ism.»

7.“? {amazww

11

73:53-93} '

m. ..

rm-“ V
.. ' -

a

.u.\.

< 4.-.-
~MD‘
«a w._ ‘v .—

8

a.

s a ' :‘z'
9“ {2‘33 35.4%:
E“;

I
‘ f;- aﬁ l , ”v3,
#5:: ‘

“a. ,
"ﬂ...
"wt.

p.

5%:
52-33% ~
‘

1.".

1‘

“ w

“a".
n...

5

9 ::€
' ~; i“ v s .
H - . .‘l~,‘!;$_ l
' u
9 32 . , 3 515;: . $53.. .
t q l {’4‘ '9, 1
. V 5 A ‘ ‘ 54
I

u I
f‘
.37.}

.4;
iirrw': 3:34
:31. "mu.“

:2?

 

520/1

This is to certify that the
dissertation entitled

Innate Immune Mechanisms Involved in Airway Inﬂammation
in Equine Recurrent Airway Obstruction (RAO)

presented by

Annerose Berndt

has been accepted towards fulfillment
of the requirements for the

Ph.D. degree in Comparative Medicine and
Integrative Biology

 

 

/___._._

 

 

/// or rofessor’s igna

"7/ 31/5/07

Date

MSU is an afﬁrmative-action, equal-opportunity employer

- -.—.—._-—.-.-.—--u-o---.—-—--c-o—-—-—

 

 

LIBRARY

lichigan State

University

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DAIEDUE

DAIEDUE

DATEDUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/07 p'lClRC/DateDueindd-pJ

 

INNATE IMMUNE MECHANISMS INVOLVED IN AIRWAY INFLAMMATION IN
EQUINE RECURRENT AIRWAY OBSTRUCTION

By

Annerose Berndt

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Comparative Medicine and Integrative Biology

2007

INNA

nc
Iri
as
lil
pr

pr

hi
ar.
CC

br

0 I.

hi;

ABSTRACT

INNATE IMMUNE MECHANISMS INVOLVED IN AIRWAY INFLAMMATION IN
EQUINE RECURRENT AIRWAY OBSTRUCTION

By

Annerose Berndt

Introduction: Recurrent airway obstruction (RAO) is characterized by
neutrophilic airway inﬂammation and obstruction, and stabling of susceptible horses
triggers acute disease exacerbations. Stable dust is rich in endotoxin and fungi, such
as Faenia rectivirgula, Thermoactinomyces vulgaris and Aspergillus ﬁxmigatus. Toll-
like receptors (TLRS) are pathogen recognition receptors that are transmembrane
proteins on a variety of airway cells (e.g. epithelial cells). Toll-like receptor 4 is the
principal recognition receptor for lipopolysaccharide (LPS), and TLR2 is the
principal recognition receptor for fungal products. In human bronchial epithelium,
TLR4 and TLR2 stimulation leads to increased production of interleukin (IL)-8, a
potent neutrophil attractant. The zinc ﬁnger protein A20 negatively regulates these
pathways. Hypotheses: We hypothesized i) that horses in stables are exposed to
higher endotoxin concentrations than on pasture; ii) that during stabling TLR4, TLR2,
and IL-8 mRNA expression are increased in RAG-affected horses compared to
controls, and that this would be paralleled by elevated neutrophil counts in
bronchoalveolar lavage ﬂuid (BAL); and iii) that A20 mRNA expression is not
increased in RAO-affected horses during stabling compared to controls. Materials
and Methods: We determined endotoxin concentrations in the breathing zone of six

horses. Furthermore, we measured the maximal change in pleural pressure (APplmax),

dt

6C

Si

0i

Vt

“I

CO

determined inﬂammatory cell counts in BAL, and quantiﬁed TLR4, TLR2, IL-8, and
A20 mRNA in bronchial epithelium by quantitative real-time polymerase chain
reaction (qRT-PCR). For these experiments, we studied six age-matched horse pairs,
each pair consisting of one RAG-affected horse and one control horse. Each pair was
studied when the RAO-affected horse had airway obstruction induced by stabling,
and after 7, 14 and 28 days on pasture. Results: Endotoxin concentrations in the
breathing zone of stabled horses were signiﬁcantly higher than in the breathing zone
of these horses on pasture. While stabling increased BAL neutophils and APplmax as
well as TLR4 and IL-8 mRNA expression signiﬁcantly, TLR2 and A20 were
unaffected. TLR4, but not TLR2, correlated with IL-8 and IL-8 was signiﬁcantly
correlated with BAL neutrophils and A20. When data were pooled, TLR2 and A20
were signiﬁcantly correlated and A20 was negatively associated with inﬂammatory
cells. Conclusions: In our stables, horses are exposed to an at least 8-fold higher
concentration of endotoxin than on pastures. Elevated TLR4 expression and lack of
A20 upregulation in bronchial epithelial cells from RAO-affected horses may
contribute to elevated IL-8 production, leading to exaggerated neutrophilic airway
inﬂammation in response to inhalation of stable dust. Stable dust exposure does not
lead to an increase in TLR2 mRNA expression. The localization or the amount of the
TLR2 receptor complex rather than the amount of TLR2 mRN A may be important in
fungal-induced airway inﬂammation. Equine epithelial-derived A20 may protect
against airway inﬂammation by decreasing the number of bronchoalveolar lavage

inﬂammatory cells, and is involved in modulation of airway inﬂammation.

COP}
Anne
2007

Copyright by
Annerose Berndt
2007

This

scan

Bert

an ir

my \

for t

scier

least

balm

This work is dedicated to the people, who, each with their own example, inspired me to
search enthusiastically for new knowledge. Foremost to my parents Jutta and Dietmar
Berndt for their true love and guidance that strengthened my steps towards developing
an independent mind and strong personality, and whose values will always last in me and
my work. In addition, to my sister, Antje Berndt, and my brother-in-law, Eric A. Stone,
for their never-ending support, advice, and patience when facing my need for endless
scientiﬁc conversations. They are the best shoulders one can fall back onto. Last, but not
least, to Tawnya L. Weaver for her unconditional friendship and her living example of

balancing an intensive work load.

l\Vol
who.
indci
u'hit‘
medi
u'isll

Rob

OUISl
ener
lnlcg
Nor
my

man
her

bth
Pall
Curr
kno.

{MEI

ACKNOWLEDGMENTS

I would like to express my deepest thanks to my major advisor Dr. Frederik J. Derlmen
who, in every instance, supported my wish and carefully guided my path to become an
independent researcher. I am not only thankful for all the scientiﬁc conversations, during
which I learned how to integrate detailed scientiﬁc work into a complex biological and
medical system, but also for his example of a truth-searching and humane scientist. I also
wish to express my thanks to my guidance committee members: Dr. N. Edward
Robinson for the many inspirational conversations and his never-ending support and
trust in my research, Dr. Susan Ewart for her advice and support regarding research and
gradate school, and for the many times she inspired me with her example as an
outstanding researcher, Dr. Vilma Yuzbasiyan-Gurkan for her Vision and everlasting
energy to establish the amazing graduate program of Comparative Medicine and
Integrative Biology (CMIB) in the veterinary college at Michigan State University, and
Norbert Kaminski for establishing opportunities and his support and advice regarding
my research. Furthermore, I am grateful for the chance to have met and worked with
many other amazing people. In particular, I would like to thank Dr. Annette Thelen for
her ability to put the sometimes stressful and draining bench work when in molecular
biology into a colorful new and exciting world. One of my deepest Thanks goes to Dr.
Patrick J. Venta for his patience in answering many of my burning questions about our
current understanding in genetic and genomic research with a never-ending source of
knowledge. In addition, I would like to express my thanks to Lisa R. Bartner for her

friendship and her talented drawings that became part of this dissertation (Figure 2-1) and

vi

papers
hchai
exprc:
Puhnt

Berni

Ihel)
Genn
acknu
Kknﬂt
Auan
Fund.
Vekxi

Al Wan

papers. Furthermore, I would like to acknowledge Victoria Hoelzer-Maddox for her
help in preparing manuscripts, presentations, and this dissertation. Last, but not least, I
express my gratitude to all the people I have had the chance to work with in the
Pulmonary Laboratory, especially Heather deFeijter-Rupp, Susan Eberhart and Cathy
Berney, without whom I would not have been able to carry through these intense
research years. My stay at MSU would not have been possible without the experiences at
the Department of Internal Medicine at the Veterinary Faculty at University of Leipzig,
Germany, under the guidance of Dr. Gerald F. Schusser. Finally, I would like to
acknowledge the funding sources that have made my research possible: foremost the
Matilda Wilson Endowment, the 2003-04 German Academic Exchange Service (DAAD)
Award, the 2005-06 grant from College of Veterinary Medicine’s Genetic Research
Fund, the 2005 Fellowship Award for PhD Training of Veterinarians from the College of
Veterinary Medicine and the 2007 Graduate School Dissertation Completion Fellowship

Award from Michigan State University.

vii

llSl
LlSl
LlSl
INTI

CH.j
LITI

CH;
ENE
IS H

TABLE OF CONTENTS

LIST OF TABLES ................................................................................... xi
LIST OF FIGURES ................................................................................ xiv
LIST OF ABBREVIATIONS .................................................................. xvii
INTRODUCTION ................................................................................... 1
CHAPTER 1
LITERATURE REVIEW ............................................................................ 6
o Recurrent Airway Obstruction - Deﬁnition and Pathogenesis ............... 6
o Endotoxin-activated Immune Mechanisms in the Lung ....................... 8
o Pathophysiology ........................................................... 9
0 Source of Endotoxin in the Environment .............................. 10
0 Measurement of Endotoxin ............................................. 10
0 Recruitment of Inﬂammatory Cells into the Lung ........................... 11
o Neutrophil Emigration into the Lung ................................. 11
o Soluble Mediators of Migration ........................................ 12
I Cytokines ........................................................ 12
I Chemoattractants ................................................ l3
0 Pathways of Innate Immune Mechanisms within the Lung ................. 16
0 Introduction ............................................................... 16
o Innate Immunity of the Airways ....................................... 17
0 Recognition of Respiratory Pathogens by the Airway
Epithelium ...................................................................................... 18
o Toll-like receptors ....................................................... 19
0 Negative Regulation of Toll-like receptor signaling ................ 26
0 Summary of Literature Review and Hypotheses ............................. 27
CHAPTER 2
ENDOTOXIN CONCENTRATIONS WITHIN THE BREATHING ZONE OF HORSES
IS HIGHER IN STABLES THAN ON PASTURE ............................................. 30
Abstract ..................................................................................... 30
Introduction ................................................................................. 3 1
Material and Methods ..................................................................... 33
Results ....................................................................................... 35
Discussion ................................................................................... 36
CHAPTER 3
PRIMER DESIGN AND QUANTITATIVE REAL TIME POLYMERASE CHAIN
REACTION (qRT-PCR) ........................................................................... 43

The Quantitative Real Time Polymerase Chain Reaction Assay 43

viii

CHA
ELE‘
EPIT
HOR

CI-I.-\
EXPQ
CELI
DL'S

CHA
A30 :
COR
BRO

CI-IA
REG]
3ti
CELL

Introduction ........................................................................ 43
Primer Design ..................................................................... 46
Veriﬁcation of ampliﬁcation product for primer pairs ...................... 47

CHAPTER 4
ELEVATED AMOUNT OF TOLL-LIKE RECEPTOR 4 mRNA IN BRONCHIAL
EPITHELIAL CELLS IS ASSOCIATED WITH AIRWAY INFLAMMATION IN

HORSES WITH RECURRENT AIRWAY OBSTRUCTION ............................... 53
Abstract ..................................................................................... 53
Introduction ................................................................................. 54
Material and Methods ..................................................................... 56
Results ....................................................................................... 61
Discussion ................................................................................... 63

CHAPTER 5

EXPRESSION OF TOLL-LIKE RECEPTOR 2 mRNA IN BRONCHIAL EPITHELIAL
CELLS IS NOT INDUCED BY EXPOSURE OF HORSES TO STABLE

DUST .................................................................................................. 73
Abstract ...................................................................................... 73
Introduction ................................................................................. 74
Material and Methods ..................................................................... 76
Results ....................................................................................... 8 1
Discussion ................................................................................... 82

CHAPTER 6

A20 mRNA EXPRESSION IN BRONCHIAL EPITHELIAL CELLS IS NEGATIVELY
CORRELATED WITH INFLAMMATORY CELL COUNTS IN EQUINE

BRONCHIAL ALVEOLAR LAVAGE FLUID ............................................... 94
Abstract ...................................................................................... 94
Introduction ................................................................................. 95
Material and Methods ..................................................................... 97
Results ..................................................................................... 101
Discussion ................................................................................. 102

CHAPTER 7

REGRESSION AND PATHWAY ANALYSIS AND MODEL ADJUSTMENT OF TLR
SIGNALING PATHWAYS IN EQUINE BRONCHIAL EPITHELIAL

CELL ................................................................................................. 1 09
Summary .................................................................................... 109
Introduction ............................................................................... l 10
Statistical analysis ......................................................................... 1 12
Results ..................................................................................... 1 14
Discussion ................................................................................. l 16

ix

C H .-'
INT

BIB

CHAPTER 8

INTERPRETATION AND OUTLOOK ....................................................... 128
Areas for future research ................................................................ 129
BIBLIOGRAPHY ................................................................................. 139

Ta
Hu

Ta
En.
pas
1116

Tal
Cli;
end
Em
H01
Ara

LIST OF TABLES

Table 1-1:
Human and murine Toll-like receptors and their ligands. ..................................... 29

Table 2-1:

Endotoxin concentrations in the breathing zone of stabled horses and horses kept on
pasture. (*) indicates signiﬁcant difference (P s 0.05) from endotoxin concentration
measured on pasture. ............................................................................... 38

Table 2-2:

Climatological data (ambient temperature and percent relative humidity) on days of
endotoxin measurements obtained from the National Climatic Data Center. E =
Environment, in which horses were maintained, S = Stable, P = Pasture, QH = Quarter
Horse, Grd = Grade Horse, TB = Thoroughbred, SB = Standard bred, Arab =
Arabian ................................................................................................. 39

Table 3-1:

The oligonucleotide primer pair sequences used for ampliﬁcation of 18S rRNA, TLR4,
TLR2, IL-8 and A20 transcripts in bronchial brushing samples. Amplicon length of the
primer pairs and Genbank accession numbers of sequences used during primer design or
sequence source are shown. ........................... . ............................................ 49

Table 4-1:

Cellular composition of bronchoalveolar lavage ﬂuid (BAL). Only horses that contained
more than 80% epithelial cells within their BB samples were included. Mean i S.E.M. of
total (loglO transformed) and differential cell counts in BAL from RAO-affected and
control horses during stabling (day 0) and after stabling (days 7, 14 and 28). (a) indicates
signiﬁcant difference (P s 0.05) between the horse groups (control, RAO) within a time
point (day 0, 7, 14, and 28). (b) indicates signiﬁcant difference (P S 0.05) from day 0.
IogTCC = loglO transformed total BAL cell count; logNeu = log10 transformed total
BAL neutrophil count; %Neu = percentage of neutrophils in BAL; logMac = loglO
transformed total BAL macrophage count; %Mac = percentage of macrophages in BAL;
logLym = loglO transformed total BAL lymphocyte count; %Lym = percentage of
lymphocytes in BAL ........................................................................... 66-67

Table 4-2:

Fold changes (PC) for the amount of TLR4 and IL-8 mRNA in bronchial brushing
samples (BBS) between RAO-affected and control horses (RAO/Control) during stabling
(day 0) and after stabling (days 7, 14 and 28) as well as between time points within the
RAD-affected (RAO/RAO) and within the control horse group (Control/Control). Only
horses that contained more than 80% epithelial cells within their BBS were included. (*)
indicates that the PC is signiﬁcant different (P S 0.05) to 1 (no change). ................... 68

xi

Table 5-1:

Clinical parameters, cellular composition of bronchoalveolar lavage ﬂuid (BAL), and
TLR2 and IL-8 mRN A expression in bronchial brushing samples (BBS). 5Mean 3: S.E.M.
of APplmax (difference between inspiratory and expiratory pleural pressure during tidal
breathing) values, #geometric means of BAL cell counts and amounts of TLR2 and IL-8
mRNA in BBS from RAD-affected and control horses during stabling (day 0) and after
stabling (days 7, 14, and 28) as well as P-values for ﬁxed effects (horse group (RAO,
control), time, horse group*time) of APplmax, BAL inﬂammatory cells (neutrophils,
macrophages and lymphocytes) as well as TLR2 and IL-8 mRNA expression in BBS. (a)
indicates signiﬁcant difference (P S 0.05) between the horse groups (control, RAO)
within a time point (day 0, 7, 14, and 28). (b) indicates signiﬁcant difference (P S 0.05)
from day 0.

TCC = Total BAL cell count (/ul); Neu = Total BAL neutrophil count (/ul); Mac = Total
BAL macrophage count (/ul); Lym = Total BAL lymphocyte count (/ul); TLR2 = Toll-
like receptor 2; IL-8 = interleukin 8 .......................................................... 86-87

Table 5-2:

Correlation matrix of Spearman correlations among total numbers of bronchoalveolar
lavage (BAL) neutrophils, and bronchial brushing (BB) TLR2 mRNA and BB IL-8
mRNA in all horses, RAD-affected or control horses.

Neu = Total BAL neutrophil count (/ul); TLR2 = Toll-like receptor 2; IL-8 = interleukin
8; SCC = Spearman correlation coefﬁcient; P = P-value; N = number of samples; n/a =
not applicable ........................................................................................ 88

Table 6-1:

Correlation matrix of Spearman correlations among total numbers of bronchoalveolar
lavage (BAL) inﬂammatory cells (neutrophils, macrophages and lymphocytes) and
epithelial-derived A20 mRNA after pooling data from RAO-affected and control horses.
Neu = Total BAL neutrophil count (/ul); Mac = Total BAL macrophage count (/ul); Lym
= Total BAL lymphocyte count (/ul); A20 = zinc ﬁnger protein A20; SCC = Spearman
correlation coefﬁcient; P = P-value; N = number of samples; n/a = not applicable
........................................................................................................ 105

Table 7-1:

Correlation matrix of Spearman correlations between total numbers of bronchoalveolar
lavage (BAL) inﬂammatory cells (neutrophils, macrophages and lymphocytes) and
epithelial-derived TLR4, TLR2, A20 and IL-8 mRNA after pooling data from RAO-
affected and control horses.

Neu = Total BAL neutrophil count (/ul); Mac = Total BAL macrophage count (/ul); Lym
= Total BAL lymphocyte count (/ul); TLR4 = Toll-like receptor 4; TLR2 = Toll-like
receptor 2; A20 = zinc ﬁnger protein A20; IL-8 = interleukin 8; SCC = Spearman
correlation coefﬁcient; P = P-value; N = number of samples; n/a = not applicable
......................................................................................................... 120

xii

Table 7-2:

Full models for regression analysis. For each outcome, a model combining possible
predictors is shown. Predictors were chosen according to the ﬂow diagram shown in
Figure 7-1. Each variable was considered as the outcome of the combination of all possible
predictors (e.g. in Figure 7-1: all variables that are at the same or a lower level than the
deﬁned outcome are considered to be predictors). The estimate, standard error, and t-value
are reported for each predictor in the model. The overall RZ-value is reported for each full
model.

Neu = Total BAL neutrophil count (/ul); Mac = Total BAL macrophage count (/ul); Lym
= Total BAL lymphocyte count (/ul); TLR4 = Toll-like receptor 4; TLR2 = Toll-like
receptor 2; A20 = zinc ﬁnger protein A20; IL-8 = interleukin 8 ...................... 121-122

Table 7-3: Stepwise regression analysis. For each full model reported in Table 7-1 a
stepwise analysis procedure was applied. The estimate, F-, P-, and Rz-values are reported
for each predictor in individual steps.

Neu = Total BAL neutrophil count (/ul); Mac = Total BAL macrophage count (/ul); Lym
= Total BAL lymphocyte count (/ul); TLR4 = Toll-like receptor 4; TLR2 = Toll-like
receptor 2; A20 = zinc ﬁnger protein A20; IL-8 = interleukin 8 ...................... 123-124

Table 8-1:

Examples of immunological studies in the ﬁeld of RAO by different research groups.
TNF-a = Tumor necrosis factor alpha; IL-lB = Interleukin 1 beta; IL-6 = Interleukin 6;
IL-8 = Interleukin 8; IL-lO = Interleukin 10; NF-KB = Nuclear factor kappa B; 1- KB =
Inhibitor of NF-KB; BBS = bronchial brushing samples; BAL = bronchoalveolar lavage;
ICAM-l = Intracellular adhesion molecule 1; GM-CSF = Granulocyte macrophage-
colony-stimulating factor; STAT-5 = Signal transducer and activator of transcription 5;
MIP-2 = Macrophage inﬂammatory protein-2; IL-4 = Interleukin 4; IFN-y = Interferon
gamma; IL-17 = Interleukin 17; IL-5R = Interleukin 5 receptor; IL-9R = Interleukin 9
receptor; TLR4 = Toll-like receptor 4; TLR2 = Toll-like receptor 2; A20 = zinc ﬁnger
protein A20; CXCLl = CXC chemokine ligand 1; G-CSF = Granulocyte-colony
stimulating factor; LPS = lipopolysaccharide; AF = Aspergillus fumigatus; HDS = Hay
dust suspension; CB = clenbuterol; PBS = Phosphate buffered saline; n.m. = not
measured; APplmax = Difference between inspiratory and expiratory pleural pressure;
EMSA = Electromobility shift assay; ISH = In-situ hybridization
.................................................................................................. 137-138

xiii

Figl
Patl
paﬂi

Fig!
Pcrs
horS
halt:
:Xna
the;
rate

hour

Iigu
Endc
horsc
burs

cndot

Figur
ASSOI
horse

1811

Ilgur
Agan
rRNA
Uane
mhph

Flgut
Akmu

Figur
geDOn

Figllr
Chnk
Onhl
Indud
blOnC}
hOl’Sc-b
mdkat
“ﬂhm

LIST OF FIGURES

Figure 1-1:
Pathway model of Toll-like receptors that are potentially involved in the RAO
pathogenesis ........................................................................................... 5

Figure 2-1:

Personal sampler device for endotoxin measurements in the breathing zone around the
horse nostril. The personal sampler consisted of a ﬁlter cassette that was attached to the
halter, and an AirCheck 52 Personal Sample Pump (kindly provided by EMSL
Analytical, Inc., Westmont, NJ) that was attached to a surcingle. The ﬁlter cassette and
the air pump were connected through a rubber hose. Endotoxin was sampled at a ﬂow
rate of 1.5 liters per minute for an average of 245 minutes (~ four
hours) ................................................................................................. 40

Figure 2-2:

Endotoxin concentrations (103 x EU/m3) in the breathing zone of stabled horses and
horses on pasture. Individual diamonds represent individual measurements. Horizontal
bars represent the geometric means. (*) indicates signiﬁcant difference (P _<_ 0.05) from
endotoxin concentration measured on pasture .................................................. 41

Figure 2-3:

Associations between endotoxin concentrations (103 x EU/m3) in the breathing zone of
horses maintained on pasture and ambient temperature (°C) (A) and relative humidity (%)
(B). Individual diamonds represent individual measurements ............................... 42

Figure 3-1:

Agarose gel electrophoreses of 3p] Lambda DNA — HindIII Digest (lane 1), 3111 18S
rRNA PCR product (lane 2), 3111 TLR4 PCR product (lane 3), 3p] TLR2 PCR product
(lane 4) and 3p] IL-8 PCR product (lane 5). 18S rRNA, TLR4, TLR2 and IL-8 were

ampliﬁed using equine genomic DNA .......................................................... 50

Figure 3-2:
Manual sequenced PCR products using the TLR2 forward (A) and reverse (B) primer...51

Figure 3-3: Cross-species MegaBLAST of human A20 mRNA against horse whole
genome sequences in Trace Database. ........................................................... 52

Figure 4-1:

Clinical parameters in RAO-affected and control horses during stabling and on pasture.
Only horses that contained more than 80% epithelial cells within their BB samples were
included. A) Lung function (APplmax, [cm H20]) and B) percentage of neutrophils in
bronchoalveolar lavage ﬂuid (BAL) (Neu) from control (grey bars) and RAO-affected
horses (black bars) during stabling (day 0) and after stabling (days 7, 14 and 28). (a)
indicates signiﬁcant difference (P S 0.05) between the horse groups (control, RAO)
within a time point (day 0, 7, I4, and 28). (b) indicates signiﬁcant difference (P S 0.05)

x1v

Cor
ant
san
51:11
ind
wit}
bet‘

Fig

Lea
that

Lea.
brus.
TLF
Neu
mRT
hors
0.05
betu

Figu
Lunt
and

bars
grou
diffc

Figu
Iota
al‘l‘c‘c
(data
bars
Indie
”Elm

Fig"
Inter]
Emu;
POOIC
meas

"'\

dilltr

from day 0. ........................................................................................... 69

Figure 4-2:

Relative mRNA quantiﬁcation of A) TLR4 and B) IL-8 in bronchial brushing cells from
control and RAG-affected horses during stabling (day 0) and after stabling (days 7, 14
and 28). Only horses that contained more than 80% epithelial cells within their BB
samples were included. Vertical bars show the 5 to 95 % range with the mean of the
standardized mRNA expression (horizontal bar) (data in antilog transformation). (a)
indicates signiﬁcant difference (P S 0.05) between the horse groups (control, RAO)
within a time point (day 0, 7, 14, and 28). (b) indicates signiﬁcant difference (P S 0.05)
between time points within a horse group. ...................................................... 70

Figure 4-3:

Least square regressions between mRNA expressions and clinical responses. Only horses
that contained more than 80% epithelial cells within their BB samples were included.
Least square regressions between relative amount of TLR4 and IL-8 mRNA in bronchial
brushing samples (BB) in (A) RAO-affected and (B) control; between relative amount of
TLR4 mRNA in BB and percentage of neutrophils in bronchoalveolar lavage (BAL) (%
Neu) in (C) RAO-affected and (D) control horses; and between relative amount of IL-8
mRNA in BB and percentage of neutrophils in BAL in (E) RAO-affected and (F) control
horses including measurements at all time points. For least square regressions with P S
0.05, Rz-values are given in the ﬁgures. Curves represent the least square regression line
between variables. ............................................................................... 71-72

Figure 5-1:

Lung function (APplmax, [cm H20]) in RAD-affected and control horses during (day 0)
and after (day 7, 14, and 28) stabling. Grey bars show RAD-affected horses and black
bars show control horses. (a) indicates signiﬁcant difference (P _<, 0.05) between the horse
groups (control, RAO) within a time point (day 0, 7, I4, and 28). (b) indicates signiﬁcant
difference (P S 0.05) from day 0. .................................................................. 89

Figure 5-2:

Total bronchoalveolar lavage (BAL) neutrophils as a function of time. Data from RAO-
affected and control horses are pooled for times during stabling (day 0) and after stabling
(days 7, 14, and 28). Individual diamonds represent individual measurements. Horizontal
bars represent the geometric means of total BAL neutrophils at each time point. (*)
indicates signiﬁcant (P S 0.05) higher than on days 7, 14, and 28. Total numbers of BAL
neutrophils are loglo transformed. .................................................................. 90

Figure 5-3:

Interleukin (IL)-8 mRNA expression in bronchial epithelial cells between the diagnostic
groups (RAO, control) and between the time points. A) Data from all time points are
pooled for RAD-affected and control horses. Individual diamonds represent individual
measurements. Horizontal bars represent the geometric means. (*) indicates a signiﬁcant
difference (P S 0.05) between the diagnostic groups. B) Data from RAD-affected and

XV

contr
and

repre
a Sig
shou

Figu
Asso
TLR
Total
exprc

Figu
Relat
from
14 at
mR.\‘

Figu;
Asso
and t
311d It

Flgm
A850.
bronc
A30 '
Curr.

Figu]
Flow
the S;

l'anal

Figu,
Asset
(BBS
rRNA

Figur
P0Pul

control horses are pooled for times during stabling (day 0) and after stabling (days 7, 14,
and 28). Individual diamonds represent individual measurements. Horizontal bars
represent the geometric means of IL-8 mRNA expression at each time point. (*) indicates
a signiﬁcant difference (P S 0.05) from days 7, 14, and 28. IL-8 mRNA expression is
shown in relation to 18S rRNA (reference gene) expression. ............................. 91-92

Figure 5-4:

Associations between total bronchoalveolar lavage (BAL) neutrophils and IL-8 (A) or
TLR2 (B) mRNA expression in bronchial brushing samples in RAO-affected horses.
Total numbers of BAL neutrophils are loglo transformed. IL-8 and TLR2 mRNA
expressions are shown in relation to 18S rRNA (reference gene) expressions. ............ 93

Figure 6-1:

Relative mRNA quantiﬁcation of the zinc ﬁnger protein A20 in bronchial brushing cells
from control and RAD-affected horses during stabling (day 0) and after stabling (days 7,
14 and 28). Vertical bars show the 5 to 95 % range with the mean of the standardized
mRNA expression (horizontal bar) (data in antilog transformation). ...................... 106

Figure 6-2:

Association between total numbers of bronchoalveolar lavage (total BAL) neutrophils
and total BAL macrophages (A), total BAL neutrophils and total BAL lymphocytes (B)
and total BAL macrophages and total BAL lymphocytes (C). .............................. 107

Figure 6-3:

Associations between epithelial-derived A20 mRNA expression and total numbers of
bronchoalveolar lavage (BAL) neutrophils (A), macrophages (B) and lymphocytes (C).
A20 mRNA expressions is shown in relation to 18S rRNA (reference gene) expressions.
Curves represent the linear associations between variables. ................................ 108

Figure 7-1:

Flow diagram for developing full models for regression analyses. All variables that are at
the same or at a lower level than the deﬁned outcome are considered to be predictors. All
variable were considered to be outcomes. ........................................................ 125

Figure 7-2:

Association between TLR2 and A20 mRNA expressions in bronchial brushing samples
(BBS) from all horses. TLR2 and A20 mRNA expressions are Shown in relation to 18S
rRNA (reference gene) expressions. ............................................................ 126

Figure 7-3: A potential model for predictors of airway inﬂammation in the general horse
population. ......................................................................................... 1 27

xvi

°/o
°/o
%.
A2
AF
Ar
BA
BB
C5:
CB
CX
DN
E31
E17
FC -
fMl
G-C
GM-
GPC
Grd
(3er
H81
hBD
HDS
H.\l(
HSP
ICA.‘
IFNI
IF N
Ig-j;
IKK
IL -;
IL-R
Hm

ISH .
I-KB .

LBP .
log“:
IOgNe
IOgTC

LIST OF ABBREVIATIONS

%Lym - percentage of lymphocytes in BAL
%Mac - percentage of macrophages in BAL
%Neu - percentage of neutrophils in BAL
A20 - zinc ﬁnger protein A20
AF - aspergillus fumigatus
Arab — arabian horse
BAL - bronchoalveolar lavage ﬂuid
BB - bronchial brushing sample
CSa - complement protein 5a
CB — clenbuterol
CXCL - CXC chemokine ligand
DNA - deoxyribonucleic aci
EMSA - electromobility shift assay
EU - endotoxin unit
FC - fold change
fMLP - forrnyl-methionylleucylphenylalanine
G-CSF - granulocyte-colony stimulating factor
GM-CSF - granulocyte macrophage-colony-stimulating factor
GPCR - G-protein-coupled receptor
Grd - grade horse
GRK - GPCR kinase
H&E - Hematoxylin and Eosin
hBD - human B-defensin
HDS - hay dust suspension
HMGB - high-mobility group box
HSP -heat shock protein
ICAM - intracellular adhesion molecule
IF NR - IFN receptor
IF N-y - interferon gamma
1g - immunoglobulin
IKK complex - inhibitor of nuclear factor-KB (IKB)-kinase complex
IL — interleukin
IL-R - interleukin receptor
IRAK - IL-1 receptor associated kinase
IRF - interferon regulatory factor
ISH - in-situ hybridization
I-KB - inhibitor of NF-KB
kDa — kilo Dalton
LBP - LPS-binding protein
logLym - log10 transformed total BAL lymphocyte count
logMac - loglO transformed total BAL macrophage count
logNeu - loglO transformed total BAL neutrophil count
IogTCC - log10 transformed total BAL cell count
xvn

LPS - lipopolysaccharide

LRR - leucine-rich repeat

LTB4 - leukotrien B4

Lym - total BAL lymphocyte count

Mac - total BAL macrophage count

MAC - macrophage antigen

MAL - MyD88 adaptor-like

MAP - mitogen-activated protein

MBL - mannan-binding Iectin

MCP - monocyte chemotactic protein

MD2 - macrophage-derived molecule 2

MIP - macrophage inﬂammatory protein

mRNA — messenger RNA

MyD88 - myleloid differentiation primary response protein 88
n.m. - not measured

n/a - not applicable

Neu - total BAL neutrophil count

NF-KB - nuclear factor kappa B

NOD — nucleotide-binding oligomerization domain
PAF - platelet-activating factor

PAR — protein-activated

PBS - phosphate buffered saline

poly[I;C] - polyinosinic-pochytidylic acid

PRR - pattern recognition receptor

QH - quarter horse

qRT-PCR - quantitative real-time polymerase chain reaction
RNA — ribonucleic acid

S.D. - standard deviation

S.E.M. — standard error of the mean

SB - standard bred

SCC - Spearman correlation coefﬁcient

SP-A - surfactant protein A

SPAOPD - summer pasture associated obstructive pulmonary disease
STAT - Signal transducer and activator of transcription
TAB - TAKl-binding protein

TAK - transforming-growth-factor-B-activated kinase
TB - thoroughbred horse

TCC - total BAL cell count

TICAM - TIR-containing adaptor molecule

TIR - toll/interleukin-l receptor

TIRAP - TIR domain-containing adaptor protein
TLR - toll-like receptor

TNFR — TNF receptor

TNF-a - tumor necrosis factor alpha

TRAF6 - TNF receptor-associated factor 6

TRAM - TRIP-related adaptor molecule

xviii

TR5
TR]
I’B(
l'E\
APp

TRAM - TRIF-related adaptor molecule

TRIF - TIR domain-containing adaptor inducing IFN-B
UBC 13 - ubiquitin-conjugating enzyme 13

UEVlA - ubiquitin-conjugating enzyme E2 variant 1
APpl..m — maximal change in pleural pressure

xix

fun
pku
mut
anx-
beet

cihz

prOI:
mOn

Shov

INTRODUCTION

Airborne pathogens, such as bacterial products (e.g., lipopolysaccharide [LPS]),
fungal (e.g., Spores, mycelium fragments, glucan), and yeast (e.g., zymosan) elements,
plant pollens, and dust mites are inhaled and subsequently deposited on the airway
mucosa. Different innate immune mechanisms involved in removing and defending the
airways against these pathogens are present in the lung. Initially, inhaled pathogens
become embedded in mucosal ﬂuid and are removed by active transportation due to the
ciliary movement of the airway epithelium. In addition, airborne particles are recognized
by soluble factors, which are released into the mucosal surface ﬂuid. Finally, pathogen
recognition receptors are expressed on cell membranes of a variety of airway cells, such
as alveolar macrophages, dendritic cells, epithelial cells, and airway smooth muscle cells.
These innate receptors belong to different families, such as the toll-like receptor (TLR)
family, the nucleotide-binding oligomerization domain (NOD) receptor family, and the
protein-activated receptor (PAR) family. In vivo animal studies and in vitro studies using
monocytes, alveolar macrophages, and airway epithelial cell cultures have been used to
Show the release of cytokines and chemokines (preferentially interleukin (IL-)IB, tumor
necrosis factor (TNF-) or, IL-6, IL-8 etc.) upon stimulation with LPS, fungal and yeast
products, dust mites, and pollen. The signaling cascades of recognition receptors are
under feed-forward and feed-back regulatory mechanisms (e.g., feed-back mechanism
through the zinc ﬁnger protein A20 of the TLR4 and TLR2 signaling pathways).

Recurrent airway obstruction (RAO) is a common airway disorder in horses

characterized by recurrent neutrophilic airway inﬂammation and airway obstruction due

10 6
cont
inve
recei
inna‘
and
inclu
ants
LPS.
the n
oftl
fanng
neuu'
exacc
Althr-
LPS.
the e)
incon
Pt
and r
Mo.
Inﬂan

f0IIOU

to exposure to stable dust. Originally, it had been suggested that RAO is an allergic
condition — especially, a type I hypersensitivity reaction has been suggested by different
investigators — in response to inhalation of allergens found in hay dust. However, in the
recent literature there is also evidence that the pathogenesis of RAO may involve aberrant
innate immune mechanisms. Stable dust is a mixture of organic (e.g., bacterial, fungal,
and viral) products and, therefore, could potentially stimulate Signaling pathways,
including those of TLR3. Toll-like receptors are the major cell membrane receptors that
are stimulated by the recognition of conserved structures found in microorganisms (e.g.,
LPS, peptidoglycan) and endogenous products produced during acute inﬂammation. In
the recent literature, inhalation of endotoxin from gram-positive bacteria and inhalation
of fungi, such as Faem'a rectivirgula, T hermoactinomyces vulgaris, and Aspergillus
fumigatus, have been shown to trigger RAO in susceptible horses. Furthermore, the
neutrophilic chemokine IL-8 is elevated in RAD-affected horses during acute
exacerbations compared to control animals or to RAO-affected horses during remission.
Although there were some attempts to study TLR4, the principal recognition receptor for
LPS, and TLR2, the principal recognition receptor for fungal products, in equine airways,
the exact mechanisms involved in pathogen recognition and subsequent signaling are still
incompletely understood.

For my dissertation, I developed a pathway model (Figure 1-1) including molecules
and receptors involved in TLR signaling pathways that could potentially be altered in
RAD-affected horses and, therefore, contribute to the development of airway
inﬂammation and obstruction during stabling. The overall organization of my dissertation

follows the ﬂow of this pathway model. Starting with the determination of endotoxin

cont
sub:
(BB
betxi
mabl

paﬂi

pope

.w

L»)

concentrations in the breathing zone of horses while stabled or maintained on pasture, I
subsequently quantiﬁed and compared gene expressions in bronchial brushing samples
(BBS) at the messenger ribonucleic acid (mRNA) level for TLR4, TLR2, IL-8, and A20
between RAO-affected and control horses during and at three different time points after
stabling. Finally, I adjusted the hypothetical pathway model for RAO to develop a
pathway that is potentially involved in airway inﬂammation in the general horse
population. The following is an overview of the topics covered in this dissertation:

1. The determination of endotoxin concentrations in the breathing zone of stabled
horses compared to horses on pasture. The mean concentrations within stables
were at least 8-fold higher than the inhaled endotoxin concentration on pasture
(chapter 2);

2. The quantiﬁcation of mRNA expression of TLR4 and IL-8 in BBS and their
correlation with each other. TLR4 and IL-8 mRNA expression were increased
during stabling compared to periods on pasture and correlated with each other
(chapter 4);

3. The quantiﬁcation of mRNA expression of TLR2 in BBS and its relation to the
IL-8 expression. TLR2 mRNA expression was not effected by stabling nor
correlated with IL-8 mRNA expression (chapter 5);

4. The quantiﬁcation of mRNA expression of A20 in BBS and its relation to
inﬂammatory cell counts in bronchoalveolar lavage ﬂuid (BAL). A20 mRNA
expression was not different between diagnostic groups and time points.
However, when data from both diagnostic groups were pooled, A20 in BBS and

inﬂammatory cells in BAL correlated negatively (chapter 6);

hly.
knov
\xritt:
mate
pubh
inder

ordr:

5. A statistical pathway analysis discussing airway inﬂammation in RAO-affected
horses and the general horse population (chapter 7);
6. An outlook for potential future research developments in studying immune
mechanisms involved in RAO’S pathogenesis (chapter 8).
My dissertation is divided into the following chapters. Chapter 1 provides the current
knowledge in the literature about the studied scientiﬁc ﬁeld. Chapters 2, 4, 5, and 6 are
written in the form of scientiﬁc peer-review papers, including abstract, introduction,
materials and methods, results, and discussion. Their modiﬁed versions have been
published or have been submitted for publication. Whereas these chapters can be read
independently of each other, chapters 3, 7, and 8 either provide background information

or draw conclusions about the research results obtained in chapters 2, 4, 5, and 6.

Fig!

M

Figure l-l: Pathway model of Toll-like receptors that are potentially involved in

RAO pathogenesis.

TLR patthays in RAO pathogenesis

Barn dust

Arte inflammationJ TLR2 T

TLR41 ;
l A20 9

 

 

 

 

1i
Incie -uLsr
I

 

 

 

7

Neutrophilic airway inflammation T ‘

 

 

 

 

 

1

Airway obstruction & Mucus accumulation T

 

 

 

Rec:

puln‘
airw
expo
diffu
bron.

\Vhic

CHAPTER 1

LITERATURE REVIEW

Recurrent Airway Obstruction - Definition and Pathogenesis

Recurrent airway obstruction (RAO), also known as heaves, is a common
pulmonary disorder in horses. Acute exacerbations are characterized by neutrophilic
airway inﬂammation, mucus accumulation, and bronchospasm, which are triggered by
exposure of susceptible horses to stable and hay dust. Recurrent airway obstruction is a
diffuse airway disease involving large conducting airways down to the terminal
bronchioles. Periods of acute exacerbation are interspersed by periods of remission,
which can be achieved by returning RAD-affected horses from stables to pasture (151).

The mechanisms responsible for airway inﬂammation in RAD-affected horses are
not well understood. Originally, it was suggested that RAO is an allergic condition (151).
Elevated immunoglobulin (lg) E levels in bronchoalveolar lavage (BAL) of affected
animals support the assumption of a type-I hypersensitivity reaction (63, 162). Increased
expression of T-helper cell (Th) type 2 (Th-2) Speciﬁc cytokines, such as IL-4 and IL-5,
as well as decreased expression of Th-l speciﬁc cytokines, such as interferon gamma
(IFN-y), further support the suggestion that adaptive immune mechanisms are involved in
the pathogenesis of RAO (18, 29, 59, 99). Interleukin-4 is important for the
differentiation of uncommitted T-cells into the Th-2 subset and, therefore, an increase in
Th-2-type cytokine production. F urtherrnore, IL-4 is required for B-cell isotype switching
to IgE antibody production, which might contribute to the local IgE production in the

lungs of RAO-affected horses during acute exacerbations (63, 162). Interleukin-5 is a

CNIO
U16 I
5UPI
inha

obst

the

sutﬁ
rece
cont
coul
the]
by ;

acat

epr

the u

cytokine involved in proliferation of granulocyte progenitors, especially eosinophils, in
the bone marrow (119). The increase in IL-5 in stabled RAD-affected horses (29, 99) is
supported by an increase in IL-5 mRNA expression in sensitized ponies after ovalbumin
inhalation challenge (1 8). However, horses affected by summer pasture associated
obstructive pulmonary disease (SPAOPD) did not exhibit a difference in IL-5 production
(77). Since RAO is characterized by a neutrophilic rather than an eosinophilic inﬂux into
the airway lumen, the reported increase of IL-5 in BAL lymphocytes needs to be the
subject of further studies. Interestingly, there was an increased expression of the IL-5
receptor (IL-5R) on peripheral neutrophils from RAO—susceptible horses compared to
controls on pasture and compared to RAO-affected horses after stabling (34). Thus, it
could potentially be interpreted that IL-5 might have a role in neutrophil recruitment into
the lung. On the other hand, mRNA expression of IL-17 — a cytokine, which is secreted
by activated T-cells and which is indirectly involved in maturation, chemotaxis and
activation of neutrophils — has been shown to be higher in the BAL of RAO-affected
horses than in control horses after stabling (4, 31). Cytokine expression proﬁles in RAO-
affected horses have been studied by different investigators and are inconsistent among
their studies (3, 29, 59, 99). For example, while Lavoie et al. showed an increase in IL-4
and IL-5 mRNA expression and a decrease in IFN-y expression in RAO-affected horses
during stabling (29, 99), Ainsworth et a1. did not ﬁnd changes in IL-4 and IL-5 mRNA
expression but rather Showed an increase in IFN-y mRNA expression (3, 4). The latter is
supported by Giguere et al., who also showed an increase in IFN-y (3, 59). One should
keep in mind that the different results might be due to either different protocol designs or

the use of different analytical methods (e.g., in situ hybridization vs. q-PCR etc.). Finally,

the
diﬂ

ada

rok
pro«
dep
’Ihe
end.
inﬂ;
ahbt
inha
susc
Chen

BAL

End(

pamu;

172».

breath

the ﬁnding that allergic skin reactions to common barn allergens lack Signiﬁcant
differences between affected and control horses (41, 115) supports the assumption that
adaptive immune mechanisms may only partially explain the RAO disease pathogenesis.
Beside adaptive immune mechanisms, innate immunity seems to play a crucial
role as well (142). Hay dust is a mixture of organic (e.g., bacterial, fungal, and viral
products) and non-organic (e.g., metals, particles) materials. Inhalation of endotoxin-
depleted hay dust Signiﬁcantly attenuates airway neutrophilia in affected horses (144).
The inﬂammatory response can be re-established by adding endotoxin back to the
endotoxin-depleted hay dust (141). These observations conﬁrm that attenuation of airway
inﬂammation is due at least partially to endotoxin. Furthermore, horse stables contain
airborne fungi (128), such as Faenia rectivirgula and Aspergillus fumigatus, that when
inhaled can induce neutrophilic airway inﬂammation and deterioration of lung function in
susceptible horses (33, 116, 143, 162). Finally, pro-inﬂammatory cytokines and
chemokines, such as IL-IB, TNF-a, and IL-8 have been shown to be increased in either
BAL cells or airway epithelial cells in RAD-affected horses during acute exacerbations

(4,13, 51, 59).

Endotoxin-activated Immune Mechanisms in the Lung

In conventional horse stables, airborne endotoxin concentrations exceed those on
pasture (117) and those that can induce airway inﬂammation in human subjects (171,
172). In our own stables, we observed that mean endotoxin concentrations in the
breathing zone of stabled horses are at least 8-fold higher than concentrations on pasture

(chapter 3). While endotoxin is present in barn dust in high concentrations, barn dust does

not

that

pulr
locc

exis

Stabl

breai

not cause exaggerated airway inﬂammation in stabled control horses. Marti et al. showed
that there is a greater risk of RAO in the offspring from two affected parents than if only
one or no parent were affected (112). This suggests that RAO is a multifactorial equine
pulmonary disorder, in which environmental and genetic factors combined cause the
phenotypical outcome.

In many stables airborne endotoxin concentrations exceed those that can induce
pulmonary inﬂammation and bronchial hyperresponsiveness in human subjects
(occupational asthma), and those that can cause bronchospasm in humans with pre-
existing pulmonary conditions (25, 26, 171, 172). It has been reported that respiratory
symptoms of humans increase when the air endotoxin concentration is greater than 250
Endotoxin Units (EU)/m3 (l ng/m3 ~ 10 EU). Therefore, the ‘Dutch Expert Committee on
Occupational Standards’ has recommended a health-based occupational exposure limit of
50 EU/m3 over an eight-hour period of endotoxin exposure. However, endotoxin levels in
stables greatly exceed these limits (177), and may be as high as 35,000 EU/m3 within the
breathing zone of stabled horses (chapter 3).

Pathophysiology: Inhalation of endotoxin has been linked to increased amount of
neutrophils in BAL (129, 140, 142-144, 169). This may be either due to an elevated
migration of neutrophils from the pulmonary vasculature into the airway lumen or due to
delayed neutrophil apoptosis and clearance from the air lumen by alveolar macrophages
(20). Endotoxin stimulates neutrophil migration on different levels. In other animal
models and in in vitro studies, potent CXC-chemokines (e.g., IL-8 and macrophage
inﬂammatory protein 2 [MIP 2]), Chemoattractants (e.g., platelet-activating factor (PAF),

leukotrien B4 (LTB4), complement protein 5a [C5a]), and pro-inﬂammatory cytokines

hura
expc

exhi

cone
(36L
pota‘
\tau:
(120
least
hnkc
prese
endo

endo'

(e.g., TNF, IL-1 and IFN-y) that are involved in modulation of adhesion molecules (e.g.,
an L-selectin ligand on endothelial cells (176), P-Selectin (114), E-selectin (88, 163), and
intracellular adhesion molecule 1 [ICAM-l]; (80)) are elevated due to endotoxin
exposure (189, 190). During acute exacerbations due to stabling, RAO-affected horses
exhibit elevated amounts of IL-8 (4, 13, 50, 51, 59), MIP-2 (51), and ICAM-1 (21).

Sources of endotoxin in the environment: Endotoxin is found at high
concentrations in agricultural dust, such as dust from grain (164), poultry (35), swine
(36), dairy cow (94), and horse housings (117, 145) as well as in cotton mills (27) and
potato plants (76). Furthermore, endotoxin exposure is also an important contaminant in
water-mist generating industries, such as the paper (158), metal (1, 150), and ﬁber-glass
(120-122) industry. It has also been suggested that the sick building syndrome results at
least partially from the endotoxin loads found in ofﬁce buildings (182). Humidiﬁers are
linked to an increased endotoxin concentration, whereas dehumidiﬁers decrease it. The
presence of a dog in family homes has been Shown to be the strongest predictor for
endotoxin levels (120). Cigarette smokers are associated with an increased exposure to
endotoxin than non-smokers (64).

Measurement of endotoxin — Endotoxin concentrations are commonly detected
by the Limulus amebocyte lysate assay (LAL assay) (180). Amebocytes, which are
nucleated, granular cells that aggregate and clot at the site of an injury, from the
hemolymph of the horseshoe crab (Limulus) are used for this test. Aggregation of the
amebocytes in the limulus is caused by the endotoxin originating from the blue-green
algae. The content from the intracytoplasmic granules from the amebocytes is used in the

LAL assay. Enzyme activation can either be determined by clot formation (clot end-

10

pent

Chrt

dilu
ﬂalo

regt

Inf]
serh
celh

and

and

Selec
I3oth
Imihr
TNF.
exPre
and E
Presel

Thobjl

point) or by color changes over time using a dye that reacts with the enzyme (kinetic
chromogenic). The latter is frequently used to determine endotoxin in environmental
samples (160). Different endotoxincollecting and analytical methods (e.g., extraction,
dilution, storage, commercially available lysates) contribute to varying results (up to 100-
fold) among laboratories and, therefore, contribute to difﬁculties in establishing safety

regulations for endotoxin exposure (28, 177).

Recruitment of Inflammatory Cells into the Lung

Neutrophil Emigration Mechanisms that are Important during Lung
Inﬂammation - Neutrophil migration from blood to the lung lumen can be seen as a
series of sequential steps, including 1) capture and rolling of neutrophils on endothelial
cells (ECS) in postcapillary venules; 2) ﬁrm adhesion, which is mediated by integrins;

and 3) transmigration.

Capture and rolling is directed by adhesion molecules, called selectins. E-selectin
and P-Selectins are expressed on endothelium and L-selectins on neutrophils. Whereas L-
selectin is important for neutrophil rolling on ECS, P-selectin is crucial for their capture.
Both, L- and P-selectin work in combination to initiate migration processes during
inﬂammation (30). Selectins are known to be autoproteolyzed in response to LPS and
TNF-o: (190). On the other hand, in vitro studies showed induction of L-Selectin ligand
expression after LPS or cytokine exposure of ECS (1 76). Endothelium-bound selectins (P-
and E-selectin) are membrane-expressed only when appropriate inﬂammatory stimuli are
present. P-selectin is stored in intracellular vesicles (Weibel-Palade bodies) and is

mobilized to the cell surface within minutes after exposure (114). E-Selectins are

11

tran:
expr
of n

Stab

by

C Di
neut
upor
endt

levc

1) n0

transcribed upon stimulation with inﬂammatory cytokines, so that their membrane
expression is highest after 4 to 6 hours of stimulation (88, 163). The earliest accumulation
of neutrophils within the airways of RAO-affected horses was reported after 5 hours

stabling (20).

During the processes of ﬁrm adhesion, neutrophils bind to ECS, which is mediated
by CD18-containing integrins (e.g., CD11a/CD18 and CD11b/CD18). The integrin
CD11b/CD18 (also termed macrophage antigen-l [MAC-1]) plays a crucial role in
neutrophil migration during inﬂammatory processes (107). It is intracellularly stored and,
upon stimulation (e.g., LPS, TNF-a, etc.), it is translocated to the cell membrane (8). An
endothelial ligand for CDllb/CD18 is ICAM-l, which in general is expressed at low
levels. However, due to stimulation of ECS by cytokines, ICAM-l becomes strongly up-
regulated (80, 88, 163). In addition, LPS and IL-8 stimulate the production of integrin-
modulating factor-1 on neutrophils, which is responsible for augmenting CD18-ligand
binding (89). ICAM-l (21) and IL-8 (4, 13, 50, 51, 59) are elevated in lungs from RAO-

affected horses during acute exacerbations.

The ﬁnal step of recruitment of neutrophils into the lung is called transmigration.
Migratory processes are accompanied by release of neutrophil-derived proteases (32). In
RAO-affected horses it has been shown that equine neutrophil elastase A2 is increased in

bronchoalveolar lavage during stabling than on pasture (20).

Soluble Mediators of Migration — There are two classes of promigratory stimuli,

i) nonchemotactic cytokines and ii) Chemoattractants.

12

expr
Chen

rehx

lL-l
hors

ofat

Cher
116111
pare
trans
rece]
neut

Prod

Cﬂokines: Inﬂammatory cytokines, for example TNF-a and IL-lB, increase
expression of adhesion molecules on neutrophils and ECS, but they do not exhibit
chemotactic functions for neutrophils. Within the lung, TNF-a and IL-IB are primarily
released by alveaolar macrophages due to stimulation with inﬂammatory stimulants, such
as LPS. Laan et al. showed that after 6 and 24 hours of stabling, production of TNF-a and
IL-lB by alveolar macrophages from RAO-affected horses was higher than in control
horses (95). TNF-0t and IL-lB activate neutrophils and ECS by stimulating the expression
of adhesion molecules (e.g., ICAM-1) and Chemoattractants (e.g., IL-8).

Chemoattractants: Beside the increase of adhesion molecule expression,

 

Chemoattractants, such as IL-8, LTB4, C5a, and PAF primarin direct migration of
neutrophils through tissue. They are released by macrophages, neutrophils, lymphocytes,
parenchymal cells, endothelial and epithelial cells, etc. The initiation of neutrophilic
transendothelial and tissue migration is initiated by binding of Chemoattractants to
receptors on the neutrophil surface. Speciﬁc receptors for Chemoattractants on
neutrophils have been shown for chemokines, such as IL-8, for C5a, LTB4 and bacterial
products (e.g., endotoxin or formyl-methionylleucylphenylalanine [IMLP]).

Chemokines are small proteins (6 to 15 kDa) and function as the primary stimulus
for neutrophilic migration. Interleukin-8 is the most studied chemokine in this group and
has been shown to be produced by neutrophils themselves as well as endothelium,
epithelium, and a variety of parenchymal cells (139). Interleukin-8 belongs to the CXC-
chemokines. Chemokines are divided into two classes due to the sequence of their two
most amino-terminal cysteines. The CXC chemokines (also termed (it-chemokines)

contain an amino acid (X) between the cysteines and attract most strongly neutrophils,

l3

whet
each
Adtht
of ot
pnne
Chen‘
pnne

CINK

Fran.
incre
hors:
subs
Para]
IL-8
hon]
that

bent

ﬁabl

whereas the cysteines in CC chemokines (also termed [iv—chemokines) are located next to
each other and elicit the stronglest chemoattractive effects for mononuclear cells.
Although IL-8 is the most potent stimulus for neutrophil migration in humans and most
of other mammals, rodents (e.g., mouse, rat) do not have IL-8. However, they have
proteins similar to growth-related oncogenes called cytokine-induced neutrophil
Chemoattractants (CINCS), for example CINC3, also called macrophage-inﬂammatory
protein-2 (MIP-2) (166). LPS-stimulation of rat macrophages leads to elevated release of
CINCS (166). It has been shown that CINCS are involved in neutrophilic response in the
lung due to bacterial and inhalation challenges (38, 52, 53).

Interleukin-8 has been studied in RAO-affected horses by several research groups.
Franchini et al. showed for the ﬁrst time that a high dust exposure was paralleled by an
increased chemotactic activity for neutrophils in BAL from RAO-affected versus control
horses. Isolated macrophages showed the secretion of two Chemoattractants, which were
subsequently identiﬁed as IL-8 and MIP-2 (51). In addition, IL-8 was increased in
parallel with the percentage of neutrophils in BAL (50). Later, Giguere et al. showed that
IL-8 protein concentrations and mRNA expression were signiﬁcantly higher in BAL
from heaves-affected horses than from controls (59). Recently, Ainsworth et al. reported
that IL-8 mRNA expression did not differ in bronchial epithelium and BAL neutrophils
between RAO-affected horses and controls during remission and after one day of
stabling. However, after 14, 35, and 49 days of stabling, RAO-affected horses exhibited
an up-regulation in IL-8 mRNA in epithelial cells (3.3-fold, 8.5-fold, and 10.3-fold,
respectively). After 1, 14, 35, and 49 days IL-8 was higher in BAL neutrophils (3.7-fold,

9.5-fold, 7.6-fold, and 14.8-fold, respectively) due to its down-regulation in controls,

14

foll
par:
lL-l
CS]
bee

alte

DEU

aim

neu-
“hi
new
beer

adm

followed by an up-regulation in RAO-affected horses. The mRNA expression was
paralleled by increased staining of IL-8 protein in bronchial biopsy samples (4). Beside
IL-8, other chemokines, such as granulocyte-macrophage colony-stimulating factor (GM-
CSF), granulocyte CSF (G-CSF), and growth-related oncogene a (CXCLl) have also
been investigated in airway epithelial cells; however, their gene expression was not
altered between affected and control horses (4).

Leukotriene B4 (LTB4) iS a product of the arachidonic acid pathway and is
released from stimulated monocytes and neutrophils. It functions in neutrophilic
chemotaxis and supports the attachment of neutrophils to the endothelial cells in the
vasculature (131). Our laboratory reported previously that stabling is associated with an
increase in LTB4 in RAO-affected horses and that this increase was correlated with the
migration of neutrophils into the airway lumen (103).

Complement protein C5a is a cleavage product of the complement protein C5 in
the classical complement as well as the alternative complement pathways. In the airways,
C5 iS produced by alveolar macrophages and type II pneumocytes (178). Therefore,
neutrophils migrate along the C5a gradient, with their highest concentration within the
airway lumen and declining concentrations towards vasculature.

Another chemoattractant is PAF, which is produced by endothelial cells, platelets,
neutrophils, and alveolar macrophages. It consists of an acetylated phosphoglycerolipid,
which originates from cell membranes. Beside chemotaxis, PAF also modulates
neutrophilic adhesion to endothelial cells (15, 93, 202). Platelet-activating factor has also
been studied in RAO-affected horses in the last decade. Marr et al. showed that

administration of a PAF receptor antagonist (WEB 2086) did not suppress the

15

net
Stal
pat
oth
an i
35}?
rece
into
add:
(IOE
afhet

inhﬂ

Path

recoé
Inobj
urider
micro
enabh
come
defens

neutrophilic migration from the vasculature into the airway lumen after seven hours of
stabling. Therefore, it was suggested that PAF may not play a central role in the
pathogenesis of neutrophilic airway inﬂammation in RAO-affected horses (110). On the
other hand, F airbain and colleagues Showed that intravenous administration of PAF led to
an immediate but transient increase of radiolabeled neutrophils in the lung of healthy and
asymptomatic RAO-susceptible horses. This response could be inhibited by the PAF
receptor antagonist. Therefore, PAF might be involved in the chemotaxis of neutrophils
into the lung due to inﬂammatory stimuli in the general horse population (42). In
addition, phosphodiesterase isoenzymes (PDE) have been shown to inhibit PAF release
(109). However, the PAF-induced adhesion of equine neutrophils obtained from RAO-
affected horses to ﬁbronectin-coated plastic could not be modulated with the PDE4

inhibitor rolipram (149).

Pathways of Innate Immune Mechanisms within the Lung

Introduction: Our understanding of the innate immune system originated with the
recognition of leukocytes in inﬂamed tissue in 1779, and Metchnikov’s discovery of
mobile ameboid cells (phagocytes) in sea anemones in 1882 (168). Our current
understanding of innate immunity includes studies about microbial binding proteins and
microbial-recognizing receptors. Throughout evolution, strategies have evolved that
enable the differentiation between airborne pathogenic and nonpathogenic particles that
come in contact with the pulmonary epithelium. In higher vertebrates, two interactive
defense mechanisms, the innate and the adaptive immunity, have been developed. The

evolutionary older system, the innate immune system, consists of i) soluble factors,

16

releas
theret
phage
nncrc
chent
Thei
and b

specn

Specil
preset
produ
expre:
(ﬂear

adapri
huUau

earlier

betuee
epithei
ab0ut]
breath e

large or

released into the airway mucosa by a variety of cells, that bind microbial products and,
therefore, make them available for clearance by phagocytic cells; ii) innate receptors on
phagocytic cells (e.g., macrophages, neutrophils, dendritic monocytes), that bind
microbial products and initiate other defense mechanisms (e.g., production of
chemokines); and iii) innate receptors on airway epithelial cells (e.g., toll-like-receptors).
The innate immunity is constantly ready to recognize and eliminate microbial products
and by recognizing common microbial patterns, the innate immune system exhibits broad
speciﬁcity.

In contrast, the adaptive immune system consists of lymphocytes that
speciﬁcally recognize peptide sequences that are presented on specialized antigen-
presenting cells (APCS). Upon recognition, these lymphocytes respond with the
production of high-afﬁnity antibodies that are capable of opsonizing microbes that
express the same speciﬁc peptide sequences. Leukocytes (e. g., macrophages, neutrophils)
clear opsonized microbial products within the airways or adjacent lymph nodes. The
adaptive immunity is further characterized by its memory (i.e., specialized lymphocytes
initiate a faster response upon recognition of microbes, which were encountered at an
earlier time point).

Innate Immunity of the Airways: The respiratory system is the largest interface
between host and environment. For humans, it has been estimated that the alveolar
epithelium alone equals about the size of a tennis court (199). At rest humans breathe
about 10,000 liters (11) of air daily and it has been established that the average horse
breathes at least 70,000 liters of air daily. Therefore, the lung is constantly exposed to a

large number of airborne pathogens. However, infectious airway diseases are relatively

l7

rare 2
ahwva
ceHul
hno 1
anhel
contp
mono
dﬂler
Ieuko

ofnﬂ

10 hﬂ
relea:
and (
recru
couk
and 1

epith

IIPOI
HO“

Still

rare and this is evidence for a highly efﬁcient defense mechanism located within the
airway mucosa. The innate immunity of the airways is a combination of humoral and
cellular components. The humoral component includes antimicrobial products released
into the epithelial lining ﬂuid, such as collectins, defensins, lysozyme, lactoferrin,
cathelicidins, complement, and immunoglobulins (e.g., IgG) (199). The cellular
component consists of leukocytes, such as neutrophils, alveolar macrophages, dendritic
monocytes, cytotoxic T cells, natural killer cells, and airway epithelial cells. Although
different innate receptor families, such as TLRS, NODS, and PARS, can be found on
leukocytes and epithelial cells, TLRs are the major cell surface receptors for recognition
of microbial and toxic endogenous products (199).

Recognition of respiratory pathogens by the airway epithelium: As a response
to inhaled pathogens, airway epithelial cells activate defense mechanisms, such as the
release of microbicidial products into the airway lumen as well as the release of cytokines
and chemokines into the submucosa. Chemokines and cytokines play a crucial role in
recruiting phagocytes for clearance of airways and airway wall of airborne pathogens that
could not be removed by the movement of the mucosal cilia. Additionally, dendritic cells
and lymphocytes are also attracted by mechanisms of epithelial cells. Therefore, airway
epithelium plays a central role in stimulating and modulating the immune response in the
respiratory system.

The response of epithelial cells to inhaled pathogenic molecules, such as
lipopolysaccharides and leipoteichoic acid, has been studied over the last decades.
However, detailed mechanisms and molecules involved in the recognition process were

still incompletely understood. Airway epithelium uses pattern recognition receptors

18

(PRR
'Ihese
or car
mann
mudk
modu
TLRs
the at
T-ccl
prodL
anon]
ﬁmctf
funga

aﬁer

(PRRS) for recognition of conserved molecular patterns presented on inhaled pathogens.
These receptors can be found as transmembrane receptors on epithelial cells (e.g., TLRS)
or can be released into the epithelial lining ﬂuid or the circulation in soluble forms (e.g.,
mannan-binding lectins [MBLS]). TLRS are recognized in almost all cells that have been
studied. Their function lies not only in initiating innate immune processes, but also in
modulating the cross-talk between the innate and the adaptive immunity. For example,
TLRS have been shown to be distributed widely on dendritic cells, where they modulate
the adaptive immune response by inﬂuencing dendritic cells in the way they interact with
T-cells. In airway epithelial cells, as in many other cell types, recognition of microbial
products by TLRS leads to the production of cytokines and chemokines, as well as
antimicrobial peptides. Additionally, TLRS have also been reported to modulate the
function of surfactant protein A. For that reason, TLRS play a crucial role in chronic
fungal and allergic asthma and inﬂammatory lung diseases, as well as allograft rejection
after lung tranplantation.

T oll-like receptors ( T LRs) are highly conserved across the animal kingdom and
were ﬁrst recognized as proteins involved in the dorsal-ventral development in
Drosophila melanogaster. Later they were found to be important in antiftmgal defenses.
Medzhitov et a1. recognized that the intracellular domain of one of these proteins, termed
Drosophila Toll, was highly homologous with the human interleukin (IL)-1 receptor (12).
Therefore, TLRS belong to the Type I transmembrane receptor family (i.e. TLR/IL-l R
superfamily), which bears a leucine-rich repeat (LRR) domain at the extracellular amino
(NH2)-terminus and a toll/interleukin-l receptor (TIR) domain at the intracellular

carboxy (COOH)-terminus, which is critical for adaptor protein binding. Currently,

19

thhte
been
hganl
intht
paﬂu:
124.
cc” c
grou;
duﬁn
such
inﬂar
macrt
bhuh
and

Prepa
Prepa

thirteen mammalian TLRS have been identiﬁed. In mice, TLRl-9 and TLR1 1-13 have
been discovered. Table 1-1 gives an overview of human and mouse TLRS and their
ligands. Among the TLRS, TLR4 and TLR2 have been the most intensively studied TLRS
in the recent literature. Their most important function is the detection and response to
pathogen associated molecules, such as LPS (TLR4) or lepoteichoic acid (TLR2) (71,
124, 146, 147). However, endogenous factors that are released during inﬂammation and
cell damage can also activate TLRS. For example, the transcription factor high-mobility
group box 1 (HMGBI) can activate TLR4 and TLR2 when it is released extracellularly
during inﬂammation (138). In addition, TLR4 responds to tissue break-down products,
such as hyaluronan oligosaccharides (181, 183), and to proteins that are common at
inﬂammatory sites, such as ﬁbrinogen (173), surfactant protein A (SP-A) that activates
macrophages by binding to TLR4, and B-defensin-2, which activates dendritic cells by
binding to TLR4. However, activation of TLR4 by heat Shock protein 60 and 70 (HSP60
and 70) is potentially caused by LPS contamination of commercially available
preparations (54, 55) as it was the case with TLR2 activation by LPS-contaminated
preparations (71). Toll-like receptors have been Shown to be expressed in a variety of
lung cells, such as alveolar macrophages, dendritic cells (183, 188), epithelial cells (61,
123), endothelial cells (10, 181), and airway smooth muscle cells (125).

T all-like receptor 4 is best known for its response to LPS. In 1998-99, research
groups identiﬁed the gene responsible for LPS resistance in two inbred strains of mice
(C3H/HeJ and C57BL/10ScCr) (78, 146). Whereas C3H/HeJ mice have a single point
mutation in the cytosolic TIR domain (see below), C57BL/IOScCr mice do not express

TLR4 mRNA. In C3H/HeJ mice a point mutation in the 712th amino acid of TLR4

20

replat
the at
seque
in the
pol yr.
TLR-
patier
\Vhicl
epitht
prote
doma
genor
allerc
CD1.”
lnltla'
mylel

dOma

and f
to be
mam!

leiDOt

LPS

replaces the proline residue with a histidine residue, which renders TLR4, and therefore,
the animals LPS unresponsive (84). However, mutations observed in the equine TLR4
sequence could not be related to an altered LPS response in this species (193). Mutations
in the signaling domain of TLR4 in these LPS-hyporesponsive mice initiated a search for
polymorphisms in the human homolog. Subsequently, polymorphisms in the human
TLR4 gene were associated with diseases (e.g., gram-negative bacterial infections in
patients in intensive care units) (66). Inhaled LPS is bound to LPS-binding protein (LBP),
which is an acute phase protein, produced not only by liver cells but also by airway
epithelium. Lipopolysaccharide, bound to LBP, is transferred to GPI-anchored or soluble
protein CD14. Beside CD14 the LPS recognition complex also consists of the LRR-
domain of TLR4 and the macrophage-derived molecule 2 (MD2). In addition to TLR4,
genomic sequence alterations of CD14 and MD2 have been shown to be associated with
altered responses to LPS in mice and humans. No alterations could be found in equine
CD14 and MD2, however (193). The LPS recognition complex (TLR4/CDl4/MD2)
initiates an intracellular signaling cascade by recruiting adaptor molecules, such as the
myleloid differentiation primary response protein 88 (MyD88), to the intracellular
domain of TLR4 (details of the signaling cascade are discussed below).

Toll-like receptor 2 recognizes a wide variety of microbial products from bacteria
and fungi. Like TLR4, it is expressed in epithelial cells. However, whereas TLR4 seems
to be localized more basolaterally, TLR2 is generally recognized on the apical cell
membrane (65). Although TLR2 is the principal receptor involved in the recognition of
leipoteichoic acid from gram-positive bacteria, TLR2 also detects structural variants of

LPS, e.g., from leptospira spp. Additionally, TLR2 is involved in the recognition of

21

that

surv
\Vuh
knoc
the c
inter

pneu

aid]
the f
Contr
TLR:

expre

mycobacterial lipoarabinomannan. It has been suggested that TLR2 might have a role in
defense against atypical mycobacterial infections and M. tuberculosis. There is evidence
that TLR2 is important in fungal pneumonias, as TLR2 knock-out mice have a decreased
survival when challenged intra-nasally with Cryptococcus neoformans when compared
with wild-type controls and TLR4 knock-outs (198). However, the survival of MyD88
knock-out mice in this study was worse in comparison to the TLR2 knock-outs, outlining
the complexity and redundancy of components involved the innate immunity (198). Most
interestingly, the pathogen involved in the most common cause of gram-positive
pneumonia (Streptococcus. pneumoniae), does not exert its pathology through TLR2
signaling (90). Although, alveolar macrophages need TLR2 to response to S. pneumoniae
in vitro, TLR2 knock-outs and control mice do not respond differently when stimulated
with S. pneumoniae intranasally. Other immune mechanisms might be involved in the
defense against this pathogen. Toll-like receptor 2 recognizes airborne pathogens, such
as products from gram-positive bacteria, mycobacteria, and fungi by heterodirnerization
with TLRI or TLR6. For example, alterations in the expression or availability of one of
the TLR2 co-receptor (e.g., TLRl or TLR6) have been shown to be important in
contributing to disease processes. Indeed, recently it has been reported that a defect of
TLR2/1 signaling may be due to decreased TLRl surface expression, while TLR2
expression was unchanged (186).

Toll-like receptor 3 knock-out mice experiments revealed that TLR3 recognizes
double stranded (ds) ribonucleic acid (RNA), which is released during viral infection (7).
A role for TLR3 in epithelial cells has been suggested in experiments using rhinovirus,

rhinovirus dsRNA, and synthetic dsRNA (polyinosinic-polycytidylic acid; poly[I;C]),

22

Ieadi

[hBE

tnous
Till
COHSt
ﬂagc

inﬂai

ininr

prodt
prote
acntn
thaw
brong
TLR:
incre;

alnk‘a

(ﬁﬁer
Others

Sln1u1

leading to increased expression of chemokines as well as defensins (human B—defensins
[hBD-2 and hBD-3]) (39, 58).

Toll-like receptor 5 recognizes bacterial ﬂagellins (67). Interestingly, human and
mouse TLRS can recognize different ﬂagellins due to an amino acid variation in the
TLRS residue 268 within the extracellular domain (9). In addition, mutational studies in
conserved surface domains proved the importance of the residues D295 and D367 for the
ﬂagellin recognition (9, 174). There is evidence that TLRS is involved in important
inﬂammatory mechanisms against invading pathogens (126, 148).

T oil-like receptor 9 recognizes bacterial deoxyribonucleic acid (DNA), resulting
in increased expression of IL-8 in colonic epithelial cells (5).

Activation of TLRS leads primarily to an increase in cytokine and chemokine
production. However, it has also been shown that expression of other antimicrobial
proteins (e.g., hBD-2) can also be stimulated through TLR signaling. Finally, TLR
activation may result in increased expression of TLR3 themselves. For example
Haemophilus inﬂuenzae stimulates TLR2, resulting in an increased TLR2 expression on
bronchial epithelial cells (81, 167). In addition, stimulation of TLR4 with LPS results in
TLR2 expression (43). Furthermore, cytokines, such as INF-y, can also lead to an
increased expression of selected TLRS (195). Finally, TLR4 expression is up-regulated in
airway epithelial cells due to stimulation with respiratory syncytial virus (123).

How can a speciﬁc response to the variety of microbes develop from ‘just’ 13
different TLRS? Some of the individual TLRS may have their own signaling pathways;
others may Share pathways, but different cell types may only express a subset of TLRS.

Simultaneous activation of different TLRS creates the unique response that is

23

charat
signal
intract
residu
consei
TIR (I
most 1
which
termin
Stimul
associz
ICCCpIL'
associa
dissoci
B‘aCllV.‘
membr,
at the
(Consist
the Ublc
CODIUga
TAK], ,
lhe IKK

C'Onslsts

characteristic for the cell type and the microorganism. In recent years, the intracellular
Signaling cascades of TLR pathways have been studied in detail. The common
intracellular TIR domain of TLRS plays a central role in TLR signaling. TIR is about 200
residues long and contains the TIR consensus motif including 3 consensus boxes (6). The
consensus box 2 reveals a proline residue, which is critical for the interaction between
TIR domains, and therefore, for the interaction between TLRS and adaptor proteins. Like
most of the TLR proteins, TLR4 is involved in the activation of a signaling cascade,
which involves MyD88 (MyD88 dependent pathway). MyD88 itself contains a carboxy-
terminal TIR domain, through which it interacts with the TIR domain of TLRS.
Stimulation of TLR4 with LPS triggers MyD88 to recruit IRAK4 (IL-1 receptor
associated kinase 4) allowing the association and phosphorylation of IRAKI. The TNF
receptor-associated factor-6 (TRAF6) is also recruited to the receptor-complex, by
association with phosphorylated IRAKI. Phosphorylated IRAK] and TRAF6 then
dissociate from the receptor to form a complex with TAKl (transforming-growth-factor-
B-activated kinase 1), TABl (TAKl-binding protein 1) and TAB2 at the plasma
membrane, which induces the phosphorylation of TAB2 and TAK]. IRAKl is degraded
at the plasma membrane, leading to the translocation of the remaining complex
(consisting of TRAF6, TAKl, TABl and TAB2) to the cytosol, where it associates with
the ubiquitin ligases UBC13 (ubiquitin-conjugating enzyme 13) and UEVlA (ubiquitin-
conjugating enzyme E2 variant 1). Ubiquitination of TRAF6 induces the activation of
TAKl, which in turn phosphorylates both mitogen-activated protein (MAP) kinases and

the IKK complex (inhibitor of nuclear factor-KB (IKB)-kinase complex). The latter

consists of IKK-a, IKK-B and IKK-y and phosphorylates IKB, which leads to its

24

ubiqu

casca‘
the e)
primal
such a
indepr
pathw:
factor

(IFNR
transcr
TLR4

phase

CITOI-gir.
additio;
‘13 TIR
inClucin
TRIF-..
M11388
differen
Signalin-
moleCult

a matter

Ubiquitination and subsequent degradation. Therefore, stimulation of the TLR4 signaling
cascade allows nuclear factor kappa B (NF-KB) to translocate to the nucleus and induce
the expression of its target genes (6). This MyD88-dependent pathway contributes
primarily to the activation of NF-KB and the production of pro-inﬂammatory cytokines,
such as TNF-a, IL-1, IL-6, and IL-8 (early-phase NF-K'B response). A second MyD88-
independent/T RIF (T IR-containing adaptor inducing interferon (INF)-ﬂ)-dependent
pathway leads to the activation of the transcription factor IRF-3 (interferon regulatory
factor 3), which controls IFN-B production (72, 197). IFN-[3 binding to IFN receptor
(IFNR) initiates signaling through STATS 1 and 2 (signal transducers and activators of
transcription 1 and 2), activating additional genes. The TRIF-dependent pathway of
TLR4 also collaborates with the MyD88-dependent pathway by activating NF-KB (late-
phase NF-ICB response) and various MAPKS pathways, promoting pro-inﬂammatory
cytokine production (6). During the discovery of the MyD88-independent pathway
additional adaptor molecules have been found: MyD88 adaptor-like (MAL; also known
as TIR domain-containing adaptor protein [TIRAP]; TIR domain-containing adaptor
inducing IFN-B (TRIF; also known as TIR-containing adaptor molecule-1 [TICAM-1]);
TRIF-related adaptor molecule (TRAM) and a- and HEAT-Armadillo motifs. Like
MyD88, these adaptor molecules have the TIR domain, but they Show structural
difference to MyD88 (132). This might be the reason for the different down-stream
Signaling due to different transducer binding. The advantage of using different adaptor
molecules is the response Speciﬁcity in the defense against different microorganisms. As

a matter of fact, it is tempting to speculate that these adaptor molecules are future

25

thera

TLRS

IL-8.‘
“SPCW
Endou
expres
the(3F
cognat
desens

migratt

knouri
lﬂay ar
PalhOgg
reSPOns
lnCI‘EaSt
Vlhich it
exhibits
Due k)

mediator
NF‘KB a

I ll 87)”

therapeutic targets, for example in ﬁghting pro-inﬂammatory cascades initiated through
TLRS, while maintaining TLR-mediated protective responses.

Beside the production of the potent chemotactic factors for neutrophils, such as
IL-8, the TLR4 Signaling pathway also cross-talks with the G-protein-coupled receptor
(GPCR) signaling pathway in neutrophils, which is used by chemokine receptors (44).
Endotoxin-activated signaling through TLR4 transcriptionally down-regulates the
expression of GRK 2 and 5 (GPCR kinases), which are cytosolic proteins that desensitize
the GPCR signaling cascade by binding to activated GPCRS, uncoupling them from their
cognate G-proteins. The reduced expression of GRKs lowers the chemokine receptor
desensitization (i.e., preserves GPCR expression) and markedly augments the neutrophil
migratory response (44).

Negative regulation of T LR signaling: Negative feedback mechanisms are
known to be important for the regulation of TLR initiated signaling, and, therefore also
play an important role in protecting the lung against excessive responses to airborne
pathogens (6, 133). A20, a zinc ﬁnger binding protein, is encoded by an immediate early
response gene (133). In primary cultured human AECS A20 expression is signiﬁcantly
increased by the binding of LPS to TLR4 (60). Expression of A20 is regulated by NF -KB,
which is mediated by two adjacent NF-KB-binding Sites in the A20 promoter (92). A20
exhibits ubiquitin ligase activity and, therefore, targets proteins fon degradation (194).
Due to the poly-ubiquitination of the receptor interacting protein (RIP), an essential
mediator of the proximal TNF receptor 1 (TNFRl) signaling complex, TNF-a-dependent
NF-KB activation is attenuated. Beside RIP, other proteins such as TRAF 6 (69) and IKK-

y (187), which are important signaling molecules along the TLR4 pathway, can also bind

26

to A;
been

cell d

acts it
down-

death (

Summ

Inﬂanl
exacer
SUch a
Toll-Ii}
pr01ejn
ep101c]
reco‘sn
ProduC.

product

to A20, therefore, clearly identifying anti-inﬂammatory characteristics of A20. It has
been demonstrated that A20'l' mice are highly susceptible to chronic inﬂammation and
cell death (100) and that A20 in human airway epithelial cells inhibits NF -t<B-dependent
TNF-a- and IL-lB- expression (82). Also, A20 is able to inhibit TLR4 and TLR2-
dependent production of IL-8 (60). In that study it was shown in detail that dominant-
negative myeloid differentiation protein (MyD88) and a mutant form of IKE-0. attenuates
the LPS activation of NF-KB and IL-8 and that over-expression of A20 inhibits activation
of NF-KB and the IL-8 promoter by LPS in a similar way. These results suggest that A20
acts in different ways as a negative regulator of the TLR4/NF-KB/IL-8 axis. Failure to

down-regulate NF-KB transcriptional activity results in chronic inﬂammation and cell

death (100).

Summary of Literature Review and Hypotheses

Recurrent airway obstruction in horses is characterized by neutrophilic airway
inﬂammation and obstruction, and stabling of susceptible horses triggers acute disease
exacerbations. Stable dust is rich in endotoxin from gram-negative bacteria and fungi,
such as Faenia rectivirgula, Thermoactinomyces vulgaris, and Aspergillus fumigatus.
Toll-like receptors (TLRS) are pathogen recognition receptors that are trans-membrane
proteins on a variety of airway cells, such as alveolar macrophages, dendritic cells,
epithelial cells, and airway smooth muscle cells. Toll-like receptor 4 is the principal
recognition receptor for LPS, and TLR2 is the principal recognition receptor for fungal
products. In human bronchial epithelium, TLR4 and TLR2 stimulation leads to increased

production of interleukin (IL)-8, a potent neutrophil attractant, through activation of the

27

 

Hana“

horse-

NF-x

pnﬂei

concer
expres
itould
and hi

nabhng

 

transcription factor NF-KB. During stabling, RAO-affected horses exhibit an elevated
NF-KB activity and IL-8 production compared to control animals and RAO-affected
horses on pasture. In laboratory animal models and cell culture studies, the zinc ﬁnger

protein A20 negatively regulates the TLR4 or TLR2/NF-KB/IL-8 axis.

Therefore, I hypothesized i) that horses in stables are exposed to higher endotoxin
concentrations than on pasture; ii) that during stabling TLR4, TLR2, and IL-8 mRNA
expression are increased in RAO-affected horses compared to controls, and that this
would be paralleled by elevated neutrophil counts in bronchoalveolar lavage ﬂuid (BAL);
and iii) that A20 mRNA expression is not increased in RAO-affected horses during

stabling compared to controls.

28

Tablt

i Rec
I TLR

irLRz

 

 

Table 1-1: Human and murine Toll-like receptors and their ligands.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Animal
Receptor Species Ligands
TLR1 Human, Mouse Tri-acyl-Iipoproteins
Associates with TLR2
TLR2 Human, Mouse Lipoproteins
Lipoteichoic acid (gram positive)
Lipoarabinomannan (mycobacteria)
Peptidoglycan
zymosan (yeast)
Aspergilus fumigatus conidia
Associates with TLR1 and TLR6
TLR3 Human, Mouse Po|y(|-C)
ds RNA (viral)
Inﬂuenza A
TLR4 Human, Mouse LPS (gram negative)
Host heat shock protein 60 and 70
Host hyaluronic acid
RSV protein F
Taxol
Aspergillus fumigatus conidia and hyphae
TLRS Human, Mouse Flagellin
TLRS Human, Mouse Di-acyI-Iipopeptides
Zymosan (yeast)
Associates with TLR2
Single-stranded viral RNA, imiquimod, resiquimod
TLR7 Human, Mouse 848
Single-stranded viral RNA, imiquimod, resiquimod
TLR8 Human, Mouse 848
TLR9 Human, Mouse Unmethylated CpG DNA
TLR10 Human Uncertain
TLR11 Mouse Uncertain
TLR12 Mouse Uncertain
TLR13 Mouse Uncertain

 

29

 

END

IS HI

 

 

conce
felam
0f Sta
fllrthe
with .
CORCe
the L.
Climar

NalIQr

CHAPTER 2

ENDOTOXIN CONCENTRATIONS WITHIN THE BREATHING ZONE OF HORSES

IS HIGHER IN STABLES THAN ON PASTURE

Abstract

Dust in horse stables is rich in respirable, pathogenic material, such as endotoxin.
The amount of respirable dust is associated with the concentration of endotoxin in stable
dust. Our laboratory has shown previously that the amount of dust in the breathing zone
of stabled horses is higher than dust concentrations in stables that were measured with
area samplers as well as dust concentrations in the breathing zone of horses on pasture.
Hay dust endotoxin causes equine neutrophilic airway inﬂammation. Endotoxin
concentrations in outdoor environments are inﬂuenced by ambient temperature and
relative humidity. We hypothesized that endotoxin concentrations in the breathing zone
of stabled horses are higher than in the breathing zone of horses kept on pasture. We
further hypothesized that the endotoxin exposure for horses kept on pasture is associated
with changes in ambient temperature and relative humidity. We measured endotoxin
concentrations in the breathing zone of six stabled horses and six horses on pasture using
the Limulus Amebocyte Lysate (LAL) Kinetic Chromogenic Assay and reported the
climatological data of ambient temperature and percent relative humidity from the

National Climatic Data Center. Mean endotoxin concentrations in the breathing zone of

30

stab]
C out
rem;
endc
ratltc

ofir.

u—
:1
no
q

]

fol-c]
lnfe

”Al

Cu
(.1)

stabled horses were signiﬁcantly higher (about 8-fold) than of horses on pasture.
Concentrations of endotoxin on pasture were not associated with changes in ambient
temperature or relative humidity. We concluded that stabled horses are exposed to higher
endotoxin concentrations than horses on pastures. Differences in individual behavior
rather than ambient climatic conditions may contribute to the actual endotoxin exposures

of individual horses in stables and on pasture.

Introduction

In the northern hemisphere it iS common to keep horses in stables, where they are
fed hay. The economical advantage of this practice, however, can be diminished by the
impairment of horse’s welfare and exercise performance due to exposure to high
concentrations of airborne pathogenic/inﬂammogenic materials. For example, we and
others have reported that stabled horses without clinical signs or a history of airway
disease experience inﬂux of inﬂammatory cells into the airway lumen (29, 31, 57, 73,
184). Furthermore, feeding hay from round bales, where horses are assumed to have a
higher dust exposure than feeding hay from traditional square bales, exhibit an about 5-
fold increased risk for >20% neutrophils in trachea (152). From these data, it can be
inferred that hay dust exposure increases the risk of equine inﬂammatory airway disease
(IAD). The prevalence of [AD in Michigan pleasure horses has been reported to be

17.3% (152).

31

accr

raci

COm
also
“(8
Stab
h’cal
90m

91C).

Airway inﬂammation has functional consequences as increased mucus
accumulation, which is associated with airway inﬂammation, is a risk factor for poor
racing performance of stabled race horses (74). Dust exposure can have even more
important consequences in horses with lung disease. For example, in horses with
Recurrent Airway Obstruction (RAO) organic substances that are frequently found in
equine stables and hay dust can cause airway inﬂammation. Speciﬁcally, endotoxin and
fungal products (from e. g., Faenia rectivirgula, Aspergillus fumigatus) within stable dust
can trigger neutrophilic airway inﬂammation and obstruction in RAO-susceptible horses
(33, 143, 144). A 12 to 50% prevalence of RAO has been reported for horses in North
America and Europe (19, 98, 118).

We have previously reported that the respirable airborne dust concentration in the
breathing zone of horses in conventional stables (straw bedding and hay feed) are
signiﬁcantly higher than in recommended stables (shavings as beddings and complete
pelleted diet) (196). McGorum et al. showed a strong correlation between dust
concentrations and airborne endotoxin concentrations (117). In this study, the authors
also Showed that the content of endotoxin within respirable dust in conventional stables
was signiﬁcantly higher than in respirable dust collected on pasture. Endotoxin in horse
stables contributes at least partially to the development of airway inﬂammation in both
healthy and RAO-susceptible horses (144). Beside the amount of dust, endotoxin
concentrations are also inﬂuenced by seasonal and climatological factors, such as
temperature and relative humidity (24).

Endotoxin in horse stables arises from different point sources (e.g., hay, manure

etc.). Due to differences in the individual behavior of horses or in stable hygiene, area

32

C 0,11

C011

brez

entl

Puln'
g€ldi

l’lOl'Sc’

Stable
bI’Eat]
‘0 ad j
measu.

 

samplers may not be adequate to reﬂect the actual endotoxin exposure of stabled horses.
Indeed, Woods et al. reported that, by using personal samplers, total and respirable dust
concentrations in the breathing zone of horses in conventional stables are higher than
concentrations measured in stables with area samplers (196).

Therefore, in the present study we determine the endotoxin concentration in the
breathing zone of horses by use of a personal sampler device. We hypothesized that the
endotoxin concentrations in the breathing zone of stabled horses are higher than of horses
on pasture. We also hypothesized that ambient temperature and relative humidity are

associated with endotoxin concentrations in the breathing zone of horses on pasture.

Materials and Methods

Animals and study design: Horses were selected from a herd maintained by the
Pulmonary Laboratory at Michigan State University. Six horses (three mares, three
geldings; 21 d: 6.7 years; two Quarter horses, one Quarter Horse/Arabian mix, one Grade
horse, one Thoroughbred horse and one Standardbred horse) were studied.

Endotoxin exposure was measured when horses were stabled in conventional
stables and subsequently on pasture. For determination of endotoxin concentrations in the
breathing zone of stabled horses, animals were brought into the stable and were allowed
to adjust to the environment for one hour before the personal sampler for endotoxin
measurement was attached (see below). The stable was cleaned and horses were fed

between 7 and 8 AM and fed again in the aftemoon between 4 and 5 PM. For the ﬁrst

33

mea:
mea:
xvas

ﬂee]
perﬁ
bytl
stab:
hors.
santj
Pers
“as

rubb
ZOne
(~ 4
Labc

[Ana
‘Naﬁc
Pohu

10:0t

disIril

measurement horses were brought into the barn between 8 and 10 AM. After the stable
measurement, horses were returned to the pasture. The second measurement (pasture)
was performed with all horses on the same pasture. Horses were allowed to move around
freely with the personal sampler attached to the halter and surcingle. The study was
performed between the 15th and the 26‘h of September, 2005. The protocol was approved
by the All-University Committee for Animal Use and Care of Michigan State University.

Determination of endotoxin concentrations in the breathing zone of horses in
stables and on pasture. For the collection of endotoxin in the breathing zone around the
horse nostril, personal samplers were attached to each horse (Figure 2-1). The personal
sampler consisted of a ﬁlter cassette that was attached to the halter, and an AirCheck 52
Personal Sample Pump (kindly provided by EMSL Analytical, Inc., Westmont, NJ) that
was attached to a surcingle. The ﬁlter cassette and the air pump were connected by a
rubber hose. Endotoxin was sampled as described previously (191). Brieﬂy, the breathing
zone air was sampled at a ﬂow rate of 1.5 liters per minute for an average of 245 minutes
(~ 4 hours). Endotoxin that was collected on the air ﬁlter was quantiﬁed by the 1H
Laboratory at EMSL Analytical, Inc. (Westmont, NJ) using the Limulus Amebocyte
Lysate (LAL) Kinetic Chromogenic Assay (104).

Climatological data: Climatological data for Lansing, M1, were obtained from the
National Climatic Data Center (http://www7.ncdc.noaa.gov/IPS/). We report four dew
point temperature and relative humidity measurements in three-home intervals (7:00AM,
10:00 AM, 1:00PM, and 4:00PM) on the days when endotoxin measurements were taken.

Statistical analysis. Raw data of endotoxin concentrations were not normally

distributed (Kolmogorov-Smirnov P S 0.05). Therefore, data were loglO-transformed.

34

SIUI

and

anal

Cart

:6
r:
i

].

shou
stab].
DESIU
fold.

Slablc‘

Show]
not 31'
18°C}
hUmid

Means of endotoxin concentrations in stables and on pasture were compared using
Student’s t-test. Data are reported as geometric means (anti-log-transformed) with the 5%
and 95% values. Mean values were calculated for the climatological data. Data were
analyzed by use of SAS version 8, SAS/STAT Software (SAS Institute Inc., Cary, North

Carolina, USA). Differences were considered signiﬁcant for P-values S 0.05.

Results

Endotoxin concentrations: Endotoxin concentrations (ECS) for each horse are
Shown in Table 2-1. Mean values of endotoxin concentrations in the breathing zone of
stabled horses (EC = 7.08 * 103 EU/m3) were Signiﬁcantly higher than of horses on
pasture (EC = 0.85 * 103 EU/m3) (Figure 2-2). The mean difference was more than 8-
fold. Endotoxin concentrations for each horse were always less on pasture than in the
stable (Figure 2-2).

Climatological data. Data for temperature and percent relative humidity are
shown in Table 2-2. The average ambient temperature during stabling (10.4 :t 1.3 °C) was
not signiﬁcantly different from the average ambient temperature on pasture (11.4 i
1.8°C) (P = 0.35). In addition, there was no signiﬁcant difference between the relative
humidity on days when horses were stabled (60.75 A: 6.66 %) as compared to days when

horses were kept on pasture (68 :l: 0.24 %) (P = 0.3).

35

Associations between endotoxin concentrations and climatological data.
Endotoxin concentrations did not correlate with the ambient temperature or the relative

humidity when horses were kept on pasture (Figure 2-3A or 2—3B, respectively).

Discussion

In this study we showed that the mean endotoxin concentration measured in the
breathing zone of stabled horses is more than 8-fold higher than that of horses kept on
pasture. Our results are in agreement with a study reported previously by McGorum et
al., in which the authors Showed an approximately 15-fold higher airborne endotoxin
concentration in conventional stables as compared to concentrations on pasture (117).
However, in that study average endotoxin concentration in stables and on pasture were
higher than in our study (1.67 mg/m3 (= 139.2 * 103 EU/m3) and 0.11 mg/m3 (= 9.2 * 103
EU/m3), respectively). While in the present study, we used a pump ﬂow rate of 1.5 l/min
over 4 hours to collect endotoxin concentrations in the horse’s breathing zone, McGorum
et al. collected dust with a rate of 2 l/min over periods of 4 to 10 hours (117).
Alternatively, the differences might be due to differences in experimental design.
Whereas McGorum et a1. obtained measurements by using one pony under three different
maintenance situations we attached personal samplers to six individual horses at the
different collection occasions. The differences in endotoxin concentrations between the
studies may, therefore, be a function of individual, behavioral differences between the

horses. This might also explain the wider range in endotoxin concentrations in our data.

36

enrh
such

due 1
11051111

inend

horses
horse.

endoto

than ar

tovan.

 

Outdoor endotoxin concentrations, when measured with area samplers, are
positively correlated with temperature and negatively with percent of relative humidity
(24). Here we reported that the endotoxin concentration in the breathing zone of horses
on pastures, which was measured with personal samplers, does not correlate with the
ambient temperature or the percent of relative humidity. This suggests that endotoxin
concentrations in the breathing zone of horses are dependent on the immediate
environment in which the horse keeps its nose, rather than on environmental conditions,
such as temperature or humidity, which might contribute to the production of endotoxin
due to bacterial growth. Therefore, differences in the individual behavior (e.g., head
positioning, snifﬁng, frequency of moving) of the animals is likely to result in variations
in endotoxin exposures.

In summary, we reported here that the use of personal samplers in a population of
horses—stabled or on pasture—allows for detection of endotoxin exposure of the individual
horse. On average, stabled horses are exposed to at least 8-fold higher concentrations of
endotoxin than horses maintained on pastures. Differences in individual behavior rather
than ambient climatological conditions, such as temperature or humidity, may contribute

to variations in endotoxin exposures between individual horses.

37

Tab
kept

COIIC

 

 

 

 

 

 

Table 2-1: Endotoxin concentrations in the breathing zone of stabled horses and horses

endotoxin

 

 

 

 

kept on pasture. (*) indicates signiﬁcant different (P S 0.05) from
concentration measured on pasture.
Endotoxin concentration
(10’ x EU/m’)
Stable Pasture

Horse 1 5.94 1.13

Horse 2 6.48 0.45

Horse 3 6.8 0.74

Horse 4 35.9 0.71

Horse 5 2.82 2.58

Horse 6 4.98 0.56
Geometric mean

7.08 (2.82; 36.3)* 0.85 (0.45; 2.57)

(5%, 95%)

 

 

38

 

 

 

Ta

En‘

Ho:

 

Hors

Horsr
Horst
Horst
Horst
Horse

Horse

Horse

 

Table 2-2: Climatological data (ambient temperature and percent relative humidity) at

days of endotoxin measurrnents obtained from the National Climatic Data Center. E =

Environment, in which horses were maintained, S = Stable, P = Pasture, QH = Quarter

Horse, Grd = Grade Horse, TB = Thoroughbred, SB = Standard bred, Arab = Arabian

 

 

 

Age Breed Temperature Relative Humidity
Date Time

(yrS-) (°C) (%)
Horse 1 19 QH 9/15/05 8:40-12:45 9.4 56.5
Horse 2 28 Grd 9/15/05 8:40-12:45 9.4 56.5
Horse 3 21 TB 9/16/05 10: 15-2: 15 12 80.75
Horse 4 28 OH 9/16/05 10:15-2215 12 80.75
Horse 5 20 SB 9/21/05 8230-12235 9.7 45
Horse 6 10 Arab/OH 9/21/05 8:30-12:35 9.7 45
Horse 1 19 OH 9/23/05 9:10-1:15 9.2 67.5
Horse 2 28 Grd 9/23/05 9:00-1:05 9.2 67.5
Horse 3 21 TB 9/17/05 1 1210-3: 15 13 68.75
Horse 4 28 QH 9/17/05 1 1:50-3:55 13 68.75
Horse 5 20 SB 9/24/05 10230-2235 12 67.75
Horse 6 10 Arab/QH 9/24/05 10:35-2:40 12 67.75

 

 

39

Figure 2-1: Personal sampler device for endotoxin measurements in the breathing zone
around the horse nostril. The personal sampler consisted of a ﬁlter cassette that was
attached to the halter, and an AirCheck 52 Personal Sample Pump (kindly provided by
EMSL Analytical, Inc., Westmont, NJ) that was attached to a surcingle. The ﬁlter
cassette and the air pump were connected through a rubber hose. Endotoxin was sampled

at a ﬂow rate of 1.5 liters per minute for an average of 245 minutes (~ four hours)

  

Filter cassette
Rubber hose

   
    
    

40

Figure 2-2: Endotoxin concentrations (103 x EU/m3) in the breathing zone of stabled
horses and horses on pasture. Individual diamonds represent individual measurements.
Horizontal bars represent the geometric means. (*) indicates Signiﬁcantly different (P S

0.05) from endotoxin concentration measured on pasture.

(II
1

(EUImAS)

 

Endotoxrn
concentration (10910)

N

 

Stable Pasture

41

Figure 2-3: Associations between endotoxin concentrations (103 x EU/m3) in the

breathing zone of horses maintained on pasture and ambient temperature (°C) (A) and

relative humidity (%) (B). Individual diamonds represent individual measurements.

Endolnxin

(10910 x EU/mt‘3)

(10910 x WW3)

 

Temperature (°C)

 

67

68 69
Relative humidity (%)

42

CHAPTER 3

PRIMER DESIGN AND QUANTITATIVE REAL TIME POLYMERASE CHAIN

REACTION (qRT-PCR)

The Quantitative Real Time Polymerase Chain Reaction Assay

 

Introduction: In the following, I describe the thought processes that led me to
choose the standard curve method of relative quantitative real-time polymerase-chain
reaction (RT-PCR) using SYBR Green chemistry for gene expression analysis in the
research described in the following chapters of this dissertation.

RT—PCR versus end-point PCR: In comparison to the end-point PCR, where data
are collected only at the end of the PCR cycle, RT-PCR allows measurement of the
accumulation of ampliﬁcation products, also termed amplicons, throughout the PCR
cycle (i.e., in real time (RT)). During RT-PCR, the amount of ampliﬁcation products is
measured at the time of ﬁrst ﬂuorescence detection rather than after a ﬁxed number of
cycles. Therefore, the higher the copy numbers within a sample, the sooner its detection.

S YBR Green I dye versus Taqman chemistry: For the detection of PCR products
two different types of chemistry are commercially available: TaqMan chemistry and
SYBR Green I dye chemistry. For the detection of these chemistries Sequence Detection
System (SDS) instruments have been developed by Applied Biosystems (e.g. SDS 7500,

7700, 7900). Whereas TaqMan chemistry uses ﬂuorogenic probes for the detection of

43

Speciﬁc PCR products, SYBR Green chemistry uses SYBR Green I dye for the detection
of double-stranded DNA. The PCR-products (i.e., amplicons) are detected by binding of
SYBR Green I dye to each ampliﬁed target sequence as they are produced by DNA
polymerase ampliﬁcation during the PCR cycles. The result is an increased ﬂuorescence
signal that is proportionate to the number of amplicons produced. SYBR Green chemistry
can be used to detect the ampliﬁcation of any double-stranded DNA sequence and, in
comparison to TaqMan chemistry, it does not require the labor and cost intensive
generation of speciﬁc probes. However, SYBR Green chemistry is error-prone, especially
for false-positive signals, due to non-speciﬁc double-stranded DNA sequences within
samples. Therefore, when using SYBR Green chemistry additional optimization
experiments become necessary.

In quantitative RT-PCR, concentrations of nucleic acid targets (i.e., RNA,
cDNA/DNA) that are ampliﬁed during each cycle are measured. In our studies, we
measured the amount of transcribed RNA of the genes of interest (i.e., TLR4, TLR2, IL-
8, A20, and 18S rRNA) after its reverse transcription into complementary DNA (cDNA).
We assumed a 100% efﬁcacy of the reverse transcription. The SDS 7900 instrument by
Applied Biosystems was used to detect the increase of ﬂuorescence Signal during each
cycle and results were shown in ampliﬁcation plots (i.e., plot of ﬂuorescence signal
versus cycle number). During the initial cycles, usually between the 5th and the 15th cycle
number, only minor changes in the ﬂuorescence signal could be detected (baseline). In
our studies, we adjusted the baseline cycles as follows: for 18S rRNA: from the 5‘h to the
10th cycle number; for TLR4: from the 3rd to the 15th cycle number; for TLR2: from the

11'h to the 17‘h cycle number; for IL-8: from the 2nd to the 15th cycle number; and for

44

A20: from the 5‘h to the 25‘h cycle number. The detection of a ﬂuorescence signal above
the baseline Speciﬁes accumulated PCR products. A Speciﬁc cycle (threshold cycle or CT)
is usually deﬁned within the linear part of the S-Shaped ampliﬁcation plot to compare
quantities of an amplicon within (e.g., standard curve) or between samples (e.g., test
samples).

Relative versus absolute quantification: There are two types of PCR product
quantiﬁcation. Whereas the absolute quantiﬁcation measures the amounts of unknown
PCR products by interpolating their amount using a standard curve with known
concentrations, relative quantiﬁcation compares the amounts of PCR products in a
sample of interest (e.g., bronchial brushing sample from RAO-affected horses) in relation
to a reference sample (e. g., bronchial brushing sample from control horses).

Standard curve versus comparative C T method: The calculation of relative
quantiﬁcation is either performed using the standard curve method or the comparative CT
method. In the present research, we decided to use the standard curve method for relative
quantiﬁcation, because i) we only detected ﬁve different genes and, therefore, the labor
and cost intensity were low enough to run the target genes (TLR4, TLR2, IL-8, and A20)
and the reference gene (18S rRNA) in separate tubes, and ii) we did not test for
ampliﬁcation efﬁcacies of each target and the reference gene over all samples. For each
target and the reference gene, standard curves (IO-fold serial dilution) were designed
using the bronchial brushing samples with the lowest Ct value (highest amount of
cDNA). Each standard curve sample was measured in triplicate. The following standard

curves were obtained during our studies:

TLR4: y = 26.1 - 3.2x (R2 =0.99),

 

TLR2: y = 29.3 - 3.6x (R2 =0.99),
IL-8: y = 28.7 - 3.7x (R2 = 0.99),
A20: y = 38.1 — 3.8x (R2=O.99) and

185 rRNA: y =17 -3.lx (R2: 0.99).

Primer design: Primer pairs were designed using Primer Express software from
Applied Biosystems. This software automatically uses a set of parameters to select primer
and probe sets. Therefore, it can be used for the design of primer pairs used in both
TaqMan and SYBR Green chemistry assays. Although no probe is needed in the SYBR
Green chemistry assay it is advantageous to automatically design a probe, in case
veriﬁcation experiments using a TaqMan assay are required to a later time point.
Furthermore, the default setting in Primer Express meets all required criteria for sequence
detection when using SYBR Green chemistry (i.e., amplicon size: 50 to 150 base pairs,
G/C content: 20 to 80%, avoids runs of identical nucleotides, especially four and more GS
in a row, melting temperature: 58 to 60°C). By using Primer Express software we also
controlled for possible primer-dimer formation.

Beside for 18S rRNA, primer pairs used in our studies were designed using
nucleotide sequences available online. The Genbank accession numbers of the sequences
used for primer pair design using Primer Express software are shown in Table 3-1.
Complete mRNA sequences for TLR4, TLR2, and IL-8 were available for Equus
caballus. The primer set for A20 was designed by aligning the known human and murine
cDNA sequences and choosing regions of high sequence identity. Care was taken that

this primer pair would span known intron-exon borders.

46

Veriﬁcation of ampliﬁcation product for primer pairs: 18S rRNA, T LR4, TLR2,
IL-8: Ampliﬁed equine genomic DNA, using the primer pairs of 18S rRNA, TLR4,
TLR2, and IL-8, yielded a single band of approximately the ampliﬁcation product sizes
reported in Table 3-1 in agarose gel electrophoresis (Figure 3-1). Therefore, primer pairs
for these genes did not span intro-exon borders. Further veriﬁcation steps were necessary
to Show that we indeed ampliﬁed the transcription products of the genes of interest.
Therefore, we subsequently performed automated (l8S rRNA, TLR4) or manual (TLR2)
sequencing of the PCR products from genomic DNA. The primer pair for 188 rRNA was
presented in another dissertation previously by the Pulmonary Laboratory (56).
Automatic sequencing resulted in the following nucleotide sequence:

ttagtccctgccctttgtacacaccgcccgtcgctactaccgattggatggtttagtgag.
A BLAST (basic local alignment search tool) analysis at the NCBI web site
(http://www.ncbi.nlm.nih.gov/BLAST/) showed that this sequence aligned with many
different species. We did not ﬁnd a horse-speciﬁc sequence in the online database for this
18S rRNA sequence. Automatic sequencing of the TLR4 ampliﬁed PCR product from
genomic DNA resulted in the following sequence:
tcctgctggactccacggtttaccatccagcaaggctttcctgagtcgtctcc.

This sequence matched the published equine TLR4 sequence in the nr database exactly.
By using BLAT at the UCSC Genome Browser (http://genome.ucsc.edu/cgi-bin/thlat)
this sequence detected the TLR4 sequence in the human genome. BLAT analysis on
DNA ﬁnds sequences of 95% or more Similarity. The ampliﬁed PCR product when using

TLR2 primer set was manually sequenced and the sequence (Figure 3-2) yielded the

47

highest BLAST hit to the equine TLR2. In BLAT this sequencing product detected the
TLR2 sequence in human genome.

A20: No sequence has been published for equine A20; therefore, the primer
pair was designed by using the published A20 human mRNA sequence (Genbank
accession number for human A20: NM_006290) in a region of high identity to the
published murine mRNA sequence (Genbank accession number for mouse A20:
NM_009397). Recently, the unassembled whole equine genome sequence became
_ available with over 31 x 106 traces published at the Trace Archive at NCBI (about 8-fold
coverage). A cross-species MegaBLAST of the human A20 against the whole equine
genome Showed high sequence identity in the region used for the A20 primer pair. One
nucleotide mismatch for the forward primer and one for the reverse primer were found
(Figure 3-3). Single nucleotide mismatches are known to have little impact on
ampliﬁcation of targeted sequences and, therefore, our A20 primer set was speciﬁc for
detecting equine A20 mRNA (79).

During sample analysis we added additional veriﬁcation steps. A dissociation
curve for each amplicon in each sample was generated to verify speciﬁcity of primer
pairs. Finally, each RNA sample was analyzed for genomic contamination by testing RT-
negative samples (use of RNAse free H20 instead of Omniscript Reverse Transcriptase in

RT reaction) for the reference gene.

48

Table 3-1: The oligonucleotide primer pair sequences used for ampliﬁcation of 188
rRNA, TLR4, TLR2, IL-8 and A20 transcripts in bronchial brushing samples. Amplicon

length of the primer pairs and Genbank accession numbers of sequences used during

primer design or sequence source are shown.

 

 

Forward primer Reverse primer Amplicon Genbank
length accession
number
185 GCA ATT ATT CCC CAT GGC CTC ACT AAA CCA 123 See (56)
GAA CG TCC AA
rRNA
TLR4 TCT GGA GAC GAC TCA GCA AGA AGC ACC TCA 91 AY005808
GGA AAG C GGA GTT T
TLR2 AGC TGT AGC AGT TGG GAC ACA TCC GGC TGA 66 AY429602
CTT AGT TCA T AAT CTA TT
1L-8 CAG CAT CTC GTC TGA AGA GCT GCA GAA AGC 73 AY184956
ACA TGA CT AGG AAG A
A20 CCT CAT TGA GCA GTC TCC ACA CTC ACC CAC 69 Human:
CAT GCT CAG TTC
NM_006290
Mouse:
NM_009397

 

49

Figure 3-1: Agarose gel electrophoreses of 3111 Lambda DNA — HindIII Digest (lane 1),
3111 188 rRNA PCR product (lane 2), 31.1.1 TLR4 PCR product (lane 3), 3111 TLR2 PCR
product (lane 4) and 3111 IL-8 PCR product (lane 5). 18S rRNA, TLR4, TLR2 and IL-8

were ampliﬁed using equine genomic DNA.

   
 
 

A
> 2,027 bp fragments
564 bp fragment

125 bp fragment ‘

  
  
  

50

Figure 3-2: Manually sequenced PCR products using the TLR2 forward (A) and reverse

(B) primer.

A)

5 ’. . .ACATCAACACCATAGAGGAAGACTCTTTTTCCTCCCTGAGGAGTCTTGAA
CATTIGGACTGTCAATAATCACTTATCTAATTTATCGTCCTCCTGGTTCAGGCC
CCTTTCTTCTTTAAAATTCTTAAACTTATGGGAAGCACTTACAAAACACTCGG
GGAAACATCTCTTTTTTTCTCATCTCACATTGCGATCTGAAGTGGAATATCAC
TCACTGAGATCAGGAAGATTGCTGGCCATCTGAGACTGGATG. . . 3 ’

B)

5 ’ . . .TAGCATCAATCTCAAGTTCTTCAAGGAAAGTGAGCCCAGCAAAATCCTTT
CCCTGAATCTCAGTGAAGTGAATATTTCCCACTTTCAGGATTCGCAAATTGGT
GAGATGAGAAAAAAGAGATGTTTCCCGAGTGI I I IGTAAGTGCTTCCCAGTA
AG. . .3’

51

Figure 3-3: Cross-species MegaBLAST of human A20 mRNA against horse whole
genome sequences in Trace Database.

A20 forward primer Shown underlined
CCTCATTGAGCAGTCCATGCT

Lineups Show human sequence on top, equine sequence below

>Rnllti] 1 233324679 G836P67577RA9.T0
Length=893
Get TBLASTX alignments

Score = 404 bits (210), Expect = 1e-108
Identities = 222/228 (97%), Gaps = 0/228 (0%)
Strand=Plus/Minus

Query 475 AAAACGCCTGTCTAGGGGCATCTCCCACGCCAGCTCCAGCATTGTTTCCCTGGCCCGGTC 534

|||]|l]llIlllllllllllllllllllllll llllllllllllllll||||||||]|
Sbjct 873 AAAACGCCTGTCTAGGGGCATCTCCCACGCCAGTTCCAGCATTGTTTCCCTGGCCCGGTC 814

Query 535 CCATGTCTCCTCCAATGGTEWvijfsxdﬁgrAGCAATGAGCACCCCCTGGAAATGCCCAT 594

||l||llllllllllllllllllllllllllllll llllllllllllllllllllllll
Sbjct 813 CCATGTCTCCTCCAATGGTGGGGGTGGGGGGAGCAGTGAGCACCCCCTGGAAATGCCCAT 754

Query 595 CTGTGCCTTCCAGCTTCCAGATCTCACTGTATACAATGAAGACTTCCGCAGCTTCATAGA 654

llllllllllll||lll|ll|l|ll|l||| |I|l|l||l|||||ll|||lllll|||||
Sbjct 753 CTGTGCCTTCCAGCTTCCAGATCTCACTGTGTACAATGAAGACTTCCGCAGCTTCATAGA 694

Query 655 GAGAGACCTCATTGAGCAGTCCATGCTGGTTGCCTTGGAACAGGCAGG 702

l||||||lll||I|| llllllllllllllll llllllllllllll
Sbjct 693 GAGAGACCTCATTGAACAGTCCATGCTGGTCGCTTTGGAACAGGCAGG 646

 

A20 reverse primer
TCCACACTCACCCACCAGTTC

GAACTGGTGGGTGAGTGTGGA — reverse complement shown underlined
>gnlltill248973705 G836P611230R110.T0

Length=933
Get TBLASTX alignments

 

 

Score = 167 bits (87), Expect = 2e-37
Identities = 99/105 (94%), Gaps = 0/105 (0%)
Strand=PluS/Plus

Query 700 AGGGCGTTTGAACTGGTGGGTGAGTGTGGATCCCACCTCTCAGAGGCTGCTTCCTTTGGC 759

lllllllllllllllllllll Illlllllllllllll lllllll lllllllllllll
Sbjct 401 AGGGCGTTTGAACTGGTGGGTAAGTGTGGATCCCACCTGTCAGAGGTTGCTTCCTTTGGC 460

 

Query 760 AACTACTGGAGATGGGAACTGCCTCCTGCATGCAGCCTCCCTTGG 804

Ill lllllllllll lllllllllllllllllllllll |]|]|
Sbjct 461 AACAACTGGAGATGGAAACTGCCTCCTGCATGCAGCCTCTCTTGG 505

52

CHAPTER 4

ELEVATED TOLL-LIKE RECEPTOR 4 mRNA LEVELS IN BRONCHIAL
EPITHELIAL CELLS IS ASSOCIATED WITH AIRWAY INFLAMMATION IN

HORSES WITH RECURRENT AIRWAY OBSTRUCTION

Abstract

Recurrent airway obstruction (RAO) is characterized by neutrophilic airway
inﬂammation and obstruction, and stabling of susceptible horses triggers acute disease
exacerbations. Stable dust is rich in endotoxin, which is recognized by toll-like receptor
(TLR) 4. In human bronchial epithelium, TLR4 stimulation leads to elevation of
interleukin (IL)-8 mRNA expression. IL-8 is a potent neutrophil chemoattractant, known
to increase BAL neutrophils within the lung. We hypothesized that TLR4 and IL-8
mRNA, and neutrophil numbers are elevated in RAOS druing stabling compared to
controls and compared to RAOS on pasture. We measured the maximal change in pleural
pressure (APplmax), determined inﬂammatory cell counts in bronchoalveolar lavage ﬂuid
(BAL), and quantiﬁed TLR4 and IL-8 mRNA in bronchial epithelium by qRT-PCR. We
studied Six horse pairs, each pair consisting of one RAO and one control horse. Each pair
was studied when the RAO-affected horse had airway obstruction induced by stabling,
and after 7, 14, and 28 days on pasture. Stabling increased BAL neutophils, APplmax and

TLR4 (4.14-fold change) signiﬁcantly. TLR4 correlated with IL-8 (R2 = 0.75). During

53

stabling, IL-8 was elevated in all horses and positively correlated with BAL neutrophils
(R2 = 0.43), however, only in RAO-affected horses. Elevated TLR4 expression in
bronchial epithelial cells from RAO-affected horses may contribute to elevated IL-8
production, leading to exaggerated neutrophilic airway inﬂammation in response to

inhalation of stable dust.

Introduction

Recurrent airway obstruction (RAO), also known as heaves, is a common
pulmonary disorder in horses. Acute exacerbations are characterized by neutrophilic
airway inﬂammation, mucus accumulation, and bronchospasm, which are triggered by
exposure of susceptible horses to stable and hay dust. RAO is a diffuse airway disease
involving large conducting airways down to the terminal bronchioles. Periods of acute
exacerbation are interspersed by periods of remission, which can be achieved by
returning RAO-affected horses from stables to pasture (151).

Although adaptive immune mechanisms were shown to contribute to the
pathogenesis of RAO (2, 29), innate immune mechanisms are also important (142). Hay
dust is rich in microbial products such as endotoxin (144, 145), and inhalation of
endotoxin-depleted hay dust Signiﬁcantly attenuates airway neutrophilia in affected
horses (144). The inﬂammatory response can be re-established by adding endotoxin back
to the endotoxin-depleted hay dust (141). These observations conﬁrm that attenuation of

airway inﬂammation iS due speciﬁcally to endotoxin. In conventional horse stables,

54

airborne endotoxin concentrations exceed those on pasture (117) and those that can
induce airway inﬂammation in human subjects (171, 172). In our own stables, we
observed that the endotoxin concentrations in the breathing zone of stabled horses are at
least 8-fold higher than concentrations on pasture (chapter 2).

Microbial-derived products, such as endotoxin, have been shown to play an
important role in human lung diseases such as asthma, acute respiratory distress
syndrome, and chronic obstructive pulmonary disease (159). Pathogen-associated
molecules are recognized by pattern recognition receptors commonly referred to as toll-
like receptors (TLRS). Toll-like receptor 4 (TLR4) is crucial for the recognition of
endotoxin, in particular lipopolysaccharide (LPS) (159). TLR4 is expressed in a variety
of cell types within the lung, including pulmonary epithelial cells (4, 68, 123), alveolar
macrophages (46), endothelial cells (10, 192), and airway smooth muscle cells (ASM)
(125), and its own expression can be stimulated by LPS itself (4, 127, 170). The TLR4

stimulation leads to production of cytokines, such as interleukin (IL)-8 (60, 123).

Little is known about TLR signaling in horses. Toll-like receptor 4 mRNA can be
found in the lung tissue of healthy horses (170). The effect of LPS on TLR4 expression in
bronchial epithelial cells of horses is unclear. Whereas, in lung tissue obtained from
unaffected horses, LPS exposure increases TLR4 mRNA expression (170), a recent
report suggests that hay dust exposure does not change the TLR4 mRNA expression in
the bronchial epithelial cells of RAO-affected horses (4). Furthermore, IL-8 mRNA
expression and protein concentration measured in bronchoalveolar lavage ﬂuid (BAL) are
elevated in RAO-affected horses compared to control animals during stabling and to

RAO-affected horses in remission (3, 4, 59).

55

In the present study, we hypothesized that the amount of TLR4 mRNA is elevated
in RAO-affected horses during stabling compared to control horses and to RAO-affected
horses on pasture. We further hypothesized that in RAO-affected horses the increased
mRNA expression of the receptor would be paralleled by an elevated expression of IL-8
mRNA. Furthermore, we hypothesized that the severity of neutrophilic airway
inﬂammation in RAO-affected horses during stabling is correlated with an increase in IL-

8 and TLR4 mRNA expression.

Materials and Methods

Animals. RAO-affected horses were selected from a herd maintained by the
Pulmonary Laboratory at Michigan State University. These animals have a history and
clinical signs compatible with a diagnosis of RAO. To enter the herd, animals fulﬁlled the
following criteria: a) Horses develop airway obstruction and inﬂammation when stabled and
fed hay; b) Airway obstruction and inﬂammation are reversed by pasturing, where horses
have no exposure to stable dust or hay; c) Airway obstruction is reversible with atropine.
Control animals did not develop airway obstruction when stabled and fed hay (151).

Study design. Six RAO-affected horses (2 mares, 4 geldings; 18.3 :I: 3.3 years) and
six control horses (3 mares, 3 geldings; 16 i 4.6 years) were studied in age-matched
horse pairs. Prior to the study, horses were kept on pasture. Acute airway obstruction was
initiated by stabling horses in a barn. Horses were brought into the barn in pairs
consisting of one RAO-affected and one control horse and were bedded on straw and fed

dusty hay. We deﬁned the beginning of the study as the day the RAO-affected horse of

56

each pair developed acute airway obstruction (day 0). When the RAO-affected horses
developed a clinical score of 5 or greater, we measured the total change in pleural
pressure during tidal breathing (APplmax). The clinical criterion to deﬁne acute
exacerbations in RAO-affected horses was a APplmax of 15 cm H20 or greater.
Furthermore, we obtained bronchial brushing samples for gene expression analyses and
bronchoalveolar lavage samples for total and differential cell count analyses. Horses were
then returned to the pasture. Subsequent measurements were obtained on days 7, 14 and
28. The protocol was approved by the All-University Committee for Animal Use and
Care of Michigan State University.

Determination of the total clinical score ( T CS): A scoring system for the
subjective clinical assessment of respiratory effort was used as previously described
(157). Nasal ﬂaring and abdominal movement were each scored separately on a scale of 1
(normal) to 4 (severe signs). To determine the TCS, scores for nasal ﬂaring and
abdominal movement were summed. Therefore, the TCS could range from 2 (normal) to

8 (severe Signs).

Measurement of the maximal change in pleural pressure during tidal breathing
(APplmaJ. The maximal change in pleural pressure was used as an indicator of airway
obstruction. Measurements were made in unsedated horses by means of an oesophageal
balloon connected via a 240-cm-long polypropylene catheter and a pressure transducer
(Validyne Model DP/45-22) to a physiograph (Dash Model 18, Astro-Med, Inc.). The
balloon was passed through the nose, and placed into the middle third of the esophagus.

Twenty breaths were averaged at each measurement period.

57

Collection of bronchial brushing samples (BB). Bronchial brushings were made
between the 3'd-and 6m-generation bronchi via bronchoscopy. Bronchoscopy was
performed with a 3-meter-long endoscope (9-mm-diameter) using a transnasal approach.
The brushing was performed by advancing a cytology brush (CytoSoﬁ Cytology Brush,
MPO) throughout the biopsy channel of the endoscope into the bronchial lumen. The brush
was gently stroked against the airway wall 15 to 20 times. Care was taken to avoid
bleeding. The brush was then withdrawn into the biopsy channel and the endoscope was
removed from the horse’s airways. The cytology brushes were then ﬂushed in 1 ml
phosphate-buffered saline (PBS) and stored on ice until further analysis. The procedure was
repeated twice in the same lung. The side of the lung chosen for brushing was alternated
between the measurement periods. The beginning lung Side was chosen randomly for
each horse.

Quantification of cells. Total and differential cell counts in BBS were performed
manually using a hemacytometer. Cell preparations were made with a cytocentrifuge and
stained with Hematoxylin and Eosin (H&E) stain. Differential cell counts were performed
by counting 200 cells per slide.

Collection of bronchoalveolar lavage ﬂuid (BAL). Bronchoalveolar lavage was
obtained by means of a 3-meter-long endoscope that was passed via the nose and wedged in
a peripheral bronchus. Three 100 ml aliquots of PBS were infused into the tube and
recovered by suction. The lavaged ﬂuids were pooled and the volume was determined.

Quantiﬁcation of inﬂammatory cells. Total and differential cell counts in BAL were

performed using a hemacytometer. Cell smears were made by use of a cytocentriﬁrge and

58

stained with Wright-Giemsa stain. Differential cell counts were performed by counting 200
cells per Slide.

RNA isolation and quantitative reverse transcription-polymerase chain reaction
(qRT-PCR). Total RNA was isolated from BBS using a phenol/guanidine isothiocyanate
mixture (TRI Reagent, Sigma) and 1-Bromo-3-Chloro-propane (Sigma-Aldrich, St.
Louis, MO). Total RNA was treated with DNA-free kit (Ambion). Yields of total RNA
were determined using NanoDrop technology (ND-1000 Spectrophotometer) and the
integrated software version 3.1.2. The quality of total RNA was assessed using the
Agilent 2100 Bioanalyzer and the integrity of the 18S and 28S rRNA was determined
visually and by determination of the 188:288 ratio. Depending on the total RNA yield of
each sample, 40 to 200 ng total RNA were used as template for reverse transcription (RT)
reaction using the OMNISCRIPT reverse transcriptase kit 200 (Qiagen), according to the
manufacturer’s protocols. The RT reaction contained 10X RT buffer, 0.5 mM of each
dNTP, 10 pM random hexamer primers (Applied Biosystems), 10 U/ul of RNAse
inhibitor and 4 U of Omniscript Reverse Transcriptase. The conditions for the RT
reaction were 37°C for 60 min, followed by 93°C for 5 min using an Eppendorf
Mastercycler. Quantitative PCR (qPCR) was performed using the ABI 7900 Sequence
Detection System (Applied Biosystems). The qPCR reaction contained 20ng of cDNA as
template, QuantiTect SYBR Green PCR Kit (Qiagen), and oligonucleotide primer pairs
Speciﬁc to each mRNA of interest (see Table 3-1). Thermal cycling conditions in the ABI
7900 were: 95°C for 15 min, followed by 40 cycles: 95°C for 15 seconds and 60°C for 60
seconds. A dissociation curve for each amplicon in each sample was generated to verify

speciﬁcity of primer pairs. In addition, each RNA sample was analyzed for genomic

59

contamination by testing RT-negative samples (use of RNAse free H20 instead of
Omniscript Reverse Transcriptase in RT reaction) for the reference gene. The
oligonucleotide primer pair sequences used for ampliﬁcation of TLR4, IL-8 and 188
rRNA transcripts, their amplicon length and their Genbank accession number or the
source for the sequence used in the study are listed in Table 3-1. 188 rRNA served as the
reference gene. Primer pairs were designed with Primer Express Software version 2.0
System (Applied Biosystems).

Statistical analysis. The relative amount of mRNA was calculated using the
relative standard curve method. Quantitative means and their standard deviations (SD) for
TLR4 and IL-8 of each BB sample were standardized to the reference gene (188 rRNA).
If standardized quantitative mRNA expression data were not normally distributed
(Kolmogorov-Smimov P S 0.05), they were loglO transformed. Only BBS with more than
80% epithelial cells were included in the statistical analysis. Standardized data were
analyzed by repeated measure ANOVA. Fold changes in mRNA expression were
determined by calculating the ratio between RAO-affected and control horses at the
different time points or between different time points within each horse group. Associations
between single variables were determined using Spearman correlation and regression
analysis. Associations were determined within a treatment group (RAO or control group) by
taking the measurements of all time points (day 0, 7, 14, and 28) into account. Data were

analyzed by use of SAS v. 9.1. Differences were considered Signiﬁcant for P-values S 0.05.

60

Results

Airway function (APplnm). In RAO-affected horses during acute exacerbation
(day 0) APplmax was signiﬁcantly elevated compared to control horses and to RAO-
affected horses during remission (Figure 4-1A). No statistical signiﬁcant differences in
APplm,x were observed between horse groups at day 7, 14, and 28 as well as between the
time points within the control group.

Cellular composition of bronchoalveolar lavage ﬂuid (BAL). Total and
differential cell counts of BAL for the two horse groups at the 4 different time points are
shown in Table 4-1. In RAO-affected horses at day 0, differential neutrophil numbers
were signiﬁcantly greater (Figure 4-lB) and differential lymphocyte counts were
Signiﬁcantly reduced compared to control horses and to RAO-affected horses during
remission. No statistically Signiﬁcant differences in the differential neutrophil and
lymphocyte numbers were observed between horse groups at days 7, 14, and 28 and
between the time points within the control group. Macrophage, eosinophil, and mast cell
counts did not differ between groups or time points, except for the day 14 measurement
period, where the differential macrophage count in BAL of RAO-affected horses was
signiﬁcantly lower than in control horses.

Quantitative mRNA expression analyses in bronchial brushing samples. TLR4
mRNA expression: RAO-affected horses during acute exacerbation showed a
Signiﬁcantly greater TLR4 mRNA expression compared to control horses (P = 0.05) and
to RAO-affected horses at days 7, 14, and 28 (P = 0.01, P = 0,005 and P = 0.002,

respectively) (Figure 4-2A). The fold changes (FCS) in mRNA expression are shown in

61

Table 4-2. The FC in TLR4 mRNA expression between RAO-affected and control horses
at day 0 was FC = 4.14 (P = 0.05). The TLR4 mRNA expression in control horses was
signiﬁcantly lower at day 28 than at day 0 (P=0.03). No statistically signiﬁcant
differences in TLR4 expression were observed between horse groups at day 7, 14, and 28
and between other time points within the RAO or the control horse group. IL-8 mRNA
expression: RAO-affected horses during acute exacerbation had a signiﬁcantly higher IL-
8 mRNA expression compared to remission values at day 14 and 28 (P = 0.04 and P =
0.04, respectively) (Figure 4-2B). The IL-8 mRNA expression in control horses was
signiﬁcantly lower at days 7, I4, and 28 than at day 0 (P = 0.009, P = 0.05 and P = 0.01,
respectively). No statistically signiﬁcant differences in lL-8 expression were observed
between horse groups at any other time points.

Associations among mRNA expression with clinical responses. We observed a
highly Signiﬁcant correlation between TLR4 and IL-8 mRNA expression in bronchial
brushing samples when data were pooled from RAO-affected and control horses (r =
0.81, P < 0.0001) as well as in RAO-affected and control horses separately (r = 0.77, P =
0.0003 and r = 0.84, P < 0.0001, respectively). The least square regression line between
TLR4 and IL-8 mRNA expression shows an rz-value of about 0.75 when data were
pooled from all horses, 0.77 in the RAO group (Figure 4-3A) and 0.78 in the control
group (Figure 4-3B). TLR4 and IL-8 mRNA expression were signiﬁcantly correlated
with the percentage of neutrophils within BAL in RAO-affected horses (r = 0.56, P =
0.02 and r = 0.61, P = 0.006, respectively) and their least square regression line showed
rz-values of about 0.32 and 0.43, respectively (Figure 4-3C and 4-3E, respectively).

However, in control horses the correlations between TLR4 or IL-8 and percent BAL

62

neutrophils were not Signiﬁcant (r = 0.34, P = 0.16 and r = 0.39, P = 0.06, respectively;

Figure 4-3D and 4-3F, respectively).

Discussion

LPS is able to increase expression of TLR4 in inﬂammatory cells (4, 127) and in
lung tissue from healthy horses (170). Our data demonstrate that this may also be true in
bronchial epithelial cells from horses, as in our control group the amount of TLR4
mRNA in BBS decreased gradually from day 0, when endotoxin exposure was greatest, to
day 7, 14, and 28 (Figures 4-2A, Table 4-2). When horses were on pasture, there was no
signiﬁcant difference in the amount of epithelial-derived TLR4 mRNA between control
and RAO-affected horses, but stable dust exposure increased amounts of TLR4 mRNA in
BBS from RAO-affected horses more than in controls (Figure 4-2A, Table 4-2). This
observation could explain the exacerbated neutrophilic airway inﬂammation of RAO-

affected horses to airborne endotoxin (141, 144).

During stabling, there was no signiﬁcant difference in the amount of epithelial-
derived IL-8 mRNA between RAO-affected and control horses (Figures 4-2B, Table 4-
2). In both horse groups, stable dust exposure was associated with Signiﬁcantly higher
amounts of epithelial-derived IL-8 as compared to pasture. We observed a strong and
highly signiﬁcant association between the expressions of TLR4 and IL-8 mRNA in BBS
in RAO-affected and control horses (Figures 4-3A and 4-3B, respectively). This suggests

a role for TLR4 in the stimulation of lL-8 production in equine bronchial epithelial cells.

63

Similarly, asthmatic patients also Show elevated production of IL-8 in bronchial epithelial
cells (113, 179) and in cell cultures of human airway epithelial cells stimulation of TLR4
with LPS leads to increased IL-8 production (60, 123). However, in contrast to IL-8
expression, we found TLR4 expression to be higher in RAO-affected horses compared to
control horses at day 0 (Figure 4-2A, Table 4-2). This suggests that the amount of
epithelial-derived TLR4 is most likely not the only factor responsible for IL-8 production
in horses’ airways. Other factors that may stimulate IL-8 expression include bradykinin
(154), TNF-01 (137), IL-10 (83), or inhaled air pollution particles (23).

Interleukin-8 is a potent neutrophil chemoattractant in human inﬂammatory
airway diseases (130), and this has also been demonstrated for RAO-affected horses (51).
Indeed, in RAO-affected horses there was a Signiﬁcant correlation between the amounts
of epithelial cell derived IL-8 mRNA and neutrophils in BAL (Figure 4-3E). However,
while there was no signiﬁcant difference in epithelial-derived IL-8 mRNA between
RAO-affected and controls at day 0, there were signiﬁcantly higher numbers of
neutrophils in RAO-affected compared to control horses at this time point (Figure 4-1B).
Therefore, epithelial-derived lL-8 mRNA is unlikely to be the only source for stimulation
of neutrophil migration into the airway lumen in horses. Other cells, such as BAL cells
(4), airway smooth muscle cells (136), endothelial cells (85), and monocytes (45), also
release IL-8 upon stimulation. Also, other neutrophil chemotactic agents such as
leukotriene B4, TNF-0r, GM-CSF, complement activation, and reduced apoptosis (62,
103, 160), may contribute to accumulation of neutrophils in the airways.

In the present study we measured the amount of TLR4 and IL-8 mRNA in

bronchial epithelial cells from RAO-affected and control horses following natural

64

challenge induced by stabling. It has been Shown previously that stable dust is rich in
endotoxin (38, 39) and that endotoxin is involved in the pathogenesis of RAO (36, 38).
Because hay dust contains other pro-inﬂammatory agents (3 8, 39) that may interact with
endotoxin to cause airway inﬂammation, we chose this natural challenge model. The
disadvantage of this approach is, however, that these other pro-inﬂammatory agents may

have contributed to airway inﬂammation in a TLR4 signaling independent manner.

In conclusion, we showed that exposure to stable dust leads to increased TLR4
mRNA expression in bronchial epithelial cells from RAO-affected horses and that the
amount of epithelial TLR4 mRNA correlates with IL-8 mRNA expression as well as
neutrophil accumulation in the airways during acute exacerbations in RAO-affected
horses induced by stabling. These data suggest that an increased TLR4 Signaling is an

important contributor to the pathogenesis of RAO.

65

Table 4-1: Cellular composition of bronchoalveolar lavage ﬂuid (BAL). Only horses that
contained more than 80% epithelial cells within their BB samples were included. Mean 3:
S.E.M. of total (log10 transformed) and differential cell counts in BAL from RAO-
affected and control horses during stabling (day 0) and after stabling (days 7, I4 and 28).
(a) indicates signiﬁcant difference (P S 0.05) between the horse groups (control, RAO)
within a time point (day 0, 7, 14, and 28). (b) indicates Signiﬁcant difference (P S 0.05)
from day 0.

IogTCC = loglO transformed total BAL cell count; logNeu = log10 transformed total
BAL neutrophil count; %Neu = percentage of neutrophils in BAL; logMac = loglO
transformed total BAL macrophage count; %Mac = percentage of macrophages in BAL;
logLym = log10 transformed total BAL lymphocyte count; %Lym = percentage of

lymphocytes in BAL

66

 

 

 

23.3.83 2.2 «3% 13:2; 138.; .22.? :8 3.39.? 14.2 “2.3. mm: 42% 53.x.
:3 h. a no a was a; 4 MS N3 3. m3 2 as a N3 36 a 3.8 m2 4 NE £5 a one 538.
mam h 2.: 33 a 3.: . a: a $5 a: a new 8.4 a 5.2 :e h 2.: e3 4 “.2 Sm a 2 82.x.
as a an 33 h ea NS 3. 88 So a 3.8 as h :6 as h Sm as 4 :3. one a ”3 2.2a...
2 w: h we... 0:. a 3N .. NE. a 8.2 ea 4 3 s :3 h E «S a cam . :12 a 3 a: h 8.2 82.x.
.2 m2 4 $4 Moo .2. NE. one a in 2.3 42 as 4 3.4 85 h 8.4 No a ”2 Be a New 328.
as a :e as h :3 No a. 33 42 a Se and h 88 25 a $8 :8 h an as h :8 oven...
92 .238 05. .2250 92 _e=so 05. .2250
an an 3 an a b5

 

67

Table 4-2: Fold changes (FC) for the amount of TLR4 and IL-8 mRNA in bronchial
brushing samples (BBS) between RAO-affected and control horses (RAO/Control) during
stabling (day 0) and after stabling (days 7, 14 and 28) as well as between time points
within the RAO-affected (RAO/RAO) and within the control horse group
(Control/Control). Only horses that contained more than 80% epithelial cells within their

BBS were included. (*) indicates that the PC is signiﬁcant different (P S 0.05) to 1 (no

 

 

change).

time TLR4 1L8

RAO/Control day 0 4.14 * 1.72
day 7 0.98 3.10

day 14 1.49 1.24

day 28 1.85 1.74

RAO/RAO day 0 — day 7 7.10 * 4.40
day 0 - day 14 9.29 * 6.83 *
day 0 — day 28 9.99 * 7.10 *

day7—day 14 1.31 1.55

day7—day28 1.41 1.61

day 14 — day 28 1.07 1.04
Control/Control day 0 — day 7 1.68 7.91 *
day 0 — day 14 3.33 4.92 *
day 0 — day 28 4.46 * 7.18 *

day 7 — day 14 1.99 0.62

day 7 —- day 28 2.66 0.91

day 14 — day 28 1.34 1.46

 

68

Figure 4-1: Clinical parameters in RAO-affected and control horses during stabling and
on pasture. Only horses that contained more than 80% epithelial cells within their BB
samples were included. A) Lung function (APplmax, [cm H20]) and B) percentage of
neutrophils in bronchoalveolar lavage ﬂuid (BAL) (Neu) from control (grey bars) and
RAO-affected horses (black bars) during stabling (day 0) and after stabling (days 7, 14
and 28). (a) indicates signiﬁcant difference (P S 0.05) between the horse groups (control,
RAO) within a time point (day 0, 7, 14, and 28). (b) indicates Signiﬁcant difference (P S

0.05) from day 0.

409 b

APplmax (cm l‘le)

 

80- ,

Neu (%)

 

 

Days

69

Figure 4-2: Relative mRNA quantiﬁcation of A) TLR4 and B) IL-8 in bronchial
brushing cells from control and RAO-affected horses during stabling (day 0) and after
stabling (days 7, 14, and 28). Only horses that contained more than 80% epithelial cells
within their BB samples were included. Vertical bars Show the 5 to 95% range with the
mean of the standardized mRNA expression (horizontal bar) (data in antilog
transformation). (a) indicates Signiﬁcant difference (P S 0.05) between the horse groups
(control, RAO) within a time point (day 0, 7, 14, and 28). (b) indicates Signiﬁcant

difference (P S 0.05) between time points within a horse group.

TLR4/183 rRNA (10"log10)
N

 

 

 

rL-enaa rRNA (10"10910)
b

 

 

 

 

70

Figure 4-3: Least square regressions between mRNA expressions and clinical responses.
Only horses that contained more than 80% epithelial cells within their BB samples were
included. Least square regressions between relative amount of TLR4 and IL-8 mRNA in
bronchial brushing samples (BB) in (A) RAO-affected and (B) control; between relative
amount of TLR4 mRNA in BB and percentage of neutrophils in bronchoalveolar lavage
(BAL) (% Neu) in (C) RAO-affected and (D) control horses; and between relative
amount of lL-8 mRNA in BB and percentage of neutrophils in BAL in (E) RAO-affected
and (F) control horses including measurements at all time points. For least square
regressions with P S 0.05, Rz-values are given in the ﬁgures. Curves represent the least

square regression line between variables.

71

 

 

 

 

 

 

E R2 = 0.7681 I
U)
2 0
=25 /
CE I . r j
(013 -2.5 /0.5 1.5
a ‘-
:’ _2 i
TLR4/18S rRNA (10910)
2 100 —
R = 0.3239
3 o
g 50 -
.\° /
I ' ‘:.‘. .. r H
-2.5 -0.5 1.5
TLR4/188 rRNA (10910)
R2 = 0.4318 80,-
3
a)
z 40 —/
°\° o
O
-2.5 0.5 1.5
lL-8/188 rRNA (10910)

72

CD

lL-8/1BS rRNA (Iogt 0)

U

% Neu

% Neu

R2 = 0.777

 

 

 

TLR4/18S rRNA (10910)

100 '

 

 

-0.5 1.5

TLR4/18$ rRNA (10910)

80-

 

 

-0.5 1.5

lL-8/l8S rRNA (10910)

CHAPTER 5

EXPRESSION OF TOLL-LIKE RECEPTOR 2 mRNA IN BRONCHIAL EPITHELIAL

CELLS IS NOT INDUCED BY EXPOSURE OF HORSES TO STABLE DUST

Abstract

Neutrophilic airway inﬂammation and obstruction are characteristic of recurrent
airway obstruction (RAO). Acute disease exacerbations can be triggered by housing in
stables. Fungi, such as Faenia rectivirgula, commonly found in stable dust, have been
reported to initiate RAO in susceptible horses. Fungal products activate innate immune
mechanisms by stimulating toll-like receptor (TLR) 2. In normal equine lung tissue TLR2
mRNA expression is increased by LPS exposure. In human airway epithelium, TLR2
stimulation leads to elevation of interleukin (IL)-8, a potent neutrophil attractant. We
hypothesized that the amount of TLR2 mRNA in bronchial epithelium is increased when
horses are stabled and decreases when they return to pasture. We further hypothesized
that TLR2 and lL-8 mRNA expression are correlated with each other and with the
numbers of neutrophils in bronchoalveolar lavage ﬂuid (BAL). We measured airway
obstruction (maximal change in pleural pressure; APplmax), determined cell counts in
BAL, and quantiﬁed TLR2 and 1L-8 mRNA in bronchial epithelial cells by qRT—PCR.
We studied six horse pairs, each consisting of one RAO and one control horse. Each pair

was studied when the RAO-affected horses had airway obstruction induced by stabling,

73

and after 7, 14, and 28 days on pasture. APplmax and neutrophils were increased in RAO-
affected horses during stabling. While stabling increased IL-8, TLR2 mRNA expression
was unaffected. TLR2 was not associated with numbers of neutrophils in BAL. While
TLR2 mRNA is detectable in bronchial epithelial cells from all horses, stable dust does
not lead to an increase in its mRNA expression. The localization or the amount of
receptor complex rather than the amount of TLR2 mRNA may be more important in

fungal—induced airway inﬂammation.

Introduction

Recurrent airway obstruction (RAO) is a common asthma-like disorder in horses.
Susceptible animals develop recurrent neutrophilic airway inﬂammation and airway
obstruction due to exposure to stable dust, and clinical remission can be re-established by
returning RAO-affected horses from the disease-initiating stable environment to the
pasture. This naturally occurring pulmonary disorder in horses closely resembles

occupational asthma in humans (151).

74

Horse stables contain airborne fungi such as Faenia rectivirgula and Aspergillus
fumigatus (128). Inhaled fungal products induce neutrophilic airway inﬂammation and
changes of lung function in susceptible horses (33, 116, 143, 162). Originally, it had been
suggested that RAO is an allergic condition as response to inhalation of, for example,
fungal pathogens (151). Elevated IgE levels in bronchoalveolar lavage (BAL) of affected
animals support a type-I hypersensitivity reaction (63, 162). Increased expression of T-
helper cell (Th) type II (Th-2) speciﬁc cytokines, such as interleukin (IL—) 4 and IL-5
mRNA (99), further supports the suggestion that adaptive immune mechanisms are
involved in the pathogenesis of RAO. However, allergic skin reactions to common barn
allergens do not differ between affected and control horses (41, 115). Furthermore,
cytokine expression proﬁles in RAO-affected horses vary among studies (3, 29, 59).
Therefore, adaptive immune mechanisms may only partially explain the disease
pathogenesis.

Innate immune mechanisms are also involved in the disease processes of RAO.
There is evidence that endotoxin within stable dust can trigger neutrophilic airway
inﬂammation in RAO-susceptible horses (144). Inhalation of endotoxin-depleted hay dust
signiﬁcantly attenuates airway neutrophilia and the inﬂammatory response can be re-
established by adding endotoxin back to the endotoxin-depleted hay dust (144). Our
laboratory has recently reported that Toll-like receptor (TLR) 4 mRNA expression in
equine airway epithelium is increased during stabling (13). Toll-like receptors are
transmembrane receptors that recognize pathogen-derived molecules, such as LPS or
lipopeptides, and once stimulated they initiate signaling cascades that culminate in

production of pro-inﬂammatory cytokines (12). TLR4 but not TLR2 mRNA is expressed

75

in normal horse lung tissue, and LPS exposure induces TLR4 and TLR2 mRNA
expression (170). Whereas TLR4 is important for the recognition of LPS, TLR2 is crucial
for recognition of A. fumigatus (108). In RAO-affected horses, inhalation of particulates
isolated from hay dust that mainly contains fungal Spores caused an increased inﬂux of
neutrophils into the airway lumen (144). Stimulation of TLR2 initiates a signaling
cascade that eventually leads to activation of nuclear factor KB (NF-KB) activity and to
the production of IL-8 (175), which is a potent neutrophil attractant (51, 130).

In the present study, we tested if the amount of TLR2 mRNA in bronchial
epithelial cells of RAO-affected horses is greater than in control animals and we
hypothesized that in both horse groups the amount of TLR2 mRNA expression is higher
during stabling than on pasture. We also hypothesized that the amounts of TLR2 and lL-8
mRNA expression and the severity of airway inﬂammation are correlated among each

other.

Materials and Methods

Animals. We studied RAO-affected and control horses in the research herd of the
Pulmonary Laboratory at Michigan State University. RAO-affected animals were chosen
because of a history of airway obstruction and inﬂammation when stabled and fed hay;
because pasturing reversed their airway obstruction and inﬂammation; and because their
airway obstruction was reversible with atropine. Control animals did not develop airway

obstruction when stabled and fed hay (151).

76

Study design. We studied RAO-affected (2 mares, 4 geldings; 18.3 :1: 3.3 years)
and control horses (3 mares, 3 geldings; 16 d: 4.6 years) in age-matched pairs, with one
RAO and one control horse per pair. Prior to the study, horses were kept on pasture.
Measurements were made at four different time points. Horses were stabled until clinical
Signs of acute exacerbation were observed in the RAO-affected horse (day 0). When the
RAO-affected horses developed a total clinical score of 5 or greater (153), we measured
the change in pleural pressure during tidal breathing (APplmax). A APplmax of 15 cm H20
or greater was used to deﬁne acute exacerbation in RAO-affected horses. In addition, we
obtained bronchial brushing samples for RNA extraction and gene expression analyses
and bronchoalveolar lavage samples for total and differential cell count analyses. Horses
were then returned to the pasture. Subsequent measurements were obtained on days 7, 14,
and 28. The protocol was approved by the All-University Committee for Animal Use and
Care of Michigan State University.

Determination of the total clinical score ( T CS): A scoring system for the
subjective clinical assessment of respiratory effort was used as previously described
(157). Nasal ﬂaring and abdominal movement were each scored separately on a scale of 1
(normal) to 4 (severe signs). To determine the TCS, scores for nasal ﬂaring and
abdominal movement were summed. Therefore, the TCS could range from 2 (normal) to

8 (severe signs).

Measurement of the maximal change in pleural pressure during tidal breathing
(APplmJ. The maximal change in pleural pressure was used as an indicator of airway
obstruction. Measurements were made in unsedated horses by means of an oesophageal

balloon connected via a 240-cm-long polypropylene catheter and a pressure transducer

77

(Validyne Model DP/45-22) to a physiograph (Dash Model 18, Astro-Med, Inc.). The
balloon was passed through the nose, and placed into the middle third of the esophagus.
Twenty breaths were averaged at each measurement period.

Collection of bronchial brushing samples (BB). Bronchial brushings were made
between the 3'd- and 6m-generation bronchi via bronchoscopy. Bronchoscopy was
performed with a 3-meter long endoscope (9-mm-diameter) using a transnasal approach.
The bronchial brushing was performed by advancing a cytology brush (CytoSoft
Cytology Brush, Medical Packaging Corporation) throughout the biopsy channel of the
endoscope into the bronchial lumen. The brush was gently stroked against the airway wall
15 to 20 times. Care was taken to avoid bleeding. The brush was then withdrawn into the
biopsy channel and the endoscope was removed from the horse’s airways. The cytology
brushes were then ﬂushed in 1 ml phosphate-buffered saline (PBS) and stored on ice until
ﬁrrther analysis. The procedure was repeated twice in the same lung. The Side of the lung
chosen for brushing was alternated between the measurement periods. The beginning
lung Side was chosen randomly for each horse.

Quantification of cells. Total and differential cell counts in BBS were performed
manually using a hemacytometer. Cell preparations were made with a cytocentrifuge and
stained with H&E stain. Differential cell counts were performed by counting 200 cells per
Slide.

Collection of bronchoalveolar lavage ﬂuid (BAL). Bronchoalveolar lavage was
obtained by means of a 3-meter long endoscope that was passed via the nose and wedged in
a peripheral bronchus. Three 100 m1 aliquots of PBS were infused into the tube and

recovered by suction. The lavaged ﬂuids were pooled and the volume was determined.

78

Quantiﬁcation of inﬂammatory cells. Total and differential cell counts in BAL were
performed using a hemacytometer. Cell smears were made by use of a cytocentrifuge and
stained with Wright-Giemsa stain. Differential cell counts were performed by counting 200
cells per slide.

RNA isolation and quantitative reverse transcription-polymerase chain reaction
(qRT-PCR). Total RNA was isolated from BBS using a phenol/guanidine isothiocyanate
mixture (TRI Reagent, Sigma) and 1-Bromo-3-Chloro-propane (Sigma-Aldrich, St.
Louis, MO). Total RNA was treated with DNA-free kit (Ambion). Yields of total RNA
were determined using NanoDrop technology (ND-1000 Spectrophotometer) and the
integrated software version 3.1.2. The quality of total RNA was assessed using the
Agilent 2100 Bioanalyzer and the integrity of the 188 and 28S rRNA was determined
visually and by determination of the 18S:28S ratio. Depending on the total RNA yield of
each sample, 40 to 200 ng total RNA were used as template for reverse transcription (RT)
reaction using OMNISCRIPT reverse transcriptase kit 200 (Qiagen), according to the
manufacturer’s protocols. The RT reaction contained 10X RT buffer, 0.5 mM of each
dNTP, 10 11M random hexamer primers (Applied Biosystems), 10 U/ul of RNAse
inhibitor and 4 U of Omniscript Reverse Transcriptase. The conditions for the RT
reaction were 37°C for 60 min, followed 93°C for 5 min using an Eppendorf
Mastercycler. Quantitative PCR (qPCR) was performed using the ABI 7900 Sequence
Detection System (Applied Biosystems). The qPCR reaction contained 20 ng of cDNA as
template, QuantiTect SYBR Green PCR Kit (Qiagen), and oligonucleotide primer pairs
speciﬁc to each mRNA of interest (see Table 3-1). Thermal cycling conditions in the ABI

7900 were 95°C for 15 min, followed by 40 cycles 95°C for 15 seconds; and 60°C for 60

79

seconds. A dissociation curve for each amplicon in each sample was generated to verify
speciﬁcity of primer pairs. In addition, each RNA sample was analyzed for genomic
contamination by testing RT-negative samples (use of RNAse free H2O instead of
Omniscript Reverse Transcriptase in RT reaction) for the reference gene. The
oligonucleotide primer pair sequences used for ampliﬁcation of TLR2, IL-8, and 18S
rRNA transcripts, their amplicon length and their Genbank accession numbers or the
source for the sequence used in the study are listed in Table 3-1. 18S rRNA served as the
reference gene. Primer pairs were designed with Primer Express Software version 2.0
System (Applied Biosystems).

Statistical analysis. The relative amount of mRNA was calculated using the
relative standard curve method. Quantitative means and their standard deviations (S.D.) for
TLR2 and IL-8 of each BB were standardized to the reference gene (18S rRNA). If
standardized quantitative mRNA expression data were not normally distributed
(Kolmogorov-Smimov P S 0.05), they were 10g10 transformed. Log-transforrned data are
reported as geometric means with the 5% and 95% range. Non-log-transforrned data are
reported as means :t S.E.M. Standardized data were analyzed by repeated measure
ANOVA. Associations between single variables were determined using Spearman
correlation analysis. Data were analyzed by use of SAS version 8, SAS/STAT Software
(SAS Institute Inc., Cary, North Carolina, USA). Differences were considered Signiﬁcant for

P-values S 0.05.

80

Results

Airway function (APplmx). Data are shown in Table 5-1. At day 0 APplmax was
signiﬁcantly increased in RAO-affected horses compared to control horses and to RAO-
affected horses returned to pasture (Figure 5-1). At days 7, 14, and 28 no statistically
signiﬁcant differences in APplm,x were observed between horse groups.

Cellular composition of bronchoalveolar lavage ﬂuid (BAL). Table 5-1 shows
total BAL cell counts for the two horse groups at the four different measurement periods.
Total number of neutrophils showed a signiﬁcant time effect (Figure 5-2). When data
from RAO-affected and control horses were pooled, the total numbers of BAL
neutrophils were signiﬁcantly increased at day 0 as compared to the BAL neutrophil
numbers during days 7, 14, and 28.

Quantitative mRNA expression analyses in bronchial brushing samples. We did
not observe Signiﬁcant group, time, or group*time effects for the TLR2 expression in
BBS (Table 5-1). There were signiﬁcant group and time effects for IL-8 expression
(Figure 5-3A and Figure 5-3B, respectively); however, the interaction between group and
time was not Signiﬁcant (Table 5-1). When pooling data from all time points, IL-8
mRNA expression was signiﬁcantly higher in RAO-affected than control horses (Figure
5-3A). When data from RAO-affected and control horses were pooled, IL-8 mRNA
expression was Signiﬁcantly higher during stabling (day 0) than on pasture (day 7, 14,
and 28) (Figure 5-3B).

Correlations between mRNA expression and BAL inﬂammatory cells. A

correlation matrix between TLR2 and lL-8 mRNA expression as well as numbers of BAL

81

neutrophils is shown in Table 5-2. Neutrophil numbers correlated with IL-8 and TLR2
mRNA expression in RAO-affected horses (Figure 5-4A and 5-4B, respectively) or when
data from all horses were pooled. Whereas the correlation between neutrophils and IL-8
(RAO: r = 0.42; pooled: r = 0.29) appeared to be positive, we observed negative
correlations between neutrophils and TLR2 (RAO: r = -0.5; pooled: r = -0.36). The
correlation of pooled data between neutrophils and IL-8 tended to be signiﬁcant at P =

0.06).

Discussion

This study demonstrates for the ﬁrst time that TLR2 mRNA is detectable in
equine bronchial epithelial cells, and that this expression is unchanged by stabling.
Furthermore, we did not observe a difference in TLR2 mRNA expression in bronchial
epithelial cells between RAO-affected and control animals (Table 5-1). It is known that
LPS, through stimulation of TLR4, leads to an increase in TLR2 expression (43).
Because stable dust is rich in LPS (144, 145), and because we have previously Shown an
increase in TLR4 expression during stabling (13), we expected an increase in TLR2
expression on day 0 and a decreasing amount during pasturing. However, a lack of a
change in TLR2 mRNA expression does not mean that the TLR2 signaling pathway is
unimportant in the pathogenesis of RAO. Altered mRNA stability (156), processes
involved in the dynamics of RNA transport and translational regulation (86) or changes

in cellular receptor protein localizations (175) could lead to unchanged amounts of

82

mRNA but increased receptor protein availability for signal transduction. Alternatively,
alterations in the expression or availability of one of the TLR2 co-receptor (e.g., TLRl or
TLR6) may be important in contributing to the phenotype of RAO. Indeed, recently it has
been reported that a defect of TLR2/1 signaling may be due to decreased TLRl surface
expression, while TLR2 expression was unchanged (186). Interestingly, the amount of
intracytoplasmatic TLRl was not different as compared to control subjects; therefore,
suggesting that posttranslational mechanisms control the formation of the membrane
receptor complex availability. The authors of this study also suggested that the threshold
of TLR2/1 involvement in intracellular signaling is elevated compared to other TLRS due
to low baseline TLRl surface expression. Therefore, in future studies it would be of
interest to determine TLR2 co-receptor mRNA and protein expression as well as to
determine the cellular localization of TLR2 in relation to its co-receptors in bronchial
epithelial cells in RAO-affected horses compared to control animals.

Our current ﬁndings are not in agreement with Singh Suri et al., who reported that
unstimulated equine lung tissue does not express TLR2 mRNA, and that LPS exposure
causes a measurable up-regulation in TLR2 mRNA expression (170). Whereas in the
present study we quantiﬁed TLR2 mRNA expression in bronchial epithelial cells after
natural challenge, Singh Suri et al. used qualitative mRNA expression analysis in whole
lung tissue after intravenous LPS exposure (170). Therefore, differences in study design

and methodology are likely explanations for these different outcomes.

In human airway epithelial cells, stimulation of the TLR2 signaling cascade leads
to an elevation of IL-8 production (175). Clinical evidence suggests that inhalation of

fungal products, such as from F. rectivirgula, is an important contributor to neutrophilic

83

airway inﬂammation in RAO-affected horses during stabling (33). Because IL-8 is a
potent neutrophilic attractant in RAO-affected horses (51), we expected a positive
association between TLR2 and IL-8 mRNA expression. The lack of correlation between
the expression of TLR2 and IL-8 mRNA (Table 5-2) further supports the contention that
the quantity of TLR2 protein or a change in the cellular localization of the receptor/co-
receptor complex may be more important for TLR2 signal transduction than the amount

of TLR2 mRNA.

We observed Signiﬁcantly more epithelial-derived IL-8 mRNA in RAO-affected
versus control horses (Figures 5-3A) and a higher IL-8 mRNA expression during stabling
than on pasture (Figure 5-3B). However, we observed a Signiﬁcant time effect (Figure 5-
2) but no group effect (Table 5-1) for numbers of neutrophils in BAL. This might explain
the correlation between IL-8 and neutrophils in RAO-affected horses (Figure 5-4A) and
the lack of correlation between those variables in control horses (Figure 5-4B). Our
laboratory and others have reported that stabled horses without clinical Signs or a history
of airway disease experience inﬂux of inﬂammatory cells into the airway lumen (29, 31,
57, 73, 184). The results we report here suggest that in control horses either epithelial-
derived IL-8 mRNA is not the only source for stimulation of neutrophil migration into the
airway lumen or that other neutrophil chemotactic agents, such as leukotriene B4, TNF-01
and GM-CSF, complement activation or reduced neutrophilic apoptosis (62, 103, 160)
may contribute to accumulation of neutrophils in the airways. Alternatively, regulators of
IL-8 production, such as the zinc ﬁnger protein A20, may lead to a down-regulation of

initially induced IL-8 expression (see below chapter 6).

84

Clinical evidence demonstrates that fungi, such as F. rectivirgula, are important in
airway inﬂammation in RAO-affected horses. In the present study, we reported that while
TLR2 mRNA expression is detectable in bronchial epithelial cells from all horses,
inhalation of stable dust does not lead to an increase in TLR2 mRNA expression. Further
studies are needed to determine whether the quantity of TLR2 protein or a change in the
cellular localization of the receptor/co-receptor complex contribute to regulation of

inﬂammation in RAO.

85

Table 5-1: Clinical parameters, cellular composition of bronchoalveolar lavage ﬂuid
(BAL) and TLR2 and IL-8 mRNA expression in bronchial brushing samples (BBS).
sMean :1: S.E.M. of APplmax (difference between inspiratory and expiratory pleural
pressure during tidal breathing) values, ”geometric means of BAL cell counts and
amounts of TLR2 and IL-8 mRNA in BBS from RAO-affected and control horses during
stabling (day 0) and after stabling (days 7, 14, and 28) as well as P-values for ﬁxed
effects (horse group (RAO, control), time, horse group*time) of APplmax, BAL
inﬂammatory cells (neutrophils, macrophages and lymphocytes) as well as TLR2 and IL-
8 mRNA expression in BBS. (a) indicates signiﬁcant difference (P S 0.05) between the
horse groups (control, RAO) within a time point (day 0, 7, 14, and 28). (b) indicates
Signiﬁcant difference (P S 0.05) from day 0.

TCC = Total BAL cell count (/ul); Neu = Total BAL neutrophil count (/ul); Mac = Total
BAL macrophage count (/ul); Lym = Total BAL lymphocyte count (/ul); TLR2 = Toll-

like receptor 2; IL-8 = interleukin 8.

86

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Day 0 Day 7 Day 14 Dav 28 P—values for ﬁxed effects
Horse
Control RAO Control RAO Control RAO Control RAO Horse group Time group*Time
,. Pplnm 3.79 :r 0.34 37.16 i 7.89“ 4.73 i 0.66 15.84 i 9.2" 3.37 i 0.32 11.74 i 7.22" 2.87 i 0.74 7.75 i 3.97h 0.0003 0.03 0.04
rcci 29.5 (1 1; 245) 21.4 (209-. 200) 40.7 (10.5; 129) 34.7 (3.16. 269) 39.8(105; 166) 21.9 (3.72; 87.1) 25.7 (4.17; 275) 1290153; 219) 0.32 0.72 0.97
Neu; 1.55 (0.28; 11.5) 9.55 (1.62: 102) 0.91 (0.21; 5.13) 0.87 (0.02; 3.98) 2.19 (0.83: 3.98) 1.38 (007-. 7.94) 0.53 (0.04; 2.75) 0.27 (0.03; 6.46) 0.79 0.02 0.37
M110“ 3.8 (0.5; 37.2) 1.74 (0.18; 26.9) 6.46 (1.05; 24.5) 4.27 (0.14; 29.5) 9.55 (1.86; 29.5) 9.55 (1.86; 29.6) 4.27 (0.47; 120) 4.27 (0.47; 120) 0.06 0.77 0.75
Lym” 18.2 (2.19; 191) 646(028; 74.1) 28.8(9.12; 110) 26.9(285; 229) 23.4 (2.14; 132) 17 (3.39: 72.4) 17.8 (3.47; 148) 10(04); 186) 0.28 047 0,9
TLR2” 0.620105; 2) 0.780144; 1.58) 0.59 (0.22; 1.48) 0.85 (0.38. 1.95) 0.54 (007; 2.45) 0.93 (9.37; 1.91) 0.72 (0.16; 1.48) 0.91 (0.49; 1.95) 0.21 0.97 0.97
11.8” 0.58 (0.15; 5.5) 1.82 (0.18; 17.38) 007 (001-037) 0.35 (0.04; 2.19) 0140101039) 0.15 (0.06; 0.36) 0.08 (0.01; 0.5) 0.14 (0.06; 0.72) 0.047 0.0013 0.54

 

87

 

 

Table 5-2: Correlation matrix of Spearman correlations among total numbers of
bronchoalveolar lavage (BAL) neutrophils, and bronchial brushing (BB) TLR2 mRNA
and BB IL-8 mRNA in all horses, RAO-affected or control horses.

Neu = Total BAL neutrophil count (/ul); TLR2 = Toll-like receptor 2; IL-8 = interleukin
8; SCC = Spearman correlation coefﬁcient; P = P-value; N = number of samples; n/a =

not applicable

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Neu TLR2 1L-8
RA 0 + SCS 1.00
Control P n/a
N 44
SCC 1.00
Neu RAO P n/a
N 24
SCC 1.00
Control P n/a
N 20
RAO ,+ SCC -0.36 1.00
Control P 0.02 n/a
N 43 47
SCC -0.50 1.00
TLR2 RAO P 0.01 n/a
N 23 23
SCC -0.27 1.00
Control P 0.25 n/a
N 20 24
SCC 0.29 0.13 1.00
333; P 0.06 0.41 n/a
N 42 46 46
SCC 0.42 -0.27 1.00
1L—8 RAO P 0.05 0.22 n/a
N 22 22 22
SCC -0.01 0.38 1.00
Control P 0.95 0.06 n/a
N 20 24 24

 

88

Figure 5-1: Lung function (APplmax, [cm H20]) in RAO-affected and control horses
during (day 0) and after (day 7, 14, and 28) stabling. Grey bars Show RAO-affected
horses and black bars Show control horses. (a) indicates signiﬁcant difference (P S 0.05)
between the horse groups (control, RAO) within a time point (day 0, 7, I4, and 28). (b)

indicates signiﬁcant difference (P S 0.05) from day 0.

O)
O
J

APpImax [cm H20]
00
O

 

 

0
Day 0 Day 7 Day 14 Day 28

89

Figure 5-2: Total bronchoalveolar lavage (BAL) neutrophils as a function of time. Data
from RAO-affected and control horses are pooled for times during stabling (day 0) and
after stabling (days 7, 14, and 28). Individual diamonds represent individual
measurements. Horizontal bars represent the geometric means of total BAL neutrophils at
each time point. (*) indicates signiﬁcantly (P S 0.05) higher than on days 7, I4, and 28.

Total numbers of BAL neutrophils are loglo transformed.

a." :

4°: *

:36)

31128 + I 4. §
_Ed>5.

8a“

82%.. ’ ’ t
’5: 9
33 Q

 

 

Day 0 Day 7 Day 14 Day 28

90

Figure 5-3: Interleukin (IL)-8 mRNA expression in bronchial epithelial cells between the
diagnostic groups (RAO, control) and between the time points. A) Data from all time
points are pooled for RAO-affected and control horses. Individual diamonds represent
individual measurements. Horizontal bars represent the geometric means. (*) indicates a
signiﬁcant difference (P S 0.05) between the diagnostic groups. B) Data from RAO-
affected and control horses are pooled for times during stabling (day 0) and after stabling
(days 7, l4, and 28). Individual diamonds represent individual measurements. Horizontal
bars represent the geometric means of IL-8 mRNA expression at each time point. (*)
indicates a Signiﬁcant difference (P S 0.05) from days 7, 14, and 28. IL-8 mRNA

expression is Shown in relation to 183 rRNA (reference gene) expression.

91

‘9 ‘3

 

 

.1

A

832...: <2! 85....

RAO

Control

.

i

 

 

- q q H

-
o 8 6 4 2 o
1

836.8: <2! «83....

B

Day 28

Day 7 Day 14

Day 0

92

Figure 5-4: Associations between total bronchoalveolar lavage (BAL) neutrophils and
IL-8 (A) or TLR2 (B) mRNA expression in bronchial brushing samples in RAO-affected
horses. Total numbers of BAL neutrophils are logm transformed. IL-8 and TLR2 mRNA

expressions are Shown in relation to 18S rRNA (reference gene) expressions.

11>

.‘L‘
%-
9: . .9 O. .
EB ° ’ 8 o
co “ .
.1" O

or 0
$2 00 "
_"’ O
S O
o
[.—

 

 

-2 -1.5 -1 -0.5 0 0.5 1 1.5
lL-8/18S rRNA (10910)

00

A
’0
cs

 

Total BAL neutrophils
(loa10 / ul)

 

0.5 0.4 0.3 0.2 0.1 o 0.1 0.2 0.3 0.4
TLR2/18S rRNA (16910)

93

CHAPTER 6

A20 mRNA EXPRESSION IN EQUINE BRONCHIAL EPITHELIAL CELLS IS
NEGATIVELY CORRELATED WITH INFLAMMATORY CELL COUNTS IN

BRONCHIAL LAVAGE FLUID

Abstract

Neutrophilic airway inﬂammation and obstruction are characteristic of recurrent
airway obstruction (RAO). In human airway epithelium, TLR4 stimulation leads to
elevation of interleukin (IL)-8, a potent neutrophil attractant. We reported previously that
in horses IL-8 mRNA expression in bronchial epithelial cells paralleled the expression of
toll-like receptor 4 (TLR4) expression (chapter 4). In other species, the TLR4 Signaling
cascade is negatively regulated by the zinc ﬁnger protein A20. Here we hypothesize that
in RAO-affected horses the epithelial-derived A20 expression is not elevated compared
to control horses. We further hypothesize that A20 is negatively associated with the
numbers of inﬂammatory cells in BAL in control horses, but not in RAO-affected horses.
We measured airway obstruction (maximal change in pleural pressure; APplm.),
determined cell counts in BAL, and quantiﬁed A20 mRNA in bronchial epithelial cells
by qRT-PCR. We studied six horse pairs, each consisting of one RAO and one control
horse. Each pair was studied when the RAO-affected horses had airway obstruction

induced by stabling, and after 7, 14, and 28 days on pasture. APplmalx and neutrophils were

94

increased in RAO-affected horses during stabling. A20 expression did not differ between
horse groups or time points. When data from RAO-affected and control horses were
pooled, A20 was negatively associated with BAL inﬂammatory cell numbers and BAL
inﬂammatory cells correlated amongst each other. Equine epithelial-derived A20 protects
against airway inﬂammation by decreasing the number of BAL inﬂammatory cells. A
lack of an increase in A20 expression as compared to control horses may contribute to the

exaggerated neutrophilic airway inﬂammation in RAO-affected horses.

Introduction

Recurrent airway obstruction (RAO) in horses is characterized by recurrent
neutrophilic airway inﬂammation and airway obstruction due to exposure to hay/stable
dust. Clinical symptoms become apparent during stabling and are reversed by returning
RAO-affected horses to pasture. Regulatory mechanisms involved in recruitment of
airway inﬂammatory cells into the lung of RAO-affected horses have been studied
previously. Sandersen et al. showed that nuclear factor kappa B (NF-KB) activity is
correlated with the percentage of neutrophils found in bronchoalveolar lavage ﬂuid
(BAL) and that NF -1<B activity in RAO-affected horses is elevated compared to control
animals during and at least 21 days after stabling (161). Nuclear factor KB activation is
crucial for production of pro-inﬂammatory cytokines, such as interleukin (IL-) 8, a potent
neutrophil attractant in horses (50, 51). Indeed, IL-8 gene expressions as well as protein

amount have been Shown to be increased in BAL and airway epithelial cells from RAO-

95

affected horses compared to control horses (4, 13, 50). Correlations between IL-8 mRNA
expression and BAL neutrophils have been reported previously (13, 59) and we have
Shown that epithelial-derived IL-8 mRNA is paralleled by the expression of toll-like

receptor 4 (TLR4) (13).

Regulation of the TLR4 signaling in equine epithelial cells is unknown. The
TLR4 Signaling cascade is under the inﬂuence of both positive and negative feedback
regulation. A variety of proteins, such as Tollip (201), suppressor of cytokine-signaling-l
(SOCSl) (87), IL-lR-associated kinase M (IRAK-M) (91), and A20 (60), are involved in
the reduction of the TLR signal transduction. The gene of the zinc ﬁnger protein A20 was
originally characterized as TNF-a-inducible (135). Subsequently, it was shown that it is a
NF-KB responsive gene, and as such A20 is inducible by a wide variety of stimuli (92).
The A20 transcript is rapidly but transiently induced, reaching its highest level within 1
hour following stimulation (135). A20-deﬁcient mice develop multi-organ inﬂammation
due to LPS stimulation (100). Therefore, A20 has been suggested to be an endogenous
regulator of LPS-induced inﬂammation. Elevated A20 expression functions as a negative
feedback loop to block NF-KB-dependent gene expression (14). For example, it has been
shown that A20 over-expression leads to inhibition of TLR4-mediated IL-8 synthesis in
human airway epithelial cells (60). Also, A20-deﬁcient ﬁbroblasts display prolonged NF-
KB activity and are unable to terminate NF-KB activation (100). A20 interferes with the
TLR Signaling pathway at the level of the tumor necrosis factor receptor (TNFR)-
associated factor (TRAF)-6 (16, 70, 105) leading to the inhibition of the phosphorylation

of l-KB kinase (IKK), which is necessary for NF-KB activation.

96

Therefore, in this study we hypothesize that epithelial-derived A20 expression is
not elevated compared to control horses. We further hypothesize that A20 is negatively
associated with the numbers of inﬂammatory cells in BAL from control horses, but not

from RAO-affected horses.

Materials and Methods

Animals and study design. Animals and study design have been described in detail
elsewhere (chapter 5). The protocol was approved by the All-University Committee for
Animal Use and Care of Michigan State University.

Determination of the total clinical score ( T CS): A scoring system for the
subjective clinical assessment of respiratory effort was used as previously described
(157). Nasal ﬂaring and abdominal movement were each scored separately on a scale of 1
(normal) to 4 (severe signs). To determine the TCS, scores for nasal ﬂaring and
abdominal movement were summed. Therefore, the TCS could range from 2 (normal) to

8 (severe Signs).

Measurement of the maximal change in pleural pressure during tidal breathing
(APplme. The maximal change in pleural pressure was used as an indicator of airway
obstruction. Measurements were made in unsedated horses by means of an oesophageal

balloon connected via a 240-cm-long polypropylene catheter and a pressure transducer

(Validyne Model DP/45-22) to a physiograph (Dash Model 18, Astro-Med, Inc.). The

97

balloon was passed through the nose, and placed into the middle third of the esophagus.
Twenty breaths were averaged at each measurement period.

Collection of bronchial brushing samples (BB). Bronchial brushings were made
between the 3'd- and 6m-generation bronchi via bronchoscopy. Bronchoscopy was
performed with a 3-meter long endoscope (9-mm-diameter) using a transnasal approach.
The bronchial brushing was performed by advancing a cytology brush (CytoSoft
Cytology Brush, Medical Packaging Corporation) throughout the biopsy channel of the
endoscope into the bronchial lumen. The brush was gently stroked against the airway wall
15 to 20 times. Care was taken to avoid bleeding. The brush was then withdrawn into the
biopsy channel and the endoscope was removed from the horse’s airways. The cytology
brushes were then ﬂushed in 1 ml phosphate-buffered saline (PBS) and stored on ice until
further analysis. The procedure was repeated twice in the same lung. The Side of the lung
chosen for brushing was alternated between the measurement periods. The beginning
lung Side was chosen randomly for each horse.

Quantiﬁcation of cells. Total and differential cell counts in BBS were performed
manually using a hemacytometer. Cell preparations were made with a cytocentrifuge and
stained with H&E stain. Differential cell counts were performed by counting 200 cells per
Slide.

Collection of bronchoalveolar lavage ﬂuid (BAL). Bronchoalveolar lavage was
obtained by means of a 3-meter long endoscope that was passed via the nose and wedged in
a peripheral bronchus. Three 100 ml aliquots of PBS were infused into the tube and

recovered by suction. The lavaged ﬂuids were pooled and the volume was determined.

98

Quantification of inﬂammatory cells. Total and differential cell counts in BAL were
performed using a hemacytometer. Cell smears were made by use of a cytocentrifuge and
stained with Wright-Giemsa stain. Differential cell counts were performed by counting 200
cells per slide.

RNA isolation and quantitative reverse transcription-polymerase chain reaction
(qRT-PCR). Total RNA was isolated from BBS using a phenol/guanidine isothiocyanate
mixture (TRI Reagent, Sigma) and 1-Bromo-3-Chloro-propane (Sigma-Aldrich, St.
Louis, MO). Total RNA was treated with DNA-free kit (Ambion). Yields of total RNA
were determined using NanoDrop technology (ND-1000 Spectrophotometer) and the
integrated software version 3.1.2. The quality of total RNA was assessed using the
Agilent 2100 Bioanalyzer and the integrity of the 18S and 28S ribosomal RNA (rRNA)
was determined visually and by determination of the 188:288 ratio. Depending on the
total RNA yield of each sample, 40 to 200 ng total RNA were used as template for
reverse transcription (RT) reaction using OMNISCRIPT reverse transcriptase kit 200
(Qiagen), according to manufacturer’s protocols. The RT reaction contained 10X RT
buffer, 0.5 mM of each dNTP, 10 uM random hexamer primers (Applied Biosystems), 10
U/ul of RNAse inhibitor and 4 U of Omniscript Reverse Transcriptase. The conditions
for the RT reaction were 37°C for 60 min, followed 93°C for 5 min using an Eppendorf
Mastercycler. Quantitative PCR (qPCR) was performed using the ABI 7900 Sequence
Detection System (Applied Biosystems). The qPCR reaction contained 20ng of cDNA as
template, QuantiTect SYBR Green PCR Kit (Qiagen), and oligonucleotide primer pairs
Speciﬁc to each mRNA of interest (see Table 1). Thermal cycling conditions in the ABI

7900 were 95°C for 15 min, followed by 40 cycles 95°C for 15 seconds; and 60°C for 60

99

seconds. A dissociation curve for each amplicon in each sample was generated to verify
speciﬁcity of primer pairs. In addition, each RNA sample was analyzed for genomic
contamination by testing RT-negative samples (use of RNAse free H2O instead of
Omniscript Reverse Transcriptase in RT reaction) for the reference gene. The
oligonucleotide primer pair sequences used for ampliﬁcation of A20 and 18S rRNA
transcripts, their amplicon length, and their Genbank accession number or the source for
the sequence used in the study are listed in Table 3-1. 188 ribosomal RNA (rRNA)
served as the reference gene. Primer pairs were designed with Primer Express Software
version 2.0 System (Applied Biosystems). The primer pair for A20 was designed by
choosing regions of high sequence identity between the human and mouse A20 mRNA
sequences.

Statistical analysis. The relative amount of mRNA was calculated using the
relative standard curve method. Quantitative mean and their standard deviation (S.D.) for
A20 of each BB were standardized to the reference gene (188 rRNA). Because the
standardized quantitative mRNA expression data and BAL inﬂammatory cell counts were
not normally distributed (Kolmogorov-Smimov P S 0.05), they were loglo transformed.
Loglo-transformed data are reported aS geometric means with the 5% and 95% values.
Standardized A20 expression data and inﬂammatory cell counts were analyzed by
repeated measure AN OVA. Associations between single variables were determined using
Spearman correlation analysis. Data were analyzed by use of SAS version 8, SAS/STAT
Software (SAS Institute Inc., Cary, North Carolina, USA). Differences were considered

signiﬁcant for P-values S 0.05.

100

Results

Cellular composition of bronchoalveolar lavage ﬂuid (BAL). BAL cell counts
for the two horse groups at the four different measurement periods have been reported
elsewhere (chapter 5, Table 5-1). Brieﬂy, when data from RAO-affected and control
horses were pooled, total numbers of neutrophils were Signiﬁcantly increased at day 0 as
compared to days 7, 14 and 28 (Figure 5-2).

Quantitative mRNA expression analyses in bronchial brushing samples. There
was no statistical signiﬁcant difference in the A20 mRNA expression in BBS between
RAO-affected and control horses or among time points (Figure 6-1).

Correlations among BAL inﬂammatory cells, and associations between A20
mRNA expression and BAL inﬂammatory cells. Since A20 mRNA expression was not
different between RAO-affected and control horses, data were pooled for correlation
analysis. A correlation matrix between A20 mRNA expression and numbers of BAL
inﬂammatory cells is shown in Table 6-1. Neutrophils in bronchoalveolar lavage ﬂuid
correlated with BAL macrophage (r = 0.62) and lymphocyte numbers (r = 0.48) (Figure
6-2A and 6-2B, respectively) and BAL macrophages correlated signiﬁcantly with BAL
lymphocytes (r = 0.83) (Figure 6-2C). Epithelial-derived A20 mRNA expression
correlated negatively with BAL neutrophils (r = -O.37), BAL macrophages (r = -0.46) and

BAL lymphocytes (r = -0.42) (Figure 6-3A, 6-3B and 6-3C, respectively).

101

Discussion

Here we showed for the ﬁrst time that equine epithelial-derived A20 may protect
against airway inﬂammation by decreasing the number of BAL inﬂammatory cells. We
observed signiﬁcant negative correlations between the epithelial-derived A20 mRNA
expression and BAL inﬂammatory cell counts (Figure 6-3). In other Species, A20 protects
against airway inﬂammation (134). For example, while A20"' mice have multiorgan
inﬂammation (100), A20 over-expression leads to an attenuation of NF-KB activity (134),
and, therefore, to decreased IL-8 production. In previous studies, we reported a positive
correlation between BAL neutrophils and bronchial IL-8 or TLR4 mRNA expression as
well as a strong association between IL-8 and TLR4 (chapters 4 and 5). Since stimulation
of TLR4 initiates a signaling cascade that eventually leads to activation of NF-KB activity
and to the production of IL-8 (175), and because A20 interferes with the TLR signaling
pathway at the level of TRAP-6 (16, 70, 105), our results in the present study support the
contention that equine A20 protects against neutrophilic airway inﬂammation by

inhibition of NF-KB activation.

Because TNF-01 in BAL from RAO-affected horses is greater during stabling than
on pasture and compared to control horses (59), and because A20 is a TNF-0t response
gene, one could assume an up-regulation of A20 mRNA expression in BBS of RAO-
affected horses during stabling and compared to controls. Instead we found that A20
expression was unaffected by stabling and did not differ between RAO-affected and
control horses (Figure 6-1). This suggests that either the mechanism responsible for A20

up-regulation is defective, or that A20 mRNA expression is down-regulated in RAO-

102

affected compared to control horses and, therefore, may contribute to the exaggerated
airway inﬂammation in RAO-affected horses during stabling. This is also supported by
the ﬁnding of a prolonged NF-tcB binding activity in RAO-affected versus control horse
after stabling (21, 22, 161). Alternatively, other mediators of inﬂammatory cell migration
into airways of RAO-affected horses, such as macrophage inﬂammatory protein-2 (MIP-
2) (51, 110, 111), platelet activating factor and leukotriene B4 (42, 47-49, 110), [32
integrin (200), and adhesion molecule CD18 (111) that may be differently regulated than
IL-8, may play a more important role in modulation of neutrophilic inﬂammation in

RAO-affected horses.

Beside neutrophils, mechanisms that attract macrophages and lymphocytes may
also be under the regulation of A20. Our observation that neutrophil, macrophage, and
lymphocyte numbers in BAL were correlated with each other (Figure 6-2) leads to the
assumption of common migratory patterns for these cells. For example, certain integrins,
such as lymphocyte-associated function antigen-1 (LFA-l or CD11a/CD18) and
macrophage antigen-1 (Mac-1 or CD1 1b/CD18), that are important for transendothelial
migration of neutrophils (190, 200), are reported to be important in migration of
lymphocytes and monocytes into the lung (101, 102, 190, 200). Furthermore, beside the
production of IL-8, production of the monocyte chemoattractant termed CCL2 (monocyte
chemotactic protein 1 [MCP-1]) can also be upregulated by stimulation of TLR4 with

LPS (125).

Here we reported for the ﬁrst time that, in horses, epithelial-derived A20 may
protect against airway inﬂammation by decreasing the amount of BAL inﬂammatory

cells, such as neutrophils, lymphocytes, and macrophages. A lack of an increase in A20

103

expression as compared to control horses may contribute to the exaggerated neutrophilic

airway inﬂammation in RAO-affected horses.

104

Table 6-1: Correlation matrix of Spearman correlations among total numbers of
bronchoalveolar lavage (BAL) inﬂammatory cells (neutrophils, macrophages, and
lymphocytes) and epithelial-derived A20 mRNA after pooling data from RAO-affected
and control horses.

Neu = Total BAL neutrophil count (/ul); Mac = Total BAL macrophage count (/ul); Lym
= Total BAL lymphocyte count (/ul); A20 = zinc ﬁnger protein A20; SCC = Spearman

correlation coefﬁcient; P = P-value; N = number of samples; n/a = not applicable

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Neu Mac Lym A-20
SCC 1.00
Neu P n/a
N 44
SCC 0.62 1.00
Mac P <0.0001 n/a
N 44 47
SCC 0.48 0.83 1.00
Lym P 0.001 <0.0001 n/a
N 44 47 47
SCC 03? -0.46 -0.42 1.00
A20 P 0.01 0.002 0.006 n/a
N 43 42 42 43

 

105

 

Figure 6-1: Relative mRNA quantiﬁcation of the zinc ﬁnger protein A20 in bronchial
brushing cells from control and RAO-affected horses during stabling (day 0) and after
stabling (days 7, 14 and 28). Vertical bars Show the 5 to 95% range with the mean of the

standardized mRN A expression (horizontal bar) (data in antilog transformation).

._3
O
J

 

 

 

A20/18$ rRNA (10010910)
01

 

 

 

0
o o '\ b- s ‘b ‘b
\ O o‘ N " \q, 3’
4° 6 (6 O .‘o O
00° Q5” 00° <25" cs .23: (Pf Q}

106

Figure 6-2: Association between total numbers of bronchoalveolar lavage (total BAL)
neutrophils and total BAL macrophages (A), total BAL neutrophils and total BAL

lymphocytes (B), and total BAL macrophages and total BAL lymphocytes (C).

A

”é,“ o
6.
0:0 3:. 3: ’

 

 

Total BAL macrophages (10910
I ul)

 

 

 

 

4- 9 0.: O
2 . . .
2 4 s 8
B Total BAL noutrophlles 00910 I 111)
3 ‘Mo 3
g: 6‘ 9 o” a '0. 0’
.. ’ t
3
§ 2 . .
" 2 4 6 8
C Total BAL neutrophils: (10910 I ul)
3 . t’ o ’
2 —_- “0‘ :0
1% a O o
z- 4 -
5
E 2 . . .
" 2 4 6 8

Total BAL macrophages (10910 I 111)

107

Figure 6-3: Associations between epithelial-derived A20 mRNA expression and total
numbers of bronchoalveolar lavage (BAL) neutrophils (A), macrophages (B) and
lymphocytes (C). A20 mRNA expressions is shown in relation to 18S rRNA (reference

gene) expressions. Curves represent the linear associations between variables.

A

 

oallllull

 

TolalPlAL eutrophils

T I
m acropbtaalplszAilogIOI

 

 

 

 

 

 

2%
:- ﬁ l I 1
'3‘ -1 -05 0 05 1
A20/18S rRNA (logIO)

108

CHAPTER 7

REGRESSION AND PATHWAY ANALYSIS AND MODEL ADJUSTMENT OF TLR

SIGNALING PATHWAYS IN EQUINE BRONCHIAL EPITHELIAL CELLS

Summag

In this dissertation, I propose a hypothetical model for innate immune mechanisms
that are potentially involved in the pathogenesis of neutrophilic airway inﬂammation of
RAO-affected horses. We have shown that RAO-affected and control horses did not
differ in their inﬂammatory cell count either during stabling or on pasture. Therefore, in
this chapter, I sought to adjust the given model for the general horse population by
combining the data obtained in previous chapters. I analyzed TLR4, TLR2, IL-8, and
A20 mRN A expressions and total bronchoalveolar lavage inﬂammatory cell counts using
Spearman correlation and stepwise regression analyses. I concluded that airway
inﬂammation in the general horse population is at least partially due to stimulation of
TLR4 leading to IL-8 production and neutrophil accumulation. The latter may contribute
to accumulations of macrophages. Finally, I propose that stimulation of TLR2 leads to
A20 production, which is able to decrease BAL macrophage stimulation, which might be
involved in protection against exaggerated airway inﬂammation in healthy control horses

during stabling.

109

Introduction

Horses affected by recurrent airway obstruction develop neutrophilic airway
inﬂammation, airway obstruction, and mucus accumulation. Since the latter two have
been Shown to be a result of inﬂammatory processes, with the research presented in this
dissertation I was particularly interested in the pathogenesis of airway inﬂammation.

Although mechanisms for adaptive immune mechanisms have been reported
elsewhere, we and others gained evidence that innate immune mechanisms are also
involved in the pathogenesis of RAO. In chapter 1, I proposed a hypothetical model for
innate mechanisms that are potentially involved in neutrophilic airway inﬂammation
(Figure 1-1). Brieﬂy, in RAO-affected horses, the production of IL-8, which is a potent
neutrophilic chemoattractant in the equine lung, might be increased due to elevated
expression of the TLR4 and TLR2, which are central in the detection of inhaled
pathogens (e.g., bacteria, fungi, yeast). Stimulation of IL-8 production due to stimulation
of TLR4 and 2 by their ligands is negatively regulated by the zinc ﬁnger protein A20,
which itself is active during early stages of acute inﬂammation. Stable dust has been
reported to contain a mixture of respirable pathogens. Furthermore, in chapter 2 we
showed that concentrations of endotoxin, which is the primary ligand for TLR4, are
higher in stables than on pasture. Finally, cell culture studies provide evidence that LPS
stimulation can lead to up-regulation of TLR4 and 2 mRNA expressions.

In the studied horse population and under the published protocol we found that
the exposure of horses to a stable environment causes neutrophilic airway inﬂammation.

The stables were shown to be richer in endotoxin in comparison to the pasture (chapter

110

2), where these horses airway inﬂammation resisted. Although Stabling caused an
increase in neutrophils in the bronchoalveolar lavage ﬂuid (BAL) of all horses (RAO-
affected and control horses), this increase was higher in affected than control horses when
considering the percentage of neutrophils (chapter 4), but did not differ when considering
the total numbers of neutrophils (chapter 5). The signiﬁcant difference in percentage of
neutrophils between RAO-affected and control horses might have been rather the result
of the accompanied decrease in the percentage of lymphocytes than an actual difference
in the neutrophil numbers. Nevertheless, in the past others have also reported that
stabling causes neutrophilic accumulation not only in the airways from RAO-affected
horses but also in those from healthy control animals.

Similarly to our observation of an increase in total BAL neutrophils, we showed
that the amount of epithelial-derived IL-8 mRNA expression was elevated in both RAO-
affected and control horses due to stabling. Interleukin-8 is a neutrophilic chemoattractant
in the airways of RAO-affected horses during acute exacerbations. Our results were
supportive as we observed a correlation between the epithelial-derived IL-8 mRNA
expression and BAL neutrophils in RAO-affected horses (chapter 4). However, this was
not true for our control horse population.

In addition to the total numbers of BAL neutrophils and the IL-8 mRNA
expression in airway epithelium, the expression of TLR4 mRNA was also elevated in
RAO-affected and control horses during stabling than compared to pasture (chapter 4).
However, the expression of TLR2 and A20 mRNA was unaffected by stabling. Toll-like
receptor 4 expression strongly correlated with IL-8 mRNA expression in both RAO-

affected and control animals. Although TLR2 expression did not correlate with IL-8, it

111

was negatively associated with the numbers of neutrophils, macrophages, and
lymphocytes when all horses were pooled. Similarly, A20 was negatively associated with
these BAL inﬂammatory cells after pooling RAO-affected and control horses.

Therefore, these studies Show evidence that the hypothetical model I developed in
chapter 1 is partially true, in that RAO-affected horses in acute exacerbation exhibit a
pronounced expression of airway epithelial TLR4 expression, which leads to increased
IL-8 production and accumulation of neutrophils in the airway lumen. In addition, we and
others have reported that stabled horses without clinical signs or a history of airway
disease experience inﬂux of inﬂammatory cells into the airway lumen (29, 31, 57, 73,
184) On the other hand, we have shown that RAO-affected horses did not differ from
control horses in their BAL inﬂammatory cell counts when they are kept in the same
environment. Therefore, in this chapter I sought to adjust the given model for the general
horse population by combining the data obtained in chapters 4, 5 and 6 in a overall

statistical analysis.

Statistical analysis

The relative amount of mRNA was calculated using the relative standard curve
method. Quantitative means and their standard deviations (S.D.) for TLR4, TLR2, IL-8 and
A20 of each BB sample were standardized to the reference gene (188 rRNA). If
standardized quantitative mRNA expression data and BAL total cell counts were not

normally distributed (Kolmogorov-Smirnov P S 0.05), they were loglO transformed.

112

Correlations between single variables were determined using Spearman correlation analysis.
Full models for regression analysis were developed according to Figure 7-1. Each variable
was considered as the outcome of the combination of all possible predictors (i.e., in Figure
7-1: all variables that are at the same or a lower level than the deﬁned outcome are
considered to be predictors). For example, numbers of neutrophils is the outcome of a
combination of TLR4, TLR2, IL-8, and A20 as well as numbers of macrophages and
lymphocytes, or IL—8 is the outcome of a combination of TLR4, TLR2 and A20. Estimates,
standard errors, t-values, and the Rz-value are reported for these full models. For each full
model, I then sought to ﬁnd subsets of predictors using a stepwise regression model to limit
the possible models to a smaller number. For each step in the stepwise models estimates, F-
and P-values are reported. For all statistical analyses, data from RAO-affected and control

horses were pooled. Data were analyzed by use of SAS v. 9.1. Differences were considered

signiﬁcant for P-values S 0.05.

113

Results

TLR4, TLR2, 1L-8, and A20 mRNA expression: Toll-like receptor 4 and 2 mRNA
expression are reported in chapter 4 and 5, respectively. Interleukin-8 mRNA expression is
reported in both chapters 4 and 5. Whereas in chapter 4 only bronchial brushing samples
with more than 80% of epithelial cells were used in data analysis, here we included all
samples with detectable TLR4 expression. There were no differences in the results of TLR4
data between the two methods. Epithelial-derived A20 expression is discussed in chapter 6.

Bronchoalveolar lavage neutrophils, macrophages and lymphocytes: Total
inﬂammatory cell counts are reported in chapters 4, 5 and 6.

Correlations between BAL inﬂammatory cell counts and bronchial brushing mRNA
expressions: Correlations of pooled data between total BAL neutrophils and TLR2 or IL-8
are shown in chapter 5 and between BAL inﬂammatory cells and A20 as well as among
inﬂammatory cells are reported in chapters 6. Beside the negative correlation between TLR2
and total BAL neutrophils (chapter 5), in the present analysis we also found negative
correlations between TLR2 and total BAL macrophages and lymphocytes (Table 7-1). As in
chapter 4, where we reported an association between epithelial-derived TLR4 and IL-8
mRNA in RAO-affected horses, we also found a very strong and highly signiﬁcant
correlation between those variables when data from all horses were pooled. Finally, we
observed a strong and highly signiﬁcant correlation between A20 and TLR2 mRNA
expression (Table 7-1 and Figure 7-2).

Regression analyses: Full models for regression analyses are shown in Table 7-2. In

the full models for the outcomes of total numbers of neutrophils and lymphocytes in BAL

114

the t-values were highest for the macrophages (t—value = 3.95 and 3.88, respectively). In the
model for the outcome of the total numbers of macrophages, t-values for BAL lymphocytes
(t-value = 3.88) and BAL neutrophils (t-value = 3.95) were higher than from other
predictors. In the models for the outcomes of epithelial-derived IL-8 and A20 mRNA, TLR4
and TLR2 mRNA, respectively, are the coefficients with the highest t-values (t-value =
10.59 and 7.2, respectively).

Stepwise models are presented in Table 7-3. Whereas the numbers of macrophages
dominated in comparison to any other predictors in the full model for the outcome of total
numbers of neutrophils in BAL, the stepwise analysis showed that the combination of BAL
macrophage numbers in combination with the amount of epithelial—derived TLR4 mRNA
resulted in a higher Rz-value than in a model with BAL macrophages alone (Rz-value = 0.64
than 0.33, respectively). The Rz-value of the former was similar to the one we observed in
the full model (Rz-value = 0.66).

When using the stepwise regression procedure for the outcome of total macrophage
numbers in BAL, we observed the highest RZ-value of a model including not only numbers
of neutrophils and lymphocytes, but also IL-8 and A20 mRN A expressions (RZ-value = 0.8).
As compared to the Rz-values in steps 1 through 3, this RZ-value was the same as for the full
model. Although there was an elevation of the Rz-value when adding A20 in the model (step
4), its P-value was only signiﬁcant at a level of 0.1. In addition, both IL-8 and A20 were
negatively associated with the macrophage number.

For the outcomes of total numbers of BAL lymphocytes and epithelial-derived IL-8

mRNA there was no additional information than those reported for the ﬁll] models. The R2—

115

values in the stepwise models are similar or the same than those in full models (Lym: R2-
value = 0.62 and 0.63, respectively; IL-8: Rz-value = 0.81 and 0.81, respectively).

A combination of TLR2 and TLR4 mRNA expression signiﬁcantly contributes to the
regression model that explains the amount of epithelial—derived A20 mRNA. Whereas the
amount of TLR2 was positively associated with the outcome, TLR4 seemed to contribute in
a negative manner. In comparison to the full model for A20, by using the stepwise
procedure TLR4 was added. However, its P-value is not as small and its F-value is not as
high as the appropriate values reported for the predictor TLR2 (TLR4: P-value = 0.02, F-
value = 5.6; TLR2: P-value = < 0.0001, F-value = 53.35). This might explain the small
increase in the Rz-value from step 1 to step 2 (RZ-value = 0.57 and 0.63, respectively).

Nevertheless, the Rz-value of the model in step 2 resembles the one for the full model.

Discussion

With the results of these analyses I developed a potential model for predictors of
airway inﬂammation in the general horse population (Figure 7—3). In chapter 4 we
reported a positive correlation between the percentage of neutrophils and TLR4 or IL-8
mRNA expression in RAO-affected horses. In the present study we found that the
correlation between the total numbers of BAL neutrophils and TLR4 mRNA expression
or IL-8 mRNA expression at least tended to be signiﬁcant as well (P = 0.07 or 0.06,
respectively). We showed in chapter 5 that total BAL neutrophils correlated negatively

with epithelial-derived TLR2 mRNA, and in chapter 6 that total BAL neutrophils also

116

correlated with BAL macrophages or BAL lymphocytes. The regression analyses we
performed here showed that the number of macrophages in BAL predicts the number of
neutrophils even when other possible predictors for neutrophils in the airways (e.g., IL-8,
TLR4, TLR2) are present as well. However, the full model’s R2-value was only achieved
when macrophages and TLR4 mRNA expression were combined during the stepwise
analysis. Therefore, besides TLR4 signaling in airway epithelium, macrophages in BAL
may also contribute to the accumulation of neutrophils within the airways. However,
since the numbers of BAL macrophages were not different during stabling than on
pasture, other characteristics than the macrophage number might be important in
determination of the amount of neutrophils, such as the release of Chemoattractants due to
activation. For example, it has been reported that MIP-2 or IL—8, which are potent
neutrophilic attractants in the horse lung, are released by equine alveolar macrophages
upon stimulation.

In our studies in chapter 6 we observed that total BAL macrophages correlate
positively with total BAL neutrophils and lymphocytes as well as negatively with
epithelial-derived A20 mRNA expression. In the present analyses we also showed that it
correlated negatively with TLR2. The Rz-value of the full regression model that explains
BAL macrophage numbers was reached during the stepwise regression analysis when
BAL neutrophils and lymphocytes were combined with IL-8 and A20. While neutrophils
and lymphocytes were positively associated with macrophage numbers, IL-8 and A20
contributed with negative associations to macrophage numbers in the selected model.
Mechanisms by which neutrophils can contribute to the accumulation of macrophages in

the airway lumen are, for example, the release of proteases, which may function in tissue

117

destruction and, therefore, would ease the trans-endothelial migration of inﬂammatory
cells. Proteases, such as equine elastase 2-A or matrix metalloproteinase (MMPs,
especially, MMP9), have been shown to be associated with airway inﬂammation in
equine airways. Macrophages are inﬂammatory cells that belong to the early
inﬂammatory response and might contribute to attraction of lymphocytes in subsequent
inﬂammatory processes, such as the presentation of pathogenic peptides to T-
lymphocytes. This might explain the inclusion of lymphocytes in the regression model
for macrophages, which is supported by the regression model for total BAL lymphocyte
numbers. Stimulation of TLR4 causes expression of the monocyte chemoattractant CCL2
(monocyte chemotactic protein 1 [MCP-1]) (111). Since TLR4 signals through activation
of NF-KB and because A20 inhibits the NF-KB translocation it might explain the negative
association between BAL macrophage numbers and A20 mRNA expression.
Investigation of production of Chemoattractants for monocytes due to TLR signaling in
airway epithelial or other airway cells should be the subject of future studies.

In chapter 4 we reported a strong correlation between TLR4 and IL-8 mRNA
expressions. Here we showed that indeed TLR4 mRNA expression explains the IL-8
production.

Finally, the regression models for A20 mRNA expression show that epithelial-
derived TLR2 mRNA is the strongest predictor, since the RZ-value in the stepwise model
that only includes TLR2 as predictor (Rz-value = 0.57) contributes over 90% to the full
model Rz-value (Rz-value = 0.63). In addition, we observed a highly signiﬁcant negative
correlation between epithelial-derived TLR2 and A20 mRNA expression (Table 7-1 and

Figure 7-2). To the authors’ knowledge this association has not been reported previously.

118

However, a strong correlation between TLR2 and A20 mRNA expression is not
unexpected. Indeed, A20 was originally characterized as a TNF-a-inducible gene (135),
and subsequently, it was shown that it is a NF-KB target gene. As such, A20 is inducible
by a wide variety of stimuli (92). While TLR2 stimulation in human airway epithelial
cells leads to IL-8 production by NF-KB activation (60), A20 is a negative regulator of
inﬂammation by inhibiting nuclear NF-KB translocation (134). Our data show that TLR2
and A20 expression are associated with each other. One may assume that A20 is involved
in modulation of TLR2-induced inﬂammation. As a matter of fact we showed here and in
the previous chapters that both A20 and TLR2 are negatively correlated with total
numbers of BAL neutrophils, macrophages, and lymphocytes. Alternatively, a common
stimulus for or the use of a common pathway that leads to TLR2 and A20 expression
could also explain a correlation between these variables.

In summary, I propose that airway inﬂammation in the general horse population is
at least partially due to stimulation of TLR4, which may cause IL-8 production and
neutrophilic airway accumulation. The latter may contribute to accumulations of
macrophages by, for example, release of mediators, such as proteases, or due to
decreasing macrophage apoptosis processes. In addition, stimulation of TLR2 may cause
A20 production, which is able to decrease BAL macrophage accumulation and which
may protect against exaggerated airway inﬂammation in healthy control horses during

stabling.

119

Table 7-1: Correlation matrix of Spearman correlations between total numbers of
bronchoalveolar lavage (BAL) inﬂammatory cells (neutrophils, macrophages and
lymphocytes) and epithelial-derived TLR4, TLR2, A20, and IL-8 mRNA after pooling
data from RAO-affected and control horses.

Neu = Total BAL neutrophil count (/ul); Mac = Total BAL macrophage count (/ul); Lym
= Total BAL lymphocyte count (/ul); TLR4 = Toll-like receptor 4; TLR2 = Toll-like
receptor 2; A20 = zinc ﬁnger protein A20; IL-8 = interleukin 8; SCC = Spearman

correlation coefﬁcient; P = P-value; N = number of samples; n/a = not applicable

 

 

 

 

 

 

 

 

 

Neu Mac Lym TLR4 TLR2 A20 lL-8
Neu SCC 1.00
P n/a
N 44.00
Mac SCC 0.62 1.00
P <0.0001 n/a
N 44.00 47.00
Lym SCC 0.48 0.83 1.00
P 0.00 <0.0001 n/a
N 44.00 47.00 47.00
TLR4 SCC 0.30 -0.24 -0.25 1.00
P 0.07 0.14 0.12 n/a
N 37.00 40.00 40.00 41.00
TLR2 SCC -0.36 -0.56 -0.53 0.21 1.00
P 0.02 <0.0001 0.00 0.19 Na
N 43.00 46.00 46.00 41.00 47.00
A20 SCC -0.39 -0.46 -0.42 -0.13 0.83 1.00
P 0.01 0.00 0.01 0.43 <0.0001 n/a
N 39.00 42.00 42.00 37.00 43.00 43.00
IL-8 SCC 0.29 -0. 14 -0.23 0.84 0.13 -0.11 1.00
P 0.06 0.36 0.13 <0.0001 0.41 0.47 n/a
N 42.00 45.00 45.00 40.00 46.00 42.00 46.00

 

 

 

 

 

 

 

 

120

 

Table 7-2: Full models for regression analysis. For each outcome a model combining
possible predictors is shown. Predictors were chosen according to the ﬂow diagram
shown in Figure 7-1. Each variable was considered as an outcome of the combination of all
possible predictors (i.e., in Figure 7-1: all variables that are at the same or a lower level than
the deﬁned outcome are considered to be predictors). The estimate, standard error, and t-
value are reported for each predictor in the model. The overall Rz-value is reported for each
full model.

Neu = Total BAL neutrophil count (/ul); Mac = Total BAL macrophage count (/ul); Lym
= Total BAL lymphocyte count (/ul); TLR4 = Toll-like receptor 4; TLR2 = Toll-like

receptor 2; A20 = zinc ﬁnger protein A20; IL-8 = interleukin 8

121

 

 

 

 

 

 

 

 

Outcomes Predictors Estimate Standard Error t-value P-value Rz-value
Neu Intercept 1.82 1.24 1.47 0.15 0.66
Mac 0.92 0.23 3.95 0.0006
Lym -0.12 0.24 -0.51 0.61
TLR4 0.79 0.41 1.93 0.06
TLR2 -0.26 0.26 —1.00 0.33
A20 0.09 0.09 1.00 0.32
lL-8 0.12 0.32 0.39 0.80
Mac Intercept 0.16 0.87 0.18 0.86 0.80
Neu 0.42 0.11 3.95 0.0006
Lym 0.49 0.13 3.88 0.0007
TLR4 -0.25 0.29 -0.85 0.40
TLR2 0.06 0.18 0.34 0.74
A20 -0.08 0.06 -1.38 0.18
lL-8 -0.19 0.21 -0.90 0.38
Lym Intercept 2.46 0.97 2.55 0.02 0.63
Neu -0.09 0.17 -0.51 0.61
Mac 0.76 0.20 3.88 0.0007
TLR4 0.05 0.37 0.14 0.89
TLR2 -0.13 0.23 -0.56 0.58
A20 0.04 0.07 0.48 0.63
IL-8 -0.03 0.27 -0.11 0.91
lL-8 Intercept 0.87 0.16 5.40 <0.0001 0.81
TLR4 1.15 0.11 10.59 <0.0001
TLR2 0.02 0.14 0.15 0.88
A20 -0.01 0.04 -0.14 0.89
A20 Intercept -0.14 0.97 -0.15 0.88 0.63
TLR4 -0.89 0.99 -0.91 0.37
TLR2 2.69 0.37 7.20 <0.0001
IoglL8 -0.10 -0.10 -0.14 0.89

 

122

 

Table 7-3: Stepwise regression analysis. For each full model reported in Table 7-1 a
stepwise analysis procedure was applied. The estimate, F-, P-, and RZ-values are reported
for each predictor in individual steps.

Neu = Total BAL neutrophil count (/ul); Mac = Total BAL macrophage count (/ul); Lym
= Total BAL lymphocyte count (/ul); TLR4 = Toll-like receptor 4; TLR2 = Toll-like

receptor 2; A20 = zinc ﬁnger protein A20; IL-8 = interleukin 8

123

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Outcome Predictors 7‘1 step 2nil step 31a step 4'" ste
Neu Intercept Estimate 1.61 1.69
F-value 2.82 5.55
P-value 0.10 0.03
Mac Estimate 0.65 0.82
F-value 14.82 39.33
P-value 0.0006 <0.0001
TLR4 Estimate 0.87
F-value 24.73
P-value <0.0001
Rz-value 0.33 0.64
Mac Intercept Estimate 0.24 -0.40 -0.17 0.51
F-value 0.10 0.31 0.06 0.46
P-value 0.75 0.58 0.80 0.50
Lym Estimate 0.86 0.73 0.55 0.50
F-value 49.17 37.41 19.90 16.66
P-value <0.0001 <0.0001 0.0001 0.0004
Neu Estimate 0.28 0.42 0.39
F-value 8.60 17.97 15.93
P-value 0.01 0.0002 0.0005
IL-8 Estimate -0.32 -0.34
F-value 8.12 9.74
P-value 0.01 0.0043
A20 Estimate -0.06
F-value 2.98
P-vaIue 0-10
RZ-value 0.62 0.71 0.77 0.80
Lym Intercept Estimate 2.20
F-value 14.13
P-value 0.0007
Mac Estimate 0.72
F—value 49.17
P-value <0.0001
Rz-value 0.62
IL-8 Intercept Estimate 0.88
F-value 46.32
P-value <0.0001
TLR4 Estimate 1.15
F-value 141.77
P-value <0.0001
RZ-value 0.81
A20 Intercept Estimate 1.08 -0.23
F-value 6.09 0.12
P-value 0.02 0.74
TLR2 Estimate 2.62 2.69
F-value 44.77 53.35
P-value <0.0001 <0.0001
TLR4 Estimate -1.01
F-value 5.60
P-value 0.02
RZ-value 0.57 0.63

 

 

 

 

 

 

 

 

124

 

Figure 7-1: Flow diagram for developing full models for regression analyses. All
variables that are at the same or at a lower level than the deﬁned outcome are considered to

be predictors. All variable were considered to be outcomes.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Model
Level 1 TLR4 TLR2
Level 2 lL-8 A20
Level 3 BAL neutrophils BAL macrophages BAL lymphocytes

 

 

 

 

 

 

 

 

 

125

Figure 7-2: Association between TLR2 and A20 mRNA expressions in bronchial
brushing samples (BBS) from all horses. TLR2 and A20 mRNA expressions are shown in

relation to 18S rRNA (reference gene) expressions.

  
 

0.5

A20/18$ rRNA (log10)
2';

TLR2/18$ rRNA (10910)

126

Figure 7-3: A potential model for predictors of airway inﬂammation in the general horse

population.

AIMI'y
epithellum
. J

 
    

 

 

I
I
I
I
l
I
I
I
I
I
I
I
I
I
I
I
I
I
V

  
   

 

   

» BAL neutrophils

    

BAL

.— -. . ' BAL lymphocytes ;
3*.“ _ . .
macrophages f a"

127

CHAPTER 8

INTERPRETATION AND OUTLOOK

The research of this dissertation contributed to the understanding of the complex
nature of immune mechanisms that are involved in the pathogenesis of RAO. Whereas in
the past, research groups focused on cytokine and chemokine changes in BAL cells (e. g.,
all cells, isolated macrophages, or lymphocytes), I demonstrated that epithelial cells may
also play a role in protecting the lung against inhaled pathogens. In addition, the current
studies showed that key players in the innate immune response (e. g., TLR-4 and IL-8) are
associated with airway neutrophilia. Finally, I observed an association between the
regulatory protein A20 with inﬂammatory cells in BAL. Despite the fact of increased
TLR4 and IL-8 mRNA expression, we did not have evidence of a parallel increase in the
amounts of their proteins. There is evidence in the literature that an increase in the TLR4
mRNA does not necessarily lead to an elevation of the TLR4 protein (125). However, an
increase in TLR4 and IL-8 mRNAs that were accompanied by an elevation of their
proteins due to LPS injection or natural challenge, respectively, has been shown for
equine lung tissue (4, 170). Therefore, it can be predicted that the changes in mRNA
expression that are reported in my dissertation are associated with quantitative
differences in the receptor protein.

Although the pathogenesis of RAO still remains ill deﬁned, it becomes more and
more clear that a genetic disease predisposition may underlie aberrant immune

mechanisms that contribute to the RAO phenotype (i.e., airway neutrophilia, mucus

128

accumulation, and hyper-responsiveness). Therefore, the challenge in the future will be to
ﬁnd genes or traits that eventually are associated with those phenotypic outcomes and to
understand the complexity of their interconnecting pathways. For these purposes whole
equine genome sequence information, advanced analytical (e.g., genome or targeted
expression chips), statistical and bioinformatic tools, and searchable databases need to
become the center of our focus. The amount of information that is gained from whole-
genome sequencing projects and expression analyses can only be understood in the
context of detailed veriﬁcation experiments on the molecular and cellular levels using

technologies, such as qRT-PCR, ELISAs, and immunohistochemistry.

Areas for future research

Our understanding that RAO can be inherited originated with the report by Marti
et al. (112) and this area will be able to be investigated in more detail with the public
online access of the complete equine genome information in the near future. Since the
beginning of sequencing efforts for the whole Equus caballus genome in March 2006,
equine sequences are currently available at the Trace Archive at NCBI under the web
address:

(http://O-wwwncbi.nlm.nih.gov.library.vu.edu.au/Traces/trace.cgi). With about 31 x 10‘5
traces of about 700 base pairs per trace there are about 2.17 x 1010 base pairs reported for

the equine genome today. Since mammalian genomes are about 3 x 109 base pairs in

129

length, the current online equine genome sequence information provides approximately
72-fold coverage.

Because of the similarity to human asthma it can be assumed that the phenotypes
of RAO are quantitative rather than discrete traits. For the former it is characteristic to be
under the inﬂuence of many genes, gene-gene, and gene-environment interactions. In
addition, quantitative traits vary over a wide range within a population, which can be the
basis for identifying disease-dominating genotypic variations (phenotype-driven
strategies for gene identiﬁcation). In contrast to mouse strains that were inbred to
generate homozygosity of their genomes, the equine population represents a pool of
phenotypic and, therefore, genotypic variation that would make it an appropriate animal
model for quantitative trait analysis (QTL analysis). Disadvantages, such as high
maintenance costs, the large size of these animals, and their long gestation and generation
time, in comparison to other laboratory animals (e.g., mice, rats) are limiting factors. The
still existing gaps in the database information for the equine genome (e.g., number of
SNPs) further limit appropriate experimental design today. However, the latter may be
overcome in the next few years, considering the fast growth of information since the
completion of the human and mouse genome projects in 2000-2003 and 2002,
respectively. Therefore, it is worthwhile to mention some theoretical considerations when
using QTLs as a technique to identify genes that are potentially involved in causing the
RAO phenotypes, such as airway neutrophilia, mucus accumulation, or airway
hyperreactivity.

A QTL study would need appropriate parents (one of the parents should be RAO

affected), back- or F2 intercrosses, pheno- and genotyping of the offsprings for

130

polymorphic markers, and regression analysis (165). The goal is to ﬁnd associations
between allelic and phenotypic variations. Single nucleotide polymorphisms (SNPs),
which are natural sequence variations, are the markers of choice in many gene mapping
experiments these days. For use in a QTL analysis SNP markers should be chosen so that
they are evenly spaced throughout the genome.

Levels of gene expression can also be considered as a quantitative trait (155).
Expression QTLs (eQTLs) can be obtained by the use of microarray data. Besides
identiﬁcation of genes that underlie a QTL, eQTLs and microarrays are especially
important in pathway analysis. The combination of data from genetic and gene-
expression analyses may lead to identiﬁcation of key regulatory genes that primarily
determine the phenotypic outcome (37). An alternative to using microarray data for
eQTL analysis is the Microsoft Excel-based software algorythm termed Global Pattern
Recognition (GPR), which is available at the following web address:
http://www.jax.org/staff/roopenian/labsite/gene_expression.html.

The amount of information we might gather from whole genome sequencing or
expression analyses cannot be interpreted without detailed studies on the molecular and
cellular level. The following recommendations for future research questions that result
from the present thesis may lead us to another level of understanding of immunological
processes in the lung of horses.

Questions resulting from chapter 2:

o How do different types of lipopolysaccharides (endotoxin) or fungi interact

with airway epithelial cells?

0 What is the effect of LPS inhalation on TLR expression?

131

Questions resulting from chapter 4:

Are TLRs other than TLR4 and TLR2 expressed in the equine lung, and is

their expression proﬁle associated with the pathogenesis of RAO?

Questions resulting from chapter 5:

What is the expression proﬁle and function of the TLR2 co-receptors TLR1
and TLR6?

Where are the TLRS localized within a cell?

Which transcription factor binding sites can be determined in the promoter
region of TLRS?

Does the TLR expression and stimulation differ in different airway cells, such
as macrophages, neutrophils, epithelial cells, dendritic cells, and airway
smooth muscle cells?

Which effector molecules in the TLR signaling pathway (e.g., MyD88,

TRAP-6 etc.) are important in the pathogenesis of RAO?

Questions resulting from chapter 6:

What are the stimulatory and regulatory mechanisms for A20?

What is the dynamic of A20 expression at early time points in RAO-affected
and control horses?

What regulatory mechanisms are involved in modulating the early inﬂux of
neutrophils into the airways in control horses?

What are the mechanisms that prevent an exaggerated airway neutrophilia in

control horses?

132

Examples of immunological studies in the ﬁeld of RAO by different research
groups are given in Table 8-1. It is apparent that comparisons and interpretations between
studies are currently difﬁcult due to the use of different protocols and methods.
Especially for the above-discussed future areas of research, study designs (animals and
protocols) and methodology standardization need to be considered. Although a
standardization of protocols and methodologies may seem to limit the scientiﬁc variety, a
broader range of information might be gained over a shorter period of time. Other factors,
such as ﬁnancial, time, and labor efforts, may also limit globally accepted agreements for
common protocols. Nevertheless, efforts for the acceptance of a common deﬁnition of the
RAO phenotype (151) could pave the way for further interactive research efforts.

Some areas in which a consensus would enable us to compare results between
studies are mentioned in the following discussion:

0 Develop a common deﬁnition of baseline. After stabling, how long do RAO-
affected horses need to be maintained on pasture to be non-differentiable from
control horses (remission)? Although it has been reported that the numbers of
BAL inﬂammatory cells and lung function do not differ between RAO-affected
and control horses after returning horses to pasture for two weeks (13), there is
evidence for continued subclinical inﬂammatory processes on pasture (102, 24).
For example, Sandersen et al. reported that the transcription factor NF -I(B activity
in RAO-affected horses does not return to baseline activity after 3 weeks on
pasture (161). Furthermore, Dewachi et al. showed that the receptors for IL-5 and
IL-9 on equine peripheral neutrophils were elevated in RAO-affected horses

compared to control horses after more than 3 months on pasture (34). Finally, it

133

will be of interest to determine if anti-inﬂammatory drugs are able to speed up or
are necessary to initiate the processes toward complete remission.

Determine biologically important time points for sample collection. Since we are
still not able to study pathological changes in whole biological systems during
disease development in real time, we have to rely on data obtained at different
time points, during which important changes in the disease pathogenesis may
occur. Using animal models of allergic lung diseases, a response to inhaled
pathogens occurs within minutes of inhalation by susceptible animals (early
response) (75). Although the amount of IgE is elevated in lungs from RAO-
affected horses (40, 63, 162), the early allergic response seems to be absent in the
RAO pathogenesis. Subsequently, the late-phase inﬂammatory reaction can be
observed after six to nine hours of inhalation challenge, which is characterized by
inﬂux of inﬂammatory cells into the airway lumen, production and release of
cytokines, and development of sustained airway obstruction (17, 106). Indeed
Fairbain et al. showed an inﬂux of neutrophils into equine airways beginning at 6
hours of exposure (42). Furthermore, time-course studies of cytokine expressions
may reveal additional time points during or after stabling, during which it might
be worthwhile to obtain data.

Using the best match of diseased and control animals. Recurrent airway
obstruction is a pulmonary disorder of primarily mature horses older than eight
years (19). Since we currently do not understand the processes that are involved in
the development of the disease, the use of age- and breed-matched horse pairs is

of advantage in case control studies. Although cost and time intensive,

134

 

establishing herds of known pedigrees could help in determination of age-related
difference, familial associations, breed associations, etc. Although laboratory
animal models, such as the mouse, are valuable for high-throughput research,
because of its small size, cost effective maintenance, and short gestation time (~
19 to 20 days), horses affected by RAO provide the advantage of their naturally
close proximity to the phenotypes of acute and occupational asthma. Gathering
information from longitudinal studies by using horses with well-deﬁned pedigree
and disease history may yield informative data about the complexity in the
pathogenesis that we might not be able to obtain from laboratory animal models

or cell culture experiments.

In conclusion, with this dissertation I contributed to our understanding of innate immune
mechanisms in equine airways. First, I showed that in our stables, animals are exposed to
higher endotoxin concentrations than on pasture and that the endotoxin receptor (TLR4)
is more abundant in airway epithelium when horses are kept in this environment as
compared to an outdoor environment. Similarly, when exposed to stable dust, IL-8
expression was also elevated in bronchial epithelial cells, most likely due to TLR4
signaling. Interestingly, all horses showed neutrophilic inﬂux into the airway lumen;
however, only the RAO-susceptible animals developed airway obstruction. Although,
RAO-affected horses had in general a higher IL-8 production, other mechanisms, such as
negative inﬂammatory regulators, like A20, or activation of other airway located
inﬂammatory cells, such as alveolar macrophages, may be important players in the

pathogenesis. Therefore, it is evident that the complex nature of these processes cannot

135

be discussed and understood in detail by studying a “tiny” subset of potentially important
genes in such a complex biological system. Future studies that will analyze expression
proﬁles and pathways on a basis of whole genome sequence availability will become
available in the near future and will possibly incorporate new challenges, such as the use

of bioinforrnatical tools.

136

Table 8-1: Examples of immunological studies in the ﬁeld of RAO by different research
groups.

TNF-a = Tumor necrosis factor alpha; IL-lB = Interleukin 1 beta; IL-6 = Interleukin 6;
IL-8 = Interleukin 8; IL-10 = Interleukin 10; NF-KB = Nuclear factor kappa B; 1- KB =
Inhibitor of NF-KB; BBs = bronchial brushing samples; BAL = bronchoalaveolar lavage;
ICAM-l = Intracellular adhesion molecule 1; GM-CSF = Ganulocyte macrophage-
colony-stimulating factor; STAT-5 = Signal transducer and activator of transcription 5;
MIP-2 = Macrophage inﬂammatory protein-2; IL-4 = Interleukin 4; IFN-y = Interferon
gamma; IL-l7 = Interleukin 17; IL-SR = Interleukin 5 receptor; IL-9R = Interleukin 9
receptor; TLR4 = Toll-like receptor 4; TLR2 = Toll-like receptor 2; A20 = zinc ﬁnger
protein A20; CXCL] = CXC chemokine ligand 1; G-CSF = Granulocyte-colony
stimulating factor; LPS = lipopolysaccharide; AP = Aspergillus fumigatus; HDS = Hay
dust suspension; CB = clenbuterol; PBS = Phosphate buffered saline; n.m. = not
measured; APplmax = Difference between inspiratory and expiratory pleural pressure;

EMSA = Electromobility shift ssay; ISH = In-situ hybridization

137

 

 

 

 

 

 

 

 

8 5 89.82.. .58... £8; 4 .u .. of .5... sea .3. o~< .m
.5 m. A 62...“... J. .3: .8...»
>88... .0... .86 $0.26
.3on .m-..
.4. «one 38 8 .8 .3 .F 228.. 85. m A 85.6.8 as. 4...: .2-..
.8 mom-.. 8.83 m s8 F 228.. 5.2 m A 85.68 .5. w-.. #2.. .2-..
.43 or. £8 E 228.. 8.2 m A 39.8 .5. mm-.. 48-...
2.8 «on; 9.3 mm 0.2me :22 v A 26.8 $9. 2...:
.8. :m. 9.8 m 2.... F 228.. 85. m A 39.» .5. >2... .3. .3.
.8. :m. 5.2 m A .5... 26.7. in... >-z.. .m-.. f...
.68... 6.9.6.. 3002. m .ocm >-z“._
.8. .880... .58... .of of .2 f... .9... .w
.5 cm A 68.9% so 2 v 65.8% J. .9-.. 9.2»
Emma 8.9.0..
69 .63... 890... 33.09... 3m...» so; w-..
m_xSoEoco
.5. «0...... .58 .. «N .5 .m .m .3 w... 8.8.... 20.8 .52. m5 2.... a-..
.8: <25 .5... 8.8.8 9.5.. .E... .288. E. m
588 8.832.... -25 80.20
.5: <msm. 9.83 m 8.8 .E... .838. .9. ...<m s. m... .”.z
.o 5ch .otm m... vm
Cw. $032.38.... 9.00; m m_m_.o .E.c 838m. >9. .mmm cc
.<m....m .o .88 .9... m... a 7.25. .9. .2
ANN. m.o_aoc...EE_ 333 m m_w_.o Ed $393 .6: Ammm
.<w.2m. .0 .85 .28 m... .VN s. m. -. 9.-..2
mo 5.3 5.9a 9.8 m .mtm
.8. «on... a... «N .o .E... 89.8 30.8.9.2. mo: .2 .mA: 2... a-.. .o
85.68 ”cos. N A -.. as... .e-.z..
.8. «.08... m... .VN .o .E... 88...... €28.25. mo: ..< .mA: 8... 5-... .o
8.....ch ”8.2 N A -... .9-.. .35
.8. 9.8805 .5... m... cm mm. 8.3.8 :8. we: ..< .8: 9-... .55
09.2.9.0 .otm mmco=mzo .o 59.0. 9......me
com 352. 5.32.3 29:3 350... 69.2.2.0 «no.3...

.0 3.38.0!

 

 

138

BIBLIOGRAPHY

1. Abrams L, Seixas N, Robins T, Burge H, Muilenberg M, and Franzblau A.
Characterization of metalworking ﬂuid exposure indices for a study of acute respiratory
effects. Appl Occup Environ Hyg 15: 492-502, 2000.

2. Aggarwal N, and Holmes MA. Characterisation of equine T helper cells:
demonstration of Th1- and Th2-like cells in long-term equine T-cell cultures. Res Vet Sci
66: 277-279, 1999.

3. Ainsworth DM, Grunig G, Matychak MB, Young J, Wagner B, Erb HN, and
Antczak DF. Recurrent airway obstruction (RAO) in horses is characterized by IFN-
gamma and IL-8 production in bronchoalveolar lavage cells. Vet Immunol Immunopathol
96: 83-91, 2003.

4. Ainsworth DM, Wagner B, Franchini M, Grunig G, Erb HN, and Tan JY.
Time-dependent alterations in gene expression of interleukin-8 in the bronchial
epithelium of horses with recurrent airway obstruction. Am J Vet Res 67: 669-677, 2006.

5. Akhtar M, Watson JL, Nazli A, and McKay DM. Bacterial DNA evokes
epithelial IL-8 production by a MAPK-dependent, NF -kappaB-independent pathway.
FasebJ 17: 1319-1321, 2003.

6. Akira S, and Takeda K. Toll-like receptor signalling. Nat Rev Immunol 4: 499-
511, 2004.

7. Alexopoulou L, Holt AC, Medzhitov R, and Flavell RA. Recognition of
double-stranded RNA and activation of NF -kappaB by Toll-like receptor 3. Nature 413:
732-738, 2001.

8. Altieri DC, and Edgington TS. A monoclonal antibody reacting with distinct
adhesion molecules deﬁnes a transition in the functional state of the receptor
CD1 1b/CD18 (Mac-l). JImmunol 141: 2656-2660, 1988.

9. Andersen-Nissen E, Smith KD, Bonneau R, Strong RK, and Aderem A. A
conserved surface on Toll-like receptor 5 recognizes bacterial ﬂagellin. J Exp Med 204:
393-403, 2007.

10. Andonegui G, Bonder CS, Green F, Mullaly SC, Zbytnuik L, Raharjo E, and

Kubes P. Endothelium-derived Toll-like receptor-4 is the key molecule in LPS-induced
neutrophil sequestration into lungs. J Clin Invest 111: 1011-1020, 2003.

139

11. Bals R, and Hiemstra PS. Innate immunity in the lung: how epithelial cells ﬁght
against respiratory pathogens. Eur Respir J 23: 327-333, 2004.

12. Basu S, and Fenton MJ. Toll-like receptors: function and roles in lung disease.
Am J Physiol Lung Cell Mol Physiol 286: L887-892, 2004.

13. Berndt A, Derksen FJ, Venta P, Ewart S, Yuzbasiyan-Gurkan V, and
Robinson NE. ELEVATED AMOUNT OF TOLL-LIKE RECEPTOR 4 mRNA IN
BRONCHIAL EPITHELIAL CELLS IS ASSOCIATED WITH AIRWAY
INFLAMMATION IN HORSES WITH RECURRENT AIRWAY OBSTRUCTION. Am
J Physiol Lung Cell Mol Physiol 2006.

14. Beyaert R, Heyninck K, and Van Huffel S. A20 and A20-binding proteins as
cellular inhibitors of nuclear factor-kappa B-dependent gene expression and apoptosis.
Biochem Pharmacol 60: 1143-1151, 2000.

15. Bhatia M, and Moochhala S. Role of inﬂammatory mediators in the
pathophysiology of acute respiratory distress syndrome. JPath01202: 145-156, 2004.

16. Boone DL, Turer EE, Lee EG, Ahmad RC, Wheeler MT, Tsui C, Hurley P,
Chien M, Chai S, Hitotsumatsu O, McNally E, Pickart C, and Ma A. The ubiquitin-
modifying enzyme A20 is required for termination of Toll-like receptor responses. Nat
Immunol 5: 1052-1060, 2004.

17. Bousquet J, Jeffery PK, Busse W, Johnson M, and Vignola AM. Asthma.
From bronchoconstriction to airways inﬂammation and remodeling. Am J Respir Crit
Care Med 161: 1720-1745, 2000.

18. Bowles KS, Beadle RE, Mouch S, Pourciau SS, Littleﬁeld-Chabaud MA, Le
Blane C, Mistric L, Fermaglich D, and Horohov DW. A novel model for equine
recurrent airway obstruction. Vet Immunol Immunopathol 87: 385-3 89, 2002.

19. Bracher V, von Fellenberg R, Winder CN, Gruenig G, Hermann M, and
Kraehenmann A. An investigation of the incidence of chronic obstructive pulmonary
disease (COPD) in random populations of Swiss horses. Equine Vet J23: 136-141, 1991.

20. Brazil TJ, and McGorum BC. Molecules and mediators of inﬂammation in
equine heaves: mechanisms and markers of disease. Equine Vet J 33: 113-115, 2001.

21. Bureau F, Bonizzi G, Kirschvink N, Delhalle S, Desmecht D, Merville MP,
Bours V, and Lekeux P. Correlation between nuclear factor-kappaB activity in
bronchial brushing samples and lung dysfunction in an animal model of asthma. Am J
Respir Crit Care Med 161: 1314-1321, 2000.

22. Bureau F, Delhalle S, Bonizzi G, Fievez L, Dogne S, Kirschvink N,
Vanderplasschen A, Merville MP, Bours V, and Lekeux P. Mechanisms of persistent

140

NF-kappa B activity in the bronchi of an animal model of asthma. J Immunol 165: 5822-
5830, 2000.

23. Carter JD, Ghio AJ, Samet JM, and Devlin RB. Cytokine production by
human airway epithelial cells after exposure to an air pollution particle is metal-
dependent. Toxicol Appl Pharmacol 146: 180-188, 1997.

24. Carty CL, Gehring U, Cyrys J, Bischof W, and Heinrich J. Seasonal
variability of endotoxin in ambient ﬁne particulate matter. J Environ Monit 5: 953-958,
2003.

25. Castellan RM, Olenchock SA, Hankinson JL, Millner PD, Cocke JB, Bragg
CK, Perkins HH, Jr., and Jacobs RR. Acute bronchoconstriction induced by cotton
dust: dose-related responses to endotoxin and other dust factors. Ann Intern Med 101:
157-163, 1984.

26. Castellan RM, Olenchock SA, Kinsley KB, and Hankinson JL. Inhaled
endotoxin and decreased spirometric values. An exposure-response relation for cotton
dust. N Engl J Med 317: 605-610, 1987.

27. Christiani DC, Ye TT, Zhang S, Wegman DH, Eisen EA, Ryan LA,
Olenchock SA, Pothier L, and Dai HL. Cotton dust and endotoxin exposure and long-
term decline in lung function: results of a longitudinal study. Am J Ind Med 35: 321-331,
1999.

28. Chun DT, Chew V, Bartlett K, Gordon T, Jacobs RR, Larsson BM, Larsson
L, Lewis DM, Liesivuori J, Michel 0, Milton DK, Rylander R, Thorne PS, White
EM, and Brown ME. Preliminary report on the results of the second phase of a round-
robin endotoxin assay study using cotton dust. Appl Occup Environ Hyg 15: 152-157,
2000.

29. Cordeau ME, Joubert P, Dewachi O, Hamid Q, and Lavoie JP. IL-4, IL-5 and
IFN-gamma mRNA expression in pulmonary lymphocytes in equine heaves. Vet
Immunol Immunopathol 97: 87-96, 2004.

30. Davenpeck KL, Steeber DA, Tedder TF, and Bochner BS. P- and L-selectin
mediate distinct but overlapping functions in endotoxin-induced leukocyte-endothelial
interactions in the rat mesenteric microcirculation. J Immunol 159: 1977-1986, 1997.

31. Debrue M, Hamilton E, Joubert P, Lajoie-Kadoch S, and Lavoie JP. Chronic
exacerbation of equine heaves is associated with an increased expression of interleukin-
17 mRNA in bronchoalveolar lavage cells. Vet Immunol Immunopathol 105: 25-31, 2005.

32. Delclaux C, Delacourt C, D'Ortho MP, Boyer V, Lafuma C, and Harf A. Role

of gelatinase B and elastase in human polymorphonuclear neutrophil migration across
basement membrane. Am J Respir Cell Mol Biol 14: 288-295, 1996.

141

33. Derksen FJ, Scott JS, Slocombe RF, and Robinson NE. Micropolyspora faeni
causes airway inﬂammation but not hyperresponsiveness in sensitized ponies. J Appl
Physiol 62: 1398-1404, 1987.

34. Dewachi O, Joubert P, Hamid Q, and Lavoie JP. Expression of interleukin
(IL)-5 and IL-9 receptors on neutrophils of horses with heaves. Vet Immunol
Immunopathol 109: 31-36, 2006.

35. Donham KJ, Cumro D, Reynolds SJ, and Merchant JA. Dose-response
relationships between occupational aerosol exposures and cross-shift declines of lung
function in poultry workers: recommendations for exposure limits. J Occup Environ Med

42: 260-269, 2000.

36. Donham KJ, Reynolds SJ, Whitten P, Merchant JA, Burmeister L, and
Popendorf WJ. Respiratory dysfunction in swine production facility workers: dose-
response relationships of environmental exposures and pulmonary function. Am J Ind

Med 27: 405-418, 1995.

37. Drake TA, Schadt EE, and Lusis AJ. Integrating genetic and gene expression
data: application to cardiovascular and metabolic traits in mice. Mamm Genome 17: 466-
479, 2006.

38. Driscoll KE, Simpson L, Carter J, Hassenbein D, and Leikauf GD. Ozone
inhalation stimulates expression of a neutrophil chemotactic protein, macrophage
inﬂammatory protein 2. Toxicol Appl Pharmacol 119: 306-309, 1993.

39. Duits LA, Nibbering PH, van Strijen E, Vos JB, Mannesse-Lazeroms SP, van
Sterkenburg MA, and Hiemstra PS. Rhinovirus increases human beta-defensin-2 and -
3 mRN A expression in cultured bronchial epithelial cells. FEMS Immunol Med Microbiol
38: 59-64, 2003.

40. Eder C, Crameri R, Mayer C, Eicher R, Straub R, Gerber H, Lazary S, and
Marti E. Allergen-speciﬁc IgE levels against crude mould and storage mite extracts and
recombinant mould allergens in sera from horses affected with chronic bronchitis. Vet
Immunol Immunopathol 73: 241-253, 2000.

41. Evans AG, Paradis MR, and O'Callaghan M. Intradermal testing of horses
with chronic obstructive pulmonary disease and recurrent urticaria. Am J Vet Res 53: 203-
208, 1992.

42. Fairbaim SM, Marr KA, Lees P, Cunningham FM, and Page CP. Effects of
platelet activating factor on the distribution of radiolabelled leucocytes and platelets in
normal horses and asymptomatic horses with chronic obstructive pulmonary disease. Res

Vet Sci 61: 107-1 13, 1996.

142

43. Fan J, Frey RS, and Malik AB. TLR4 signaling induces TLR2 expression in
endothelial cells via neutrophil NADPH oxidase. J Clin Invest 112: 1234-1243, 2003.

44. Fan J, and Malik AB. Toll-like receptor-4 (TLR4) signaling augments
chemokine-induced neutrophil migration by modulating cell surface expression of
chemokine receptors. Nat Med 9: 315-321, 2003.

45. Fernandez N, Renedo M, Garcia-Rodriguez C, and Sanchez Crespo M.
Activation of monocytic cells through Fc gamma receptors induces the expression of
macrophage-inﬂammatory protein (MIP)-1 alpha, MIP-1 beta, and RANTES. J Immunol
169: 3321-3328, 2002.

46. Fernandez S, Jose P, Avdiushko MG, Kaplan AM, and Cohen DA. Inhibition
of IL-10 receptor function in alveolar macrophages by Toll-like receptor agonists. J
Immunol 172: 2613-2620, 2004.

47. Foster AP, Cunningham FM, and Lees P. Inﬂammatory effects of platelet
activating factor (PAF) in equine skin. Equine Vet J 24: 208-214, 1992.

48. Foster AP, Lees P, Andrews MJ, and Cunningham FM. Effects of WEB 2086,
an antagonist to the receptor for platelet-activating factor (PAF), on PAF-induced
responses in the horse. Equine VetJ 24: 203-207, 1992.

49. Foster AP, Lees P, and Cunningham FM. Platelet activating factor is a
mediator of equine neutrophil and eosinophil migration in vitro. Res Vet Sci 53: 223-229,
1992.

50. Franchini M, Gill U, von Fellenberg R, and Bracher VD. Interleukin-8
concentration and neutrophil chemotactic activity in bronchoalveolar lavage ﬂuid of

horses with chronic obstructive pulmonary disease following exposure to hay. Am J Vet
Res 61: 1369-1374, 2000.

51. Franchini M, Gilli U, Akens MK, Fellenberg RV, and Bracher V. The role of
neutrophil chemotactic cytokines in the pathogenesis of equine chronic obstructive
pulmonary disease (COPD). Vet Immunol Immunopathol 66: 53-65, 1998.

52. Frevert CW, Farone A, Danaee H, Paulauskis JD, and Kobzik L. Functional
characterization of rat chemokine macrophage inﬂammatory protein-2. Inﬂammation 19:
133-142, 1995.

53. Frevert CW, Huang S, Danaee H, Paulauskis JD, and Kobzik L. Functional

characterization of the rat chemokine KC and its importance in neutrophil recruitment in
a rat model of pulmonary inﬂammation. J Immunol 154: 335-344, 1995.

143

54. Gao B, and Tsan MF. Endotoxin contamination in recombinant human heat
shock protein 70 (Hsp70) preparation is responsible for the induction of tumor necrosis
factor alpha release by murine macrophages. J Biol Chem 278: 174-179, 2003.

55. Gao B, and Tsan MF. Recombinant human heat shock protein 60 does not

induce the release of tumor necrosis factor alpha from murine macrophages. J Biol Chem
278: 22523-22529, 2003.

56. Gerber V. Nature and causes of mucus accumulation in equine lower airway
disease. Dissertation Thesis 2003.

57. Gerber V, Lindberg A, Berney C, and Robinson NE. Airway mucus in
recurrent airway obstruction--short-term response to environmental challenge. J Vet
Intern Med 18: 92-97, 2004.

58. Gem JE, French DA, Grindle KA, Brockman-Schneider RA, Konno S, and
Busse WW. Double-stranded RNA induces the synthesis of speciﬁc chemokines by
bronchial epithelial cells. Am J Respir Cell Mol Biol 28: 731-737, 2003.

59. Giguere S, Viel L, Lee E, MacKay RJ, Hernandez J, and Franchini M.
Cytokine induction in pulmonary airways of horses with heaves and effect of therapy
with inhaled ﬂuticasone propionate. Vet Immunol Immunopathol 85: 147-158, 2002.

60. Con Y, Asai Y, Hashimoto S, Mizumura K, Jibiki I, Machino T, Ra C, and
Horie T. A20 inhibits toll-like receptor 2- and 4-mediated interleukin-8 synthesis in
airway epithelial cells. Am J Respir Cell Mol Biol 31: 330-336, 2004.

61. Guillot L, Medjane S, Le-Barillec K, Balloy V, Dane] C, Chignard M, and Si-
Tahar M. Response of human pulmonary epithelial cells to lipopolysaccharide involves
Toll-like receptor 4 (TLR4)-dependent signaling pathways: evidence for an intracellular
compartmentalization of TLR4. J Biol Chem 279: 2712-2718, 2004.

62. Guo RF, and Ward PA. Mediators and regulation of neutrophil accumulation in
inﬂammatory responses in lung: insights from the IgG immune complex model. Free
Radic Biol Med 33: 303-310, 2002.

63. Halliwell RE, McGorum BC, Irving P, and Dixon PM. Local and systemic
antibody production in horses affected with chronic obstructive pulmonary disease. Vet
Immunol Immunopathol 38: 201-215, 1993.

64. Hasday JD, Bascom R, Costa JJ, Fitzgerald T, and Dubin W. Bacterial
endotoxin is an active component of cigarette smoke. Chest 115: 829-835, 1999.

65. Hauber HP, Tulic MK, Tsicopoulos A, Wallaert B, Olivenstein R, Daigneault

P, and Hamid Q. Toll-like receptors 4 and 2 expression in the bronchial mucosa of
patients with cystic ﬁbrosis. Can Respir J 12: 13-18, 2005.

144

66. Hawn TR, Verbon A, Janer M, Zhao LP, Beutler B, and Aderem A. Toll-like
receptor 4 polymorphisms are associated with resistance to Legionnaires' disease. Proc
Natl Acad Sci U S A 102: 2487-2489, 2005.

67. Hayashi F, Smith KD, Ozinsky A, Hawn TR, Yi EC, Goodlett DR, Eng JK,
Akira S, Underhill DM, and Aderem A. The innate immune response to bacterial
ﬂagellin is mediated by Toll-like receptor 5. Nature 410: 1099-1103, 2001.

68. Hertz CJ, Wu Q, Porter EM, Zhang YJ, Weismuller KH, Godowski PJ,
Ganz T, Randell SH, and Modlin RL. Activation of Toll-like receptor 2 on human

tracheobronchial epithelial cells induces the antimicrobial peptide human beta defensin-Z.
J Immunol 171: 6820-6826, 2003.

69. Heyninck K, and Beyaert R. The cytokine-inducible zinc ﬁnger protein A20
inhibits IL-l-induced NF-kappaB activation at the level of TRAF6. FEBS Lett 442: 147-
150, 1999.

70. Heyninck K, De Valck D, Vanden Berghe W, Van Criekinge W, Contreras R,
Fiers W, Haegeman G, and Beyaert R. The zinc ﬁnger protein A20 inhibits TNF-
induced NF-kappaB-dependent gene expression by interfering with an RIP- or TRAF2-
mediated transactivation signal and directly binds to a novel NF-kappaB-inhibiting
protein ABIN. J Cell Biol 145: 1471-1482, 1999.

71. Hirschfeld M, Ma Y, Weis JH, Vogel SN, and Weis JJ. Cutting edge:
repuriﬁcation of lipopolysaccharide eliminates signaling through both human and murine
toll-like receptor 2. J Immunol 165: 618-622, 2000.

72. Hoebe K, Du X, George] P, Janssen E, Tabeta K, Kim SO, Goode J, Lin P,
Mann N, Mudd S, Crozat K, Sovath S, Han J, and Beutler B. Identiﬁcation of Lps2
as a key transducer of MyD88-independent TIR signalling. Nature 424: 743-748, 2003.

73. Holcombe SJ, Jackson C, Gerber V, Jefcoat A, Berney C, Eberhardt S, and
Robinson NE. Stabling is associated with airway inﬂammation in young Arabian horses.
Equine Vet J 33: 244-249, 2001.

74. Holcombe SJ, Robinson NE, Derksen FJ, Bertold B, Genovese R, Miller R,
de Feiter Rupp H, Carr EA, Eberhart SW, Boruta D, and Kaneene JB. Effect of

tracheal mucus and tracheal cytology on racing performance in Thoroughbred racehorses.
Equine Vet J38: 300-304, 2006.

75. Holgate ST. The role of mast cells and basophils in inﬂammation. Clin Exp
Allergy 30 Suppl 1: 28-32, 2000.

145

76. Hollander A, Heederik D, and Kauffman H. Acute respiratory effects in the

potato processing industry due to a bioaerosol exposure. Occup Environ Med 51: 73-78,
1994.

77. Horohov DW, Beadle RE, Mouch S, and Pourciau SS. Temporal regulation of
cytokine mRNA expression in equine recurrent airway obstruction. Vet Immunol
Immunopathol 108: 237-245, 2005.

78. Hoshino K, Takeuchi O, Kawai T, Sanjo H, Ogawa T, Takeda Y, Takeda K,
and Akira S. Cutting edge: Toll-like receptor 4 (TLR4)-deﬁcient mice are
hyporesponsive to lipopolysaccharide: evidence for TLR4 as the Lps gene product. J
Immunol 162: 3749-3752, 1999.

79. Housley DJ, Zalewski ZA, Beckett SE, and Venta PJ. Design factors that
inﬂuence PCR ampliﬁcation success of cross-species primers among 1147 mammalian
primer pairs. BMC Genomics 7: 253, 2006.

80. Iigo Y, Suematsu M, Higashida T, Oheda J, Matsumoto K, Wakabayashi Y,
Ishimura Y, Miyasaka M, and Takashi T. Constitutive expression of ICAM-1 in rat
microvascular systems analyzed by laser confocal microscopy. Am J Physiol 273: H138-

147, 1997.

81. Imasato A, Desbois-Mouthon C, Han J, Kai H, Cato AC, Akira S, and Li JD.
Inhibition of p38 MAPK by glucocorticoids via induction of MAPK phosphatase-1
enhances nontypeable Haemophilus inﬂuenzae-induced expression of toll-like receptor 2.
J Biol Chem 277: 47444-47450, 2002.

82. Jaattela M, Mouritzen H, Elling F, and Bastholm L. A20 zinc ﬁnger protein
inhibits TNF and IL-1 signaling. J Immunol 156: 1166-1173, 1996.

83. John M, Au BT, Jose PJ, Lim S, Saunders M, Barnes PJ, Mitchell JA, Belvisi
MG, and Chung KF. Expression and release of interleukin-8 by human airway smooth
muscle cells: inhibition by Th-2 cytokines and corticosteroids. Am J Respir Cell Mol Biol
18: 84-90, 1998.

84. Kaisho T, and Akira S. Dendritic-cell function in Toll-like receptor- and
MyD88-knockout mice. Trends Immunol 22: 78-83, 2001.

85. Karakurum M, Shreeniwas R, Chen J, Pinsky D, Yan SD, Anderson M,
Sunouchi K, Major J, Hamilton T, Kuwabara K, and et al. Hypoxic induction of

interleukin-8 gene expression in human endothelial cells. J Clin Invest 93: 1564-1570,
1994.

86. Kindler S, Wang H, Richter D, and Tiedge H. RNA transport and local control
of translation. Annu Rev Cell Dev Biol 21: 223-245, 2005.

146

 

87. Kinjyo I, Hanada T, Inagaki-Ohara K, Mori H, Aki D, Ohishi M, Yoshida H,
Kubo M, and Yoshimura A. SOCSl/JAB is a negative regulator of LPS-induced
macrophage activation. Immunity 17: 583-591, 2002.

88. Klein CL, Bittinger F, Kohler H, Wagner M, Otto M, Hermanns I, and
Kirkpatrick CJ. Comparative studies on vascular endothelium in vitro. 3. Effects of
cytokines on the expression of E-selectin, ICAM-1 and VCAM-l by cultured human
endothelial cells obtained from different passages. Pathobiology 63: 83-92, 1995.

89. Klugewitz K, Ley K, Schuppan D, Nuck R, Gaehtgens P, and Walzog B.
Activation of the beta2 integrin Mac-1 (CD1 1b/CD18) by an endogenous lipid mediator
of human neutrophils and HL-60 cells. J Cell Sci 110 (Pt 8): 985-990, 1997.

90. Knapp S, Wieland CW, van 't Veer C, Takeuchi O, Akira S, Florquin S, and
van der PollvT. Toll-like receptor 2 plays a role in the early inﬂammatory response to
murine pneumococcal pneumonia but does not contribute to antibacterial defense. J

Immunol 172: 3132-3138, 2004.

91. Kobayashi K, Hernandez LD, Galan JE, Janeway CA, Jr., Medzhitov R, and
Flavell RA. IRAK-M is a negative regulator of Toll-like receptor signaling. Cell 110:
191-202, 2002.

92. Krikos A, Laherty CD, and Dixit VM. Transcriptional activation of the tumor
necrosis factor alpha-inducible zinc ﬁnger protein, A20, is mediated by kappa B
elements. J Biol Chem 267: 17971-17976, 1992.

93. Kruse-Elliott KT, Albert DH, Summers JB, Carter GW, Zimmerman JJ, and
GrosSman JE. Attenuation of endotoxin-induced pathophysiology by a new potent PAF
receptor antagonist. Shock 5: 265-273, 1996.

94. Kullman GJ, Thorne PS, Waldron PF, Marx JJ, Ault B, Lewis DM, Siege]
PD, Olenchock SA, and Merchant JA. Organic dust exposures from work in dairy
barns. Am Ind Hyg Assoc J 59: 403-413, 1998.

95. Laan TT, Bull S, Pirie R, and Fink-Gremmels J. The role of alveolar
macrophages in the pathogenesis of recurrent airway obstruction in horses. J Vet Intern
Med 20: 167-174, 2006.

96. Laan TT, Bull S, Pirie RS, and Fink-Gremmels J. The anti-inﬂammatory
effects of IV administered clenbuterol in horses with recurrent airway obstruction. Vet J
171: 429-437, 2006.

97. Laan TT, Bull S, Pirie RS, and Fink-Gremmels J. Evaluation of cytokine

production by equine alveolar macrophages exposed to lipopolysaccharide, Aspergillus
fumigatus, and a suspension of hay dust. Am J Vet Res 66: 1584-1589, 2005.

147

 

98. Larson VL, and Busch RH. Equine tracheobronchial lavage: comparison of
lavage cytologic and pulmonary histopathologic ﬁndings. Am J Vet Res 46: 144-146,
1985.

99. Lavoie JP, Maghni K, Desnoyers M, Taha R, Martin JG, and Hamid QA.
Neutrophilic airway inﬂammation in horses with heaves is characterized by a Th2-type
cytokine proﬁle. Am J Respir Crit Care Med 164: 1410-1413, 2001.

100. Lee EG, Boone DL, Chai S, Libby SL, Chien M, Lodolce JP, and Ma A.
Failure to regulate TNF-induced NF-kappaB and cell death responses in A20-deﬁcient
mice. Science 289: 2350-2354, 2000.

101. Li X, Abdi K, Rawn J, Mackay CR, and Mentzer SJ. LFA-l and L-selectin
regulation of recirculating lymphocyte tethering and rolling on lung microvascular
endothelium. Am J Respir Cell Mol Biol 14: 398-406, 1996.

102. Li XC, Miyasaka M, and Issekutz TB. Blood monocyte migration to acute lung
inﬂammation involves both CD1 l/CD18 and very late activation antigen-4-dependent
and independent pathways. J Immunol 161: 6258-6264, 1998.

103. Lindberg A, Robinson NE, Nasman-Glaser B, Jensen-Waern M, and
Lindgren JA. Assessment of leukotriene B4 production in leukocytes from horses with
recurrent airway obstruction. Am J Vet Res 65: 289-295, 2004.

104. Lindsay GK, Roslansky PF, and Novitsky TJ. Single-step, Chromogenic
Limulus amebocyte lysate assay for endotoxin. J Clin Microbiol 27: 947-951, 1989.

105. Liu YC, Penninger J, and Karin M. Immunity by ubiquitylation: a reversible
process of modiﬁcation. Nat Rev Immunol 5: 941-952, 2005.

106. MacKay RJ, King R, Dankert JR, Reis KJ, and Skelley LA. Cytotoxic tumor
necrosis factor activity produced by equine alveolar macrophages: preliminary
characterization. Vet Immunol Immunopathol 29: 15-30, 1991.

107. Malik AB, and Lo SK. Vascular endothelial adhesion molecules and tissue
inﬂammation. Pharmacol Rev 48: 213-229, 1996.

108. Mambula SS, Sau K, Henneke P, Golenbock DT, and Levitz SM. Toll-like
receptor (TLR) signaling in response to Aspergillus fumigatus. J Biol Chem 277: 39320-
39326, 2002.

109. Manabe H, Akuta K, Okamura K, and Ohmori K. KF 19514, a

phosphodieterase 4 and 1 inhibitor, inhibits PAF-induced lung inﬂammatory responses
by inhaled administration in guinea pigs. Int Arch Allergy Immunol 114: 389-399, 1997.

148

110. Marr KA, Fairbairn SM, Page CP, Lees P, and Cunningham FM. A study of
the effect of a platelet activating factor (PAF) receptor antagonist on antigen challenge of

horses with chronic obstructive pulmonary disease. J Vet Pharmacol T her 19: 233-237,
1996.

111. Marr KA, Lees P, and Cunningham FM. Agonist-induced adherence of equine
neutrophils to ﬁbronectin- and serum-coated plastic is CD18 dependent. Vet Immunol
Immunopathol 71: 77-88, 1999.

112. Marti E, Gerber H, Essich G, Oulehla J, and Lazary S. The genetic basis of
equine allergic diseases. 1. Chronic hypersensitivity bronchitis. Equine Vet J 23: 457-
460, 1991.

113. Mattoli S, Marini M, and Fasoli A. Expression of the potent inﬂammatory
cytokines, GM-CSF, 1L6, and IL8, in bronchial epithelial cells of asthmatic patients.
Chest 101: 27S-29S, 1992.

114. McEver RP, and Cummings RD. Role of PSGL- 1 binding to selectins in
leukocyte recruitment. J Clin Invest 100: S97- 103,1997.

115. McGorum BC, Dixon PM, and Halliwell RE. Evaluation of intradermal mould
antigen testing in the diagnosis of equine chronic obstructive pulmonary disease. Equine
Vet J 25: 273-275, 1993.

116. McGorum BC, Dixon PM, and Halliwell RE. Responses of horses affected with
chronic obstructive pulmonary disease to inhalation challenges with mould antigens.
Equine VetJ 25: 261-267, 1993.

117. McGorum BC, Ellison J, and Cullen RT. Total and respirable airborne dust
endotoxin concentrations in three equine management systems. Equine Vet J 30: 430-
434, 1998.

118. McPherson EA, Lawson GH, Murphy JR, Nicholson JM, Fraser JA, Breeze
RG, and Pirie HM. Chronic obstructive pulmonary disease (COPD): identiﬁcation of
affected horses. Equine VetJ 10: 47-53, 1978.

119. Menzies-Gow A, and Robinson DS. Eosinophils, eosinophilic cytokines
(interleukin-5), and antieosinophilic therapy in asthma. Curr Opin Pulm Med 8: 33-38,
2002.

120. Milton DK, Amsel J, Reed CE, Enright PL, Brown LR, Aughenbaugh GL,
and Morey PR. Cross-sectional follow-up of a ﬂu-like respiratory illness among

ﬁberglass manufacturing employees: endotoxin exposure associated with two distinct
sequelae. Am J Ind Med 28: 469-488, 1995.

149

121. Milton DK, Walters MD, Hammond K, and Evans JS. Worker exposure to
endotoxin, phenolic compounds, and formaldehyde in a ﬁberglass insulation
manufacturing plant. Am Ind Hyg Assoc J57: 889-896, 1996.

122. Milton DK, Wypij D, Kriebel D, Walters MD, Hammond SK, and Evans JS.
Endotoxin exposure-response in a ﬁberglass manufacturing facility. Am J Ind Med 29: 3-
13, 1996.

123. Monick MM, Yarovinsky TO, Powers LS, Butler NS, Carter AB,
Gudmundsson G, and Hunninghake GW. Respiratory syncytial virus up-regulates
TLR4 and sensitizes airway epithelial cells to endotoxin. J Biol Chem 278: 53035-53044,
2003.

124. Morath S, Stadelmaier A, Geyer A, Schmidt RR, and Hartung T. Synthetic
lipoteichoic acid from Staphylococcus aureus is a potent stimulus of cytokine release. J
Exp Med 195: 1635-1640, 2002.

125. Morris GE, Whyte MK, Martin GF, Jose PJ, Dower SK, and Sabroe I.
Agonists of toll-like receptors 2 and 4 activate airway smooth muscle via mononuclear
leukocytes. Am J Respir Crit Care Med 171: 814-822, 2005.

126. Murthy KG, Deb A, Goonesekera S, Szabo C, and Salzman AL. Identiﬁcation
of conserved domains in Salmonella muenchen ﬂagellin that are essential for its ability to
activate TLR5 and to induce an inﬂammatory response in vitro. J Biol Chem 279: 5667-

5675, 2004.

127. Muzio M, Bosisio D, Polentarutti N, D'Amico G, Stoppacciaro A, Mancinelli
R, van't Veer C, Penton-Rol G, Ruco LP, Allavena P, and Mantovani A. Differential
expression and regulation of toll-like receptors (TLR) in human leukocytes: selective
expression of TLR3 in dendritic cells. J Immunol 164: 5998-6004, 2000.

128. Nardoni S, Mancianti F, Sgorbini M, Taceini F, and Corazza M. Identiﬁcation
and seasonal distribution of airborne fungi in three horse stables in Italy. Mycopathologia
160: 29-34, 2005.

129. Nevalainen M, Raulo SM, Brazil TJ, Pirie RS, Sorsa T, McGorum BC, and
Maisi P. Inhalation of organic dusts and lipopolysaccharide increases gelatinolytic matrix
metalloproteinases (MMPs) in the lungs of heaves horses. Equine Vet J 34: 150-155,
2002.

130. Nocker RE, Schoonbrood DF, van de Graaf EA, Hack CE, Lutter R, Jansen

HM, and Out TA. Interleukin-8 in airway inﬂammation in patients with asthma and
chronic obstructive pulmonary disease. Int Arch Allergy Immunol 109: 183-191, 1996.

150

131. Nohgawa M, Sasada M, Maeda A, Asagoe K, Harakawa N, Takano K,
Yamamoto K, and Okuma M. Leukotriene B4-activated human endothelial cells
promote transendothelial neutrophil migration. J Leukoc Biol 62: 203 -209, 1997.

132. O'Neill LA, Fitzgerald KA, and Bowie AG. The Toll-IL-l receptor adaptor
family grows to ﬁve members. Trends Immunol 24: 286-290, 2003.

133. O'Reilly SM, and Moynagh PN. Regulation of Toll-like receptor 4 signalling by
A20 zinc ﬁnger protein. Biochem Biophys Res Commun 303: 586-593, 2003.

134. Onose A, Hashimoto S, Hayashi S, Maruoka S, Kumasawa F, Mizumura K,
Jibiki I, Matsumoto K, Gon Y, Kobayashi T, T akahashi N, Shibata Y, Abiko Y,
Shibata T, Shimizu K, and Horie T. An inhibitory effect of A20 on NF-kappaB
activation in airway epithelium upon inﬂuenza virus infection. Eur J Pharmacol 541:
198-204, 2006.

135. Opipari AW, Jr., Boguski MS, and Dixit VM. The A20 cDNA induced by
tumor necrosis factor alpha encodes a novel type of zinc ﬁnger protein. J Biol Chem 265:
14705-14708, 1990.

136. Pang L, and Knox AJ. Bradykinin stimulates IL-8 production in cultured human
airway smooth muscle cells: role of cyclooxygenase products. J Immunol 161: 2509-
2515, 1998.

137. Pang L, and Knox AJ. Synergistic inhibition by beta(2)-agonists and
corticosteroids on tumor necrosis factor-alpha-induced interleukin-8 release from
cultured human airway smooth-muscle cells. Am J Respir Cell Mol Biol 23: 79-85, 2000.

138. Park JS, Svetkauskaite D, He Q, Kim JY, Strassheim D, Ishizaka A, and
Abraham E. Involvement of toll-like receptors 2 and 4 in cellular activation by high
mobility group box 1 protein. J Biol Chem 279: 7370-7377, 2004.

139. Pease JE, and Sabroe I. The role of interleukin-8 and its receptors in
inﬂammatory lung disease: implications for therapy. Am J Respir Med 1: 19-25, 2002.

140. Pirie RS, Collie DD, Dixon PM, and McGorum BC. Evaluation of nebulised
hay dust suspensions (HDS) for the diagnosis and investigation of heaves. 2: Effects of
inhaled HDS on control and heaves horses. Equine Vet J 34: 337-342, 2002.

141. Pirie RS, Collie DD, Dixon PM, and McGorum BC. Inhaled endotoxin and
organic dust particulates have synergistic proinﬂammatory effects in equine heaves
(organic dust-induced asthma). Clin Exp Allergy 33: 676-683, 2003.

142. Pirie RS, Dixon PM, Collie DD, and McGorum BC. Pulmonary and systemic

effects of inhaled endotoxin in control and heaves horses. Equine Vet J 33: 311-318,
2001.

151

 

143. Pirie RS, Dixon PM, and McGorum BC. Endotoxin contamination contributes

to the pulmonary inﬂammatory and functional response to Aspergillus fumigatus extract
inhalation in heaves horses. Clin Exp Allergy 33: 1289-1296, 2003.

144. Pirie RS, Dixon PM, and McGorum BC. Evaluation of nebulised hay dust
suspensions (HDS) for the diagnosis and investigation of heaves. 3: Effect of
fractionation of HDS. Equine VetJ 34: 343-347, 2002.

145. Pirie RS, McLachlan G, and McGorum BC. Evaluation of nebulised hay dust
suspensions (HDS) for the diagnosis and investigation of heaves. 1: Preparation and
composition of HDS. Equine Vet J 34: 332-336, 2002.

146. Poltorak A, He X, Smirnova 1, Liu MY, Van Huffel C, Du X, Birdwell D,
Alejos E, Silva M, Galanos C, Freudenberg M, Ricciardi-Castagnoli P, Layton B,
and Beutler B. Defective LPS signaling in C3H/HeJ and C57BL/IOScCr mice:
mutations in Tlr4 gene. Science 282: 2085-2088, 1998.

147. Qureshi ST, Lariviere L, Leveque G, Clermont S, Moore KJ, Gros P, and
Malo D. Endotoxin-tolerant mice have mutations in Toll-like receptor 4 (Tlr4). J Exp
Med 189: 615-625, 1999.

148. Rhee SH, Im E, Riegler M, Kokkotou E, O'Brien M, and Pothoulakis C.
Pathophysiological role of Toll-like receptor 5 engagement by bacterial ﬂagellin in
colonic inﬂammation. Proc NatlAcad Sci USA 102: 13610-13615, 2005.

149. Rickards KJ, Page CP, Lees P, and Cunningham FM. Differential inhibition
of equine neutrophil function by phosphodiesterase inhibitors. J Vet Pharmacol Ther 24:
275-281, 2001.

150. Robins T, Seixas N, Franzblau A, Abrams L, Minick S, Burge H, and Schork
MA. Acute respiratory effects on workers exposed to metalworking ﬂuid aerosols in an
automotive transmission plant. Am J Ind Med 31: 510-524, 1997.

151. Robinson NE. International Workshop on Equine Chronic Airway Disease.
Michigan State University 16-18 June 2000. Equine VetJ 33: 5-19, 2001.

152. Robinson NE, Karmaus W, Holcombe SJ, Carr EA, and Derksen FJ. Airway
inﬂammation in Michigan pleasure horses: prevalence and risk factors. Equine Vet J 38:
293-299, 2006.

153. Robinson NE, Olszewski MA, Boehler D, Berney C, Hakala J, Matson C, and
Derksen FJ. Relationship between clinical signs and lung function in horses with
recurrent airway obstruction (heaves) during a bronchodilator trial. Equine Vet J 32: 393-

400, 2000.

152

 

154. Rodgers HC, Pang L, Holland E, Corbett L, Range S, and Knox AJ.
Bradykinin increases IL-8 generation in airway epithelial cells via COX-2-derived
prostanoids. Am J Physiol Lung Cell Mol Physiol 283: L612-618, 2002.

155. Rosa GJ, de Leon N, and Rosa AJ. Review of microarray experimental design
strategies for genetical genomics studies. Physiol Genomics 28: 15-23, 2006.

156. Ross J. mRNA stability in mammalian cells. Microbiol Rev 59: 423-450, 1995.

157. Rush BR, Raub ES, Rhoads WS, Flaminio MJ, Matson CJ, Hakala JE, and
Gillespie JR. Pulmonary function in horses with recurrent airway obstruction after
aerosol and parenteral administration of beclomethasone dipropionate and

dexamethasone, respectively. Am J Vet Res 59: 1039-1043, 1998.

158. Rylander R, Thorn J, and Attefors R. Airways inﬂammation among workers in
a paper industry. Eur Respir J 13: 1151-1157, 1999.

159. Sabroe 1, Parker LC, Wilson AG, Whyte MK, and Dower SK. Toll—like
receptors: their role in allergy and non-allergic inﬂammatory disease. Clin Exp Allergy
32: 984-989, 2002.

160. Sampson AP. The role of eosinophils and neutrophils in inﬂammation. Clin Exp
Allergy 30 Suppl 1: 22-27, 2000.

161. Sandersen C, Bureau F, Turlej R, Fievez L, Dogne S, Kirschvink N, and
Lekeux P. p65 Homodimer activity in distal airway cells determines lung dysﬁmction in
equine heaves. Vet Immunol Immunopathol 80: 315-326, 2001.

162. Schmallenbach KH, Rahman I, Sasse HH, Dixon PM, Halliwell RE,
McGorum BC, Crameri R, and Miller HR. Studies on pulmonary and systemic
Aspergillus fumigatus-speciﬁc IgE and IgG antibodies in horses affected with chronic
obstructive pulmonary disease (COPD). Vet Immunol Immunopathol 66: 245-256, 1998.

163. Scholz D, Devaux B, Hirche A, Potzsch B, Kropp B, Schaper W, and Schaper
J. Expression of adhesion molecules is speciﬁc and time-dependent in cytokine-
stimulated endothelial cells in culture. Cell Tissue Res 284: 415-423, 1996.

164. Schwartz DA, Thorne PS, Yagla SJ, Burmeister LF, Olenchock SA, Watt JL,
and Quinn TJ. The role of endotoxin in grain dust-induced lung disease. Am J Respir

Crit Care Med 152: 603-608, 1995.

165. Sen S, and Churchill GA. A statistical framework for quantitative trait mapping.
Genetics 159: 371-387, 2001.

166. Shibata F. [The role of rat cytokine-induced neutrophil Chemoattractants (CINCS)
in inﬂammation]. Yakugaku Zasshi 122: 263-268, 2002.

153

167. Shuto T, Imasato A, Jono H, Sakai A, Xu H, Watanabe T, Rixter DD, Kai H,
Andalibi A, Linthicum F, Guan YL, Han J, Cato AC, Lim DJ, Akira S, and Li JD.
Glucocorticoids synergistically enhance nontypeable Haemophilus inﬂuenzae-induced
Toll-like receptor 2 expression via a negative cross-talk with p38 MAP kinase. J Biol
Chem 277: 17263-17270, 2002.

168. Silverstein AM. History of immunology. Cellular versus humoral immunity:
determinants and consequences of an epic 19th century battle. Cell Immunol 48: 208-221,
1979.

169. Simonen-Jokinen T, Pirie RS, McGorum B, and Maisi P. Dose responses to
inhalation of endotoxin, hay dust suspension and Aspergillus fumigatus extract in horses
as measured by levels and activation of matrix metalloproteinase-9. Equine Vet J 37: 155-
160, 2005.

170. Singh Suri S, Janardhan KS, Parbhakar O, Caldwell S, Appleyard G, and
Singh B. Expression of Toll-like receptor 4 and 2 in horse lungs. Vet Res 37: 541-551,
2006.

171. Smid T, Heederik D, Houba R, and Quanjer PH. Dust- and endotoxin-related
acute lung function changes and work-related symptoms in workers in the animal feed
industry. Am J Ind Med 25: 877-888, 1994.

172. Smid T, Heederik D, Houba R, and Quanjer PH. Dust- and endotoxin-related
respiratory effects in the animal feed industry. Am Rev Respir Dis 146: 1474-1479, 1992.

173. Smiley ST, King JA, and Hancock WW. Fibrinogen stimulates macrophage
chemokine secretion through toll-like receptor 4. J Immunol 167: 2887-2894, 2001.

174. Smith KD, Andersen-Nissen E, Hayashi F, Strobe K, Bergman MA, Barrett
SL, Cookson BT, and Aderem A. Toll-like receptor 5 recognizes a conserved site on
ﬂagellin required for protoﬁlament formation and bacterial motility. Nat Immunol 4:
1247-1253, 2003.

175. Soong G, Reddy B, Sokol S, Adamo R, and Prince A. TLR2 is mobilized into
an apical lipid raft receptor complex to signal infection in airway epithelial cells. J Clin
Invest 113: 1482-1489, 2004.

176. Spertini O, Luscinskas FW, Kansas GS, Munro JM, Grifﬁn JD, Gimbrone
MA, J r., and Tedder TF. Leukocyte adhesion molecule-1 (LAM-1, L-selectin) interacts
with an inducible endothelial cell ligand to support leukocyte adhesion. J Immunol 147:
2565-2573, 199].

154

177. Stijkel A, van Eijndhoven JC, and Bal R. Drafting guidelines for occupational
exposure to chemicals: the Dutch experience with the assessment of reproductive risks.
Am JInd Med 30: 705- 717, 1996.

178. Strunk RC, Eidlen DM, and Mason RJ. Pulmonary alveolar type II epithelial
cells synthesize and secrete proteins of the classical and alternative complement
pathways. JClin Invest 81: 1419-1426, 1988.

179. Takizawa H. Bronchial epithelial cells in allergic reactions. Curr Drug Targets
Inﬂamm Allergy 4: 305-311, 2005.

180. Tamura H, Arimoto Y, Tanaka S, Yoshida M, Obayashi T, and Kawai T.
Automated kinetic assay for endotoxin and (1-->3)-beta-D-glucan in human blood. Clin
Chim Acta 226: 109-112, 1994.

181. Taylor KR, Trowbridge JM, Rudisill JA, Termeer CC, Simon JC, and Gallo
RL. Hyaluronan fragments stimulate endothelial recognition of injury through TLR4. J
Biol Chem 279: 17079-17084, 2004.

182. Teeuw KB, Vandenbroucke-Grauls CM, and Verhoef J. Airborne gram-
negative bacteria and endotoxin in sick building syndrome. A study in Dutch
governmental ofﬁce buildings. Arch Intern Med 154: 2339-2345, 1994.

183. Termeer C, Benedix F, Sleeman J, Fieber C, Voith U, Ahrens T, Miyake K,
Freudenberg M, Galanos C, and Simon JC. Oligosaccharides of Hyaluronan activate
dendritic cells via toll-like receptor 4. J Exp Med 195: 99-111, 2002.

184. Tremblay GM, Ferland C, Lapointe JM, Vrins A, Lavoie JP, and Cormier Y.
Effect of stabling on bronchoalveolar cells obtained from normal and COPD horses.
Equine VetJ 25: 194-197, 1993.

185. Turlej RK, Fievez L, Sandersen CF, Dogne S, Kirschvink N, Lekeux P, and
Bureau F. Enhanced survival of lung granulocytes in an animal model of asthma:
evidence for a role of GM-CSF activated STATS signalling pathway. Thorax 56: 696-
702,200].

186. van Duin D, Mohanty S, Thomas V, Ginter S, Montgomery RR, Fikrig E,
Allore HG, Medzhitov R, and Shaw AC. Age-associated defect in human TLR-l/2
function. J Immunol 178: 970-975, 2007.

187. Van Huffel S, Delaei F, Heyninck K, De Valck D, and Beyaert R.

Identiﬁcation of a novel A20-binding inhibitor of nuclear factor-kappa B activation
termed ABIN-2. J Biol Chem 276: 30216-30223, 2001.

155

188. Visintin A, Mazzoni A, Spitzer JH, Wyllie DH, Dower SK, and Segal DM.
Regulation of Toll-like receptors in human monocytes and dendritic cells. J Immunol
166: 249-255, 2001.

189. Wagner JG, and Roth RA. Neutrophil migration during endotoxemia. J Leukoc
Biol 66: 10-24, 1999.

190. Wagner JG, and Roth RA. Neutrophil migration mechanisms, with an emphasis
on the pulmonary vasculature. Pharmacol Rev 52: 349-374, 2000.

191. Walters M, Milton D, Larsson L, and Ford T. Airborne environmental
endotoxin: a cross-validation of sampling and analysis techniques. Appl Environ
Microbiol 60: 996-1005, 1994.

192. Wassef A, Janardhan K, Pearce JW, and Singh B. Toll-like receptor 4 in
normal and inﬂamed lungs and other organs of pig, dog and cattle. Histol Histopathol 19:
1201-1208, 2004.

193. Werners AH, Bull S, Vendrig JC, Smyth T, Bosch RR, Fink-Gremmels J,
and Bryant CE. Genotyping of Toll-like receptor 4, myeloid differentiation factor 2 and
CD-14 in the horse: an investigation into the inﬂuence of genetic polymorphisms on the
LPS induced TNF-alpha response in equine whole blood. Vet Immunol Immunopathol
111: 165-173, 2006.

194. Wertz IE, O'Rourke KM, Zhou H, Eby M, Aravind L, Seshagiri S, Wu P,
Wiesmann C, Baker R, Boone DL, Ma A, Koonin EV, and Dixit VM. De-
ubiquitination and ubiquitin ligase domains of A20 downregulate NF-kappaB signalling.
Nature 430: 694-699, 2004.

195. Wolfs TG, Buurman WA, van Schadewijk A, de Vries B, Daemen MA,
Hiemstra PS, and van 't Veer C. In vivo expression of Toll-like receptor 2 and 4 by
renal epithelial cells: IFN-gamma and TNF-alpha mediated up-regulation during
inﬂammation. J Immunol 168: 1286-1293, 2002.

196. Woods PS, Robinson NE, Swanson MC, Reed CE, Broadstone RV, and
Derksen FJ. Airborne dust and aeroallergen concentration in a horse stable under two
different management systems. Equine VetJ 25: 208-213, 1993.

197. Yamamoto M, Sato S, Hemmi H, Hoshino K, Kaisho T, Sanjo H, Takeuchi
O, Sugiyama M, Okabe M, Takeda K, and Akira S. Role of adaptor TRIF in the
MyD88-independent toll-like receptor signaling pathway. Science 301: 640-643, 2003.

198. Yauch LE, Mansour MK, Shoham S, Rottman JB, and Levitz SM.

Involvement of CD14, toll-like receptors 2 and 4, and MyD88 in the host response to the
fungal pathogen Cryptococcus neoformans in vivo. Infect Immun 72: 5373-5382, 2004.

156

199. Zaas AK, and Schwartz DA. Innate immunity and the lung: defense at the
interface between host and environment. Trends Cardiovasc Med 15: 195-202, 2005.

200. Zhang CH, Davis WC, Grunig G, and Antczak DF. The equine homologue of
LFA-l (CD11a/CD18): cellular distribution and differential determinants. Vet Immunol
Immunopathol 62: 167-183, 1998.

201. Zhang G, and Ghosh S. Negative regulation of toll-like receptor-mediated
signaling by Tollip. J Biol Chem 277: 7059-7065, 2002.

202. Zimmerman GA, Elstad MR, Lorant DE, Mantyre TM, Prescott SM,

Topham MK, Weyrich AS, and Whatley RE. Platelet-activating factor (PAF):
signalling and adhesion in cell-cell interactions. Adv Exp Med Biol 416: 297-304, 1996.

157

lliliiiiiiiiililiiillililliii