The Role of cytokines in Sendai virus-induced bronchiolar fibrosis

MISSING IMAGE

Material Information

Title:
The Role of cytokines in Sendai virus-induced bronchiolar fibrosis
Physical Description:
vi, 152 leaves : ill. ; 29 cm.
Language:
English
Creator:
Uhl, Elizabeth Whitford, 1957-
Publication Date:

Subjects

Subjects / Keywords:
Respirovirus -- immunology   ( mesh )
Respirovirus -- pathogenicity   ( mesh )
Pulmonary Fibrosis -- etiology   ( mesh )
Pulmonary Fibrosis -- physiopathology   ( mesh )
Transforming Growth Factor beta -- physiology   ( mesh )
Transforming Growth Factor beta -- metabolism   ( mesh )
Tumor Necrosis Factor -- physiology   ( mesh )
Tumor Necrosis Factor -- metabolism   ( mesh )
Gene Expression Regulation   ( mesh )
Rats, Inbred BN   ( mesh )
Rats, Inbred F344   ( mesh )
Cell Line   ( mesh )
Macrophages, Alveolar   ( mesh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1996.
Bibliography:
Includes bibliographical references (leaves 135-151).
Statement of Responsibility:
by Elizabeth Whitford Uhl.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
oclc - 49818329
ocm49818329
System ID:
AA00011214:00001

Full Text










THE ROLE OF CYTOKINES IN SENDAI VIRUS-INDUCED BRONCHIOLAR
FIBROSIS













By

ELIZABETH WHITFORD UHL


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA













ACKNOWLEDGMENTS


I cannot begin to thank my family for their support: my parents for their

example and their encouragement, my sisters, Jean and Mary, and brother

Charlie, for never letting me feel too sorry for myself when going got tough, and

especially my husband Frank for his unfailing support and for covering for the

many times I was not there. Special thanks go to William Castleman for giving

me the opportunity to do the work, and for setting and maintaining high

standards. I thank my the members of my committee Claus Buergelt, Mary

Brown, Anthony Barbet and especially Rosalia Simmen for their interest and

assistance. I would also like to acknowledge Lyle Moldawer for providing the

TNF receptor fusion protein and for his enthusiasm. Roberta O'Connor, Karen

Dukes, Allison Mays, Christine Klein and especially Tracy Jack provided both

help with the laboratory work and friendship, without which this project would

have been much more difficult.









TABLE OF CONTENTS



ACKNOWLEDGMENTS................................................ ii

ABSTRACT............................ ......... ...... ........................ v

CHAPTER 1 LITERATURE REVIEW

Viral Respiratory Infections and Asthma ........................................ 1
Sendai Virus Infection in Rats ....................... ..................... 4
The Role of Cytokines in Pulmonary Disease .................................. 6
T G F ........................................................... ............... .... 7
TN F-a ...................................................... .................... 12
P D G F ..................................................... ........................... 16

CHAPTER 2: RESEARCH PROTOCOL

Hypothesis and Specific Aims ....................... ........................ 19
Background/Significance .................................................................... 20
Experim ental Design .......................................................................... 21

CHAPTER 3: INCREASED TRANSFORMING GROWTH FACTOR Pi
(TGF-0P) GENE EXPRESSION PRECEDES SENDAI
VIRUS-INDUCED BRONCHIOLAR FIBROSIS

S um m ary ....................................................................... ............... 34
Introduction ......................................................... ................... 35
Materials and Methods ................... .. ................. 38
R e su lts ............................................................................................... 4 3
D discussion .......................................................................................... 4 5

CHAPTER 4: INCREASED TUMOR NECROSIS FACTOR-a (TNF-a)
GENE EXPRESSION IN MACROPHAGES PRECEDES
BRONCHIOLAR FIBROSIS IN VIRUS-SUSCEPTIBLE BN RATS

S um m ary .............................................. ........................ 55
Introduction .......................................................... .................. 56
M materials and M ethods ........................................ ....................... 59
Results ................... .. ........................... 65
D discussion ....................................... ............................................. 67









CHAPTER 5: FIBROBLAST PROLIFERATION IN THE DEVELOPMENT OF
SENDAI VIRUS-INDUCED BRONCHIOLAR FIBROSIS IN BN
RATS IS INHIBITED BY A SOLUBLE TNF RECEPTOR

Summary ....................................... ....... .................. 77
Introduction ................................... ......... ....................... 78
Materials and Methods ................................... ......................... 80
Results ............................................... ........................ 85
D discussion ........................................................ .......................... 89

CHAPTER 6: SENDAI VIRUS DIRECTLY UPREGULATES TUMOR NECROSIS
FACTOR-a (TNF-a), BUT NOT TRANSFORMING GROWTH
FACTOR-3, (TGF-p ) GENE EXPRESSION IN RAT ALVEOLAR
MACROPHAGES

Sum m ary ............................................................. .................. 102
Introduction ..................................... ..... ................. ................ 103
Materials and Methods ........................................... ..................... 105
R results ............................................................ .................... ....... 109
D discussion ......................................................... .................... 111

CHAPTER 7: GENERAL SUMMARY....................................................... 123

APPENDIX ............................................ 129

REFERENCES .............................................. 135

BIOGRAPHICAL SKETCH .................................................... 152









Abstract of Dissertation Presented to the Graduate School of the University of
Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of
Philosophy

THE ROLE OF CYTOKINES IN SENDAI VIRUS-INDUCED BRONCHIOLAR
FIBROSIS

By

Elizabeth Whitford Uhl

May 1996

Chairperson: Dr. William L. Castleman
Major Department: Veterinary Medicine

Viral bronchiolitis in children may induce persistent alterations in lung

function including increased airway resistance and airway hyperresponsiveness.

Similar pulmonary functional abnormalities induced in rats following infection

with parainfluenza type I (Sendai) virus are associated with bronchiolar mural

fibrosis. This observation led to the hypothesis that virus-induced bronchiolar

mural fibrosis is mediated through virus-induced increases in gene expression of

transforming growth factor-B, (TGF-13), platelet-derived growth factor-B (PDGF),

andlor tumor necrosis factor-a (TNF-a) in the bronchiolar microenvironment.

Pulmonary TGF-B, and TNF-a mRNA and protein expression were assessed

in rat strains that are susceptible (BN) and resistant (F344) to virus-induced

bronchiolar fibrosis. Northern analysis, in situ hybridization and protein

immunocytochemistry demonstrated that virus-inoculated BN rats had

significantly higher increases in TGF-B, and TNF-a pulmonary mRNA and

increased numbers of bronchiolar macrophages and fibroblasts expressing









TGF-B, and TNF-a protein than virus-inoculated F344 rats. Pulmonary PDGF

mRNA levels were not altered by viral infection in either rat strain.

The significance of TNF-a in the development of virus-induced bronchiolar

fibrosis was determined by using a 55 kD soluble TNF receptor-immunoglobulin

G fusion protein (sTNFR-IgG) to inhibit TNF-a bioactivity in virus-inoculated BN

rats. Treated rats had fewer proliferating bronchiolar fibroblasts, as detected by

bromodeoxyuridine incorporation, compared to virus-inoculated control rats.

There was also significantly increased mortality in p55sTNFR-lgG-treated virus-

inoculated rats associated with increased viral replication and decreased

numbers of macrophages and lymphocytes in bronchioalveolar lavage fluid.

To assess the mechanism of the virus-induced increase in cytokine

expression in macrophages, a rat alveolar macrophage cell line (NR8383) was

inoculated with Sendai virus, and the levels of TNF-a and TGF-p, mRNA and

protein were determined. TNF-a mRNA and protein levels were significantly

elevated at 24, 48, and 72 hours after virus-inoculation. Cells incubated with

UV-inactivated virus did not have any detectable TNF-a mRNA 24 hours after

inoculation. TGF-p, mRNA and protein levels in virus-inoculated cells were not

consistently elevated over those from control cells. The results of these

experiments indicate that TGF-31 and TNF-a are important mediators of virus-

induced bronchiolar fibrosis. They also demonstrate that Sendai virus can

directly upregulate TNF-a expression, and that TNF-a is critical to the immune

and inflammatory response to the virus.














CHAPTER 1
LITERATURE REVIEW



Viral Respiratory Infections and Asthma



The majority of individuals who develop viral respiratory infections recover

with normal airway function. However, following clinical resolution of acute viral

infections, a subgroup of patients develop airway functional abnormalities

ranging from persistent increases in airway resistance and airway

hyperresponsiveness to asthma.17 The potentiating impact of viral respiratory

infections on asthma appears to be dependent on the presence of underlying

risk factors, with young children, especially boys, being the most likely to

develop wheezing with a viral respiratory infection. Other risk factors include: a

family history of allergic disease, the development of an IgE-dominated immune

response, the severity of both the infection and the underlying asthma, and the

type of virus.17

Epidemiological studies have associated infection with respiratory syncytial

virus, parainfluenza viruses, adenoviruses, or rhinoviruses with the development

of asthma, and several pathogenic mechanisms have been proposed to explain

the development of virus-induced pulmonary functional abnormalities.1'78 In










children with acute respiratory syncytial virus (RSV) bronchiolitis, for example,

prospective studies have shown that titers of RSV-specific IgE are independent

predictors of airway hyperresponsiveness,O and rising titers correlate with

increased leukotriene C4 (LTC4) levels in nasopharyngeal washes during

subsequent RSV infections.10 These results suggest that RSV itself may act as

an allergen in susceptible patients. Viral respiratory infections also may

facilitate sensitization to non-viral antigens in genetically predisposed individuals

since virus-induced epithelial damage permits exposure of allergens to the host

immune system."

A further consequence of epithelial damage is exposure of nerves in the

airway wall. There is an extensive network of afferent nerves situated

immediately below the epithelial tight junctions.12 Exposure of these nerves to

the environment can result in the release of neuropeptides such as substance P

and neurokinin A, both of which can stimulate smooth muscle contraction.12.13

Viral infections also can have effects on neuropeptide release and metabolism

as well as nerve conduction and receptor function.14'15

Inflammation of the airways is a characteristic feature of asthma, and

increased recruitment and persistence of inflammatory cells in airways may be

potential mechanisms by which viral infections contribute to the development of

airway functional abnormalities.8'6 Studies using animal models of virus-

induced hyperresponsiveness have correlated the appearance of airway

hyperresponsiveness induced by viruses with the influx of inflammatory cells.'718









3

Most of these inflammatory cells were macrophages, which have the capacity to

release a vast array of inflammatory mediators and cytokines including

leukotrienes, thromboxanes, platelet activating factor, tumor necrosis factor-a

(TNF-a), interleukin-2 (IL-2), and prostaglandin F 2 that can modulate airway

smooth muscle contraction, affect mucus secretion and alter vascular

permeability.1920,21

Viral infections may also induce abnormalities in pulmonary structure. Airway

wall thickening with fibrosis is a common lesion in asthmatics that can contribute

to the development of increased airway resistance by altering the mechanical

properties of airways.2224 Viral infections may enhance the development of

airway fibrosis by causing epithelial necrosis, which leads to the build up of

necrotic debris, inflammatory cells, mucus, and plasma exudate in airways.

Subsequent organization of this material leads to the deposition of fibrous

connective tissue and can result in irreversible peripheral airway loss and airway

obstruction. The stimulation of aberrant or exuberant repair processes following

viral infection may also result in mural fibrosis, which can be significant even if

airways are not completely occluded, since the deposition of connective tissue in

airway walls dramatically affects the degree of luminal narrowing during smooth

muscle contraction."

The development of persistent pulmonary functional abnormalities associated

with viral respiratory infections is dependent upon a complex interplay of

mechanisms that include IgE-dependent reactions, enhanced inflammation,










epithelial damage, autonomic nervous system dysfunction, modulation of the

immune response, and the induction of structural abnormalities. The purpose of

the studies presented here was to investigate the potential role of fibrogenic

cytokines in the pathogenesis of virus-induced bronchiolar fibrosis using

parainfluenza type I (Sendai) virus infection in rats as the experimental model.



Sendai Virus Infection in Rats: Summary of Previous Studies



Sendai virus is a common pathogen of rodents that can induce either clinical

disease or inapparent infection in laboratory rats, mice, guinea pigs, and

hamsters.227 Spontaneous disease associated with Sendai virus infection is

well documented,28 and the pathological features have been defined

experimentally.213 Following aerosol infection, CD (Sprague-Dawley) rats

develop multifocal necrotizing bronchiolitis associated with viral replication in

bronchiolar epithelial cells. Infection of neonatal (5 day old) rats produces the

most severe pneumonia at seven days post infection (dpi), with epithelial repair

and resolution of the bronchiolitis being largely complete at 17 dpi.29 3 Neonatal

CD rats infected with Sendai virus develop long term pathological sequelae

consisting of bronchiolar hypoplasia and fibrosis, alveolar dysplasia, and

bronchiolar mastocytosis.30 These morphological changes are associated

with pulmonary functional abnormalities including increased pulmonary

resistance, decreased arterial PO2, and airway hyperresponsiveness to

aerosolized methacholine.35.











Inbred rat strains that are susceptible ( Brown Norway: BN) and resistant

(Fischer 344: F344) to Sendai virus induced bronchiolar hypoplasia, fibrosis,

mastocytosis and the associated pulmonary function abnormalities have been

identified.37 At 7 days after Sendai virus infection, BN rats had higher pulmonary

viral titers, and increased numbers of neutrophils in the bronchoalveolar lavage

fluid compared to F344 rats.37 Significantly increased numbers of bronchioles

with aggregates of lymphocytes and macrophages were observed in BN rats at

14 days post infection.37 Neonatally infected BN rats also had an over 3-fold

increase in the area of bronchiolar wall per unit length of bronchiolar basement

membrane by 65 days dpi." This peribronchiolar increase in tissue was largely

accounted for by a 1.5-fold increase in the number of fibroblasts per mm2 of

bronchiolar basement membrane (b.m.) as well as by a 4-fold increase in

collagen fiber density, and a 4.7 fold increase in extracellular matrix (um3 lum2

b.m.).3 Active proliferation of fibroblasts, assessed by bromodeoxyuridine

labeling, demonstrated significant increases (p<0.05) in the percentages of

labeled bronchiolar fibroblasts in BN rats at 7, 10, & 14 dpi (unpublished data).

In contrast, neonatally infected F344 rats did not have evidence of diffuse

bronchiolar fibrosis.3

The results of these studies indicate that bronchiolar fibrosis is an important

component of Sendai virus-induced increases in bronchiolar wall thickness and

therefore could explain the persistently increased airway resistance noted in

neonatally infected BN rats. Comparison of BN and F344 rats allows an unique










opportunity to determine which of several potentially important mediators are

responsible for the dysplastic, fibrotic, and hyperresponsive changes induced by

neonatal viral bronchiolitis. The studies outlined in this proposal are designed to

investigate the role of three potent mediators of pulmonary fibrosis, transforming

growth factor-B (TGF-B), tumor necrosis factor-a (TNF-a), and platelet-derived

growth factor (PDGF), in the pathogenesis of Sendai virus-induced bronchiolar

fibrosis, and to assess the ability of Sendai virus to directly induce expression of

these cytokines.



The Role of Cvtokines in Pulmonary Disease



In recent years the role of cytokines in the regulation of both physiologic and

pathologic processes occurring in the lung has been extensively investigated."

5 Studies on naturally occurring and chemically induced pulmonary fibrosis

indicate that TGF-B,TNF-a, and PDGF are important mediators of pulmonary

fibrosis.'4" In addition, the ability of infectious agents to directly effect the

expression of inflammatory and fibrogenic mediators, including TGF-B and TNF-

a, has been demonstrated and is believed to be an important aspect in the

pathogenesis of several viral diseases."0










Transforming Growth Factor-B (TGF-B )



TGF-B is a multifunctional growth factor, having growth regulatory activities

on essentially all cell types. The cellular response to TGF-B is complex, being

dependent on cell type, the presence of other growth factors, level of

differentiation, and the state of cell activation.61 It is a potent inhibitor of

epithelial cell growth,6 probably through inhibition of the c-myc proto-

oncogene,6 while in combination with PDGF it is mitogenic for cells of

mesenchymal origin." TGF-B is also an important regulator of differentiation for

mesenchymal cells."

In the normal lung, TGF-B is postulated to have a role in the stabilization of

the structural cell populations;5 however, if regulation of its biosynthesis is

disrupted, high protein levels could suppress repair processes by inhibiting

epithelial cell proliferation. All three isoforms of TGF-B (-B,,-r2,-B3) are

constitutively expressed in the lung. mRNA transcripts are found in

macrophages, smooth muscle cells and fibroblasts, while protein expression

also may be associated with the bronchiolar epithelium.65 There are cell type-

specific differences in the number and type of TGF-B receptors as well as

subtypes of receptors with different affinities for the different TGF-13 isoforms.

Rat lung fibroblasts, for example, have receptors with a higher affinity for TGF-B,

than for TGF-112." This finding correlates well with the fact that TGF-R, is the

isoform most implicated in fibrosis.67








8

TGF-B is a potent inducer of tissue fibrosis," with a variety of both paracrine

and autocrine effects. It can increase the tissue levels of collagen and other

matrix proteins" by increasing the transcription of fibronectin and procollagen

genes,68'7 through postranscriptional regulation of steady-state levels of

fibroblast collagen and fibronectin mRNA,7 and by inhibiting the degradation of

existing extracellular matrix by both decreasing the biosynthesis of proteases"

and increasing the synthesis of protease inhibitors.46 TGF-B, along with TNF-a,

can also enhance the development of pulmonary fibrosis by impairing the ability

of lung fibroblasts to degrade fibrin,n thus stabilizing a transitional matrix on

which fibroblasts can proliferate and secrete collagen. In addition, TGF-1 has

effects on the expression of other fibrogenic growth factors including; PDGF,

TNF-a, and basic fibroblast growth factor (bFGF).74-7

Evidence that TGF-B is one of the most potent fibrogenic agents known has

come from studies of several different models of tissue fibrosis (see reference 67

for review). Investigations of fetal wounds, which heal without the formation of a

scar, revealed that unlike in neonates or adults, TGF-B was not associated with

the wound, while the injection of TGF-B into fetal wounds led to scarring.47 The

significance of this finding was followed up by the observation that the injection

of antibodies to TGF-B into wounds in adults resulted in healing without the

formation of a scar.7 In models of pathological fibrosis increased TGF-B

expression has been observed in humans with liver cirrhosis, idiopathic

pulmonary fibrosis,7 or systemic sclerosis,78 in mice with hepatic fibrosis,79 and










in rats with experimental glomerulonephritis.a' The importance of TGF-B in the

pathogenesis of glomerulonephritis in rats was confirmed by the ability of

injected neutralizing antibodies against TGF-I to suppress the disease.

Studies of chemically induced and naturally occurring pulmonary

fibrosis',12 have confirmed the importance of TGF-B in the pathogenesis of the

lung lesions. Bleomycin treatment of rat lung fibroblast cultures resulted in an

increase of TGF-B mRNA and an increased secretion of TGF-B protein. The

response could be blocked by treatment with an inhibitor of protein synthesis,

and by an inhibitor of RNA synthesis." In an in vivo study, Hoyt and Lazo4

demonstrated that the amount of TGF-B mRNA extracted from the lungs of mice

chronically exposed to bleomycin was five-times the amount extracted from the

lungs of control animals. This increase preceded changes in gene expression

for collagens and fibronectin and coincided with areas of active fibrosis. Of

particular interest is the observation that a mouse strain that does not develop

pulmonary fibrosis after bleomycin exposure had a decrease in total mRNA that

included a drop in TGF-I mRNA.' Bleomycin also induced increases in lung

collagen accumulation which could be inhibited by the administration of

antibodies to TGF-B, confirming TGF-B has an important role in the etiology of

bleomycin induced pulmonary fibrosis." Immunocytochemical analysis has

identified macrophages as the most common pulmonary cell type expressing

TGF-B protein in bleomycin-treated rats.81 Patients with idiopathic pulmonary

fibrosis also have increased pulmonary expression of TGF-B protein; however, it










is predominantly in bronchiolar and alveolar epithelial cells and deposited

extracellularly in subepithelial areas adjacent to the fibrotic lesions." Finally,

neutralization of TGF-3, in a mouse model of immune-induced lung fibrosis

caused by repeated intranasal exposure to heat-killed bacillus Calmette-Guerin

(BCG) resulted in substantial decreases in lung fibrosis and granulomatous

inflammation, which implies a role for TGF-B, in inducing inflammation and lung

fibrosis in response to an immune stimulus."

In addition to its role in tissue fibrosis, TGF-B is an important modulator of the

immune response, with both immunosuppressive and proinflammatory effects

depending upon the context in which it is presented. In experimental acute and

chronic arthritis, for example, systemic administration of TGF-1 nearly

obliterates the disease process"7 whereas local administration exacerbates it."

Systemically delivered TGF-B initially encounters the endothelium and its

primary action may be to decrease endothelial cell adhesion molecule

expressionO"' thus impairing leukocyte recruitment. TGF-B also can inhibit

activated T-cell proliferation and deactivate macrophages." These

immunosuppressive effects may not be predominant in all situations, however,

since TGF-B is strongly chemotactic for monocytes and macrophages and can

induce undifferentiated monocytes to express a variety of inflammatory

mediators including: IL-1, TNF-a, IL-6, and PDGF.9'

The complex modulation of the immune response by TGF-B has been

partially characterized by studies of transgenic and gene knockout mice.










Targeted disruption of the TGF-1, gene in mice results in a multifocal

inflammatory disease in many tissues, especially the heart and lungs.'"10' The

nature of the inflammatory response varied depending upon the tissue, for

example: the lungs had interstitial inflammatory infiltrates consisting of T and B

lymphocytes associated with increased vascular expression of major

histocompatibility complex class I and II antigens, while macrophages were the

major type of inflammatory cell in the heart. Surprisingly, inflammatory lesions

were also prominent in transgenic mice overexpressing TGF-13.7 TGF-B can

activate immature monocytes and T lymphocytes while inhibiting macrophages

and activated lymphocytes.76 In the normal inflammatory response this means

immature cells are turned on to participate in inflammation, while mature,

activated cells are switched off, promoting resolution of the inflammatory process

and tissue repair. The unexpected similarities in the phenotypes of the

transgenic and gene knockout mice is probably the result of TGF-I3's ability to

both accelerate and inhibit the inflammatory response.

In several infectious diseases, the immunosuppressive effects of TGF-1

predominate, resulting in prolonged survival of the pathogen and increased

tissue damage.55, In Leishmania infection the down-regulation of macrophages

by TGF-I1 is a virulence factor that determines in vivo susceptibility.9-102 During

acute Trypanosoma cruzi infection in mice, TGF-B1 inhibited the protective,

macrophage activating effects of interferon-y and exacerbated infection.10 In

Mycobacterium tuberculosis infection in human monocytes, neutralizing antibody









12
to TGF-B, reduced the intracellular bacterial growth, while treatment with TGF-B,

abrogated the bacteriostatic effects of TNF-a and interferon-y.14 TGF-1 also

has been implicated as having a role in the pathogenesis of viral infections. In

acquired immunodeficiency syndrome (AIDS), a viral protein (tat) appears to

directly upregulate the production of TGF-1B,'1 setting up a potential feedback

loop since TGF-B enhances HIV-1 replication in some cell lines.108 TGF-B has

also been demonstrated to increase human cytomegalovirus replication in

human lung fibroblasts; in contrast, basic fibroblast growth factor repressed viral

replication. 107

Considering the wide-ranging effects of TGF-B on both the inflammatory

response and the pathogenesis of fibrosis, excessive production of TGF-B in the

susceptible BN rats could be contributing both to the increased viral titers

observed in these rats at 7 dpi, and to the peribronchiolar fibrosis and the

epithelial hypoplasia.



Tumor Necrosis Factor-a



TNF-a is a pro-inflammatory cytokine produced by monocytes and

macrophages.'" Its effects are complex, as it appears to play an important role

in both acute injury and in the fibroproliferative response.. '44S. 110 In the

inflammatory response, TNF-a stimulates the adhesion of leukocytes to the

endothelial surface by inducing the expression of adhesion molecules for










neutrophils, monocytes and lymphocytes.1' Other pro-inflammatory effects on

endothelial cells include alterations in the transcription and cell surface

expression of class I MHC antigens and induction of IL-1 secretion."' TNF-a

also has direct stimulatory effects on neutrophils, acting to enhance the

phagocytic capacity, antibody-dependent cytotoxicity, degranulation, and

production of reactive oxygen species."' In the repair process, a role for TNF-a

is suggested by its growth stimulatory effect on fibroblasts, which is thought to

be mediated by the increased cell surface expression of epithelial growth factor

receptors, and by its ability to induce angiogenesis.11

Several of the pro-inflammatory effects of TNF-a, for example: its ability to

increase the adhesion of both neutrophils and eosinophils to vascular

endothelium,11'112 the potentiation of interleukin-3 and granulocyte/macrophage-

colony stimulating factor induced proliferation of hemopoietic progenitor cells,"3

the enhancement of eosinophil cytotoxicity,"4 and its mechanistic role in the

activation of dermal endothelium triggered by mast cell degranulation,"'5 suggest

TNF-a may be important in the pathogenesis of allergic inflammation. In the

lung, this importance has been implicated by studies of airway

hyperresponsiveness.

Bronchial hyperresponsiveness and inflammation, induced by the

administration of aerosolized endotoxin to rats, has been demonstrated to be

mediated by TNF-a.11 Pretreatment of rats with antibodies against TNF-a

decreased the hyperresponsiveness, while administration of recombinant TNF-a










induced it. The same study also demonstrated that the increase in neutrophil

numbers observed after the administration of endotoxin was partially inhibited by

antibodies to TNF-a."1 Treatment with antibodies to TNF-a was also used in a

mouse model of hypersensitivity pneumonitis induced by intranasal inoculation

with the actinomycete Faenia rectiviroula (farmer's lung).116 Such treatment

blocked inflammatory cell recruitment to the lungs, completely abolished TNF-a

secretion by bronchiolar associated lymphoid tissue, and prevented the

development of fibrosis."1 At the cellular level, in vitro studies have

demonstrated the production of TNF-a in pulmonary mast cells, alveolar

macrophages, and bronchiolar epithelial cells in lung fragments following IgE

receptor triggering.117

Several studies have indicated the importance of TNF-a in fibrotic pulmonary

disorders. For example, significant increases in TNF-a mRNA have been

observed in the lungs of treated mice susceptible to bleomycin induced

pulmonary fibrosis, but not in the lungs of treated non-responder mice.10 In rats,

the importance of bleomycin-induced increases in the pulmonary expression of

TNF-a was further emphasized by the finding that the development of fibrosis

could be blocked by the administration of antibodies to TNF-a.11" In bleomycin-

induced pulmonary fibrosis there is a strong correlation between collagen

deposition and platelet trapping, which is probably due to the fact that platelets

are rich sources of fibrogenic mediators such as TGF-P and PDGF. TNF-a may

contribute to bleomycin-induced pulmonary fibrosis through platelet trapping by










upregulating endothelial expression of CD54, an adhesion molecule for

platelets.1" An association between increased levels of TNF-a mRNA and the

development of silicosis has been demonstrated in mice. In this model the

fibrotic changes were not only inhibited by antibodies to TNF-a, but also

enhanced by the infusion of recombinant TNF-a.120

Several viruses,121,1".0 including Sendai virus,57 directly induce the

expression and production of TNF-a; however, its role in the pathogenesis of

viral diseases is complex. TNF-a can inhibit the replication of several DNA and

RNA viruses including herpes simplex type II, vesicular stomatitis virus,

encephalomyocarditis virus and adenovirus type-2," and can kill virus-infected

cells originally resistant to its cytolytic actions." The importance of TNF-a as an

antiviral cytokine has also been emphasized by the discovery that several

viruses produce virulence factors that specifically inhibit TNF-a activity.'22124

TNF-a is not protective in all viral infections, however, as it can reactivate latent

herpes simplex infection in monocytes,125 and HIV infection in cell lines of T-cell

and monocyte origin.126 111 Viral infections also may act to prime cells for the

production of TNF-a. For example, influenza A virus infection of murine

macrophages resulted in a massive accumulation of TNF-a mRNA, including a

unique high molecular weight transcript, that was not translated to the bioactive

protein until the cells were exposed to small amounts of endotoxin."

Based upon the studies demonstrating the importance of TNF-a in the

pathogenesis of airway fibrosis, its probable role in the development of airway









16
hyperresponsiveness, its importance in viral pathogenesis, and the observation

that Sendai virus induces the production of TNF-a in monocytes, TNF-a can be

predicted to play a role in the pathogenesis of the Sendai virus-induced

bronchiolar fibrosis and airway hyperresponsiveness observed in the susceptible

BN rats. The studies presented here were designed to test that prediction.



Platelet-Derived Growth Factor



PDGF is a highly cationic protein dimer composed of two chains, A&B, that

can exist as AA, AB, or BB depending upon the species, cellular source," and

transcriptional regulator. PDGF is secreted by platelets, endothelial cells,

activated monocytes, macrophages, and mesenchymal cells.8,127 It is a

competence or priming type growth factor for cells of mesenchymal origin,'28 and

a chemoattractant for fibroblasts and smooth muscle cells.'1 PDGF stimulates

collagen synthesis, and may act to carry out the fibro-proliferative actions of

TGF-R since cells that cannot both synthesize and respond to PDGF remain

inhibited by TGF-B.4~50 Stimulation of fibroblast proliferation by both IL-1 and

TNF-a may also be mediated by the induction of PDGF.'30,131

Expression of PDGF in normal cells is tightly regulated; it acts locally in low

concentration and is rapidly cleared. In the lung, PDGF accounts for most of the

mitogenic activity of alveolar macrophages,51 and there is evidence that it is

important in fetal lung development.'32 At sites of tissue injury, PDGF-BB










enhances the inflammatory phase by: increasing the directed migration of

macrophages, inhibiting natural killer cell activity,13 and stimulating T

lymphocytes to secrete IL-1, while decreasing their secretion of IL-4, IL-5, and

interferon.'" In the repair phase of tissue injury, PDGF triggers an earlier, more

sustained influx of procollagen type I fibroblasts than that produced by TGF-B.52

PDGF-BB has also been found to activate eosinophils, and has been speculated

to play a role in the pathogenesis of asthma.13

Clinical and experimental evidence of the importance of PDGF in the

pathogenesis of pulmonary fibrosis consists of investigations of its role in

idiopathic and chemically-induced pulmonary fibrosis.'" 137', Patients with

idiopathic pulmonary fibrosis (IPF) were demonstrated to have a 3-fold increase

in the percentages of interstitial macrophages expressing PDGF compared to

normal individuals.1l PDGF was also aberrantly expressed in the alveolar

epithelial cells of IPF patients.1'3 Pulmonary PDGF levels have also been

measured in patients with asbestosis and silicosis and those positive for PDGF

experienced progression of their disease." There is also evidence that

chrysotile asbestos may stimulate a PDGF-AA-mediated autocrine loop leading

to hyperplasia in rat lung fibroblasts.1'4

In the rat, measurement of pulmonary PDGF mRNA levels indicate that it is

constitutively expressed at a low level. However, during continuous exposure to

85% oxygen, the steady state levels of PDGF-BB mRNA rose 2.5 fold above

control levels. This increase preceded an increase in DNA synthesis and









18

implies a causal role for PDGF in the fibroblast proliferation induced by

hyperoxic injury to the lung."

The role of PDGF in the pathogenesis of viral-induced tissue damage has not

been investigated. However, the prolonged induction of PDGF production either

by direct transcriptional upregulation by the virus, or some other factor (ie TGF-

B), may be contributing to the airway fibrosis observed in BN rats following

Sendai virus infection.














CHAPTER 2
RESEARCH PROTOCOL


Hypothesis and Specific Aims



The goal of this research was to identify cytokines critical to the development

of bronchiolar fibrosis following bronchiolitis induced by parainfluenza (Sendai)

virus infection during early life. The hypothesis tested was that bronchiolar

mural fibrosis following viral bronchiolitis is mediated by virus-induced increases

in transcription and/or translation of transforming growth factor-3, (TGF-P,),

platelet-derived growth factor-B (PDGF), andlor tumor necrosis factor-a (TNF-a)

in infiltrating bronchiolar and peribronchiolar macrophages. There were 4

specific objectives:



1) Determine whether Sendai virus-induced bronchiolar fibrosis in virus-

susceptible BN rats is preceded by increased gene expression of

fibrogenic cytokines TGF-P,, PDGF, and/or TNF-a in macrophages and

in other bronchiolar cells.

2) Determine whether rats susceptible to virus-induced bronchiolar fibrosis

(BN) have higher cellular expression of fibrogenic cytokines in

bronchioles after virus infection relative to virus-resistant (F344) rats.










3) Determine if the development of virus-induced bronchiolar fibrosis in BN

rats can be blocked by neutralizing cytokine activity.

4) Determine whether parainfluenza virus directly or indirectly induces gene

transcription and/or translation for TGF-P,, PDGF, and/or TNF-a in

pulmonary macrophages.



Background/Significance



Epidemiological studies have associated childhood viral bronchiolitis,

especially as the result of infection with respiratory syncytial virus, parainfluenza

viruses, adenoviruses, or rhinovirus, with the subsequent development of

persistent increases in airway resistance and recurrent wheezing or airway

hyperresponsiveness.'8 A rat model of parainfluenza virus-induced bronchiolar

hyperresponsiveness and increased airway resistance has been characterized,

and both susceptible and resistant inbred rat strains have been identified.33'335,

3637 In this model, the increased airway resistance observed in the susceptible

rats is associated with collagen deposition in the bronchiolar walls, and

increases in the number of peribronchiolar fibroblasts."

Airway wall thickening and fibrosis has also been identified as a common

lesion occurring in human asthma and may contribute to exaggerated airway

luminal obstruction following smooth muscle contraction.8.22.23 In asthmatics,

collagen deposition is most likely the result of fibroblast activation, because it










consists of interstitial structural proteins (collagens III & V), rather than of

basement membrane components (type IV collagen, and laminin).23 This study

investigated the role of fibrogenic cytokines in the development of viral-induced

bronchiolar fibrosis, and assessed the direct effects of virus infection on the

expression of these cytokines.



Experimental Design and Methods



Experiment 1: Pulmonary expression of TGF-1,. PDGF. and TNF-a in Sendai
virus-infected BN and F344 rats


Objectives



1) To determine the magnitude and duration of pulmonary expression of TGF-

a,, PDGF, and TNF-a in control and Sendai virus-infected BN and F344

rats.

2) To determine the number and type of bronchiolar cells expressing

cytokine mRNA and protein in Sendai virus-infected BN and F344 rats.



Rationale



Increases in pulmonary mRNA for TGF-B, PDGF and TNF-a have been

demonstrated in several models of chemically-induced pulmonary fibrosis." ''-70
48, 77,81, 8218,12,9 52 Northern analyses were used to assess the overall magnitude










and duration of pulmonary cytokine expression in infected rats. In situ

hybridization and immunohistochemistry were used to identify the pulmonary cell

types expressing cytokine mRNA and protein after viral infection, and to

determine their association with areas of inflammation and fibrosis.



Design and Methods



Twenty-two day old weanling male BN and F344 rats were inoculated with

Sendai virus or sterile chorioallantoic fluid, and studied according to the

schedule outlined in Table 2-1.

Necropsy and tissue Drocessing. The rats were deeply anesthetized with

sodium pentobarbital and killed by exsanguination via intracardiac puncture. For

Northern analyses, the lungs were frozen in liquid nitrogen after removal and

stored at -80"C until processed. Lungs were harvested and fixed by 2 hour

perfusion via the trachea with 4% paraformaldehyde-PBS (pH 7.4) and

embedded in paraffin, for in situ hybridization and immunohistochemistry.

cDNA probes. 1) A 985 bp cDNA fragment including the major coding

region of rat TGF-1, precursor (supplied by Dr. Su Wen Quian, Laboratory of

Chemoprevention, NIH). 2) PDGF-B: cDNA (human) 2.6 kb ATCC; and 3) TNF-

a: cDNA (rat supplied by Dr. T. Shirai, Laboratory for Chemical Research,

Asahi Chemical Industry Co. Ltd, Tagata-gun, Shizuoka).










Northern analysis. Lung tissue was lysed in guanidium thiocyanate, and

total RNA extracted and precipitated.13 Preliminary Northerns for TGF-p,, TNF-

a, and PDGF were performed on poly (A) RNA pooled from 3-5 animals per

group. Virus-induced increases in pulmonary TGF-J, and TNF-a RNA were

present (data not shown), but pulmonary PDGF RNA levels were not increased

in the virus-inoculated rats of either strain (see Figure 2-1). Northern analysis for

TGF-0, and TNF-a RNA was performed on total RNA extracted from the lungs of

individual animals.

In situ hybridization. The TGF-P, and TNF-a cDNA probes used above

were subcloned and used as templates for the production of the RNA probes.

Digoxigenin- (TNF-a) and 3S-(TGF-P,) labeled antisense riboprobes were made

and used to qualitatively determine which bronchiolar cell types were expressing

cytokine mRNA in lung sections taken from virus- or sham-inoculated rats.

Control sections were incubated with labeled sense riboprobes.

Immunohistochemistry. Antibodies against TGF-B, and TNF-a were used

to detect bronchiolar cell expression of cytokine protein using

immunocytochemical techniques. The labeled bronchiolar macrophages and

fibroblasts were quantitated and the numbers of cells positive for cytokine

protein were compared between treatment groups.

Data analysis. Comparisons of pulmonary cytokine mRNA levels between

BN and F344 rat strains pre- and postinfection, were made by comparing the

relative optical densities (ROD) of the bands on the Northern autoradiographs.









24
The RODs of the total RNA from individual lungs were compared, both between

and within groups, to assess the amount of individual animal variation.

Statistical analysis was performed using a microcomputer-based statistical

program (Systat version 4, Evanston IL). Density of mRNA and protein labeling

were assessed by counting and classifying the number of labeled cells per mm

bronchiolar basement membrane in 25 high power (400X) fields per slide. Three

slides per animal were counted. Differences in the density of stained cells

between the infected and control, BN and F344 rats were compared using

ANOVA and Tukey tests.



Experiment 2: The effects of cytokine inhibition on the development of viral-

induced bronchiolar fibrosis.



Objective



1) To determine if the development of Sendai virus-induced bronchiolar

fibrosis in BN rats can be inhibited by blocking the activity of a

single highly expressed cytokine identified in the above experiments.



Rationale



TNF-a was chosen to block because: 1) Its transcription is directly induced

by the virus, 2) TNF-a pulmonary mRNA is elevated earlier in both rat strains









25
(3 dpi) than is TGF-13 mRNA, 3) Inhibition of TNF-a has been more effective in

blocking the development of chemically-induced pulmonary fibrosis. The basic

plan was to treat infected BN rats with a 55 kD soluble TNF-a receptor

immunoglobulin G fusion protein (p55 sTNFR-IgG), a treatment which has been

used to successfully block TNF-a activity in animal models of gram-negative

sepsis, and to assess the amount of fibroblast proliferation present at 10 dpi by

counting BrdU labeled fibroblasts.



Design and Methods



Experimental Design. One group of eighteen 25-day-old, male BN rats

were infected with Sendai virus, while a second group of 12 were sham-

inoculated with CAF (see Figure 2 -2). Six rats from the control and 9 from the

virus-inoculated rats were injected intraperitoneally (ip) with 10 mg/kg of p55

sTNFR-lgG1 fusion protein on days 0, 2, 4, 6, & 8 days post inoculation (dpi).

The remaining rats were injected ip with an equivalent volume of human IgG

according to the same schedule. At 7dpi, bronchoalveolar lavage fluid (BAL)

was taken from 3 of the virus-inoculated, p55 sTNFR-IgG1 fusion protein-treated

rats, and 3 of the virus-inoculated, human IgG treated animals. The TNF-a

protein content of this fluid was determined using a mouse TNF-a ELISA kit and

by bioassay. On day 9 post inoculation, 6 rats from each of the four groups were

given 200 pg/g 5-bromo-2' -deoxyuridine (BrdU, 2 mg/ml in PBS) ip 12 hours

before they were sacrificed at 10 dpi.










Tissue processing. All rats were anesthetized with sodium pentobarbital

(approximately 200 pg/g body weight) and were killed by exsanguination via

intracardiac puncture. BAL's were performed on the rats at the time they were

sacrificed, total and differential cell counts and the TNF-a protein content of the

BAL fluid was determined. Lungs were fixed by tracheal perfusion at 30 cm H20

pressure with 4% paraformaldehyde for 2 hours.

BrdU Labeling. Four transverse sections were taken from the fixed left

lungs of rats treated with BrdU: at the hilus, between the hilar section and the

cranial lung margin, and at points equidistant caudal from the hilar section and

cranial from the caudal lung margin. Tissue blocks were embedded in paraffin,

sectioned and immunocytochemical staining to detect BrdU was performed.

Data Analysis. Fibroblast proliferation was assessed by counting the

number of BrdU-labeled fibroblasts (identified by their spindloid shape) in 25

high power (400X) fields per slide. Three slides per animal were counted.

Differences in the density of labeled cells between sham- and virus-inoculated

and p55 sTNFR-lgGI and human IgG treated animals were compared using

ANOVA and Tukey tests.









27

Experiment 3: Assessing the ability of Sendai virus to induce increased cvtokine

gene expression.



Objectives



The original objectives for experiment 3 were as follows:

1) To determine whether Sendai virus induces TGF-iB, TNF-a, andlor PDGF

gene transcription and translation in pulmonary macrophages isolated from

BN rats and whether the induction of cytokine gene transcription is

dependent on protein synthesis, and/or is associated with an increase in

the stability of cytokine mRNA.

2) To determine if pulmonary macrophages isolated from F344 rats have the

same Sendai virus-induced pattern of cytokine expression as BN rats.

Because of the high number of animals required to obtain adequate cell

numbers for these experiments, and the fact that PDGF mRNA levels were not

elevated in the virus-inoculated rats, the objectives were revised:

1) Determine whether Sendai virus directly or indirectly induces gene

expression for TGF-p, and TNF-a in a rat alveolar macrophage cell line

NR8383 by: comparing cytokine mRNA and protein levels in virus infected

NR8383 cells with those incubated with chorioallantoic fluid (CAF).

2) Determine whether live virus is required for any increase in TGF-B, and

TNF-a gene expression observed in virus-inoculated cells.









28

3) Determine whether there is an increase in the stability of cytokine mRNA in

NR8383 cells associated with Sendai virus infection.



Rationale:



Several viruses, including Sendai virus, have been shown to directly

upregulate the production of cytokines.127'131'"'74 Macrophages are major

sources for TGF-11, and TNF-a, and can be infected with Sendai virus. Nuclear

run on assays were attempted (see Appendix), but produced inconsistent

results. The objectives were addressed by using Northern analysis and enzyme-

linked immunoabsorbant assays.



Design and Methods



Baseline Studies. NR8383 cells were infected with Sendai virus at

multiplicity of infection (MOI) of 1:1 and 1:10. The viral titer of culture

supernatants was determined at 4, 12, 24, 48, 72, 96, and 120 hours after

infection by performing plaque assays. These assays were done to verify that

the cells could support a productive viral infection, and to determine the time of

peak viral replication. Immunoperoxidase staining for viral antigens was also

performed on the infected cells to confirm infection.

Northern analysis. Northern analysis for TGF-P, and TNF-a was

performed on cytoplasmic RNA isolated from chorioallantoic fluid (CAF) and











virus-inoculated cells (MOI = 1) at 4, 8, 24, 48, and 72 hours post inoculation.

Blots were re-probed with a "P-labeled probe for human 3-actin to assess the

equality of loading.

Studies with UV-inactivated Sendai virus. UV-inactivated virus was

produced by exposing virus to a 30 watt UV lamp overnight. Plaque assays

were used to confirm viral inactivation. Cells were inoculated with inactivated

virus at an MOI of 1:1, and cytoplasmic RNA extracted 24 hours after virus-

inoculation.

Actinomvcin-D Experiments. Cells were incubated with virus (MOI = 1:1) or

CAF for 1 hour, media containing a final concentration of 5 mg/ml of

actinomycin-D (Sigma, St. Louis, MO) was added, and the cells were incubated

at 37* C, 5% CO2) for 24 hours. Cytoplasmic RNA was extracted and the

relative abundance of TNF-a and TGF-P RNA was determined using Northern

analysis.

Protein Assays. The secretion of TNF-a and TGF-p protein from CAF and

virus-inoculated macrophages was monitored by ELISA.

Data Analysis. Cytoplasmic TNF-a, TGF-p,, and p-actin RNA levels were

evaluated by comparing the relative optical densities (ROD) of bands on

Northern autoradiographs. Comparisons of band densities were only made

between bands in RNA run on the same gel. ROD values, TNF-a and TGF-p,

protein levels were normalized to cell counts. Statistical analysis of group

means was performed by one-way analysis of variance (one-way ANOVA) using








30

a computer-based statistical program (SigmaStat, Jandel Corporation, San

Rafael, CA). Differences between individual means were assessed by

Bonferroni's method of all pairwise multiple comparison."'













Northern Analysis


31



Immunocvtochemistr


Days BN Rats F344 Rats BN Rats/ F344 Rats

After Control Virus Control Virus Control Virus

Inoculation

3 6 6 4

7 6 6 6 6 4 4

10 6 6 4

14 6 6 6 6 4 4

30 6 6 4 4



The blots were re-probed wherever possible



Table 2-1. Experimental design for Northern analysis and immunocytochemical

studies


















m


BN F344 BN F344


S-PDGF



BN


7V 10V 10V 14S 14V 14S 14V


Figure 2-1: Northern analysis of PDGF pulmonary RNA levels in sham- and
virus-inoculated BN and F344 rats. Each lane was loaded with 5 pg of pooled
mRNA from sham- and virus-inoculated rats. No significant differences in band
densities were noted in either rat strain at 7, 10, or 14 days after inoculation
(F344 expression at 7 days is not shown). Based upon these results, PDGF
gene expression was not studied further.






33




BN
Virus-Inoculated Sham-Inoculated
Days 0, 2, 4, 6, 8 Days 0, 2, 4, 6, 8

Con IgG p55 sTNFR- IgG p55 sTNFR- IgG Cn
10 mg/kg 10 mg/kg 10 mg/kg 10 m/kg

TNF assay

10 days
TNF Assay
"6= n=6 BRDU Fibroblast Labelling [n l



Figure 2-2: Experimental design for Experiment 3.














CHAPTER 3
INCREASED TRANSFORMING GROWTH FACTOR B, (TGF-B1) GENE
EXPRESSION PRECEDES SENDAI VIRUS-INDUCED BRONCHIOLAR
FIBROSIS



Summary



Rats infected early in life with parainfluenza type 1 (Sendai) virus develop

bronchiolitis that is followed by long-term increases in airway resistance and

hyperresponsiveness. These virus-induced functional abnormalities are closely

associated with bronchiolar mural fibrosis and persistent increases in

bronchiolar inflammatory cells including macrophages and mast cells. Brown

Norway (BN) rats are highly susceptible to virus-induced bronchiolar fibrosis and

airway function abnormalities, whereas Fischer-344 (F344) rats are resistant.

We hypothesized that bronchiolar fibrosis was regulated by the production of

TGF-p, by bronchiolar mural macrophages during the acute and chronic

inflammatory response induced by virus. Pulmonary and bronchiolar TGF-B,

gene expression was studied in BN and F344 rats following Sendai virus

infection by combined methods of Northern analysis and in situ hybridization for

TGF-B, mRNA and by immunocytochemistry for protein. Steady-state pulmonary

TGF-P, mRNA was increased in both BN and F344 rats at 7 days after

34










inoculation (p<0.001). BN rats had over two-fold greater increases than F344

rats (p<0.05). At 10 and 14 days after inoculation, pulmonary TGF-B, mRNA

remained elevated in BN rats, but not in F344 rats (p<0.001). TGF-B, mRNA

was identified by in situ hybridization in bronchiolar mural macrophages and

fibroblasts. Seven days after inoculation, BN rats had over two-fold greater

numbers of bronchiolar mural macrophages positive for TGF-B, mRNA than did

control BN rats (p<0.05). Pulmonary TGF-R, protein expression was assessed

by immunocytochemistry at 3, 7, 10, 14, and 30 days after inoculation. BN rats

had over a two-fold increase in numbers of bronchiolar macrophages expressing

TGF-B, protein at 7 days after inoculation as compared to both sham-inoculated

BN and virus-inoculated F344 rats (p<0.0001). At 10 days after inoculation,

virus-inoculated BN rats had three-fold increases in numbers of bronchiolar

macrophages and fibroblasts expressing TGF-B, protein as compared to the

control animals (p<0.0001). We conclude that virus-induced bronchiolar TGF-11

gene expression precedes bronchiolar mural fibrosis and may play an important

pathogenetic role in virus-induced bronchiolar fibrosis and airway function

abnormalities.



Introduction



Clinical and epidemiological studies have linked severe viral bronchiolitis

during infancy with subsequent long-term pulmonary function abnormalities








36
including increased airway resistance and hyperresponsiveness.25 Viral lower

respiratory tract disease is also identified as a risk factor in the development of

asthma.3,6 However, it remains to be determined what direct role viral infection

plays in the pathogenesis of pulmonary function abnormalities and what genetic

factors may be contributing to adverse sequelae from viral and/or immunologic

lung injury during early life.e

Parainfluenza type 1 (Sendai) virus infection in young rats has been

developed as an animal model of bronchiolitis in children. Sendai virus induces

acute necrotizing bronchiolitis and pneumonia that is followed by persistent

alterations in pulmonary function. Functional alterations include elevated

respiratory resistance, decreased dynamic compliance, and

hyperresponsiveness to methacholine. "361 37 38. Susceptibility to virus-induced

functional abnormalities has been demonstrated to be influenced both by the

age and by the strain of the rat. In rats inoculated with virus as 25 day old

weanlings, persistent airway lesions associated with long-term pulmonary

function abnormalities include bronchiolar mural fibrosis and inflammation

characterized by mural aggregates of macrophages, mast cells, and

eosinophils.3 Both BN and LEW rats appear to be more susceptible to virus-

induced alterations in pulmonary structure and function than are F344 rats.36.'8

37, 835 Comparative viral pathogenesis studies with virus-susceptible BN rats

and resistant F344 rats indicate that genetic differences between the strains

influence important factors in host-virus interaction including duration of viral









37

replication, magnitude of immune response, and severity of acute and persistent

inflammatory cell response." 37. W

Bronchiolar mural fibrosis and associated airway wall thickening is a

prominent feature of virus-induced lung injury that is associated with lung

dysfunction in susceptible weanling BN rats.3 Fibrosis develops as early as 15

to 30 days after viral inoculation. Airway wall thickening and fibrosis has also

been identified as a common lesion occurring in human asthma and may

contribute to exaggerated airway luminal obstruction following modest smooth

muscle contraction."224

A growing body of research has demonstrated that several cytokines,

including transforming growth factor-p (TGF-1), tumor necrosis factor-a (TNF-a),

and platelet-derived growth factor ( PDGF), play critical roles in regulating

pulmonary fibrosis following alveolar injury by toxins such as bleomycin or in

less well-defined spontaneous fibrosis.4'1, 'o', 1. 84 Furthermore, there is

evidence that genetic susceptibility to develop pulmonary fibrosis in response to

chemically-induced damage may be controlled at the level of TGF-P

expression.4 TGF-p upregulates both fibroblast proliferation and the synthesis

of collagens and glycosaminoglycans in the lung." '"1

Based on the preceding information, we hypothesized that virus-induced

airway fibrosis and thickening in virus-susceptible BN rats was due to greater

expression of TGF-3 in airway walls following viral infection than in virus-

resistant F344 rats. The objectives of this study were to determine if Sendai









38
virus infection induces increased expression of TGF-B, in bronchiolar walls prior

to fibrosis and to determine whether TGF-P, gene expression induced by viral

bronchiolitis is greater in BN rats that are susceptible to virus-induced fibrosis

than in F344 rats that are resistant to this virus-induced effect.



Materials and Methods

Animals



Twenty-two day old weanling male BN and F344 rats were purchased from

Harlan Sprague Dawley Inc., Indianapolis, IN. They were negative for serum

antibodies to Sendai virus, pneumonia virus of mice, mycoplasma, Kilham rat

virus, and sialodacryoadenitis virus/rat corona virus, and for respiratory aerobic

bacteria. The control and experimental groups were separated and housed in

adjacent isolation cubicles.



Experimental Protocol



Three days after arrival, the rats in the experimental group were inoculated

with parainfluenza type 1 (Sendai) virus at a concentration of 1-3 plaque forming

units per ml of gas in an aerosol exposure apparatus (Tri-R Instruments,

Rockville Center, NY) for 15 minutes. Viral assays to calibrate exposure were

performed as previously described.332 Sendai virus strain P3193 produced in










embryonated chicken eggs was used. Control rats were exposed to sterile

chorioallantoic fluid also harvested from embryonated chicken eggs. Virus-

inoculated BN and F344 rats were processed at 3, 7, 10, 14 and 30 days after

inoculation. Sham-inoculated animals were sacrificed at 7, 14, and 30 days after

inoculation. The rats were deeply anesthetized with sodium pentobarbital and

killed by exsanguination via intracardiac puncture. For Northern analysis, the

lungs from six rats from each group were frozen in liquid nitrogen and stored at

-80*C until processed. Lungs from four animals in each group were processed

for in situ hybridization and immunohistochemistry. These lungs were fixed for 2

hours via airway perfusion at 30 cm fixative pressure with 4%

paraformaldehyde-phosphate buffered saline (pH 7.4) and embedded in paraffin.



Northern Analysis



Lung tissue was lysed in guanidine thiocyanate, and total RNA was extracted,

precipitated and further processed to recover polyadenylated mRNA as

described.1' Twenty pg of total RNA from each lung or 5 pg of mRNA extracted

from pooled lung tissue was electrophorectically separated on 1.2% (wt:vol)

agarose gels containing 6% (vol:vol) formaldehyde, transferred to a positively

charged nylon membrane (Boehringer Mannheim Corporation, Indianapolis, IN),

and fixed by UV-crosslinking. Equality of RNA loading between lanes was

evaluated by visually comparing the intensity of ethidium bromide staining of the










18s and 28s ribosomal RNA bands on gels and membranes. A 985 bp cDNA

fragment containing the major coding region of the rat TGF-B, gene (supplied by

Dr. Su Wen Quian, Laboratory of Chemoprevention, NIH, Bethesda, MD) and a

2.6 kb human cDNA fragment of platelet derived growth factor (PDGF)

(American Type Culture Collection, Rockville, MD) were labeled with 32P-dCTP

using the random primer method as described.4' Prehybridization (15 min,

60-65"C) and hybridization (overnight, 65C) was in 0.5 M Na2HPO4, 7% SDS,

1% BSA, and 10mM EDTA as described.'41 The membranes were washed at the

hybridization temperature in 20mM phosphate, 1% SDS, and used to expose

Kodak XAR-5 film with Dupont Cronex intensifying screens. Film band densities

were quantified using a computerized video image analysis unit (MicroComputer

Imaging Device, Imaging Research Inc., St. Catharines, Ontario).



In Situ Hybridization and Immunocytochemistry



For in situ hybridization, the rat TGF-B, cDNA probe in pBluescript II KS+

(Stratagene, La Jolla, CA) was linearized and used as a template for the

production of "S-labeled RNA sense and antisense probes by a modified

version of the Riboprobe Gemini System II kit (Promega Biotech, Madison,

WI).142 To synthesize the antisense probe, plasmids were linearized with Xba I

and incubated with T3 polymerase. The sense probe was made by linearizing

with Hind III and incubating with T7 polymerase. Lung sections were









41
prehybridized, hybridized, and washed as described.142 Briefly, the slides were

initially de-paraffinized in xylene, re-hydrated through a graded series of ethanol

washes, and denatured in 0.2 N HCI and in 2X SSC at 70C. Prehybridization

consisted of incubation in 0.001% proteinase K (30 min, 37C), acetylation in

acetic anhydride, and dehydration in ascending concentrations of ethanol.

Hybridization was performed at 57C overnight in a humidified chamber. The

slides were washed in 4X SSC, incubated in RNase A (30 min, 37"C) and

washed in descending concentrations of SSC. A final high stringency wash was

performed using 0.1X SSC at 65C for 30 min. The slides were then dehydrated

and dried. Signal was detected by coating the slides with Ilford K.5D nuclear

emulsion (Polysciences, Inc., Warrington, PA), and development in Kodak D-19

developer (Kodak, Rochester, NY) after 7-14 days. 42 Autoradiograms were

stained with hematoxylin and eosin. Specificity of the probes was determined by

comparison with slides hybridized with labeled sense transcripts. Only lungs

from BN rats were processed for in situ hybridization.

For immunocytochemistry, lung sections from BN and F344 rats were de-

paraffinized and re-hydrated as described above for in situ hybridization, washed

in 0.01 M Tris-buffered saline, blocked with normal goat sera, and incubated with

either chicken IgG against recombinant human TGF-B, (R&D Systems,

Minneapolis, MN), or normal chicken IgG (Sigma, St. Louis, MO), both at 0.01

mg/ml. Antibody binding was detected using an avidin-biotin alkaline

phosphatase system (Vectastain ABC-AP and substrate kits, Vector Lab,

Burlingame, CA) as described.81










Quantitation Methods and Data Analysis



Evaluation of pulmonary cytokine mRNA levels in control and virus-

inoculated BN and F344 rat strains was made by comparing the relative optical

densities (ROD) of bands on Northern autoradiograms. The density of cells

expressing mRNA and protein was assessed by counting and classifying the

number of labeled cells per millimeter of bronchiolar basement membrane in 25

consecutive fields at 400X per slide." Macrophages and fibroblasts were the

cells most consistently labeled by the TGF-P, immunocytochemistry and in situ

hybridization procedures. Vascular and airway smooth muscle cells also

contained TGF-P, mRNA and protein, while epithelial cells were occasionally

positive for TGF-p, protein. Macrophages and fibroblasts were counted in the

bronchiolar wall. In addition, macrophages in immediately adjacent alveolar

spaces were counted in the same microscopic fields. Three lung sections from

each rat were evaluated in the immunocytochemistry procedures. In more

limited quantitative procedures for the in situ hybridization of BN rats, one slide

per rat was counted. Cells that had round profiles, central round to ovoid nuclei,

and abundant eosinophilic cytoplasm were counted as macrophages. Cells

were counted as fibroblasts when they had elongated ovoid nuclei and

elongated, spindle-shaped cytoplasm. Statistical analysis among group means

was performed using ANOVA and Tukey's HSD multiple comparison

procedure'" in a computer-based statistical program (Systat version 4,

Evanston, IL).














Effect of Sendai virus infection on total lung levels of TGF-B, mRNA



At 7 days after inoculation, viral infection resulted in significant increases in

pulmonary TGF-1, mRNA in both rat strains (Figures 3-1 and 3-2). Virus-

susceptible BN rats had a greater than two-fold increase in TGF-B, mRNA as

compared to the peak elevation observed in the virus-resistant F344 rats

(p<0.05). At 10 and 14 days after inoculation, pulmonary TGF-B, mRNA levels

remained elevated in the virus-inoculated BN rats, whereas in virus-inoculated

F344 rats the levels were not significantly increased over those of control rats.

The level of pulmonary TGF-B, mRNA in lungs of virus-inoculated BN rats was

equivalent to that of the control animals by 30 days after inoculation. Pulmonary

PDGF mRNA was not increased in either rat strain at 7, 10, & 14 days after

inoculation (see Figure 2-1).



Identification of pulmonary cell types expressing TGF-B, mRNA



Macrophages and fibroblasts were the main cell types expressing TGF-1,

mRNA as detected by in situ hybridization in both control and virus-inoculated

rats (Figure 3-3). Macrophages in bronchiolar walls and alveolar macrophages

hybridized antisense probe. Bronchiolar smooth muscle cells as well as arterial










and venous smooth muscle cells were also positive for TGF-P, mRNA. In

control BN rats, the density of bronchiolar mural macrophages positive for TGF-

B, mRNA was 6.3 1.1 (mean sem) per mm bronchiolar basement membrane.

The density of alveolar macrophages positive for mRNA was 70.7 7.2 per mm

(data not shown). At 7 days after inoculation, there were approximately two-fold

increases in the density of bronchiolar mural macrophages positive for TGF-B,

mRNA as compared to control rats (11.7 1.8 per mm, p<0.05), and over 30-fold

increases in labeled alveolar macrophages (197.3 13.6 per mm, p<0.005)

(data not shown). Mean numbers of labeled macrophages at 10 days after

inoculation and labeled fibroblasts at 7 or 10 days after inoculation did not differ

significantly from controls (data not shown). Non-specific labeling was assessed

in tissue sections hybridized with the sense riboprobe.



Pulmonary TGF-B1 protein expression



Seven days after viral inoculation there were significantly increased

numbers of macrophages expressing TGF-1, protein in the lungs of both BN and

F344 rats (Figures 3-4 & 3-5). In infected BN rats at 7 days after inoculation, the

increase was ten-fold (p<0.0001) over control BN rats and greater than two-fold

(p<0.0001) over the number observed in infected F344 rats at 7 days after

inoculation. At 10 days after inoculation, BN rats had three-fold increases in

macrophages and fibroblasts expressing TGF-B, protein as compared to










controls (p<0.0001). At 10 days after inoculation in F344 rats, the numbers of

macrophages and fibroblasts positive for TGF-131 were not significantly different

than sham-inoculated animals (Figures 3-5 & 3-6). TGF-P, protein was also

detected in bronchiolar, venous, and arterial smooth muscle in both BN and

F344 rats.





Discussion



The goal of this study was to determine whether TGF-P, expression induced

by viral bronchiolitis might be an important stimulus for bronchiolar mural fibrosis

and thickening. The bronchiolar fibrosis that follows acute virus-induced injury

and inflammation is associated with long-term airway dysfunction 36 and,

therefore, is an important pathologic endpoint to investigate. Two main

observations from this study are consistent with the interpretation that increased

TGF-P, gene expression may be an important stimulus for virus-induced

bronchiolar fibrosis. First, increases in pulmonary steady-state mRNA levels

and in density of bronchiolar cells expressing mRNA and protein precede the

bronchiolar fibrosis that develops by 15 or 30 days following inoculation.3 38

Second, BN rats (high responders to Sendai virus-induced bronchiolar fibrosis)

have higher and more prolonged levels of increased TGF-P, protein and mRNA

expression in lung and airways than do F344 rats (low responders).3









46
Although the results are consistent with TGF-P, being an important regulator

of virus-induced bronchiolar fibrosis, they do not delineate the specific

mechanism for increased gene expression. Increased quantities of TGF-P,

mRNA in whole lung homogenates were associated with increased numbers of

macrophages and fibroblasts in airway walls and surrounding alveoli expressing

cytoplasmic mRNA and protein. Increased mRNA levels in lung could be due to

increased infiltration of cells containing TGF-P, mRNA, virus-induced gene

transcription and protein expression in macrophages and fibroblasts, or a

combination of the two mechanisms. Although the data do not clearly

distinguish the relative importance of these mechanisms, several findings

suggest that the increased TGF-P, mRNA and protein expression in cells around

virus-infected bronchioles may be due at least in part to virus-induced increases

in gene transcription and/or protein synthesis.

First, increased numbers of peribronchiolar fibroblasts expressing TGF-0P

protein were observed at 7 days after inoculation and prior to appreciable

increases in numbers of bronchiolar mural fibroblasts. Increases in fibroblast

numbers are observed at 15 to 30 days after viral inoculation. Second, if all the

increases in TGF-0, mRNA were due to infiltration of macrophages, we would

expect to see similar increases in mRNA for PDGF, another cytokine associated

with macrophages.4 Pulmonary steady-state levels of PDGF mRNA did not

increase following infection in either BN or F344 rats at 7, 10, and 14 days after

inoculation (Figure 2-1).










Transcriptional upregulation of TGF-11, gene expression could be a direct

result of viral infection. The ability of viruses to directly increase expression of

inflammatory cytokines is considered to be an important mechanism in the

pathogenesis of several viral diseases. 105.57 In acquired immunodeficiency

syndrome (AIDS), a viral protein (tat) appears to directly upregulate the

production of TGF-B.1"0 More complex viral effects on host cell gene regulation

are also recognized. Influenza A virus infection of murine macrophages not only

increases transcription of tumor necrosis factor-a (TNF-a), but also induces the

production of a unique TNF-a transcript.1' Sendai virus infection of cultured

macrophages has been shown to directly stimulate the production of TNF-a.57 In

Sendai virus infected rats, pulmonary viral titers peak around 5 days after

inoculation,3 before the peak expression of TGF-B at 7 days after inoculation.

Direct upregulation of TGF-P, gene transcription and protein synthesis by

Sendai virus infection could explain some of the differences in the pathologic

response of BN and F344 rats to viral infection. Pulmonary Sendai viral

replication persists for several days longer in weanling BN rats than in F344

rats.37 Persisting viral replication could be an important stimulus of persisting

cytokine expression and more severe fibrosis in BN rats.

TGF-1, is a major regulator of fibrosis and other tissue repair mechanisms. It

contributes to increasing tissue levels of collagen and other matrix proteins by

several mechanisms. TGF-1, increases transcription of fibronectin and

procollagen genes", post-transcriptionally regulates fibroblast, collagen and










fibronectin mRNA," and inhibits the degradation of existing extracellular

matrix.71 Studies of naturally occurring and chemically-induced fibrotic lung

lesions indicate that TGF-B, is an important mediator of pulmonary fibrosis.4 ,'"
48, -oo.08. In this study, the prolonged upregulation of TGF-B, in the susceptible

rats implies that such increases may also be important in the regulation of virus-

induced bronchiolar fibrosis and airway function abnormalities.

In addition to its regulatory effects on repair and fibrosis, TGF-B, can

suppress both T- and B-lymphocytes, decrease cytotoxic and lymphokine-

activated killer cell production, and down-regulate macrophage activity.925 In

several models of infectious disease, these immunosuppressive effects have

resulted in direct correlations between the level of TGF-B, expression, the

prolonged survival of infectious agents and the extent of tissue damage. 9'2-102.

103 It is reasonable to speculate that prolonged and excessive TGF-P,

expression in infected BN rats may be contributing to less effective immunity and

to virus-induced tissue damage and abnormal repair marked by fibrosis.

In conclusion, the results of this study demonstrate that virus-induced

bronchiolar TGF-B, gene expression precedes bronchiolar mural fibrosis and

suggest that it may play an important pathogenetic role in virus-induced

bronchiolar fibrosis and airway function abnormalities.
















3- E



8 l. We
* I00a)

'o-






S4)














E E
I~_ .a

o o
-a-









-: E-C
<-


c2 .E -z

.:, A oo -,E











e O F344
S* p< 0.001
E A BNp



S 4
O
0




Z

0' o' '-------
0 2 4 6 8 10 12 14 30
Days After Inoculation



Figure 3-2: Graphic presentation of the results of Northern analysis for
TGF -P1. Data points and bars represent the mean relative optical band
density (ROD) and standard error of the mean (SEM) for 4-6 rats.
Statistical comparisons of ROD's were only made between sham- and
virus-inoculated groups represented on the same blot.









































I-












Figure 3-3. TGF-1B mRNA in lung sections was detected by in situ hybridization.
Top panel: brightfield and darkfield micrographs from a sham-inoculated BN rat.
The bottom panel shows a bronchiole from an infected BN rat at 10 days after
inoculation. (x200).






















Irbi- AW I S
Vd. -- ^




















Figure 3-4. TGF-a, protein in bronchiolar macrophages. Both micrographs are
of sections taken from a virus-inoculated BN rat at 7 days after inoculation. The
top was incubated with 0.01 mg/ml normal chicken IgG and the bottom with 0.01
mg/ml chicken anti-human TGF-I, IgG (x600).














Sp< 0.0001
og 0 F344 Control
2 g V F344 Infected
0 f BN Control
S40 \ BN Infected

*E n- p< o.ooo1



z

0 5 10 18 20 25 0
Days After Inoculation



Figure 3-5: Density of bronchiolar mural macrophages expressing
TGF-01 protein. Data points represent mean + SEM (n=4).















L* p< 0.0001
a 0 F344 Control
2 / \ V F344 Infected
S\ BN Control
0 y BN Infected
m


0 -
2-


zM

0 5 10 15 20 25 30
Days After Inoculation




Figure 3-6: Density of bronchiolar mural fibroblasts expressing TGF-P1
protein. Data points represent mean + SEM (n=4).














CHAPTER 4
INCREASED TUMOR NECROSIS FACTOR-a (TNF-a) GENE EXPRESSION IN
MACROPHAGES PRECEDES BRONCHIOLAR FIBROSIS IN VIRUS-
SUSCEPTIBLE BN RATS


Summary


Parainfluenza type I (Sendai) virus infection in genetically susceptible young

rats induces persistent increases in airway resistance and airway

hyperresponsiveness. In virus-susceptible BN rats, the pulmonary function

abnormalities are associated with bronchiolar mural fibrosis and recruitment and

persistence of inflammatory cells including mast cells and macrophages. We

hypothesized that TNF-a gene expression is an important regulatory event in

virus-induced bronchiolar fibrosis. Pulmonary TNF-a mRNA and protein

expression were assessed at 3, 7, 10, 14, and 30 days after virus-inoculation in

rat strains that are susceptible (BN) and resistant ( F344) to virus-induced

bronchiolar fibrosis. Northern analysis revealed elevated pulmonary TNF-a

mRNA levels in both rat strains at 3, 7, and 10 days after inoculation (p<0.01).

At 14 days after inoculation only the virus-susceptible BN rats had increased

levels of TNF-a mRNA in their lungs (p<0.01). Macrophages and fibroblasts

were identified as being the main bronchiolar cell types containing TNF-a mRNA










and protein by in situ hybridization and immunocytochemistry. Macrophage

expression of TNF-a protein peaked at 7 days after inoculation in both rat

strains (p<0.0001). Virus-inoculated BN rats had almost twice the density of

TNF-a-positive macrophages in their bronchiolar walls as infected F344 rats

(p<0.0001). The number of macrophages positive for TNF-a protein remained

significantly elevated in infected BN rats at 10 dpi (p<0.05). Over two-fold

increases in the number of labeled bronchiolar fibroblasts were observed in

infected rats of both strains at 7 dpi (p<0.001) compared to the control rats. At

10 dpi the mean number of fibroblasts positive for TNF-a protein had increased

to over 5-fold that of the control in the infected BN rats (p<0.0001), but was not

significantly elevated in the infected F344 rats. The data suggest that increased

TNF-a gene expression may be important in the pathogenesis of virus-induced

bronchiolar fibrosis.





Introduction



Viral bronchiolitis in early childhood has been linked with persisting

pulmonary functional abnormalities including increased airway resistance and

airway hyperresponsiveness and has been identified as a risk factor for the

development of asthma.2" The development of asthma-like symptoms following

viral infection appears to be associated with an aberrant, IgE-dominated








57

immunological response that is probably determined by a combination of genetic

and viral factors.2" Young rats develop similar pulmonary functional

abnormalities after infection with parainfluenza virus type I (Sendai) and are

being studied as an animal model to identify genetic and viral factors important

in the pathogenesis of virus-induced lung injury. 1, .35

The development of virus-induced pulmonary function abnormalities in rats is

associated with infection as neonates or weanlings and genetic susceptibility.

Brown Norway (BN) rats develop virus-induced alterations in pulmonary

structure and function, while Fischer 344 (F344) rats are relatively resistant."'"

' Comparative viral pathogenesis studies indicate there are differences in

duration of viral replication, magnitude of immune response, and severity of both

the acute and persistent inflammatory response between the virus-susceptible

BN and virus-resistant F344 rats.3" The virus-induced abnormalities in

pulmonary function consist of increased respiratory resistance, decreased

dynamic compliance and hyperresponsiveness to methacholine. These

functional abnormalities are associated with bronchiolar aggregates of

macrophages, mast cells, and eosinophils, and bronchiolar mural fibrosis.3

Bronchiolar mural fibrosis and the resulting airway wall thickening is

prominent in BN rats 15 to 30 days after being infected with Sendai virus.

Bronchial fibrosis is a common finding in human asthmatics.22-24 4 The

importance of mural fibrosis in the development of increased airway resistance is

indicated by airway modeling studies that demonstrate that moderate amounts of










airway wall thickening markedly alter mechanical properties of airways and

accentuate airway luminal narrowing during smooth muscle contraction.24

Several cytokines including tumor necrosis factor-a (TNF-a), transforming

growth factor-3 (TGF-P), and platelet-derived growth factor (PDGF) stimulate

pulmonary fibroblast proliferation and collagen synthesis and are upregulated in

both naturally occurring and chemically-induced pulmonary fibrosis.40'6 "-'0, 85,
84.11820 Pulmonary expression of TGF-P, and PDGF during Sendai virus

infection has been assessed in both virus-susceptible BN and virus-resistant

F344 rats (see Chapter 3 and Figure 2-1). Virus-induced increases in

pulmonary PDGF mRNA were not observed. However, prolonged increases in

pulmonary TGF-P, mRNA, as well as increased numbers of bronchiolar

macrophages and fibroblasts expressing TGF-P, protein were present in infected

BN rats before the development of virus-induced fibrosis.

Increased pulmonary expression of TGF-P appears to be important in the

pathogenesis of some, but not all forms of pulmonary fibrosis. The

administration of antibodies to TGF-1 significantly reduced bleomycin-induced

increases in lung collagen." In contrast, pulmonary TGF-1 mRNA was not

increased in lungs during the development of silica-induced pulmonary

fibrosis.'20 TNF-a has been shown to be increased in both silica-induced

pulmonary fibrosis and bleomycin-induced fibrosis."' 120 The importance of

TNF-a in these two forms of fibrosis further emphasized by studies showing that

TNF-a inhibition with antibody reduces severity of fibrosis.11 "2 Furthermore,










the infusion of recombinant TNF-a accentuated the severity of pulmonary

fibrosis in silica-treated mice.12

Based upon these observations, we hypothesized that the development of

Sendai virus-induced bronchiolar fibrosis in susceptible BN rats also may be

regulated by bronchiolar expression of TNF-a. The objectives of this study were

to determine if Sendai virus infection induces increases in pulmonary TNF-a

mRNA levels prior to the development of fibrosis and to determine whether rats

susceptible to virus-induced bronchiolar fibrosis (BN) have increased numbers

of bronchiolar cells expressing TNF-a protein compared to the virus-resistant

(F344) rats.



Materials and Methods


Animals



Twenty-two day old weanling male BN and F344 rats were purchased from

Harlan Sprague Dawley Inc., Indianapolis, IN. They were negative for serum

antibodies to Sendai virus, pneumonia virus of mice, mycoplasma, Kilham rat

virus, and sialodacryoadenitis virus/rat corona virus, and for respiratory aerobic

bacteria. The control and experimental groups were separated and housed in

adjacent individually ventilated isolation units.
















Three days after arrival, the rats in the experimental group were exposed to

parainfluenza type I (Sendai) virus at a concentration of 1-3 plaque forming

units (PFU) per ml of gas for 15 minutes in a Tri-R aerosol exposure apparatus

(Tri-R Instruments, Rockville Center, NY). Viral assays to calibrate exposure

were performed as previously described."37 Sendai virus strain P3193

produced in embryonated chicken eggs was used. Control rats were exposed to

sterile chorioallantoic fluid. Virus-inoculated BN and F344 rats were processed

at 3, 7, 10, 14 and 30 dpi. Sham inoculated animals were sacrificed at 7, 14,

and 30 dpi. The rats were deeply anesthetized with sodium pentobarbital and

killed by exsanguination via intracardiac puncture. For Northern analysis, the

lungs of six rats from each group were frozen in liquid nitrogen and stored at

-80*C until processed. Lungs from four animals in each group were processed

for in situ hybridization and immunohistochemistry. These lungs were fixed for 2

hours via airway perfusion with 4% paraformaldehyde-PBS (pH 7.4), at 30 cm

perfusion pressure, and embedded in paraffin for in situ hybridization and

immunocytochemistry.










Northern Analysis



Lung tissue was lysed in guanidine thiocyanate, and total RNA extracted and

precipitated as described.1' Twenty micrograms (pg) of total RNA from each

lung was electrophorectially fractionated on a series of 1.2% (wt:vol) agarose

gels containing 6% (vol:vol) formaldehyde, transferred to a positively charged

nylon membrane (Boehringer Mannheim Corporation, Indianapolis, IN), and

fixed by UV-crosslinking. Equality of RNA loading between lanes was assessed

by visually comparing the intensity of ethidium bromide staining of ribosomal

RNA bands in the gels and membranes. A cDNA probe for rat TNF-a (supplied

by Dr. T. Shirai, Laboratory for Chemical Research, Asahi Chemical Industry Co.

Ltd, Tagata-gun, Shizuoka, Japan) was labeled with "P-dCTP using the random

primer method as described.14 Prehybridization (15 min, 58C) and

hybridization (overnight, 58C) was in 0.5 M Na2HPO4, 7% SDS, 1% BSA, and

10mM EDTA as described.141 The membranes were washed at the hybridization

temperature in 40mM phosphate, 1% SDS, and exposed to Kodak (Rochester,

NY) XLS-5 film with Dupont Cronex intensifying screens. Film band densities

were quantified using a computerized video image analysis unit (MicroComputer

Imaging Device, Imaging Research Inc., St Catharines, Ontario).










In Situ Hybridization/Immunocytochemistry



In situ hybridization was performed by subcloning the rat TNF-a cDNA probe

into pBluescript II KS+ (Stratagene, La Jolla, CA) as a template for the

production of digoxigenin-labeled RNA probes. Sense and antisense riboprobes

were made according to manufacturer's instructions (The Genius System,

Boehringer Mannheim, Indianapolis, IN). Lung sections were initially de-

paraffinized in xylene, re-hydrated through a graded series of ethanol washes,

and denatured in 0.2 N HCI. Prehybridization treatment consisted of: incubation

in 0.001% proteinase K (30 min, 37C), acetylation in acetic anhydride, and

dehydration in ascending concentrations of ethanol. The probe was added to

the hybridization solution (62.5% formamide, 12.5% dextran sulfate, 0.3 M NaCI,

0.025X Denhart's solution [Sigma, St. Louis, MO] 12.5 mM Tris-HCI pH 8.0, 1.25

mM EDTA pH 8.0) and denatured at 65* C for 10 minutes. Approximately 50 ng

of probe was added to each slide, coverslips were applied and the slides were

placed in a 65 C oven for 10 minutes. Hybridization was then performed at

37C overnight in a humidified chamber. The slides were washed in 2 X SSC

(10 min, 37"C), 1 X SSC (10 min, 37C), and 0.1 X SSC (10 min, 37"C). The

sections were blocked in 2% normal sheep sera (30 min, room temperature) and

incubated (2 hours, room temperature) with an alkaline phosphatase-conjugated

antibody to digoxigenin (Boehringer Mannheim, Indianapolis, IN). Signal was

detected using nitroblue tetrazolium (4-8 hours, room temperature) (Boehringer










Mannheim, Indianapolis, IN). Specificity of the probes was determined by

comparison with slides hybridized with equivalent quantities of labeled sense

transcripts.

For immunocytochemistry, lung sections from BN and F344 rats were de-

paraffinized and re-hydrated as described above for in situ hybridization, washed

in 0.01 M Tris-buffered saline, blocked with normal goat sera, and incubated with

either rabbit sera (Sigma, St. Louis, MO) or polyclonal rabbit anti-mouse TNF-a

(Genzyme, Cambridge, MA) both at 0.001 mg/ml. Antibody binding was

detected using an avidin-biotin alkaline phosphatase system (Vectastain ABC-

AP and substrate kits, Vector Lab, Burlingame, CA).



Protein Analysis on BAL Fluid



Bronchioalveolar lavage fluid (BAL) was collected from sham- and virus-

inoculated BN and F344 rats at 7 and 10 dpi using two phosphate-buffered

saline lavages as described.5 The amount of rat TNF-a protein in each sample

was determined by enzyme-linked immuno-absorbance assays (ELISA) using an

immunoassay kit according to the manufacture's instructions (BioSource

International, Inc. Camarillo, CA).










Quantitation Methods and Data Analysis



Pulmonary TNF-a mRNA levels were evaluated by comparing the relative

optical densities (ROD) of the bands on Northern autoradiographs.

Comparisons of TNF-a bands detected in pulmonary RNA were only made

between bands in RNA run on the same gel. ROD's of the bands were

expressed as percentages of the band density of a positive standard (30 pg of

pooled pulmonary poly-A mRNA from Sendai virus-inoculated BN rats harvested

at 7 dpi) run on each gel. The density of cells expressing TNF-a protein was

assessed by counting and classifying the number of labeled cells per millimeter

of bronchiolar basement membrane in 25 high power (400X) fields per slide.

Round cells with central round to ovoid nuclei and abundant cytoplasm were

counted as macrophages, cells with elongated ovoid nuclei and spindle-shaped

cytoplasm were counted as fibroblasts, and culsters of round cells with central

round nuclei and sparse cytoplasm were identified as lymphocytes. Labeled

macrophages in the bronchiolar wall and immediately adjacent alveolar spaces

and labeled fibroblasts in the bronchiolar wall were counted in the same

microscopic field. Three slides from each of 4 animals per group were counted.

Statistical analysis of group means was performed using ANOVA and Tukey's

HSD multiple comparison procedurel" in a computer-based statistical program

(Systat version 4).










Results



Sendai virus infection causes an increase in total lung levels of TNF-a mRNA



TNF-a transcripts were not detected in pulmonary mRNA extracted from

sham-inoculated animals. Viral infection resulted in increased total lung levels

of TNF-a mRNA in both rat strains at 3, 7, and 10 days after inoculation

(p<0.01). Pulmonary TNF-a mRNA levels peaked at 3 dpi in the F344 rats and

at 7 dpi in the BN rats. Virus-susceptible BN rats still had increased levels of

TNF-a mRNA in their lungs 14 days after inoculation (p<0.01), while in the virus-

resistant F344 rats levels were not significantly above those of the sham-

inoculated rats. Pulmonary TNF-a mRNA levels in the BN rats did not return to

baseline levels until 30 days after inoculation (Figures 4-1 & 4-2).



Identification of pulmonarv cell types expressing TNF-a mRNA and protein



In situ hybridization identified macrophages and fibroblasts as the main

bronchiolar cells expressing TNF-a mRNA in infected BN rats (Figure 4-3).

Epithelial cells and occasional scattered aggregates of lynmphocytes around

bronchi (BALT) were also labeled in sections hybridized with the anti-sense

TNF-a riboprobe.









66
The number of bronchiolar macrophages positive for TNF-a protein peaked

at 7 days after inoculation in both rat strains (p<0.0001, Figures 4-4 & 4-5).

Infected virus-sensitive BN rats had almost twice as many labeled macrophages

in their bronchiolar walls as infected F344 rats (p<0.0001). The number of

labeled macrophages was still elevated in infected BN rats at 10 days after

inoculation (p<0.05). The number of bronchiolar fibroblasts positive for TNF-a

protein increased over two-fold in both rat strains at 7 days after inoculation

(p<0.001, Figure 4-6). The largest number of labeled fibroblasts observed in the

virus-resistant F344 rats was at 7 dpi, while in the virus-susceptible BN rats, the

numbers increased until 10 days after inoculation (p<0.0001, Figure 4-6). TNF-

a protein was also observed in epithelial cells and in BALT lymphocytes.



Effects of Sendai virus infection on TNF-a protein levels in BAL fluid



TNF-a protein levels in the BAL fluid harvested from infected BN rats at 7 dpi

were elevated at 346.7 47.5 pg/ml (mean sem) compared to 55.83 36.38

pg/ml for infected F344 rats (p<0.05). At 10 dpi, the levels of TNF-a protein

were not significantly different between the two rat strains (BN = 104.7 24.3

pg/ml, F344 = 74.5 41.5 pg/ml).














The deposition of collagen in bronchiolar walls increases wall thickness and

may be a major factor accounting for persistent increases in airway resistance

observed in BN rats following viral infection.""1 Two main observations of this

study support a potential role for TNF-a in the pathogenesis of this virus-

induced airway fibrosis. First, pulmonary TNF-a mRNA and protein expression

are elevated before the development of bronchiolar fibrosis at 14-30 days after

virus inoculation." Second, although increases in TNF-a mRNA and protein

were induced by viral infection in both rat strains, higher and more prolonged

TNF-a gene expression was observed in the rats that develop bronchiolar

fibrosis (BN) compared with that in the rats that do not (F344).

These observations are similar to those made in studies of chemically-

induced airway fibrosis in which susceptibility and resistance to the development

of bleomycin-induced pulmonary fibrosis in mice was associated with the level of

TNF-a expression. As was observed in the virus-susceptible and -resistant rats

in this study, mice with a high level of TNF-a expression after exposure

developed fibrosis, while those exhibiting a low level did not.109 The significance

of elevated TNF-a mRNA and protein in the pathogenesis of pulmonary fibrosis

has been demonstrated in rodent studies in which both the development of

bleomycin-induced pulmonary fibrosis and the lesions of silicosis were

prevented by the administration of antibodies to TNF-a."'1 2










The results of this study imply that TNF-a is also important in the

pathogenesis of virus-induced airway fibrosis. However, the mechanism of the

increased gene expression is unclear. The initial increase in the amount of

TNF-a mRNA present in whole lung homogenates was observed at 3 days after

inoculation, while increases in the numbers of cells (mostly macrophages and

fibroblasts) expressing cytoplasmic protein were not noted until 7-10 days post

inoculation. This finding indicates that not all of the increased expression was

the result of the inflammatory cell infilitrate, especially since macrophages do not

inflitrate until 4-5 days after virus-inoculation.37 Epithelial cells and and

lymphocytes in the airway associated lymphoid tissue (BALT) were also positive

for TNF-a mRNA and protein and are the most likely sources of the initial

increase in TNF-a mRNA levels observed in the whole lung homogenates.

Epithelial cells are also the first cells infected by Sendai virus, and the induction

of TNF-a expression may be a direct effect of infection.

Several viruses can directly increase TNF-a mRNA and protein expression.".
111.112, 145. 1 Influenza A virus infection of murine macrophages results in a

massive accumulation of TNF-a mRNA, including a unique high molecular

weight transcript that is not translated into protein until the cells are exposed to a

secondary trigger, such as small amounts of endotoxin.14 Although unique

species of TNF-a mRNA were not observed in this study, Sendai virus has been

found to stimulate TNF-a protein production in monocytes57 and direct

upregulation of TNF-a transcription may contribute to the increased expression









69
observed in infected rats. Direct upregulation of TNF-a gene transcription and

translation by Sendai virus infection could also account for some of the

differences in the pathologic responses of BN and F344 rats. Viral replication

persists for several days longer in the lungs of weanling BN rats than in those of

F344 rats.37 Persisting viral replication could be an important stimulus for

prolonged cytokine expression and result in airway fibrosis in BN rats.

The mechanism by which TNF-a contributes to the development of

pulmonary fibrosis is not completely understood. TNF-a can upregulate the

inflammatory response through its ability to activate macrophages, stimulate

lymphocyte proliferation, and increase the adhesion of both neutrophils and

eosinophils.147' 1' It has also been shown to induce necrosis of alveolar

epithelial and endothelial cells when given intravenously."1'.4 Both of these

effects contribute to the development of fibrosis by increasing the extent of

tissue damage. In addition, TNF-a can directly stimulate fibroblast proliferation

and collagen deposition, and induce the expression of other fibrogenic cytokines

such as PDGF.13

Recently, it has been found that pulmonary platelet trapping strongly

correlates with collagen deposition in bleomycin-induced pulmonary fibrosis.11

The same study also found that antibodies to TNF-a, as well as other treatments









70
mediate the development of fibrosis."1 The importance of platelet trapping in

the pathogenesis of Sendai virus-induced bronchiolar fibrosis is uncertain.

Elevations in pulmonary TGF-P mRNA and protein expression were found in the

infected susceptible BN rats, but the increases were associated with the peak

macrophage infiltration at 7 days after inoculation, and pulmonary PDGF mRNA

levels were not increased by viral infection.

The results of this study demonstrate that virus-induced bronchiolar TNF-a

gene expression precedes bronchiolar mural fibrosis and suggest that it may

play an important pathogenetic role in bronchiolar fibrosis and the associated

airway function abnormalities.
























18s -. ... : TNF-a


SHAM ][ 3 DPI ][ 7 DPI ]



Figure 4-1: Northern blot of rat pulmonary total RNA hybridized with a "P-
labeled cDNA probe for rat TNF-a. This blot contains pulmonary RNA extracted
from sham-inoculated F344 rats at 7 days after inoculation (Sham), and from
virus-inoculated F344 rats at 3 & 7 days after inoculation. Each lane was loaded
with 20 pg of RNA from an individual animal. The upper panel shows the
ethidium bromide-stained gel.















CO
0: 60
S E

Z2
W 20
E ao


0 2 4 6 8 10 12 14 30
Days After Inoculation


Figure 4-2: Graphic presentation of the results of Northern analysis for
TNF-a. Data points and bars represent the mean % maximum relative
optical band density (ROD) and standard error of the mean (SEM) for
4-6 rats. The maximum ROD for each blot was determined from
positive control RNA. Statisical comparisons of ROD's were only made
between bands from sham- and virus-inoculated rats present on the
same autoradiograph.














































riboprobe (x600).
M %








Figure 4-3: TNF-a mRNA in bronchiolar macrophages. These serial sections
were taken from a virus-inoculated BN rat at 7 days after inoculation. The top
section was incubated with 50 ng of a digoxigenin-labeled antisense riboprobe
for rat TNF-a and the bottom with 50 ng of the digoxigenin-labeled sense
riboprobe (x600).












,-AW' %


-Ac


Figure 4-4: TNF-a protein in bronchiolar macrophages. Both micrographs are of
sections taken from a virus-inoculated BN rat at 7 days after inoculation. The
upper section was incubated with 0.001 mg/ml normal rabbit sera and the lower
with 0.001 mg/ml polyclonal rabbit anti-mouse TNF-a (x600).


























s- T 15 20 25
Days After Inoculation


Figure 4-5: Density of bronchiolar mural macrophages expressing
TNF-a protein. Data points represent group mean + SEM (n=4).














P, U.UUUI
0 F344 Control
V F344 Infected
S 4 0 BN Control
g 1 BN Infected


o < / I
0 -V




0 5 15 20 25 30
Days After Inoculation



Figure 4-6: Density of bronchiolar fibroblasts expressing TNF-a
protein. Data points represent group mean + SEM (n=4).
















CHAPTER 5
FIBROBLAST PROLIFERATION IN THE DEVELOPMENT OF SENDAI VIRUS-
INDUCED BRONCHIOLAR FIBROSIS IN BN RATS IS INHIBITED BYA
SOLUBLE TNF RECEPTOR


Summary



Sendai virus infection in virus-susceptible BN rats induces bronchiolar

fibrosis associated with airway hyperresponsiveness and persistent increases in

airway resistance. This study investigated the significance of TNF-a in the

pathogenesis of Sendai virus-induced bronchiolar fibrosis by determining

whether the numbers of proliferating bronchiolar fibroblasts induced by virus

infection could be reduced by inhibition of TNF-a bioactivity. Virus-susceptible

BN rats were treated with a 55 kD soluble TNF receptor-immunoglobulin G

fusion protein (p55sTNFR-IgG) following either sham- or virus-inoculation.

Control sham- and virus-inoculated animals were treated with human IgG.

Proliferating bronchiolar fibroblasts were labeled with bromodeoxyuridine

(BrdU), detected by immunocytochemical staining and quantitated. At 8 to 9

days after virus-inoculation, rats treated with p55sTNFR-lgG had decreased

numbers of bronchiolar fibroblasts incorporating BrdU compared to virus-









78
inoculated IgG-treated rats (p<0.05). There was significantly increased mortality

in p55sTNFR-IgG-treated virus-inoculated rats associated with increased viral

replication and decreased numbers of macrophages and lymphocytes (p<0.05),

in bronchioalveolar lavage fluid. In conclusion, TNF-a is an important mediator

of bronchiolar fibrosis and has a critical role in the termination of Sendai viral

replication in the lung.



Introduction



Viral bronchiolitis in early childhood is associated with the development of

chronic respiratory disease and has been identified as a risk factor for the

development of asthma.26 After inoculation with parainfluenza virus type I

(Sendai), weanling rats develop persistent increases in airway resistance,

decreased dynamic compliance and airway hyperresponsiveness; and are a

useful model to determine which genetic, inflammatory and immunological

factors are important in the pathogenesis of virus-induced airway

abnormalities.'8 33.35

The development of virus-induced pulmonary function abnormalities in

rats is determined by age at infection and genetic susceptibility.18'3S" Neonatal

and weanling Brown Norway (BN) rats are susceptible to persistent virus-

induced alterations in pulmonary structure and function. Fischer 344 (F344) rats

are resistant.18' 3 Bronchiolar inflammation composed of aggregates of










macrophages, mast cells and eosinophils, and bronchiolar mural fibrosis" is

associated with the virus-induced abnormalities in pulmonary function.

Bronchiolar mural fibrosis and the resulting airway thickening is prominent in

BN rats 15 to 30 days after being infected with Sendai virus." Airway modeling

studies have shown that the dynamics of increased airway resistance can be

profoundly affected by mural fibrosis since even moderate amounts of airway

wall thickening can greatly accentuate airway luminal narrowing during smooth

muscle contraction." Bronchiolar fibrosis is a common finding in asthmatics,"22

and has been linked to airway obstruction.

The pathogenesis of airway fibrosis is poorly understood, however, tumor

necrosis factor-a (TNF-a) has been implicated as an important mediator of

chemically-induced pulmonary fibrosis.' 1", 1' 2' In models of both silica- and

bleomycin-induced pulmonary fibrosis, the administration of antibodies to TNF-a

prevented the development of fibrotic lung lesions.11""12 In the silicosis model,

lung lesions were also enhanced by the infusion of recombinant TNF-a.'12

The results of previous studies indicate that the development of Sendai virus-

induced bronchiolar fibrosis in BN rats is preceded by prolonged increases in

pulmonary TNF-a mRNA, and increases in the numbers of bronchiolar

macrophages and fibroblasts expressing TNF-a protein (see Chapter 4). The

objective of this study was to determine whether the inhibition of TNF-a

bioactivity during Sendai virus infection would prevent the development of

bronchiolar fibrosis in BN rats. Since preliminary studies indicated that










increases in the number of proliferating bronchiolar fibroblasts precede the

development of Sendai virus-induced airway fibrosis, bronchiolar fibroblast

proliferation was used as an indicator of fibrosis. The bioactivity of TNF-a was

inhibited using a fusion protein composed of the soluble part of the p55 TNF

receptor linked to the human IgG heavy chain constant region, a construct that

has been successfully used to block TNF activity in animal models of

endotoxemia.'5-152



Materials and Methods


Animals. Infection Protocol



Twenty-two day old weanling male BN rats were purchased from Harlan

Sprague Dawley Inc., Indianapolis, IN. They were negative for serum antibodies

to Sendai virus, pneumonia virus of mice, mycoplasma, Kilham rat virus, and

sialodacryoadenitis virus/rat corona virus, and for respiratory aerobic bacteria.

The sham- and virus-inoculated groups were separated and housed in

individually ventilated isolation units located in separate rooms. Three days

after arrival, the rats in the virus-inoculated groups were exposed to

parainfluenza type 1 (Sendai) virus at a concentration of 1-3 plaque forming

units (PFU) per ml of gas for 15 minutes as previously described."32 Sham-

inoculated rats were exposed to sterile chorioallantoic fluid (CAF).













Experimental Design



One group of eighteen 25-day-old, male BN rats were inoculated with Sendai

virus, while a second group of 12 were sham-inoculated with CAF. Six rats from

the sham and 9 from the virus-inoculated rats were injected intraperitoneally (ip)

with 10 mg/kg of p55 sTNFR-IgG fusion protein (generously supplied by

Hoffman- La Roche, Basel, Switzerland) on 0, 2, 4, 6, & 8 days post inoculation

(dpi). The remaining animals: 6 sham-inoculated and 9 virus-inoculated rats,

were injected ip with10 mg/kg of human IgG according to the same schedule.

Three rats from each of the virus-inoculated groups were sacrificed at 7 dpi.

The rest of the rats were given 200 .g/g 5-bromo-2' -deoxyuridine (BrdU, Sigma,

St Louis, MO, 2 mg/ml in PBS) ip 12 hours before they were processed at 9 dpi

(Figure 2-2).

In a separate infection run, an additional group of 12 virus-inoculated rats

received 1 mg/kg of p55 sTNFR-IgG on the same schedule as above. These

animals were injected with BrdU 12 hours before being sacrificed at 7 dpi (n=6)

and 8 dpi (n=6). All rats were anesthetized with sodium pentobarbital

(approximately 200 pg/g body weight) and were killed by exsanguination via

intra cardiac puncture. Bronchoalveolar lavages (BAL) were performed, and the

lungs were fixed by tracheal perfusion with 4% paraformaldehyde at 30 cm H20

pressure for 2 hours.










Bronchoalveolar Lavage



Phosphate-buffered saline (PBS) was used to perform bronchoalveolar

lavage (BAL) as previously described.53 Total cell counts were made using a

hemacytometer. Cytocentrifuge cell samples of lavage fluid were made with a

Cytospin 2 (Shandon Southern Instruments, Inc, Swickley, PA), fixed in

methanol, and stained with Giemsa solution. A cell differential was calculated

after evaluating 200 cells. Total cell numbers recovered per lung were

calculated (macrophages, neutrophils, and lymphocytes) and normalized per

100 g body weight. The levels of TNF-a protein and bioactivity in the BAL fluid

were determined as described below.



TNF-a Protein and Bioassays



The amount of TNF-a protein present in the BAL fluid samples was

determined using an enzyme-linked immunoabsorbance assay (ELISA) kit

(Biosource International Inc. Camarillo, CA) performed according to the

manufacter's instructions. The level of TNF-a bioactivity in BAL fluid samples

was determined by cytotoxicity assays.'"'" Briefly, 50,000 cells from the

mouse cell line Wehi 164, a BALB/c fibrosarcoma'5 were plated per microtiter

well in 200 pl RPMI media with Glutamine, penicillin (50 IU/ml) streptomycin (50

mcg/ml) and incubated over-night at 37*, 5% CO2. Eighty pI of a 1.2 pg/ml










actinomycin D/RPMI solution, followed by 20 jl of the appropriate standard or

sample was then added to each well and the plates were incubated over-night at

37", 5% CO2. Ten pl of a freshly prepared 6 mg/ml MTT solution in RPMI was

then added to each well and the plates were incubated for 4 hours at 37', 5%

CO,. At the end of this incubation, 100 pl of 2-propanol was added to each

well, the plates were shaken at room temperature for 30 min, and 100 / of

water/well was added. The optical densities of the standard and sample wells

was read at 570/690 nm, and the amount of TNF-a bioactivity present in the

samples was determined from the standard curve.



BrdU Labeling



Four transverse sections were taken from the fixed left lungs of rats treated

with BrdU. Sections were taken at the hilus, between the hilar section and the

cranial lung margin, and at points equidistant caudal to the hilar section and

cranial to the caudal lung margin. Tissue blocks were embedded in paraffin the

same day. Five pm-thick sections were cut, mounted on slides, deparaffinized in

xylene, rehydrated, and washed in distilled water. Immunocytochemical staining

for BrdU was performed using a method modified from that previously

described.'17 Slides were placed in 3.0% H202 for 10 minutes to quench

endogenous peroxidase, washed twice in PBS, and pretreated in both 2N HCI

(30 min, 37* C) and 0.1% w/v trypsin in PBS (20 min, 37" C). Sections were









84
rinsed in PBS, covered with antibody diluent (1.0% BSA and 0.5% Tween 20 in

PBS) for 30 min to block nonspecific binding, blotted and incubated for 1 hour in

mouse anti-BrdU monoclonal antibody (Becton-Dickinson, Mountain View, CA)

diluted 1:100 in antibody diluent. Sections were washed and covered with

peroxidase-conjugated goat anti-mouse IgG (Fc specific, Sigma, St Louis, MO)

diluted 1:100 in antibody diluent for 1 hour. Bound peroxidase-conjugated

antibody was detected by development in the chromogen diaminobenzidine

(Sigma, St. Louis, MO) in 0.25 mg/ml in 0.025 M Tris, pH 7.6 and 0.01% H202 for

20 min. The slides were counter-stained with hematoxylin and eosin.



Virology



Following bronchoalveolar lavage, the right lungs of virus-inoculated rats

treated with 1 mg/kg of p55 sTNFR-IgG were collected aseptically, frozen in

liquid nitrogen, and stored at -80 C until viral assays were run. Plaque assays

for infectious virus were performed on lung homogenates using Madin-Darby

bovine kidney cells as previously described.32 The presence of viral antigens in

paraffin-embedded lung sections from infected p55sTNFR-lgG-treated rats was

detected by immunocytochemistry. Tissue sections were pretreated as

described above for the detection of BrdU, and incubated with a mouse

monoclonal antibody to parainfluenza type 1 virus (Chemicon, Temecula, CA).

Application of peroxidase-conjugated Cappel goat anti-mouse antibody and the









85
detection of bound peroxidase-conjugated antibody was performed as described

above.





Data Analysis



Fibroblast proliferation was assessed by counting the number of BrdU-

labeled bronchiolar fibroblasts in 100 high power (400X) fields per lung. Cells

counted as fibroblasts had elongated spindloid profiles, oval nuclei, and

eosinophilic cytoplasm. Only cells with nuclei within the plane of section were

counted. Mean values between groups were compared by one-way analysis of

variance (one-way ANOVA) using a microcomputer-based statistical program

(SigmaStat, Jandel Corporation, San Rafael, CA). Differences between

individual means were assessed by Bonferroni's method of all pairwise multiple

comparison.14 BAL fluid data were subjected to logarithmic transformation prior

to analysis.



Results



p55sTNFR-lgG Inhibits Virus-induced Proliferation of Bronchiolar Fibroblasts



Nine days after virus-inoculation, the mean number of BrdU-labeled

fibroblasts in the bronchiolar walls of virus-inoculated, p55sTNFR-lgG-treated










rats was significantly lower than that of virus-inoculated IgG-treated controls

(p55sTNFR-lgG: 12.3 2.3, IgG: 48.7 5.2, p<0.05) (Figure 5-1). Virus-

inoculated IgG-treated rats also had higher numbers of BrdU-labeled bronchiolar

fibroblasts than the sham-inoculated rats (p<0.05). The number of labeled

fibroblasts in the virus-inoculated p55sTNF-lgG treated rats was not significantly

different from that of the sham-inoculated rats. The numbers of BrdU-labeled

bronchiolar fibroblasts in the p55sTNF-lgG and IgG treated sham-inoculated rats

were not significantly different.



Bronchoalveolar Lavaae Fluid: TNF-a Protein and Bioassays



The mean pg/ml of TNF-a protein in the BAL fluid of virus-inoculated

p55sTNFR-lgG-treated rats was over four-fold higher than that of the virus-

inoculated IgG-treated rats at 7 days after inoculation (p< 0.003, Figure 5-2).

However, the level of TNF-a bioactivity in the BAL fluid of virus-inoculated rats

treated with 10 mg/kg of p55sTNF-IgG was significantly lower at 7 dpi than that

present in the virus-inoculated IgG-treated controls (p55sTNF-IgG: 2.5 1.41

pg/ml, IgG: 95.5 54.58 pg/ml, mean sem, p<0.05) (Figure 5-3). TNF-a

bioactivity was also decreased in the virus-inoculated rats treated with 10 mg/kg

p55sTNF-lgG at 9 dpi, but the means between the two virus-inoculated groups

were not significantly different. The virus-inoculated rats given 1 mg/kg

p55sTNF-IgG also had decreased, but not significantly lower, TNF-a bioactivity









87
in their BAL fluid at 7 dpi compared to the IgG-treated virus-inoculated rats (data

not shown). At 9 dpi, the surviving virus-inoculated rats in the 1 mg/kg

p55sTNF-lgG treated group had slightly but not significantly increased levels of

TNF-a bioactivity in their BAL fluid (data not shown). TNF-a bioactivity was not

detected in the lavage fluid of any of the rats in the sham-inoculated groups

(Figure 5-3).



Bronchoalveolar Lavage Fluid: Cell Counts



The data from total and differential cell counts performed on BAL fluid

samples are presented in Figures 5-4, 5-5, 5-6, 5-7. The counts from the virus-

inoculated low and high dose p55sTNFR-lgG-treated groups were pooled.

Virus-inoculated rats treated with p55sTNF-lgG had 3.5 fold fewer total cells/g

present in their BAL fluid at 7 days after inoculation compared to virus-

inoculated rats receiving IgG (p55sTNF-lgG: 12714 2004 IgG: 44886 5415,

mean sem, p<0.05). Sixfold decreases in total cell numbers were present at 9

days after inoculation (p55sTNF-lgG: 8217 1542, IgG: 52782 13632, p<0.05)

(Figure 5-4). Differential counts at seven days after inoculation, revealed virus-

inoculated p55sTNF-lgG treated rats had fivefold fewer numbers of

macrophages/g (p55sTNF-IgG: 4036 720, IgG: 20981 2433, p<0.05) and

almost sixfold fewer lymphocytes/g (p55sTNF-lgG: 586 + 145, IgG: 3948 1371,

p<0.05) in their BAL fluid compared to virus-inoculated IgG-treated rats (Figures









88
5-5 & 5-6). Ninefold fewer numbers of macrophages (p55sTNF-IgG: 2674 384,

IgG: 26153 6042, p<0.05) and fourteen-fold fewer lymphocytes (p55sTNF-IgG:

233 66, IgG: 3381 1068, p<0.05) were observed at 9 days after virus-

inoculation (Figures 5-5 & 5-6). Neutrophil counts were also two- and sixfold

lower in the p55sTNF-lgG treated virus-inoculated rats at 7 and 9 dpi

respectively, however the difference was not statistically significant (Figure 5-7).

The virus-inoculated, IgG-treated rats had more total cells, with increased

numbers of macrophages, lymphocytes, and neutrophils in their BAL fluid than

did the rats in either of the sham-inoculated groups (p<0.05). In contrast, virus-

inoculated, p55sTNF-lgG treated rats did not have increased numbers of

macrophages or lymphocytes in their BAL fluid compared to the sham-inoculated

rats. However, they did have increased numbers of neutrophils compared to

both groups of sham-inoculated rats, and higher total cell numbers than the

p55sTNF-lgG treated sham-inoculated rats (p<0.05). There were no significant

differences in the BAL fluid total or differential cell counts between the p55sTNF-

IgG and the IgG-treated sham inoculated rats.



Increased Mortality and Pulmonary Lesions in the Sendai Virus-Inoculated

p55sTNF-lIG Treated Rats



All of the virus-inoculated rats treated with 10 mg/kg p55sTNF-lgG died or

were found moribund and were sacrificed 8-9 days after virus-inoculation. Four










of the six virus-inoculated rats given 1 mg/kg p55sTNF-IgG also died 8-9 days

after virus-inoculation. Three rats from these two groups (one received 10

mg/kg, two were treated with 1 mg/kg) were labeled with BrdU. Unexpected

mortality did not occur in the other experimental groups. Compared to virus-

inoculated control animals, many of the airways in the virus-inoculated

p55sTNFR-lgG-treated rats were characterized by hyperplastic bronchiolar and

alveolar epithelial cells, which were heavily labeled in the BrdU-stained sections.

Virus-inoculated p55sTNFR-lgG rats also had subjectively fewer inflammatory

cells with proportionally fewer lymphocytes than the virus-inoculated controls.

Viral antigens were detected in the hyperplastic epithelial cells using

immunocytochemistry.



Pulmonary Viral Titers



The virus-inoculated rats treated with 1 mg/kg of p55sTNFR-lgG had mean

pulmonary viral titers that were 5-fold higher at both 7 and 9 dpi than those of

comparable groups of untreated virus-inoculated rats (Figure 5-8), however the

increase did not reach statistical significance.



Discussion



TNF-a has been implicated as an important mediator in the pathogenesis of

the pulmonary fibrosis induced by pneumotoxins such as silica and











bleomycin.108'1"'120 This study investigated the significance of TNF-a in a rat

model of virus-induced bronchiolar fibrosis. We wanted to determine whether

inhibition of TNF-a bioactivity in virus-inoculated BN rats abrogated virus-

induced increases in bronchiolar fibroblast proliferation. Fibroblast proliferation

precedes virus-induced bronchiolar fibrosis, and is a well-characterized effect of

TNF-a.'1

Although the results of this study are complicated by the profound effects the

inhibition of TNF-a bioactivity had on the immune/inflammatory response to the

virus, treatment with p55sTNFR-IgG blocked TNF-a activity and partially blocked

the virus-induced BrdU incorporation. Significantly fewer bronchiolar fibroblasts

were labeled with BrdU in the virus-inoculated p55sTNFR-lgG-treated rats

compared to the number labeled in the IgG-treated virus-inoculated animals.

This is consistent with the interpretation that TNF-a is stimulating fibroblast

proliferation.

The results of the TNF-a bioassays on BAL fluid from the virus-inoculated,

p55sTNFR-lgG-treated animals confirm that p55sTNFR-IgG did effectively

inhibit TNF-a's biological activity, even though TNF-a protein levels as detected

by ELISA were elevated. The increased protein levels in the p55TNFR-lgG-

treated rats was probably due to decreased rate of breakdown of the bound

inactivated protein. The low level of TNF-a bioactivity observed at 9 days after

virus-inoculation in the rats given the low dose of p55sTNFR-lgG (1 mg/kg)

indicates that TNF-a was not completely inhibited in these animals at this dose.










The mechanism by which TNF-a may be inducing fibroblast proliferation in

this model is unclear. TNF-a is a mitogen for certain fibroblasts,1' and could

therefore be directly stimulating bronchiolar fibroblast proliferation. TNF-a could

also be affecting fibroblast proliferation indirectly through several different

mechanisms. It is a potent activator and chemoattractant for macrophages,"'111

and induces expression of adhesion molecules for platelets,119 both of which are

important sources of other fibroblast mitogens such as transforming growth

factor-p (TGF-p), and platelet-derived growth factor (PDGF). The results of

previous studies have identified an influx of macrophages expressing TGF-p,

mRNA and protein into bronchioles as a prominent feature of rats developing

virus-induced bronchiolar fibrosis (see Chapter 3), and the depression in

macrophage numbers observed in the BAL fluid of the infected p55sTNFR-IgG-

treated rats confirms that TNF-a is critical to the virus-induced macrophage

recruitment. Other indirect effects of TNF-a on fibroblast proliferation and the

development of bronchiolar fibrosis include its ability to increase the expression

of adhesion molecules for both neutrophils and eosinophils,'11"2 and to depress

fibrinolysis.7 These effects enhance the inflammatory response, increase the

extent of tissue damage, and therefore contribute to fibroblast proliferation and

fibrosis.

TNF-a is a potent antiviral cytokine.1~'" The results of in vitro studies using

several different cell and virus systems indicate TNF-a has both the ability to

enhance the killing of virus-infected cells and protect uninfected cells from










infection.l", le1,6 In addition, several viruses produce virulence factors that

specifically inhibit TNF-a.16385 For example, adenoviruses inhibit the action of

TNF by blocking signal transduction from its receptor'" while several pox

viruses contain genes that encode a soluble TNF receptor similar to the one

used in this study.'" TNF-a is also able to activate macrophages independent

of T and B cells in murine cytomegalovirus and herpes simplex virus

infections.'"

The importance of TNF-a in the inflammatory response to Sendai virus

infection is indicated by the high mortality in the virus-inoculated, p55sTNFR-

IgG-treated rats. The increased mortality was associated with significant

reductions in the numbers of macrophages and lymphocytes in the BAL fluid, the

tendency towards higher pulmonary viral titers and the presence of numerous

hyperplastic epithelial cells that were positive for viral antigens. These findings

suggest that inhibition of TNF-a activity decreased the inflammatory cell

response to the virus, prevented the destruction of infected epithelial cells and

allowed prolonged viral replication. Paradoxically, however, BN rats have

prolonged elevations of pulmonary TNF-a expression but also higher pulmonary

viral titers compared to Fischer 344 rats, which are resistant to virus-induced

airway abnormalities.37 These observations suggest that TNF-a may be

protective early in the course of Sendai virus infection, while prolonged

expression may be harmful.

There is evidence that the role of TNF-a in T-cell-mediated inflammation is








93
dependent upon the Thl/Th2 cytokine profile. In studies using Mycobacterium

tuberculosis in dosing protocols designed to elicit a predominantly Thi

response, TNF-a was found to act as an additional macrophage-activating

factor.167 However, when the dosing regime was altered to elicit a predominantly

Th2 response, TNF-a contributed to tissue damage.1' Studies using T cells

stimulated by individual proteins of respiratory syncytial virus to secrete either a

Thi or Th2 cytokine profile indicate that different forms of viral bronchiolitis can

be caused by functionally distinct types of antiviral T cells.'" These findings

imply that the underlying cytokine profile may determine the role of TNF-a in

viral pathogenesis and may explain the higher viral titers observed in BN rats in

the face of high pulmonary expression of TNF- a.

In conclusion, the results from this study indicate that TNF-a is a potentially

important mediator of virus-induced bronchiolar fibrosis, and is critical to the

immune/inflammatory response to parainfluenza type-1 infection.
















Sso5- I r *p < 0.05

SS4 Mean + SE
E0


20
S30-


0

10-



Sham Virus: 9 dpi


Figure 5-1: Density of BrdU-labeled bronchiolar mural fibroblasts in
sham- and virus-inoculated BN rats (n=5 or 6 for sham and infected
IgG-treated groups; n=3 for the infected p55sTNFR-lgG-treated group).