory pathogens which are in most cases encapsulated and spread from one host to another by means of aerosols. A capsule may also significantly affect access of molecules to the bacterial cell envelope and cytoplasmic membrane in a way that they are not accessible to antibacterial agents such as antibiotics (Costerton et al., 1987).

Capsules may mediate adherence of organisms including mycoplasmas to host tissues. Direct evidence that mycoplasma capsules may promote adherence of the organisms to host cells is lacking and derived mostly from morphological observations

(Green and Hanson, 1973; Wilson and Collier, 1976; Robertson and Smock, 1976; Taylor-

Robinson, 1981; Tajima and Yagihashi, 1982; Rosenbusch, 1992). Recently, however,

Yoimg et al. (1996), using a microtiter adherence assay, reported that Af. hyopneumoniae purified capsular polysaccharide possessed an inhibition effect on the adhesion of the organism to swine cilia. This demonstrated, perhaps for the first time, evidence of direct 27 involvement of polysaccharide capsules in the interaction of mycoplasmas and their host cells.

Participation of capsules in microbial adhesion is relatively controversial because of their composition. Capsules are negatively charged, and are best known for their repulsive properties, therefore contributing to antiphagocytic activity. Possible explanations as to how negatively charged bacterial capsules can mediate adherence include the formation of a latch effect provided by multiple binding sites as a result of repeating structural entities of polysaccharides. In addition, enzymes or proteins, supposedly bound to or buried in the membranes may extend across the extracellular capsule and neutralize the negative charges of the capsule or provide important binding sites in negative to negative-charged surface attachment (reviewed by Almeida, 1988). Based on the presence of lectins in mammalian host cells (Roberson et al., 1981; McArthur and Ceri, 1983), there has been speculation that capsules may promote adhesion through specific interaction between host cell lectins and capsular sugar moieties (Marcus and Baker, 1985). Perhaps the best indication that bacterial capsules can act as adhesion molecules is lectinophagocytosis, a nonspecific phagocytosis of bacteria in the absence of serum opsonins mediated by lectins on the phagocytic cell membrane that bind specifically to carbohydrate on the bacterial surface.

For instance, Klebsiella pneumoniae undergoes phagocytosis mediated by capsular exopolysaccharides which are recognized by a specific lectin of macrophages (Athamna et al., 1991). This also would not be inconsistent with the ability of many mycoplasma 28

species to adhere to the surface of phagocytic cells, although xjptake occurs only in the

presence of specific antibodies (Howard and Taylor, 1985).

A common theme in bacteriology regarding the role of polysaccharide capsules is

their antiphagocytic effect associated with their electronegative charges and with their capacity to nonspecifically interfere with complement activation. It has been shown that capsular polysaccharides inactivate the amplification loop of the complement system either by binding to factor H or decreasing the affinity of factor B for C3b, resulting in the blockage of the alternative pathway (Kasper, 1986). This strategy of limiting complement activation has been described for capsules containing sialic acid such as group B streptococci (Edwards et al., 1982) and K1-encapsulated E. coli (Stevens et al., 1978), and those that do not contain sialic acid such as type 7 and type 12 pneumococci (Joiner et al.,

1984). Other mechanisms by which the capsule has been reported to protect against phagocytosis include formation of microcolonies immersed in a glycocalix of joined individual exopolysaccharides which creates a protective niche against host enzymatic degradation (Isenberg, 1988), and masking binding receptors on the bacteria for opsonic factors such as specific antibodies and complement components (Absolom, 1988).

Mycoplasmas, in general, are known to resist phagocytosis. For instance, many in vitro studies have shown that in the absence of specific antibodies several mycoplasma species can colonize the surface of phagocytic cells without stimulating phagocytosis (Howard and

Taylor, 1985). This phenomenon has been described for both encapsulated and unencapsulated organisms of M. dispar (Almeida et al., 1992), and for the unencapsulated 29

M. bovis (Howard, 1980). Thus, it appears that with mycoplasmas, production of a capsule per se is not necessarily a prerequisite for restricting phagocytosis, at least for the organisms themselves. Yet, mycoplasma capsules have been reported to interfere with the phagocytosis and the killing of a second bacterial target by phagocytic cells. For example, both encapsulated M. dispar and its purified capsule reduce bovine alveolar macrophage metabolism and suppress phagocytosis of a second bacterial target (Almeida et al., 1992).

Mycoplasma capsules may have toxic effects. The capsular galactan of M. mycoides subsp. mycoides has detrimental effects on host cells; these effects are expressed as lymphatic and capillary thrombosis, pulmonary edema and infiltration of lymphatic spaces of the affected lobules (Nicholas and Bashiruddin, 1995). Such changes were reproduced in calves following intravenous injection of purified galactan into cattle (Buttery et al.,

1976). Capsular galactan induced lethal effects for chicken embryos when inoculated in the chorioallantoic cavity (Gourlay and Howard, 1982). Another toxic effect associated with capsular galactan of M. mycoides subsp. mycoides is the deposition of fibrin in chronic lung lesions of cattle, thus promoting lung lesions and persistent mycoplasmemia (Gourlay and

Howard, 1982).

In general, bacterial polysaccharide capsules are poorly immunogenic, are T cell independent antigens, fail to induce development of memory cells and immune responses may not develop in the very young (Stein, 1985; Maxon and Kroll, 1990). Also, capsules may express immunosuppressive activity by preventing an appropriate immune response.

All of these features have also been described for immune responses against some of the 30

mycoplasma capsules (Hudson et al., 1967; Howard and Gourlay, 1983; Bansal et al.,

1995). Paradoxically, however, convalescent animals may have titers of anticapsular

antibodies in their serums (Buttery, 1970; Rurangirwa et al., 1987). There are several

reasons why polysaccharide capsules are poor immimogens. The majority of bacterial

polysaccharide capsules lack complexity and are made up of repeating single (linear

homopolymers) or a few (linear heteropolymers) monosaccharide types. This property may

also explain their T cell independency. The repeating antigenic determinants should

contribute to an efficient cross-linking to the surface immunoglobulin receptors on B cells

and provide activation signals to B cells (reviewed by Bansal, 1994). Polysaccharide

capsules are also generally high molecular weight and are slowly metabolized since the

animal may lack the appropriate catabolic enzymes. This could contribute to the ability to

induce immunological tolerance. In many situations, there is structural mimicry between

polysaccharide capsules and host cell constiments and it is believed that this relationship is

responsible for establishing an initial inmiunological tolerance. A well known example of

this mimicry is the capsular galactan of M. mycoides subsp. mycoides which has been shown to possess a serological similarity to pneumogalactan, a product of normal lung epithelial cells (Bashiruddin and Nicholas, 1995). Structural mimicry has also been described in other bacterial capsules including K1 and K5 polysaccharides of E. coli

(Finne, 1982; Navia et al., 1983). 31

Interaction of Mycoplasmas with the Host Immune System

The interaction of mycoplasmas with the imTnnnp. system has been thoroughly reviewed elsewhere (Messier, 1986; Almeida, 1991; Niang, 1992); therefore, only pertinent aspects wiU be cited.

One of the main features of mycoplasma species is their ability to survive and even multiply in cultures of phagocytic cells in the absence of specific antibodies. This phenomenon has been observed with several respiratory mycoplasmas including

M. pneumoniae, M. pulmonis, M. dispar and Af. bovis (Howard and Taylor, 1985), and

M. ovipneumoniae (Al-BCaisi and Alley, 1983). Mycoplasmas are also recognized for their ability to suppress the ability of phagocytic cells to ingest and kill a second bacterial target.

Examples include M. hominis, M. anhritidis (Simberkoff et al., 1971; Thomsen and

Heron, 1979), M. bovis (Howard and Taylor, 1983), M. dispar (Almeida et al., 1992), and

M. hyopneumoniae (Carusso and Ross, 1990). Our previous in vitro studies have also demonstrated decreased uptake of a second bacterial agent by sheep alveolar macrophages when these macrophages were pretreated with M. ovipneumoniae (Niang, 1992). The mechanism by which mycoplasmas inhibit the phagocytic capacity of phagocytes is still unclear. However, a role for polysaccharide capsules has been demonstrated (Almeida et al., 1991). Also, it has been speculated that the ability of mycoplasmas to inhibit phagocytosis is due to their ability to bind firmly to the cell membranes of these phagocytic cells. Thus, this action may mask or prevent stimulation of appropriate receptors (Howard and Taylor, 1985). Another in^)ortant characteristic of mycoplasmas is their ability to interact

nonspecifically with lymphocytes as first demonstrated by Ginsbm-g and Nicolet (1973)

with M. pulmonis. Following this report, many other mycoplasma species were reported to

possess mitogenic activity for either T or B cells or both (Cole et al., 1985; Messier et al.,

1990). This mitogenic activity has been demonstrated with whole organisms, culture supematants, and membrane preparations but the nature of mycoplasmal mitogens has not

been well characterized (Cole et al., 1985). In the case of M. arthritidis, a factor characterized as a superantigen has been implicated in the T cell-mitogenic activity of the organism (Cole, 1991). Mitogenic activity of mycoplasmas and their products has also been observed with in vivo experimentation (Cole et al., 1985; Messier et al., 1990) and has led to speculation that this activity may contribute to disease pathogenesis .

While mycoplasma species have been reported to activate lymphocytes, suppressive effects on these cells have also been reported. Both viable and killed nonfermenting mycoplasmas have been reported to suppress lymphocyte blastogenic responses to phytohaemagglutinin (PHA) lectin. Barile and Leventhal (1968) suggested that the suppressive factor in these arginine requiring mycoplasmas was depletion of arginine required for the nutrition of cells. More recently, arginine deiminase was characterized as the factor responsible for inhibition of lymphocyte blastogenesis by these mycoplasmas

(Sugimura et al., 1989). In addition, other experiments with glucose fermenting mycoplasma species including M. mycoides subsp. mycoides, M. hyopneumoniae, M. bovis and M. gallisepticum (Thomas et al., 1990), and M. dispar (Finch and Howard, 1990) and 33

M. hyopneumoniae (Kishima and Ross, 1985) also demonstrated suppression of lymphocyte blastogenic responses to lectins. These suppressive effects have also been demonstrated by in vitro culture and lectin stimulation of lymphocytes from infected animals.

Suppressive effects on both himioral and cell-mediated responses by mycoplasma species have been reported. Kaklamanis and Pavlatos (1972) reported that simultaneous injections of Af. arthritidis and viral antigens resulted in complete or partial suppression of humoral responses against the antigens. Similarly, Af. dispar has also been shown to induce immimosuppression when injected in combination with other antigens (Howard and

Gourlay, 1983). Impaired development of delayed type hypersensitivity has been reported with M. hyopneumoniae (Adegboye, 1978), Af. pneumoniae (Biberfeld and Sterner, 1976) and Af. pulmonis (Lai et al., 1989). Suppression of both humoral and cell-mediated responses were reported with Af. bovis in calves (Bennett and Jaspar, 1977) and

M. pneumoniae in humans (Sabato et al., 1981). Evidence of mycoplasma induced- immunosuppression can also be implied by their ability to potentiate bacterial infections.

For example, the potentiating effect of M. pneumoniae on the development of pneumococcal septicemia in hamsters by Diplococcus pneumoniae was demonstrated by

Lui et al. (1972).

Concerning M. ovipneumoniae, studies conducted in this laboratory showed that responses of sheep peripheral blood lymphocytes to various antigen preparations of the organism were inconsistent. These inconsistent responses in blastogenic assays were noted between animals and even in the same animal at different times. Also, lymphocytes from 34

normal lambs gave moderate responses in these assays (unpublished data). This suggests

that Af. ovipnewnoniae or its product could elicit regulatory influences on immune

responses of sheep.

Role of the Host Immune Responses in the Pathogenesis of Mycoplasma Diseases

The activation of cells of the immune system is not always beneficial for the

host, and the concept that the pathologic lesions seen in mycoplasma diseases are host-

mediated immune responses is plausible. Indeed, a feature common to many mycoplasmas- induced respiratory and arthritic diseases is the marked lymphoid-cell infiltration at the site of primary involvement. Many of the respiratory mycoplasmas mainly colonize the ciliated epithelial cells lining the respiratory tract and induce characteristic lesions, partly comprised of a lymphocyte infiltration in a form of marked peribronchial and perivascular accumulations (Gourlay and Howard, 1982; Howard and Taylor, 1985; Howard et al.,

1987). Many of the lymphocytes observed in these lesions are lymphocytes producing immunoglobulins that were, in some cases, specific for the invading mycoplasmas (Gourlay and Howard, 1982; Howard and Taylor, 1985). Additional support for this view was provided by the studies of Denny et al. (1972) in mice infected with M. pulmonis and

Taylor et al. (1974) in hamsters infected withM. pneumoniae who showed that treatments which suppressed the immune response, such as thymectomy, irradiation, antilymphocyte serum treatment, and cyclophosphamide injections, reduced the extent of the lesions formed in the lungs of these animals. Furthermore, athymic mice were found to develop less lung lesions and less specific immunoglobulins than normal mice when both were infected with M. pulmonis (Cassell and McGhee, 1975). However, experimental studies in gnotobiotic calves with M. dispar failed to demonstrate the characteristic lymphoc)rtic accumulations observed with other respiratory mycoplasmas. A possible explanation was that this species is poorly immunogenic and has an innnunosuppressive effect in young calves (Howard et al., 1987).

Evidence regarding the mechanisms involved in the formation of these lymphoid lesions varies. A local specific humoral immune response to persistent mycoplasmal antigen stimulation on the surface mucosa appears to play a major role in the development of these lesions. Cassell et al. (1974) examined the nature of the lymphocyte infiltration in mouse lungs after infection with M. pulmonis and observed that one week after infection the majority of the lymphocytes were IgM producing cells and a few IgA producing cells were also present. By 3 weeks post-infection, the accumulated lymphocj^es were predominantly IgA producing cells but large numbers of IgGl and IgG2 producing cells were also present. Similar observations, but with a different antibody isotype predominating, were made on lungs of M. hyopneumoniae-mSs.ctt6. pigs (Messier, 1990),

Af. pneumoniae-witc\&i. hamsters (Femald et al., 1972), and M. dispar- and M. bovis- infected calves (Howard et al., 1987). Evidence for a direct role of host specific immunity in the pathogenesis of respiratory mycoplasmas remains an open question, since evidence of cross-reactivity of these specific mycoplasma antibodies with host cells is lacking. 36

However, indirect evidence, such as progressive and persistent lymphoid hyperplasia causing occlusion of airways and alveolar collapse, is overwhelming.

The mitogenic activity for lymphocytes exhibited by numerous respiratory mycoplasmas is an alternative explanation for lymphoid responses seen in mycoplasma respiratory lesions. This nonspecific stimulation of lymphocytes by the mycoplasmas also may contribute to disease pathogenesis by induction of antibodies with various specificities.

Support for the idea that mycoplasma mitogenicity plays a role in disease pathogenesis comes from evidence that mitogenicity of certain mycoplasmas is not only an in vitro phenomenon but is also expressed in vivo. Inoculation of mycoplasmal membranes into the respiratory tract of experimental animals led to significant lymphocyte infiltration and lung lesions that were indistinguishable from that produced by viable mycoplasmas (Cole et al.,

1985; Messier et al., 1990). Furthermore, the release of a variety of soluble products including interferons and interieukins by mycoplasma mitogen-activated lymphocytes could induce damage to host cells by recruiting and stimulating differentiation or activation of other cells (Cole et al., 1985).

Other evidence suggests that some type of activation of the host immune system plays a role in pathogenicity of mycoplasmas, eventually leading to an autoimmune response and thereby to immunopathological damage. An autoimmune response to various tissue antigens associated with mycoplasmal infection has been described in humans

(Biberfeld, 1985), in catde (Cassell et al., 1985, Nicholas and Bashiruddin, 1995), in pigs

(Suter et al., 1985, Baumgarter and Nicolet, 1984; Roberts and Little, 1970) and in rats 37

(Runge et al., 1990; Kirtchoff et al., 1989). In addition to nonspecific lymphocyte activation, reviewed above, molecular mimicry was proposed to explain the occurrence of

these immunopathological responses during microbial infections. The capsular galactan of

M. mycoides subsp. mycoides is antigenically related to pneumogalactan, a product of normal lung epithelial cells (Nicholas and Bashiruddin, 1995). This suggests that an immune response to the mycoplasma galactan could induce autoreactive antibodies with consequential formation of immune complexes in the lymph nodes leading to lymphatic and capillary thrombosis, edema and infiltration of lymphatic spaces of the affected lobes

(Nicholas and Bashiruddin, 1995). Antigenic mimicry has also been associated with the occurrence of autoantibodies in himians infected with M. pneumoniae (Biberfeld, 1985) and

M. genitalium (Bisset, 1994). Possible induction of an autoimmune response as a consequence of aberrant expression of MHC molecules on recognized antigen presenting ceils, or on cells that normally do not present antigen is another proposed mechanism by which microorganisms can elicit autoantibodies. In this regard, the in vitro upregulation of

MHC molecules on antigen presenting cells by several mycoplasma species has recently been demonstrated and proposed to be responsible of the generation of autoimmime pathology (Stuart et al., 1989; Stuart et al., 1990). The relevance of this is indicated by several autoimmune conditions in which abnormal expression of HLA class n molecules by specific cells of the target organs has been demonstrated (Bottazzo et., 1986; Smith and

Allen, 1992). A possible role for mycoplasmas in the pathogenesis of asthma and allergic airway

hypersensitivity is beginning to emerge. Serological studies on asthmatic patients infected

with M. pneumoniae indicated a high level of serum IgE specific for the organism,

suggesting involvement of IgE-mediated hypersensitivity (Tipimeni et al., 1980; Yano et al., 1994). Yano et al. (1994) speculated on the possibility that the production of this M.

pneumoniae-s^ciSic IgE may arise from epitopes that are cross-reactive with components of the airway. Nagayama et al. (1987) also demonstrated a correlation with infection by observing decreased levels of plasma IgE in 78.4% of children with M. pneumoniae

infection between the acute and convalescent phases. Mycoplasmal infection was also a predisposing factor for IgE production to other allergens (Yano et al., 1994). This effect could be due to destruction of respiratory mucosal cells facilitating allergen penetration of the mucosa and also to influences on regulatory mechanisms leading to IgE synthesis

(Frick, 1986). 39

CHAPTER 3. DEMONSTRATION OF A CAPSULE ON MYCOPLASMA OVIPNEUMONIAE

A paper to be submitted to the Am J Vet Res

Mamadou Niang,' MS; Ricardo F. Rosenbusch,^ DVM, PhD; John J. Andrews,^ DVM, Ph.D; Merlin L. Kaeberle/ DVM, PhD.

Objective

To examine Mycoplasma ovipneumoniae for the presence of a capsule and the potential role of the capsule as a virulence factor.

Samples

Seventeen isolates of M. ovipneumoniae and two isolates of M. arginini, isolated from respiratory tracts of sheep with respiratory disease, were used. The smdy also included a reference strain and a high passage level isolate of M. ovipneumoniae.

Procedure

Mycoplasmas were cultured in modified Friis broth mediirai, ovine fetal lung cells or ovine tracheal ring explants. Pelleted mycoplasmas or mycoplasma-infected ring cultures

^ Department of Microbiology, Immunology and Preventive Medicine, Iowa State University, Ames, lA 50011.

^ Veterinary Research Instimte, College of Veterinary Medicine, Iowa State University, Ames, lA 50011.

^ Veterinary Diagnostic Laboratory, College of Veterinary Medicine, Iowa State University, Ames, LA 50011. 40 were treated with rutheniuin red or polycationic ferritin and visualized by transmission electron microscopy. Also, the reactivity of several lectins with the mycoplasmas was studied by a microtiter agglutination test.

Results

Electron microscopic analysis revealed a large number of M. ovipneumoniae cells covered with an electron dense-stained amorphous material. This dense-staining material was absent when unlabeled mycoplasma preparations were examined, suggesting that it was a capsule. Marked variations in the thickness of the capsule were observed for a single M. ovipneumoniae isolate grown under the different culture conditions. Multiple passages of the microorganisms in modified Friis broth medium decreased the thickness of the capsule but not the percentage of cells encapsulated. When the different Af. ovipneumoniae isolates grown in modified Friis broth medium or co-cultured with ovine fetal lung cells were compared for capsular thickness or for percentage of encapsulation, marked differences were observed among them. However, no apparent differences in the percentage of encapsulation were observed for any of these isolates when growth in modified Friis broth mediimi was compared with co-culturing in ovine fetal lung cells, but there were differences in the capsular thickness. In thin sections of ruthenium red-stained tracheal ring cultures, the mycoplasmas appeared to be in close contact with cilia through their capsule.

All isolates of M. ovipneumoniae reacted strongly with wheat germ agglutinin lectin.

Variable reactivity of the M. ovipneumoniae isolates with the other lectins was observed. 41

Conclusions

M. ovipneumoniae expresses a polysaccharide capsule with variable size expression in different cultures. Strain heterogeneity could be further indicated by variable reactivity of this capsular material with several different lectins. Morphological observations suggest that this capsule facilitates adherence of the organism to ciliated epithelium.

Clinical Relevance

Strain variation associated with this capsule and its possible association with attachment to cilia may provide a versatile interaction of M. ovipneumoniae with its environment; thus, a long term colonization of the host by the organism.

Introduction

M. ovipneumoniae is the most commonly isolated mycoplasma from sheep with respiratory disease. However, the potential pathogenic role of this agent still remains unresolved. Experimental transmission of the organism has yielded inconsistent evidence of pathogenicity. There is lack of evidence relative to its ability to produce certain factors that may be suggestive of pathogenic potential such as a toxin, adhesion molecules, or a capsule. A capsule may be an important factor that enables pathogenic microbes including mycoplasmas to colonize the host respiratory tract. The presence of a capsule has been reported on several mycoplasma species that are proven respiratory tract pathogens 42 including M. pneumoniae, M. mycoides subsp. mycoides, M. dispar, M. hyopneumoniae,

M. gallisepticum, and M. pulmonis.

With many other mycoplasma species causing respiratory diseases (e.g.,

M. pneumoniae* M. dispar,^ M. hyopneumoniae,^'^ M. gallisepticum,^ andM. pulmonis^, a positive correlation between increased capsule production and the ability to colonize the host's respiratory tract has been suggested. More recently, the capsule of

M. dispar has been reported to induce a poor immune response in calves and mice.'° This association of mycoplasma capsules with failure of an immune response that might inhibit attachment and colonization of host respiratory cells suggests that the capsule may be an important virulence factor. Based on these observations, it was of interest to examine

M. ovipneunwniae for the presence of a capsule since the agent has also been observed to attach to cilia, to colonize the upper respiratory tract of sheep,' and to induce a poor immune response in naturally infected lambs.''

Materials and Methods

Mycoplasma isolates. Nineteen isolates of M. ovipneumoniae including a reference strain, a high passage level isolate, and two isolates of M. arginini were used in this study.

All the mycoplasma isolates, except the reference strain which was provided by the second author, were isolated in this laboratory from lung tissues or nasal swabs of sheep with respiratory disease and identified as either M. ovipneumoniae or M. arginini.^ All the mycoplasma isolates, except the high passage level one, were utilized below the 11th 43 passage level. The high passage isolate was originally derived from a low passage isolate by successive in vitro growth in modified Friis broth medium (MFBM). For investigation of in vitro passage effects on capsule production, both the high passage and its parent low passage isolate were included in various experiments. A stock culture of each of the mycoplasma isolates was prepared by growing the organisms in MFBM. Cultures were harvested and washed once in phosphate buffered saline solution (PBS, pH 7.4) by centrifligation at 10,000 g for 30 minutes. Washed cells were resuspended and concentrated in PBS and aliquots were stored at -70PC.

Growth conditions. Mycoplasmas were cultured for various experiments in

MFBM, ovine tracheal ring explants or ovine fetal lung (OFL) cells. For growth in

MFBM, organisms from a stock aliquot were inoculated into MFBM, cultures were incubated at 37°C in 0.5% CO2 in air. On attaining optimal growth, organisms were harvested by centrifugation at 10,000 g for 30 minutes at 4®C.

Ovine tracheal ring cultures were prepared from the trachea of Iambs obtained by caesarean section. The trachea was removed, stripped of extraneous tissues and sliced into rings which were placed in 12-well tissue culture platesP containing 0.4 ml of Medium 199

(M199),'' supplemented with 5% fetal bovine serum'' and 200 /ig/ml of bacitracin''. Tracheal rings were infected with M. ovipneumoniae (isolate R19) at 105 CFU/ml and plates were incubated at 37°C in 0.5% COj. Spent medium was removed every other day and replaced with fresh M199 medium until the experiment was terminated. 44

Ovine fetal lung (OFL) cells,® used below flie 15th passage level, were cultured in

Dulbecco's modified Eagle's medium'' (DMEM) supplemented with 10% fetal bovine serum in 25 cra^ tissue culture flasks."^ Cell monolayers were infected with selected isolates of M. ovipneumoniae at approximately 25 mycoplasmas per cell and flasks were incubated at 3>TC in 0.5% COj for 30 hours. Thereafter, the flasks were rinsed with cold PBS and supematants were saved. The monolayer cells were scraped from the flasks and added to the corresponding supematants. The suspensions were centrifiiged into pellets.

Fixation and labelling. Seven M. ovipneumoniae isolates cultured in MFBM were pelleted by centrifiigation. The pellets were fixed with equal parts of 3 % glutaraldehyde in

0.1 M sodium cacodylate buffer (pH 7.2) with and without 1% rutheniimi reef for 2 hours on ice and washed 3 times (15 miDutes each) with 0.1 M sodium cacodylate buffer. Pelleted cells co-cultured in OFL were also fixed with and without ruthenium red.

Only isolate R19 grown in MFBM was labelled with polycationic ferritin.

Organisms were fixed with 3% glutaraldehyde in 0.1 M sodium cacodylate buffer and labelled with polycationic ferritin^ as previously described." Fixed mycoplasmas were resuspended in 0.1 M of sodium cacodylate buffer with polycationic ferritin (1 mg/ml) for

30 minutes at 4°C and the reaction was stopped by ten-fold dilution with cold PBS.

Mycoplasmas were resuspended in 4% agar and washed 3 times in cold PBS by centrifiigation. 45

Freshly collected tracheal rings infected with isolate R19 at two and six days post­

infection were fixed with and without ruthenium red.

Transmission electron microscopy. Samples were postfixed with 1 % osmiimi

tetroxide in 0.1 M sodium cacodylate buffer for 2 hours at room temperature, washed with

distilled water, dehydrated in a graded series of 50, 75, 95 and 100% acetone and then embedded at 60°C in Epon 812 resin. Thin sections cut at 70-90 mn using a diamond knife

were stained with 2% uranyl acetate in 50% methanol followed by Reynold's lead citrate, and then examined with an electron microscope (Hitachi H500) at 75 kv.

For estimation of the percentage of encapsulation, a total of 100 mycoplasma cells were coimted randomly under the transmission electron microscope for each of the seven mycoplasma isolates grown either in MFBM or co-cultured with OFL cells."

Capsular thickness of the mycoplasma isolates, grown under the different culture conditions, was estimated by using a photomicrograph scale marker® on enlarged photomicrograph prints following manufacturer's instructions.

Lectin agglutination. The reactivity of several lectins with the nineteen

M. ovipneumoniae isolates and the two M. arginini isolates grown in MFBM was smdied by a microtiter agglutination test as previously described" with some modifications.

Briefly, serial two-fold dilutions of the lectins'* (starting concentration of 1 mg/ml) in 25 /zl saline were made in 96-well microtiter plates with U-shaped bottoms (Inomulon 1 46 microplates,).' Twenty five fil of a constant amount of each washed mycoplasma suspension in saline was added to each well. After mixing, plates were incubated overnight or up to 24 hours at room temperature. Agglutination was detected visually and the agglutination titer of each lectin was recorded as the last dilution causing agglutination. For determination of lectin specificity, inhibition of agglutination with various carbohydrates' was assayed by mixing 25 fil of a two fold dilution series of 1 % stock sugar solutions with an equal volimie of each lectin solution (containing at least 2 agglutinating units). After 15 minutes mcubation at 37°C, an equal volume of the mycoplasma suspension was added and the reaction was incubated and read.

Statistical analysis. Estimates of the percentage of encapsulation and those of the capsular thickness were analyzed using analysis of variance and differences were considered significant when p values < 0.05 were obtained.

Results

Electron microscopic evaluation of MFBM-grown M. ovipneumoniae isolate R19 fixed with glutaraldehyde showed a thin fuzzy layer of material on the surface of all mycoplasma cells (Figure 1). In thin sections of the same mycoplasma preparation fixed in the presence of ruthenium red or labelled with polycationic ferritin, a large niraiber of mycoplasma cells could be observed covered with a dense-staining capsular material

(Figures 2, 3, 4). Variations in the thickness of the capsule were observed for organisms 47 grown under different culture conditions. Isolate R19 organisms cultured in ovine tracheal ring explants had a much thicker capsule when compared to organisms grown either in

MFMB or co-cultured with OFL cells. At the same time, mycoplasma organisms co- cultured with OFL cells had a thicker capsule than those grown in MFBM (Table 1,

Figures 2, 3 and 5). Also, when the different mycoplasma isolates grown in MFBM or co- cultured with OFL cells were compared, marked differences in capsular thickness were observed among them (Table 2). There was no apparent differences in the percentage of encapsulation for any isolate when grown either in MFBM or co-cultured with OFL cells.

However, significant differences in the percentage of encapsulation were observed between different isolates when grown ui MFBM or co-cultured with OFL cells (Table 3).

Following multiple in vitro passages of M. ovipneumomae R19 in MFBM, a marked decrease in capsular thickness but not in the percentage of encapsulation was also noted

(Figure 6 and Tables 2, 3).

In thin sections of rutheniimi red-stained ovine tracheal ring ciiltures and OFL cells, the mycoplasmas appeared to be in close contact with the cilia or the membrane of the lung epithelial cells through their capsule (Figure 7). This was in contrast with thin sections fixed without ruthenium red where the organisms appeared to be separated from the cells

(Figure 8).

Variable reactivity of the M. ovipneumoniae isolates with different lectins was observed (Table 4). However, all the isolates reacted strongly with wheat germ agglutinin lectin which has specificity for N-acetyl-D-glucosamine residues. Most of these isolates 48

reacted also to a lesser degree with concanavalin-A which binds specifically to D-glucose or D-mannose. A few isolates showed low reactivity with phytohemagglutinin erythroagglutinin, Pissium sativum agglutinin, Ricinus communis-l agglutinin. Lens culimris agglutinin or soybean agglutinin lectins. There was no reactivity with Dolichos biflorus agglutinin, peanut agglutinin or phytohemagglutinin leukoagglutinin lectins.

Neither of the M. arginim isolates was agglutinated by the lectins. Inhibition of

M ovipneumoniae agglutination was observed with N-acetyl-D-glucosamine, D-glucose and

D-mannose. Decreased agglutinability of M. ovipneumoniae was observed following multiple in vitro passages of the organism in MFBM; however, the lectin specificity was not affected.

Discussion

To the best of our knowledge, this is the first report of a capsule associated with M. ovipneumoniae and the first study that describes strain variation of a mycoplasma species based on capsule characteristics. Among the mycoplasma species, several well known human and animal respiratory tract pathogens have been reported to produce a capsule.

These include M. pneumoniae, M. hyopneumoniae, M. dispar, M. gallisepticum,

M. pulmonis, M. mycoides subps. mycoides and M. pulmonis}'^ As in these previous studies, we used positively charged electron dense components such as rutheniimi red and polycationic ferritin, which react to acidic polysaccharides on the surface of the organism, to demonstrate presence of a capsule on M. ovipneumoniae. The methods gave similar 49

resiilts and additional evidence for the expression of a capsule on M. ovipneumoniae was obtained by lectin reactivity.

Multiple in vitro passages of the organism in MFBM decreased significantly the thickness of the capsule. In addition, when various isolates were grown in the same culture conditions, the resultant capsule was not always equal in thickness, indicating that the binding of ruthenium red to the cell surface is unlikely to be mediated by nonspecific adsorption of culture medium components on the surface of the organism. These observations are in agreement with similar studies conducted on several other mycoplasma species.'"

Growing M. ovipneumoniae under different culture conditions affected the thickness of the capsule but not the percentage of encapsulation. For instance, co-culturing the organism with OFL cells increased the thickness of the capsule from 13 nm (when grown in MFBM) to 26 nm. Likewise, growth in tracheal organ cultures increased the size of the capsule to 45 nm. This increase in the capsular thickaess is unlikely due to deposition of host cellular components on the surface of the organisms since in the same mycoplasma preparations many unencapsulated mycoplasma cells were observed. This variable size expression of the capsule imder different culture conditions could be explained, at least in part, by the fact that tracheal ring explants mimic the natural environment and the in vivo growth of the mycoplasmas more than OFL cells or MFBM. Likewise, OFL cells mimic the in vivo conditions more than MFBM. Other workers have reported effects of culture conditions on capsule production. Variations in the thickness of the capsule of M. hyopneumoniae^ and M. gallisepticim^'^^ have been reported. From their studies,

Tajima and Yagihashi® reported decreases in the thickness and poor staining of the capsule of an in vitro-passaged strain of M. hyopneumoniae. Similar results have also been reported for M. pulmonis.^ Cytoadsorptive virulent strains of M. pulmonis have more capsular material than strains that have been passaged in vitro until they are no longer virulent.

These observations have also been described for certain capsule producing bacteria other than mycoplasmas.^^ While these observations seem consistent with our findings, we also, however, observed that the percentage of encapsulation was not affected by multiple in vitro passage and culture conditions. Ahneida and Rosenbusch" reported that the percentage of encapsulated M. dispar co-cultured with bovine lung fibroblasts was significantly higher than that of MFBM-grown mycoplasmas. These results differ from our findings but could be explained by species variation.

Lectins have been used for detection of carbohydrate residues on the membrane surfaces of many microorganisms including mycoplasma species.^ This specific interaction of lectins and carbohydrates can also provide some basic information on the sugar composition of the capsular exopolysaccharides. In the present smdy, the strong agglutination of the M. ovipneumoniae organisms with wheat germ agglutinin lectin indicates that carbohydrates containing N-acetyl-D-glucosamine and/or D-glucose and/or

D-mannose constitute the main sugar residues of M. ovipneumoniae capsule. Despite the fact that all the strains that were tested reacted strongly with wheat germ agglutinin, marked variable reactivity with other lectins was evident, possibly indicating strain variation among the different M. ovipneumoniae isolates. Another possible explanation could be the failure of agglutination despite binding of the lectin to cell surfaces. Binding of lectins to cell surfaces without agglutination is a well known phenomenon.^' However, if that was the case here, the agglutination titers would be expected to be about equally low or high for all the lectins when reacted with a single mycoplasma isolate. Consistent with this idea is the fact that both the high passage isolate and its parent low passage isolate had the same lectin specificity. Strain variation among M. ovipneumoniae isolates has been reported previously on the basis of antigenic and DNA analyses Strain variation has been predominantly recognized as a mechanism for evasion of the host inmiune response.

This would not be inconsistent with our earlier findings that lambs namrally infected with

M. ovipneumoniae had low antibody titers against capsule of the organism (Chapter 5 in this dissertation).

The present experiments suggest that the capsule may have a role in adherence of the organism to host cells. In ruthenium red-stained thin sections, the organisms were observed in close association with either the cilia or the lung cells through their capsule, but when non-stained sections were examined, a gap seemed to separate the organisms from the host cells. Direct experimental proof of the role of capsule in the adherence of mycoplasmas to host cells has yet to be provided. However, as in the present study, several reports of morphological observations have suggested participation of capsule in the attachment process of mycoplasma species to host cells.^ "'"® ''' '^ More recently, Young^ et al. demonstrated that purified M. hyopneumoniae capsular polysaccharides possess an 52 inhibitory effect on the binding of the organism to swine cilia. In addition, the capsules of certain bacteria other than mycoplasmas have also been proposed to mediate their adhesion to host cells.

Participation of capsules in microbial adhesion is relatively controversial because they are commonly composed of carbohydrates. Being negatively charged, capsules are best known for their repulsive properties, and therefore antiphagocytic activity. Adherence

is probably a multifactorial event involving both specific and nonspecific mechanisms.

Specific microbial adhesion is most commonly described as interaction between microbial surface proteins (adhesins) and complementary sugar moieties (receptors) on host cells.

However, in some instances, proteins on the host cells can act as specific receptors for other microbial surface components including capsules. The involvement of

lipopolysaccharide in the adhesion of Actinobacillus pleuropneumoniae to porcine tracheal

rings has been reported.'^ Human plasma fibronectin has also been shown to bind

Streptococcus pyogenes via lipoteichoic acid.^° Exopolysaccharide capsules of Pasteurella

Pseudomonas aeruginoscr^-^ and Streptococcus sui^ have been described as

imponant components in the adhesion of these bacteria to host cells. There have been

reports of the presence of a lectin-like material in the lungs of rats.^''-" Based on that,

Marcus and Baker^' speculated about specific Interactions between host cell lectins and

P. aeruginosa capsular sugar residues.

Perhaps the best example that bacterial capsules can act as adhesion molecules is the

lectinophagocytosis process in which there is phagocytosis of bacteria in the absence of 53 specific serum opsonins. Adherence is mediated by lectins on the phagocytic cell membrane that bind specifically to carbohydrate on the bacterial surface. Klebsiella pneumoniae undergoes phagocytosis mediated by capsular expolysaccharides which are recognized by a specific lectin of macrophages.^^ This would not be inconsistent with the ability of many mycoplasma species to adhere to the surface of phagocytic cells, although uptake occurs only in the presence of specific antibodies.

Possible explanations as to how a negatively charged mycoplasma capsule can mediate adhesion include the formation of latch effect provided by the multiple binding sites as a result of repeating structural components of polysaccharides. In addition, enzymes or proteins supposedly bound to or buried in the membranes and arising across the extracellular capsule may neutralize the negative charges of the capsule or provide important binding sites in negative to negative-charged surface attachment.' It could also be speculated, on the basis of the afformentioned lectin reports, that another possible explanation would be through specific interaction between capsular sugar moieties and host cell proteins (lectins).

Another suggested role for mycoplasma capsules in pathogenesis is their modulating activity on phagocytic cells. Encapsulated M. dispar or its purified capsule was shown to depress bovine alveolar macrophage metabolism and suppress phagocytosis of a second bacterial target."' Our previous in vitro smdies have demonstrated that M. ovipneumoniae organisms have marked effects on characteristics and functions of ovine alveolar macrophages."' Whether or not the capsule of the organism has something to do with these 54 effects was not investigated at that time but an analogy with M. dispar capsule could be implied.

Results of the present smdy clearly indicate that M. ovipneumoniae produces a capsule in which N-acetyl-D glucosamine and probably D-glucose and D-mannose are the predominant sugar residues. Although the exact functions of this capsule have yet to be clearly defined, the morphological observations suggest its possible association with attachment of the organism to the host cells. In addition, variable reactivity of this capsule with different lectins may be suggestive of strain variation associated with its expression.

This may provide a versatile interaction of the organism with its environment as well as a means to evade host immune responses. The consequences of all these effects would be long term colonization of the host by the organism. It thus appears that expression of capsule by M. ovipneumoniae may be related to its pathogenicity.

Acknowledgments

The authors thank Jean Olsen and Dr. Paul G. Eness for technical support; and Jose

Lopez-Virella, Lie-Ling Wu and Dr. Kevin Flaming for help in statistical analysis.

This work was supported in part by grants from the USDA and the Iowa Livestock Health

Advisory Council. 55

®Khan MA. Respiratory infection of lambs with M. ovipneumoniae. PhD thesis. Iowa State

University, Ames, lA, 1993.

•"Niang, et al: uiq)ublished data, 1996.

""Coming Glass Works, Coming.NY.

''Gibco Laboratories, Inc., Grand Island, NY.

^Courtesy of Dr. H.Lehmkul, National Animal Disease Center, Ames, lA.

^Sigma Chemical Co., St. Louis, MO.

®Ted Fella, Inc., Tustin, CA.

•TCN Biochemicals, Inc., Aurora, OH.

'Dynatech Laboratories Inc., Alexandria, VA.

^Rosenbusch RF. Mucosal colonization by capsulated mycoplasmas (abstr). Proceedings.

9th Int Congr. Int Org Mycoplasmology. Ames, lA, 1992;60.

'Toung TF, Erickson BZ, Rosenbusch RF, Ross RF. Capsular polysaccharide of

Mycoplasma hyopneumoniae (abstr). Proceedings. 11th Int Cong. Int Org

Mycoplasmology. Orlando, Florida, 1996;391. 56

'Almeida RA. Studies on Mycoplasma dispar capsular material. MS thesis. Iowa State

University, Ames, LA, 1988.

"^iang M. Interaction of Mycoplasma ovipnewnoniae with sheep alveolar macrophages.

MS thesis. Iowa State University, Ames, lA, 1992.

References

1. Minion CF, Rosenbusch RF. Extramembranoiis structure in Mycoplasmas. In; Rottem S, Kahane I, eds. Subcellular Biochemistry. Vol 20. New York: Plenum Press, 1993; 189-201.

2. Razin S. The Mycoplasmas. Microbiol Rev 1978;42;414-470.

3. Rosenbusch RF, Minion CF. Ceil envelope: morphology and biochemistry. In: Maniloff J, McEIhaney RN, Finch LR, and Baseman J, eds. Mycoplasmas, molecular biology and pathogenesis. ASM, Washington, DC, 1992;73-77.

4. Wilson MH, Collier AM. Ultrastructural smdy of Mycoplasma pneumoniae in organ culture. J Bacteriol 1976;125:332-339.

5. Allan EM, Pine HM. Electron microscopy observations on mycoplasmas in pneumonic calf lungs. J Med Microbiol 1977;10:469-472.

6. Tajima M, Yagihashi T. Interaction of Mycoplasma hyopneumoniae with the porcine respiratory epithelium as observed by electron microscopy. Infect Immun 1982;37:1162-1169.

7. Tajima M, Yagihashi T, Nunoya T. Ultrastructure of mycoplasmal capsules as revealed by stabilization with antiserum and staining with ruthenium red. Jpn J Vet Sci 1985;47:217-223.

8. Tajima M, Yagihashi T, Yuriko M. Capsular material of Mycoplasma gallisepticum and its possible relevance to the pathogenic process. Infect Immun 1982;36:830-833. 57

9. Taylor-Robinson D, Furr PM, Davies HA et al. Mycoplasmal adherence with particular reference to the pathogenicity of Mycoplasma pulmonis. Isr J Med Sci 1981;17:599-603.

10. Bansal P, Adegboye DS, Rosenbusch FR. Immune responses to the capsular polysaccharide of Mycoplasma dispar in calves and mice. Comp Immun Microbiol fafectDis. 1995;18:259-268.

11. Almeida RA, Rosenbusch RF. Capsulelike surface material of Mycoplasma dispar induced by in vitro growth in culture with bovine cells is antigenically related to similar structures expressed in vitro. Infect Immun. 1991;59:3119-3125.

12. Yagihashi T, Nunoya T, Tajima M. Isr J Med Sci 1987;23:517.

13. Kasper DL, Onderdonk AB, Reinap BG et al. Variations of Bacterioides fragilis with in vitro passage: presence of an outer membrane-associated glycan and loss of capsular antigen. J Infect Dis 1980;142:750-756.

14. Green HI F, Hanson M. Ultrastructure and capsule of Mycoplasma meleagridis. J Bacteriol 1973;116:1011-1018.

15. Robertson J, Smook E. Cytochemical evidence of extramembranous carbohydrates on Ureaplasma urealyticum (T-Strain Mycoplasma). J Bacteriol 1976;128:658-660.

16. Baselga R, Albizu I, Amorena B. 1994. Staphylococcus aureus capsule and slime as virulence factors in ruminant mastitis. A review. Vet Microbiol 39:195-204.

17. Tnzana TJ. Capsules and virulence in the HAP group of bacteria. Can Vet Res 1990;54:S22-S27.

18. Pijoan C, Trigo F. Bacterial adhesion to mucosal surfaces with special reference to Pasteurella multocida isolates from atrophic rhinitis. Can J Vet Res 1990;54:S16- S21.

19. Belanger M, Dubreuil D, Harel J, et al. Role of lipopolysaccharides in adherence of Actinobacillus pleuropneumoniae to porcine tracheal rings. Infect Immun 1990;58:3523-3530.

20. Stanislawski L, Simpson WA, Hasty D, et al. Role of fibronectin in attachment of Streptococcus pyogenes and Escherichia coli to human cell lines and isolated oral epithelial cells. Infect Immtm 1985;48:257-259. 58

21. Marcus H, Baker NR. Quantitadon of adherence of mucoid and nomnucoid Pseudomonas aeruginosa to hamster tracheal epithelium. Infect lomiun 1985;47:723- 729.

22. Ramphal R, Pier GB. Role of Pseudomonas aeruginosa mucoid exopolysaccharide in adherence to tracheal cells. Infect Immun 1984;47:1-4.

23. Gottschalk M, Petitbois S, Higgins, R et al. Adherence of Streptococcus suis capsular type 2 to porcine lung sections. Can J Vet Res 1991;55:302-304.

24. McArthur HA, Ceri H. Interaction of a rat lug lectin with exopolysaccharides of Pseudomonas aeruginosa. Infect Immun 1983;42:574-578.

25. Roberson MM, Ceri H, Shadle PJ, et al. Heparin inhibitable lectins: marked similarities in chicken and rat. J Supramolec Struct Cell Biochem 1981;15:395-402.

26. Athamna A, Ofeck I, Keisrari Y et al. Lectinophagocytosis of encapsulated Klebsiella pneumoniae mediated by surface lectins of guinea pig alveolar macrophages and human monocyte-derived macrophages. Infect Immun. 1991;50:1673-1682.

27. Almeida RA, Wannemuehler MJ Rosenbusch RF. Interaction of Mycoplasma dispar with bovine alveolar macrophages. Infect Immun, 1992;60:2914-2919.

28. Kahane I, Tully GJ. Binding of plant lectins to mycoplasma cells and membranes. J Bacteriol 1976;128:1-7.

29. lonas G, Norman NG, Clarke JK, et al. A study of the heterogeneity of isolates of Mycoplasma ovipneumoniae from sheep in New Zealand. Vet Microbiol 1991;29:339-347.

30. Mew AJ, lonas G, Clarke JK, et al. Comparison of Mycoplasma ovipneumoniae isolates using bacterial restriction endonuclease DNA analysis and SDS-PAGE. Vet Microbiol 1985;10:541-548. Figure 1. Electron micrograph of Mycoplasma ovipneumoniae grown in modified Friis broth medium, unstained. A thin layer of amorphous material is seen on the surface of all mycoplasma cells. Bar = 0.5 fim.

Figure 2. Electron micrograph of Mycoplasma ovipneumoniae grown in modified Friis broth medium as for Fig. 1, stained with rutheniimi red. A large number of mycoplasma cells are covered with a dense ruthenium red-stained capsular material (C). Bar = 0.5 ^m.

Figure 3. Electron micrograph of Mycoplasma ovipneumoniae co-cultured with fetal ovine lung cells, stained with ruthenium red. A dense ruthenium red-stained capsular material (C)

is evident outside the limiting cj^oplasmic membrane (CM) of the mycoplasma. Bar = 0.1

fxni.

Figure 4. Electron micrograph of Mycoplasma ovipneumoniae grown in modified Friis broth medium, labelled with polycationic ferritin. A large number of mycoplasma cells are covered with a dense-stained capsular material. Bar = 0.5 (im.

67

Table 1. Estimates of capsular thickness of Mycoplasma ovipneumoniae isolate (R19) when grown in modified Friis broth medium (MraM), on ovine tracheal ring explants (OTR) or co-cultured in fetal ovine lung cells (FOLC).

Growth Conditions Mean-Thickness^+.SEM MFBM^ 17±2.185 FOLC^ 30+2.333 OTR^ 43jil.0{)0 ^Statistically different (p<0.05) from each other. Vverage of 3 capsular thickness measurements in nanometers from 3 separate experiments. Figure 5. Electron micrograph of Mycoplasma ovipneumoniae organisms cultured in ovine tracheal ring explants, stained with ruthenium red. Mycoplasma cells are covered with a very dense ruthenium red-stained capsular material (C). Bar = 0.1 /zm.

70

Table 2. Estimates of capsular thickness of several Mycoplasma ovipnewnomae isolates when grown in modified Friis broth medium (MFBM) or co-cultured with fetal ovine lung cells (FOLC).

Mean-Thickness'^'^+SEM Isolates MFBM FOLC 1-Lomax^ 19+1.000^^* 24-h3.500^* 2-R19 19-t-0.500^^ 31-f-2.500^'^'^ 3-R19^ 04+1.000^^^^^ 10+1.500^^^"^ 4-MLA2 13-1-1.500^^^^ 18±2.000^® 5-P21 26+3.000^^''^^ 31 + 1.000^"^ 6-Hill 17+0.500^^ 29-1-0.500^'^ 7-S6 19Jil.000^^ 20+3.500^^^ ^Reference strain. ''Multiple passages in MFBM. Values represent average of 3 capsular thickness measurements in nanometers from 2 separate experiments. Significant differences (p < 0.05) between growth in MFBM and co-culturing with FOLC for Hill and R19 isolates. *Superscript numbers in each column represent differences (p<0.05) among isolates 71

Table 3. Percentage of encapsulated Mycoplasma ovipneumoniae organisms of several isolates when grown in modified Friis broth medium (MFBM) or co-cultured with fetal ovine lung cells (FOLC).

Percent Encapsulation Isolates MFBM FOLC 1-Lomax^ 25±1.000^'^^®''* 23 +3.000^'^^^''* 2-R19 36±0.500^^^ 40±3.500^^^ 3-RI9'' 28±1.500^^'' 29ii3.500^'^^ 4-MLA2 39 + 1.000^^^ 37±0.000^^^ 5-P21 54±3.500^^^'^ 52^0.500^^^"^ 6-HILL 35±3.500^^ 36±.0.000^^ 7-S6 33±2.000^^ 36ii0.000^^ Values represent mean percentages C+.SEM) of 2 separate experiments. ^Reference strain. 'Multiple passages in MFBM. •Superscript numbers in each column represent differences (p<0.05) among isolates. Figure 6. Electron micrograph of high passage of Mycoplasma ovipneumoniae grown in modified Friis broth medium, stained with ruthenium red. A marked decreased in capsular

(C) thickness is noted when compared with the parent low passage as shown in Fig. 2. Bar

= 0.5 nm.

Figure 7. Electron micrograph of a fetal ovine tracheal ring explant infected with

Mycoplasma ovipneumoniae for 2 days, unstained. Mycoplasama organisms (M) are not in close contact with the cilia (CI) and a gap between both membranes are clearly discemable

(arrow). Bar = 0.5 /tm.

Figure 8. Electron micrograph of a fetal ovine tracheal ring explant infected with

Mycoplasma ovipneumoniae as for in Fig. 7, stained with ruthenimn red. The capsular material appears to bridge the gap between the mycoplasmas (M) and the cilia (CI).

Bar = 0.5 ixm.

78

Table 4. Microtiter agglutination of Mycoplasma ovipneumoniae and M. arginini isolates with weat germ agglutinin (WGA), concanavalin-A (CONA), phytohemagglutinin erythoagglutinin (PHAE), Pissium sativam agglutinin (PSA), Ricinus communis-l agglutinin (RCAI), Lens cuUnaris agglutinin (LCA), soybean agglutinin (SBA), Dolichos b^orus agglutinin (DBA), peanut agglutinin (PNA), or phytohemagglutinin leukoagglutinin (PHAL) lectins.

Titers of Agglutination with Different Lectins Isolates WGA" CONA' PHAE PSA RCAI LCA SBA DBA PNA PHAL LOMAX* 128 128 32 8 0 0 0 0 0 ND' R19 >256 64 32 8 4 8 0 0 0 0 RIQ" 128 32 16 4 4 4 0 0 0 ND 27 >256 >256 64 16 64 16 0 0 0 ND P21 >256 >256 64 32 128 32 0 0 0 ND HILL-N >256 >256 >256 16 8 8 0 0 0 ND ENESS >256 64 64 8 128 128 2 0 0 ND 53 >256 64 32 16 32 2 0 0 0 ND WL >256 >256 32 8 2 32 0 0 0 ND HILL-L >256 64 16 8 128 64 128 0 0 0 64 >256 128 128 8 0 0 0 0 0 0 219 128 128 128 8 16 4 0 0 0 0 764 128 128 16 4 0 0 0 0 0 0 S6 128 128 4 4 8 0 0 0 0 0 50 >256 0 0 0 0 0 0 0 0 0 47 128 >256 32 4 8 8 0 0 0 ND 288 128 128 32 4 0 0 0 0 0 ND 2033 128 >256 32 16 32 16 0 0 0 ND 126 >256 64 64 8 64 2 0 0 0 ND 221' 0 0 0 0 0 0 0 0 0 0 WNS' 2 0 0 0 0 0 0 0 0 0 'Reference strain. ""Multiple passages in modified Friis broth medium. "M. arginini isolates. Agglutination was inhibited by N-acetyl-D-glucosamine (1%). 'Agglutination was inhibited by D-glucose (1%), D-mannose (1 %). •Not done. 79

CHAPTER 4. CYTOPATmC EFFECTS OF MYCOPLASMA OVIPNEUMONIAE

FIELD ISOLATES IN OVINE TRACHEAL ORGAN CULTURES

A paper to be submitted to the Am J Vet Res

Mamadou Niang,' MS; Mary C. Debey, DVM, Ph.E)^; Yosiya Niyo,^ DVM, PhD; John J. Andrews,^ DVM, Ph.D; Merlin L. Kaeberle,^ DVM, PhD.

Objective

To investigate the cytopathic potential of different isolates of Mycoplasma ovipneumoniae on ovine tracheal ring explants and factors that contribute to in vitro cytopathic effects.

'Department of Microbiology, Immunology and Preventive Medicine, College of Veterinary Medicine, Iowa State University, Ames, lA 50011.

^USDA, National Animal Disease Center, Ames, lA 50010.

^Department of Veterinary Pathology, College of Veterinary Medicine, Iowa State University, Ames, LA 50011.

^Veterinary Diagnostic Laboratory, College of Veterinary Medicine, Iowa State University, Ames, lA 50011. 80

Samples

Four isolates of M. ovipneumoniae including a high passaged isolate, and one

isolate of Mycoplasma arginini were used in this study.

Procedure

Ovine tracheal rings were prepared from tracheas of unborn lambs obtained from

sheep by caesarean section. Tracheal rings were randomly placed in a 12-well tissue culture

plate (one ring/well) and wells were inoculated with each of the mycoplasmas in M199

supplemented with 5% fetal calf serum and bacitracin. Tracheal rings were examined every

over day under an inverted microscope for evaluation of ciliary activity. Infected and

noninfected tracheal rings were randomly selected at various incubation times for light and electron microscopic examinations. Also, samples of spent mediimi from infected and

noninfected rings were collected at various incubation times for determination of extracellular calmodulin content and for presence of hydrogen peroxide.

Results

Infected tracheal rings showed significantly decreased ciliary activity as compared

to the noninfected control rings. There were, however, marked differences between isolates

in the onset and severity of the effects which correlated with their ability to produce

hydrogen peroxide. Infected tracheal rings released more calmodulin than the noninfected controls. The amount of calmodulin released also varied between isolates and somewhat

reflected the degree of loss of ciliary activity in the corresponding rings induced by the 81 different isolates. Light and electron microscopic examinations of infected tracheal rings revealed disorganization and sloughing of the epithelium, and association of mycoplasmas only with the cilia. Following repeated in vitro passages, the organisms had reduced ability to inhibit ciliary activity which correlated with decreased hydrogen peroxide production.

Addition of catalase to the organ cultures delayed loss of ciliary activity.

Conclusions

M. ovipneumoniae induced ciliostasis in ovine tracheal ring explants which correlated with hydrogen peroxide production. M. ovipneumoniae is an extracellular pathogen that colonizes the ciliated epithelium.

Clinical relevance

M. ovipneumoniae colonization of respiratory epithelial cells contributes to the pathogenesis of sheep respiratory disease.

Introduction

M. ovipneumoniae has been routinely recovered from the respiratory tract of sheep with respiratory disease worldwide.' Recently, Khan® consistently isolated both M. ovipneumoniae and M. arginini from Iambs exhibiting a clinical coughing syndrome. This respiratory disease is a major problem in the midwestem states of the United States of

America due to widespread occurrence in young lambs with resultant predisposition to 82 rectal prolapses, to secondary bacterial pneumonias, and reduction in the rate of gain.

M. ovipneumoruae can be consistently demonstrated in association with the cilia of the upper respirator>' tract of the affected lambs.

The pathogenicity of M. ovipneimoniae has been extensively studied by experimental infection of sheep; these smdies have yielded conflicting results.^ It is generally assimied that this may be due to variations in pathogenicity of different strains, attenuation during isolation and cloning on artificial medium, and nature of the experimental animals.

Evaluation of the effects of mycoplasma species on tracheal organ cultures is another approach to study their pathogenicity since interference with the mucociliary clearance mechanism is an important consideration in the induction of respiratory disease.®"

" In these instances, subjective measurement of ciliary activity, light and electron microscopic observations, and measurement of metabolic activities (e.g., calmodulin levels, dehydrogenase enzymes, ATP content, and oxygen utilization*-^) have been used to assess epithelial cell damage.

The possibility of hydrogen peroxide (HjOj) or hemolysin produced by several mycoplasma species as cytotoxic factors responsible for cellular injiuies is controversial because of conflicting reports in the literature. However, there are several lines of evidence that suggest correlation between H2O2 production and ciliostasis induced by mycoplasma species.'^ " Also, recently published data indicated that both high and low concentrations of 83 exogenous H2O2 alone disrupt airway epithelial cells and inhibit ciliary activity in sheep"* as well as in guinea pig** and rat'^ tracheal organ cultures.

Previous work has demonstrated ciliostasis and loss of cilia caused by

M. ovipneumoniae and M. arginini in ovine tracheal explants,'® but possible factors and variability in pathogenicity related to these cytopathic changes were not reported. The purpose of the present smdy was to investigate die cytopathic potential of different isolates of Af. ovipneumoniae on ovine tracheal explants for determination of variability in pathogenicity among the different isolates of the organism. Factors that could potentially contribute to the in vitro cytopathic effects caused by the organisms were also evaluated.

Materials and Methods

Mycoplasma isolates. M. ovipneumoniae isolates R19, Hill and MLA2, and M. arginini isolate WNS Gow passage) were isolated from pneumonic sheep lungs in this laboratory. M. ovipneumoniae R94 (a high passage strain) was derived from the R19 isolate by 94 in vitro passages in modified Friis broth medium (MFBM).

The mycoplasmas were grown in MFBM. Cultures were harvested and washed once in phosphate buffered saline (PBS, pH 7.4) by centrifiigation at 10,000 g for 30 minutes at

4°C. Washed cells were resuspended in PBS, the number of organisms/ml was determined by the plate count method and aliquots were stored at -70PC until used. The viability of the stock preparation was determined at the time of use in each experiment. 84

Organ cultures. The method of preparation of the organ explants has been fully

described elsewhere.® Unborn lambs were obtained from sheep by caesarean section. The

trachea was removed aseptically and placed in sterile M199 medium (M199).'' Comiective

tissues were trimmed from the outer surface of the trachea and the organ washed with

M199. The trachea was cut into separate rings and each ring was randomly placed in a well

in a 12-well tissue culture plates'* containing 0.4 ml of M199s (M199 supplemented with

5% fetal bovine serum® and 200 /xg/ml of bacitracin). After incubation for 12 hours at 37C

in 0.5% COj in air, wells were either inoculated with 50 /nl of 1 x l(f colony forming units

(CFU) per ml of each of the mycoplasma in M199s or M199s alone. Spent medium was

removed every other day and replaced with fresh M199s until the experiment was

terminated. The spent medium from each well was cultured for mycoplasmas and other

bacteria, and was either used immediately for specific assays or frozen at -70'C until further usage.

Evaluation of ciliary activity. Tracheal rings were examined after 12 hours in cultures (just prior to inoculation) and every other day after inoculation by using an

inverted microscope (lOOX or 800X magnification). Treatment ring identity was obscured

to avoid bias. Quantitation of ciliary activity was determined as previously described.'

Briefly, the proportion of the ring surface bearing motile cilia (0 to 100%) was multiplied

by the vigor of ciliary movement (0 to 3+) and these results were expressed as relative ciliary activity. 85

Light and electron microscopy. Light and electron microscopic examinations were conducted on tracheal rings infected with the M. ovipneumoniae isolate R19 and uninfected control rings. Infected and noninfected tracheal rings were randomly selected at various incubation times. For light microscopy, rings were fixed in 10% buffered neutral formalin, processed by routine paraffin embedding, microsectioned and stained with hematoxylin and eosin. Formalin-fixed tissue sections were also examined for the presence of M. ovipneumoniae antigen using an indirect immunohistochemical staining procedure described elsewhere®. For electron microscopy, rings were fixed with or without ruthenium red® at 4°C in 3% glutaraldehyde in 0.1 M sodium cacodylate buffer (pH 7.2), processed as previously described,® and thin sections were examined by transmission electron microscopy (TEM) using a Hitachi H500 TEM at 75 kV.

Determination of extracellular calmodulin content. Samples of spent mediimi firom infected and noninfected rings (in triplicate) collected at various incubation times were homogenized in an equal volume of homogenizing buffer (0.25 M sucrose, 0.3 mM

EDTA, 1 mM 2-mercaptoethanol, pH 7.2).^ The homogenates were centrifiiged at 105,000

X g for 15 minutes and designated as tissue culture medium (TCM). Prior to calmodulin assay, prepared TCM samples were heated at lOCyC for 2 minutes and their protein content was determined with a Bio-Rad Protein Assa/ following the manufacturer's instructions.

The calmodulin content of test samples was assayed by measuring inorganic phosphate release from AMP in a calmodulin-activated enzyme cascade described by Sharma and Wang" except that, instead of reading the assay tubes in a spectrophotometer, a total volume of 200 fd from each tube was transferred in duplicate to individual wells of

96-well, microtiter flat bottom plates (Immulon)® and read in a microtiter plate reader*" at

650 nm. The estimation of calmodulin content was made by plotting the absorbance values of test samples against a standard curve of known amoimts of purified calmodulin.

Calmodulin concentrations in fig were calculated with reference to mg protein content.

Detection of hydrogen peroxide in tracheal ring cultures. The presence of 1^02 in tracheal ring cultures was determined with the benzidine test of Cherry and Taylor-

Robinson" with some modifications. Benzidine-blood-agar (BBA) plates were prepared aseptically by mixing mycoplasma agar base mediirai' with 4% washed sterile sheep red blood cells, 100 U/ml of bacitracin and 0.1% benzidine hydrochloride.'

Spent medium from infected and noninfected rings (in triplicate) collected at 2, 4, 6 and 8 days postinfection was centrifiiged at low speed for 4 minutes and 100 fil of supernatant fluid was placed in wells cut into the BBA plates. Plates were incubated 2 to 4 days aerobically at 37°C in a humidified container, if necessary overnight at 4°C, and examined for presence of black coloration around wells as an indicator of the presence of

HjOo.

To determine whether mycoplasma isolates used in this experiment produced HjO,, the organisms were actively grown in MFBM, centrifiiged at low speed for 4 minutes and

100 ^1 of the supernatant fluid was placed in wells cut into the BBA plates. A 30% 87 solution of H2O2' and fresh MFBM were included as positive and negative controls, respectively.

Effect of catalase on benzidine reaction. Actively growing mycoplasma cultures in MFBM, a 30 % solution of HjOj or fresh MFBM were placed in wells cut into either

BBA plates or BBA plates supplemented with 2,000 U/ml of catalas^. The plates were incubated and examined for presence of H2O2.

Effect of catalase on cytotoxicity induced by mycoplasmas. M. ovipneumoniae isolate, R19, which caused the most ciliostasis, was chosen for this experiment. Tracheal ring explants prepared as before were inoculated with 50 /xl of 1 x IC? of the mycoplasma preparation or M199s. To some inoculated rings (triplicate), catalase was added. Ciliary activity and HjOj production were determined.

Results

Tracheal ring cultures were examined 12 hours after establishment and only those rings showing strong ciliary activity were used. All mycoplasma isolates readily grew in the tracheal ring cultures and a large number of the organisms could be recovered from the spent medium throughout the experimentation (data not shown). Ciliostasis, expressed as percent ciliary activity, of infected and noninfected tracheal ring explants is reported in

Figure 1. Tracheal rings infected with the different mycoplasma isolates showed a significant decrease ^<0.05) in ciliary activity at all times postinfection as compared to control rings. However, isolates showed marked differences in the rapidity and severity of the effects. Tracheal rings infected with isolates R19 or Hill showed a drastic drop in ciliary activity within 2 days postinfection which persisted until day 8, when there was almost no detectable ciliary activity. In contrast, in tracheal rings infected with isolates

MLA2, WNS or P94, this decrease was moderate and there appeared to be a small recovery at day 6 postinfection, particularly in rings infected with WNS, which is a

Af. arginini isolate. Rings infected with isolates R19 or Hill had significantiy (p<0.05) less ciliary activity at all times postinfection than did rings infected with isolates MLA2,

WNS or the high passage isolate, exclusive of isolate WNS at day 2.

The potential for HjOj to be a cause of ciliary damage was the basis for testing for this material in the tracheal explant cultures. HjOj was detectable in ring cultures infected with isolates R19 or Hill but not in those infected with isolates MLA2, WNS or P94 and noninfected control rings. The ability of the different mycoplasma isolates to produce ItO, was also tested and only isolates R19 and Hill produced detectable quantities in the assay utilized (Figure 2a). Further association between H2O2 production by the mycoplasma isolates and their ability to induce loss of ciliary activity was investigated. Extensive in vitro passage caused a significant loss of HjO, production associated with a significant decrease in the ability of this isolate to induce ciliostasis (Figure 1). Addition of catalase to the BBA plate inhibited the benzidine reaction (Figure 2b). Also, addition of catalase to the organ cultures delayed significantly the loss of ciliary activity (Figure 3). 89

When infected tracheal ring sections with isolate R19 were observed by light

microscopy, cellular damage consisting of disorganization and sloughing of the ciliated

epithelium from the basement membrane was evident by day 2 postinfection. Often, a new

single layer of low cuboidal to flat epithelial cells with no cilia was observed replacing the

damaged epithelium at day 8 postinfection. However, this change was not apparent during

evaluation of ciliary activity. Conversely, control noninfected tracheal ring sections usually

appeared normal with intact and organized epithelium (data not shown).

Indirect immimohistochemical staining detected M. ovipneumoniae antigen on the

ciliated epithelium of infected tracheal ring sections. Positive staining was not observed in

any other areas of the sections or on noninfected tracheal ring sections. Examination of

tracheal ring sections using TEM (Figures 4, 5, 6) confirmed the cellular damage in the

form of loss of epithelium and ftirthermore revealed close association of mycoplasma organisms only with the remaining cilia. However, at later stages of the infection, a few organisms were observed near the microvilli (Figure 5). No mycoplasmas were found in association with the basement membrane, within epithelial cells, or within the lamina

propria. In addition, attachment of the organisms to the cilia was never observed as being mediated by a terminal structure. When stained with ruthenium red, a dense thick layer of capsular material was seen not only bridging the mycoplasmas and the cilia but also adjacent mycoplasmas in the form of microcolonies (Figure 4).

Metabolic changes as measured by intracellular calmodulin release are indicated in

Table 1. With few exceptions, the mean calmodulin concentrations detected in samples of 90 spent medium of the infected tracheal rings were significantly higher (p<0.05) than in the medium firom noninfected explants, especially at days 6 and 8 postinfection. Overall, these values varied significantly from one isolate to another and reflected the degree of ciliostasis induced by each isolate.

Discussion

The present study confirmed an earlier report that M. ovipneumoniae induces ciliostasis in ovine tracheal ring explants'® and extended our understanding of the in vitro colonization process of respiratory epithelium by the organism. We also investigated factors that might contribute to pathogenicity and attempted to compare these in vitro cytopathic effects on tracheal ring explants with variability in pathogenicity among the different isolates of the organism.

When tracheal rings infected with isolate R19 were examined for histopathological changes, cellular damage consisting mainly of disorganization and sloughing of ciliated epithelium firom the basement membrane was evident at day 4 postinfection. A new single layer of flat epithelial cells without cilia was present replacing the damaged epitheliimi at day 8 postinfection. Similar observations have been reported for M. hyopneumoniae- infected porcine tracheal organ cultures.® While not investigated here, these changes may indicate a healing process and suggest that the effects of the HjO, or other toxic factor produced by the organism may be reversible. This is supported by the observations of 91

Kobayashi et al." who reported that a low concentratioii of HjOj induced reversible damage on ovine tracheal organ cultures.

Immtmohistochemical staining demonstrated M. ompneumoniae antigens associated only with the ciliated epithelium of the infected tracheal rings. Positive staining was not observed in any other areas, suggesting that M. ovipneumoniae colonizes only the ciliated epithelium. This is in agreement with the observations made by Khan," but contrasts with those made by Haziroglu et al.'® who demonstrated the presence of M. ovipneumoniae antigens within the epithelial cells with an inmmnoperoxidase technique in pneumonic ovine lungs. Further evidence corroborating our findings were observations by TEM which showed close association of mycoplasma organisms only with the cilia. No mycoplasmas could be seen near or below the basement membrane, although at late stages of the infection, a few organisms were noted close to the microvilli. These observations suggest that M. ovipneumoniae, like most other mycoplasmas, is not invasive. In addition, when tracheal ring sections were stained with ruthenium red, a dense thick layer of capsular material was observed bridging the mycoplasma organisms and the cilia, as well as adjacent mycoplasmas in the form of microcolonies. Although the exact fimction of mycoplasma capsules is not ftiUy known, a role in adhesion of mycoplasma to host cells has been suggested.''-^'

All mycoplasma isolates induced ciliostasis; however, the various isolates showed marked differences in the rapidity and severity of the effects. These differences between the isolates was probably not a result of organism concentrations since all the isolates were recovered in about equal numbers from the corresponding tracheal ring cultures.

Differences in ciliostatic activity correlated with the presence of HjOj in the tracheal ring cultures suggesting that this material might contribute to pathogenic effects. Since no ftOz was detected in noninfected control tracheal ring cultures, it is likely that this ftO, was a product of the mycoplasmas. Consistent with this idea was the fact that only isolates R19 and Hill were found to produce HjOj when grown in MFBM. Extended passage of the isolate R19 resulted in a significant loss of HjO, production and ciliostatic activity.

Further evidence suggesting that H2O2 is a contributor to the ciliostatic effect was provided by experiments where catalase was added to the tracheal ring cultures. Catalase delayed significantly the loss of ciliary activity induced by the isolate R19. Catalase can inhibit effects of HjOj generated by the mycoplasmas since addition of catalase to the BBA plate inhibited the benzidine test reaction.

Hydrogen peroxide has been implicated as a possible factor in inducing host cell damage by several mycoplasma species.Additionally, some workers have reported that both low and high amounts of exogenous HjOj alone disrupted airway epithelial cells and mhibited ciliary activity in sheep" as well as in rat'^ and guinea pig" tracheal organ cultures. These studies reenforce the importance of HjOj as a mediator of host cellular injuries.

While our observations are consistent with previous reports, the results also suggest that H2O2 production may not be solely responsible for the cytopathic effects. Tracheal rings infected with isolates that did not produced detectable quantities of H,02 also had 93 significant decreases in ciliary activity as compared with the noninfected tracheal rings. It is possible, however, that these isolates were producing tmdetectable amounts of

Because of the intimate contact of the organisms with the host cell membrane, production of a small amoimt of HjOj by the organisms could result in host cell injuries. Catalase production by these isolates, although not investigated in the present study, would be another possible explanation.

There are pathogenic mycoplasmas, like M. hyosynoviae that do not produce

H202,^ and nonpathogenic mycoplasmas like, Acholeplasma laidlawii, that produce HjO,."®

Also, other researchers who compared in vivo effects of virulent and homologous attenuated strains of M. pneumoniae failed to show a consistent relationship between pathogenicity and peroxide production." However, HjOj produced by M. pneumoniae and other mycoplasmas can be rapidly destroyed by host catalase activity. Therefore, for the

H2O2 to exert its toxic effects, either host catalase must be depleted or mycoplasmas must adhere close enough to host cell surface to maintain a toxic, steady state concentration of

H2O2 sufficient to cause direct damage to the cell membrane.^® Similarly, Sobeslavsky et al.^' have suggested a correlation between cytoadsorption and peroxide production in association with the pathogenesis of M. pneumoniae. Kahane'" and Almagor et al.^ presented evidence that the effectiveness of H2O2 may depend on both the intimate contact of the organism with the host cell membrane and production of superoxide anions which inactivate host cell catalase, thus permitting the oxidative action of ItOj produced by both the organism and the host cells. Furthermore, the importance of attachment in enabling 94 mycoplasmas to deliver their toxic factors, including HjOj, into host cell membranes has been stressed by Stadtlander and Kirchhoff.^

In addition, other toxic substances produced by some mycoplasmas have been implicated in the cytopathic effects on organ cultures. However, once again, the relevance of such toxic factors to the cytopathic effects of some of these mycoplasmas for organ cultures has yet to be proven. For instance, a toxin has been described for M. bovis;^^ however, its relevance to the cytopathic effects of this organism for organ cultures could not be demonstrated.'® Recently, Debey and Ross^ related the cytopathic effects of M. hyopneumoniae on porcine ciliated cells to disturbance of normal cell membrane functions and excluded the involvement of a diffusible toxin because organisms in indirect contact with the tracheal rings through a 0.1 ptm membrane were not capable of inducing ciliostasis. Other studies have suggested that proteins associated with the mycoplasma cell membrane are the toxic factors responsible for these cytopathic effects. Both M. pneumoniae live organisms and membrane preparations induced ciliostasis in organ cultures of adult hamster trachea.^^ Similarly, whole cells and purified membranes of M. hyopneumoniae were reported to be cytotoxic for porcine lung fibroblasts.^^ In contrast to these reports, however, Hu et al.^ were unable to induce ciliostasis or necrosis in tracheal organ cultures treated with M. pneumoniae membranes.

We observed, on the basis of onset and severity of ciliary activity reduction, that there were marked differences between isolates. Since all the isolates were recovered in high number fi-om the infected tracheal rings throughout the experimentation, these 95

differences should not be attributed to the number of organisms. Although isolates were not

conq)ared for histopathological changes, there is reason to believe that strain variations in

virulence, which is related in this case to H2O2 production, accounted for these differences.

There are reports in the literature of such variations among different strains of

mycoplasma species. Stadtlander et al.^^ reported a high degree of heterogeneity in the

ciliostatic effect of different strains of M. fermentans on rat tracheal organ cultures.

Levishon et al.' reported similar observations for different strains of M. gallisepticwn on

chicken tracheal ring cultures. Using equine uterine mbes, Bermudez et al.^® noted observable differences of ciliostatic effects between different isolates of M. equigenitalium, although these differences were not statistically significant.

Levinshon et al.' also showed that multiple passages in artificial medium resulted in reduced pathogenicity of M. gallisepticum for chicken tracheal ring cultures. More recently, Debey and Ross^ reported that ciliostasis induced by M. hyopneumoniae in porcine tracheal ring cultures diminished with in vitro passage of the organism. This is consistent with our observations but differ from the results reported by Stadtlander et al.' who did not find any observable differences of cytopathic potential between a high passage

M. pulmonis strain and its parent low passage strain. Debey and Ros^ also related their results to loss of ability of the organism to colonize host cells. While our data did not rule out such a possibility, the experimentation revealed loss of HjOj production by the organism. There is one report in the literanire which indicates that a mutant strain of the T1 vaccine strain of M. mycoides subsp. mycoides (Small Colony) derived by repeated 96

passages, produced reduced amount of and lacked a H202-producing enzyme.

Interestingly, the strain had reduced pathogenicity.^' Nicholas and Bashiruddin?" suggested

that this reduced ability of H2O2 production may partly explain the loss of virulence of the

organism after either extensive subculture or egg passage. In addition, Lipman and Clyde

Jr^' previously reported variations in H2O2 production between different strains of

M. pneumoniae, although no relationship to virulence or degree of attenuation was found.

In the present study, all mycoplasma-infected tracheal rings released significantly

more calmodulin than the noninfected controls. However, significant differences were also observed between isolates which reflected the degree of ciliary loss induced by the isolates.

This somewhat agrees with the results reported by Bermudez et al.^ and contrast with those of Debey and Ross.®

Also in the present study, it was found that M. arginini induced loss of ciliary activity in tracheal organ cultures, but the onset was less rapid and the effect was less

pronounced than those of the M. ovipneumoniae isolates, in particular isolates R19 and

Hill. These results contrast with those reported by Jones et al.'® who showed that the onset of M. arginini was more rapid than that of M. ovipneumoniae. The reason for such discrepancy is not known but it may be related to the nature of the isolate used. However, our findings were consistent with the previous report in regard to an apparent recovery of ciliary activity after the 4"* day of infection of tracheal rings with Af. arginini.

It is apparent that the spectrum of cytopathic effects induced by mycoplasma species

in organ cultures is complex, involving diverse and interrelated mechanisms such as 97

attachment, liberation of various toxic factors and enzymes, rate and mode of delivery of

these toxic factors into host cells, and disturbance of normal host cell membrane functions.

However, published reports indicate that what is true for one species is not necessarily tme for an another closely related species. Our findings suggest diat ability of

M. ovipneumoniae to induce cUiostasis on ovine tracheal ring explants is related, at least in part, to HjOj production. However, factors other than HjO, may be involved. Finally, these smdies suggest that M. ovipneumoniae is an extracellular pathogen that colonizes only the cilia of the respiratory tract and there is marked variability in virulence.

Acknowledgments

The authors thank Dr. Paul G. Eness, Jean Olsen and David Cavanaugh for technical support; and Jose Lopez-Virella, Lie-Ling Wu and Dr. Kevin Flaming for help in statistical analysis. This work was supported in part by grants from the United States Deparment of

Agriculture and the Iowa Livestock Health Advisory Council.

"Khan MA. Respiratory infection of Iambs with Mycoplasma ovipneumoniae. PhD thesis.

Iowa State University, Ames, lA, 1993. 98

"TKhan Ar, Bengtson B, Laitinen LA. Influence of HjOz on the ciliary activity and the

contractile response of the guinea pig trachea (abstr). Am Rev Respir Dis

1990;141(SuppI.):A532.

"^Gibco Laboratories, Inc., Grand Island, NY.

''Coming Glass Works, Coming, NY.

®Sigma Chemical Corporation, St. Louis, MO.

"Bio-Rad Laboratories, Inc., CA.

®lDynatech Laboratories, Inc., Alexandria, VA.

"Molecular Devices Corporation, Menlo Park, CA.

'Becton Dickson, Microbiology Systems, Cocheysville, MD.

^Fisher Scientific, Fair Lawn, NJ. 99

"Toung TF, Erickson BZ, Rosenbmch RF, Ross RF. Capsular polysaccharide of

Mycoplasma hyopneumoniae (abstr). Proceedings. 11th Int Congr. Int Org

Mycoplasmology. Orlando, Florida. 1996;391.

References

1. Davies DH. Aetiology of pneumoniae of young sheep. Prog Vet Microbiol 1985;1:229-249.

2. Jones GE, Gilmour JS, Rae AG. The effects of different strains of Mycoplasma ovipneumoniae on specific pathogen-free and conventionally-reared lambs. J Comp Pathol 1982;95:477-487.

3. lonas G, Norman NG, Clarke JK, et al. A study of the heterogeneity of isolates of Mycoplasma ovipneumoniae from sheep in New Zealand. Vet Microbiol 1991;29:339- 3347.

4. Gabridge MG. Differentiated respiratory epithelium as an in vitro model of mycoplasma infections. In: Schiff LJ, ed. In vitro models of respiratory epithelium. CRC press. Inc., Boca Raton, Florida. 1986:85-102.

5. Bermudez VM, Miller RB, Rosendal S, et al. Measurement of the cytotoxic effects of different strains of Mycoplasma equigenitalium on the equine uterine tube using a calmodulin assay. Can J Vet Res 1992;56:331-338.

6. Debey MC, Ross RF. Ciliostasis and loss of cilia induced by Mycoplasma hyopneumoniae in porcine tracheal organ cultures. Infect Immun 1994;62:5312-5318.

7. Levinsohn S, Dykstra MJ, Lin MY, et al. Comparison of in vivo and in vitro methods for pathogenicity evaluation for Mycoplasma gallisepticum in respiratory infection. Avian Pathol 1986;15:233-246.

8. Gabridge MG. Respiratory tract organ cultures to assay attachment and pathogenicity of mycoplasmas. Ann Microbiol 1984;135A:33-38. 100

9. Stadtlander CTK-H, Watson HL, Simecka JW, et al. Cytopathic effects of Mycoplasma pulmonis in vivo and in vitro. Infect Immun 1991;59:4201-4211.

10. Thomas LH, Howard CJ, Parsons KR, et al. Growth of Mycoplasma bovis in organ cultures of bovine foetal trachea and comparison with Mycoplasma dispar. Vet Microbiol 1987;13:189-200.

11. Chery JD, Taylor-Robinson D. Growth and pathogenesis of Mycoplasma mycoides wax. capri in chicken embryo tracheal organ cultures. Infect Immun 1970;4:431-438.

12. Tryon W, Baseman JB. Pathogenic determinants and mechanisms. In: Maniloff J, McElhany RN, Finch LR, Baseman j, eds. Mycoplasmas: molecular biology and pathogenesis. ASM, Washington, DC. 1992:457-471.

13. Lipman, RP and Clyde, WA Jr. The interrelationship of virulence, cytadsorption, and peroxide formation in Mycoplasma pneumoniae. Proc Soc Exp Biol Med 1969;131:1163-1167.

13. Rosendal S. Mycoplasma. In: Carlton LG and Thoen CO, eds. Pathogenesis of Bacterial Infections in Animals, Second Ed, Iowa State University Press, Ames. 1993:297-311.

14. Kobayashi K, Salathe M, Pratt MM, et al. Mechanism of hydrogen peroxide-induced inhibition of sheep airway cilia. Am J Respir Cell Mol Biol 1992;6:667-673.

15. Burman WJ, Martin n WJ. Oxidant-mediated ciliary dysfimction: possible role in airway disease. Chest 1986; 89:410-413.

16. Jones, GE, Keir WA, Gilmour, JS. The pathogenicity of Mycoplasma ovipneumoniae and Mycoplasma arginini in ovine and caprine tracheal organ cultures. J Comp Pathol 1985;95:477^87.

17. Sharma RK, Wang JH. Preparation and assay of the Ca?'^-dependent modulator protein. Adv Cyclic Nucleotide Res 1979;10:187-198.

18. Haziroglu R, Diker KS, Turkarslan J, Gulbahar MY. Detection of Mycoplasma ovipneumoniae and Pasteurella haemolytica antigens by an immunoperoxidase technique in pneumonic ovine lungs. Vet Pathol 1996;33:74-76. 101

19. Minion FC and Rosenbusch RF. Extramembranous structure in mycoplasmas. In; Rotten S and Kahane I, eds. Subcellular biochemistry, Vol 20, Plenum press. New York. 1993:189-201.

20. Nicholas RAJ, Bashiruddui JB. Mycoplasma mycoides subspecies mycoides (Small Colony Variant): the agent of contagious bovine pleuropneumonia and member of the "Mycoplasma mycoides Cluster". J Comp Pathol 1995;113:1-27.

21. Almagor M, Yatziv S, Kahane I. Inhibition of host cell catalase by Mycoplasma pneumoniae: a possible mechanism for cell injury. Infect Immun 1983,41:251-256.

22. Almagor M, Kahane I, Yatziv S. Role of superoxide anion in host cell injury by Mycoplasma pneumoniae infection. J Clin Invest 1984;73:842-847.

23. Stadtlander C, Kirchhoff H. The effects of Mycoplasma mobile 163 K on the ciliary epithelium of tracheal organ cultures. Zbl Bakt Hyg A 1988;269:355-365.

24. Brennan PC, Feinstein RN. Relationship of hydrogen peroxide production by Mycoplasma pulmonis to virulence for catalase-deficient mice. J Bacteriol 1969;98:1036-1040.

25. Pijoan C. Secretion of hydrogen peroxide by some common pig mycoplasmas. Vet Rec 1979;86:216-217.

26. Razin S. The mycoplasmas. Microbiol Rev 1978;42:414-470.

28. Cohen G and Somerson NL. Glucose dependent secretion and destruction of hydrogen peroxide by Mycoplasma pneumoniae. J Bacteriol 1969;98:547-551.

29. Sobeslavsky O, Chanock RM. Peroxide formation by mycoplasmas which infect man Proc Soc Exptl Biol Med 1968;129:531-535.

30. Kahane, I. In vitro smdies on the mechanism of adherence and pathogenicity of mycoplasmas. Isr J Med Sci 1984;9:210-216.

31. Geary SJ, Tourtelotte ME, Cameron, JA. Inflammatory toxins from Mycoplasma bovis: isolation and characterization. Science 1981;212:1032-1033.

32. Gabridge MG, Johnson CK, Cameron AM. Cytotoxicity of Mycoplasma pneumoniae membranes. Infect Immun 1974;10:1127-1134. 102

33. Geary SJ, Walczack EM. Cytopathic effect of whole cells and purified membranes of Mycoplasma hyopneumoniae. Infect Immun 1983;41:132-136.

34. Hu PC, Collier AM, Baseman JB. Interaction of virulent Mycoplasma pneumoniae with hamster tracheal organ cultures. Infect Immun 1976;14:217-224.

35. Stadtlander CTK-H, Watson HL, Simecka JW. Cytopathogenicity of Mycoplasma fermentans (including strain Incognitus). Clin Infect Dis 1993;17(Suppl 1):S289-301.

36. Bermudez V, Miller R, Rosendal S, et al. In vitro cytopathic effect of Mycoplasma equigemtaliian on the equme uterine tobe. Zbl Bakt 1990;20 (suppl.):419-428.

37. Waldher BJ, Henderson CI, Miles RJ, et al. A mutant of Mycoplasma mycoides subsp. mycoides lacking the hydrogen peroxide-producing enzyme L-alpha glycerophosphate. FEMS Microbiol Letters 1990;72:127-130. 103

• Uninfect. -•-R19(ab) -A-R94(a) -e-HILUab) -*-MLA2(a) -•-WNS(a)

Day after inoculation

Figure 1. Percent ciliary activity in ovine tracheal rings uninfected (uninfect.) or infected with Mycoplasma ovipneumoniae isolates R19, R94, HILL, MLA2 or M. arginini isolate WNS. Values represent means percentages (+. SEM) of 2 separate experiments (N = 2-3 tracheal rings per experiment). (a)Significantly different (p<0.05) from control values. (b)Significantly different (p<0.05) when R19 or HILL was compared with the remaining isolates, exclusive of WNS at day 2. Figure 2a. Benzidine reaction of the different Mycoplasma ovipneumoniae isolates (R19,

R94, Hill, MLA2 and WNS). Isolates were grown in modified Friis broth medium and placed in wells cut into the benzidine-blood-agar plates. A 30% solution of HjOz and fresh

Friis broth medium were included as positive and negative controls, respectively. Plates were incubated and examined for black coloration around wells as an indicator of the presence of HiOj.

Figure 2b. Inhibitory effect of catalase on benzidine reaction of Mycoplasma ovipneumoniae, grown in modified Friis broth medium. Isolate R19 was grown as for Fig.

2a and placed, along with a 30% solution of HjOj or fresh Friis broth medium, in wells cut mto benzidine-blood-agar plate supplemented with catalase. The plate was incubated and examined for presence of black coloration. 105

R19

B Hill

Friis 106

Uninfect. R19 R19+Cat

Day after inoculation

Figure 3. Effect of catalase (Cat.) on the ciliary activity of ovine tracheal rings infected with Mycoplasma ovipneumoniae isolate R19. Values represent mean percentages of one experiment (N= 2-3 tracheal rings). Uninfect. = uninfected rings. Figure 4. Electron micrograph of a fetal ovine tracheal ring explant, after 4 day infection

with Mycoplasma ovipneumoniae and stained with ruthenium red. The epithelial cells have

lost cilia and a mycoplasma (M) is seen closely associated with remaining cilia (CI). A

dense thick layer of capsular material apparently bridges not only the mycoplasma and the cilia but also adjacent mycoplasmas in the form of a microcolony. Bar = 0.5 im.

Figure 5. Electron micrograph of a fetal ovine tracheal ring explant, after 8 day infection

Mycoplasma ovipneumoniae. Numerous mycoplasmas (M) are among the cilia (CI). A few mycoplasmas are seen near the microvilli (mv). No mycoplasmas are found associated with the basement membrane or within the lamina propria. Bar = 0.5 /zm.

Figure 6. Electron micrograph of a noninfected fetal ovine tracheal ring explant, after 8 day cultivation. Section appears normal with intact and organized epithelium. Bar = 0.5

Ijm.

113

Table 1. Calmodulin levels in tissue culture medium from uninfected ovine tracheal rings and rings infected with Mycoplasma ovipneumoniae isolates R19, R94, Hill, MLA2 or M. arginini isolate WNS at 2, 6 and 8 days after inoculation.

Day after inoculation Isolate 2 6 8 None 1.791Ji0.000 1.229 +0.000 0.840+0.000 R19 2.738ii0.032 3.274+0.000'' 1.832±0.000'' R94 1.8874:0.003 1.684+0.113'"' LSSO+O-OOO"^ Hill 2.417+.0.142 2.867±0.003'' 2.010+0.014" MLA2 1.338ji0.000^'' 1.404+0.000^ 1.268+0.363^'' WNS 2.603+.0.001 2.088 +0.011^ 0.386+0.008^ Results are means (+SD) from two separate ejqperiments (N = 3 pooled tracheal ring spent medium). ^Significantly lower (p<0.05) than the respective R19 and Hill values. ''Significantly lower (p<0.05) than the respective WNS value. '^Significantly higher (p<0.05) than the respective control value. '^Significantly lower (p < 0.05) than the respective Hill value. 114

CHAPTER 5. IMMUNE RESPONSE IN LAMBS NATURALLY INFECTED WITH MYCOPLASMA SPECIES

A paper to be submitted to the J. Vet. Diagn. Invest.

Mamadou Niang, MS; Jose Lopez-Virella, BS; Merlin L. Kaeberle, DVM, Ph.D

Abstract

Antigens shared by Mycoplasma ovipneumoniae with other species of mycoplasma contribute to difficulty in interpreting serologic results. Cross reactivity ^iih. Mycoplasma arginini was confirmed and led to development of a serological test specific for

M. ovipneumoniae. A crude capsular extract obtained from M. ovipneumoniae was found to be specific for detecting antibodies to this microorganism.

This crude capsular material and sodium dodecyl sulfated-treated M. ovipneumoniae and M. arginini cells were used in ELISA tests to evaluate antibody responses to M. ovipneumoniae and M. arginini in serum samples of lambs with chronic respiratory disease. Sera were obtained firom lambs in several flocks at various stages of the clinical disease. There were low levels of antibody to M. ovipneumoniae in flocks sampled at the early stages of infection whereas increased levels of antibody were present in lambs from flocks that had apparently recovered from the clinical disease. Slowly rising titers of circulating antibodies to M. ovipneumoniae was confirmed by sequential bleeding of lambs during the course of the clinical disease. On the other hand, antibody levels to M. arginini 115 were more likely to increase earlier in the disease process. The results suggest that the chronic nature of this disease may result from the failure of the immune system to produce antibodies that are protective against M. ovipneumoniae infection. At such a time that appreciable levels of specific antibodies appear in the serum (several weeks following infection) lambs seem to recover from the clinical disease.

Introduction

M. ovipneumoniae (MO) is one of the most commonly isolated microorganisms from sheep with respiratory disease worldwide (Davies, 1985; Malone et al., 1988).

Recently in this laboratory, this microorganism as well as M. arginini (MA) have been routinely recovered from young Iambs with a respiratory disease that we have termed the

"coughing syndrome" since a severe paroxysmal cough leading to rectal prolapses is the primary clinical response. This syndrome is widespread in the midwestem states of the

United State of America (USA) with morbidity and severity variable among flocks. The disease is chronic and persists for several weeks in most affected lambs. There is some evidence that the extended persistence of the MO organism in the respiratory tract of these lambs and the chronic nature of the disease may be due to failure of the immune system to generate protective immunity (Chapter 6 in this dissertation), suggesting that infection with the microorganism may have manipulative effects upon the respiratory immune system of lambs. Reasons for that are not known but marked variation in the MO organisms isolated from sheep (Khan, 1993) and expression of a polysaccharide capsule by the organisms 116

(Chapter 3 in this dissertation) are recognized. This study was undertaken to determine the levels of antibodies to MO and MA antigens in the serum of naturally-infected lambs to better understand the nature of the immune response against these organisms.

Because of its simplicity, ELISA has been extensively used in detecting antibodies to mycoplasma species. However, concerns exist about its specificity because of cross reactions among several mycoplasma species inherent in the nature of antigen preparations used (Bereiter et al., 1990). Several attempts have been made to overcome such specificity problems by using different methods of antigen preparation (Nicolet and Paroz, 1980,

Kazama et al., 1989). With regard to MO, Donachie et al. (1982) used solubilized whole cell antigens and made the same observations. They concluded that a more specific antigen was needed for future use. Consequently, a crude capsular material was extracted from MO for use as a diagnostic antigen and confirmed the specificity of this reagent in ELISA and western immunoblotting procedures. We were able to distinguish humoral response to MO from the response to MA as measured with crude capsular material extracted from MO.

Materials and Methods

Animals and sample collection

Full details of the lambs used and collection of sera and nasal swabs from these lambs are described elsewhere (Chapter 6 in this dissertation).

Hyperimmune sera against MO and MA were raised in horses as described elsewhere (Khan, 1993). 117

Preparation of mycoplasma antigens

MO and MA organisms were grown in modified Friis broth medimn for 3 days at

37°C (Khan, 1993). Cultures were harvested and washed 3 times m phospliate buffered saline (PBS, pH 7.4) by centrifiigation at 10,0(K) g and 4°C for 30 minutes. The protein concentration of washed mycoplasmas was determined with a Bio-Rad Protein Assay Kit

(Bio-Rad Laboratories, Hercules, CA) following the instructions of the manufacturer, and aliquots were stored at -70°C.

SDS-treated MO and MA antigens were prepared by mining the respective washed organisms with SDS (1 mg/mg of cell protein) at 37'C for 30 minutes (Piffer et al., 1984).

These preparations were dialyzed in 0.01 M Tris-NaCl buffer (pH 7.2) for 48 hours, and aliquots were stored at -20°C.

Crude capsular material of MO was prepared according to the method of Almeida

(1991) with some modifications. Washed MO organisms were incubated in PBS at 5&C for

30 minutes and centrifuged as before. The supernatant containing the capsular material was filtered through a 0.22 fim membrane filter. The protein content of the preparation was determined as before and aliquots were stored at -2CPC. Removal of the capsule was indicated by a experiment demonstrating that the decapsulated cells could not be agglutinated by any of the lectins specific for the organism. 118

Enzyme-linked immunosorbent assay (ELISA)

Both SDS-treated MO and MA antigens were diluted in carbonate-bicarbonate

buffer to 6 fig protein/mi. Crude capsular MO antigen was diluted in PBS to 15 fig

protein/ml.

One hundred fil of each antigen preparation was coated onto individual wells of 96

well, flat bottom microtiter plates (Immulon 1, Dynatech Laboratories, Inc., Alexandria,

VA). The plates were incubated overnight at 4°C. The coated wells were washed 9 times

with an ELISA wash buffer (0.13 mM sodium chloride, 150 mM potassium phosphate

dibasic, 320 mM phosphate monobasic, 270 mM potassium chloride and 0.05% Tween

20, pH 7.2) to remove unabsorbed antigens. Then, 100 fil of ammonium chloride (0.1 M)

was added to each well to block unreacted sites. After 1 hour incubation at 3TC, the plates

were again washed 9 times with ELISA wash buffer. One hundred fil of each sheep test

serum, diluted 1:800 in PBS containing 0.05% Tween 80 and 1% horse serum, was added

in duplicate to individual wells except for 2 wells in each plate that served as background controls. Positive and negative control sera were also included in each plate. The plates

were incubated 2 hours at room temperature and washed again 9 times with ELISA wash

buffer. Then, 100 fil of rabbit anti-sheep IgG (H+L) conjugated to horseradish peroxidase

CBvirkegaard & Perry Laboratories Inc., Gaithersburg, MD, KPL), diluted 1:780 in PBS containing 1 % horse serum, was added to each well. The plates were incubated and washed as described above. The wells were developed with 100 /tl/well of one component 2.2'- azino-di 3-ethyl-benzthioazoline sulfonate peroxidase substrate (ABTS, KPL) solution at 119 room temperature for approximately 12 minutes and the reaction stopped by adding 100 nl of 1% SDS solution.

The optical density (OD) values were measmred with an automatic microplate reader

(Dynatech MR600 microplate reader, Dynatech) at 405 mn. The net OD values of individual samples were determined by subtracting the OD value of background wells from that of each sample.

The reactivity of the different antigens against horse hyperimmune serum to each mycoplasma was tested as described except that each serum was diluted at 1:200, 1:400,

1:800 and 1:1600 and incorporation of rabbit normal serum instead of horse normal serum in the diluents. Peroxidase labeled goat anti-horse IgG (gamma) (KPL) was used as the secondary antibody.

Western immunoblotting

Washed MO or MA organisms were adjusted to a final concentration of 1 mg of protein per ml with PBS. Crude capsular material of MO was adjusted to a final concentration of 4 mg protein per ml. Antigens were separated under reducing conditions using 4% stacking gel and 10% running gel (Laemmli, 1970) and a vertical slab mini-gel apparatus (Bio-Rad). Prestained SDS-polyacrylamide gel electrophoresis standards (broad range, Bio-Rad) were included in each gel as molecular mass standards. The electrophoresed proteins along with the molecular mass standards were electrophoretically transferred onto 0.45 fxm. pore size nitrocellulose membranes (Bio-Rad) at 4?C for 90 120

minutes at 100 volts in transfer buffer consisting of 25 mM Tris, 192 mM glycine (Bio-

Rad, pH 8.3) and 20% v/v methanol.

The membranes containing the separated proteins were blocked overnight at 4'C with

1 % gelatin in TBS (500 mM NaCl, 20 mM Tris, pH 7.5). The membranes were subsequently washed with gentle agitation 3 times (5 minutes each) in TTBS (TBS containing 0.05% Tween 20) and reacted 1 hour at room temperature with the appropriate test Iamb or horse sera diluted in diluent solution (TTBS containing 1 % gelatin). The membranes were again washed as before and incubated at room temperature for 1 hour with gentle agitation in peroxidase-labelled anti species IgG (L+H) antibody (KPL) diluted in diluent solution. After washing, one component TMB membrane peroxidase substrate solution (KPL) was used for color development. After 10-15 minutes incubation at room temperature, the reaction was stopped by rinsing the membranes in deionized water.

Bacterial examination

Isolation and identification of mycoplasmas and bacteria of pathogenic importance from the collected nasal swabs are ftilly described elsewhere (Chapter 6 in this dissertation).

Results

The presence of both MO and MA organisms in lambs in the various flocks was indicated by the presence of serum antibodies and isolation of the agents from nasal swabs. 121

The results in Table 1 highlight the reactivity of SDS-treated MO or MA antigens, or the crude capsular material of MO to horse hyperimmune sera against each organism. At various serum dilutions, there was substantial cross-reaction of anti-MA serum with SDS- treated MO antigen, less cross-reaction of anti-MO serum with SDS-treated MA antigen and no cross-reaction of the anti-MA serum with crude capsular material of MO. Several convalescent sheep sera reactive to MA antigens were tested for their reactivity against the

SDS-treated or the crude capsular MO antigens. While these sera possessed reactivity with

SDS-treated MO, there was no detectable reactivity with capsular MO antigen (Table 2).

Reactivity of the same antigens with each of the horse hyperimmune sera or the convalescent sheep sera was also evaluated with immunoblotting. Each antigen preparation reacted with the homologous hyperimmune serum (Figure 1). While several bands developed to SDS-treated MO antigen when reacted with anti-MA sera, no reactivity could be seen when the crude capsular material of MO was reacted with the same anti-MA sera.

A few bands were present when SDS-treated MA antigen was reacted with anti-MO hyperimmune serum (Figures 1, 2).

The results of serological testing of lambs from the 10 flocks are provided in Tables

3, 4, 5. Clinical disease was not evident in two flocks (HD, MY and ESI) at the time of bleeding while flocks KP, SN, ES2 and AB were at mid-stage of the disease and only a small percentage of lambs was still coughing in flocks HR, MR and RB. While there was extensive variability in titers among lambs in each flock, interesting trends were noted.

Appreciable levels of antibodies to MA were present in lambs without clinical disease. 122

High levels of antibody to MA were present in lambs at the mid-stage of the disease process. Low levels of antibody to MO were present in lambs in the miaffected flocks and there was only a moderate rise during the first several weeks of clinical disease. However, lambs had developed high levels of antibody to MO late in the disease process when most lambs had apparently recovered.

A similar pattern of antibody production to MO and MA was observed when lambs in a flock in which the disease is endemic were sequentially bled prior to, during, and after the disease process (Figure 3). Many of the Iambs in the flock developed a paroxysmal cough at 10 to 12 weeks of age. At that time, the mean antibody levels to both MO and

MA were relatively low, but antibody levels against MO antigens were lower than for MA antigen. As the disease progressed, more lambs seroconverted and the mean antibody levels to MO antigens rose dramatically late in the disease process when most lambs had apparently recovered at approximately 25 to 27 weeks of age. In contrast, the mean antibody levels to MA increased early during the disease process and rapidly declined during the recovery phase. Both MO and MA were routinely isolated from nasal swabs taken from some of the Iambs in the flock throughout the duration of the experimentation.

Other bacteria that were frequently isolated from the same samples included N. ovis,

Pasteurella sp. and Escherichia coli. 123

Discussion

ELISA and western inmunoblotting performed in the present study indicated

significant cross-reactivity between the SDS-treated antigens of MO and MA when horse

hyperimmune sera of each specificity or convalescent sheep sera were tested against these

antigens. The crude capsular material provided specificity in detecting MO antibodies.

Cross-reactivity has previously been demonstrated between MO and other species of

mycoplasmas including M. hyopneumoniae, M. dispar and M. floculare by immunoblotting

(Thirkell et al., 1991). Cross-reactivity has also been observed with ELISA reactions of

MO whole ceU antigen solubilized in carbonate-bicarbonate buffer with hyperimmune sera

of a variety of mycoplasma species (Donachie and Jones, 1982; Thirkell et al., 1991). Such cross-reactions would certainly interfere with the specificity of any serological assay, but in

particular the ELISA. This is particularly important when detecting antibody to MO and

MA since both organisms frequently coexist in sheep.

Cross-reactions in ELISA between Af. hyopneumoniae and a nonpathogenic M. floculare in pigs (Bereiter et al., 1990) is widely recognized and is a limiting factor for its use in the serodiagnosis of mycoplasmal pneumonia of swine. Several attempts have been

made to overcome this specificity problem in ELISA by using different methods for antigen

preparation.

Several workers used purified Tween 20 extracts of M. hyopneumoniae and found superior specificity (Nicolet and Paroz 1980; Kazama et al. 1989). These improved ELISA antigens have been described as suitable for the practical serodiagnosis of mycoplasmal 124 pneumonia of swine (Yagihashi et al., 1993) and are available in a commercial

M. hyopneumoniae ELISA test kit.

Likewise, there was a need to develop a more specific ELISA antigen to assist in the detection of mycoplasmal pneumonia of sheep and the crude capsular material described here should be smtable as a reliable ELISA antigen in detecting MO specific antibodies.

We further observed that storage at -20PC for at least 2 years did not affect the activity and the specificity of the antigen.

The results of this experimentation indicated that lambs readily produce antibodies in response to MA but the response to MO was delayed. This slow development of circulating antibodies to MO could be confirmed by sequential bleeding of lambs during the course of the disease.

There is an apparent failure to generate an immune response to MO that will eliminate the organisms fi-om the respiratory tract until late in the clinical disease. At such a time that levels of specific antibodies rise in the serum, lambs seem to recover from the clinical disease. This finding corroborates our previous observations that infection with the microorganism has manipulative effects upon the respiratory system of lambs leading to development of aberrant immune responses such as immediate hypersensitivity and ciliary autoantibodies during early stages of the disease process (Chapter 6 in this dissertation).

This view can be further supported by reports of temporary abatement of clinical signs in lambs treated with antihistamines early during the clinical disease process (Kaeberle, 125

personal observations) and demonstration of respiratory allergy in association with

mycoplasmal infection in people (Yano et al., 1994).

Acknowledgements

The authors thank Roman Pogranichnyy, Don Hummel and Susan L. Seiler for

technical assistance. This work was supported in part by a grant from the USDA, Iowa

Sheep and Wool Promotion Board and the Iowa Livestock Health Advisory Council.

References

1. Almeida RA, Rosenbusch RF: 1991, Capsulelike surface material of Mycoplasma dispar induced by in vitro growth in culture with bovine cells is antigenically related to similar structures expressed in vivo. Infect Immun 59:3119-3125.

2. Bereiter M, Young TF, Joo HS, Ross, RF: 1990, Evaluation of the ELISA and comparison to the complement fixation test and radial immunodiffusion enzyme assay for detection of antibodies against Mycoplasma hyopneumoniae in swine serum. Vet Microbiol 25:177-192.

3. Davies DH: 1985, Aetiology of pneumonias of yoimg sheep. Prog Vet Microbiol Immun 1:229-248.

4. Donachie W, Jones GE: 1982, The use of ELISA to detect antibodies to Pasteurella haemofytica A2 and Mycoplasma ovipneumoniae in sheep with experimental chronic pneumonia. In: The ELISA: Enzyme-Linked Immimosorbent Assay in Veterinary Research and Diagnosis, eds. Wardly RC, Crowther JR, pp. 102-111. Martinus Nijhoff Publishers, The Hagues.

5. Kazama S, Yagihashi T, Seto K: 1989, Preparation of Mycoplasma hyopneumoniae antigen for the enzyme-linked unmunosorbent assay. Can J Vet Res 53:176-181.

6. Laemmli UK: 1979, Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680-685. 126

7. Malone FE, McCuUough ST, McCullough MF, BaU H, et al: 1988, Infectious agents in respiratory disease of housed, fattening lambs in Northern Ireland. Vet Rec 122:203-207.

8. Nicolet J, Paroz P: 1980, Tween 20 soluble proteins of Mycoplasma hyopneumoniae as antigen for an enzyme linked immunosorbent assay. Res Vet Sci 29:305-309.

9. Piffer lA, Young TF, Petenate A, Ross RF: 1984, Comparison of complement fixation test and enzyme-linked immunosorbent assay for detection of early infection with Mycoplasma hyopneumoniae. Am J Vet Res 45:1122-1126.

10. Thirkell D, Spooner RK, Jones GE, Russel WC, et al: 1991, Cross-reacting antigens between Mycoplasma ovipneumoniae and other species of mycoplasma of aniTnal origin, shown by ELISA and immunoblotting with reference antisera. Vet Microbiol 26:249-261.

11. Yagihashi T, Kazama S, Tajima M: 1993, Seroepidemiology of mycoplasmal pneumonia of swine in Japan as surveyed by an enzyme-linked immunoabsorbent assay. Vet Microbiol 34:155-166.

12. Yano T, Ichikawa Y, Komatu S et al: 1994, Association of Mycoplasma pneumonia antigen with initial onset of bronchial asthma. Am J Resp Crit Care Med 149:1348- 1353. 127

Table 1. Reactivity in ELTSA of crude capsular material of M. ovipneumoniae (MO- CAP) and sodium dodecyl sulfate-treated MO (MO-SDS) and M. arginim (MA-SDS) antigens against horse hyperimmune serum to each mycoplasma.

OD Values (+SD) Antigens Hyperimmune serum Dilutions MO-CAP MO-SDS MA-SDS Negative 1 200 0.157+0.040 0.196+0.050 0.135 +0.025 1 400 0.090+0.020 0.119±0.054 0.111Ji0.030 1 800 0.054+0.010 0.056±0.033 0.079Ji0.028 1 1600 0.041+0.010 0.046±0.015 0.075Ji0.025

Anti-MO 1 200 1.693 +0.221 1.734+0.200 0.497 4i0.155 1 400 1.535+0.116 1.656+0.162 0.266+l0.029 1 800 1.400±0.013 1.604+0.135 0.165+0.032 1 1600 1.177+0.105 1.500+0.104 0.100Ji0.027

Anti-MA 1 200 0.189+0.091 1.161+0.163 1.735+0.197 1 400 0.173 +0.044 0.890+0.164 1.590+0.139 1 800 0.096+0.023 0.544+0.165 1.579+0.164 1 1600 0.062 +0.013 0.295 +0.074 1.493+0.145

Values represent mean optical densities (OD) of 3 separate experiments. 128

Table 2. Reactivity in ELISA of various convalescent (S4-352 to S5-560) and negative (C1-C3) sheep sera with crude capsular material of M. ovipneumomae (MO-CAP) and sodium dodecyl sulfate-treated MO (MO-SDS) and Af. arginini (MA-SDS) antigens.

OD Values (+SD) Antigens Serum# MO-CAP MO-SDS MA-SDS S4-352 0.084±0.013 0.382 +0.033 0.755+0.014 S4-359 0.213-^0.008 0.571+0.010 1.190+0.030 S5-485 0.174+0.001 0.348 +0.005 0.769 +0.009 S5-560 0.191+0.007 0.390+0.018 0.559+0.043 CI 0.117+0.005 0.164+0.018 0.168+0.013 C2 0.087 +0.012 0.112+0.000 0.109+0.016 C2 0.156+0.014 0.123 +0.003 0.115+0.006

Values represent mean optical densities (OD) in duplicate wells. 129

Figure 1. Immunoblots of sodium dodecyl sulfate (SDS)-treated Mycoplasma ovipneumoniae whole cells (lane 2), crude capsular material prepared from the same M. ovipneumoniae organisms (lane 3) and SDS-treated M. arginini whole cells (lane 4) developed with horse hyperimmune serum prepared against M. ovipneumoniae (panel A) and horse hyperimmune serum prepared against M. arginini (panel B). Lane 1 represents molecular mass standards in kDa. Peroxidase-conjugated rabbit anti-horse IgG whole molecule was used as a secondary antibody. While multiple bands developed to SDS- treated M. ovipneumoniae whole cell antigens when reacted with anti-M. arginini serum, no staining bands could be seen when the crude capsular material from M. ovipneumoniae was reacted with the same anti-M. arginini serum. Staining bands were present when the SDS-treated M. arginini antigen was reacted with anti-Af. ovipneumoniae serum. 130

Figure 2. Immunoblots of sodium dodecyl sulfate (SDS)-treated Mycoplasma ovipneumoniae whole cells (lane 2), crude capsular material prepared from the same M. ovipneumoniae (lane 3) and SDS-treated M. arginini whole cells (lane 4) developed with convalescent lamb sera (panels A and B) which were positive by ELISA for both SDS- treated mycoplasma whole cell antigens but negative for M. ovipneumoniae crude capsular material. Lane 1 represents molecular mass standards in kDa. Peroxidase conjugated rabbit anti-sheep IgG whole molecule was used as secondary antibody. Multiple cross reactive bands developed between the two SDS-treated mycoplasma antigens. However, bands were absent or very weekly stained in the crude capsular material from M. ovipneumoniae. 131

Table 3. Antibody levels to sodium dodecyl sulfate-treated M. ovipneumoniae in lambs from flocks sampled at different stages of disease.

Flock # Tested Mean TiterC±SD) Range of Titer Disease Stage HD 12 19.496±4.196 11.411-27.500 Early MY 12 19.390+5.395 10.847-27.258 Early ESI 8 28.757+16.414 13.989-58.199 Early KP 20 35.284+9.717 25.021-58.285 Middle SN 8 27.932+10.783 14.552-48.266 Middle ES2 8 46.142+22.146 16.612-84.239 Middle AB 6 42.478+15.430 25.638-69.282 Middle HR 15 48.874+25.241 18.606-105.913 Late MR 10 35.284±9.717 25.021-58.285 Late KB 6 70.380 +20.930 38.783-97.872 Late

Values represent mean titer expressed in S/P ratios. 132

Table 4. Antibody levels to crude capsular material of M. ovipneumoniae in lambs from flocks sampled at different stages of disease.

Flock # Tested Mean TiterC±SD) Range of Titer Disease Stage HD 12 11.573 +4.715 5.508-20.469 Early MY 12 18.296+10.181 4.946-38.910 Early ESI 8 18.493±12.375 10.086-40.393 Early KP 20 25.510+13.752 6.453-57.342 Middle SN 8 26.777+12.456 13.433-51.102 Middle ES2 8 48.012+24.404 17.088-83.737 Middle AB 6 36.547+16.862 18.378-62.312 Middle HR 15 56.686iil6.139 29.562-90.069 Late MR 10 50.874+21.556 33.696-107.325 Late RB 6 62.246+28.201 21.930-98.993 Late

Values represent mean titer expressed in S/P ratios. 133

Table 5. Antibody levels to sodium dodecyl sulfate-treated M. argimni in lambs from flocks sampled at different stages of disease.

Hock # Tested Mean Titer(±SD) Range of Titer Disease Stage HD 12 26.100±3.462 22.050-32.264 Early MY 12 32.299+12.521 11.153-50.379 Early ESI 8 21.515+12.709 12.054-50.8260 Early KP 20 53.529 +27.910 14.731-113.342 Middle SN 8 40.813jil3.470 18.204-62.611 Middle ES2 8 45.223 +9.0603 26.809-56.472 Middle AB 6 38.772±13.980 25.647-59.2772 Middle HR 15 44.952±24.468 20.109-96.468 Late MR 10 35.786iil7.518 12.720-64.292 Late RB 6 35.044+13.136 20.530-59.138 Late

Values represent mean titer expressed in S/P ratios. 134

60t

-MO-SDS •MO-CAP •MA-SDS

« 20

3 4 5 6 7 8 9 10 11 12 13 14 Bleeding Period (2-3 weeks per interval)

Figure 3. Antibody levels to M. ovipneumoniae (MO) and M. arginini (MA) in sera of lambs in an affected flock. (MO-SDS) Sodium dodecyl sulfate-treated M. ovipneumoniae whole cells. (MO-CAP) M. ovipneumoniae crude capsular material. (MA-SDS) Sodium dodecyl sulfate-treated M. arginini whole cells. 135

CHAPTER 6. OCCURRENCE OF AUTOANTIBODIES TO CILIA IN LAMBS

WITH A "COUGHING SYNDROME"

A paper to be submitted to the Vet. Immun. Immunopath.

Mamadou Niang, MS; Ricardo R. Rosenbusch, DVM, Ph.D; John J. Andrews, DVM, Ph.D; Merlin, L. Kaeberle, DVM, Ph.D.

ABSTRACT

A respiratory disease of lambs that has been termed the "coughing syndrome" has been observed in the mid-west region of the United States of America. Mycoplasma ovipneumoniae and Mycoplasma arginini were routinely isolated from the respiratory tract of lambs with this disease. Using an enzyme-linked immunoabsorbent assay, a high level of antibodies reactive with ovine cilia of the upper respiratory tract was detected in the sera from many of the lambs in affected flocks but not in sera of lambs from unaffected flocks.

The reactivity of these antibodies with cilia was confirmed by indirect immimofluorescent staining. These antibodies were predominantly of the IgG isotype. They were distinct from cold agglutinins and could be absorbed from the sera with cilia but not with antigens of common bacterial pathogens of the sheep respiratory tract including M. ovipneumoniae, M. arginini, Pasteurella heamotytica, Pasteurella multocida or Neisseria ovis. In addition, their occurrence appeared to be independent from the specific antibodies to M. ovipneumoniae and M. arginini. The demonstrated antibodies were determined by Western 136 immunoblottiiig to be directed primarily against antigens with apparent molecular weights ofSOkDa.

One flock from which serial serum samples were collected from the same lambs over a ten month period indicated that the antibodies developed before the onset of the clinical disease and persisted for a period of several months until most of the lambs had apparently recovered. However, colonization of the respiratory tract of the lambs by

M. ovipneumoniae preceded the production of these antibodies. Sequential serum samples taken from another flock, with no known history of this coughing, showed no such antibodies throughout the bleeding period. On the basis of these results it is suggested that an immunopathologic mechanism involving production of autoantibodies directed against ciliary antigens of the lambs could be a contributing factor to the pathogenesis of this clinical syndrome. Furthermore, the possible association of mycoplasmal infection with the production of these autoantibodies is discussed.

1. Introduction

M. ovipneumoniae (MO) has been routinely recovered from the respiratory tract of sheep with respiratory disease worldwide (Davies, 1985). Recently, in this laboratory, this organism as well as M. arginini (MA) have been isolated from a significant number of lambs with a respiratory disease that we have termed the "coughing syndrome" because of the nature of its clinical signs which include mainly paroxysmal cough predisposing to rectal prolapses and reduced weight gain. This syndrome is widespread in the mid-west 137 states of the United States of America (USA) with morbidity and severity variable among flocks. The disease is chronic and persists for several weeks, even months, in most affected flocks. MO can be consistently demonstrated in association with the cilia of the upper respiratory tract of these affected lambs. While investigating the blocking effect of convalescent sera from coughing lambs on the attachment of the agent to ovine cilia we noted high reactivity of these sera with ciliary antigens. Thus, it was hypothesized that an immunopathologic mechanism involving production of autoantibodies against cilia could contribute to the pathogenesis of this disease.

Autoimmune responses associated with mycoplasmal infection, though overlooked, has been known for some time. The occurrence of autoantibodies to various tissue antigens has been demonstrated in humans after infection with M. pneumoniae (Biberfield, 1985) and in pigs infected withM. hyopneumoniae (Baimigartner and Nicolet, 1984; Roberts and

Littie, 1970; Suter et al., 1985). A role for autoimmune reactions in the pathogenesis of pleuropneumonia in cattle infected withAf. mycoides (Cassell et al., 1985; Nicholas and

Bashiruddin, 1995), as well as polyarthritis in rats infected with M. arthritidis (Kirchhoff et al., 1989; Runge et al., 1990) have been suggested.

The purposes of the present study were to investigate the occurrence of autoantibodies to cilia in sera of lambs with the coughing syndrome and to determine the clinical significance of these autoantibodies. 138

2. Materials and methods

2.1. Animals and sample collection

Lambs from twenty different sheep flocks around the state of Iowa, USA were included in this study. The lambs were of various breeds and most were approximately 10 to 12 weeks of age when bled for serum during the period of 1989-1995. The disease status, as determined on the basis of clinical signs which included nasal discharge, cough, prolapses and poor growth, for each flock was recorded at the time of the bleeding by the investigators or as reported by the owner or attending veterinarian. In one flock with a history of clinical disease, three week-old lambs were sequentially bled and nasal swabbed at 2-3 week intervals over a period of approximately 10 months. In a second flock (with no observable disease) sequential samples were obtained over an eight-month period. In addition, blood samples were taken from nine lambs in another flock with no history of clinical respiratory disease and were free of mycoplasmas at the time of bleeding; these served to provide reference serum samples.

Sera were harvested within 24 hours of bleeding and stored at -20'C until tested.

Nasal swabs were processed immediately for mycoplasmas and bacterial isolation.

2.2. Preparation of cilia antigen

Tracheas were obtained from new-bom lambs that either died immediately after birth or were sacrificed by intravenous injection of Beuthanasia-D special containing 350 mg pentabarbitone/ml (Schering-Plough Animal Health Corp., Kenilworth, NJ). The tracheas 139 were immediately exposed and tied below the larynx with a piece of cord and separated from the Imigs by cutting at the bifurcation. Connective tissues were removed from the outer surfece. Thereafter, tracheas were washed 3 times with sterile M199 medium (Gibco

BRL, Life Technologies, Inc., Grand Island, NY) and cut into small fragments. For cilia extraction and piirificatiou, the methodology of Zhang et al. (1994) originally described by

Tuomanen et al (1988) was used. The purity of the cilia was determined by light microscopy after staining with trypan blue. The protein concentration of the cilia was determined by a Bio-Rad Protein Assay Kit (Bio-Rad Laboratories, Hercules, CA) following the instructions of the manufacturer, and aliquots were stored at -70PC.

Sodium dodecyl sulfate (SDS)-treated cilia antigen was prepared by incubating purified cilia with SDS (1 mg/mg cilia protein) at 3TC for 45 minutes (Zhang et al., 1994), and aliquots were stored at -20°C.

2.3. Enzyme-linked immunosorbent assay (ELISA)

2.3.1. Detection of autoantibodies to cilia antigen

The assay was performed using 96 well, flat bottom microtiter plates (Immulon 2,

Dynatech Laboratories, Inc., Alexandria, VA). SDS-treated cilia were diluted in carbonate- bicarbonate buffer (pH 9.5) to a final dilution of 10 /xg/ml. Some wells of the microtiter plates were coated overnight at room temperature with 100 [il of the antigen preparation while the remaining wells were left uncoated and served as controls. Plates were washed 4 times with PBS to remove unabsorbed antigen. Then, 200 fil of PBS containing 1 % gelatin 140 was added to each coated or non-coated well and plates were incubated for 2 hours at 37C.

After 4 washings with PBS, 100 n\ of each test serum, diluted 1:400 in PBS containing

0.5% gelatin, was added to duplicate cilia-coated or noncoated wells. Positive and negative control sera were also included in each plate. The plates were incubated for 70 minutes at

37°C and washed again 4 times with PBS. Then, 100 ^1 of rabbit anti-sheep IgG (H-l-L) conjugated to horseradish peroxidase (Kirkegaard & Perry Laboratories Inc. (KPL),

Gaithersburg, MD), diluted 1:700 in PBS containing 0.5% gelatin, was added to each well.

The plates were incubated and washed as described above. The wells were developed with

100 ^1/well of one component 2.2'-azino-di 3-ethyl-benzthioazoline sulfonate peroxidase substrate (ABTS, KPL) solution at room temperatiure for 14 minutes and the reaction stopped by adding 100 ptl of 1 % SDS solution to each well.

The optical density (OD) values were measured with an automatic microplate reader

(Molecular Devices Corporation, Menlo Park, CA) at 405 nm. The net OD value of each serum sample and the positive and negative control sera was determined by subtracting from the mean OD value of the serum wells the mean OD value of the corresponding background control wells. Then, the serological cut-off for each test serum was determined according to the method of Morris et al. (1995) with some modifications. The mean OD value for each test serum (S) was divided by the mean OD value of the positive control sera

(P), the resulting value multiplied by 100, and results were expressed as S/P ratios. Values in excess of two standard deviations (SD) from the average S/P ratio values of the nine senmi samples obtained from a herd with no known history of coughing, as described 141 earlier, were considered as positive. Thus, seropositive lambs for autoantibodies to cilia were defined as those with a S/P ratio greater than 23.822, which is an average S/P ratio of

14.782 +9.04.

For absorption of autoantibodies, positive sera containing autoantibodies were diluted

1:10 in PBS and mixed with 2 parts of a concentrated suspension of either purified cilia or various bacterial whole washed cells including MO, MA, P. haemolytica (PH), P. multocida (PM) and N. ovis (NO). The mixtures were incubated overnight at 4?C with gentle agitation, centrifuged at 48,(XX) g for 30 minutes at 4?C, and the supernatant representing the absorbed sera harvested. Antibody activity to cilia, MO and MA antigens of absorbed and the corresponding unabsorbed sera was determined by the ELISA. The OD values of individual samples were determined by subtracting the OD value of background wells from that of each sample.

2.3.2. Autoantibody isotyping

Levels of the different immunoglobulin classes of autoantibodies in different positive and negative sera were determined by the ELISA protocol described above except that backgroxmd controls consisted of both cilia-coated and noncoated wells. Unlabeled rabbit anti-sheep IgG heavy-chain specific or IgM-mu chain specific or IgA-alpha chain specific antibodies (Bethyl Laboratories, Inc., Montgomery, Tx), each diluted 1:150 were used as secondary antibody. Peroxidase-labelled donkey anti-rabbit IgG (H+L, Bethyl

Laboratories, Inc., Montgomery, TX), diluted 1:2000 was the detecting conjugate. The net 142

OD values of each serum sample was determined by subtracting the mean OD value of the background control wells from the mean OD value of the serum wells.

2.3.3. Detection of specific antibodies to mycoplasmas

Antigens and protocol used to test for specific antibodies to MO and MA are fiilly described elsewhere (Chapter 5 in this dissertation).

2.4. Detection of cold agglutinins

Selected positive lambs for autoantibodies, as determined by the ELISA results, were bled. Freshly collected sera were tested for cold agglutinins by the method described by

Baumgartner and Nicolet (1984).

2.5. Indirect immunofluorescence staining

Tracheal rings from newborn lambs were either frozen at -70PC in embedding mediimi for frozen tissue sections (O.C.T. Compound, Miles Inc., Diagnostics Division,

Elkhart, IN), sectioned and mounted on polylysine-coated slides, or fixed in 10% buffered neutral formalin, processed by a routine paraffin embedding technique, and microsectioned. The formalin-fixed tracheal ring section slides were deparaffinized by a standard method while the fresh frozen sections slides were fixed with acetone at 4'C for

10 minutes. 143

For indirect immunofluorescence staining, sections were treated overnight at 4°C with an autoantibody-positive or -negative serum, each diluted 1:40 in PBS. After washings fluorescein-labelled rabbit anti-sheep IgG (H-f-L, BCPL), diluted 1:20 in PBS, was added onto the slides and incubated at room temperature for two hours. The slides were mounted in glycerinated PBS (9 parts of glycerol plus 1 part of PBS) and observed under a fluorescence microscope.

2.6. SDS-polyacryiaimde gel electrophoresis and western immunoblotting

Aliquots of purified cilia, washed MO or MA organisms were adjusted to a final concentration of 1 mg of protein per ml with PBS. Antigens were separated under reducing conditions using 4% stacking gel and 10% miming gel (Laenunli, 1979) with a vertical slab mini-gel apparatus (Bio-Rad). Prestained SDS-PAGE standards (broad range, Bio-Rad) were included in each gel as molecular mass standards. The electrophoresed proteins along with the molecular mass standards were electrophoretically transferred onto 0.45 ^m pore size Immobilon membranes (Millipore, Millipore Corporation, Bedford, MA) according to the instructions provided with the product.

Blocking and subsequent washing and labelling of membranes containing separated proteins was as described in the Immobilon Technical Protocol (Millipore, Millipore

Corporation, Bedford, MA) except that the membranes were blocked for six hours at room temperature and reacted overnight at 4°C with the appropriate test sera. Alkaline phosphatase-conjugated rabbit anti-sheep IgG (H-l-L, KPL) and alkaline phosphatase 144 membrane substrate solution (BCIP/NBT, KPL) were used as secondary antibody and for color development, respectively.

2.7. Bacteriological eacaiiiiiiation

Immediately after collection, nasal swabs were placed in tubes containing brain heart infusion broth without antibiotics for transport to the laboratory. Ten-fold serial dilutions were made in modified Friis broth medium culture, and the mycoplasmas isolated and identified as described previously (Khan, 1993). All the nasal swabs were also streaked onto 5% blood agar and McConkey's agar plates for isolation and identification of bacteria of pathogenic importance.

3. Results

While investigating the potential of convalescent sera to block attachment of MO to ovine cilia (unpublished data), we noticed high reactivity of these sera with the cilia antigen. This apparent presence of antibodies to cilia was tested by developing an ELISA utilizing purified cilia firom newborn lambs as the antigen. Specificity of this ELISA was determined in preliminary experimentation by reactivity of various sera including sheep sera that were either positive or negative to MO and MA, horse anti-MO and anti-MA hyperimmune sera, and convalescent swine serum positive to M. hyopneumoniae (courtesy

DR. R.F. Ross, Veterinary Medical Research Institute, College of Veterinary Medicine,

Iowa State University, Ames, LA). Among the different sheep sera tested, some were 145

positive while others were negative irrespective of their reactivity against the mycoplasma

antigens (Table 1). In addition, neither of the horse sera nor the swine serum reacted with

the cilia antigen. Further testing of sera of lambs in several flocks indicated the prevalence

of these autoantibodies (Table 2). The results also indicated that these autoantibodies were

present in sera of many lambs in affected flocks but not in sera of lambs in unaffected

flocks.

The apparent association between the presence of autoantibodies and the development of the clinical disease led to investigation of the time of appearance and persistence of the autoantibodies. Sequential serum samples from lambs in two separate flocks, one in which the disease has occurred annually and the other with no history of the disease, were tested.

Lambs were initially bled at three weeks of age and a paroxysmal cough developed at 10-12 weeks of age in one group of lambs whereas no clinical evidence of respiratory disease was noted in the second flock throughout the duration of the experimentation. While all lambs in the diseased flock lacked detectable autoantibodies at 3 weeks of age, a few lambs were positive at the second bleeding (five weeks of age) and additional lambs seroconverted and the mean autoantibody levels increased at each bleeding until the seventh bleeding at approximately 15-17 weeks of age. This level was maintained imtil the 11'*' bleeding and then significantly declined when most lambs had apparently recovered fi:om the clinical disease (Figure 1). The percentage of positive lambs also steadily increased at each bleeding until almost 90% of the lambs had seroconverted at the ll"* bleeding and then declined significantly after the n"" bleeding when most lambs had apparently recovered 146 from the clinical disease (Figure 2). The autoantibodies were detected prior to the onset of clinical disease; however, MO was isolated from nasal swabs from some of the lambs before the detection of autoantibodies. MA was also isolated from lambs in the same flock, but not until the mid-point of the experimentation. In contrast, lambs in the healthy flock had an undetectable level of autoantibodies in their serum throughout the duration of the experimentation (Figure 3). Although both MO and MA were recovered from Iambs in the healthy flock, this was inconsistent and late in the course of the experimentation. Other bacteria that were frequently isolated from both flocks included NO, PH, PM, Escherichia coli and Klebsiella sp.

Reactivity of these autoantibodies with the ovine cilia was demonstrated by indirect immunofluorescence with either frozen or formalin-fixed trachea and lung tissue sections obtained from newborn lambs. Strong fluorescence was present when sections were treated with serum from coughing lambs and ELJSA-positive for autoantibodies (Figure 4). This fluorescence was strongly localized to the basal bodies of the cilia. In contrast, no fluorescence was observed on any part of the surface epithelium of the sections treated with serum from normal lambs or ELISA-negative serum (Figure 4). Sera containing autoantibodies were tested by immunoblotting for their reactivity against the cilia antigen.

The results indicated that the predominant reactivity of the autoantibodies was with a molecule having an apparent molecular weight of 50 kDa (Figure 5). A few weakly staining bands with apparent moleciilar weights higher or lower than the 50 kDa were also present. 147

The potential cross-reactivity of the ciliary autoantibodies with antigens of common bacterial pathogens including MO, MA, Pasteurella sp and NO, was investigated with absorption studies. Autoantibody positive sera were absorbed with the various agents or cilia and the absorbed sera along with their corresponding nonabsorbed sera were tested by

EUSA. Results are shown in Table 3. Cilia antigen markedly reduced levels of autoantibodies in all sera. Bacterial whole cells did not influence reactivity with cilia while reducing significantly levels of antibodies specific to these organisms. Strain variation is recognized among different MO isolates and that could be related to a specific cross- reactive antigen with the cilia. To rule out such a possibility, several sera containing autoantibodies were absorbed with different MO isolates, but no significant reduction in the ciliary autoantibody activity was observed when these absorbed sera were tested against the cilia antigen (Table 4). Cross-reactivity of the autoantibodies with mycoplasma antigens was also evaluated with an immunoblotting assay utilizing cilia and mycoplasma antigens.

Two bands that developed to MO antigen indicated antigens with a very similar molecular weight to the 50 kDa component in cilia antigen (Figure 6). However, absorption of these sera with cilia antigen removed or reduced significantly the intensity of the 50 kDa band in cilia antigen but not the reactivity with MO antigen. Also, absorption with MO antigen removed reactivity for MO but not for the cilia antigen.

Potential cross-reactivity was evaluated further by testing a group of sera with autoantibodies for antibodies to MO and MA. Marked variations were noted in the 148

reactivity of these sera with some being positive while others negative for one or both

mycoplasma organisms (Table 1).

Determination of the isotypes of the autoantibodies indicated that these autoantibodies

were predominantly of the IgG class (Figure 7). However, low levels of IgM and traces of

IgA could also be demonstrated in these sera. The type of antibody did not seem to change

with time as indicated by testing a pool of sequential serum samples positive for autoantibodies (Figure 8).

4. Discussion

A respiratory disease of lambs that we have termed the coughing syndrome continues to be a major problem in the mid-west states of the USA. The major clinical feamres in lambs affected with the disease are a paroxysmal cough leading to rectal prolapses and poor weight gain. We have found that a high percentage of lambs with the disease have a high level of antibodies in their serum that react with cilia of the upper respiratory tract. These antibodies can be regarded as autoantibodies and probably contribute to the disease process.

Association of viruses with this disease is unlikely on the basis of evidence available, including absence of serological responses in paired serum samples to several common sheep viruses during outbreaks of the disease and absence of histopathological lesions consistent with viral infections (unpublished observations). On the other hand, previous studies in this laboratory indicate prevalence and possible association of mycoplasmas, in particular MO, with the disease process (Khan, 1993). 149

An autoinumine response to various tissue antigens associated with mycoplasmal infection has been described in humans (Biberfeld, 1985), in cattle (Cassell et al., 1985;

Nicholas and Bashiruddin, 1995), in pigs (Suter et al., 1985; Baumgarter and Nicolet,

1984; Roberts and Litfle, 1970) and in rats (Runge et al., 1989; BCirchhoff, et al., 1990).

Therefore, occurrence of autoantibodies seems to be a rather common finding in mycoplasmal infection. However, opinions differ about their clinical relevance during the development of the disease, and there are suggestions that such antibodies result from a nonspecific inflanmiatory response.

This experimentation clearly demonstrated that autoantibodies to cilia develop during the course of a respiratory disease in lambs and possibly contribute to the disease process.

The results indicated a high level of autoantibodies in the sera of many lambs in the affected flocks but not in sera from the imaffected flocks. This clearly suggests a correlation between presence of autoantibodies and occurrence of the clinical disease. This was confirmed by clinical observation and serologic smdies on lambs in two separate flocks. Fortunately, many lambs in the flock with history of the disease developed the clinical disease at 10-12 weeks of age, which is the usual age when the disease starts.

Lambs in the second flock remained healthy throughout the duration of the experimentation. Sequential serum samples taken from the two flocks showed a high level of autoantibodies in the serum of lambs in the diseased flock but not in serum of lambs in the healthy flock. Interestingly, the mean autoantibody levels in these lambs rose before the start of the clinical disease, persisted for a period of approximately 8 weeks and 150 significantly declined when most lambs had apparently recovered jfrom the clinical disease.

MO was isolated from the nasal swabs of these lambs prior the appearance of the autoantibodies. Therefore, based on the circumstantial evidence presented here, it is tempting to suggest that colonization of the respiratory tract of young Iambs by MO and maybe MA modulates the immune response of lambs resulting in the production of autoantibodies against the cilia, leading to inflammation.

However, this view is apparently inconsistent with the finding that both MO and MA were also isolated from the lambs in the normal flock. Isolation of MO from the respiratory tract of normal lambs has been reported previously (Bakke and Nostvold, 1982; Bakke,

1982; lonas and Mew, 1985), but as in the present study, the isolation rate in these normal animals was inconsistent and not overwhelming. In addition, our study showed isolation of the organism from the normal lambs was possible only in the latter stages of the experimentation. Therefore, it would be arguably possible that early and massive colonization of the respiratory tract of lambs is necessary for the events that lead to development of autoimmunity. Massive colonization of the upper respiratory tract of lambs by MO shortly after birth has been demonstrated (Jones, et al., 1979) and suggested to play a potentiating effect on the future development of lung lesions in young lambs (Bakke and

Nostvold, 1982). Other explanations could be strain variation or environmental factors as has been suggested (Bakke and Nostvold, 1982, lonas and Mew, 1985; lonas et al., 1991); marked variation in the MO organisms isolated from sheep is recognized (Khan, 1993). In any event, it seems that the ability to elicit autoantibody production may be linked to the 151 differences in pathogenicity of the agent, the background of the lambs, and perhaps the environment. This may also partially explain the difficulties encountered when trymg to reproduce the clinical disease.

The results suggest that these autoantibodies are not cross-reactive with antigens of several cormnon bacterial pathogens of the sheep respiratory tract since they could be absorbed from the sera with cilia but not with MO, MA, PM, PH or NO whole cells.

Additional evidence comes from the observation that presence of these autoantibodies did not correlate with MO or MA specific antibodies. In fact, when the sequential serum samples taken from lambs in the diseased flock were tested for presence of specific antibodies to these mycoplasmas (data to be reported elsewhere), it was noticed that the autoantibodies were present long before the appearance of these specific antibodies. Based on this difference in time of appearance, it is likely that autoantibodies to cilia and specific antibodies to the mycoplasmas are distinct populations. The present observations are consistent with those made by Neu et al. (1987) who showed by absorption and serum neutralization tests that autoantibodies to heart myosin were not cross-reactive with

Coxsackie-virus B3 which has been implicated in the induction of autoimmune disease including myocarditis in hmnans. Also, tissues autoantibodies that are not cross-reactive with Af. pneumoniae have been reported to occur in people following infection with the agent (Lind et al., 1988; Biberfield, 1985). However, we recognize that our data do not completely exclude the possibility of shared epitopes between the cilia and the bacteria examined here. The observations also do not exclude the possibility that there are other 152 autoantigens in addition to cilia. As Neu et al. (1987) suggested, one way to explore this question is by production of monoclonal antibodies to cilia and to the organisms

Several mechanisms have been proposed to explain the occurrence of immunopathological responses during microbial infections. Among these, molecular mimicry and nonspecific polyclonal B-cell activation have been emphasized. Alternatively, the possible induction of autoimmime response as a consequence of aberrant expression of

MHC molecules on recognized antigen presenting cells, or on cells that normally do not present antigen such as epithelial cells, during infection has been suggested. The relevance of this theory is indicated by several autoimmune conditions in which abnormal expression of HLA class n molecules by specific cells of the target organs has been demonstrated

(Bottazzo et al., 1986; Smith and Allen, 1992). More recently, the in vitro upregulation of

MHC molecules on macrophages by several mycoplasmas has been demonstrated and proposed to be responsible of the generation of autoimmune pathology (Stuart et al., 1989;

Smart et al., 1990). Therefore, it is highly conceivable that during infection of lambs with

MO and MA there is concurrent activation of alveolar macrophages, increase in MHC antigen expression on these macrophages, and local tissue damage, induced by the organisms. This inappropriate expression of MHC antigens may enable these macrophages to present self antigens to autoreactive T cells during their attempt to eliminate the organisms as well as the damaged tissues. Thus the sequence of the events that lead to the clinical disease in these lambs infected with the mycoplasmas may have been early and massive colonization of the respiratory tract during the early age, ciliostasis with 153 concurrent activation of alveolar macrophages and direct or indirect aberrant MHC antigens expression on these cells. The consequences of all this would be the potential for self antigen presentation to autoreactive T cells leading to autoimmunity. Indeed such a scenario has already been postulated for mycoplasmal infections by Stuart et al. (1990).

Experimental evidence for similar suggestions have also been provided by Neu et al. (1987) in heart-specific autoantibodies induced by Coxsackie-virus B3 in a murine model. A second likely scenario would be inappropriate expression of MHC antigens by the epithelial cells, in particular ciliated cells, which enables these cells to present their own surface antigens to autoreactive T cells, therefore by-passing the requirement for antigen presenting cells as has been reported in several organ-specific autoimmune diseases (Bottazo et al.,

1986).

Polyclonal B and T cell activation by several mycoplasma species has also been demonstrated and suggested as mechanisms for development of autoimmimity. While the possibility exists that MO or MA organisms are mitogenic for B and T cells, the present fmdings imply a role for T-cell help in the production of these autoantibodies since they belong predominantly to the IgG class. While some level of IgM antibody was present, an increasing level of IgG but not IgM was evident when sequential sera from positive lambs were isotyped. In addition, unlike cases of M. pneumoniae infection in human, these autoantibodies are distinct from cold agglutinins (Data not shown). However, as we pointed out before, this does not rule out the possibility of cross-reaction with other tissues as has been demonstrated in M. pneumoniae infection (Biberfeld, 1985). 154

Another fundamental question that needs to be addressed here is how the

autoimmunity contributes to disease? An inq)ortant aspect of pathologic consequences of

autoimmunity is the formation of immune complexes with potential activation of

complement. The inflammatory processes that arise from the complement activation include cellular cytotoxicity, but most importantly, at least in our case, could be the generation of

anaphylatoxic fragments that possess various important biological activities including degranulation of mast cells and basophils. While it is apparent that IgE is the primary mediator in triggering mast cell and basophil degranulation, otiier immunoglobulin classes may also contribute through activation of complement.

Apart from complement activation, simple binding of an autoantibody to its target autoantigen can affect the host cells through several mechanisms including induction of signal transduction across host epithelial cell membranes with possible physiological effects on biologically active molecules (Sell, 1987). Other potential allergic airway reaction mediators include arachidonic metabolites that could act on mast cells or basophils. For instance, activated macrophages are known to release a variety of inflammatory mediators which in turn initiate the activation of membrane-bound phospholipase A2 (PLA2) in mast cells or basophils and a variety of other target cells, leading to release of prostaglandins, leukotrienes and platelet activating factor (Pruzanski and Vadas, 1991). Alternatively, soluble PLA2 can be released extracellularly from a variety of activated cells including macrophages and exert their pro-inflammatory effects through reactive oxygen radicals

(Pruzanski and Vadas, 1991). Therefore, since there is potential for immune complex 155

formation during the stage of autoimmuniQ^, and macrophages are known to have receptors

that interact with boimd IgG in soluble immune complexes (Levy et al., 1991), it is not

unreasonable to think that uptake of these inomune complexes by the macrophages would

activate and induce release of a variety of inflammatory products including soluble PLA2

which in turn initiate the activation of membrane-bound PLA A2 of different lung cell

types. Indeed, there are experimental reports in guinea pigs which support such a

possibility (Al-Laith et al., 1993). Additionally, Pruzanski and Vadas (1991) reported

release of soluble PLA2 from activated macrophages isolated from the lymph of sheep and

from mouse and rabbit peritoneal exudate.

The possibility also exists that IgE-mediated hypersensitivity may be involved in the

pathogenesis of this syndrome as has been suggested in asthmatic patients infected with M. pneumoniae (Tipimeni et al., 1980). More recently, Yano et al. (1994) found also high level of IgE specific to M. pneumoniae in the serum of an asthmatic patient and suggested involvement of IgE-mediated hypersensitivity. The authors speculated on the possibility that the production of these M. pneumoniae-s^tciSic IgE may arise from epitopes that are cross-reactive with components of the airway. Whether some of the autoantibodies demonstrated here belong to IgE class or whether MO elicits production of specific or cross-reactive IgE class antibodies in these lambs was not determined due to lack of specific reagent. However, oiu- previous smdies have demonstrated marked immediate skin reaction to intradermal injection of MO antigen in many of the coughing lambs

(unpublished observations). In any event, it is very tempting to speculate that development 156

of aberrant inunune responses associated with the mycoplasmal infection is the contributing

factor for the pathogenesis of Uiis clinical syndrome.

Acknowledgements

The authors thank Jose Lopez-Virella, Roman Pogranichnyy, Don Hummel, Susan L.

Seiler and David L. Cavanaugh for technical assistance. This work was supported in part by grants from the United States Department of Agriculmre, Iowa Sheep and Wool

Promotion Board and the Iowa Livestock Health Advisory Council.

References

Al-Laith, M., Weyer, A., Havet, N., Dumarey, C. Vargaftig, B.B. and Bachelet, M. 1993. Immunoglobulin-G-dependent stimulation of guinea pig lung mast cells and macrophages. Allergy, 48: 608-614.

Bakke, T. 1982. The occurrence of mycoplasmas and bacteria in lungs from sheep in Southern Norway. Acta Vet. Scand., 23: 235-247.

Bakke, T., and Nostvold, S. 1982. An investigation of ovine pneumonia in four herds from Central Norway. I. Prevalence of pneumoniae and microbiological findings. Acta. Vet. Scand., 23: 248-258.

Baumgartner, A. and Nicolet, J. 1984. Detection of antibodies against membrane- glycoproteins of swine erythrocytes after experimental infection with Mycoplasma hyopneumoniae. Vet. Immunol. ImmunopathoL, 7: 73-79.

Biberfeld, G. 1985. Infection sequelae and autoimmune reactions \n. Mycoplasma pneumoniae infection. In: S. Razin and F.M. Barile (editors). The mycoplasmas, vol. rv. Academic Press, Inc., New York. pp. 293-311.

Bottazzo, G.F., Todd, I., Mirakian, R., Belfiore, A. and Pujol-Borrell, R. 1986. Organ- specific autoimmmiity: A 1986 overview. Immunol. Rev., 94: 137-169. 157

Cassell, G.H., Clyde, W.A. and Davis, J.K. 1985. Mycoplasmal respiratory infections. In: S. Razin and M.F. Barile (editors). The Mycoplasmas, Vol. IV. Academic Press, Inc., New York. pp. 65-106.

Davies, D.H. 1985. Aetiology of pneimionias of yoimg sheep. Prog. Vet. Microbiol. Immun., 1: 229-249. lonas, G. and Mew, A.J. 1985. Colonization of the respiratory tract of lambs by strains of Mycoplasma ovipneumoniae. Vet. Microbiol., 10: 533-539. lonas, G., Norman, N.G., Clarke, J.K. and Marshall, R.B. 1991. A study of the heterogeneity of isolates of Mycoplasma ovipneumoniae from sheep in New Zealand. Vet. Microbiol., 29: 339-3347.

Jones, G.E., Buxton, D. and Harker, D.B. 1979. Respiratory infections in housed sheep, with particular reference to mycoplasmas. Vet. Microbiol., 4: 47-59.

Khan, M.A. 1993. Respiratory infection of lambs with Mycoplasma ovipneumoniae. Ph.D. Thesis, Iowa State University, Ames Iowa, USA.

Kirchhoff, H., Binder, A., Runge, M., Meir, B., Jacobs, R. and Busche, K. 1989. Pathogenetic mechanisms in the Mycoplasma arthritidis polyarthritis of rats. Rheumatol. Int., 9: 193-1996.

Laemmli, U.K. 1979. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature. 227:680-685.

Lind, K., Hoier-Madsen, M. and Wiik, A. 1988. Autoantibodies to the mitotic spindle apparatus m. Mycoplasma pneumoniae disease. Infect. Immun., 56: 714-715.

Morris, C.R., Gardner, I.A., Hietala, S.K., Carpenter, T.E., Anderson, R.J. and Parker, K.M. 1995. Seroepidemiologic smdy of natural transmission of Mycoplasma hyopneumoniae in a swine herd. Prev. Vet. Med., 21: 323-337.

Neu, N., Craig, S.W., Rose, N.R., Alvarez, F. and Beisel, K.W. 1987. Coxsachievirus induced myocarditis in mice: cardiac myosin autoantibodies do not cross-react with virus. Clin. Exp. Immim., 69: 566-574.

Nicholas, R.A. and Bashiruddin J.B. 1995. Mycoplasma mycoides subspecies mycoides (Small Colony Variant): the agent of contagious bovine pleuropneumonia and member of the "Mycoplasma mycoides cluster". J. Comp. Path., 113: 1-27. 158

Pruzanski, W. and Vadas, P. 1991. Phosphalipase A2 - a mediator between proximal and distal effectors of inflammation. Immmiol. Today, 12: 143-146.

Roberts, D.H. and Little, T.W.A. 1970. An autoimmime response associated with porcine enzootic pnemnonia. Vet. Rec., 80: 328.

Rimge, M., Binder, A., Gurr, E., Fischer, M., Meir, J. and Kirchhoff, H. 1990. Epitope- sharing between Mycoplasma arthritidis and chondrocytes demonstrated by monoclonal antibodies. Int. J. Med. Microbiol., 273: 531-538.

Sell, S. 1987. Inactivation or activation of biologically active molecules. In: Immunology, Immunopathology and Immunity. Elsevier Science Publishing Company, Inc. New York. pp. 323-348.

Smith, S.C. and Allen, P.M. 1992. Expression of myosin-class n major histocompatibility complexes in the normal myocardium occurs before induction of autoimmune myocarditis. Proc. Nad. Acad. Sci. USA., 89: 9131-9135.

Stuart, P.M., Cassell, G.H. and Woodward, J.G. 1990. Induction of class II MHC antigen expression in macrophages by Mycoplasma species. J. LnmimoL, 142: 3392-3399.

Stuart, P.M., Cassell, G.H. and Woodward, J.G. 1990. Mycoplasma induction of class n MHC antigens, possible role in autoimmunity. Zbl. Bakt., 20 (Suppl.): 575S-577S.

Suter, M., Kobisch, M. and Nicolet, J. 1985. Stimulation of immunoglobulin-containiag cells and isotype-specific antibody response in experimental Mycoplasma hyopneumoniae infection in specific-pathogen -free pigs. Infect. Immun., 49: 615- 620.

Tipimeni, P., Moore, B.S., Hyde, J.S. and Schauf V. 1980. IgE antibodies to Mycoplasma pneumoniae in asthma and other atopic diseases. Ann. Allergy, 45: 1-7.

Tuomanen, E., Towbin, H., Rosenfelder, G., Braim, D., Larson, G., Hanson, G.C. and Hill, R. 1988. Receptor analogs and monoclonal antibodies that inhibit adherence of Bordetellapertussis to human ciliated respiratory epithelial cells. J. Exp. Med., 168: 267-277.

Yano, T., Ichikawa, Y., Komatu, S., Arai, S. and Oizumi, K. 1994. Association of Mycoplasma pneumoniae antigen with initial onset of bronchial asthma. Am. J. Resp. Crit. Care Med., 49: 1348-1353. 159

Zhang, Q., Young, T.F. and Ross, R.F. 1994. Microliter plate adherence assay and receptor analogs for Mycoplasma hyopnewnoniae. Infect, and Immun., 62: 1616-22. 160

Table 1. Reactivity of various sheep sera (S1-S13), anti-M. ovipneumomae hyperimmune horse serum (H-MO), anti-ilf. arginini hyperimmune horse serum (H- MA) and anti-Af. hyopneumoniae convalescent swine serum (SW) with ovine cilia, AT. ovipneumomae (MO) or M. arginini (MA) antigens.

EUSA Value (OD) Serum# Cilia MO MA SI 1.780 0.090 0.174 S2 1.968 0.151 0.153 S3 1.281 0.182 0.723 S4 1.068 0.174 0.769 S5 0.185 0.198 1.907 S6 1.180 1.487 0.162 S7 1.852 2.068 1.926 S8 1.996 1.085 0.726 S9 1.947 0.402 0.452 SIC 0.144 0.198 0.077 811 0.143 0.187 0.098 S12 0.112 0.790 1.043 S13 0.308 1.179 0.398 HS-MO 0.097 1.336 0.168 HS-MA 0.104 0.094 1.591

SW 0.123 - -

Values represent mean optical densities (CD) in triplicate wells. (-)Not tested. 161

Table 2. Occurrence of ciliary autoantibodies (AA) in sera of Iambs in different flocks.

Flock of Origin Mean AA Titers Positive Lambs Disease Status (S/P Ratio)

NC 12.175 0(8)* -

VR 13.149 0(13) - IFl 46.23 14(16) + + IF2 12.025 0(7) - ES 21.297 2(9) + MY 45.025 17(22) + + KP 32.554 7(10) + + KR 30.328 8(10) + + LD 21.712 3(9) + m 33.665 4(10) + + AM 34.420 5(5) + + ER 32.145 4(9) + + HR 40.493 7(10) + + GN 43.469 5(10) + + MC 33.426 10(10) + + SK 50.824 7(8) + +

VL 20.712 1(10) - SN 42.783 3(6) + + HN 20.233 3(10) +

HL 13.223 0(8) -

•Numbers in parenthesis indicate total lambs sampled per flock. (-)Absent. (+)Mild. (++)Severe. 162

5 30

0 T 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Bleeding Period (2-3 weeks per interval)

Figure 1. Mean ciliary autoantibody (AA) levels in sera of lambs in an affected flock during the course of the clinical disease (N = 16-20 lambs for each bleeding). 163

90 n

80 -

70 -

oCfl a.

20 H

10 •

1 2 3 4 5 6 7 8 9 10 11 12 13 14 Bleeding Period (2-3 weeks per interval)

Figure 2. Percentage of lambs with ciliary autoantibodies in an affected flock during the course of the clinical disease (N = 16-20 lambs for each bleeding). 164

60n

-Diseased •Normal

2 3 4 5 6 7 Bleeding Period (2-3 weeks per interval)

Figure 3. Antibody levels to ovine cilia in sera of lambs in flock with and without clinical disease (N = 16-20 and 5 lambs for each bleeding period, respectively). Figure 4. Indirect immunofluoscence of formalin-fixed trachea and lung sections from a new bom lamb treated with serum firom coughing lambs and ELISA-positive for autoantibodies (A) or with serum from normal lambs and ELISA-negative for autoantibodies (B). 166

A

B 167

Figure 5. Immunoblots of ovine cilia antigen developed with different convalescent sheep sera and ELISA-positive for autoantibodies (lanes 2, 3), serum firom normal Iamb and ELISA-negative for autoantibodies (lane 4) and serum as in lane 2 but absorbed with cilia antigen (lane 5). Lane 1 represents molecular mass standards in kDa. Alkaline phosphatase conjugated rabbit anti-sheep IgG whole molecule was used as a secondary antibody. The autoantibody-positive sera reacted predominantly with a molecule having an apparent molecular weight of 50 kDa. 168

Table 3. Reactivity of sera with cilia, Mycolasma ovipneumoniae (MO) or M. arginini (MA) antigens following absorbtion with cilia, MO, MA, Neisseria oyds (NO), Pasteurella multocida (PM) or P. haemofytica (PH) antigens.

OD values Serum# Absorbent Cilia MO MA S5-595 None 1.773 1.085 0.726 Cilia 0.283 0.869 0.839 MO 1.645 0.358 0.710 MA 1.639 0.858 0.087

NO 1.643 - -

PM 1.744 - - PH 1.590 - - S5-557 None 1.026 0.338 0.294 CUia 0.222 0.229 0.195 MO 0.705 0.135 0.185 MA 0.859 0.458 0.070

NO 0.913 - - PM 0.963 - -

PH 1.024 - - S5-561 None 1.677 0.621 0.595 Cilia 0.161 0.485 0.476 MO 1.114 0.198 0.325 MA 1.239 0.349 0.066

An ELISA was used to determine the serum activity before and after absorption with each of the indicated antigens. Values represent mean optical densities (OD) in duplicate wells. (-)Not tested. 169

Table 4. Reactivity of sera with cilia following absorption with different Mycoplasma ovipneumoniae (MO) isolates.

OD values Serum# None MO (R88) MO (G76) MO (R19) S5-557 1.046 0.786 0.749 0.827 S5-561 1.296 1.219 1.165 1.077 S5-572 0.763 0.679 0.644 0.713 S5-574 1.036 1.076 1.017 0.931 S5-595 1.585 1.446 1.578 1.098

An EUSA was used to determine the serum activity to cilia antigen before and after absorption with each of the indicated isolates. Values represent mean optical densities (OD) in duplicate wells. 170

A B C 1234 12 34 1234 203 — 118 — 86 —

51.6

29 — 19.2 7.5

Figure 6. Determination of cross-reactivity of the ciliary autoantibodies with mycoplasma antigens using an immmioblotting assay. (A) convalescent serum and ELISA-positive for autoantibodies was reacted with ovine cilia (lane 2), sodiimi dodecyl sulfate (SDS)-treated Mycoplasma ovipneumoniae Qane 3) and SDS-treated M. arginini (lane 4). The same convalescent serum absorbed with cilia antigen (B) or with Af. ovipneumoniae antigen (C) was reacted with the same antigens as in A. Lanes 1 represent molecular mass standards. Alkaline phosphatase conjugated rabbit anti-sheep IgG whole molecule was used as a secondary antibody. One band that developed to M. ovipneumoniae antigen indicated a molecule with a very similar molecular weight to the 50 kDa component in cilia antigen. However, absorption with the cilia antigen did not remove reactivity with the band in M. ovipneumoniae antigen. Likewise, absorption with M. ovipneumoniae did not remove the band in cilia antigen. 171

0.8 1

• Positive sera 3v • Negative sera 13

IgG* IgM* IgA* Ig classes

Figure 7. Levels of different immunoglobulin (Ig) classes of ciliary autoantibody in positive and negative sera. Values represent mean optical densities (OD) C±SEM) of 2 separate experiments (N = 4 positive or negative sera). *(p<0.05) when mean OD values of positive sera were compared with that of negative sera for each Ig class. 172

0.6

0.4

-•-IgG -•-IgM -^IgA

0.2

0.0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Bleeding period (2-3 weeks per interval)

Figure 8. Measurement of different immunolglobulin classes of ciliary autoantibody in sera of lambs during the course of the clinical disease. Values represent mean optical densities (OD) from a pool of two sera at each bleeding. 173

CHAPTER 7. GENERAL CONCLUSIONS

A chronic respiratory disease of lambs that we have termed the coughing syndrome continues to be a major problem in sheep flocks in the mid-west states of the United States of America. The clinical disease is mainly characterized by a severe paroxysmal cough that contributes to a reduced rate of gain and predisposes to secondary bacterial pneumonias and rectal prolapses. Initial experimentation suggested prevalence and possible association of mycoplasmas, in particular M. ovipneumoniae, with outbreaks of the disease (Kaeberle and

Eness, 1988; Eness etal., 1990; Niang, 1992; Khan, 1993). Consequently, our experimentation has concentrated on the potential role of the mycoplasmas in the pathogenesis of the disease which seems to be very complex. Although infection and limg lesions can be induced by intratracheal exposure of lambs to live M. ovipneumoniae, we have not been able to routinely reproduce the clinical syndrome. However, this failure cannot be accepted as a basis for dismissing the causal relationship. Fulfillment of Koch's posmlates has been difficult to achieve with mycoplasmas even with M. mycoides subsp. mycoides, the most pathogenic of mycoplasmas (Gourlay and Howard, 1982; Stalheim,

1983; Cassell et al., 1985).

In assessing the etiological significance of such microorganisms, suggestions have been made to look for evidence relative to their ability to produce certain factors that may be suggestive of pathogenic potential, or their ability to cause damage to any host tissue, or their ability to manipulate the elements of the host immune system (Gourlay and Howard, 174

1978; Gouriay and Howard, 1982). Consequently, we have undertaken investigations aiming at identifying M. ovipneumoniae specific factors or components that may contribute to its disease producing potential. This led to demonstration of a polysaccharide capsule on the microorganism and possible role of this capsule in the colonization process of ovine respiratory ciliated cells by the organism. N-acetyl-D-glucosamine and probably D-glucose and D-mannose are the predominant lectin-reactive sugar residues of this capsule. Also, extensive strain variation among isolates of the microorganism was associated with the reactivity of the capsule with different lectins. Expression of a polysaccharide capsule by mycoplasmas can be suggestive of virulence since only those species that are proven pathogens are known to express such a component structure. A positive correlation between increased capsule production, ability to attach to host cells, and perhaps to colonize the host respiratory tract has been suggested for several respiratory mycoplasma pathogens (Wilson and Collier, 1976; Allan and Pirie, 1977; Taylor-Robinson et al., 1981;

Tajima et al., 1982; Tajima and Yagihashi, 1984). In addition, mycoplasma capsules have been reported to lack immunogenicity (Bansal et al., 1995), to interfere with phagocytosis and killing of a second bacterial target by phagocytic cells (Almeida et al., 1992), and to have toxic effects (Nicholas and Bashiruddin, 1995). More recently, evidence of a direct involvement of mycoplasma capsules in the adherence of the organisms to host cells has been reported (Yoimg et al., 1996). It is quite possible, therefore, to imply that expression of a capsule by M. ovipneumoniae may be related to its pathogenicity. 175

The second part of this work examined the cytopathic effect of Af. ovipneumordae on ovine tracheal ring cultures. It was found that M. ovipneumoniae induced loss of ciliary activity in these tracheal organ cultures. There were marked differences between isolates of the microorganism in the onset and severity of the effects which correlated with their ability to produce hydrogen peroxide. In vitro passages decreased hydrogen peroxide production and this also correlated with diminished ability of the organisms to induce ciliostasis. These findings suggest that the ability of M. ovipneumoniae to induce ciliostasis on ovine tracheal ring explants is related, at least in part, to hydrogen peroxide production, although this does not rule out a role for other factors. The mucociliary clearance mechanisms of the mucosal surfaces play an important role in the host defenses by trapping and propelling infectious agents along these surfaces. Therefore, this adverse effect of

M. ovipneumoniae on tracheal ciliary movement has implications for pathogenic properties, as has been suggested for several mycoplasma respiratory pathogens (Gabridge et al.,

1974; Levishon et al., 1986; Stadtlander et al., 1991; Debey and Ross, 1994). Presence of

M. ovipneumoniae organisms or antigens associated only with the ciliated epithelium of the infected tracheal rings was detected by transmission electron microscopy and histochemical staining, respectively. This suggests thatM. ovpneumoniae, like many other respiratory mycoplasma pathogens, is not invasive and colonizes only the ciliated epithelium.

The third part of this work was related to the characteristics of the immune responses associated with the mycoplasmal infection in lambs with this clinical syndrome. It was of interest to conduct a serological study to gain a better understanding of the nature of 176 specific iimnune response against the mycoplasmas. However, substantial cross-reaction between M. ovipneumoniae and a variety of mycoplasma species has been recognized

(Donachie and Jones, 1982; Thirkell et al., 1991) and been a liniiring factor in the development of a reliable serological test, which could be used in serodiagnosis of mycoplasmal pneumonia in sheep. Therefore, our efforts originally focussed on overcoming such a cross-reaction. A crude capsular material prepared from

M. ovipneumoniae organisms was used as antigen in an ELISA system and was found to be more specific in detecting M. ovipneumoniae antibodies than the conventional whole cell antigens. Results from this serological investigation revealed a somewhat interesting observation that the specific response to M. ovipneumoniae in lambs with the clinical disease is delayed and associated with the recovery phase. This suggested that infection with the microorganism has a manipulative effect upon the respiratory system of lambs and could result in development of aberrant immune responses as has been demonstrated in

M. pneumoniae infection in himians (Tipimeni et al., 1980; Biberfeld, 1985; Yano et al.,

1994). Also, this idea was supported by the fact that, when investigating the blocking effect of convalescent sera from Iambs with the disease on the attachment of M. ovipneumoniae to ovine cilia, high reactivity of these sera with the cilia was noted.

By the application of ELISA, IFA and western immunoblotting, we were able to demonstrate the occurrence of autoantibodies to cilia in the sera of many lambs in affected flocks but not in sera of lambs from unaffected flocks. These autoantibodies were predominantly of the IgG isotype and seemed not to be cross-reactive with antigens of 177 common bacterial pathogens of sheep respiratory tract including the mycoplasmas. These observations could imply a role for T-cell help in the production of these autoantibodies.

Interestingly, sequential serum samples taken firom one flock in which the disease has occurred annually indicated that these autoantibodies developed before the onset of the clinical signs, persisted for the duration of the clinical disease, and significantly declined during the recovery phase when levels of specific antibodies to M. ovipneumoniae increased. Although evidence for the involvement of the autoantibodies in the pathogenesis of the disease remains debatable, their relevance, based on the circumstantial evidence presented here, merits special attention.

Investigations into possible mechanisms to explain the occurrence of the autoantibodies indicated a significant increase in MHC antigen expression on sheep alveolar macrophages induced by both M. ovipneumoniae and M. arginini. This upregulation of MHC antigens may enable the macrophages to present self antigens to auto­ reactive T cells during their attempt to eliminate the organisms as well as the damaged host cells produced during the ciliostasis event. The in vitro upregulation of MHC molecules on antigen presenting cells by several mycoplasma species has recentiy been demonstrated and proposed to be responsible of the generation of autoimmune pathology (Smart et al., 1989;

Stuart etal., 1990).

Furthermore, we initiated limited experimentation to investigate the possibility of involvement of an immediate hypersensitivity in the pathogenesis of this clinical syndrome.

This was based on reports of temporary abatement on clinical signs in lambs treated with 178 antihistamines (Kaeberle, personal commumcations) and also demonstration of respiratory allergy in association with mycoplasmal infection in people (Yano et al., 1994). Results from this study indicated a marked, immediate skin reaction when lambs with the clinical disease were tested for their skin reactivity to M. ovipneumoniae antigens.

Based on the information gathered in the different experiments of this smdy, the following hypothesis as to the possible role of M. ovipneumoniae, and maybe M. arginini, in the pathogenesis of the respiratory infection in lambs is suggested. Mediated by the polysaccharide capsule, there is massive colonization of the respiratory tract of lambs by the mycoplasmas during the very early age. It is assumed that shortly after infection, the toxicity of the hydrogen peroxide produced by mycoplasmas causes ciliostasis with massive local cellular necrosis. Also, there is a concomitant activation of alveolar macrophages and increased MHC molecule expression on these macrophages. Since activated macrophages are not specific in their phagocytic functions they would engulf local damaged tissues in their attempt to eliminate the mycoplasmas. The inappropriate expression of MHC molecules may enable these macrophages to present self antigens to auto-reactive T cells.

The logical consequences of all this would be production of auto-reactive antibodies leading to inflammation. Also, there is additionally the suggestion of involvement of an immediate type hypersensitivity as another contributor to this clinical syndrome.

We do recognize that numerous aspects concerning the present model need clarification. Additional experimentation should be directed at purification and chemical characterization of the polysaccharide capsule for fimctional activities. Another study that 179 may help would be to deteimine how MHC molecule expression is modulated by the mycoplasmas. Also, evidence of immune complex and complement deposition in Imig tissues from lambs with the clinical disease warrants investigation. 180

APPENDIX A. ASSOCIATED OBSERVATIONS: EXPRESSION OF MHC CLASS I AND n ANTIGENS ON SHEEP ALVEOLAR MACROPHAGES

Abstract

Experiments were conducted to investigate the modulation of sheep alveolar macrophage MHC molecule expression by Mycoplasma ovipneumoniae and Mycoplasma arginini, isolated from lambs with respiratory diseases. Macrophages were cultured alone, or in the presence of live or heat killed organisms, and reacted with anti-MHC class I or class n monoclonal antibodies followed by labelling with fluorescein- or R-phycoerythrin- conjugated second antibody. Fluorescence measurements indicated substantial increases in the expression of both MHC classes by the mycoplasma-induced macrophages when compared with the control unstimulated macrophages. Live mycoplasmas were more effective in inducing expression of both MHC classes than killed mycoplasmas. This ability of mycoplasmas to upregulate the expression of MHC molecules may result in increased ability of the macrophages to present self antigens to autoreactive T cells. This mechanism may be a contributing factor in the pathogenesis of disease induced by these organisms.

Short Communication

Autoimmune responses associated with mycoplasmal infection, although overlooked, have been known for some time (Biberfield, 1985; Baumgartner and Nicolet, 1984;

Kirchhoff et al, 1989; Runge et al, 1990; Rosendal, 1994). We have recently demonstrated the occurrence of autoantibodies to cilia in the sera of lambs with a respiratory disease 181 associated with mycoplasmal infection (Chapter 6 in this dissertation). These autoantibodies, based on absorption studies, seemed not to be cross-reactive with antigens of several common bacterial pathogens, including M. ovipneumoniae and M. arginini, isolated from the respiratory tract of these lambs. Thus, the possibility that the production of these autoantibodies could arise as a consequence of immunomodulatory effects induced by the mycoplasmal infection was raised.

Studies on infection and autoimmunity have emphasized antigenic epitopes shared by the pathogen and the host, and polyclonal B-cell activation, as the possible mechanisms contributing to inamunopathological responses. Alternatively, the possible induction of an autoimmune response as a consequence of aberrant expression of MHC molecules on host cells has been suggested. In support of this theory, several mycoplasma species have been reported to upregulate the expression of MHC molecules on macrophages (Stuart et al.,

1989; Stuart et al., 1990) and were proposed to be responsible for the generation of autoimmime pathology.

The purpose of this smdy was to investigate the effect of M. ovipneumoniae and

M. arginini, isolated from lambs with a respiratory disease, on the expression of MHC molecules on normal sheep alveolar macrophages. Findings might explain a possible immimological relationship between mycoplasmal infection in lambs and the production of the autoantibodies to cilia.

M. ovipneumoniae and M. arginini, isolated from lung tissues of a lamb with respiratory disease, were grown in modified Friis broth medium (Khan, 1993). Cultures 182 were harvested and washed twice in phosphate buffered saline (PBS, pH 7.4) by centrifugation at 10,000 g for 30 minutes at 4°C. Washed organisms were resuspended in

PBS and aliquots were stored at -70°C.

Details of methods for collection and culturing sheep alveolar macrophages have been fully described elsewhere (Niang, 1992). Briefly, bronchial lavage of whole normal lungs obtained from freshly slaughtered Iambs was performed by infusing 500 ml of sterile warm

PBS containing heparin and antibiotics into the lungs through the trachea. Lungs were free of mycoplasmas as attested by bacterial culture of tracheal swabs taken deep into the trachea just before the lavage. The obtained lavage fluid was filtered through a double layer of cheesecloth, washed by centrifugation, and cultured in flasks in RPMI1640 medium

(Sigma Chemical Co., St Louis, MO) supplemented with 10% fetal calf serum (FCS) and antibiotics. The flasks were incubated overnight at 37C in a 5% CO2 himudified atmosphere to allow macrophages to adhere. The medium was removed from the flasks to eliminate nonadherent cells. Fresh mediimi was added and flasks were reincubated for an additional 24 hours. At that time, the adherent macrophages were harvested from the flasks by using 0.2% trypsin (Difco Labs., Detroit, MI). The number of macrophages was coimted in a hemocytometer and their viability when checked by trypan blue exclusion was greater that 95%.

Two himdred ^1 of the macrophage suspensions in RPMI 1640 medium without antibiotics at a concentration of about 1 X 10^ cells per ml were placed in each well of a

96-weIl, flat bottom sterile microliter plates (Coming Glass Works, Coming, NY) in the 183 presence or absence of live or heat killed preparations of each mycoplasma to give a macrophagermycoplasma ratio of about 1:50. After an incubation of 1 or 2 days, the medium was removed from the wells and macrophages were fixed with cold acetone for 5 minutes and washed 3 times (5 minutes each) with warm Hanks balanced salt solution without phenol red (HBSS, Sigma Chemical Co., St. Louis, MO) containing 5% FCS. All subsequent washes were performed in this manner, and all dilutions were made in the same

HBSS.

For the fluorescence assay, 100 /zl of anti-mouse class I and class U monoclonal antibodies species cross-reactive (VMRD, Inc., Pullman, WA), each diluted at 1:100, were added to an individual well. After an incubation of 90 minutes at 37'C, the plates were washed, and labelled with 100 lA of either fluorescein-conjugated (BCirkegaard & Perry

Laboratories Inc., Gaithersburg, MD) or R-phycoerythrin-conjugated (Sigma Bio Sciences,

St. Louis, MO), goat anti-mouse antibody (IgG gamma-chain specific) diluted at 1:50 and

1:11, respectively. After incubation and washing, the plates were read in a fluorometric microtiter plate reader (Cambridge Technology Inc., Watertown, MA) at appropriate excitation and emission filter wavelengths for each fluorochrome. Results were reported as fluorescent values. Comparisons were made between values of unstimulated control macrophages and those of mycoplasma-induced macrophages, or values of live mycoplasma-induced macrophages and those of heat killed mycoplasma-induced macrophages. To rule out nonspecific binding, several background control wells were included in each plate. These consisted of wells with mycoplasmas or unstimulated 184

macrophages, reacted with each of the secondary conjugated antibodies alone in the case of

the macrophages, or with each of the monoclonal antibodies and the secondary conjugates

in the case of the mycoplasmas.

Fluorescent values of the background control wells were quite similar to those of

empty wells in each plate, suggesting that nonspecific binding of the reagents was

reasonably minimal. Of the two conjugates used, R-phycoerythrin appeared to be the most

sensitive. Therefore, results were concentrated on the fluorescence meastirements with the

R-phycoerythrin conjugate. In a separate experiment, unstimulated macrophages were cultured on Teflon-coated microscope slides (Cel-line, Newfield, NJ), and monolayers

were subsequently reacted with the monoclonals and the fluorescein conjugate and examined under a fluorescence microscope. The number of fluorescent cells in the monolayers that had been reacted with the anti-class I antibodies was far greater than that in the monolayers reacted with the anti-class n antibodies (data not shown). This suggests that the level of MHC class I expression on normal sheep alveolar macrophages could be

higher than that of class n. Also, great variation in the intensity of the positively stained cells was noticed between cells reacted with each monoclonal, suggesting that the extent of expression of both classes may vary between individual cells.

Although there was a noticeable increase on the expression of both classes when

macrophages were stimulated by the mycoplasmas, the rate of induction differed markedly depending on the mycoplasma species (Tables 1 and 2). Live M. ovipneumoniae organisms

induced a rapid increase in both class I and class n molecules. That rise was more than a 185

onefold increase at day 1 as compared with levels of expression on respective unstimulated

macrophages. By day 2, however, expression was decreased, in particular for class I, to a

level below that of unstimulated controls. In contrast, mayitnal induction of both MHC

classes by M. arginini was at day 2. Also, there was an indication that live M.

ovipneumoniae and M. arginini organisms, at all times, were more effective in inducing

expression of both classes than killed organisms.

The data in this report on the ability of M. ovipneumoniae and M. arginini to

positively influence MHC molecule expression on sheep alveolar macrophages could account for our previously observed prevalence of autoantibodies to cilia that develop during the course of a respiratory disease in lambs infected with these organisms (Chapter

6 in this dissertation). That production of these autoantibodies was the result of molecular mimicry or nonspecific polyclonal B-cell activation was imlikely. This was based on our failure to absorb these autoantibodies from sera with antigens of several common bacterial pathogens of the sheep respiratory tract including the mycoplasmas and the fact that they predominantly belong to the IgG class. It is likely, therefore, that their production may arise as a consequence of immunomodulatory effects induced by the mycoplasmal infection as demonstrated in this study.

A direct relation between the increased ability of macrophages to present antigen and an increased level of MHC molecule expression on these macrophages has been reported

(Ziegler et al., 1984). However, defective macrophage antigen presentation associated with the reduced expression of MHC class U antigens has also been reported (Ayala et al.. 186

1990). In view of this, it could be argued that uncontrolled expression of these molecules on antigen presenting cells could also lead to autoantigen presentation as suggested in the present study. The in vitro upregulation of MHC molecules on macrophages by several mycoplasmas including M. arginini, M. arthritidis, M. pulmonis, M. hominis and

M. fermentans, has been reported and suggested to be responsible of the generation of autormmime pathology (Stuart et al., 1989; Stuart et al., 1990).

Based on the aforementioned reports, and on our previous experiments (Chapters 3-6 in this dissertation) and on the results presented herein, we suggest the following pathologic events that lead to development of immunopathological responses observed in lambs in association with mycoplasmal infection, in particular M. ovipneumoniae. These include massive colonization of respiratory tract of young lambs during their very early age by the mycoplasmas, probably with the aid of a polysaccharide capsule. It is assumed that shortly after infection, the toxicity of hydrogen peroxide produced by the organism causes ciliostasis with massive local cellular necrosis. Also, there is a concomitant activation of alveolar macrophages and increased in the expression of MHC molecules on these macrophages. Since activated macrophages are not specific in their phagocytic functions, in their attempt to eliminate the mycoplasmas they would take up local damaged tissues. As a result of the high level of MHC expression on these cells, self antigens are presented to autoreactive T lymphocytes. The logical consequences of all this would be production of auto-reactive antibodies leading to inflammation. 187

References

Baumgartner, A. and Nicolet, J. 1984. Detection of antibodies against membrane- glycoproteins of swine erythrocytes after experimental infection with Mycoplasma hyopneumoniae. Vet. Immunol. Immunopathol., 7: 73-79.

Biberfeld, G. 1985. Infection sequelae and autoimmune reactions in Mycoplasma pneumoniae infection. In: S. Razin and F.M. Barile (editors). The mycoplasmas, vol. IV. Academic Press, Inc., New York. pp. 293-311. iChan, M.A. 1993. Respiratory infection of Iambs 'Vi'vih. Mycoplasma ovipneumoniae. Ph.D. Thesis, Iowa State University, Ames Iowa, USA.

Kirchhoff, H., Binder, A., Runge, M., Meir, B., Jacobs, R. and Busche, K. 1989. Pathogenetic mechanisms in the Mycoplasma arthritidis polyarthritis of rats. Rheimiatol. Int., 9: 193-1996.

Neu, N., Craig, S.W., Rose, N.R., Alvarez, F. andBeisel, K.W. 1987. Coxsachievirus induced myocarditis in mice; cardiac myosin autoantibodies do not cross-react with virus. Clin. Exp. Immun., 69: 566-574.

Nicholas, R.A. and Bashiruddin J.B. 1995. Mycoplasma mycoides subspecies mycoides (Small Colony Variant): the agent of contagious bovine pleuropneumonia and member of the "Mycoplasma mycoides cluster". J. Comp. Path., 113: 1-27.

Roberts, D.H. and Little, T.W. A. 1970. An autoimmune response associated with porcine enzootic pneumonia. Vet. Rec., 80: 328.

Runge, M., Binder, A., Gurr, E., Fischer, M., Meir, J. and BCirchhoff, H. 1990. Epitope- sharing between Mycoplasma arthritidis and chondrocytes demonstrated by monoclonal antibodies. Int. J. Med. Microbiol., 273: 531-538.

Smart, P.M., Cassell, G.H. and Woodward, J.G. 1990. Mycoplasma induction of class II MHC antigens, possible role in autoimmunity. Zbl. Bakt., 20 (SuppL): 575S-577S.

Stuart, P.M., Cassell, G.H. and Woodward, J.G. 1990. Induction of class n MHC antigen expression in macrophages by Mycoplasma species. J. Immunol., 142: 3392-3399.

Suter, M., Kobisch, M. and Nicolet, J. 1985. Stimulation of immunoglobulin-containing cells and isotype-specific antibody response in experimental Mycoplasma 188

hyopneumoniae infection in specific-pathogen -firee pigs. Infect. Immun., 49: 615-620.

Niang, M. 1992. Interaction of Mycoplasma ovipneumoniae with sheep alveolar macrophages. MS Thesis, Iowa State University, Ames Iowa. pp. 102.

Ziegler, H.K., Staffileno, L.K., and Wentworth, P. 1984. Modulation of macrophage la- expression by lipopolysaccharide. I. Induction of la expression in vivo. J. Immimol. 133:1825-1835. 189

Table 1. Effects of Mycoplasma ovipneumoniae (MO) and Mycoplasma arginini (MA) on expression of MHC class n on sheep alveolar macrophages.

Ruorescence values Day Treatment Experiment 1 Experiment 2 Experiment 3 Average* Control 729 634 776 713.000 MO-Live 2163 1824 2133 2040.000 MO-HK 1419 1148 1409 1325.333 MA-Live 1455 1400 1472 1442.333 MA-HK 1274 1114 1291 1226.333

Control 358 288 506 384.000 MO-Live 643 493 719 618.333 MO-HK 388 361 486 411.667 MA-Live 1156 948 706 936.667 MA-HK 885 681 566 710.667

Macrophages were cultured alone (control), or in the presence of live or heat killed (HK) mycoplasma organisms. After 1 or 2 days incubation, macrophages were reacted with anti-MHC class n monoclonal antibodies followed by labeling with R-phycoerythrin conjugated second antibody. * Average values of the 3 separate experiments. 190

Table 2. Effects of Mycoplasma ovipneumoniae (MO) and Mycoplasma arginim (MA) on expression of MHC class I on sheep alveolar macrophages.

Ruorescence values Day Treatment Experiment 1 Experiment 2 Experiment 3 Average* Control 2094 1726 2023 1947.667 MO-Live 3972 3055 4046 3691.000 MO-HK 2377 1945 2466 2262.667 MA-Live 2432 1786 2477 2231.667 MA-HK 2173 1816 2270 2086.333

Control 1324 959 1197 1160.000 MO-Live 1189 942 1099 1076.667 MO-HK 897 721 1079 899.000 MA-Live 2018 1563 1230 1603.667 MA-HK 1560 1274 1574 1469.333

Macrophages were cultured alone (control), or in the presence of live or heat killed (HK) mycoplasma organisms. After 1 or 2 days mcubation, macrophages were reacted with anti-class I monoclonal antibodies followed by labeling with R-phycoerythrin conjugated second antibody. * Average values of the 3 separate experiments. 191

APPENDIX B. ASSOCIATED OBSERVATIONS; TYPE I HVPERSENSmVITY IN LAMBS WITH COUGHING SYNDROME

Summary

Lambs from two flocks with a chronic respiratory disease characterized by paroxysmal cough and rectal prolapses were tested for their skin reactivity to Mycoplasma ovipneumoniae (MO) and M. arginini (MA) antigens (Ags). There was a marked, immediate skin reaction to intradermal injection of MO Ag in many of the tested lambs.

Some of these positive lambs also reacted to MA Ag. Phosphate buffered saline (PBS) used as a negative control gave no skin reaction in any of the tested animals. In addition, simultaneous serological tests revealed low antibody levels against MO and MA. However, both agents could be routinely isolated from nasal swabs of the affected Iambs. There is reason to suspect that an immediate type hypersensitivity to MO Ag and perhaps to other allergens that develops in association with the mycoplasmal infection contributes to this coughing syndrome.

Brief Report

M. ovipneumoniae has been routinely recovered from the respiratory tract of sheep with respiratory disease worldwide. Recently, in our laboratory. Khan (1993) also consistently isolated both MO and MA from lambs with a coughing syndrome. This respiratory disease is a major problem in Iowa and neighboring states due to widespread 192

occurrence in young lanabs predisposing to rectal prolapses and secondary bacterial

pneumonia, as well as reducing the rate of gain.

The mechanisms of pathogenesis of MO infection in sheep still remain unclear.

However, we believe that an immediate type hypersensitivity to the organism contributes to the clinical sjoidrome. This was based on reports of temporary abatement of clinical signs in lambs treated with antihistamines (Kaeberle, personal communication) and also demonstration of respiratory allergy in association with mycoplasmal infection in people

(Yano et al., 1994). Production of IgE antibodies specific for the allergen dimng the infection is responsible for the allergic reactions involving mast cells or basophils.

Diagnosis of an allergy is usually based upon detection of specific IgE antibodies or immediate skin reactions following intradermal injection of the allergen.

Since an antibody specific for sheep IgE is not available the only suitable means of allergy diagnosis in sheep is skin testing. This is a classical test that has been widely used for indirect demonstration of IgE antibodies specific for allergens. Our objective was to investigate the immediate skin reactivity of lambs with coughing syndrome to MO and MA

Ags as a demonstration of allergic respiratory disease in these animals.

Coughing lambs in the Iowa State University McNay sheep research farm or purchased from the same flock (Table 1) and housed in our facilities were utilized for this experimentation. These lambs were approximately 10 to 12 weeks of age and of mixed breeding. As Ags, cultures of MO and MA organisms were washed, sonicated and diluted to 5 ixg protein/ml. Lambs were injected intradermal on the inside of the thigh with 0.1 ml 193 of each Ag preparation and PBS as a control preparation. To improve the observation of the skin reaction, some animals were injected intravenously with 5 ml of a suspension of

0.5% Evans blue just before testing. The sites were observed for at least 30 minutes for development of blue stained wheals indicative of positive skin reactions. In lambs which were not injected with Evans blue, development of a wheal and flare response (swelling, redness) was considered as a positive skin reaction. Reactions were graded on the basis of size as a moderate reaction of 5-10 mm diameter (1 +) and a strong reaction of 10-25 mm diameter (2+).

Also, sera collected from the lambs were tested with an ELISA using sodium dodecyl sulfate (SDS)-treated MO and MA organisms and capsular material of MO as Ags.

The results of the skin tests on animals originating at McNay farm are shown in

Table 1. There was an immediate skin reaction in about 56 % (19/34) of the lambs injected with MO Ag, whereas only 29 % (10/34) reacted to MA Ag. All the lambs that were positive to MA Ag reacted also to MO Ag. This response usually began within 5 minutes after injection and reached its maximum within 30 minutes. Although some of the wheals lasted for more than 6 hours, no induration at the site of injection was observed at 48 hours post-injection indicating absence of delayed hypersensitivity.

ELISA tests conducted on serum of animals in group B demonstrated low circulating antibody levels in many of the tested animals . However, this lack of response was more evident against MO capsular Ag (average mean OD values = 0.385, range 0.152 to 0.785) 194 and MO SDS treated Ag (average mean OD values = 0.385, range 0.135 to 0.863) than

MA Ag (average mean OD values = 0.456, range 0.054 to 1.026) (data not shown).

Sheep from a second flock of purebred Suffolks were also skin-tested. Strong reactions to MO AG were observed in some coughing lambs of approximately 10 to 12 weeks of age and also in some yearling ewes that had apparently recovered from the clinical disease.

Results of this stody indicate that an unmediate type of hypersensitivity specific for

MO can develop during infection of lambs with this microorganism. The reaction is apparently specific, since not all the Iambs injected showed positive reaction, the resultant wheals in various lambs were not always equal in size and PBS injection induced no reaction. Whether or not the responses to MA are specific is unclear since lambs responding to Ag from this microorganism also responded to MO Ag. Therefore, the possibility of cross reactivity exists.

Skin testing has been routinely used to demonstrate allergic responses of humans and animals to allergens from microorganisms (Yano at al., 1994; Chen et al., 1990; Gershwin and Craig, 1984). The characteristics of the response observed in this smdy are quite similar to those previously described in sheep hypersensitivity induced by Ascaris suum extract (Chen et al., 1990) and in the human allergic reaction to M. pneumoniae (Yano et al., 1994), including formation of raised blebs immediately after injection with no late reaction. 195

The relatively low levels of IgG specific antibodies to both MO and MA during the

early stages of the clinical disease, as seen here, is likely an indication of a predominant

IgE response. In such a situation, this IgE may play a role in the development of allergic airway reactions in these lambs involving release of histamine and serotonin mainly from

mast cells and basophils as has been suggested in asthmatic people infected with M. pneumoniae (Yano et al., 1994). Unfortunately however, due to lack of anti-sheep IgE we cannot at the moment measure the level of IgE antibodies in these sheep.

Selected References

1. Chen, W. et al. 1990. N. Z. Vet. J. 38:57-61.

2. Kaeberle, M.L. 1995. Unpublished observations.

3. Khan, M.A. 1993. PhD. Thesis, Iowa State University, Ames, lA. 226pp.

4. Gershwin, L.J. and Craig, L.O. 1984. Am. J. Vet. Res. 10:2135-2142.

5. Yano, T. et al. 1994. Am. J. Respir. Crit. Care Med. 149:1348-1353. 196

Table 1. Skin-test reactions to Mycoplasma ovipneumoniae (MO) and Mycoplasma argimni (MA) antigens.

Lamb Group #Tested #Reacting to MO #Reacting to MA #Reacting to PBS 2-i-' 1+" 2+ 14- A 19 5 7 1 5 0 B 15 2 5 1 3 0

a- 2+, 10-20 mm diameter, b- 1 +, 5-10 mm diameter. 197

LITERATURE CITED

Absolom, D.R. 1988. The role of bacterial hydrophobicity in infection: bacterial adhesion and phagocytic ingestion. Can. J. Microbiol. 34:287-298.

Adegboye, D.S. 1978. A review of mycoplasma-induced immunosuppression. Br. Vet. J. 134:556-560.

Adler, K.B. 1986. Mucin secretion by explants of respiratory in vitro. In Schiff, L.J. (ed.). In Vitro Models of Respiratory Epithelium. CRC Press, Inc., Boca Raton, Horida. pp. 27-51.

Al-E[aisi, A., and Alley, M.R. 1983. Electron microscopic studies of the interaction between ovine alveolar macrophages and Mycoplasma ovipneumoniae in vitro. Vet. Microbiol. 8:571-584.

Alley, M.R., and Clarke, J.K. 1977. The influence of microorganisms on the severity of lesions in chronic ovine pneumonia. N. Z. Vet. J. 27:217-220.

Alley, M.R. 1975. The bacterial flora of the respiratory tract of normal and pneumonic sheep. N. Z. Vet. J. 23:113-118.

Almagor M., Kahane I., and Yatziv S. 1984. Role of superoxide anion in host cell injury by Mycoplasma pneumoniae infection. J. Clin. Invest. 73:842-847.

Almagor, M., Yatziv, S., and Kahane, I. 1983. Inhibition of host cell catalase by Mycoplasma pneumoniae: a possible mechanism for cell injury. Infect. Immun. 41:251- 256.

Almeida, R.A. 1991. Pathogenesis of respiratory infections caused by Mycoplasma dispar. Ph.D. Thesis, Iowa State University, Ames. pp. 140.

Almeida, R.A. 1988. Smdies on Mycoplasma dispar capsular material. MS Thesis, Iowa State University, Ames. pp. 82.

Almeida, R.A., and Rosenbusch, R.F. 1994. Impaired tracheobronchial clearance of bacteria in calves infected -fiiih. Mycoplasma dispar. J. Vet. Med. Bacterial. 41:473-482.

Almeida, R.A., and Rosenbusch R.F. 1991. Capsule-like surface material of Mycoplasma dispar induced by in vitro growth in culture with bovine cells is antigenically related to similar structures expressed in vivo. Infect. Inomun. 59:3119-3125. 198

Almeida, R.A., Wannemuehler, M.J., and Rosenbusch, R.F. 1992. Interaction of Mycoplasma dispar with bovine alveolar macrophages. Infect. Tmmnn 60:2914-2919.

Araake, M. 1982. Comparison of ciliostasis by mycoplasmas in motise and chicken tracheal organ cultures. Microbiol. Immunol. 26:1-14.

Athamna, A., Ofeck, I., Keisrari, Y., Markowitz, S., Dutton, G.G.S., and Sharon, N. 1991. Lectinophagocytosis of encapsulated Klebsiella pneumoniae mediated by surface lectins of guinea pig alveolar macrophages and human monoc3^-derived macrophages. Infect. Immun. 50:1673-1682.

Bakke, T., and Nostvold, S. 1982. An investigation of ovine pneumonia in four herds from Central Norway. I. Prevalence of pneumoniae and microbiological findings. Acta. Vet. Scand. 23: 248-258.

Bakke, T. 1982. The occurrence of mycoplasmas and bacteria in limgs from sheep in Southern Norway. Acta. Vet. Scand. 23: 235-247.

Bansal, P., Adegboye, D.S., and Rosenbusch, F.R. 1995. Inmiime responses to the capsular polysaccharide of Mycoplasma dispar in calves and mice. Comp. Immunol. Microbiol. Infect. Dis. 18:259-268.

Bansal, P. 1994. Chemical characterization of Mycoplasma dispar capsule and a study of its immune response in calves and mice. MS Thesis, Iowa State University, Ames. pp. 105.

Barile, M.F., and Leventhal, B.G. 1968. Possible mechanism for mycoplasma inhibition of lymphocyte transformation induced by phytohaemagglutinin. Nature (London) 219:751- 752.

Baselga, R., Albizu, I., and Amorena, B. 1994. Staphylococcus aureus capsule and slime as virulence factors in ruminant mastitis: A review. Vet. Microbiol. 39:195-204.

Baseman, K.B., Morrison-Plummer, J., Drouillard, D., Tryon, V.V., and Holt, S.C. 1987. Identification of a 32 kDa protein of mycoplasma pnemnoniae associated with hemadsorption. Isr. J. Med. Sci. 23:474-479.

Basemen, J.B., Cole, R.M., Krause, D.C., and Leith, D.K. 1982. Molecular basis for cytadsorption of Mycoplasma pneumoniae. J. Bact. 151:1514-1522. 199

Baumgartner, A., and Nicolet, J. 1984. Detection of antibodies against membrane- glycoproteins of swine erythrocytes after experimental infection with Mycoplasma hyopneumoniae. Vet. Immunol. Immmiopathol. 7: 73-79.

Bayer, M.E. 1990. Visualization of the bacterial polysaccharide capsule. Curr. Top. Microbiol. Immun. 150:129-157.

Belanger, M., Dubreuil, D., Harel, J., Girard, C., and Jacques, M. 1990. Role of lipopolysaccharides in adherence of Actinobacillus pleuropneumoniae to porcine tracheal rings. Infect. Inmmn. 58:3523-3530.

Bennett, R.H., and Jasper, D.E. 1977. Immunosuppression of humoral and cell-mediated responses in calves associated with inoculation of Mycoplasma bovis. Am. J. Vet. Res. 38:1731-1738.

Bereiter, M., Young, T.F., Joo, H.S., and Ross, R.F. 1990. Evaluation of the ELISA and comparison to the complement fixation test and radial immimodiffusion enzyme assay for detection of antibodies against Mycoplasma hyopneumoniae in swine serum. Vet. Microbiol. 25:177-192.

Bermudez, V., Miller, R., Rosendal, S., and Johnson, W. 1990. In vitro cytopathic effect of Mycoplasma equigenitaliwn on the equine uterine mbe. Zbl. Bakt. 20 (suppl.):419-428.

Bermudez, V.M., Miller, R.B., Rosendal, S., Fernando, A.M., Johson, W.H., and O'Brien, P.J. 1992. Measurement of the cytotoxic effects of different strains of Mycoplasma equigenitalium on the equine uterine mbe using a calmodulin assay. Can. J. Vet. Res. 56:331-338.

Biberfeld, O.E., and Sterner, G. 1976. Tuberciilin anergy in patients with Mycoplasma pneumoniae mfection. Scand. J. infect. Dis. 8:71-73.

Biberfeld, G. 1985. Infection sequelae and autoimmime reactions m Mycoplasma pneumoniae infection. In: Razin, S., and Barile, M.F. (eds.). The mycoplasmas, vol. 4. Academic Press, Inc., New York. pp. 293-311.

Bisset, L.R. 1994. Molecular mimicry in the pathogenesis of AIDS: the HIV/MHC/mycoplasma triangle. Med. Hypotheses. 43:388-396.

Bottazzo, G.F., Todd, I., Mirakian, R., Belfiore, A., and Pujol-Borrell, R. 1986. Organ-specific autoimmunity: A 1986 overview. Inomimol. Rev. 94: 137-169. 200

Brennan, P.C., and Feinstein, R.N. 1969. Relationship of hydrogen peroxide production by Mycoplasma pulmonis to virulence for catalase-deficient mice. J. Bacterial. 98:1036- 1040.

Brogden, K.A., Rose, D., Cutlip, R.C., Lehmkuhl, H.D., and Tully, J.G. 1988. Isolation and Identification of mycoplasmas firom the nasal cavity of sheep. Am J. Vet. Res. 49:1669-1672.

Buddie, B.M., Herceg, M., and Davies, D.H. 1984. Comparison of virulence of ovine respiratory mycoplasmas in the mouse mammary gland. Vet. Microbiol. 9:367-374.

Buman, W.J., and Martin n, W.J. 1986. Oxidant-mediated ciliary dysfunction; possible role in airway disease. Chest. 89:410-413.

Buttery, S.H. 1970. Hapten inhibition of the reaction between Mycoplasma mycoides polysaccharide and bovine antisera. Immunochemistry. 7:305-310.

Buttery, S.H., Lloyd, L.C., and Titchen, D.A. 1976. Acute respiratory, circulatory from Mycoplasma mycoides subsp. mycoides. J. Med. Microbiol. 9:379-391.

Carmichael, L.E., St. George, T.D., SuUivan, N.D., and Horsfall, N. 1972. Isolation, propagation and characterization smdies of an ovine mycoplasma responsible for proliferative interstitial pneumonia. Cornell Vet. 62:654-679.

Carusso, J.P., and Ross, R.F. 1990. Effects of Mycoplasma hyopneumoniae and Actinobacillus (haemophilus) pleuropneumoniae infections on alveolar macrophage fimctions in swine. Am. J. Vet. Res. 2:227-231.

Casseil, G.H., Clyde, W.A., and Davis, J.K. 1985. Mycoplasmal respiratory infections. In: Razin, S., and Barile, M.F. (eds.). The Mycoplasmas. Vol. 4. Academic Press, Inc., New York. pp. 65-106.

Casseil, G.H., and McGhee, J.R. 1975. Pathogenic and immunologic responses of nude mice following intranasal inoculation of Mycoplasma pulmonis. J. Retic. Soc. 18:426.

Casseil, H.G., Lindsey, J.R., and Baker, H.J. 1974. Immune response of pathogen-free mice inoculated intranasally with Mycoplasma pulmonis. J. Immunol. 112:124-136.

Chery, J.D., and Taylor-Robinson, D. 1970. Growth and pathogenesis of Mycoplasma mycoides var. capri in chicken embryo tracheal organ cultures. Infect. Immun. 4:431-438. 201

Clyde, W.A.Jr. 1963. Hemolysis in identifying Eaton's pleuropnemnnia like organisms. Science. 139:55.

Cohen, 6., and Somerson, NX. 1969. Glucose dependent secretion and destruction of hydrogen peroxide by Mycoplasma pneumoniae. J. Bacterial. 98:547-551.

Cole, B.C. 1991. The inmnmology of Mycoplasma arthritidis and its superantigen MAM. Curr. Top. Microbiol. Tmmimol. 174:107-119.

Cole, B.C., Noat, Y., Stanbridge, E.C., and Wise, K.S. 1985. Interactions of mycoplasmas and their products with lymphoid cells in vitro. In: Razin, S., and Barile, M.F. (eds.). The Mycoplasma. Vol. 4. Academic Press, New York. pp. 203-257.

Cole, B.C., and Atkin, C.L. 1991. The Mycoplasma arthritidis T-cell mitogen, MAM: A model superantigen. Immunol. Today. 12:271-276.

Collier, AM. 1979. Mycoplasma in organ cultures. In: TuUy, J.G., and Whitcomb, R.F.(eds.). The Mycoplasmas. Vol. 2. Academic Press, Inc., New York. pp. 475-493.

Collier, A.M., and Clyde, Jr. W.A. 1969. Mycoplasma pneumoniae disease pathogenesis studies in tracheal organ culture. Fed. Proc. 28:616.

Collier, A.M., and Baseman, J.B. 1973. Organ culture techniques with mycoplasmas. Ann. N. Y. Acad. Sci. 225:277-289.

Collier, A.M., Hu, P.C., and Clyde, W.A. 1985. The changing pathogenicity of Mycoplasma pneumoniae shared by other pathogenic mycoplasma species. Infect. Immun. 51:690-692.

Costerton, J.W., Cheng, K.J., Greesey, G.G., Ladd, T.I., Nickel, J.C., Dasgupta, M., and Marie, T.J. 1987. Bacterial biofilms in nature and disease. Ann. Review Microbiol. 41:435-464.

Davies, D.H., Jones, B.A.H., and Thurley, D.C. 1981. Infection of specific-pathogen- free lambs with parainfluenza virus type 3, Pasteurella haemolytica and Mycoplasma ovipneumoniae. Vet. Microbiol. 6:295-308.

Debey, M.C., and Ross, R.F. 1994. Ciliostasis and loss of cilia induced by Mycoplasma hyopneumoniae in porcine tracheal organ cultures. Infect. Immun. 62:5312-5318. 202

Davies, D.H. 1985. Aetiology of pneumoniae of young sheep. Prog. Vet. Microbiol. 1:229-249.

Debey, M.C. 1992. Pathogenic effects of Mycoplasma hyopnewnoniae on porcine tracheal organ cultures and neutrophils in vitro, and porcine respiratory goblet cells in vivo. Ph.D. Thesis, Iowa State University, Ames. pp. 237.

Denny, F.W., Taylor-Robinson, D., and Allison, A.C. 1972. The role of thymus- dependent immunity in Mycoplasma pulmonis infections of mice. J. Med. Microbiol. 5:327-336.

Dudman, W.F. 1977. The role of surface polysaccharides in natural environments. In: Sutherland, I.W. (ed.). Surface Carbohydrates of prokaryotic cells. Academic Press, New York. pp. 357-415.

Edwards, M.S., Kasper, D.L., Jennings, H.J., Baker, C.J., and Nicholson-Weller, A. 1982. Capsular sialic acid prevents activation of the alternative complement pathway by type in, group B Streptococci. J. Immunol. 128:1278-1283.

Eness, P.G., Kaeberle, M.L., and Hummel, D. 1990. Association of Mycoplasma ovipneumoniae with respiratory disease in feeder lambs. Beef-Sheep Research Report. Iowa State University, Ames. pp. 44-45.

Femald, G.W., Clyde, W.A., and Bienenstock, J. 1972. Immunoglobulin-containing cells in lungs of hamsters infected with Mycoplasma pneumoniae. J. Immunol. 108:1400- 1408.

Howard, C.J., and Gourlay, R.N. 1983. Immune response of calves following the inoculation of Mycoplasma dispar and Mycoplasma bovis. Vet. Microbiol. 8:45-56.

Finch, J.M., and Howard, C.J. 1990. Inhibitory effect of Mycoplasma dispar and Mycoplasma bovis on bovine immune responses in vitro. Intern. J. Med. Microbiol. 20(Suppl.):563-569.

Finne, J. 1982. Occurrence of unique polylialosyl carbohydrate units in glycoproteins of developing brain. J. Bacterial. Chemistry. 257:11966-11970.

Foggie, A., Jones, G.E., and Buxton, D. 1976. The experimental infection of specific pathogen free lambs with. Mycoplasma ovipneumoniae. Res. Vet. Sci. 21:28-35. 203

Foggie, A., and Angus, K.W. 1972. Observation on the distribution of Mycoplasma arginini as a respiratory tract infection in sheep and its pathogenicity for specific pathogen free lambs. Vet. Rec. 90:312-313.

Freudt, E.A. 1958. The pleuropneimionia group of organisms: morphology, biology and taxonomy. In: The Mycoplasmataceae. Munksgaard, Copenhagen.

Frick, O.L. 1986. Effect of respiratory and other virus infections of IgE immunomodulation. J. Allergy Clin. Immunol. 78:1013-1018.

Gabridge, M.G. 1984. Respiratory tract organ cultures to assay attachment and pathogenicity of mycoplasmas. Ann. Microbiol. 135A:33-38.

Gabridge, M.G. 1986. Differentiated respiratory epithelium as an in vitro model for mycoplasma infections. In: Schiff, L.J. (ed.). In Vitro Models of Respiratory Epithelium. CRC Press, Inc., Boca Roton, Florida, pp. 85-102.

Gabridge, M.G. 1982. A review of the morphological and biochemical features of the attachment process in infections -with Mycoplasma pneumoniae. Review Infect. Dis. 4(S):179-184.

Gabridge, G., Chandler, D.K.F., and Daniel, M.J. 1985. Pathogenicity factors in mycoplasmas and spiroplasmas. In: Razin, S., and Barile, M.F. (eds.). The Mycoplasmas. Vol. 4. Academic Press, Inc., New York. pp. 313-351.

Gabridge, M.G., Johnson, C.K., and Cameron, A.M. 1974. Cytotoxicity of Mycoplasma pneumoniae membranes. Infect. Immun. 10:1127-1134.

Geary, S.J., and Walczack, E.M. 1983. Cytopathic effect of whole cells and purified membranes of Mycoplasma hyopneumoniae. Infect. Immun. 41:132-136.

Geary, S.J., Tourtelotte, M.E., and Cameron, J.A. 1981. Inflammatory toxins from Mycoplasma bovis: isolation and characterization. Science. 212:1032-1033.

Gilmour, J.S., Jones, G.E., and Rae, A.G. 1979. Experimental smdies of chronic pneumonia of sheep, comp. Immun. Microbiol. Infect. Dis. 1:285-293.

Ginsburg, H., and Nicolet, J. 1973. Extensive transformation of lymphocytes by a mycoplasma organism. Nature (London), New Biol. 246:143-146. 204

Gk>rski, F., and Bredt, W. 1977. Studies on the adherence mechanism of Mycoplasma pneumoniae. FEMS Microbiol. Lett. 1:265-267.

Gourlay, R.N., and Howard, C.J. 1982. Respiratory mycoplasmosis. Adv. Vet. Sci. Comp. Med. 23:289-332.

Gourlay, R.N., and Howard, C.J. 1982. Respiratory mycoplasmosis. Ad. Vet. Sci. Comp. Med. 26:289-332.

Gourlay, N.R., and Howard, C.J. 1978. Isolation and pathogenicity of mycoplasmas from the respiratory tract of calves. In: Martin, W.B. (ed.). Respiratory Diseases in Cattle. Martinus Nijhoff, Boston, pp. 295-304.

Gourlay, R.N., and Howard, C.J. 1978. Isolation and pathogenicity of mycoplasmas from the respiratory tract of calves. In: Martin, W.B. (ed.). Current Topics in Veterinary Medicine. Vol 3., Respiratory diseases in cattle. Martinus Niijoff, Boston, pp.295-304.

Green HI, F., and Hanson, M. 1973. Ultrastructure and capsule of Mycoplasma meleagridis. J. Bacterial. 116:1011-1018.

Greig, A.S. 1955. The isolation of Pleuropneumonia-like organisms from the respiratory tract of sheep. Can. J. Comp. Med. 265-271.

Hamdy, A.H., and Pounden, W.D. 1959. Experimental production of pneumonia in lambs. Am. J. Vet. Res. 20:78-83.

Howard, C.J., and Taylor, G. 1985. Immime responses to mycoplasma infections of the respiratory tract. Vet. Immun. Inmiunupathol. 10:3-23.

Howard, C.J. 1980. Variation in susceptibility of bovine mycoplasmas to killing by the alternative complement pathway in bovine serum. Immunol. 41:561-568.

Howard, C.J., and Taylor, G. 1983. Interaction of mycoplasmas and phagocytes. Yale J. Biol. Med. 56:643-648.

Howard, C.J., Thomas, L.H., and Parsons, K.R. 1987. Immune response of cattle to respiratory mycoplasmas. Vet. Immxm. Inmiunopathol. 17:401-412.

Howard, C.J., and Gourlay, R.N. 1983. Immune response of calves following the inoculation of Mycoplasma dispar and Mycoplasma bovis. Vet. Microbiol. 8:45-56. 205

Hu, P.C., Collier, A.M., and Baseman, J.B. 1976. Interaction of virulent Mycoplasma pneumoniae with hamster tracheal organ cultures. Infect. Inmiun. 14:217-224.

Hu, P.C., Collier, A.M., and Baseman, J.B. 1977. Surface parasitism by Mycoplasma pneumoniae of respiratory epithelium. J. Exp. Med. 145:1328-1343.

Hu, P.C., Cole, R.M., Huang, Y.S., Graham, J.A., Gardner, D.E., Collier, A.M., and Clyde, Jr., W.A. 1982. Mycoplasma pneumoniae infection: role of a smface protein in the attachment organelle. Science. 216:313-315.

Hudson, J.R., Buttery, S., and Cottew, G.S. 1967. Investigations into the influence of the galactan of Mycoplasma mycoides on experimental infection with the organism. J. Pathol. Bacterial. 94:257-273.

Inzana, T.J. 1990. Capsules and virulence in the HAP group of bacteria. Can. Vet. Res. 54:S22-S27. lonas, G., and Mew, A.J. 1985. Colonization of the respiratory tract of lambs by strains oi Mycoplasma ovipneumoniae. Vet. Microbiol. 10: 533-539.

Isenberg, H.D. 1988. Pathogenicity and virulence: Another view. Clin. Microbiol. Reviews. 1:40-53.

Jacobs, E., Stuhlert, A., Drews, M., Pumpe K., Schaefer, H.E., Kist, M., and Bredt, W. 1988. Host reactions to Mycoplasma pneumoniae infections in guinea pigs preimmunized systematically with the adhesion of this pathogen. Microb. Pathogen. 5:259- 265.

Jacobs, E., Drews, M., Stuhlert, A., Buttner, C., klein P.J., Kist, M., and Bredt, W. 1988. Immunological reaction of guinea-pigs following intranasal Mycoplasma pneumoniae infection and immimization with the 168 kDa adherence protein. J. Gen. Microbiol. 134:473-479.

Joiner, K.A., Brown, E.J., and Frank, M.M. 1984. Complement and bacteria: chemistry and biology in host defense. Ann. Reviews Immunol. 2:461-492.

Jones, G.E., Gilmour, J.S., and Rae, A.G. 1982a. The effects of different strains of Mycoplasma ovipneumoniae on specific pathogen-free and conventionally-reared lambs. J. Comp. Pathol. 95:477-487. 206

Jones, G.E., Gilmour, J.S., and Rae, A.G. 1982b. The effect of Mycoplasma ovipneumoniae and Pasteurella haemofytica on specific pathogen-free and conventionally- reared lambs. J. Comp. Pathol. 92:261-266.

Jones, G.E., Foggie, A., Mould, D.L., and Liyitt, S. 1976. The comparison and characterization of glycolytic mycoplasmas isolated from the respiratory tract of sheep. J. Med. Microbiol. 9:39-52.

Jones, G.E., Keir, W.A., and Gilmour, J.S. 1985. The pathogenicity of Mycoplasma ovipneumoniae and Mycoplasma arginini in ovine and caprine tracheal organ cultures. J. Comp. Path. 95:477^87.

Kaeberle, M.L., and Eness, P. 1988. Serologic analysis of pneumonia in feeder Iambs. Beef-Sheep Research Report. Iowa State University, Ames. pp. 54-55.

Kahane, I. 1984. In vitro smdies on the mechanism of adherence and pathogenicity of mycoplasmas. Isr. J. Med. Sci. 9:210-216.

Elahane, I., and Tully, G.J. 1976. Binding of plant lectins to mycoplasma cells and membranes. J. Bacterial. 128:1-7.

Kaklamanis, E., and Pavlatos, M. 1972. The immunosuppressive effect of mycoplasma infection. I. Effect on the htmaoral and cellular responses. Immunology. 22:695-702.

Kapatais-Zoumbos, K., Chandler, D.K.F., and Barile, M.F. 1985. Survey of immunoglobulin A protease activity among selected species of ureaplasma and mycoplasma: specificity for immxmoglobulin A. Infect. Immun. 47:704-709.

Kasper, D.L., Onderdonk, A.B., Reinap, B.G., and Lingberg, A.A. 1980. Variations of Bacteroides fragilis with in vitro passage: presence of an outer membrane-associated glycan and loss of capsular antigen. J. Infect. Dis. 142:750-756.

Kiian, M.A. 1993. Respiratory infection of lambs with Mycoplasma ovipneumoniae. Ph.D. Thesis, Iowa State University, Ames Iowa, USA. pp. 226.

Khan, A.R., Bengtson, B., and Latinen, L.A. 1990. Influence of on the ciliary activity and the contractile response of the guinea-pig trachea. Am. Rev. Respir. Dis. 141(Suppl.):A532. 207

Kirchhoff, H., Binder, A., Rimge, M., Meir, B., Jacobs, R., and Busche, K. 1989. Pathogenetic mechanisms in the Mycoplasma arthritidis polyarthritis of rats. Rheumatol. Int. 9:193-1996.

Kishima, M., and Ross, R.F. 1985. Suppressive effect of nonviable Mycoplasma hyopneimoniae on phytohemagglutin-induced transformation of swine lymphocytes. Am. J. Vet. Res. 46:2366-2368.

Kluve, B., Merrick, W.C., Stanbridge, E.J., and Gershman, H. 1981. Mycoplasmas induce collagenase in BALB/c 3T3 cells. Nature. 292:855-857.

Kobayashi, K., Salathe, M., Pratt. M.M., Cartagena, N.J., Soloni, Felix, Seyboid, Z.V., and Wanner, A. 1992. Mechanism of hydrogen peroxide-induced inhibition of sheep airway cilia. Am. J. Respir. Cell. Mol. Biol. 6:667-673.

Krause, D.C., and Baseman, J.B. 1982. Mycoplasma pneumoniae proteins that selectively bind to host cells. Infect, immun. 37:382-386.

Lai, W.C., Pakes, S.P., Owusu, I., and Wang, S. 1989. Mycoplasma pulmonis depresses humoral and cell-mediated responses in mice. Lab. Animal Sci. 39:11-15.

Levinsohn, S., Dykstra, M.J., Lin, M.Y., and Kleven S.H. 1986. Comparison of in vivo and in vitro methods for pathogenicity evaluation for Mycoplasma gallisepticum in respiratory infection. Avian. Path. 15:233-246.

Lipman, R.P., and Clyde, W.A. Jr. 1969. The interrelationship of virulence, cytadsorption, and peroxide formation'm. Mycoplasma pneumoniae. Proc. Soc. Exp. Biol. Med. 131:1163-1167.

Luft, J.H. 1971. Rutheniirai Red and Violet. I. Chemistry, purification, methods of use for electron microscopy and mechanism of action. Anat. Rec. 171:347-368.

Lui, C., Jayanetra, P., Voth, D.W., Muangmanee, L., and Cho, C.T. 1972. Potentiating effect of Mycoplasma pneumoniae infection on the development of pneumococcal septicemia in hamsters. J. Infect. Dis. 125:603-612.

Malone, F.E., McCullough., S.T., McCuUough, M.F., Ball, H.J., O'Hagan, J., and Neil, S.D. 1988. Infectious agents in respiratory disease of housed, fattening Iambs in Northern Ireland. Vet Rec. 122:203-207. 208

Marcus, H. and , Baker, N.R. 1985. Quantitation of adherence of mucoid and nonmucoid Pseudomonas aeruginosa to hamster tracheal epithelium. Infect. Immun. 47:723-729.

Maxon, E.R., and Kroll, J.S. 1990. The role of bacterial polysaccharide capsules as virulence factors. Curr. Top. Microbiol. Immunol. 150:65-95.

McArthur, H.A., and Ceri, H. 1983. Interaction of a rat lug lectin with exopolysaccharides of Pseudomonas aeruginosa. Infect. Immun. 42:574-578.

Messier, S.J. 1986. Interactions between Mycoplasma hyopneumoniae and the porcine immune system. Ph.D. Thesis, Iowa State University, Ames. pp. 146.

Messier, S.J., Ross, R.F., and Paul, P.S. 1990. Humoral and cellular immune responses of pigs inoculated with Mycoplasma hyopneumoniae. Am. J. Vet. Res. 51:52-58.

Minion, F.C., and Rosenbusch, R.F. 1993. Extramembranous structure in mycoplasmas. In: Rotten, S., andKahane, I. (eds.). Subcellular biochemistry. Vol. 20. Plenum press. New York. pp. 189-201.

Minion, F.C., Cassell, G.H., Pnini, S., Kahane, I. 1984. Multiphasic interaction of Mycoplasma pulmonis with erythrocytes defined by adherence and hemagglutination. Infect. Immun. 44:394-400.

Nagayama, Y., Sakurai, N., Kojima, S., and Funabashi, S. 1987. Total and specific IgE responses in the acute and recovery phases of respiratory infections in children. J. Asthma. 24:159-166.

Navia, J.L., Riesenfeld, J., Vann, W.F., Lindahl, U., and Roden, L. 1983. Assay of N-acetylheparosan deacetylase with a capsular polysaccharide firom Escherichia coli K5 as substrate. Analytic Biochemistry. 135:134-140.

NeyroUes, O., Brenner, C., Prevost, M.C., Fontaine, T., Montagnier, L., and Blanchard, A. 1996. Isolation and characterization of a novel capsule-like material from Mycoplasma penetrans. 11th International Congress of the International Organization for Mycoplasmology. Abst. No OS 12-2. pp. 366. Orlando, Rorida.

Niang, M. 1992. Interaction of Mycoplasma ovipneumoniae with sheep alveolar macrophages. M.S. Thesis, Iowa State University, Ames Iowa, USA. pp. 102. 209

Nicholas, R.A., and Bashiruddin, J.B. 1995. Mycoplasma mycoides subspecies mycoides (Small Colony Variant): the agent of contagious bovine pleuropneumonia and member of the "Mycoplasma mycoides cluster". J. Comp. Path. 113: 1-27.

NostYoId, S., and Bakke, T. 1982. An investigation of ovine pneumonia in four herds from Central Norway, n. Relation between pathomorphology and presence of micro­ organisms. Acta Vet. Scand. 23:259-274.

Pfeffer, A., Thurley, D.C., Boyes, B.V., and Davies, G.B. 1983. The prevalence and microbiology of pneumonia in a flock of lambs. N. Z. Vet. J. 31:196-202.

Pijoan, C., and Trigo, F. 1990. Bacterial adhesion to mucosal surfaces with special reference to Pasteurella muUocida isolates from atrophic rhinitis. Can. J. Vet. Res. 54:S16-S21.

Ramphal, R., and Pier, G.B. 1985. Role of Pseudomonas aeruginosa mucoid exopolysaccharide in adherence to tracheal cells. Infect. Lmmim. 47:1-4.

Razin, S. 1978. The mycoplasmas. Microbiol. Rev. 42:414-470.

Razin, S. 1985. Mycoplasma adherence. In: Razin, S., and Barile, M.F. (eds.). The Mycoplasmas. Vol. 4. Academic Press, Inc., New York. pp. 161-202.

Roberson, M.M., Ceri, H., Shadle, P.J., andBarondes, S.H. 1981. Heparin inhibitable lectins: marked similarities in chicken and rat. J. Supramolec. Struct. Cell. Biochem. 15:395-402.

Roberts, D.H., and Little, T.W.A. 1970. An autoimmune response associated with porcine enzootic pneumonia. Vet. Rec. 80: 328.

Robertson, J.A., Stemler, M.E., and Stemke, G.W. 1984. Immunoglobulin A protease activity of Ureaplasma ureatyticum. J. Clin. Microbiol. 19:255-258.

Robertson, J., and Smook, E. 1976. Cytochemical evidence of extramembranous carbohydrates on Ureaplasma ureatyticum (T-Strain Mycoplasma). J. Bacterial. 128:658- 660.

Rosenbusch, R.F. 1992. Mucosal colonization by capsulated mycoplasmas. 9th International Congress of the International Organization for Mycoplasmology, Ames, Iowa. Abstract No W9/3. pp. 60. 210

Rosenbusch, R.F., and Minion, C.F. 1992. Cell envelope; morphology and biochemistry. In: Maniloff, J., McElhaney, R.N., Finch, L.R., and Baseman, J. (eds.). Mycoplasmas, molecular biology and pathogenesis. ASM, Washington, DC. pp. 73-77.

Rosendal, S. 1993. Mycoplasma. In: Carlton, L.G., and Thoen, C.O. (eds.). Pathogenesis of Bacterial Infections in Animals. Second Ed. Iowa State University Press, Ames. pp. 297-311.

Rosendai, S. 1994. Ovine and caprine mycoplasmas. In: Whitford, H., Rosenbusch, R.F., and Lauerman, L.H. (eds.). Mycoplasmosis in Animals: Laboratory Diagnosis. Iowa State University Press, Ames, Iowa. pp. 84-96.

Rimge, M., Binder, A., Gurr, E., Fischer, M., Meir, J., and Kirchhoff, H. 1990. Epitope-sharing between Mycoplasma arthritidis and chondrocytes demonstrated by monoclonal antibodies. Int. J. Med. Microbiol. 273:531-538.

Rurangirwa, F.R., McGuire, T.C., Magnuson, N.S., Kibor, A., and Chema, S. 1987. Composition of a polysaccharide from mycoplasma (F-38) recognized by antibodies from goats with contagious pleuropneumonia. Res. Vet. Sci. 42:175-178.

Sabato, A.R., Cooper, D.M., and Thong, Y.H. 1981. Transitory depression of immune function following Mycoplasma pneumoniae infection in children. Pediatr. Res. 15:813- 816.

Sabin, A.B. 1938. Identification of the filtrable transmissible nemrotoxic agent isolated from toxoplasma-infected tissues as a new pleuropneumonia-like microbe. Science. 88:575- 576.

Saylers, A.A., and Whitt, D.D. 1994. Virulence factors that promote colonization. In: Bacterial Pathogenesis: A Molecular Approach. ASM Press, Washington, D.C. pp.30-46.

Simberkoff, M.S., and Elbach, P. 1971. The interaction in vitro between polymorphonuclear leukocytes and mycoplasma. J. Exp. Med. 134:1417-1430.

Simon, R.H., Scoggin, C.H., and Patterson, D. 1981. Hydrogen peroxide causes the fatal injury to himian fibroblasts exposed to oxygen radicals. J. Biol. Chem. 256:7181- 7186.

Smith, S.C., and Allen, P.M. 1992. Expression of myosin-class n major histocompatibility complexes in the normal myocardimn occurs before induction of autoimmune myocarditis. Proc. Natl. Acad. Sci. USA. 89: 9131-9135. 211

Sobeslavslg^, O., and Chanock, R.M. 1968. Peroxide fomiation by mycoplasmas which infect man. Proc. Soc. Exp. Biol. Med. 129:531-535.

St. George, T.D., SuUiyan, N. D., Love, J.A., and Horsfall, N. 1971. Experimental transmission of pneumonia in sheep with a mycoplasma isolated from pnemnonic sheep Imig. Aust. Vet. J. 47:282-283.

Stadtlander, C.T.K-H, Watson, H.L., Simecka, J.W., and Cassell, G.H. 1991. Cytopathic effects of Mycoplasma pulmonis in vivo and in vitro. Infect. Inmimi. 59:4201- 4211.

Stadtlander, C.T.K-H., Watson, H.L., and Simecka, J.W., and Cassell, G.H. 1993. Cytopathogenicity of Mycoplasma fermentans (including staialncogmtus). Clin. Infect. Dis. 17(Suppl):S289-301.

Stadtlander, C., and Kirchhoff, H. 1988. The effects of Mycoplasma mobile 163 K on the ciliary epithelium of tracheal organ cultures. Zbl. Bakt. Hyg. A. 269:355-365.

Stalheim, O.H.V. 1983. Mycoplasmal respiratory diseases of ruminants: a review and update. J. Am. Vet. Med. ^soc. 182:403-406.

Stanislawski, L., Simpson, W.A., Hasty, D., Sharon, N., Beachey, E.H., and Ofek, Itzhak. 1985. Role of fibronectin in attachment of Streptococcus pyogenes and Escherichia coli to human cell lines and isolated oral epithelial cells. Infect. Immun. 48:257-259.

Stein, K.E. 1985. Network regulation of the immune response to bacterial polysaccharide antigens. Curr. Top. Microbiol. Tmmnnnl 119:57-74.

Stevens, P., Huang, S.N.H., Welch, W.D., and Young, L.S. 1978. Restricted complement activation by Escherichia coli with K-1 capsular serotype: a possible role in pathogenicity. J. Immunol. 121:2174-2180.

Stuart, P.M., Cassell, G.H., and Woodward, J.G. 1989. Induction of class n MHC antigen expression in macrophages by Mycoplasma species. J. Immunol. 142: 3392-3399.

Stuart, P.M., Cassell, G.H., and Woodward, J.G. 1990. Mycoplasma induction of class n MHC antigens, possible role in autoimmunity. Zbl. Bakt. 20(Suppl.): 575S-577S.

Sugimura, K., Ueda, Y., Takeda, K., Fukuda, S., Tsukahara, K., Habu, Y, Fujiwara, H., and Azimia, I. 1989. A cytokine, lymphocj^ blastogenesis inhibitory factor (LBIF), arrests mitogen-stimulated T lymphocytes at early G1 phase with no 212 influence on interleukin 2 production and interleukin 2 receptor light chain expression. Eur. J. Immunol. 19:1357-1364.

Sullivaii, N.D., St. George, T.D., and HorsfaU, N. 1973. A proliferative interstitial pneumonia of sheep associated with mycoplasma infection. 1. Natural history of the disease in a flock. Aust. Vet. J. 49:57-62.

Suter, M., Kobisch, M., and Nicolet, J. 1985. Stimulation of immnnnglnhu1 in-cnntaining cells and isotype-specific antibody response in experimental Mycoplasma hyopneumoniae infection in specific-pathogen-free pigs. Infect. Immun. 49: 615-620.

Tajima, M., Yagihashi, T., and Yuriko, M. 1982. Capsular material of Mycoplasma gallisepticwn and its possible relevance to the pathogenic process. Infect, immun. 36:830- 833.

Tajima, M., and Yagihashi, T. 1982. Interaction of Mycoplasma hyopneumoniae with the porcine respiratory epithelium as observed by electron microscopy. Infect. Immun. 37:1162-1169.

Tajima, M., Yagihashi, T., and Nunoya, T. 1985. Ultrastructure of mycoplasmal capsules as revealed by stabilization with antiserum and staining with rutheniimi red. Jpn. J. Vet. Sci. 47:217-223.

Taylor, G., Taylor-Robinson, D., and Femald, G.W. 1974. Reduction in the severity of Mycoplasma pneumoniae-mduc&i pneumonia in hamsters by immimosuppressive treatment with anti-thymocyte sera. J. Med. Microbiol. 7:343-348.

Taylor-Robinson, D., Furr, PM., Davies, H.A., Manchee, R.J., Mouches, C., and Bove, J.M. 1981. Mycoplasmal adherence with particular reference to the pathogenicity of Mycoplasma pulmonis. Isr. J. Med. Sci. 17:599-603.

Thirkell, D., Spooner, R.K., Jones, G.E., Russel, W.C., and Voice, M.W. 1991. Cross-reacting antigens between Mycoplasma ovipneumoniae and other species of mycoplasma of animal origin, shown by ELISA and immunoblotting with reference antisera. Vet. Microbiol. 26:249-261.

Thomas, L. 1969. Mechanism of pathogenesis in mycoplasma infection. The Harvey Lectures. 63:73-98. 213

Thomas, C.B., Mellter, J., Sharp, P., Jensen-Kestenbader, J., and Schultz, R.D. 1990. Mycoplasma bovis suppression of bovine lymphocyte response to phytohemagglutinin. Vet. Lumunol. Inimunopathol. 26:143-155.

Thomsen, A.C., and Heron, I. 1979. Effect of mycoplasmas on phagocytosis and immunocompetence in rats. Acta Pathol. Microbiol. Scand. 87C:67-71.

Tipimeni, P., Moore, B.S., Hyde, J.S., and Schauf, V. 1980. Ig£ antibodies to Mycoplasma pneumoniae in asthma and other atopic diseases. Ann. Allergy. 45: 1-7.

Tryon, V.V., and Baseman, J.B. 1992. Pathogenic determinants and mechanisms. In: Maniloff, J., McElhany, R.N., Finch, L.R., and Baseman, J.B. (eds.). Mycoplasmas: molecular biology and pathogenesis. ASM, Washington, DC. pp. 457-471.

Uppal, P.K., and Chu, A.P. 1977. Attachment of Mycoplasma gallisepticum to tracheal epithelium of fowls. Res. Vet. Sci. 22:259-260.

Whitescarver, J., Castillo, F., and Fumess, G. 1975. The preparation of membranes of some human T-mycoplasmas and the analysis of their carbohydrate content. Proc. See. Exp. Biol. Med. 150:20-22.

Whitfield, C. 1988. Bacterial extracellular polysaccharides. Can. J. Microbiol. 34:415- 420.

Wilson, M.H., and Collier, A.M. 1976. Ultrastructural stody of Mycoplasma pneumoniae in organ culture. J. Bacterial. 125:332-339.

Yagihashi, T., Nunoya, T., and Tajima, M. 1987. Isr. J. Med. Sci. 23:517.

Yano, T., Ichikawa, Y., Komatu, S., Arai, S., and Oizmni, K. 1994. Association of Mycoplasma pneumoniae antigen with initial onset of bronchial asthma. Am. J. Resp. Crit. Care Med. 149: 1348-1353.

Young, T.F., Erickson, B.Z., Rosenbusch, R.F., and Ross, R.F. 1996. Capsular polysaccharide of Mycoplasma hyopneumoniae. 11th International Congress of the International Organization for Mycoplasmology. Abst. No PS 5-224. pp. 391. Orlando, Florida.

Zhang, Q., Yomig, T.F., and Ross, R.F. 1994. Microtiter plate adherence assay and receptor analogs for Mycoplasma hyopneumoniae. Infect. Inomun. 62: 1616-22. 214

Zielinski, G.C. 1991. Adherence to host tissues, siirface hydrophobici^, and pathogenic and phenotypic characteristics of Mycoplasma hyopneumoniae. Ph.D. Thesis. Iowa State University, Ames. pp. 155.

Zielinski, G.C., and Ross, R.F. 1992. Morphologic features and hydrophobicity of the cell surface of Mycoplasma hyopneumoniae. Am. J. Vet. Res. 53:11191124. 215

ACKNOWLEDGMENTS

I wish to express my sincere gratitude to my major professor Dr. Merlin L. Kaeberle for his financial support, guidance and contributions throughout the course of these smdies.

I would also like to thank Drs. Ricardo F. Rosenbush, Yosiya Niyo, John J. Andrews,

Donald C. Beitz and Michael Waimemuhler for serving on my committee.

My deepest appreciation goes to my wife and children for their love, understanding and sacrifices dming all these years that took me to complete my studies. My gratimde is also extended to my parents, all members of my family and my fnends for their patience and support throughout my stay at Iowa State University.

Special thanks go Jose Lopez-Virella, Roman Pogranichnyy, Susan L. Seiler, Mattew

D. Sternberg, David L. Cavanaugh, Joann M. Kinyon, Dr. Paul G. Eness and the staff of

Dr. Roth's laboratory without whose help this work would not have been possible. My sincere gratimde goes to Dawne Buhrow, Dr. Ada Mae Lewis, Dr. Moustapha A. Gabal,

Dr. Yosiya Niyo, Dr. Loren A. Will, Dr. Dagmar E. Frank, Robert Cowly,

Mrs. Kaeberle, Deb Vance and Linda Metz for help in various ways.

Finally, I would like lo thank the Malian Government for granting me a leave of absence for the completion of my Ph.D degree.