. .
33w:

.. .
7’:- .w

.-
9 r t8 .5!
l 17

5:55:

. .
ﬂaw“. up ...

V 4'ta‘-~w\.5- -

_~
.

.u. onﬂlgu‘

 

......m.,.‘-. .-»a.:- un- .n. a .‘.-.n

 

...Q.l'» ...,-mn

 

.‘ I .4 w. |.r , v ‘ 25.. £033.34 1H,}

 

NATURE AND CAUSES OF MUCUS ACCUMULATION
IN EQUINE LOWER AIRWAY DISEASE

By

Vincent Gerber

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Department of Large Animal Clinical Sciences

2003

ABSTRACT
NATURE AND CAUSES OF MUCUS ACCUMULATION
IN EQUINE LOWER AIRWAY DISEASE
By

Vincent Gerber

The purpose of my work was to quantify airway mucus accumulation and
investigate its causes in RAG-affected horses and clinically healthy controls. I
hypothesized that mucus accumulation is a function of the interaction of disease status
(RAO), environment (dust exposure) and age of the horse. My investigations showed that
endoscopic scoring is a reliable tool to quantify mucus accumulation. Using endoscopic
scoring to compare RAG-affected and clinically healthy horses, I found very large mucus
accumulations to be limited to RAD-affected horses, especially during acute
exacerbation. There was less mucus accumulation in controls, and I found no difference
between age groups. Variation between individuals was high, however, with many
clinically healthy horses showing Signs of IAD. Overall, mucus accumulation was
associated with neutrophilic airway inﬂammation. I then proposed that increased
accumulation is a consequence of unfavorable rheological changes and/or increased
mucin production. I found that increases of mucus accumulation in RAO exacerbation are
accompanied by a large rise of mucus viscoelasticity, which in turn is at least partly due
to F-actin ﬁbers, likely released from degenerate neutrophils. These rheological changes
could not explain the increased mucus accumulation in RAO remission or the overall

high variability, however. Therefore, I identiﬁed equine homologues of gel-forming

mucin genes. EqMUCSAC, but not eqMUC2 showed robust expression in large and
small airway generations of both RAD-affected and clinically healthy horses. A
preliminary semi-quantitative study on pooled samples suggested that eqMUCSAC
expression is increased in RAO. I identiﬁed equine homologues of hCACCl and EGFR,
key signaling elements in mucin gene regulation, and developed quantitative RT-PCR
assays for eqMUCSAC, equine hCACC 1-like, EGFR and EGF. This allowed me to
accurately determine mRNA levels in individual samples and investigate the association of
eqMUCSAC expression with the expression of these key signaling elements, with stored
intraepithelial mucosubstance and with neutrophilic inﬂammation. My ﬁndings did not
support the hypothesis that eqMUCSAC is up-regulated and intraepithelial
mucosubstances are increased in RAG-affected compared to clinically healthy control
horses. EqMUCSAC, equine hCACCl-like, EGFR or EGF mRNA levels showed
remarkable consistency across different lung lobes within individuals, indicating that the
underlying pathogenesis is a diffuse process. Conversely, differences were considerable
between horses. Equine hCACCl-like and EGFR mRNA levels as well as neutrophil
percentages were associated with eqMUCSAC mRNA levels in both groups. In
conclusion, gross airway mucus accumulation is increased in RAD-affected horses,
particularly during exacerbation. The increase during exacerbation can partly be
explained by unfavorable rheological changes. The roles and pathogenesis of increased
mucin production and mucus cell hyperplasia in RAO remain unresolved. My ﬁndings
indicate, however, that EqMUCSAC, equine hCACCl-like, EGFR and neutrophilic
inﬂammation may be involved in mucus production in the airways of horses suffering

ﬁ'om RAO as well as of horses with milder degrees of airway inﬂammation.

Dedicated to my father, Heinz Gerber

iv

ACKNOWLEDGMENTS

I wish to express my gratitude to my major professor, Dr. N. E. Robinson, the
best mentor one could wish for, and my thanks to my guidance committee, Drs. F.
Derksen, J. Hotchkiss, J. Harkema and H. Schott. Many others have given me crucial
help and advice in speciﬁc areas: Drs. S. Ewart, P. Gehr, J. Hauptman, V. ImHof, M.
King, M. Kiupel, E. Marti, P. Venta, D. Schneider, R. Straub and J. Wagner. Absolutely
central to the success of my thesis work was the help of Cathy Bemey and Heather
deFeijter-Rupp. I would also like to acknowledge the assistance of D. Aebischer, L.
Bartner, G. Bleuer, D. Boruta, E. Curat, U. Du Pasquier, S. Eberhart, B. Feller, A. Jefcoat,
J. Klumpp, A. Lindberg, M. Ricklin , J. Rawson and F. Uehlinger. As an “outsider”, my
brother Manuel Gerber has given me very helpful advice on many manuscripts, and the
help of Victoria Hoelzer-Maddox has been invaluable preparing manuscripts for
submission and in meeting the laughably excessive and largely pointless formatting
requirements of this thesis. Faith Peterson has helped me navigate graduate school
throughout my PhD candidate years. My ﬁancée Shannon Axiak deserves my thanks for
her understanding and support when I wasn’t always the most pleasant to be with during
all that work. I also thank my mother for her untiring help and support across the
Atlantic. Finally, I would like to acknowledge the funding agencies of USDA Animal
Health and Well-Being; the Matilda Wilson Endowment, 3M, USA and Novartis,
Switzerland that have helped ﬁnd this research. I am especially indebted to the Swiss
National Science Foundation for the fellowship that allowed me to come to Michigan State

University ﬁve years ago, and thus started it all.

TABLE OF CONTENTS

LIST OF FIGURES ................................................................................. v

INTRODUCTION .................................................................................... 1

CHAPTER 1 -

LITERATURE REVIEW ............................................................................ 4
Nature of airway mucous secretions .................................................... 4
Description, definitions and significance of RAO and [AD ........................ 5

Description and deﬁnitions ............................................... 5
Clinical signiﬁcance ....................................................... 6
Clinical and pathophysiological significance of mucus accumulations
- Is the phenomenon worth studying? ................................................. 8
Comparative aspects ...................................................... 8
Significance of mucus accumulation in equine
lower airway diseases ..................................................... 9
Causes and mechanisms of mucus accumulation .................................. 11
Clearance - primarily compromised or just overwhelmed? ...... 12
Ciliary function and dysfunction .............................. 14
Mucus biophysical properties and their alterations
in disease ......................................................... 15
Production and secretion ............................................... l9
Mucins — production and secretion.» .......................... 21
Origin and mechanisms of mucin hypersecretion ........... 24
Pathophysiological & immunological context of mucus
h ypersecretion ............................................................ 25
Innate immunity .................................................. 25

The "yin-yang” "T helper (h)2 — Th] hypothesis............26
Most cytokines play pleiotropic roles in

down-stream effector pathways ................................ 27
Pathophysiological and immunological mechanisms
in equine lower airway disease ......................................... 28
Conclusions of literature review and derived hypotheses ........................ 31

CHAPTER 2
ENDOSCOPIC SCORING OF MUCUS QUANTITY AND QUALITY - OBSERVER
AND HORSE RELATED VARIANCE AND RELATIONSHIP WITH MEASURED

VARIABLES ........................................................................................ 34
Abstract ...................................................................................... 35
Introduction .................................................................................. 36
Materials and methods ..................................................................... 37
Results ........................................................................................ 42
Discussion ................................................................................... 43

vi

CHAPTER 3
AIRWAY MUCUS ACCUMULATION — COMPARISON OF RAD-AFFECTED
AND CLINICALLY HEALTHY CONTROL HORSES BEFORE AND DURING

ENVIRONMENTAL CHALLENGE ....................... g ..................................... 54
Abstract ...................................................................................... 55
Introduction .................................................................................. 56
Materials and methods ..................................................................... 57
Results ........................................................................................ 59
Discussion ................................................................................... 61

CHAPTER 4

COMPARISON OF AIRWAY MUCUS ACCUMULATION IN TWO AGE GROUPS
OF CLINICALLY HEALTHY WELL-PERFORMING SPORT HORSES IN A

CONTROLLED ENVIRONMENT ............................................................... 68
Abstract ...................................................................................... 69
Introduction .................................................................................. 70
Materials and methods ..................................................................... 71
Results ........................................................................................ 74
Discussion ................................................................................... 75

CHAPTER 5

MUCUS VISCOELASTICITY AND PREDICTED CLEARABILITY IN RAO-
AFFECTED AND CLINICALLY HEALTHY CONTROL HORSES BEFORE AND

DURING ENVIRONMENTAL CHALLENGE ................................................ 84
Abstract ......................... 85
Introduction .................................................................................. 86
Materials and methods ..................................................................... 88
Results ........................................................................................ 92
Discussion ................................................................................... 96

CHAPTER 6

PRELIMINARY STUDY ON THE MUCOLYTIC EFFECTS OF DNase AND
GELSOLIN ON EQUINE MUCUS RHEOLOGY — F ~ACTIN MAY CONTRINUTE TO

INCREASED VISCOELASTICITY DURING RAO EXACERBATION ................ 104
Abstract ..................................................................................... 105
Introduction ................................................................................ 106
Materials and methods .................................................................... 107
Results ....................................................................................... 109

CHAPTER 7

IDENTIFICATION OF EQUINE HOMOLOGUES OF GEL-FORMING MUCIN
GENES AND SEMI-QUANTITATIVE MEASUREMENT OF OF eqMUCSAC AND
eqMUC2 mRNA LEVELS IN POOLED SAMPLESA FROM DIFFERENT AIRWAY
GENERATIONS OF RAO-AF F ECTED AND HEALTHY HORSES .................... 115

vii

Abstract ..................................................................................... 116

Introduction ................................................................................ 1 17

Materials and methods .................................................................... 118

Results ....................................................................................... 122

Discussion .............................................. ' .................................... 1 24
CHAPTER 8

MUCSAC mRNA LEVELS, BUT NOT INTRAEPITHELIAL MUCOSUBSTANCE,
ARE ASSOCIATED WITH EQUINE CACCl-LIKE AND EGFR mRNA LEVELS
AND WITH INTRALUMINAL NEUTROPHILS IN SMALL CARTILAGINOUS

AIRWAYS OF RAO — AFFECTED AND CONTROL HORSES .......................... 133
Abstract ..................................................................................... 134
Introduction ................................................................................ 135
Materials and methods .................................................................... 137
Results ....................................................................................... 145
Discussion .................................................................................. 147

CHAPTER 10

SUMMARY, CONCLUSIONS AND OUTLOOK ........................................... 160

Background .............................................................................. 160
Hypotheses .............................................................................. 161
Airway mucus accumulation ........................................................... 162
Role of mucus rheology ................................................................ 165
Role of mucin production .............................................................. 166
Interpretation and outlook...................................... ........................ 169
BIBLIOGRAPHY .............................................................................. 173-88

viii

LIST OF TABLES

Table 1-1:
Characteristics of recurrent airway obstruction (RAO) and inﬂammatory airway
disease (IAD) .......................................................... , ................................ 7

Table 2-1:
Variance attributable to observers (020), repetitions (021;), time (ozT), and horse (62“)
for mucus accumulation, localization, apparent viscosity and color scores ................. 46

Table 2-2:
Spearman Rank Order correlation within and between observers ............................ 47

Table 2-3:

Means i SD (0, 6, 12, and 24 hours), ranges (all time points) and overall means i SD

(all time points) of all observers and repetitions for mucus accumulation, localization,
apparent viscosity and color scores for each horse ............................................. 48

Table 3-1:

Mucus grades and bronchoalveolar lavage ﬂuid cytology in RAO-affected horses and
control horses before (0 hours) and during (6, 24, 48 hours) environmental challenge:
medians with quartiles (25"‘, 75th percentiles) for mucus grade (MG), total cell count (tcc;
x cells/pl), percentages of lymphocytes (%lym), macrophages (%mac), neutrophils
(%neu), eosinophils (%eos), mast cells (%mas), and epithelial cells (%epi) ............... 65

Table 4-1:
Individual and group signalment, mucus scores and broncho-alveolar lavage cytology..80

Table 5-1:

Biophysical properties of individual mucus samples: Viscoelasticity (log G"; dyn/cmz)
and tangent 8, measured before (0 hours) and during (6, 24, 48 hours) stabling in each
RAO horse (1-7) and each control (8-14) ....................................................... 101

Table 6-1:

Change of Viscoelasticity (Alog G*, dyn/cmz) without any substance added (sham), after
saline treatment and alter mucolytic treatment with nacystelyn, gelsolin and thNase on

RAO-exacerbation (R), control (C) and cystic ﬁbrosis (CF) mucus samples ............ 113
Table 7-1:

Primer sequences, product size and references ................................................. 128
Table 7-2:

Nucleotide homology of 01S) of eqMUC2 and eqMUCSAC with other mucins. . . . . ....129
Table 7-3:

EqMUCSAC and ZO-l on pooled RAO and pooled control airway samples ............. 130

ix

Table 8-1: ,
Real-time PCR primer sequences (5 ’ to 3’) and predicted product size in basepairs. . . 154

Table 8-2:

Airway secretion cytology, intraepithelial mucosubstance and gene expression in RAO-
affected and control horses: Number of observations, mean values i SD for percentages
(%) of macrophages, neutrophils, lymphocytes, mast cells and eosinophils, volume

density (Vs, n1 / mm2 basal lamina) of intraepithelial mucosubstance (IMS), and relative
mRNA levels of eqMUCSAC, equine hCACCl-like, EGFR and EGF .................... 155

LIST OF FIGURES
Images in this dissertation are presented in color

Fig. 1-1: .
Types of bonds in a mucous gel and respective mucolytic agents ........................... 18

Fig. 1-2:
My illustrated view of some immunological and pathophysiological mechanisms
that may be involved in equine lower airway disease .......................................... 30

Fig. 1-3:
Basic model hypothesis: Mucus accumulation is caused by increased production
and/or decreased clearance, in turn these are consequences of airway inﬂammation. . ....33

Fig. 2-1:
Scoring system for mucus accumulation (A), localization (B), apparent viscosity (C)
and color (D) ......................................................................................... 49

Fig. 2-2:
Intraobserver (observer 1, round 1 and 2) correlations for mucus accumulation (A),
localization (B), apparent viscosity (C) and color (D) scores ................................. 50

Fig. 2-3:

Associations of mucus accumulation scores with tracheo-bronchial secretion (TBS)
neutrophil percentages (A) at 0 hours, and with bronchoalveolar lavage ﬂuid (BALF)
neutrophil percentages (B) and BALF neutrophil absolute numbers (C) at 24 hours. . ....51

Fig. 2-4:
Relationship between mucus accumulation score and volumes of “artiﬁcial mucus”
(Jell-O®') ............................................................................................. 52

Fig. 2-5:

Relationships between mucus accumulation (A), apparent viscosity (B) and

color (C) scores and measured Viscoelasticity (log G“ at 10 radian/s; dyn/cmz) and

trend (P = 0.051) towards a signiﬁcant difference between Viscoelasticity (D) of

mucus samples collected ventrally vs. samples collected dorsally in the trachea at

the same time ........................................................................................ 53

Fig. 3-1:

Boxplots of mucus grades (A) and bronchoalveolar lavage ﬂuid neutrophil numbers (B)
in recurrent airway obstruction (RAO)-affected (grey boxes) and control horses (white
boxes) before and 6, 24, and 48 hours after stabling. *Signiﬁcantly different (P < 0.05)
from 0 hours. #Signiﬁcantly different from control at same time point; +trend toward a
difference compared to control at that time point (P = 0.054) ................................. 66

xi

Fig. 3-2:

Distribution histogram of mucus grades (MGs). Graph shows relative frequency (%) of
each MG (0-5) in controls (clinically healthy horses; white columns) and in recurrent
airway obstruction (RAO)-affected animals (gray columns). Numbers of observations are
listed in the attached table. There was a signiﬁcant difference in MGs between control
and RAO-affected horses (main group effect). MG 0-1 had a good Speciﬁcity (0.92), but
low sensitivity (0.5) for controls. Large amounts of mucus (MG 4-5), on the other hand,
showed a high speciﬁcity (0.98), but also low sensitivity (0.56) for RAO. The
intermediate MGs 2-3 were a non-speciﬁc ﬁnding, where clinically healthy and RAO-
affected horses completely overlap ............................................................... 67

Fig. 4-1:

Interindividual variability- mucus scores (A) and BALF neutrophil percentages (B) of
each horse of both age groups. Ages of the 13 younger and 13 older horses on X-
axis .................................................................................................... 81

Fig. 4-2:

Comparison between age groups (meaniSD)- Proportions (%) of BALF cell populations
were not signiﬁcantly different between the two age groups (A), but younger horses had
signiﬁcantly (*P < 0.05) higher total BALF cell counts (B) than the older group. No

difference was observed in mucus scores (C) ................................................... 82
Fig. 4-3:

Relationships of mucus quantity with broncho-alveolar lavage (BALF) neutrophil
percentages (A) and absolute numbers (B) .................. 83
Fig. 5-1:

Vector diagram illustrating the Viscoelasticity of mucus (G*), the viscous (G’ ’) and
elastic (G’) components and their ratio deﬁned by tangent 5 ................................ 102
Fig. 5-2:

A) Viscoelasticity at 10 radian/s on a logarithmic scale (log G*; dyn/cmz), B) mucociliary
clearability index (MCI) and C) cough clearability index (CCI) of mucus samples in
RAO-affected (principal group) and control horses (control group) before (0) and after
environmental challenge (6, 24 and 48 hours). # Indicate signiﬁcant differences between
principal and control group at a time point. * Indicates a signiﬁcant difference between
the time point and baseline in the same group ................................................. 103

Fig. 6-1:

Viscoelasticity (log G“) before and after treatment. Negative (Sham, A and saline, B) and
positive (nacystelyn, C) control treatments on RAO-exacerbation (R), control (C) and
cystic ﬁbrosis (CF) samples. RhDNase treatment on R (D), C (E) and CF (F) samples.
Gelsolin treatment on R (G), C (H) and CF (1) samples. Signiﬁcant (P < 0.05) and non-
signiﬁcant (ns) effects are shown in the graphs. RhDNase and gelsolin effects on CF
control samples (F and 1, respectively) were not statistically analyzed (na). ............. 114

xii

Fig. 7-1:

Tissue speciﬁcity eqMUC2 and eqMUCSAC mRNA detected by RT-PCR.
Representative samples from airway generation 1 (G1) from control and recurrent airway
obstruction (RAO)- affected horses, as well as from stomach and colon tissue. Ethidium
bromide stained PCR products were run on 3 % agarose gels. ., ............................ 131

Fig. 7-2:

EqMUCSAC and ZO-l RT-PCR cDNA products at each airway generation and graphic
representation of the ratios of the band volumes. All cDNA products were run on the
same gel (ethidium bromide stained 3 % agarose), lanes shown were cut and pasted with
graphic software. Pooled samples for control and recurrent airway obstruction (RAO)-
affected horses from airway generations (G) l, 5, 10 and 15; ﬁ'om small airways (S) of
approximately 1 mm diameter; and from parenchyma (P) without macroscopically visible
airways .............................................................................................. 132

Fig. 8-1:

Volume density of intraepithelial mucosubstance (A) in lung lobes a, b, c and d of 5
RAO-affected (I-V; black bars) and 5 clinically healthy control (i-v; white bars) horses.
*In three airways volume density could not be measured due to excessive artefacts.
Volume densities did not differ between groups or between lung lobes. Within a horse
some airways could appear to store mucosubstance in rounded goblet cells (B; horse V,
lobe c), while others showed goblet cells expressing mucus (C; horse V, lobe a), or were
almost completely devoid of mucosubstance (horse V, lobe b). ............................ 156

Fig. 8-2:

Distribution of relative mRNA levels of eqMUCSAC(A), CACC1(B), EGFR (C) and
EGF (D) in the 4 lung lobes (a, b, c and d) of 5 RAO-affected (black bars) and 5
clinically healthy control (white bars) horses. All results are expressed as relative x-fold
levels of mRNA compared with the sample that was lowest for all target genes, Ivc,
which is set as l-fold. *Six RNA samples were not available for measurements ...... 157-8

Fig. 8-3:

Partial correlation analysis (r = partial correlation coefﬁcient): signiﬁcant correlations
between eqMUCSAC mRNA levels and equine hCACCl-like (A) and EGFR (B) mRNA
levels as well as neutrophil percentages (C). Fig. D illustrates the loglinear relationship
between mean (a, b, c, d lobes) eqMUCSAC mRNA levels and equine hCACCl-
like .................................................................................................... 159

xiii

INTRODUCTION

In European classical times, respiration was thought to be a cooling mechanism
for the blood, and nasal mucus a discharge from the brain. While Arabic medicine
already possessed a much clearer understanding of physiology in medieval times, the
credit of pointing out the pathophysiological signiﬁcance of respiratory secretions to
Western medicine belongs to Laennec. Excess mucous accumulation became one of the
cardinal Signs and a recognized problem of many respiratory diseases of humans.

The airway mucus apparatus (i.e., mucous secretory cells, ciliated cells, and
progenitor cells) is a component of the innate mucosal defense system, which can
respond rapidly to inhaled toxicants, microbial pathogens, and particulate matter by
increasing mucus production and clearance in order to trap and remove the offending
agents. If production and secretion exceed clearance capacity, however, excessive mucus
accumulation is observed in the airways.

Excess mucus accumulation in the airways is a hallmark of equine recurrent
airway obstruction (RAO) [1-3]. When exposed to conventional stable environment, i.e.,
hay feeding and bedding on straw, RAO-affected horses develop overt respiratory
distress, i.e., heaves. In these severely affected animals the presence of large quantities of
thick sticky-appearing mucus in large and small pulmonary airways is a common
observation. Bronchoscopy, however, has expanded the recognition of excessive mucus
accumulation to a wider range of horses. In particular, mucus accumulation has become a
hallmark of inﬂammatory airway disease (IAD), a milder form of airway disease that is

more difﬁcult to deﬁne.

Three major pathophysiological components of equine lower airway disease are
recognized: mucus accumulation, airway inﬂammation and bronchospasm/airway
hyperreactivity. Compared with the latter two components, the nature and causes of
mucus accumulation have received much less investigative attention and are therefore
less well understood. In the following literature review, I examine the available
information on mucus accumulation in equine lower airway diseases with a focus on
RAO and IAD. Comparative aspects are included when relevant to develop the
conceptual model and resulting hypotheses that form the basis of my PhD work:

The focus of my investigations is the excess mucus accumulation that occurs in
equine RAO. In order to determine if this excess airway mucus is caused either by
increased production and/or by decreased clearance, I did the following:

- validated an endoscopic scoring system and used it to test the hypothesis that
increased mucus accumulation is a ﬁrnction of disease status (RAO) and the environment.
Further, that increased mucus accumulation is associated with neutrophilic airway
inﬂammation and the age of the horse (chapters 2-4).

- used the microrheometric method to test the hypothesis that increased mucus
accumulation is a result of and associated with unfavorable alterations of mucus physical
properties, i.e., increased Viscoelasticity and decreased predicted clearability (chapters 5-
6).

- identiﬁed equine homologues of gel-forming mucin genes and measured mRNA
levels in conjunction with morphometric methods to test the hypothesis that increased
mucus accumulation is a result of and associated with increased production of mucus

(chapters 7-9).

The individual chapters are written and referenced so that they can be read
independently. At the beginning of chapters 2-8 describing the original research, I brieﬂy
state the thoughts behind the individual study in the context of my thesis. The studies
themselves are then reported in the ﬁrst-person plural form, i.e., ‘Vve performed this
study,” since indeed I did not perform these studies alone, but with the help of many

others, and also because I had already written up some of the work as research papers.

CHAPTER 1

LITERATURE REVIEW

Nature of airway mucous secretions

Airway mucous secretions are a heterogeneous mix of water, electrolytes, lipids
and proteins, as well as cells and cellular debris [4]. All these components form the
complex bioﬁlrn most often called “mucus”. Of particular functional importance are
mucins, however. More strictly deﬁned, respiratory mucus represents only the gel-
forrning mucin products derived from secretion of epithelial goblet cells and submucosal
glands. Tracheobronchial secretions, on the other hand, include the mucus plus all the
other components listed above. For the purpose of this work, the terms “mucus
(accumulation)” and “respiratory (or airway) secretions” are used interchangeably to
describe the entity of gel-like secretions with all their components.

Respiratory mucus has important physiological roles: trapping, inhibition,
destruction and removal by mucociliary clearance of inhaled materials as well as
humidiﬁcation of air and prevention of ﬂuid loss. Excess mucus accumulations may
enhance some of these physiological functions. They can also lead to undesired
functional consequences, such as airﬂow obstruction, cough and bacterial colonization.
As in human chronic bronchitis, asthma and cystic ﬁbrosis and in animal models of these
diseases, mucus accumulation in horses is generally viewed as a characteristic, but non-
speciﬁc Sign of various airway diseases [5]. In this review of the literature and in the

following investigations, I will focus on equine RAO and IAD.

Description, definitions and significance of RAO and IAD

Description and deﬁnitions—Non-infectious lower airway disease is a major
health problem of stabled horses. In conventional stable environments in which horses
are bedded on straw and fed hay, the air in the breathing zone of horses contains very
high levels of organic dust with allergens and non-speciﬁc irritants like endotoxin [6-8].
Horses housed permanently indoors in such a conventional stable environment Show high
incidences of non-infectious lower airway disease. Published estimates range ﬁ'om 54%
in Switzerland [9] to 79% in the Netherlands [10]. Thus, environmentally induced equine
lower airway disease is a major health problem of horses, and understanding the
underlying mechanisms is crucial for the improvement of environmental and medical
management. Because the use of various terms referring to environmentally-induced
equine lower airway disease has led to confusion, the state-of-the-art of our knowledge
was reviewed and working deﬁnitions were proposed in two recent workshops [11, 12].
RAO refers to the severe, debilitating disease characterized by clinically evident
increased breathing effort due to cholinergic bronchospasm, coughing, and airway
hyperreactivity as well as neutrophil and mucus accumulation in the airways [1, 3, 13-
22]. In most RAO-affected horses clinical exacerbation is induced by stabling and hay
feeding, and clinical remission is often achieved within days to weeks when horses are
kept at pasture or given corticosteroids [1, 23-28]. RAO has been well characterized, and
pathogenic mechanisms involved in this disease Show striking similarities to human
asthma.

In contrast to horses with RAO, however, many horses Show less severe forms of
lower airway disease without clinically evident increased breathing effort when exposed
to a conventional stable environment. These horses are now categorized as having IAD

5

[29], which may be associated with airway inﬂammation, mucus accumulation and
airway hyperreactivity and clinically manifest with cough and exercise intolerance [18,
29, 30]. The pathogenesis of IAD is much less well understood than that of RAO, but the
above-cited surveys [9, 10] suggest that it is very common among stabled mature horses.

Clinical signiﬁcance-Signs of RAO in exacerbation are obvious and its clinical
signiﬁcance as a debilitating disease is undisputed. IAD, on the other hand, may oﬁen go
undetected unless one speciﬁcally looks for its signs. The clinical signiﬁcance of IAD as
a cause of “poor performance” is far from clear. In a standardized exercise test to fatigue
in Thoroughbred and Standardbred racehorses with poor performance, the predicted
exercise capacity correlated best with indices of oxygen uptake and transport capacity
[31]. Even mild airway disease leading to peripheral airway obstruction and
ventilation/perfusion mismatching that is subclinical at rest, may signiﬁcantly
compromise oxygen uptake during maximal exercise. However, the percentages of “poor
performance” cases attributable to IAD in clinical studies differ widely. Several authors
[32-35] cite high incidences of 22-50% of IAD in racehorses. In contrast, Morris and
Seeherman (1991 [36]) diagnosed only 4% of 275 “poor performance” cases as lower
airway disease based on thoracic radiographs and ventilation / perfusion scintigraphy. In
another study of 348 cases of poor performance, IAD (or lower airway disease) was not
even mentioned, while upper airway, cardiac and musculoskeletal problems constituted
most of the 73% of cases in which a deﬁnitive diagnosis was reached [37]. The same
group, however, found a high proportion of IAD and/or EIPH in 42 “poor performance”
horses in another study [35]. Rather than indicating the true causative effect of IAD,

these wide variations seem to reﬂect the complex nature of “poor performance”,

Table 1-1:

Characteristics

inﬂammatory airway disease (IAD)

of recurrent airway obstruction (RAO) and

 

RAO

IAD

 

Typical signalment

Older (> 8 years of age)
horses; no gender or known
breed preference, but a

genetic predilection has been

demonstrated

In particular, younger racehorses,
but also horses of any age, use
and breed.

 

Causes

Allergens (e. g. mold spores)
in combination with non-
speciﬁc irritants from hay

Indoors stable environment;
Possibly bacterial colonization
especially in younger horses;

 

and exposed to hay

dust and indoors stable otherwise largely unknown

environment

Overt respiratory distress Coughing, poor performance;
Main clinical Signs and coughing when stabled may not Show any clinical signs

 

Main additional

Bronchospasm is reversible
with bronchodilator;

>5% neutrophils in BALF;
increased endoscopic mucus

 

 

 

 

 

diagnostic >20% neutrophils in BALF; grades;

characteristics increased endoscopic mucus increased responsiveness to
grades inhaled histamine or metacholine
Cholinergic bronchospasm; Neutrophilic, eosinophilic or

Main neutrophilic inﬂammation; mast cell -type inﬂammation;

pathophysiological moderate to large mucus moderate to large mucus

characteristics accumulation in trachea accumulation in trachea;

airway hyperresponsiveness

Avoid dust exposure; Avoid dust exposure; various

Treatment Corticosteroids; treatments of unproven efﬁcacy
bronchodilators

 

Broncho-alveolar lavage ﬂuid (BALF)

 

diagnostic difﬁculties, and the speciﬁc focus of the respective study. The role of IAD and

associated excess airway mucus in causing “poor performance” is thus presently unclear.
Despite the confusion on deﬁnitions and the conﬂicting data on signiﬁcance, it is

clear that mucus accumulation is generally regarded as a hallmark of both RAO [3, 11]

and IAD [12, 32, 33] and is often attributed a causative role in airway obstruction.

Clinical and pathophysiological significance of mucus accumulations: IS the
phenomenon worth studying?

Increased mucus accumulation can directly cause bronchial obstruction as well as
effectively increase resting airway wall thickness. This latter effect ampliﬁes to the 4th
power the lumen-narrowing effect of bronchoconstriction [3 8]. That is, the same amount
of mucus will for simple geometrical reasons obstruct the lumen increasingly as the inner
diameter of the airway is decreased by contraction of the surrounding airway smooth
muscle. The logic of this argument is simple and convincing, and the obstructive effect of
excessive secretions seems intuitively clear. The available data, however, presents a
much more complex and confusing picture.

Comparative aspects—The volumes of excessive airway mucus accumulations,
i.e., sputum, and its unfavorable rheological properties are thought to be associated with
the severity of human airway diseases. Mucus accumulation may contribute to the
morbidity of human airway diseases by causing patient discomfort, airﬂow obstruction,
and predisposition to respiratory infection [39]. In large population based studies, the
presence of chronic mucus secretion is associated with excess decline in forced
expiratory volume in one second [40] and with an increased risk of mortality [41].

However, the magnitude of a direct causative effect of excessive airway mucus on airway

obstruction, patient morbidity and mortality, as well as the beneﬁt of “rnucoactive
therapy” are all controversial and, at best, difﬁcult to assess [39]. Less controversial than
the signiﬁcance of excess secretions for airway obstruction, is their role as a predisposing
factor for pulmonary bacterial infections in humans. Exacerbations in patients with cystic
ﬁbrosis and chronic obstructive pulmonary disease are associated with colonization of
mucus by H. inﬂuenzae, M. catarrhalis, S. pneumoniae, H. parainﬂuenza, and, less
commonly, P. aeruginosa and K. pneumoniae [39].

Signiﬁcance of mucus accumulation in equine lower airway diseases—
Following administration of bronchodilators to RAO-affected horses in clinical
exacerbation, airway resistance rapidly decreases, but still remains greater than in control
horses [42, 43]. In another study, an inhaled bronchodilator improved the distribution of
radionucleotide aerosol in all tested horses with acute RAO, but in 5 of 6 affected horses
some degree of airway obstruction was still detectable [44]. Robinson et al. (1996 [3])
suggested that these remaining lung function deﬁcits and the inconsistent effect of
bronchodilators on dynamic compliance may be due to mucus accumulation and airway
wall thickening. For lack of a reliable and practical way to quantify mucus accumulation
its relative contribution has not been deﬁned, however.

In one study the volumes of secretions aspirated transtracheally (obviously
without prior infusion of saline) were directly measured. Four horses without lower
airway disease had less than lml aspirable mucus. Out of 22 horses with IAD or RAO 5
had < lml, 3 had 1-3 ml and 15 had > 3ml of tracheal secretions [45]. Since then, only
subjective endoscopic scoring of mucus accumulation has been employed to quantify
mucus accumulation. Various systems with scores from 0-3 [18] and 0-5 [46, 47] have

been used.

Endoscopic mucus scoring has yielded interesting results. In a retrospective
clinical study, Dixon et al. (1995 [5]) clearly demonstrates the non-Speciﬁc nature of
mucus accumulation in horse airway diseases: compared with a group of healthy control
horses, mucus was increased in RAO (n=148), “undifferentiated pulmonary disease”
(n=18), “infectious pulmonary disease” (n=45), “S. zooepidemicus pulmonary infection”
(n=7), “Lungworm infestation” (n=7), “exercise-induced pulmonary hemorrhage” (n=16)
and “miscellaneous pulmonary disease” (n=20). RAO horses Showed the highest score
(median score of 3), but this was not signiﬁcantly different from all the other groups.
Mucus scores were also increased in RAO-affected horses compared with IAD and with
clinically healthy animals [46, 48]. Furthermore, endoscopically observed increased
mucus accumulation is associated with coughing [49-51], decreased racing performance
[33] and decreased lung function [48].

Increased mucus accumulations are also associated with bacterial colonization of
the airways. When horses are prevented from lowering their head for a prolonged period
of time, mucus pooling with excessive bacterial grth occurs in the distal trachea [52].
This may increase the risk of bacterial pleuro-pneumonia in immunocompromized
animals. Furthermore, inﬂammation scores, which include mucus accumulation, are
highly correlated with positive bacterial cultures from tracheobronchial secretions in IAD
of young racehorses [53]. Molecular interactions between bacteria and mucin
glycoproteins are a highly interesting ﬁeld, which is only beginning to be explored [54].
However, although even in healthy horses S. zooepidemicus and other upper airway
commensals are frequently isolated from tracheal secretions [55], exacerbations
associated with bacterial infections are not a clinical problem recognized in RAO-

affected horses [11].

10

Based on the available information, I conclude that mucus accumulation is a
clinical problem in IAD, RAO and other lung diseases, can directly and indirectly affect
lung function and performance and can be associated with bacterial colonization. I also
note, however, that the magnitude of these effects is unclear, that the effect on
performance has not been studied in sport horses, and that endoscopic scoring has not
been validated. No one grading scale is generally accepted, and, therefore, comparisons,
between studies and observers are difﬁcult at best. In some studies a mucus score is
compounded into a general examination score based on various other parameters [53],
and it is therefore impossible to extract mucus speciﬁc information. The repeatability of
mucus scores both within an individual horse and within and particularly between
observers may be problematic in a measure that is both variable and subjective by nature,
for example, a horse may cough and clear its trachea of mucus, exercise could increase or
reduce visible mucus, and “a few” versus “many” mucous ﬂakes may not mean the same
to different observers.

The diagnostic potential of a standardized mucus score and, possibly
immunological tools for quantifying mucins [54], must be investigated and validated. The
signiﬁcance of mucus accumulation in sport horses should be evaluated. Moreover, in
order to better understand, and ultimately treat this phenomenon, its causes and

underlying mechanisms must be addressed.

Causes and mechanisms of airway mucus accumulation
The amount of mucus present in the lumen of airways is the result of a dynamic
process that involves speciﬁc mucin gene expression, the production, storage and

secretion of mucins into the airway lumen, and the subsequent clearance by cilia lining

11

the conducting airways. In addition to mucins, all other components of mucous secretions
can be present in excess, and must also be removed by mucociliary clearance. Clearance
capacity, probably genetically determined, can therefore be reduced by alterations in the
ciliary apparatus and/or in mucus physical properties that inﬂuence its clearability.

This is a simple concept, but our understanding of its fundamental cellular and
molecular regulatory mechanisms in normal and diseased airways is as yet very
incomplete. Although production logically precedes clearance, I will discuss the latter
ﬁrst, since more is known about it in the horse.

Clearance: primarily compromised or just overwhelmed?—Mucociliary clearance
in humans increases from the periphery to the central airways, is reduced during sleep
and decreases with age. Physiologic regulation is primarily exercised by autonomic
mechanisms: B-adrenergic agonists are potent stimulants of clearance, and cholinergic
antagonists potent depressants, while the effect of cholinergic agonist drugs is variable
[56].

In healthy horses, physiologic regulation of clearance appears to be similar to
other species. While atropine (0.02 mg/kg IV) signiﬁcantly decreases tracheal clearance
rate [57], a Bz-agonist (Clenbuterol 0.8 ug/kg IV) signiﬁcantly increases tracheal mucus
velocity in vivo [58]. Water vapor-saturated air therapy, exercise and mild dehydration by
ﬁrrosemide have no effect on mucociliary clearance in healthy horses [59]. Gravitational
force can decrease or accelerate tracheal mucus velocity in vivo [60] and in vitro [61].

A variety of airborne pollutants, such as S02, N02, 03 and smoke ﬁ'om cigarettes,
have deleterious effects on mucociliary clearance. After cessation of smoking,

mucociliary clearance improves after 3 months [56]. In donkeys, exposure to sulfuric acid

12

 

(H2SO4) leads to a decreased mucociliary clearance that takes up to three months to
return to normal [62].

Humans with chronic bronchitis, emphysema, CF and asthma can have reduced
rates of clearance. A Single allergen challenge reduces mucociliary clearance rate by 28%
in asthmatics, and by 46-64% for 7 days in sheep [56, 63]. Airway inﬂammation appears
to play a major role in this decline of mucociliary clearance [63]. Similarly, a decrease in
mucus clearance associated with airway inﬂammation could contribute to the
accumulation of airway sections in horses with allergic lower airway disease [2].

Investigations of the effectiveness of mucociliary clearance in RAO-affected
horses have yielded conﬂicting results. Clearance has been determined by use of a
radioactive marker that was followed up the airway by a scintigraphic method. In RAO-
affected animals, clearance was decreased by 24% compared with controls, and
radioactive marker was less likely to move as a discrete bolus [64]. Another investigation
that tracked the movement of a surface marker by use of endoscopy found a tracheal
mucus velocity of 11.41il.09 mm/min in RAO-affected animals vs. 21.41:] .03 mm/min
in controls [58]. Others, in contrast, found no difference between RAO-affected and
healthy horses [65]. No studies have investigated mucociliary clearance rate in IAD.
Infection with rhinovirus has no effect on mucociliary clearance, but herpesvirus reduces
clearance in half of the infected horses. In contrast, mucociliary clearance is severely
depressed in all horses aﬁer inﬂuenza infection and takes four weeks to return to normal
[66], underlining the importance of adequate rest periods and good air hygiene after
clinically apparent respiratory viral infections.

The studies described in the previous paragraph also found a wide variation of
transport rates among healthy control animals. Validation of an endoscopic marker

13

method showed good repeatability within 20 healthy horses, but considerable differences
between individual horses [67]. Such wide interindividual variation also has been
observed in humans, and twin studies suggest an individual, genetically determined
mucociliary clearance capacity [68].

When the amount of mucus exceeds the mucociliary clearance capacity,
secretions progressively accumulate, and must then be cleared by coughing [69]. The
association of excess mucus accumulation and cough has been shown in a clinical study
in horses [49]. In airway disease, mucociliary clearance capacity may be compromised
due to changes in the ciliary apparatus or unfavorable physical properties of the mucus
gel [2, 69].

Ciliary function and dysfunction —Ciliary amplitude and beat frequency determine
the maximal velocity at the tips of the cilia, and hence the maximal forward force
transmitted to the mucus layer. The density of spacing between the cilia will also affect
clearance rate, since more energy is lost with increasing distance between the cilia. In
smaller airways, the cilia are generally shorter, and less densely packed than in the large
bronchi, which partly explains the decrease of transport velocity from the central airways
towards the periphery.

Ciliary dysfunction in disease can result in severe mucus stasis and loss of the
natural defense against inhaled particles. Immotile cilia and ciliary dyskinesia syndromes
in humans demonstrate the potentially lethal effects of ciliary dysfunction [56]. These
congenital diseases have not been diagnosed in the horse. Respiratory viral diseases,
however, can also seriously affect ciliary function. In humans, reduced mucociliary
transport rate following viral infections correlates with the presence of acquired ciliary
defects. About 50% of cilia must be lost or damaged to cause a measurable decrease of

14

clearance [56], indicating the high reserve capacity of the ciliary apparatus. Differences
in severity of epithelial ciliary damage caused by rhino-, herpes- and inﬂuenza viruses,
explain the differences in mucociliary clearance decrease observed after infection with
these agents [66].

Compared with viral diseases, the extent and signiﬁcance of damage to the ciliary
apparatus is less clear in chronic non-infectious airway diseases of human [56] and horse:
Kaup et al. (1990 [70]) reported a loss of ciliated cells and abnormal ciliary structure in
large conducting airways of RAO-affected horses. Undifferentiated cells replaced ciliated
cells in a hyperplastic epithelium. These changes differ greatly in their severity and extent
between RAO-affected horses. While the extent of ciliary loss is associated with the
severity of clinical disease, the degree of ciliary malformations appears unrelated to
disease state and severity. Galati et al. (1991 [71]) point out that such structural ciliary
defects occur in about 5% (a frequency comparable to other species) of cilia in airway
epithelium of healthy horses. Considering the high reserve capacity of the ciliary
apparatus, the clinical signiﬁcance of the changes observed in non-viral lower airway
diseases is unclear.

Studies on excised pieces of trachea from healthy horses have shown that
variations of ciliary beat frequency at physiological temperatures have only a minor
effect on mucociliary clearance rate. In contrast, physical properties of mucus are very
important [72].

Mucus biophysical properties and their alterations in disease—Mucous factors
affecting mucociliary clearance are the depth of the mucous layer and its biophysical
properties. Mucus that is deep, i.e., more than the physiological 5-10 microns, is well
suited for cough clearance, but unfavorable for mucociliary transport [69]. Mucus

15

biophysical properties can be characterized and quantiﬁed as visco-elasticity, spinnability
and adhesivity. All of these affect clearance, but viscosity and visco-elasticity are the best
studied in respiratory pathophysiology. They are also the only mucus physical properties
that have been investigated in the horse.

A balance between viscosity and elasticity must be maintained for optimal
mucociliary clearance: The elasticity of mucus is important for clearance by cilia because
elastic mucus efﬁciently transmits energy. The viscosity of mucus results in energy loss,
but is necessary so that mucus can be displaced. The essential visco-elastic nature of
mucus is a result of its three-dimensional, cross-linked mucin network structure. Fig. 1
illustrates how this three-dimensional network is in turn dependent on a number of
molecular forms of bonding: l) disulﬁde bonds join glycoprotein subunits into extended
macromolecular chains; 2) because of their Size, mucin polymers readily form
entanglements among each other; 3) oligosaccharide sidechains, which. make up 80 % of
the mucin weight, form hydrogen bonds with complementary sugar units on neighboring
mucins; 4) ionic interactions (negatively charged sugar units and positively charged
amino acids) extend the macromolecular conformation; 5) inﬂammation can add an extra
network of high molecular weight DNA and actin ﬁlaments released from dying
leukocytes. Interactions with water, electrolytes, hydrogen ions (pH), lipids, enzymes and
other (serum) proteins will inﬂuence these mucin gel-bonds, resulting in the physical
properties of mucus.

The sheer complexity of this system makes it a difﬁcult object of study.
Furthermore, healthy airways contain only small amounts of secretions available for
sampling and analysis. Only the development of methods that allow for the investigation

of microliter quantities of mucus—microrheometry (King and Macklem, l977)—has made

16

the comparison of “diseased” with “healthy” mucus possible [73]. Hence, the literature is
still heavily biased by data from disease states without healthy controls. This helps
explain why the role of unfavorable rheological changes in chronic bronchitis, cystic
ﬁbrosis and asthma is acknowledged, but its importance is still controversial. Patients
suffering from these diseases may have mucus of very high, normal or, uncommonly, low
Viscoelasticity. Overall, high Viscoelasticity is associated with increased purulence of
secretions [39, 69].

Not surprisingly, the available data on mucus rheology in equine airway disease is far
from complete or conclusive. Based only on the physical appearance of tracheobronchial
secretions, Schatzmann et al. (1972 [45]) proposed that mucus viscosity increases with
the severity of clinical Signs of airway obstruction in RAO-horses. However, when these
authors measured viscosity of mucus in 5 RAO-affected horses they found it to be lower
than in human patients with severe chronic bronchitis [21]. Hajer (1979 PhD thesis, cited
in [58]) found increased viscosity of tracheal mucus in RAO horses. No study has
directly compared mucus from RAO-affected to that of healthy horses, however.

It is important to note that there is indirect evidence that factors other than
Viscoelasticity may negatively inﬂuence clearability in RAO-affected horses. Adhesivity
and wettability, which contribute to the optimal interface properties of mucus, are
primarily inﬂuenced by its phospholipid content and composition [74]. Exogenous
surfactant increases transport velocity ex vivo [75] and in vivo [76]. Surfactant is present
in amounts sufﬁcient to signiﬁcantly reduce surface tension in the trachea of healthy
horses [77], but clinical observation (reduced foam of BAL) and biochemical analysis of
airway secretions suggest that a deﬁcit and altered composition of surfactant are present

in RAO [78], and after horses have been subjected to long periods of transport [79].

17

 

 

Fig. 1-1: Types of bonds in a mucous gel and respective mucolytic agents
Entire ﬁgure ﬁ'orn M. King, Pulmonary Research Group, University of Alberta, Canada.

     

. BANE WOZUM Em:

. . . man—:83
. $03385 mac—Baa Ho gram . .
E g mango—0% m-m wanna Oman—mac:

. HOZHO wOZUw

. Beam: Engage—3E8 was we? .
. @3548 BE now—5.43 muom arm—don. . Uaxg
5.3%. an... 26» _o om gunman . . . -

 

. EMUWOQHZ WOZUm . I, . .
. mica—mm mar 90 _ $38.5
ommogoornlmo aaoég ., .

. <>2 DH” 3%. wOHNOHw

. Funnmmmmnwmoa c0333
ammogo armada Seaman
33. co r3335»

. EEQEQ
. Emman— an—m—goau

caning 35am: avenge—3&8

l8

 

 

wen: cm 580 368 cm
coax—m mm 2.3 on» 8
3:3un

  

. MEG—mg” 62>
@mfPOHHZ

. wan-=0— aonﬂonr moan—9mg F F? amon—

 

 

 

 

 

 

 

 

Our current understanding of equine mucus rheology in the horse is patchy at
best, but it provides a rational basis to hypothesize that unfavorable mucus rheology is a
cause of the decreased mucus clearance in RAO-affected horses. Characterization of
mucus rheology in RAO and investigation of underlying pathophysiological mechanisms
Should ultimately lead to better treatment. I therefore set out to compare Viscoelasticity
between RAO-affected and healthy control horses before and during envirorunental
challenge.

Production and secretion -The physiologically heterogeneous mucus is a mix of
water (~95%), electrolytes (1%), lipids (1%), proteins (2-3%): enzymes, other (serum)
proteins and high molecular weight O-linked glycoproteins referred to as mucins, as well
as very variable amounts of epithelial and inﬂammatory cells [4]. There is very little
published information on the relative proportions of these components in health and
disease [80]. Nonetheless it is clear, that the concentration of all of these components can
be altered, and all these alterations will affect the biophysical properties and,
consequently the rheology of mucus. In contrast, only an increase in inﬂammatory cell
inﬂux, serum protein exudation and mucin secretion can be expected to Signiﬁcantly
increase the volume of the secretions to result in their accumulation. Proteins, in
particular mucin glycoproteins, will inﬂuence the total volume by raising oncotic
pressure, i.e., drawing and retaining water in the mucus gel. Inﬂammatory cells may
directly contribute to the volume of mucus, i.e., “mucopus”.

The BALF total cell counts are consistently increased during RAO exacerbations.
The quantiﬁcation of cells or substances in the pulmonary lining ﬂuid of the lower
respiratory tract, as obtained by BAL, is not precise because of the variable dilution of

19

the ELF by the instilled lavage ﬂuid. In order to estimate the number of cells per volume
of ELF, cell numbers have been adjusted to the concentration of albumen or urea. In
horses, the total cell count per volume of ELF is moderately higher in RAO-affected
animals (approximate median 45’000 cells / ul) than in controls (~30’000 cells / pl),

when adjusted for albumen, but slightly lower (~17’000 cells / ul in RAO vs. ~20’000

cells / u] in controls) when adjusted for urea concentration [81]. These results do not
indicate substantial differences in cell content of airway secretions. The use of urea and
albumin concentrations in such calculations is fraught with methodological problems,
however [82]. Furthermore, direct microscopic inspection of undiluted RAO tracheo-
bronchial secretions oﬁen shows a high cellularity compared to mucus sampled from
control horses, suggesting that the volumetric contribution of inﬂammatory cells to
mucdpurulent (sic!) airway secretions may be substantial.

Persson (1995 [83]) proposed that airway inﬂammation causes plasma exudation
from postcapillary venules, earlier than and independent from cellular extravasation.
According to this model, the exudate causes a bulk ﬂow of non-sieved plasma into the
airway lumen rather than airway wall edema. After activation of proteolytic cascades, the
multiplication of plasma-derived protein molecules attracts and retains progressively
more ﬂuid, thereby further increasing the volume of the mucus. Winder and von
Fellenberg (1986 [84]) immunohistochemically demonstrated extravasated ﬁbrinogen in
tissues of horses with RAO, indicating that plasma exudation does take place in RAO.
Lower BALF albumen: urea ratios in RAO than in controls [81], however, speak against

signiﬁcant plasma exudation, which should increase albumen relative to urea.

20

While the relative contributions of cellular and plasma protein components to the
volume of airway mucus secretions in RAO are presently unknown, I will focus on the
main constituents of mucus, the large, cross-linked mucin glycoproteins.

Mucins: production and secretion—Mucins, also called mucous glycoproteins, are
high molecular weight glycoconjugates. Numerous oligosaccharide side chains are 0-
glycosidically linked to threonine and serine of the peptide core, which represents the
primary gene product, the apomucin. The threonine- and serine-rich regions are formed
of tandem repeat sequences, which are the common feature of epithelial mucins. By
convention, human mucin genes (products) are designated by MUC, mouse by Muc, and
rat by rMuc. Epithelial mucins are subdivided into membrane-associated and secretory
products. MUC7, a salivary mucin, is the only soluble secretory form, while all other
secretory mucins identiﬁed so far contain non-repetitive cysteine-rich regions, believed to
allow intra- and intermolecular cross linking through disulﬁde bond formation. The major
gel-forming mucins expressed in human and rodent airways (MUC2, MUCSAC and
MUCSB) are encoded by members of a large family of mucin genes [85-87]. Also present
are highly glycosylated, monomeric WC] and MUC4 mucins Shed from the cell surface
as well as additional N and/or O-glycosylated proteins, lipids, and glycolipids [88].
Synthesis, storage and release of mucins are central properties of the innate mucosal
defense against inhaled irritants and pathogens.

“Too much of a good thing” aptly describes the situation where the airway
mucosa produces and releases so much mucin glycoprotein as to overwhelm the capacity
of the mucociliary transport system. Such a hypersecretory state is found in human
chronic bronchitis, asthma exacerbation and cystic ﬁbrosis [89]. MUCSAC and MUCSB
are the major oligomeric mucins and airways mucus contains variable amounts of these

21

 

glycoproteins [90-92]. By contrast, the MUC2 mucin comprised only 2.5% of the weight
of the gel-forming mucins, indicating that MUC2 is a minor component in human
sputum. Finally, the amounts and glycosylated variants of the MUCSAC and MUCSB
mucins can be altered Signiﬁcantly in diseased airways with, for instance, an increase in
the low-charge form of the MUCSB mucin in cystic ﬁbrosis and chronic obstructive
pulmonary disease mucus [90-92].

MUCSB is expressed predominantly in submucosal glands, and an increase is
thought to occur through gland enlargement [93]. One study found up-regulation of
MUCSB expression in a mouse model of asthma [94], but others report that this mucin is
constitutively expressed and does not directly respond to irritant and inﬂammatory
stimuli [95]. In contrast, many studies have shown MUCSAC up-regulation in response
to various stimuli and mediators -notably BAL from asthmatics- has been shown in a
variety of models of airway inﬂammation [95, 96]. MUCSAC expression also precedes
and accompanies epithelial mucous cell metaplasia [97]. Increased expression of MUC2,
normally a predominantly intestinal mucin, may play a role in cystic ﬁbrosis [98, 99].
Presently, MUCSAC is regarded as the main “signature” mucin in experimental and
natural models of airway disease.

Based on reactivity with polyclonal antibodies raised against the respective human
mucins, it has been proposed that a homologue of MU CSAC is present in equine stomach
[100], and that homologues of MUCSAC and MUCSB may be found in equine airway
secretions [101, 102]. These ﬁndings are based on reactivity with antibodies that have
been raised against human mucins, however, and genetic identiﬁcation of equine
homologues of the main gel-forming mucins is a priority for the investigation of mucin

hypersecretion in RAO.

22

Jefcoat et al. (2001 [54]) assessed relative differences in amount of mucin
glycoproteins between control and RAO-affected horses using a carbohydrate side chain-
speciﬁc monoclonal antibody (4E4) in an enzyme-linked immunosorbent assay, and by
carbohydrate-speciﬁc enzyme-linked lectin assays. Signiﬁcantly increased levels of 4E4-
immunoreactive glycoprotein and the mucin-associated carbohydrates fucose (oz-1,2
linkage), N-acetyl-glucosamine, and N-acetyl-galactosamine were detected in RAO
horses in acute disease. RAO horses in remission also showed increased levels of 4E4-
recognized mucin, a-1,2 fucose, and N-acetyl-glucosamine, while N-acetyl-
galactosanrine levels were equivalent to those of controls. Although not measures of
absolute amounts of mucin glycoproteins or mucin carbohydrates in the airways, these
ELISA and ELLA assays Show persistent alteration of mucin oligosaccharide side-chains
in RAO-affected horses. Mucin oligosaccharide side-chain structure has ﬁrnctional
importance, imparting speciﬁc binding activity toward structures such as bacterial
adhesin molecules [103, 104], and contributing to the degree of Viscoelasticity of the
mucus layer [69, 105]. Individual mucin sugars, such as fucose, have also been used as
markers of tracheobronchial mucus production in humans [106], and a-1,2 fucose may be
associated with MUCSAC, in particular (J efcoat, personal communication).

Origin and mechanisms of mucin hypersecretion -Mucin gene products are
produced and secreted by epithelial mucous secretory goblet cells, which produce
MUCSAC and MUCSB, and by submucosal glands, which produce mainly MUCSB
[107]. In the horse, submucosal glands are Sparse even in the large airways [108], while
the submucosal glands can contribute a large fraction of total secreted mucins in humans.
In natural and experimental airway diseases, increased amounts of luminal airway mucus

are correlated with the histological appearance of numerous goblet cells in the surface

23

epithelium and submucosal glands in airways that normally contain only some mucous
cells (i.e., mucous cell hyperplasia) and in regions that normally contain few or no
mucous cells (i.e., mucous cell metaplasia) [97, 109-111].

Several histopathological and ultrastructural studies describe mucous cell
hyperplasia, metaplasia, and hypertrophy in puhnonary airways of RAO-affected horses
(e.g., [70]; reviewed in [2]). None of these reports are based on quantitative
morphometric techniques, however. In contrast, Hotchkiss (1998 [112]) reported
preliminary morphometric results that surprisingly showed no signiﬁcant differences in
the amount of stored mucosubstance between RAO and control horses at any airway
generation.

Such measurements are based on the assessment of “a moment in time” within the
dynamic process of mucin production, storage, and secretion, however. Current
understanding of inﬂammatory airway pathogenesis indicates that after injury, immediate
release of stored mucosubstance is followed by up-regulation of mucin production within
hours, and an increase in goblet cell numbers within days [113].

Since horses have few submucosal glands and MUCSAC appears to be the main
inducible mucin of goblet cells, I propose that in RAO-affected horses the bulk of
secreted mucins are MUCSAC products derived from surface epithelial goblet cells [2,
70]. Excessive amounts of airway mucus in RAO-affected horses could thus be due to
elevated production and secretion of mucins by the same number of mucous cells found
in normal horses, i.e., increased production and secretion per cell. Alternatively, there
may be increased numbers of airway epithelial mucous cells without a change in
production or secretion per cell. Concurrently, total volumes of stored mucosubstance in

the epithelimn can be increased or decreased. Storage may increase especially during

24

clinical remission and “low-level secretion”, while it may, at least temporarily decrease in
“high-through-put states” during acute exacerbation.

Control of mucus (hyper)secretion can be further divided into many more steps,
involving both transcriptional and post-transcriptional mechanisms [96] that regulate
mRN A formation and degradation (stability), protein translation and carbohydrate
modiﬁcation at the molecular level; granule formation and exocytosis on the cellular
level, as well as the above discussed changes on the epithelial level. Furthermore, these
mechanisms must be viewed in the pathophysiological and, in particular, the
immunological context of the disease.

Pathophysiological and immunological context of mucus hypersecretion —Innate
immunity provides broad, but relatively nonspeciﬁc host defenses [114]. Macrophages
and neutrophils as well as epithelial cells are regulator and effector cells of the innate
inﬂammatory response. Two of the most important early—response cytokines in innate
immunity are interleukin (IL)-lB and tumor necrosis factor (TNF)-a. While toll-like-
receptors (TLRs) mediate pathogen recognition by the innate immune system, cytokines
are necessary for the full development of the innate host defense and the transition to
adaptive immunity.

The "yin-yang” T helper(h)2-Thl hypothesis postulates two distinct patterns of
cytokine secretion by stimulated CD4+ T cells [115]. The lists of known Th1 (interferon
[IFN]-y IL-2, -12, -18 and TNF-a) and Th2 (IL-4, -5, -6, -10, and -13) cytokines
continues to expand, and some of these cytokines are not exclusively produced by one
type of Th cells, but also by other T cells and cells unrelated to leukocytes. Th1 type cells
primarily confer immunity against intracellular pathogens, whereas Th2 cells mediate

responses against extracellular parasites [116]. Th2 type cytokines are also key factors in

25

asthma and experimental models of allergic airway disease [117-119]. Like innate
immunity reactions, Th1 responses can be initiated by inhalation of microbial pathogens,
endotoxin and particulate material in organic dusts. Th1 type proﬁles are observed in
human chronic bronchitis. Both Th1 and innate immunity may oppose Th2-mediated
responses, but can also exacerbate them or cause inﬂammatory airway disease in their
own right [120].

Speciﬁc proﬁles and polarization (Th1 vs. Th2) depend on the composition,
timing, dose and route of the challenge as well as the genetic background of the species,
breeds and strains investigated [121-124]. For instance, in (early) asthma Th1- and Th2
type cytokines are often co-expressed [125] and their proﬁles as well as the associated
pathophysiological and morphological features overlap in people suffering from both
chronic bronchitis and asthma [126].

Most cytokines play pleiotropic roles in down-stream eﬂector pathways with
functional divergence, convergence, “cross talk” and redundancy [127-129]. IL—l3, in
contrast, has been identiﬁed as a critical regulator and very direct effector of the allergic
response (hyperresponsiveness; mucus secretion) in animal models [127, 130]. Recently,
a calcium-dependent chloride channel (CACCI; gob-5 in the mouse) has been identiﬁed
as an important down-stream effector element of IL-l3, mediating airway
hyperresponsiveness and overproduction of mucus in animal models and, possibly, in
asthmatic subjects [131-133]. IL-13 as well as Thl-type and innate immunity cytokines
can also up-regulate and activate epidermal grth factor receptor (EGFR) through

epidermal growth factor (EGF), transforming growth factor (TGF)-0t and other ligands.

Neutrophils can up-regulate EGFR expression through TNF-a and subsequently activate

EGFR in a ligand-independent fashion through the release of oxidative radicals [134,

26

135]. TNF-a can also directly increase MUCSAC mRNA half-life, indicating that
transcript stabilization in addition to up-regulation of expression may be an important
mechanism leading to increased MUCSAC mRNA levels. 'In the bronchial epithelium,
both CACCl and EGFR pathways converge on up-regulation of MUCSAC, the gel-
forming, dominant mucin gene expressed in goblet cells, with subsequent mucous
remodeling and hypersecretion [136]. Endotoxin and other microbial components can
potentiate this effect through toll-like receptors (TLR-2 and-4) [137] and activation of the
NF-KB pathway [138]. Less well studied, but as important as the induction of mucin
hypersecretion, is what happens afterwards: There is evidence that the Th2 cytokine IL-
13 delays apoptosis of goblet cells leading to persistence of mucus cell metaplasia, while
the Th1 cytokine IFN-y leads to the resolution of these Th2-associated changes [139,
140]

Much insight regarding mucous secretory cell biology has been gained in rodent
models, and the above cited “yin-yang" example of regulation is very appealing.
Nonetheless, extrapolation to other species demands caution. It is important to recognize
that mice and many rat strains normally have few mucous secretory cells in their trachea,
bronchi, and bronchioles that are instead dominated by a Clara cell lineage system [141,
142]. Conversely, in human beings [142] as well as in horses [2, 70], mucous secretory
cells are found in the surface epithelium of many airway generations. Therefore, more
mucous secretory cells in a mouse most likely represents metaplastic conversion of Clara
cells into mucous cells, whereas in people and likely in horses, mucous secretory cell
hyperplasia is predominant. This important pathogenetic difference may translate to the
molecular mechanistic level: For instance, IL—4 and IL-13 induce mucus cell metaplasia

and Muc5ac overexpression in genetically manipulated mice [143, 144], but decrease

27

MUCSAC expression in normal human nasal and bronchial epithelial cell cultures [145-
147]

Pathophysiological and immunological mechanisms in equine lower airway
disease—Cholinergic bronchospasm is directly responsible for most of the measurable
(~80%) airway obstruction during RAO crisis [42]. Neural control of mucus secretion is
likely less important in the horse than in species with many submucosal glands.
Mediators of “neurogenic inﬂammation” such as substance P, however, may induce
mucus hypersecretion in horses [148]. Mast cell mediators increase smooth muscle
excitability [149], and airway hyperreactivity is associated with increased mast cell
percentages in broncho-alveolar lavage ﬂuid (BALF) [30]. In contrast, neutrophils, the
predominant inﬂammatory cell type in RAO, appear to have no direct effects on
bronchospasm [149], but may exacerbate obstruction by inducing mucus accumulation
[150, 151]. Neutrophils produce potent secretagogues that are increased in RAO: reactive
oxygen species [152], proteases in particular elastase (Jefcoat, personal communication)
and LTB4 [153]. Immunological causes and mechanisms remain ill-deﬁned, but there is
evidence for IgE- and IgG- mediated type I and III hyperreactivity [154, 155] against
mold allergens, as well as non-speciﬁc responses to endotoxin [156]. The full clinical
severity and pathophysiological changes can only be reproduced by natural challenge
with hay and straw [157]. These ﬁndings are in accordance with recent evidence for
mixed T112 and Th1 cytokine proﬁles in RAO. A Th2 type bias is apparent particularly
during crisis: increased numbers of interleukin IL-4 and IL-5 positive (in situ
hybridization) cells in BALF and increased mRNA levels of IL-4 and IL-13 in BALF
[158-160]. However, Th1 type cytokines IFN-y, IL-lB, TNF-a mRNA levels in BALF

can also be increased in crisis and may persist longer during remission [159, 161].

28

Furthermore, irritants and toxins, in particular endotoxin that is abundant in the stable
environment [7], can both directly and indirectly increase and potentiate airway mucin
expression and mucus cell hyper- and metaplasia [97, 162, 163], and may therefore play a
role in RAO- and IAD-affected as well as in clinically healthy horses.

When RAO horses are removed ﬁom conventional stable enviromnent,
bronchospasm decreases and clinical signs wane within days. Residual lung function
deﬁcits, neutrophil and mucus accumulation persist, however, and are highly correlated
with increased NF-KB activity and ICAM-1 expression of epithelial cells [164]. NF-KB is
an important intracellular signaling element that is involved in mucin gene up-regulation
in vitro [96]. Sustained NF-KB activation in equine RAO is driven by granulocytes and
mediated by IL-1 [3 and TNF-a [165], both of these cytokines increase MUCSAC mRNA
levels in vitro [166]. While some of the cytokines have been investigated in the horse, the
role of epithelial factors in equine RAO is unknown. I propose that EGFR may be central

to mucin gene up-regulation in RAO, and that CACCl could also play a role.

29

 

Figure 1-2: My illustrated view of some immunological and pathophysiological

mechanisms that may be involved in equine lower airway disease

90mg

ma\

6 9 $02; 902--

q
13‘ II”-
.L,’ -‘

HIHHIh‘h .BHID!.r1

Some signaling element graphics from J .A. Hotchkiss, Laboratory for Experimental and

 

Toxicologic Pathology, Department of Pathobiology and Diagnostic Investigation, Michigan

State University, USA.

30

Conclusions of literature review and derived hypotheses

The literature review Shows that increased airway mucus accumulation is a
clinically signiﬁcant problem in equine lower airway disease, both in recurrent airway
obstruction (RAO) and in the milder inﬂammatory airway disease (IAD). It is also
apparent that there is no critically validated tool to quantify mucus accumulation in
equine RAO and IAD, and furthermore, the causes of airway mucus accumulation are
largely speculative. There is evidence, however, that mucus clearance is reduced in RAO.
Findings in human airway diseases indicate that unfavorable alterations of mucus
Viscoelasticity can cause mucus stasis in the airways. Neutrophils, the predominant
inﬂammatory cell in RAO, can cause both unfavorable rheological alterations by
neutrophil breakdown products like DNA and f-actin. Several authors have also
suggested that mucus cell hyper- and metaplasia play a role in equine RAO, but this
claim has thus far not been substantiated by morphometric measurements. Since
submucosal glands are sparse in horse airways, changes in epithelial goblet cells must
account for any increased production and secretion of mucins in equine lower airway
disease. Gel-forming mucins (MUC2, MUCSAC, MUCSB) have been identiﬁed in
several other Species, and their expression and secretion is regulated within complex
immunological networks. Neutrophil products as well as Th1 and Th2 cytokines that are
involved in equine RAO can inﬂuence the expression of mucins, particularly of
MUCSAC, in airway diseases of other species. Recent evidence from experimental
models of airway disease ﬁrrther suggests that MUCSAC up-regulation converges on ‘

EGFR and CACCl pathways in the airway epithelium.

31

Neither mucus accumulation, nor mucus viscoelasticity, nor mucin production
and storage have been critically studied in equine RAO. The purpose of my work was
therefore to quantify airway mucus accumulation and investigate its causes in RAO-
affected compared with clinically healthy control horses.

To achieve this goal, the following hypotheses were developed:

1) Increased mucus accumulation is a function of disease status (RAO) and age of
the horse as well as the environment. Further, increased mucus accumulation is
associated with neutrophilic airway inﬂammation.

2) Increased mucus accumulation is a result of and associated with
unfavorable alterations of mucus physical properties, i.e., increased Viscoelasticity and
decreased predicted clearability and/or

3) increased production of mucus, i.e., increased mucin gene mRNA levels and
storage of mucosubstrance in the airway epithelium.

This dissertation describes: a) the validation of an endoscopic scoring system for
mucus accumulation and its use to test the ﬁrst hypothesis (chapters 2-4), b) the use of
the microrheometric method to test the second hypothesis (chapters 5-6), and c) the
identiﬁcation of equine homologues of gel-forming mucin genes and quantiﬁcation of
mRNA levels in conjunction with morphometric methods to test the third hypothesis
(chapters 7-9).

The individual chapters are written and referenced so that they can be read

independently.

32

Fig. 1-3: Basic model hypothesis: Mucus accumulation is caused by increased

production and/or decreased clearance, in turn these are consequences of airway

 

 

   
    

inflammation
Mucus Accumulation
Decreased . , _. - Increased
Clearance ', 53;“ Production
- -~\-._i-""’.JA ’
Inflammation

 

 

 

“Angry epithelium” from J .R. Harkema, Laboratory for Experimental and Toxicologic Pathology,

Department of Pathobiology and Diagnostic Investigation, Michigan State University, USA

33

CHAPTER 2
ENDOSCOPIC SCORING 0F MUCUS QUANTITY AND
QUALITY—OBSERVER AND HORSE RELATED VARIANCE AND

RELATIONSHIP WITH MEASURED VARIABLES

Increased mucus accumulation and altered mucus rheological properties have
been associated with equine lower airway disease. Endoscopic scoring has been
previously used to assess changes in both mucus quantity and quality in horse airways.
The validity of such scoring systems, however, has not been investigated.

In order to determine if endoscopic scoring could be used to investigate airway
mucus in horses, I performed the following study assessing observer and horse variance

of endoscopic scores and their relationship with measured variables.

34

Abstract

Reasons for performing the study—Endoscopic scoring of airway mucus quantity
and quality has not been critically assessed.

Objectives—To evaluate mucus scores for 1) observer- and horse-related variance;
2) association with inﬂammation, mucus Viscoelasticity and measured amounts.

Methods—Variance of scoring within and between observers and over time within
horses for mucus accumulation, apparent viscosity, localization and color, and
correlations of mucus accumulation scores with neutrophils in secretions were
determined. The relationship of accumulation score to measured volumes of “artiﬁcial
mucus” was investigated. Correlations of mucus accumulation, apparent viscosity and
color scores with measured Viscoelasticity were tested. Viscoelasticity was compared
between ventrally and dorsally collected samples.

Results—Mucus accumulation scoring showed excellent interobserver agreement
and moderate horse-related variance, was related to measured volumes of “artiﬁcial
mucus”, and correlated well with neutrOphilic airway inﬂammation. Scores of mucus
viscosity, color and localization showed high observer-related variance. Mucus
accumulation, apparent viscosity and color scores did not correlate with measured mucus
viscoelasticity, but dorsally localized mucus showed 2-fold higher measured
Viscoelasticity than ventrally localized samples.

Conclusions and relevance—Endoscopic scoring of mucus accumulation is a
reliable clinical and research tool. In contrast, apparent viscosity, localization and color

scores Should be interpreted with caution.

35

Introduction

Endoscopy is an important tool for both clinical work-up and scientiﬁc
investigation of equine lower airway disease. Endoscopic scoring systems have been used
for subjective, semi-quantitative assessment of accumulation [5, 46] and appearance [46,
167] of airway secretions.

However, both the observer and the subject may be important sources of variance
in such scoring systems. For example, “a few” vs. “many” mucous ﬂakes may not mean
the same to different observers or a horse may cough before an endoscopic exam and
thereby momentarily clear mucus from its trachea. Moreover, relationships between
subjective endoscopic scoring variables (e.g., accumulation and apparent viscosity
scores) and corresponding measured variables (e.g., mucus volume and Viscoelasticity)
are unknown. The lack of validation of endoscopic mucus scores makes interpretation of
results, comparison between studies and communication between clinicians difﬁcult and
uncertain.

The goal of the present study was to evaluate the validity of endoscopic scores
used to assess mucus accumulation and character. Speciﬁcally, we asked the following
questions:

1. Does intra- and interobserver variance limit the usefulness of endoscopic

scores?

2. How much do endoscopic scores vary within a horse over the course of a day?

3. Does mucus accumulation score correlate with neutrophilic airway

inﬂammation?

4. Does mucus accumulation score correlate with measured mucus volumes?

36

5. Do accumulation, apparent viscosity, and color scores correlate with measured
Viscoelasticity?

6. Is the localization of tracheal mucus related to its Viscoelasticity?

Materials and methods

Deﬁnitions—In this paper, the terms “mucus (accumulation)” and “(airway)
secretions” are used interchangeably to describe the gel-like mucous secretions with all
their components. The horses with recurrent airway obstruction (RAO) that were used in
this study consistently developed respiratory distress (increased breathing effort and rate)
and cough when stabled and exposed to hay, in accordance with workshop consensus
deﬁnitions [11]. Inﬂammatory airway disease (IAD) in this paper does not refer to the
condition in young racehorses but to older “chronic coughers”, in accordance with an
expanded deﬁnition proposed at the International Workshop on IAD [12]. When stabled
and exposed to hay, the IAD-affected horses coughed, but did not develop clinically
manifest respiratory distress, while the healthy control horses showed no clinical signs.
All horses in the study showed no clinical signs when at pasture and/or in low-dust
indoor environment (peat moss bedding and haylage feeding). Experimental studies were
approved by respective university committees for animal use of the University of Beme
and of Michigan State University.

Overview of experimental protocols —Questions 1-3 (see Introduction) were
addressed in the ﬁrst experimental protocol, question 4 in the second, question 5 in the
third, and question 6 in protocol 4.

Protocol 1: observer and horse variation and association with airway

inﬂammation—Endoscopic examinations (Olympus1 CF-0140L) were performed in 9

37

horses (2 without clinical signs, 4 with IAD and 3 with RAO; 9 to 24 years old) in the
morning (0 hours), afternoon (6 hours), evening (12 hours) and on the following morning
(24 hours). One day before and during the experiment, horses were not exercised, were
kept in-doors bedded on shavings and were fed concentrate and/or good-quality hay.
Each endoscopic examination was recorded on videotape (Sonyz DVCAM DRS-
20MDP) and video clips were digitized (DVgate Motion (acquired in .avi)- version
2.2.00 DVgate Assemble (convert to .mpeg) - version 2.2.00 [copyright Sony2 Corp.
2000]).

Tracheo-bronchial secretion (TBS) collection was performed at 0, 6, 12 and 24
hours by introducing a teﬂon-coated PCV catheter through the working channel of the
endoscope, instilling 10 ml of phosphate-buffered saline (PBS) through the catheter in
front of the tracheal bifurcation and immediately re-aspirating the PBS mixed with
secretions. From the recovered ﬂuid, 100 and 200 uL aliquots were centrifuged for
cytospin preparations. At the end of the experiments (24 hours), bronchoalveolar lavage
ﬂuid (BALF) was collected through the working channel of the endoscope, which was
wedged in a peripheral bronchus [15]. Six 50-ml aliquots of PBS were infused into the tube
and recovered by suction. The lavaged ﬂuids were pooled. 100 and 200 uL aliquots were
centrifuged for cytospin preparations. Total cell counts in BALF were performed manually
using a hemacytometer. After staining with May-Gritnwald-Giemsa, differential cell
counts were obtained from TBS and BALF cytospin preparations. Differential cell counts
were expressed as percent of total cells by counting 200 cells using standard morphologic
criteria under a light-microscope. Absolute numbers of BALF cells were calculated from

total cell counts and the proportion of each cell type.

38

The digitized video clips of the endoscopic examinations were viewed on
personal computer screens and scored for mucus accumulation, apparent viscosity,
localization and color according to the continuous scales scoring system illustrated in Fig.
lA-D. Scoring was performed in randomized order independently by three blinded
observers, and repeated by each observer at least three weeks after the ﬁrst scoring series.

Protocol 2: mucus accumulation scores and measured volumes of “artificial
mucus ”—In order to relate endoscopically scored mucus accumulation to measured
volumes of mucus, “artiﬁcial mucus” made from Jell-O® (Kraft Foods3) was deposited
with a syringe and attached catheter into a full-length excised horse trachea, which was
rinsed before each procedure. The endoscope was introduced into the cranial end of the
excised trachea. Increasing volumes of “artiﬁcial mucus” were deposited during
continuous endoscopic observation of the trachea’s full length. Deposited volumes (mL)
of “artiﬁcial mucus” were recorded when the endoscopically observed accumulation
corresponded to scores 0, 0.1, 1, 2, 3, 4 and 5 assessed during the endoscopic procedure
by observer 1 according to Fig. 1A. A score of 0.1 was one single small, but
endoscopically visible bleb of “artiﬁcial mucus”. The procedure was performed three
times.

Protocol 3: Mucus accumulation, apparent viscosity and color scores and
measured viscoelasticity-Fiﬁy—six endoscopic scores of mucus accumulation, viscosity
and color were correlated to corresponding measurements of mucus Viscoelasticity in
clinically healthy and RAO-affected horses in an environmental challenge protocol.
Brieﬂy, 7 RAO-affected (9 to 26 years old) and 7 healthy control (7 to 26 years old)
horses were kept at pasture and their diet was supplemented with pellets until all animals

had no clinical signs of airway obstruction. Endoscopic mucus accumulation, apparent

39

viscosity and color scores and samples of mucus were obtained at baseline (0 hours) and
6, 24 and 48 hours after environmental challenge by stabling in stalls with straw bedding
and feeding hay. Mucus was obtained from the ventral part of the trachea by means of a
cytology brush passed via an endoscope. After sampling, the brush was retracted 2-3 cm
into the working channel and the endoscope was withdrawn from the animal. The brush
was again protruded and mucus removed from the brush to be immediately stored under
light mineral oil at -80°C. The magnetic microrheometer technique was used to measure
the bulk viscosity and elasticity as described [168]. Mucus accumulation, apparent
viscosity and color were scored during the endoscopic procedure by observer 1 (VG)
according to Fig. 1 A, C and D, respectively. Viscoelasticity data during environmental
challenge can be found in Gerber et al. (2000 [169]). The associations of mucus
accumulation, viscosity and color scores with measured Viscoelasticity have not been
previously reported.

Protocol 4: mucus Viscoelasticity of samples from the ventral and the dorsal
trachea—The relationship of the localization of mucus to its Viscoelasticity was
investigated by comparing Viscoelasticity of samples collected ventrally to samples
concurrently collected dorsally in the tracheae of six IAD-affected horses (5 to 16 years
old). A dorsal and a ventral sample each were collected separately in randomized order
with a clean, dry cytology brush passed via an endoscope. Mucus samples were stored at
-80 C until analysis by magnetic microrheometry as outlined above and described
previously in detail [168]. We report Viscoelasticity (G* = vector sum of viscosity and

elasticity) as log G* at 10 radian/s; dyn/cmz.

40

Statistics—Statistical analyses were performed using SigmaStat for Windows,
version 2.038 statistical software by SPSS4 and NCSS5 6.0.22. After accepting normal
distribution of scored data (Martinez-Iglowicz test for normality), data from protocol 1
were analyzed by means of a four-way ANOVA according to the model:

Y=u+H+O+HO+T+HT+OT+Error
where Y was the response variable (e. g., accumulation score), it was the overall mean, H
was the random effect of Horse, 0 was the random effect of Observer, T was the random
effect of Time, HO, HT and OT were interactions of those random factors, and Error was
the error due to repetition (within observer). Variance (S2 = standard deviation2 = oz) was
calculated by setting the Estimated Mean Square equal to the Mean Square calculated by
ANOVA, and solving for 02. Because there was no signiﬁcant effect of time when it was
considered a ﬁxed factor, it was considered a random factor for the purpose of calculating
oz. Between-observer variance (0'20) is reported as the sum of the variances due to
observer, horse*observer, and time*observer. Within-observer variance (repetition, 02R)
is reported as 62mm. Within-horse variance over time (GZT) is reported as the sum of the
variances due to time and time*horse. Between-horse (62H) variance is reported as the
variance due to horse.

Coefﬁcients of variance ( [standard deviation / mean] * 100; %) were calculated
for each of the 3 repeated volume measurements at mucus accumulation scores 0, 0.1, l,
2, 3, 4, 5 in protocol 2. We report the overall mean of the coefﬁcients of variance.

Within- and between-observer correlations for all scores (protocol 1), correlations
of mucus accumulation, viscosity and color scores with mucus Viscoelasticity (overall

and separate by time point; protocol 3), and of mucus accumulation scores with TBS and

41

BALF neutrophil percentages and numbers (protocol 1) were tested with Spearrnan Rank
Order tests. Mucus accumulation scores were correlated with TBS (sampled at 6, 12 and
24 hours) and BALF (sampled only at 24 hours) airway neutrophils in the nine horses of
protocol 1. Likely due to local trauma from the repeated procedures, TBS neutrophil
percentages Signiﬁcantly increased at 6, 12 and 24 hours compared to 0 hours (results not
shown). Therefore, mucus accumulation scores were correlated with BALF neutrophil
percentages and absolute numbers (at 24 hours, only time point sampled) and with TBS
neutrophil percentages at 0 hours (since the values at the later time points may have been
inﬂuenced by iatrogenic inﬂammation). Viscoelasticity of dorsal vs. ventral paired mucus
samples (protocol 4) was compared by Wilcoxon Signed Ranks test. Signiﬁcance limit

for all tests was set at P < 0.05.

Results

1 Contributions of observer (between observer variance; 0'20), repetition (within
observer variance; 02R), time (within individual horse variance; 0'21") and horse (between
horse variance; oz”) to the total variance for mucus accumulation, viscosity, localization
and color scores are presented as % of total variance (Table 1). Accumulation scores

20 and 0%,), and,

showed the lowest observer related variance (combined 0
correspondingly, correlated best within and between observers (Table 2; Fig. 2) of all the
tested scores. In contrast, observer-related variance was highest for apparent viscosity
scores (combined 020 and 02R = 80% of total; Table l), with complete disagreement
between observer 2 and the other observers (Table 2). Mucus color and localization

scores showed intermediate observer-related variance (Table 1) and correlations (Table 2;

Fig. 2). Within-horse variance over time was moderate for all scores (Table 1). The range

42

was less than 1 score unit in most horses (Table 3) for apparent viscosity, localization and
color scores, and only two of nine horses showed a range of more than 1.7 score units for
accumulation scores (Table 3).

In the 9 horses, there were signiﬁcant correlations between mucus accumulation
score (mean of all time points, observers and repetitions) and TBS neutrophil percentage
(r2 = 0.74; Fig. 3 A), and BALF neutrophil percentage (r2 = 0.59; Fig. 3 B) and absolute
numbers (r2 = 0.88; Fig. 3 C).

The experiment with artiﬁcial mucus showed that mucus accumulation scores
were associated with volumes of mucus within the trachea (Fig. 4). The repeatability of
this experiment was good, with a mean coefﬁcient of variance of 3 1% of the 3 repetitions
of 7 measurements at mucus accumulation scores 0, 0.1, 1, 2, 3, 4 and 5.

Endoscopic scores of mucus accumulation, apparent viscosity and color (n = 56;
Fig. 5 A, B and C) were not signiﬁcantly correlated with the corresponding
measurements of mucus Viscoelasticity in clinically healthy and recurrent airway
obstruction (RAO)- affected horses. Conversely, a strong trend (P = 0.051) towards a
Signiﬁcant difference was observed between Viscoelasticity (log G* at 10 radian/s;
dyn/cmz) of mucus samples collected ventrally (2.0 i 0.9) and dorsally (2.4 i 1.1) in the
trachea (n = 6 paired samples; Fig. 5 D). This difference (0.4x log units) corresponds to a

2.5-fold difference in Viscoelasticity on a linear scale.

Discussion

Mucus accumulation scores were the most reliable and informative of the
evaluated scores: highly reproducible within and between observers (Table 1 and 2; Fig.
2), with moderate variance over time within horses (Table l and 3), and related to

43

measured volumes of mucus (Fig. 4). The use of the illustrated scoring scale (Fig. 1 A)
and endoscopic video clips, which could be watched repeatedly and in a quiet
environment, likely optimized agreement within and between observers. The present
results support the validity of previous ﬁndings of mucus accumulation scoring [5,46,47].

The available data [5,46,47] indicate that “no mucus or multiple small blobs”
(accumulation score < 2; Fig. 1A) can be regarded as normal. Increased mucus
accumulation, in contrast, is associated with coughing [5, 50, 51] and with poor racing
performance [33]. Mucus accumulation scores are not speciﬁc for RAO. They can be
observed in various lower airway diseases [5]. It is also interesting that in clinically
healthy horses, we found no increase of accumulation scores with stabling and no
correlation with airway inﬂammation [47].

In contrast, mucus accumulation scores correlated well with TBS neutrophil
percentages and BALF neutrophil numbers in the 9 horses of protocol 1 that evenly
represented the severity Spectrum of non-infectious equine lower airway disease.
Neutrophilic inﬂammation may cause mucus accumulation in horse airways through
increased production and secretion of mucins [135, 170] and/or by unfavorably altering
physical properties and clearability of the secretions [171, 172].

Unfortunately, measurements of Viscoelasticity and other physical properties that
inﬂuence clearability are technically demanding and often unavailable. Therefore,
reliable endoscopic scoring systems that correlate with mucus physical properties could
be useful tools for clinical and research purposes. Scores of apparent viscosity are
reported to correlate with measurements of lung function [167], and were found to
decrease with oral N-acetylcysteine treatment in horses with RAO in a blinded study
[173].

44

In the present study, within-horse variance over time of mucus viscosity,
localization and color scores was moderate (Table 1) and the range was less than 1 score
unit in most horses (Table 3). Mucus viscosity scores did not correlate with measured
Viscoelasticity (Fig. 5B) and the combined within- and between—observer variance was
80% of the total (Table 1) with complete disagreement between observer 2 and the other
observers (Table 2). Similarly, mucus color did not correlate with measured
Viscoelasticity (Fig. 5C) and, possibly due to different monitors used by the observers for
scoring, showed quite high observer-related variance (Table 1 and 2). It is important to
note here that other physical properties such as spinnability and adhesiveness of mucus
that are also important determinants of clearability [69] were not measured or correlated
with endoscopic scores in this study. Also, it may still be possible to subjectively
differentiate the extremes recorded in some of the scores (Fig. 4) despite relatively high
observer-related variance. Regardless of the scores assessing mucus quality, only
localization is somewhat supported by the results of this study. Dorsally vs. ventrally
located mucus showed a 2-fold difference of Viscoelasticity (Fig. 5D), and these extremes
of localization can be endoscopically differentiated (Tables 1 and 2; Fig. ZB).

In conclusion, this study showed that endoscopic scoring of mucus accumulation
is a reliable clinical and research tool. Due to high observer-related variance, however,
endoscopic scores of mucus viscosity, localization and color should be interpreted with

caution.

45

Table 2-1: Variance attributable to observers (030), repetitions (0'23), time (021), and
horse (01") for mucus accumulation, localization, apparent viscosity and color

scores

 

Variance (sd2 = 0’) Percent of total variance

 

Response Variable Mean 0'20 0'2 R 011' 0'2“ 0'20 013 011‘ 0'2"

 

Accumulation 2.14 0.12 0.20 0.30 1.28 6% 10% 16% 68%

 

Localization 2.03 0.10 0.26 0.15 0.32 12% 32% 17% 39%

 

Apparent viscosity 2.66 0.32 0.24 0.07 0.08 45% 34% 10% 11%

 

Color , 3.34 0.38 0.32 0.23 0.16 35% 29% 21% 15%

 

 

 

 

 

 

 

 

 

 

 

 

46

Table 2-2: Spearman Rank Order correlation within and between observers

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1 2 3
Accumulation scores (0-5) .
l 0.86; 36 0.85; 36 0.77; 36
2 - 0.88; 36 0.90; 36
3 - - 0.71; 36
Localization scores (1-5)
1 0.27; 36 0.20; 35 0.44; 36
2 - 0.59; 35 0.34; 36
3 - - 0.58; 36
Apparent viscosity scores (1-5)
1 0.46; 36 0.08; 36 (NS) 0.42; 36
2 - 0.35; 35 0.03; 36 (NS)
3 - - 0.44; 36
Color scores (1-5)
1 0.59; 36 0.45; 35 0.52; 36
2 - 0.53; 35 0.56; 35
3 - - 0.62; 36

 

Correlations are based on means of the two repetitions for each observer. Matrix: observers 1

(VG), 2 (ER) and 3 (AI). Each cell shows Spearman rank order r2; number of valid observations.

All correlations are signiﬁcant, except those marked NS, not signiﬁcant.

47

 

Table 2—3: Means :1: SD (0, 6, 12, and 24 hours), ranges (all time points) and overall
means :1: SD (all time points) of all observers and repetitions for mucus
accumulation, localization, apparent viscosity and color scores for each horse.

 

48

Fig. 2-1: Scoring system for mucus accumulation (A), localization (B), apparent

viscosity (C) and color (D).

A: Accumulation

      

,3” ’i ' .
draw

1 I vr-. _ ,r ' ' . :
Hr. mv-‘w—x "1‘?“le

 

1234567895

1 L j 4
j

0123456789]123456789212345678931134567894
a l I l l l 1

none little moderate marked large extreme
clean singular, multiple small blobs larger blobs, conﬂuent stream fonning pool forming profuse amounts

8: Locallzatlon and stickiness C: Apparent viscoslty

12453214683246I424635

v. llaid fluid intermediate viscous v. viscous

D: Color

124632246332463424635 124682246832468424685

r a I ‘r

 

‘— I

ventral lateral dorsal threading yellow white colorless

49

Fig. 2-2: Intraobserver (observer 1, round 1 and 2) correlations for mucus accu-

mulation (A), localization (B), apparent viscosity (C) and color (D) scores.

Score - Round 2

Score - Round 2

 

 

A: Accumulation

 

 

’e..
e
'8
e
0...:
. 3
'e
e
.00
. a
so
a. 9:036
1 2 3 4

 

 

 

 

3 as as
P=OA6
2 3 4

Score - Round 1

50

 

 

 

 

 

 

 

 

B: Localization .
O
O O
O : O O
I O
O
O
O O O
O C.
O O. . I O
O
'3” ', 8:027
1 2 3 4 5
D: Color
0 O O O
O
.‘O 0
O O O
O O O
O
. O O
O
O O I O
O O
l O
f=059
. I Y
1 2 3 4 5

Score — Round 1

Fig. 2-3: Associations of mucus accumulation scores with tracheo-bronchial

secretion (TBS) neutrophil percentages (A) at 0 hours, and with bronchoalveolar

lavage ﬂuid (BALF) neutrophil percentages (B) and. BALF neutrophil absolute

numbers (C) at 24 hours.

 

 

 

 

 

 

 

 

3‘? 100
8 so '
5
C 0
g so
a
E ‘°‘ .
O o
g 20 .
a 8:014
:3 0 o a . ' I
E
0 1 2 3 4 5
Accumulation score

a 10000
a C
2 , o 0
3 1000
E . '
g 100 . .

1 ,
2 ° '
u. r’=o.sa
"' 1
< . .
m o 1 2 3 4 5

Accumulation score

51

BALF neutrophil percentages (%)

100

60~

40‘

20

 

 

 

 

1 2 3 4 5

Accumulation score

Fig. 2-4: Relationship between mucus accumulation score and volumes of “artificial

mucus” (Jen-0m).

50

Measured amounts of “artificial mucus” (ml)

 

Accumulation score

52

Fig. 2-5: Relationships between mucus accumulation (A), apparent viscosity (B) and
color (C) scores and measured Viscoelasticity (log G* at 10 radian/s; dyn/cmz) and
trend (P = 0.051) towards a significant difference between Viscoelasticity (D) of

mucus samples collected ventrally vs. samples collected dorsally in the trachea at the

 

 

 

 

 

 

 

 

 

same time.
3.6 - o
«A 3,4 l A 3 6 B .
E 3 2 ‘ Q . “T" 34 .
2 i j e . . E 3'2 .
‘5'. 3'0 . I . ° 3 o ' .
3 2.8 ~ . . f 2 2‘8 1. I g .
. 2.6 - ' ; . ° 3 ' . r a
o 24 ° ' 1 ' . 2'6 I I
3 2'2 . . o : o 2.4 . . ' 0 .
J ' Q . O . . I 22 . :
2.0 . . . . § 2 0 I
1.8 1 . . 1 8 o O 8
0 1 2 3 4 5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Accumulation score'
Apparent viscosity score
4 .. w- .__...._..__
35 1 ' D 1
34 « C «3"
('7‘ ' O . E i
g 3.2 . o 0 3 i
E 3.0 i 0 ! . g
g. 2 8 l ; . t
.. 2.6 - ’ ° " 2 I
. ° ' o o 0 . l
0 2.4 ‘ i O . o l
J 18 i ' ' l l l
' ’ - l P - 0.051 7 '
- __.. -_.__.___. o '
1.0 1.5 2.0 2.5 3.0 3.5 40 4.5 5.0 ventral dorsal
Color score Localization of mucus in trachea

53

CHAPTER 3
AIRWAY MUCUS ACCUMULATION-COMPARISON OF RAO-AFFECTED
AND CLINICALLY HEALTHY CONTROL HORSES BEFORE AND DURING

ENVIRONMENTAL CHALLENGE

Gerber, V., Lindberg, A., Berney, C., Robinson, N.E. Airway mucus in RAO — short-
term response to environmental challenge. (2003). Accepted for publication in J Vet Int
Med.

In the previous study (chapter 1) I showed that endoscopic scoring is a reliable
tool to quantify airway mucus accumulation in horses.

In order to test the hypothesis that airway mucus accumulation is a function of
disease status and/or environment, I compared mucus accumulation scores in RAO-
affected and clinically healthy control horses before (when they were. at pasture) and after
(6, 24 and 48 hours) stabling on straw bedding and hay feeding.

Several alternative outcomes were possible: No difference of mucus accumulation
scores between RAO-affected and control horses and no difference before and after
stabling. Signiﬁcant effects of disease group (i.e., difference between RAO and controls)
and/or environment (i.e., difference before and aﬁer stabling); signiﬁcant interaction
effects of disease group and environment (i.e., change in mucus accumulation with

stabling in one group but not the other).

54

Abstract

Mucus accumulation and neutrophilic inﬂammation in the airways are hallmarks
of RAO. Endoscopically visible mucus accumulations, however, have not been studied
during exposure to dusty hay and allergens (i.e., environmental challenge). We
hypothesized that a) RAO-affected horses have increased mucus accumulation compared
to controls, b) mucus accumulations increase in RAO-affected horses during
environmental challenge, and 0) environmental challenge also induces neutrophilic
inﬂammation and mucus accumulation in control horses. Mucus accumulation was
graded endoscopically (mucus grades [MG] 1-5) and airway inﬂammation was evaluated
by bronchoalveolar lavage ﬂuid (BALF) cytology before (0 hours) and during (6, 24, 48
hours) environmental challenge. Large amounts of mucus (MG 4—5) were speciﬁc for
RAO-affected horses in this study. Variation among controls was considerable, however,
and intermediate grades (MG 2-3) were non-speciﬁc, showing complete overlap between
the 2 groups. Mucus accumulations [median (25“, 75th percentile)] increased in RAO-
affected horses from 2.5 (1.5, 3.5) MG at baseline to 3.5 (2.0, 4.0), 4.0 (3.0, 4.0) and 4.0
(4.0, 4.0) MG at 6, 24, and 48 hours, respectively. MG did not increase in controls,
overall 1.0 (1.0, 2.0) MG, even though controls also showed a moderate increase of
BALF neutrophils. In conclusion, mucus accumulations, before and especially after
exposure to dust and allergens, are increased in RAO-affected horses compared to
controls. Healthy controls show considerable variability in mucus accumulation, but
despite an inﬂux of neutrophils into the airways, no increase of mucus accumulation after

exposure to hay dust.

55

Introduction

Mucus accumulation and neutrophilic inﬂammation in the airways are hallmarks
of recurrent airway obstruction (RAO) [3]. The changes in BALF cytology in response to
stabling and hay feeding (i.e., environmental challenge) in RAO have been well
characterized [15]. Most prominent is neutrophilic inﬂammation, and neutrophils
accumulate in the lung within 7 hours of environmental challenge [14]. Increases in
BALF neutrophils during environmental challenge, however, are not limited to RAO
horses, but also have been reported in clinically healthy animals [47, 174]. In other
species, airway inﬂammation, especially neutrophils and their products, can cause
increased mucin gene expression and mucus hypersecretion [135, 150, 175, 176] as well
as unfavorable changes in mucus rheology [171, 177]. Changes in mucus rheology may
lead to decreased clearance of secretions as reported in RAO [58, 64]. Both increased
production and decreased clearance can lead to mucus accumulation in the airways. To
date, however, the amount of endoscopically visible mucus accumulations in airways
before and during environmental challenge has not been investigated in horses. We
hypothesized that RAO-affected horses have increased mucus accumulation compared to
controls and that mucus accumulations increase during environmental challenge. We
further hypothesized that environmental challenge also induces neutrophilic airway
inﬂammation and mucus accumulation in clinically healthy control horses. This study
investigates changes in mucus accumulation and BALF cytology in RAO-affected and

control horses before (0 hours) and during (6, 24, 48 hours) environmental challenge.

56

Materials and methods

Animals and study design -The protocol was approved by the All-University
Committee for Animal Use and Care of Michigan State. University. Twelve horses (6
mares and 6 geldings), 9 to 25 years old (18.1 i 5.2 years, mean :1: SD), with a previous
diagnosis of RAO and manifesting respiratory distress during environmental challenge
served as the RAO group. Eleven horses (4 mares and 7 geldings) aged 7 to 27 years
(15.2 i 7.9 years, mean :1: SD) and without clinical signs of respiratory tract disease
before and during environmental challenge, served as the control group. The horses in
both groups were selected and monitored according to criteria based on history and on
numerically graded clinical signs of respiratory distress as previously described [178].
Both groups were drawn from closed research herds in the Department of Large Animal
Clinical Sciences, Michigan State University that develop (RAO group) or do not
develop (control group) signs of obstructive pulmonary disease when. stabled and fed hay.
All horses were vaccinated against equine inﬂuenza, rhinopneumonitis, and Eastern and
Western equine encephalitis. Horses were dewonned within 2 months of the study. The
study was performed between the months of August and October. Before entering the
study, the horses were kept at pasture and their diet was supplemented with pellets when
grass was sparse until all animals showed no clinical signs of airway obstruction. Horses
were transported to the laboratory in an open-sided stock trailer for less than 10 minutes
and without their heads tied. Clinical assessment, endoscopic mucus grades (MG) and
BALF were obtained at baseline (0 hours) and after 6, 24, and 48 hours of environmental

challenge by stabling in stalls with straw bedding and feeding of dusty hay.

57

Bronchoalveolar lavage ﬂuid (BALF) sampling and endoscopic mucus
scoring—BALF was collected, processed and analyzed according to guidelines set forth in
the International Workshop on RAO [1 1]. Brieﬂy, the endoscope (Olympus GIF 300,
custom made) was directed to a different site within the lung for each collection (0, 6, 24,
and 48 hours). After the endoscope was wedged in an airway, 3 IOO—ml aliquots of
phosphate-buffered saline were instilled into the airway and then aspirated. The aspirates
were pooled and mixed. Total cells in BALF were counted by use of a hemacytometer.
Differential cell counts were made on 200 cells of a cytocentrifuge preparation stained
with Diff-Quick (Dade Bering) and performed by one observer (Lindberg). Total
neutrophil counts were calculated from total and differential cell counts.

The amount of mucus visible in the trachea was graded on a validated endoscopic
scale of 0—5 MG [179]: 0 (no visible mucus), 1 (singular small blobs), 2 (multiple blobs
only partly conﬂuent), 3 (mucus ventrally conﬂuent), 4 (large ventral pool), 5 (profuse
amounts of mucus occupying more than 25% of tracheal lumen). The entire trachea was
assessed and gradings were determined according to the largest amount observed in any
part of the trachea, which usually was the most caudal aspect cranial to the biﬁircation.
Two observers (VG and AL) graded MG in 7 RAO-affected horses and 6 controls and in
5 RAO-affected horses and 5 controls, respectively.

Statistical analysis-All statistical analyses were performed using the SAS System
for Windows, version 8.02. MG (dependent measure of interest) and were analyzed by
non-parametric repeated measures analysis of longitudinal data (SAS-macros).
Independent factors included 2 levels each of the between subjects factors, observer (VG

vs. AL) and group (control vs. RAO horses), and 4 levels (0, 6, 24, and 48 hours) of the

58

repeated measure, time. Subanalyses were performed using Wilcoxon-sign-ranks test (for
paired comparison within groups over time) and Kruskal-Wallis (Wilcoxon-rank-surn; for
non-paired comparison between groups) when main effect, interaction effect, or both
were signiﬁcant. BALF total cell counts and differential cell percentages and numbers
(total cell count x percentage of each cell type) did not satisfy the model assumptions of
equal variance and normal distribution for parametric analysis and therefore were
analyzed like the MG data, except that there was no observer factor. Sensitivity [true
positive / (false negative + true positive)] and speciﬁcity [true negative / (false positive +
true negative)] of low MG values (<2) for clinically healthy horses and of high MG
values (>3) for RAO-affected animals were calculated. All data were expressed as
medians and quartiles (i.e., 25th, 50th and 75th percentiles). Signiﬁcance limit was set at P

< 0.05.

Results

Endoscopic MG—A complete set of observations (n = 92) of MG was obtained
and summary data are presented in Table l. Mucus accumulations over time also are
illustrated in Fig. 1A, and the overall distributions in both groups are shown as
histograms in Fig. 2. There was no effect of the observer (VG vs. AL), but a signiﬁcant
difference in MG between control and RAO-affected horses (main group effect), as well
as signiﬁcant time and interaction effects. Subanalysis determined that MG were different
between groups at time points 6, 24 and 48 hours, and showed a strong trend (P = 0.054)
for a difference at baseline (0 hours). MG signiﬁcantly increased over time in RAO (6,

24, and 48 hours differed from baseline), but not in control horses. Overall (Fig. 2),

59

observations of “no mucus or a few single mucus droplets” (MG O-l) had high speciﬁcity
(0.92), but low sensitivity (0.52) for clinically healthy controls. Large amounts of mucus
(MG 4-5) also showed high speciﬁcity (0.98), but low sensitivity (0.56) for RAO. In
contrast, intermediate MG ("multiple only partly conﬂuent blobs to ventrally conﬂuent
mucus"; MG 2-3) completely overlapped between clinically healthy and RAO-affected
horses.

BALF cytology changes and relationship with tracheal mucus grades—A
complete set of BALF samples (n = 92) was available for cytological evaluation.
Summary data of BALF cytology (total cell counts and differential cell percentages) are
presented in Table 1. BALF differential cell percentages, total cell counts and differential
cell numbers (total cell count x percentage of each cell type) were analyzed statistically.
There was a signiﬁcant effect of group, time, and interaction on BALF lymphocyte,
neutrophil, macrophage, mast cell, and epithelial cell percentages, respectively. Results
of differential cell percentages subanalysis are presented in Table 1. There was a
signiﬁcant effect of group, time and interaction on BALF total cell counts and on
neutrophil cell counts. Subanalysis determined that total cell counts and neutrophil cell
counts were different between groups at time points 6, 24, and 48 hours, but not at
baseline (0 hours). Total cell counts increased signiﬁcantly over time in RAO (6, 24, and
48 hours differed from baseline), but not in control horses. Fig. 13 illustrates how
neutrophil numbers increased signiﬁcantly over time in RAO (6, 24, and 48 hours
differed from baseline) and also in control horses (24 and 48 hours differed from
baseline). Further, there was a signiﬁcant effect of time on BALF epithelial cell numbers,

due to a decrease of epithelial cell numbers at 24 hours compared to baseline in control

60

horses. No main or interaction effects were observed for lymphocyte, eosinophil, mast

cell, and macrophage numbers.

Discussion

RAO-affected horses had more mucus accumulation at all times compared to
control horses (Fig. 1A). Even when at pasture, removed from allergen exposure and in
clinical remission, RAO-affected horses showed a strong trend for increased mucus
accumulation. We have previously demonstrated persistent alterations of mucin-side-
chain-associated carbohydrates (increased concentrations of al,2-fucose) in BALF
during RAO-remission [54]. These complementary ﬁndings suggest that in RAO the
mucus apparatus undergoes changes that persist even during clinical remission when
clinical signs, bronchospasm, and inﬂammation have waned or are undetectable. Within 6
hours of exposure to the stable environment, however, mucus accumulation further
increased in RAO-affected animals (Fig. 1A). Overall (all time points; Fig. 2), we noted
that none of the RAO-affected horses were without visible mucus at any time, and only
8% of the observations showed just singular small blobs. More than two thirds (77%) of
the endoscopic evaluations showed completely conﬂuent secretions or more mucus (MG
3—5) with large ventral pools seen in almost half of the observations. In contrast, only 1
control horse had a large ventral pool and none showed profuse amounts of mucus
occupying more than 25% of the tracheal lumen. Still, almost half of the observations in
controls (21 of 44; 48%) showed more than a few blobs of mucus (MG > 1). This
demonstrates that although controls had less mucus accumulation than RAO-affected

animals, the amount of tracheal secretions can vary much more in clinically healthy

61

horses than previously reported [5]. Our results, however, did not support the hypothesis
that mucus accumulation would increase with stabling in control horses. No change in
mean MG was observed over time (Fig. 1A).

The distribution of mucus accumulation in both groups (Fig. 2) thus showed that
“no mucus or a few single mucus droplets” in the trachea (MG 0-1) rarely is seen in RAO
horses and can be regarded as normal. Large amounts of mucus (MG 4-5), on the other
hand, had a high speciﬁcity for RAO in the 2 groups we studied. Intermediate MG
("multiple only partly conﬂuent blobs to ventrally conﬂuent mucus"; MG 2-3), however,
were a non-speciﬁc ﬁnding, for which clinically healthy and RAO-affected horses
completely overlapped.

At baseline, BALF neutrophil percentages were not signiﬁcantly different in the
RAO group, 4.3 (2.4, 15.1) as compared to controls, 5.0 (1.6, 6.0). The 75th percentile
shows, however, that increased neutrophil percentages were present in some RAO-
affected horses at baseline, even though all horses were in clinical remission when they
entered the study. Within 6 hours of environmental challenge, neutrophils increased to
over 100 per uL in RAO-affected horses, and remained at those concentrations by 24 and
48 hours after stabling. In control horses, BALF neutrophils also increased during
environmental challenge (Fig. 1B, Table 1). Similar degrees of neutrophilic inﬂammation
in response to stabling have been described previously in clinically healthy horses [47,
180]. When compared to normal values proposed by the International Workshop on IAD
[12], in all horses, one or more variables (visible mucus accumulation, percent
neutrophils, eosinophils or mast cells; means in Table 1) over time could be considered

abnormally increased, and thus signs of IAD. Also, mast cell, lymphocyte and

62

macrophage percentages (but not absolute numbers) signiﬁcantly decreased in RAO
horses aﬁer stabling (Table 1). This ﬁnding likely represented a relative change due to
the large increase in neutrophils in this group. There was. no evidence that the increases
“above normal values” and relative changes in these cell types were related to RAO or
parasitism. IAD is as yet an ill—deﬁned syndrome, and it seems cautious not to over-
interpret our ﬁndings, especially because neutrophilic inﬂammation in controls could
have been caused partly by the repeated procedures. Nevertheless, the present results
agree with those of previous studies [9, 10, 47, 180] that indicate that subclinical levels of
IAD are present in a majority of stabled horses. The signiﬁcance of these subclinical
levels of IAD, which may be a “normal” (i.e., common) reaction to the high dust loads
that horses are exposed to in a conventional stable environment [6, 8], warrants further
investigation.

The repeated BAL procedures potentially could have inﬂuenced not only BALF
cytology but also MG. Possible inﬂuences could be “washing out” of mucus, iatrogenic
inﬂammation caused by the BAL procedures and, because the observers were not blinded
to the diagnoses of the horses, hiased MG. However, MG in the present study agree with
those observed when mucus was graded by a blinded observer in 6 control and 6 RAO
horses before (1.8 at 0.9 and 2.2 3: 1.1, respectively) and after 3 days (1.6 i 1.1 and 3.5 i
7.6, respectively) of stabling, without any invasive procedure between these
measurements (Robinson, unpublished data).

In conclusion, both mucus accumulations and BALF neutrophils increase in
RAO-affected horses and are considerably higher than in control horses. The distribution

of mucus accumulation in this study (Fig. 2) suggests that “no mucus or a few single

63

mucus droplets” in the trachea (MG 0-1) can be regarded as normal. Large amounts of
mucus (MG 4-5), on the other hand, were speciﬁc for RAO in this study, but could
certainly also be caused by other airway diseases in a. random population of horses.
Clearly, intermediate amounts of mucus ("multiple only partly conﬂuent blobs to
ventrally conﬂuent mucus"; MG 2-3) were a non-speciﬁc ﬁnding, for which clinically
healthy and RAO-affected horses completely overlapped. Persistent up-regulation of
mucus production in remission and increased accumulation of secretions in exacerbation
(likely due to decreased clearance and increased secretion) appear to characterize the

mucus apparatus in RAO.

Table 3-1: Mucus grades and bronchoalveolar lavage fluid cytology in RAO-affected
horses and control horses before (0 hours) and during (6, 24, 48 hours)
environmental challenge: medians with quartiles (25“, 75"I percentiles) for mucus
grade (MG), total cell count (tcc; x cells/pl), percentages of lymphocytes (%lym),
macrophages (%mac), neutrophils (%neu), eosinophils (%eos), mast cells (%mas),

and epithelial cells (%epi).

 

RAO-affected horses (n = 12)

 

hours 0 6 24 48
MG 2.5 (1.5, 3.5)+ 3.5 (2.0, 4.0)*# 4.0 (3.0, 4.0)*# 4.0 (4.0, 4.0)*#
tcc 73.2 (39.8, 123.8) 210.0 (130.0, 260.8 (161.7, 176.5 (119.5,
%lym 48.5 (36.6, 64.8) 14.1 (10.8, 36.2)*# 14.3 (8.3, 27.8)*# 10.8 (4.2, 19.3)*#
%mac 29.9 (22.3, 39.3) 10.4 (6.3, 16.5)*# 9.3 (4.8, 13.0)*# 14.5 (6.3, 20.3)*#
%neu 4.3 (2.4, 15.1) 75.1 (21.8, 87.6)*# 75.3 (56.3, 85.0)*# 69.5 (55.5, 81.5)*#
%eos 0.0 (0.0, 0.2) 0.0 (0.0, 0.5) 0.0 (0.0, 0.3) 0.0 (0.0, 0.5)
%mas 5.2 (3.0, 7.0) 1.5 (0.6, 4.0)*# 0.9 (0.3, 3.0)* 1.0 (0.0, 6.0)
%cpi 0.5 (0.0, 1.7) 0.0 (0.0, 2.5) 0.0 (0.0, 0.2)* 0.0 (0.0, 1.7)
Control horses (n = 11)
MG 1.0 (1.0, 2.0) 2.0 (1.0, 2.8) 2.0 (1.0, 2.8) 1.0 (1.0, 2.0)
tcc 61.0 (20.3, 95.6) 41.7 (19.4, 95.6) 67.5 (27.9, 118.1) 95.0 (23.9, 96.9)
%lym 42.0 (35.5, 53.2) 52.0 (36.9, 57.9) 46.5 (35.0, 50.4) 47.0 (31.6, 54.7)
%mac 35.0 (28.6, 43.7) 23.9 (17.3, 34.6) 28.0 (19.2, 36.9) 27.0 (20.7, 37.6)
%neu 5.0 (1.6, 6.0) 9.0 (3.9, 19.5)* 16.2 (7.0, 30.7)* 18.0 (15.0, 27.5)*
%eos 0.0 (0.0, 0.93) 0.2 (0.0, 0.2) 1.0 (0.2, 2.6) 0.5 (0.0, 1.0)
%mas 5.8 (2.5, 11.5) 5.5 (4.1, 7.1) 2.2 (1.0, 8.4)* 3.0 (2.6, 10.3)
%epi 1.5 (1.0, 5.3) 0.5 (0.1, 1.9)* 0.0 (0.0, 0.9)* 0.5 (0.0, 0.9)*

 

’“Signiﬁcantly different (P < 0.05) from baseline (0 hours); #signiﬁcantly different 0’ < 0.05)

from control at same time point; +trend toward a difference compared with control at that time

point (P = 0.054).

65

Fig. 3-1: Boxplots of mucus grades (A) and bronchoalveolar lavage fluid neutrophil
numbers (B) in recurrent airway obstruction (RAO)-affected (grey boxes) and
control horses (white boxes) before and 6, 24, and 48 hours after stabling.
*Significantly different (P < 0.05) from 0 hours. #Significantly different from

control at same time point; +trend toward a difference compared to control at that

time point (P = 0.054).

 

Mucus Grades
0.)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0 2 o
0 6 24 48 hours
[:1 controls
A B *ﬁ
.5! 1000 . E] RAO 1 . ﬂ
8 1004 . ' .
E T '2; ' ,, ﬂ
3 ..',.:
C . i. t .
E 104 L i
a *1 --
e » 4.
g 1 * D i o
O
l .
0.1 '

24 48 hours

0
O)

66

Fig. 3-2: Distribution histogram of mucus grades (MGs). Graph shows relative
frequency (%) of each MG (0-5) in controls (clinically healthy horses; white
columns) and in recurrent airway obstruction (RAO)-affected animals (gray
columns). Numbers of observations are listed in the attached table. There was a
significant difference in MGs between control and RAO-affected horses (main
group effect). MG 0-1 had a good specificity (0.92), but low sensitivity (0.5) for
controls. Large amounts of mucus (MG 4-5), on the other hand, showed a high
specificity (0.98), but also low sensitivity (0.56) for RAO. The intermediate MGs 2-3
were a non-specific ﬁnding, where clinically healthy and RAO-affected horses

completely overlap.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

§ 60 A

E [:3 controls (44 observations)

,2 5° [:1 RAO (48 observations) .—

(0

g 40 - r

U)

.Q

53 30

o .—

§ 20 ~

(1)

3

8' 10 H 4

LL 0 ‘l 54 ﬂ
4

o 1 2 '3
Mucus Grades

numberof 121‘ 5 1811 9 1 1

E
1

observations 11: 80’ 7 > 4 7 10 23

l

 

A10

 

67

CHAPTER 4
COMPARISON OF AIRWAY MUCUS ACCUMULATION IN TWO AGE
GROUPS OF CLINICALLY HEALTHY WELL-PERFORMING SPORT

HORSES IN A CONTROLLED ENVIRONMENT

Gerber, V., Robinson, N.E., Luethi, S., Marti, E., Wampfler, B., Straub, R. Airway
inﬂammation and mucus in two age groups of asymptomatic well-performing sport

horses. (2003). Equine Vet. Journal, 35, 491-495.

The previous study (chapter 3) showed that airway mucus accumulation is a
function of disease status, i.e., mucus was increased in recurrent airway obstruction
(RAO)—affected horses overall. In addition, an interaction of disease status and
environment was observed, i.e., mucus increased with exposure to hay and straw in
RAO-affected, but not in clinically healthy horses.

In order to test the hypothesis that airway mucus accumulation is also a function
of age, I compared mucus accumulation scores in two age groups of clinically healthy
well-performing sport horses that lived in the same controlled and constant stall
environment.

Several alternative outcomes were possible: no difference between age-groups;
increased mucus in the older horses, as suggested by research in human smokers that
produce more sputum with increasing pack-years; or, increased mucus in younger horses,

as observed in racehorses entering training facilities.

68

Abstract

Mucus quantity and quality and BALF differential cytology were investigated in
thirteen younger horses (Y) with a mean age of 5 years and 13 older horses (0) with a
mean age of 15 years, without historical or clinical evidence of lower airway disease. The
only differences between the age groups were increased BALF total and lymphocyte cell
counts in the younger horses. We found that clinically healthy older horses, having been
exposed 10 more years to conventional stable environment without developing clinical
signs of lower airway disease, do not show increased subclinical airway inﬂammation or
mucus accumulation. Overall, however, large interindividual variations were observed in
mucus quantity scores as well as BALF neutrophil, eosinophil, mast and epithelial cells,
and the ranges and means of the variables were larger than previously described as
normal. If inﬂammatory airway disease (IAD) is simplistically deﬁned as >10% BALF
neutrophils and/or more than just a few blobs of mucus in the trachea, then the majority
of these asymptomatic horses were suffering from subclinical IAD. Why, then, was no
negative effect noted on the present and past performance in any of the sport horses we
examined? We propose that mild to moderate degrees of airway inﬂammation and mucus
accumulation are a normal (i.e., observed in the majority of individuals) reaction to
conventional stable environment, and that the signiﬁcance of such ﬁndings should be
interpreted according to the level of respiratory performance expected from an individual.
Moreover, a deﬁnition of IAD based on airway inﬂammation and mucus alone without a

measure of airway obstruction and hyperreactivity may be inadequate for clinical

purposes.

69

Introduction

Horses housed permanently indoors, bedded on straw and fed hay (conventional
stable environment) show high incidences of chronic obstructive pulmonary disease
(COPD), 54% [181] and even 79% [10]. However, the term “COPD” as used in these
studies may encompass very variable degrees of lower airway disease. The common
denominators are airway inﬂammation and mucus accumulations. Recently [11, 12],
more rigid diagnostic criteria have been proposed to differentiate the severe airway
disease known as recurrent airway obstruction (RAO) from milder forms of airway
inﬂammation. Brieﬂy, the presence of recurrent and reversible airway obstruction
distinguishes RAO from inﬂammatory airway disease (IAD), which is characterized by
airway inﬂammation, mucus accumulation and hyperreactivity in the absence of dyspnea.

IAD, even in the absence of clinical signs such as cough or abnormal lung sounds,
may negatively affect racing performance [33]. Most studies on IAD have focused on
young racehorses. However, the above-cited surveys [10, 181] suggest that milder
degrees of airway disease are also common among stabled sport horses of various ages. It
has been proposed, but not shown, that subclinical degrees of IAD may affect the
performance of sport horses. We chose to test this hypothesis by determining what
degrees of airway inﬂammation and mucus occur in asymptomatic horses performing
well in dressage and show jumping.

The causes and pathogenesis of IAD are incompletely understood and may
include inhaled irritants as well as infectious agents [53]. There is evidence, that bringing
asymptomatic horses of varying age groups from pasture into a conventional stable

environment causes acute neutrophilic airway inﬂammation in asymptomatic horses [47,

70

180]. In contrast, we showed that mucus accumulations, even though they vary
considerably between asymptomatic horses, do not seem to increase after short-term
stabling [182]. There is no information on the effects of long-term exposure, however. In
smokers, the duration of exposure (i.e., pack-years) to inhaled irritants (i.e., cigarette
smoke causing neutrophilic airway inﬂammation) correlates with mucus accumulation
and viscosity [183]. We hypothesized that similar changes may occur in horses that had
been stabled for years and that even though older horses were asymptomatic, they would
have increased degrees of subclinical lower airway disease, and that mucus accumulation
would be related to neutrophil numbers in the airways. .

We tested the above hypotheses in two age groups of clinically and historically
healthy sport horses that were performing well. All the horses were housed in the same
conventional stable environment. The present study 1) investigated what degrees of
airway inﬂammation and mucus are present in asymptomatic well-performing sport
horses, 2) examined the relationship between airway inﬂammation and mucus, and 3)

compared older to younger individuals.

Materials and methods

Stable environment and management conditions of horses—The Swiss National
Equine Centre is stocked with over 200 Warmblood horses. They typically enter the
facility at 2-4 years of age, and thereafter live permanently in a conventional stable
environment: Horses are bedded on straw and housed 20-22 hours per day in box stalls or
tie stalls. All horses are used daily (except for one resting day per week) for dressage

and/or show jumping, most of them compete regularly. A combination of good quality

71

hay and haylage is fed three times per day along with individual rations of concentrated
feed. Primary health care is provided at the Swiss National Equine Centre, all medical
records are kept with the resident veterinarian.

Selection of horses and design of the study—Initially, 16 horses from among
those younger than 8 years of age and 18 from among those older than 12 years were
randomly chosen. Predeterrnined exclusion criteria were then applied to eliminate horses
with historical and/or clinical evidence of lower airway disease: Historical criteria for
exclusion were based on medical records and included a past history of chronic lower
airway disease for more than three months, and/or infectious lower airway disease in the
previous three months. Clinical exclusion criteria included evidence of lower airway
disease such as cough, mucoid nasal discharge and/or abnormal lung sounds and/or
evidence of infection. Clinical signs were evaluated before, during (coughing) and
immediately after an exercise test consisting of walk, trot and canter for 10 minutes each
(examined by Straub). Evidence of infection was provided by physical examination as
well as complete blood count and serum chemistry proﬁle. Four horses older than 12
years of age were eliminated because of a history of chronic lower airway disease, and 1
horse younger than 8 years had suffered from recent infectious lower airway disease. A
further 3 horses (2 older than 12 years of age, 1 younger than 8 years) were eliminated
from the study because of clinical signs of lower airway disease. Thirteen younger horses
(younger group; Y), 4 ﬁllies and 9 geldings, with a mean age of 5 years (i1, SD) and 13
older horses (older group; 0), all geldings, with a mean age of 15 (i2, SD) qualiﬁed for

the study (Table 1).

72

Bronchoalveolar lavage ﬂuid (BALF) sampling and analysis—BALF was
collected by means of a l97-cm endoscope (CF-VL Video colonoscope, Olympusl) that
was passed via the nares and wedged in a peripheral bronchus. Six SO-mL aliquots of
phosphate buffered saline (PBS) were inﬁrsed into the tube and recovered by suction. The
lavaged ﬂuids were pooled and the total recovered volume determined. Total cell counts
in BALF were performed manually using a hemacytometer. Cell smears were made with
a cytocentrifuge and stained with Diff-Quick. Differential cell counts of lymphocytes,
macrophages, neutrophils, eosinophils, mast cells and epithelial cells were performed by
counting 200 to 240 cells per sample.

Mucus scores—The amount of mucus visible in the trachea (mucus quantity
scores) was graded on a scale of 0 (no visible mucus), l (singular small blobs), 2
(multiple blobs only partly conﬂuent), 3 (mucus ventrally conﬂuent), 4 (large ventral
pool) to 5 (profuse amounts of mucus occupying more than M: of tracheal lumen). We
have previously shown that these mucus accumulation scores are reproducible within and
between observers, and within horses and correlate with measured amounts [179]. One
observer (Gerber) graded all mucus scores.

Statistical analysis—Statistical analyses were performed using SigmaStat for
Windows, version 2.03S statistical software by SPSSZ. Absolute numbers of BALF cells
were calculated from total cell counts and the proportion of each cell type. Grouped data
were described with mean, standard deviation (SD), and coefﬁcient of variation (CV). All
data were analyzed using non—parametric tests. Mann-Whitney Rank Sum Tests were
used for comparisons between age groups. Spearman Rank Order correlations were used

to test for associations of age with mucus scores and BALF cytology parameters.

73

Correlations of mucus scores with BALF neutrophil percentages and numbers were also

tested with Spearman Ranks Order correlations. Signiﬁcance limit was set at P < 0.05.

Results

Interindividual variability-Complete sets of observations of mucus scores as well
as of BALF cytological variables are shown in Table l. The variation (CV) of mucus
scores was high overall (65.3%; all individuals combined) and in either age group (Table
1; illustrated in Fig. 1A). The CVs of BALF cytological parameters overall (all
individuals) were 37.7% for total cell counts, 24.6% for lymphocytes, 21.5% for
macrophages, 86.5% for neutrophils, 392.1% for eosinophils, 54.1% for mast cells and
65.3% for epithelial cells. The distribution of neutrophil percentages in both age groups is
illustrated in Fig. 13.

Comparison between age groups—Mucus scores and BALF cell percentages did
not differ between Y and 0 (Fig. 2A & C). The only signiﬁcant differences in BALF
cytology between the groups were a higher total cell count (Fi g. ZB) and higher numbers
of lymphocytes in Y (42 i 15 cells/pl) compared to O (28 i 13 cells/pl). All other
absolute numbers of BALF cells did not differ between the two age groups. Variations
(CV, Table l) of all mucus and BALF cytology parameters were very similar in Y and 0.

Correlations between age, mucus scores and BALF cytology—Age was
negatively correlated with total cell counts (r = -O.4) and lymphocyte numbers (r = -0.6).
No other BALF cytology variables or mucus scores correlated with age. Mucus scores

were not signiﬁcantly correlated with BALF neutrophil percentages and absolute

74

numbers (Fig. 3A and B). Yet, Fig. 3B shows a trend for the mucus scores to increase

with BALF neutrophil numbers.

Discussion

The purpose of this study was to 1) investigate what degrees of airway
inﬂammation and mucus accumulation are present in asymptomatic well-performing
sport horses, 2) examine the relationship between airway inﬂammation and mucus, and 3)
compare older to younger individuals.

Subclinical IAD was present in a majority of the healthy well-performing sport
horses we studied: Variability and means of airway mucus accumulation and certain
BALF cell percentages were considerably higher than normal ranges and means proposed
in a recent international workshop on IAD [12]. Because of the absence of eosinophils in
most horses and a very high percentage in one horse, this cell type showed by far the
highest variation. More than 1% eosinophils has been regarded as an abnormal ﬁnding,
but increases are often transient and their clinical signiﬁcance is unclear [184]. Since we
did not use a special stain for mast cells, we may have underestimated percentages of this
cell type. Nevertheless, mast cell percentages were also higher in our population than
observed in healthy racehorses without exercise intolerance [184]. Furthermore, epithelial
cell percentages varied considerably between individuals, but no ciliacytophtorea that
would indicate subclinical viral infection [185] was detected.

Neutrophils are the hallmark of airway inﬂammation in RAO [l 1] and increase
with stabling in asymptomatic horses [47, 180], but their role in IAD is unclear [184]. We

observed a large variation of this cell type between individuals, independent of their age

75

(Table 1; Fig. 1A). Both relative (Table 1; Fig. 1B, 2A and 3A and C) and, in particular,
absolute numbers (Fig. 3B) were much lower than observed in RAO horses in
exacerbation [3]. Compared to recently proposed normal values [12], however, BALF
neutrophil percentages in the majority of our subjects would be considered abnormally
increased. Our present results suggest that increased airway neutrophils may not only be
an acute response to stabling [47, 180], but can be observed in a large proportion of
asymptomatic horses housed long-term in a conventional stable environment. We also
observed markedly more variation of airway mucus accumulations (Table 1; Fig. 1B)
than reported by Dixon et al. (1995 [5]) in healthy control horses. This may be due to
differences in selection criteria. In the latter study, endoscopic and cytological criteria
were used to select healthy controls (McGorum, personal communication), while our
present ﬁndings are based solely on clinical and historical selection criteria.

Based on our previous ﬁndings [179], we had hypothesized that we would ﬁnd an
association between mucus scores and BALF neutrophils (Fig. 1A and B; Table 1), which
are known to produce potent secretagogues such as elastase and oxidant radicals [135,
175]. Contrary to our expectations, mucus scores were not signiﬁcantly correlated with
BALF neutrophil percentages (Fig. 3A and B). Fig. 3B suggests, however, that higher
mucus scores tended to be associated with BALF neutrophil numbers above 50 cells/uL.
We have observed a similar threshold increase in studies with control and RAO-horses
[51]. Since most of the horses in this study had less than 100 neutrophils/uL, the
relationship may not become statistically apparent. It is possible that the activity of the
mucus apparatus is largely independent from lower grade airway inﬂammation, i.e., < 100

neutrophils/uL. The epithelium may react directly to irritants, and there may be important

76

regional (i.e., large vs. small airways) differences in the intensity of this reaction. This
would explain the large variations of mucus accumulation independent from BALF
cytological variations we observed in this study as well as the previously reported
discrepancy between tracheal aspirate and BALF cytology [186]. Alternatively, BALF
cytology, which provides no information on the activity of the observed cells, may
simply be too crude a measure of airway inﬂammation to detect an existing association
between subclinical degrees of mucus secretion and inﬂammatory events in IAD.

Our ﬁndings and those of previous survey studies [10, 181] suggest that
(subclinical) IAD is present in a majority of stabled horses and may be a normal reaction
to the high loads of inhaled irritants [6, 8]. We don’t know whether increased levels of
airway inﬂammation and mucus persist in certain individuals. Further studies, which
follow horses over time, are needed to address this question. In a clinical setting,
however, diagnostic evaluation and treatment decisions are often based on a single
examination, similar to that performed in the present study. Even in the absence of
clinical signs, BALF neutrophil percentages above 10% and/or more than a few mucus
blobs in the trachea (mucus accumulation score > 1) may lead to a diagnosis of IAD.
Based on such a simplistic deﬁnition, more than half of the examined horses could be
diagnosed with IAD. The variation and means of neutrophil, mast cell and eosinophil
percentages we observed do not agree with those previously proposed as normal [12].
Indeed, four individuals had neutrophil percentages of more than 15% and three horses
had large mucus accumulations.

IAD, even at subclinical levels, is not only an accepted cause of poor performance

in racehorses [33], but has also been proposed to affect the ability of sport horses to jump

77

and perform in dressage [187]. Quite in contrast to that, we found that sport horses
performing well over years showed a large range of airway mucus accumulation and
BALF inﬂammatory cell cytology. Consequently, we propose that the signiﬁcance of
different degrees of IAD must be deﬁned separately for different disciplines and levels of
equine performance, according to the respective demands placed on the respiratory
system and founded on scientiﬁc data.

Even though airway inﬂammation and mucus varied considerably among
individuals, we found fewer differences between the two age groups than we had
expected. Age is a well-documented factor in the onset of RAO [3], but very little is
known about the association of age with milder degrees of airway disease in horses. We
hypothesized that years of living in a conventional stable environment, which exposes
horses to high dust loads [6, 8], would exacerbate subclinical lower airway disease
detectable by endoscopic and BALF examination. However, the present ﬁndings lead us
to reject this hypothesis. BALF cytology did not show increased inﬂammation in the
older horses (Table 1; Fig. 2A & B), which on average had spent ten more years in the
same conventional stable environment than the younger group. Neither did we observe
age-related increases of mucus accumulation or apparent viscosity (Table 1; Fig. 2 C), a
characteristic ﬁnding of long-term irritant exposure in humans [183]. Furthermore, the
variations of all BALF and mucus variables were also very similar in the two age groups
(CV; Table 1). By deﬁnition of our selection criteria, we had eliminated all horses with
historical or clinical evidence of lower airway disease from the study. These data

therefore suggest that those older horses in their teens, which have remained clinically

78

healthy despite long-term dust expose, do not suffer increased subclinical airway
inﬂammation and mucus accumulation.

The younger horses, on the other hand, did show higher BALF cell counts and
absolute lymphocyte numbers than the older group, a difference reﬂected in the negative
correlation of total cell counts and lymphocyte numbers with age. A lymphocytic type of
IAD has been described in young horses and has been compared to the acute form of
hypersensitivity pneumonitis [184]. In our view, this may simply reﬂect an increased
“immunological activity”, similar, for instance, to the follicular pharyngitis commonly
observed in younger horses. Chapman et al. (2000 [53]) reported increased “airway
inﬂammation scores”, associated with bacterial numbers, in two- and four-year-old
compared to older racehorses. However, these “airway inﬂammation scores” actually
represented airway neutrophilia and mucus, variables that showed no age association in
our study.

We conclude that older horses, which have remained clinically healthy, did not
show increased subclinical airway inﬂammation or mucus accumulation, while younger
horses had moderately higher lymphocyte cell counts. Large interindividual variation of
mucus accumulations, and BALF neutrophils, mast cells and eosinophils were observed.
Based on previously proposed endoscopic and BALF criteria more than half of the
examined healthy and well-performing sport horses had subclinical evidence of IAD.
Inclusion of measures of airway obstruction and hyperreactivity may be needed to arrive
at a more speciﬁc and functionally signiﬁcant deﬁnition of IAD. We further propose that
the clinical signiﬁcance of different degrees of IAD should be judged according to the

level of respiratory performance expected from an individual.

79

Table 4-1: Individual and group signalment, mucus scores and broncho-alveolar

lavage cytology.

 

Young (Y) and older (0) individuals. Geldings (G) and ﬁllies (F). Mucus scores (MS). Broncho-
alveolar lavage total cell counts (tcc; cells/pl), and percentages of lymphocytes (%lym),
macrophages (%mac), neutrophils (%neu), eosinophils (%eos), mast cells (%mas) and epithelial
cells (%epi). ND: not determined.

80

Fig. 4-1: Interindividual variability- mucus scores (A) and BALF neutrophil

percentages (B) of each horse of both age groups. Ages of the 13 younger and 13

older horses on X-axis.

MUCUS score

Percent (%) BALF neutrophils

 

25-

20-

 

13‘. 1.1
. 11 ‘11.
. 11 .1
3’ 1;,
.11}
1.1. .
“ 1
’ 1 ‘1']
1
111‘. 1‘ .1
1'1 1.11
‘1‘ ‘111‘
M 11“" ‘ ‘1
11 1
. 1 ..
1 11'1 1
‘ l ‘1 1‘
. 1 1 1‘1
111 1 1 1 1 .
l “E .1 ‘1‘ “ 1
‘11] .1 1.
. 1 11
1
1 ‘ 1.
1

 

3555555666667

 

 

 

 

.3131313131414141415171919age

 

younger group

81

older group

Fig. 4-2: Comparison between age groups (meaniSD)- Proportions (%) of BALF
cell populations were not significantly different between the two age groups (A), but
younger horses had significantly (*P < 0.05) higher total BALF cell counts (B) than

the older group. No difference was observed in mucus scores (C).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

60 "
,3 A - lvmohocvte
i 1 8‘8““ macroohqoe
50 1 . WI" neutroohi
8 8: _ eosinophil
% 8 8 — mast cells
.. . 8 8 m eoithelia
= 40 § §
8 8 8
\ 8
u. 8 8
:r 8
8 8
- 201 8 8
° 8 8
c 8 8 l
0’ - 8 8
9 1° ' 8 r 8 e
m 8 2 8 2
a 8 a 1 8 a
old young
3
5
a
\ 160 *
32 B 93 4 C
3 120- 8
5 80 . g
s
_. 4O
.9 E 1 1
O
l— 0. 0 l

 

 

 

 

 

 

 

 

 

   

old young old young

82

Fig. 4-3: Relationships of mucus quantity with broncho-alveolar lavage (BALF)

neutrophil percentages (A) and absolute numbers (B).

mucus score

 

 

 

 

-A .
0 510152025%

BALF neutrophil percentages

83

 

 

 

 

O O O- O O O O
O O
10 50 100 cells/UL

BALF neutrophil numbers

CHAPTER 5
MUCUS VISCOELASTICITY AND PREDICTED CLEARABILITY IN RAO-
AFFECTED AND CLINICALLY HEALTHY CONTROL HORSES BEFORE

AND DURING ENVIRONMENTAL CHALLENGE

Gerber, V., King, M., Schneider, D.A., Robinson, N.E. Tracheobronchial mucus
Viscoelasticity during environmental challenge in horses with recurrent airway

obstruction. (2000). Eq Vet Journal.

The previous studies (chapter 1-3) I showed that airway mucus accumulation is a
function of disease status interacting with environment, but not of age. Based on these
results, estimated volumes of mucus transported along the trachea in clinically healthy
horses are similar to healthy humans and dogs, but are increased 2-fold in RAO-affected
homes in remission and more than 4-fold during exacerbation.

In order to test the hypothesis that these increases in airway mucus accumulation
in remission as well as during exacerbation may be caused by unfavorable alterations of
mucus rheological properties, I compared mucus Viscoelasticity (and predicted
clearability) in RAO-affected and clinically healthy control horses before (when they

were at pasture) and after (6, 24 and 48 hours) stabling on straw bedding and hay feeding.

84

 

Abstract

The goal of this study was to compare the rheological properties of mucus from
horses with recurrent airway obstruction (RAO) to that from healthy controls during
environmental challenge by stabling in stalls with straw as bedding and hay as feed. We
determined and report Viscoelasticity (log 6* dyn/cmz, at 10 radian/s) and calculated and
report mucociliary clearability index (MCI) and cough clearability index (CCI), which
are derivative parameters of G* and the ratio of viscosity and elasticity measured at 1 and
100 radian/s, respectively. We also investigated the solids content of mucus, and cytology
of bronchoalveolar lavage ﬂuid (BALF). Samples were obtained before (0 hours) and 6,
24 and 48 hours aﬁer environmental challenge. The central ﬁndings were rheological
changes in airway mucus, which occurred over time in RAO-affected animals, but not in
controls. Mucus rheology was similar in both groups at 0 and 6 hours. In RAO-affected
horses, mucus viscoelasticity, as measured by log 6*, increased ﬁom 2.49 d: 0.18
dyn/cm2 (mean i: s.e.m.) at 0 hours to 3.05 :t 0.13 dyn/cm2 at 24 hours after
environmental challenge, and was accompanied by signiﬁcant decreases in MCI and CCI.
Percent solids of mucus did not signiﬁcantly differ between the two groups nor over time.
Rheological values did not correlate with BALF cytology. We conclude, that viscoelastic
properties of tracheal mucus samples from RAO-horses in remission do not differ from
those of normal horses. However, environmental challenge causes clinical signs of acute
airway obstruction and a concurrent increase in mucus Viscoelasticity only in RAO-
horses. Therefore, we infer that unfavorable changes in mucus rheology may contribute
to stasis and accumulation of mucus in RAO-horses in exacerbation, but not in clinical

remission.

85

 

Introduction

Excess mucus accumulation in the airways is a hallmark of recurrent airway
obstruction (RAO) [11]. We have shown that mucus accumulation scores are higher in
RAO than in controls, and increase further in RAO-affected animals when they are
stabled and exposed to hay [182]. These accumulations may be due to increased secretion
and/or decreased clearance. In turn, alterations in the ciliary apparatus or in the physical
properties of mucus ultimately result in changes of mucociliary clearance [69, 188]. The
effectiveness of mucociliary clearance in horses with RAO, however, is controversial.
Some investigators demonstrated that the rate of mucociliary clearance is decreased in
RAO-affected animals [5 8, 64], while others found no difference compared to healthy
horses [65]. Studies on excised pieces of trachea from healthy horses have shown that
variations of ciliary beat frequency at physiological temperatures have only a minor
effect on mucociliary clearance rate. In contrast, the properties of mucus are very
important [188].

We use the term “mucus” to describe the entity of gel-like secretions that are
present in the airways and available for sampling and rheological analysis. The
composition of airway mucus is heterogeneous, composed of water, mucin glycoproteins,
lipids, as well as cells with their degradation products and serum-derived proteins. This
heterogeneity imparts both liquid-like (viscous) and solid-like (elastic) properties to this
complex viscoelastic gel. Fig. 1 illustrates the relationships between Viscoelasticity (6*),
the vector sum of viscosity (6”) + elasticity (G’), and the ratio of viscosity / elasticity
(tan 5). Alterations in these viscoelastic properties of mucus result in changes in its

clearability by ciliary and cough mechanisms [69]. A critical factor affecting the

86

Viscoelasticity of mucus is hydration, i.e., its relative solids and water content [189, 190].
Another important factor is the presence of inﬂammatory cells. In particular, high
molecular weight DNA [171, 191] and ﬁlamentous actin [192] released by degrading
neutrophils increase mucus Viscoelasticity in cystic ﬁbrosis and possibly also in chronic
bronchitis and asthma. In these human airway diseases rheological alterations of mucus
are thought to be associated with the severity of disease (reviewed by Kim 1997 [39]).

In RAO-affected horses, neutrophils invade the lung within 7 hours of
environmental challenge [14] and accumulate in the airways in large numbers [15].
Airway obstruction develops concurrently. Although stasis and accumulation of mucus in
the airways is thought to play an important role in the pathogenesis of RAO [3], there is
little information on the rheology of equine respiratory mucus [21, 172].

The purpose of this study was to investigate the rheological properties of equine
respiratory mucus. We hypothesize that in RAO-affected horses in exacerbation, airway
inﬂammation is associated with compromised clearability of mucus. Therefore, we
compared the rheological properties of mucus from horses with RAO and healthy
controls, before and after environmental challenge by stabling in stalls with straw as
bedding and hay as feed. We measured Viscoelasticity of tracheal mucus samples,
computed the derivative parameters of predicted mucociliary and cough clearability and
determined the solids content. Furthermore, we investigated the association of rheological
properties with the solids content of mucus and with the cytology of bronchoalveolar

lavage ﬂuid (BALF).

87

Materials and methods

Animals and study design —This study was approved by the All-University
Committee for Animal Use and Care of Michigan State University. Seven horses (3
mares and 4 geldings), 9 to 26 years old (mean i s.e.m., 17.3 :t 2.4 years), with a
previous diagnosis of RAO, served as the principal group. Seven horses (5 mares and 2
geldings), aged between 7 and 26 years (mean i s.e.m., 13.4 i 2.8), and with no evidence
of respiratory tract disease, served as the control group. Both groups of horses were
drawn from closed research herds of the Department of Large Animal Clinical Sciences,
Michigan State University, that have been maintained for several years and develop
(principal group) or do not develop (control group) signs of obstructive pulmonary
disease when stabled and fed hay. All horses were vaccinated against equine inﬂuenza,
rhinopneumonitis, Eastern and Western equine encephalitis.

Clinical signs of airway obstruction were numerically scored as deﬁned below.
Horses were kept at pasture and their diet was supplemented with pellets until all animals
had no clinical signs of airway obstruction. Clinical assessment and samples of mucus
and BALF were obtained at baseline (0 hours) and 6, 24 and 48 hours after environmental
challenge by stabling in stalls with straw bedding and feeding hay.

Clinical score (CS) —One observer (Gerber) using a numerical scoring system
clinically assessed severity of airway obstruction. Each horse was examined at rest,
before sample collection. Nostril ﬂare was scored on a scale of 0 (least severe) to 2 (most
severe) and abdominal compression on a scale of 0 (least severe) to 3 (most severe). For
nostril ﬂare: 0 = no or little movement on inspiration; 1 = ﬂare during inspiration,

returning to normal as inspiration ends; 2 = ﬂare during inspiration and exhalation. For

88

abdominal lift: 0 = no or little movement in the ventral region of the ﬂank; 1 = slight
abdominal ﬂattening in the ventral region of the ﬂank; 2 = obvious abdominal ﬂattening
and “heave line” extending no more than half way between the cubital joint and the tuber
coxae; 3 = obvious abdominal lift and “heave line” extending beyond halfway between
the cubital joint and the tuber coxae. CS was calculated by adding the scores from nostril
ﬂare and abdominal lift. In order to enter the study, all animals had to have CS _<. 1 at 0
hours; i.e., RAO horses were in clinical remission at the beginning of the study. By
deﬁnition, CS had to remain S 1 at all time points (0, 6, 24 and 48 hours) for all animals
of the control group. Horses belonging to the principal group had to develop CS 2 2 at
some time point after stabling (6, 24, 48 hours).

Collection and analysis of mucus samples—Mucus was obtained in the trachea by
means of a cytology brush passed via an endoscope, similar to the technique used in
humans [73]. The sheath of a soft—haired bronchial brush (Olympus1 BC-23Q) was
removed, the wire of the brush glued into a teﬂon catheter and thus modiﬁed for use with
a 3 meter-endoscope (Olympusl GIF 300, custom made). The working channel of the
endoscope was thoroughly ﬂushed and then completely dried with compressed air each
time before the brush was introduced. As previous results had shown that equine mucus
sampled in the dorsal vs. the ventral part of the trachea differs in its average rheological
properties [179], we aimed to reduce the variability by sampling mucus only in the
ventral half of the trachea. After sampling, the brush was retracted 2-3 cm into the
working channel and the endoscope was withdrawn from the animal. The brush was

again protruded and mucus removed from the brush to be immediately stored under light

89

mineral oil at -80°C. For the subsequent rheological measurements the samples were
rapidly thawed to room temperature.

The magnetic microrheometer was used to measure the bulk viscosity and
elasticity [168]. A 100 um steel ball was carefully positioned in a 2 -10pL sample of
mucus and oscillated by means of an electromagnetic ﬁeld gradient. The motion of this
sphere was tracked with the aid of a photocell. Plots of ball displacement versus magnetic
force were used to determine the Viscoelasticity (G*; Fig. l) and tangent 8 (tan 8: Fig. 1).
G* (dyn/cmz) and tan 8 were measured at frequencies of 1, 10 and 100 radian/s. We
report Viscoelasticity as log“), (log G“) at 10 radian/s. For purposes of comparison, G“ of
water is 0.01 (log -2) dyn/cm2 and that of resin is 100 (log 2) to 1000 (log 3) dyn/cmz.
For a perfect solid, tan 8 is zero; for a perfect liquid, tan 8 is inﬁnite. From these
measured rheological properties the derived parameters mucociliary clearability index
(MCI) and cough clearability index (CCI) were computed based on observations of
clearance in model studies [69]. MCI, indicating clearability by normal ciliary function,
was calculated based on log 6* and tan 8 measured at 1 radian/s, and CCI was
calculated based on log G“ and tan 8 at 100 radian/s. Both indices relate negatively with
log G“; MCI also relates negatively with tan 8, but CCI relates positively with it. The
respective formulae are as follows:

MCI = 1.62 - (0.22 * log G*1) - (0.77 * tan 81) (1)

CCI = 3.44 - (1.07 * log G*1oo) + (0.89 * tan 8100) (2)

After rheological analysis, the solids content of mucus was determined. Samples
were rinsed in petroleum ether to remove adherent parafﬁn oil, weighed on glass slides

and evaporated to dryness in a microwave oven (30 min at 750 W). The ratio of dry to

90

 

wet weight was calculated and expressed as % solids content. Samples less than 2 mg
were excluded from the % solids analysis.

Bronchoalveolar lavage ﬂuid (BALF) sampling and analysis-BALF was
collected by means of a 3-meter endoscope (Olympusl GIF 300, custom made) that was
passed via the nares and wedged in a peripheral bronchus. Six 50-mL aliquots of
phosphate buffered saline (PBS) were infused into the tube and recovered by suction. The
lavaged ﬂuids were pooled and the total recovered volume determined. Total cell counts
in BALF were performed manually using a hemacytometer. Cell smears were made with
a cytocentrifuge and stained with Diff-Quick. Differential cell counts of lymphocytes,
macrophages, neutrophils, eosinophils, mast cells and epithelial cells were performed by
counting 100 cells per sample.

Statistical analysis—A11 statistical analyses were performed using a statistical
software program (SAS2 for Windows95, version 6.12). Four dependent measures on
tracheal mucus were analyzed by multivariate ANOVA for repeated measures using the
general linear model. Mucus Viscoelasticity (log G“ measured at 10 Hz), the two
rheological derivatives (MCI and CCI) and the % solids were the dependent measures of
interest. Independent factors in the model included two levels (principal and control
horses) of the between subjects factor, group, and four levels (0, 6, 24, and 48 hours) of
the repeated measure, time. Subanalyses were performed as predetermined speciﬁc linear
contrasts or as simple effects and interactions when main or interaction effects were
signiﬁcant. BALF total and differential cell counts (total cell count X percentage of each
cell type) were expressed as the base 10 logarithmic transformation to satisfy the model

assumptions of equal variance and normal distribution. A constant was ﬁrst added to

91

these data in order to make all data non-zero for logarithmic transformation. The
transformed BALF differential cell counts were then similarly analyzed. Pearson’s partial
correlation analysis was used to probe the relationships of three rheological measures (log
6*, MCI, and CCI) and the % total solids of tracheal mucus with the differential cell
counts from BALF. Partial variables included the two levels of group and four levels of
time. For all analyses, signiﬁcant differences were declared if P < 0.05. All data is herein

expressed as the mean ( s.e.m.).

Results

Clinical score (CS)—The fourteen horses entering the study in the principal (n=7) or
control group (n=7) qualiﬁed according to the respective criteria for CS as deﬁned above.
When entering the study, all animals had CS S 1 (0 hours). All individuals of the
principal group developed CS 2 2 at some time point after stabling; at 0 hours CS was
0.57 :l: 0.20, at 6 hours 1.14 :t 0.26, at 24 hours 2.86 i 0.55 and at 48 hours 2.57 :1: 0.43.
CS remained S 1 in all animals of the control group at all time points (0.14 d: 0.14, 0.0 d:
0.0, 0.14 i 0.14 and 0.0 i 0.0 at 0, 6, 24 and 48 hours, respectively).

Biophysical properties of mucus secretions-The rheological measures, log 6*
and tan 8 at 10 radian/s, for each horse are presented in Table 1. All parameters of
mucus physical properties are summarized in Table 2. Figs. 2A, 28 and 2C illustrate log
G“ at 10 radian/s, MCI and CCI, respectively.

A complete set of measurements of log G* and tan 5 at 10 radian/s from 56
samples was obtained. In the principal group baseline values (0 hours) of log G“

averaged 2.49 :t 0.18 dyn/cm2 (Table 2; Fig. 2A). At 6 hours values had decreased to 2.32

92

:1: 0.13 dyn/cmz. With the exception of horse 5, the highest values of each individual were
observed at 24 (Table 1: horses 1, 2 and 3) and 48 hours (Table 1: horses 4, 6 and 7). This
resulted in an increase of the average to 3.05 i 0.13 and 2.83 i 0.10 dyn/cmz,
respectively (Table 2; Fig. 2A). In the control group (Table 1: horses 8-14), with the
exception of horse 11, values of log 6* at 10 radian/s were below 2.9 dyn/cm2 and no
values of 3 dyn/cm2 or above were observed. On average log G“ (all control group values
pooled 2.40 i 0.06) was similar to the principal group baseline values and remained so
during stabling (Table 2; Fig. 2A). There was a signiﬁcant difference between principal
and control log 6* values (main group effect, P = 0.0187) which depended on time
(interaction effect of group and time, P = 0.0161). Subanalysis detennined that log 0*
values were not different between groups at baseline (P = 0.759) and that log G“ did not
change over time in control horses (P = 0.746). In contrast, log G" was increased in
principal horses at 24 hours when compared to baseline (P = 0.041) and at 24 and 48
hours when compared to time-matched control values (P = 0.002 and 0.003,
respectively).

Values of tan 8 were similar between the groups and over time (Table 1 and 2).
None of the main or interaction effects were signiﬁcant. The average of all tan 8 values at
10 radian/s, principal and control group pooled, was 0.37 i 0.01.

A complete set of measurements of log 6* and tan 8 at 1 radian/sec from 56
samples was available to compute MCI. (Table 2 and Fig. 2B). There was a signiﬁcant
difference between principal and control MCI values (main group effect, P = 0.031).
However, a signiﬁcant main or interaction effect of time was not detected. Subanalysis

determined that MCI values were not different between groups at baseline (P = 0.935)

93

and that MCI values did not change over time in control horses (P = 0.901). The
difference between principal and control MCI values was signiﬁcant at 24 hours (P =
0.011) and the difference between principal and control MCI values at 24 hours was
signiﬁcantly greater than at baseline (P = 0.043). A signiﬁcant simple effect of time on
the principal MCI values was not detected (P = 0.209).

CCI was computed for 49 samples, 7 measurements of log G* were not possible
at 100 radian/s due to high Viscoelasticity of the samples. Only 6 control and 3 principal
horses were retained in the analysis due to these missing values. There was a signiﬁcant
difference between principal and control CCI values (main group effect, P = 0.012).
However, a signiﬁcant main or interaction effect of time was not detected. Subanalysis
determined that CCI values were not different between groups at baseline (P = 0.616) and
that CCI values did not change over time in control horses (P = 0.478, n = 6). The
difference between principal and control CCI values was signiﬁcant at 24 hours (P =
0.028) and the difference between principal and control CCI values at 24 hours was
signiﬁcantly greater than at baseline (P = 0.043, n = 5). At 48 hours, the P-value for the
difference between principal and control CCI values was 0.061. A signiﬁcant simple
effect of time on the principal CCI values was not detected (P = 0.316).

The % solids content was measured in 54 samples (Table 2). Two samples of less
than 2 mg wet weight were discarded, and due to missing values one control horse was
excluded from this analysis. There were no signiﬁcant main or interaction effects. The
average of all pooled samples (n=54) was 9.8 i 0.4 % solids. No signiﬁcant correlations

were observed between % solids and log 6*, MCI and CCI.

94

BALF cytology-Fifty-three BALF samples were available for cytological
evaluation. Summary data are presented in Table 3. Total cell count increased more than
three-fold (0 vs. 48 hours) and % neutrophils more than four-fold (0 hours vs. 48 hours)
in the principal group during environmental challenge. The following statistical results
were obtained on log10 transformed BALF cell counts. There was a signiﬁcant effect of
group (P = 0.01) and time (P = 0.014) on the total cell count. The P-value of the
interaction effect of group and time was 0.064. Subanalysis determined that the total cell
counts were not different between groups at baseline (P = 0.167) and that total cell counts
did not change over time in control horses (P = 0.873). In contrast, the total cell count
increased in principal horses at 6 hours when compared to baseline (0.003), and at 6 and
24 hours when compared to time matched control values (P = 0.032 and 0.022,
respectively). There was a signiﬁcant effect of group (P = 0.001) on the neutrophil count,
which depended on time (interaction effect of group and time, P = 0.001). Subanalysis
determined that the neutrophil counts were not different between groups at baseline (P =
0.258) and that neutrophil counts did not change over time in control horses (P = 0.13). In
contrast, the neutrophil count was increased in principal horses at 6, 24, and 48 hours
when compared to baseline (P = 0.002, P = 0.001, and P < 0.001, respectively). For
lymphocyte, eosinophil, mast, macrophage, and epithelial cell counts, there were no main
or interaction effects. There were no signiﬁcant partial correlations of BALF cytology

with either log G“, MCI, or CCI.

95

Discussion

To our knowledge, no studies have directly compared mucus from RAO-affected
horses and healthy controls. While other rheological methods require relatively large
quantities of mucus, microrheometry allows for investigation of mucus rheology even in
healthy subjects, whose airways contain only small amounts of secretions [73].

At the start of the protocol and six hours after the horses had been stabled and fed
hay, Viscoelasticity and clearability of mucus did not differ signiﬁcantly between RAO-
affected horses and controls. The values were in the range of those previously observed
in healthy human subjects [73] and in healthy dogs, but lower than in smaller species, i.e.,
ferrets, rabbits and rats [193]. Since mucus rheology was not different from controls in
the RAO-affected animals before stabling, accumulations of secretions observed when
RAO-horses are in remission are likely be due to other than rheological factors (e.g.,
hypersecretion). However, twenty-four hours after the environmental challenge, the
Viscoelasticity of mucus had signiﬁcantly increased, log 6* averaging 3.0 :1: 0.3 dyn/cm2
(Table 2; Fig. 2A). This change of log 0.5 represents a 3-fold increase of Viscoelasticity
on a linear scale. Furthermore, log G" values of 3 and more are in the range found in
human patients suffering from severe airway obstruction caused by cystic ﬁbrosis [194].

The ratio of viscosity / elasticity did not differ between controls and RAO-horses
and was not signiﬁcantly affected by stabling and hay challenge. This means that the
observed increase of Viscoelasticity in the principal group was due to proportional
changes of both viscosity and elasticity. Values for tan 8 were similar to those reported in

healthy humans [73] and in other species [193].

96

As a result of the changes in viscoelasticity, both predicted mucociliary and cough
clearability in RAO-affected horses were signiﬁcantly affected by the environmental
challenge, MCI reﬂecting the changes in log G“ more closely (table 2; Fig. 2A, B and C).
These results suggest that the changes in viscoelasticity, which occur at the onset of
clinical signs, affect both ciliary and cough clearability of mucus and, may therefore
contribute to accumulation and stasis of mucus in the airways of RAO-affected horses in
acute exacerbation. It is important to note that, while our ﬁndings suggest unfavorable
changes in mucus clearability based on calculated MCI and CCI, factors other than
Viscoelasticity such as the spinnability or the adhesivity of mucus for instance, which we
have not investigated in this study, may signiﬁcantly inﬂuence clearability. Also, the
changes in predicted MCI and MCI were most pronounced at 24 hours of environmental
challenge. After another day in the stable, both Viscoelasticity and mucociliary
clearability values showed a trend of returning towards baseline. The very high values of
Viscoelasticity we observed may be a transient phenomenon, of acute exacerbation
occurring after stabling highly susceptible horses.

We have previously investigated mucus rheology in horses with only mild to
moderate clinical signs of airway obstruction (i.e., chronic accumulation of mucus in the
airways, coughing, but no dyspnea) [195]. The results of this investigation contrast with
the ﬁndings of the present study. These horses with mild to moderate clinical signs
showed remarkably lower average Viscoelasticity (mean :1: SD: 1.96 :1: 46) and better
clearability (MCI mean :t SD: 0.99 i 0.11) than the principal and even the control horses
in the present study. Our principal group, of course, was composed of severely RAO-

affected animals, which by deﬁnition showed marked clinical signs when stabled and fed

97

 

hay. Also, the mean age of both the principal and control horses of the present study was
7 and 3 years higher, respectively, than that of the mild to moderately affected horses.
Interestingly, in heavy smokers an age (pack-years)-dependent decrease followed by an
increase in Viscoelasticity has been observed [183]. Age may play a role in equine mucus
rheology as well. In the present study, however, age was not a statistically signiﬁcant
factor either between or within the principal and the control group (results not shown).

Since the mucus samples were harvested, stored and analyzed by the same
methods, the disparity in the rheology of the mucus between the present and the previous
ﬁndings must result from either the above mentioned differences in severity of disease
and age of the animals or in the differing design of the two studies. While the horses in
the present study were exposed to hay after having been at pasture for an extended
period, the mild to moderately affected animals in the previous study were stabled and
fed bay for at least three weeks prior to sampling mucus.

Based only on the physical appearance of tracheobronchial secretions, Schatz-
mann et al. (1972 [45]) proposed that mucus viscosity is proportional to the severity of
clinical signs of airway obstruction in RAO-horses. When these authors measured
viscosity of mucus in 5 RAO-affected horses in a subsequent study they found it to be
lower than in human patients with severe chronic bronchitis [21]. However, prior to the
trial, hay had been eliminated from the diet of these animals [21]. A recent report
indicated signiﬁcantly higher values of mucus viscosity in horses with RAO than in
horses affected with inﬂammatory airway disease [172]. The authors proposed that the
difference was due to the presence of high molecular weight DNA in the mucus of the

RAO-affected animals. Unfortunately, due to differing methods and limited published

98

information no direct comparison is possible with these two reports [21, 172]. The
available data on equine respiratory mucus rheology suggest, however, that alterations in
mucus rheology in horses may be a function of the intensity and duration of exposure to
irritants and allergens as well as the severity and duration of airway disease in the
individual. This could explain contradictory ﬁndings reported in studies of mucociliary
clearance in RAO-affected horses [58, 64, 65], since clearance may be either normal or
compromised depending on individual and environmental factors. Further studies are
needed to determine the respective inﬂuences and interaction of intensity and duration of
exposure as well as the severity and duration of disease in the individual on changes in
mucus rheology.

The present study was not designed speciﬁcally to identify mechanisms
responsible for changes in mucus rheology. Hydration of mucus (i.e., % solids) was not
signiﬁcantly affected by stabling and did not seem to play a role in the rheological
changes we observed. BALF cytology before and during environmental challenge
showed the typical differences between RAO-affected horses and controls [3]. However,
BALF and tracheal aspirate cytology lack a strong correlation [186]. This, and the
stringency of the statistical test we employed, may explain why we did not observe any
signiﬁcant correlations between parameters of BALF cytology and tracheal mucus
rheology. Therefore, our results do not support any predictive or diagnostic value of
BALF cytology for rheological changes induced by the environment; but neither do they
rule out that inﬂammatory cells, or more speciﬁcally their degradation products, may
have an effect on mucus rheology. Interestingly, preliminary mucolytic data based on a

limited number of samples from this study suggest that extracellular ﬁlamentous actin is

99

present in mucus from RAO-affected animals and, possibly in combination with DNA
ﬁbers, increases its Viscoelasticity (Gerber, unpublished results).

The central ﬁnding of this study was that environmental challenge by stabling and
hay diet unfavorably alters mucus rheology in RAO-affected horses, but not in controls.
We conclude, that viscoelastic properties of tracheal mucus samples from RAO-horses in
remission do not differ from those of normal horses. During the onset of clinical signs of
airway obstruction, however, mucus Viscoelasticity in RAO-horses increases to values
observed in severe obstructive lung disease of humans, reducing both ciliary and cough
clearability. Therefore, unfavorable changes of mucus rheology may contribute to stasis
and accumulation of mucus in the airways of RAO-affected horses in exacerbation, but

not in remission.

100

Table 5-1: Biophysical properties of individual mucus samples: Viscoelasticity (log
G*; dyn/cmz) and tangent 8, measured before (0 hours) and during (6, 24, 48 hours)

stabling in each RAO horse (1-7) and each control (8214).

 

 

Viscoelasticity (log 6* at 10 radian/s; tangent 8 (at 10 radian/s)
dyn/cmz)
hours 0 6 24 48 0 6 24 48
after
stabling
Principal group

horse 1 2.39 2.47 3.32 2.30 0.23 0.36 0.32 0.24
horse 2 1.89 2.57 3.07 2.91 0.29 0.31 0.47 0.33
horse 3 2.58 2.65 3.61 2.86 0.34 0.31 0.32 0.30
horse 4 2.22 1.87 2.8 2.93 0.29 0.41 0.51 0.54
horse 5 3.27 2.66 2.81 2.71 0.33 0.54 0.4 0.32
horse 6 2.95 2.13 3.16 3.17 0.36 0.29 0.41 0.19
horse 7 2.15 1.91 2.57 2.92 0.36 0.48 0.37 0.43
Control group

horse 8 2.49 2.57 2.11 2.51 0.32 0.46 0.36 0.35
horse 9 2.66 2.70 2.31 2.15 0.39 0.28 0.34 0.29
horse 10 2.66 2.38 2.24 2.52 0.35 0.37 0.35 0.39
horsell 2.34 2.99 2.90 2.54 0.35 0.36 0.38 0.32
horse 12 1.74 2.52 2.50 2.48 0.30 0.50 0.43 0.40
horse 13 2.89 1.68 2.50 2.00 0.39 0.56 0.35 0.68
horse 14 2.16 2.53 2.08 2.16 0.36 0.42 0.37 0.35

 

101

Fig. 5-1: Vector diagram illustrating the Viscoelasticity of mucus (G*), the viscous

(G”) and elastic (G’) components and their ratio deﬁned by tangent 8.

G99 (3*

 

 

G9

102

Fig. 5—2: A) Viscoelasticity at 10 radian/s on a logarithmic scale (log G*; dyn/cmz), B)
mucociliary clearability index (MCI) and C) cough clearability index (CCI) of
mucus samples in RAO-affected (principal group) and control horses (control
group) before (0) and after environmental challenge (6, 24 and 48 hours). # Indicate
significant differences between principal and control group at a time point. *
Indicates a significant difference between the time point and baseline in the same

group.

 

3.5 4 A prInCIpal group +
control group -0-

 

 

 

# *
3.08

log 6' (dyn/cmz)

2.5 ~

 

2.0/

\\

 

0.0

 

1.0. B

MCI

0.8 .

0.6 J

\\

 

0.0
1.6 T

 

CCI

1.2 .

0.8 .

0.4 - #

 

0.0

 

O 6 24 48 hours

103

CHAPTER 6
PRELIMINARY STUDY ON THE MUCOLYTIC EFFECTS OF DNase AND
GELSOLIN ON EQUINE MUCUS RHEOLOGY-F—ACTIN MAY
CONTRINUTE TO INCREASED VISCOELASTICITY DURING RAO

EXACERBATION

In the previous study (chapter 5) I showed that airway mucus Viscoelasticity
increased 3-fold in RAO-affected horses during exacerbation. This increase coincides
with a dramatic inﬂux of neutrophils in the airways of RAO-affected horses.
Neutrophils in the airways degenerate and release breakdown products, speciﬁcally
DNA and/or ﬁlamentous (F )-actin that can unfavorably alter mucus rheological
properties.

In order to test the hypothesis that these breakdown products cause
unfavorable rheological changes, i.e., the observed increased mucus Viscoelasticity
during RAO exacerbation, I investigated the effects of DNase, which degrades DNA,
and gelsolin, which degrades F-actin, on airway mucus from RAO-affected in

exacerbation compared with control mucus samples in a preliminary mucolytic study.

104

Abstract

We demonstrated that the Viscoelasticity of mucus in recurrent airway
obstruction (RAO)-affected horses in remission is similar to that of healthy controls.
In acute exacerbation after stabling and hay exposure, however, mucus Viscoelasticity
increases 3-fold in RAO-affected horses, but remains normal in controls. This increase
in Viscoelasticity coincides with a large neutrophil inﬂux in the airways. High
molecular weight DNA ﬁbers and ﬁlamentous (F) -actin from degrading neutrophils
increase mucus Viscoelasticity in human cystic ﬁbrosis (CF). The importance of DNA
ﬁbers in RAO mucus is unclear, however, and F -actin has not been investigated. The
aim of this preliminary study was to investigate the role of these neutrophil
degradation products in the unfavorable rheological changes observed during RAO
exacerbation. We assessed the mucolytic effects of gelsolin, a capping protein that
severs F-actin ﬁlaments, and thNase, which cleaves high molecular weight DNA
molecules. Concentrations of gelsolin and thNase previously shown effective on CF
sputum were tested on mucus from RAO-affected horses in exacerbation compared
with control mucus from RAO-affected horses in remission and from clinically
healthy horses. Negative and positive control treatments (sham, saline and nacystelyn;
DNase and gelsolin on CF mucus) conﬁrmed the validity of mucolytic experiments.
RhDNase treatments did not alter Viscoelasticity in either RAO or control mucus.
Gelsolin had a signiﬁcant mucolytic effect on mucus of RAO horses in exacerbation,
effectively reducing Viscoelasticity almost 3-fold to normal levels. Unlike the control
treatment with nacystelyn, gelsolin had no effect on control mucus. This indicates that

F-actin plays a role in unfavorable rheological changes observed in mucus of RAO-

105

affected horses during exacerbation. In contrast, DNA ﬁbers either do not play a
similarly important rheological role in RAO mucus as in CF mucus, or, alternatively,
they may be protected to a larger degree from mucolytic action. F-actin ﬁbers have
been shown to form an entangled network with DNA ﬁbers, not only causing
increases in viscoelasticity, but also protecting DNA ﬁbers from mucolytic action by
thNase. Further studies are needed to elucidate the role of DNA and F-actin ﬁber
interactions in RAO mucus. In conclusion, this preliminary study showed that RAO
mucus is similarly susceptible to the action of gelsolin as CF sputum, while higher

concentrations of thNase may be necessary to achieve a mucolytic effect.

Introduction

Equine recurrent airway obstruction (RAO) is characterized by bronchospasm
and intense neutrophilic inﬂammation with inﬂammatory cells and cellular debris,
serum proteins and mucus glycoproteins accumulate in the airways [2, 3]. All these
components of mucous secretions can inﬂuence the Viscoelasticity of mucus, which in
turn is associated with mucus clearability by mucociliary and cough mechanisms [69,
196-198]. We demonstrated that the Viscoelasticity of mucus in RAO affected horses
in remission is similar to that of healthy controls. In acute exacerbation, however,
mucus Viscoelasticity in RAO increases 3-fold [169] to levels otherwise only observed
in human cystic ﬁbrosis (CF) [194]. This increase in Viscoelasticity coincides with an
increase of endoscopically observed mucus accumulation [182] and with neutrophil
inﬂux [169] in the airways, but is not associated with changes in the hydration of the

mucus [169].

106

High molecular weight DNA [171] and ﬁlamentous (F) -actin [192] released
by degrading neutrophils increase mucus Viscoelasticity in CF and possibly also in
chronic bronchitis and asthma. DNA and F-actin can form a network within the
existing mucin network. The relative importance of this DNA/f-actin network in
human airway diseases is unclear, but sputum stained to differentiate mucin from
DNA and actin showed that there are large amounts of mucin and little DNA in
chronic bronchitis sputum, but relatively less mucin and much more DNA and actin in
CF sputum (M. King, personal communication).

In the horse, Schatzmann et a1 (1973 [21]) found that free DNA was not
present in excess in mucus of RAO-affected horses compared with sputum of human
chronic bronchitis patients. In contrast, Pietra et a1. (2000 [172]) have shown a
mucolytic effect of human(h) recombinant(r) DNase on mucus of RAO-affected
horses. Thus, the importance of DNA ﬁbers is unclear and F-actin has not been
investigated in equine airway mucus.

The aim of this preliminary study was therefore to investigate the role of these
neutrophil degradation products in the unfavorable rheological changes observed
during RAO exacerbation. We assessed the mucolytic effects of gelsolin, a capping
protein that severs F-actin ﬁlaments, and thNase, which cleaves high molecular
weight DNA molecules. Concentrations of gelsolin and thNase effective on CF

sputum [192, 194] were tested on mucus from RAO-affected horses in exacerbation.

Materials and methods
Horses, mucus sampling and rheological measurements—Seven RAO-

affected (9 to 26 years old) and 7 healthy control (7 to 26 years old) horses were kept

107

at pasture and their diet was supplemented with pellets until all animals had no
clinical signs of airway obstruction. Mucus samples were obtained at baseline (0
hours) and 6, 24 and 48 hours after environmental challenge by stabling in stalls with
straw bedding and feeding hay. Mucus was sampled ﬁom the ventral part of the
trachea by means of a cytology brush passed via an endoscope [73] and immediately
stored under light mineral oil at -80°C as described [169]. The magnetic
microrheometer technique was used to measure Viscoelasticity (the vector sum of
viscosity and elasticity: G" (dyn/cm2)) and tan (the ratio of viscosity and elasticity)
at frequencies of 1, 10 and 100 radian/s as described [168].

Mucolytic treatment-Gelsolin (Biogen, Inc. Cambridge, MA) and thNase
(Pulmozyme®, Genentech, Inc. San Francisco, CA) were applied in saline at 10% of
sample weight. Gelsolin and thNase concentrations were 200nM, as previously
shown effective in CF sputum [199]. Nacystelyn (SMB pharmaceuticals, Belgium),
which acts by breaking up disulﬁde bonds between mucin molecules, was used as a
positive control treatment [200]. The negative control treatments were saline at 10%
sample weight and sham, i.e., no substance added to mucus. Samples were divided
into control (C) samples and RAO exacerbation (R) samples. C-samples (n = 21) were
from control horses at any time point and from RAO-affected horses at time points 0
and 6 hours, since Viscoelasticity does not change in control horses during
environmental challenge and is normal at 0 and 6 hours in RAO-affected animals
[169]. R—samples were from RAO-affected horses 24 or 48 hours after environmental
challenge when the increase of mucus Viscoelasticity occurs [169]. Large R-samples

were carefully split for several measurements so that a sufﬁcient number (n = 31) was

108

obtained for all treatments. Microrheological measurements were performed before
and after 30 minutes incubation at 37°C with gelsolin, thNase, nacystelyn, saline or
no treatment (sham). Samples were randomly assigned to treatments, which were
performed in randomized order. As a further positive control, each treatment was also
performed on a small number of cystic ﬁbrosis (CF) mucus samples.

Statistical analysis and interpretation—A11 statistical analyses were performed
using SPSS for Windows (SPSS Inc., 233 S. Wacker Drive 11th Floor, Chicago, IL
60606, USA). Paired one-tailed t-tests were used on Viscoelasticity (log G*) and tan 8
values of paired samples before and after treatments with gelsolin, thNase,
nacystelyn, saline or no treatment (sham) to assess signiﬁcance of mucolytic effects.
For paired t-test analysis, samples were grouped together (C, R and CF) for negative
(sham, saline) and positive (nacystelyn) control treatments, and grouped separately (C,

R, CF) for endpoint treatments with gelsolin and thNase.

Results and discussion

Control treatments—The sham procedure on R, C and CF samples had no
effect and showed very little variation (Table 1; Fig. 1A). Similarly, the saline control
showed no signiﬁcant change and only slightly more variation (Table 1; Fig. 1B). The
positive control treatment nacystelyn showed a signiﬁcant combined mucolytic effect
on R, C and CF samples (Table 1). The overall decrease in Viscoelasticity was more
than 3-fold on a linear scale (Table 1). Fig. 1C shows that this was mainly due to a
large effect on samples with high starting Viscoelasticity. In contrast, samples that had
a logG* of less than 2.5 dyn/cm2 showed no effect. A paired t-test was therefore

performed separately on control samples, since they are of inherently lower

109

Viscoelasticity. The test conﬁrmed that nacystelyn also had a signiﬁcant effect on
control samples (Table 1). This was expected since nacystelyn, a thiol reducing agent
and mucolytic similar to N-acetyl-L-cysteine, indiscriminately breaks mucin disulﬁde
bonds that make up the mucus gel network. Three CF samples each were used as
positive controls for the mucolytic treatment with gelsolin and DNase. Although no
statistical analysis was performed, Figs. 1F & I illustrate that the concentrations used
had a strong mucolytic effect on CF samples in this study, as expected based on
previous reports [199]. No effects on tan 8 were observed with any of the control
treatments (results not shown). The control treatments conﬁrmed the validity of our
mucolytic experiments.

Mucolytic effects of DNase on equine mucus samples-RhDNase treatments
did not alter Viscoelasticity of either mucus from RAO-affected horses in exacerbation
or of control samples (Table 1). No effects on tan 8 were observed with thNase in
any of the sample groups (results not shown). Although the power of the test in the
thNase/control sample group was limited due to the sample size, Figs. 1D & E do
not show any trend of a mucolytic effect in either sample group. This is in agreement
with Schatzmann et a1 (1973 [21]), who concluded that free DNA was not present in
excess in RAO mucus compared with sputum from humans with chronic bronchitis.
On the other hand, our results are in apparent conﬂict with those of Pietra et al. (1998
[172]) who found that RAO mucus, in contrast to that of horses affected by small
airway inﬂammatory disease, had a high DNA content and was susceptible to the
mucolytic effect of thNase. Although the methodology cannot be directly compared,

Pietra et al. (1998 [172]) used a higher concentration of thNase, which could explain

110

the discrepancy in our ﬁndings. Our methodology was identical, however, with that
previously used on CF-sputum [199], and we can therefore conclude that thNase
concentrations that are effective on CF mucus do not work on RAO exacerbation or
control equine samples. Thus, DNA ﬁbers either do not play a similarly important
rheological role in RAO mucus as in CF mucus, or, alternatively, they may be
protected to a larger degree from mucolytic action. F-actin ﬁbers have been shown to
form an entangled network with DNA ﬁbers [201], causing increases in Viscoelasticity
[171, 192], and also protecting DNA ﬁbers from mucolytic action [199, 202].

Mucolytic effects of gelsolin on equine mucus samples—The gelsolin
treatments offered evidence that F-actin may indeed play a role in RAO mucus
rheology. As illustrated in Fig. 16, gelsolin had a signiﬁcant mucolytic effect on
mucus from RAO horses in exacerbation, reducing Viscoelasticity almost 3-fold (on a
linear scale) to normal values. Unlike the nacystelyn control treatment, gelsolin did
not change Viscoelasticity of control samples (Table 1; Fig. 1H). This strong selective
effect of gelsolin on RAO exacerbation mucus indicates that F-actin ﬁbers may be
involved in the unfavorable rheological changes we have observed in mucus from
RAO-affected horses after stabling and hay exposure [169]. In order to better
understand these mechanisms, however, further studies are needed. In particular, the
effects of these mucolytic agents may have to be studied separately at different
concentrations as well as in combination, since a synergistic effect of gelsolin and
thNase has been demonstrated in CF sputum [199].

DNA/F-actin network- synergistic effects of DNase and gelso-

lin—Fluorescence microscopy of CF sputum has shown that F-actin and DNA co-

111

localize to form large aggregates in CF sputum and that DNA forms in vitro ﬁber
aggregates only in the presence of actin [201]. These ﬁndings suggest that increases in
mucus Viscoelasticity may not simply be due to the presence of either F-actin and/or
DNA ﬁbres, but rather to an interaction of both ﬁber types to form large aggregates.
The demonstration of a synergistic effect of gelsolin and thNase in CF sputum [199]
provided ﬁirther evidence for an interaction of the two degradation products. We
propose that following neutrophilic inﬂammation in equine RAO, aggregates of F-
actin and DNA ﬁbers form in airway secretions and interact to cause the dramatic
increases in mucus Viscoelasticity observed during acute RAO exacerbation. Since the
cost of the available gelsolin and thNase products are prohibitive in equine
medicine, we cannot expect to derive direct proﬁt for therapeutic application of these
products in the near future. Nevertheless, investigations of mucolytic effects, such as
performed in this preliminary study, can offer insight into the mechanisms of mucus
rheological changes and accumulation of secretions in equine RAO.

Conclusions—Our results indicate that mucus from RAO-affected horses in
exacerbation is similarly susceptible to the action of gelsolin as CF sputum. Severing
F-actin ﬁlaments reduced Viscoelasticity of RAO exacerbation mucus to normal
values. In contrast, a higher concentration of thNase than is effective in CF sputum
may be necessary to achieve a mucolytic effect in RAO mucus [172]. Thus, F-actin
ﬁber aggregates, possibly interacting with DNA ﬁbres, appear to play a role in the

unfavorable rheological changes of mucus from RAO-affected horses in exacerbation.

112

Table 6-1: Change of Viscoelasticity (Alog G*, dyn/cmz) without any substance
added (sham), after saline treatment and after mucolytic treatment with nacystelyn,
gelsolin and thNase on RAO-exacerbation (R), control (C) and cystic fibrosis (CF)

mucus samples.

 

 

 

 

 

 

 

 

 

 

 

 

treatment sample groups number of mean i SD of change in
used samples Viscoelasticity
(Alog G*, dyn/cmz; P)
sham R + C + CF 9 (6 + l + 2) +0.02 i 0.22; ns
saline R + C + CF 10 (6 + 2 + 2) -0.001 i 0.35; ns
nacystelyn R + C + CF 15 (5 + 7 + 3) -0.53 i 0.56; P < 0.05
nacystelyn C 7 -0.48 d: 0.47; P < 0.05
gelsolin R 7 -0.40 i 0.21; P < 0.01
gelsolin C 7 +0.18 1: 0.47; ns
gelsolin CF 3 -0.56 i: 0.37; na
thNase R 7 +0.08 i 0.30; ns
thNase C 4 +0.09 :1: 0.28; ns
thNase CF 3 -0.67 i: 0.61; na

 

 

 

 

 

P > 0.1 was considered not signiﬁcant (us). The 3 CF samples in gelsolin and DNase treatments

were not statistically analyzed (na)

ll3

Fig. 6-1: Viscoelasticity (log G*) before and after treatment. Negative (sham, A and
saline, B) and positive (nacystelyn, C) control treatments on RAO-exacerbation (R),
control (C) and cystic fibrosis (CF) samples. RhDNase treatment on R (D), C (E)
and CF (F) samples. Gelsolin treatment on R (G), C (H) and CF (1) samples.
Significant (P < 0.05) and non-significant (ns) treatment effects are shown in the
graphs. Effects of thNase and gelsolin on CF control samples (F and 1,
respectively) were not statistically analyzed (na).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

3.5 3.5 4
3.5 ~
a 3 r a) 3 ' c»
.2 2 2
V v V 3,
f, 2.5, g 2.5- g
g to m
s 2 .————- *3 2- s 2,
0 u 0
.2 .2 .2
>15. / >1.5- 9"" >151
ll! '
1 1 “5 1 P<0.05
before after belore alter before after
sham sallne nacystelyn
A 4 A 4 A 4
to E9 '
3.3.5~ D 3,3.5~ 5\ 33.5. F
c a 8,
a? 3 1 2‘ 3 ‘ 3; 3 .
.2 E E
g 2.5. jg 2.5« / g 2.5.
3 - _. _
g 2 . ./—4 § 2 - § 2 ~
.22 ,2 m
> > , g .
1.5 . / 1.5 m 1.5
ns na
1 1 .\
before atler before alter 1 before alter
thNase IR thNase IC thNase ICF
A 4 4 4
{9 G A H r l
g 3.5 " as 4 C; 3.5 _‘
c 3 .9
3‘ 3 1 V 3 < V 3
:§ :8 5 l
a; 2.5 « 13.5. g 2.5-
8 2 , \ g 2‘ § 2 ‘
.2 > ,9
> 1.5‘ 1.58 > 1.54
P<0.05 ns na
1 1
before after w” “I“ 1 before alter
gelsolln I R gelsolin I C gelsolin I CF

114

CHAPTER 7
IDENTIFICATION OF EQUINE HOMOLOGUES OF GEL-FORMING MUCIN
GENES AND SEMI-QUANTITATIVE MEASUREMENT OF OF eqMUCSAC
AND eqMUC2 mRNA LEVELS IN POOLED SAMPLESA FROM DIFFERENT

AIRWAY GENERATIONS OF RAO-AFFECTED AND HEALTHY HORSES

Gerber, V., Robinson, N.E., Venta, P.J., Rawson, J., Jefcoat, A.M., Hotchkiss, J.A.
Mucin genes in horse airways: MUCSAC, but not MUC2 may play a role in recurrent

airway obstruction. (2003). Eq Vet Journal.

In the ﬁrst studies (chapter 2-4) I showed that airway mucus accumulation is
increased 2-fold in RAO-affected in remission compared with healthy control horses and
further increases more than 2-fold in RAO-affected horses during exacerbation. Further
studies (chapter 5 and 6) showed that mucus Viscoelasticity is normal in RAO-affected
horses in remission, but increases 3-fold during exacerbation. This increase is due at least
in part to breakdown products, in particular F-actin, from degenerate neutrophils.

Unfavorable rheological properties leading to decreased mucus clearance can only
partly explain the increased mucus accumulation in RAO, however. Particularly during
remission, but potentially also in exacerbation, increased production of mucus must also
play a role. Based on ﬁndings in other species, any or all of the secreted, gel-forming
mucins, MUC2, MUCSAC and/or MUCSB, could be expressed in horse airways and may

be involved in increased mucin production in RAO.

115

 

In order to test the hypothesis that increased production of mucus is a cause of the
observed increased mucus accumulation, I identiﬁed equine homologues of gel-forming
mucins and investigated the regional distribution of mRNA-levels of equine mucin gene

homologues at different airway generations.

Abstract

Increased mucin gene expression may be an important cause of mucus accumulation
observed in recurrent airway obstruction (RAO)-affected horses. To date, however, no
mucin gene sequences are available for the horse. Our goal was to identify equine
homologues of gel-forming mucins and investigate their expression at different airway
generations of healthy and RAO-affected horses. Two equine homologues were identiﬁed
by cloning and sequencing fragments of equine (eq)MUC5AC and eqMUC2. Semi-
quantitative RT-PCR on RNA from airways (generations 1, 5, 10, 15; small airways and
parenchyma), stomach (glandular), and colon revealed that eqMUCSAC is expressed in
equine stomach and in all of the airway samples. In contrast, eqMUC2 steady-state mRNA
levels were detected in colon and very faintly in stomach, but not in airway tissue.
EqMUCSAC expression was also compared to that of ZO-l, a tight junction protein, and
eqMUCSAC/ZO—l ratios were higher in RAO-affected compared to control horses at all
airway generations. We conclude that eqMUCSAC is expressed in horse airways, but any
expression of MU C2 is undetectable and unlikely to be of physiological consequence.
EqMUCSAC up-regulation may be a primary mechanism responsible for mucus

hypersecretion and accumulation in RAO.

116

Introduction

In equine recurrent airway obstruction (RAO), excess airway mucus may be due to
hypersecretion and/or decreased clearance [72]. Based on histological observation of goblet
cell hyperplasia, hypersecretion has long been associated with RAO (reviewed in [2, 3], and
we found mucin glycoprotein alterations in bronchoalveolar lavage ﬂuid (BALF) of RAO
horses [54]. Furthermore, we showed that mucus accumulation in RAO can only partly be
explained by decreased mucus clearability [169].

Therefore, we propose that up—regulation of speciﬁc mucin gene(s) may be a
primary mechanism leading to hypersecretion of speciﬁc mucins, subsequent mucus
accumulation, and possibly changes in glycosylation patterns. In order to investigate these
hypotheses and their implications it is ﬁrst necessary to characterize speciﬁc mucins and
their expression in equine airways.

Three gel-forming mucins, which have been (partly) sequenced, are present in
human and rodent airways: MUC2, MUCSAC and MUCSB [87, 93, 203-207]. Based on
reactivity with polyclonal antibodies raised against the respective human mucins, it has been
proposed that a homologue of MUCSAC is present in equine stomach [100], and that
homologues of MUCSAC and MUCSB may be found in equine airway secretions [101,
102]. Since these ﬁndings are based on reactivity with antibodies that have been raised
against human mucins, genetic identiﬁcation of equine homologues of the main gel-
forming mucins is a priority for the investigation of mucin hypersecretion in RAO.

To date, no equine mucin gene products have been genetically characterized. The
aims of this study were, therefore, to identify equine homologues of human and rodent

mucin genes, to develop RT-PCR—based assays to determine which gel-forming mucins are

117

expressed in airways of healthy and RAO-affected horses, and to investigate the regional

distribution of mucin gene expression in the airway tree.

Materials and methods

Animals, tissue collection, and processing for RNA samples—Seven control
horses (4 mares and 3 geldings; 8i7 years of age), without a history of and without
clinical signs of chronic airway disease, and 5 horses (4 mares and l gelding; l7i6 years
of age), with a history and clinical signs of RAO, were exposed to an indoor stable
environment and hay-feeding for several days. Within an hour after euthanasia (by
barbiturate overdose in accordance with guidelines set forth by the All-University
Committee on Animal Use and Care at Michigan State University), 1-2 cm segments of
conducting airways were collected from airway generations (G) 1, 5, 10 and 15; from
small airways (S) of approximately 1 mm diameter; and from parenchyma (P) without
macroscopically visible airways. In addition, stomach and colon tissues of two of these
animals were sampled. Immediately after excision, the tissues were submerged in 2—3 mL
of TriReagent (Molecular Research Center, Inc., Cincinnati, OH) and stored at —80 °C.
Total RNA was isolated from homogenized samples based on a previously described
method [208] and resuspended in nuclease-free water containing rRNasin (40
units/lOOuL). RNA concentrations were determined with a ﬂuorescent RNA-binding
assay (RiboGreen; Molecular Probes, Eugene, OR), using a SpectraMax GEMINI
spectroﬂuorometer (Molecular Devices Corp., Sunnyvale, CA), and individual samples
were diluted to a working concentration of SOng/uL. For each experimental group (i.e.,

control and RAO horses) equal amounts of RNA from each individual were combined to

118

form pooled samples of each lung sample site (i.e., G1, G5, G10 and G15; S; and P). The
pooling was necessary because the amount of RNA extracted from some of the individual
samples was insufﬁcient to run RT-PCR of MUC2, MUCSAC and ZO-l in duplicate
reactions. Electrophoresis of aliquots ﬁom each of the pooled samples was performed on
a 2% agarose/fonnaldehyde gel (NuSieve 3:1; FMC Bioproducts, Rockland, ME) to
insure the integrity of the RNA samples.

Identiﬁcation of equine mucin genes eqM U C2 and eqM U C 5A C—Adopting a
strategy previously employed by Guzman et al. (1996 [209]), we used PCR primers
(panspecies MUC primers; Table 1) which correspond to two octapeptide domains in the
C-terminal region that are conserved between the human and rat MUCSAC and MUC2
homologues in order to detect homologous equine sequences. Reverse transcriptase-
polymerase chain reaction (RT-PCR) was performed starting with a volume of 20 uL
containing PCR buffer plus 5mM MgClz, 1 mM each dNT P, 10 units rRNasin, 125 ng
oligo(dT) (Becton Dickinson, Bedford, MD), 100 ng total RNA from airway, stomach
and colon samples, and 40 units of MMLV reverse transcriptase (Promega). All RNA
samples were then incubated at 42°C for 15 min, followed by incubation at 95°C for 4
min (RT- reaction). A PCR master-mix (PCR buffer, 4 mM MgC12, 6 pmol each of
panspecies MUC forward and reverse primers [Table 1], and 1.25 units T aq DNA
polymerase [added after 85 0C pre-heating]) was heated to 85 °C for 5 min and then added
to the cDNA (double-stranded complementary DNA from RT-reaction) samples, for a
ﬁnal volume of 50 uL. Samples were then immediately heated to 95°C for 3 min and then
cycled 32 times at 95 °C for 308, 56 °C for 605, and 72 °C for 60s, after which an

additional ﬁnal extension step at 72 °C for 10 min was included. Electrophoresis of PCR

119

products (10 uL) was performed on 3 % agarose gels (NuSieve 3:1; FMC Bioproducts)
stained with ethidium bromide. PCR products of expected size (Table 1) were excised
from the gel and puriﬁed using the Wizard® PCR Preps DNA puriﬁcation system
(Promega) according to the manufacturer’s instructions, and then cloned in One Shot'l’M
competent E.coli (InvitroGen, Invitrogen Corporation, Carlsbad, California) using the
pCR2.lTOPOTM (InvitroGen) vector according to the manufacturer’s instructions.
Plasmid DNA was puriﬁed using the Wizard® Plasmid Preps (Promega) and 1000 ng in
6uL of water was submitted to the MSU sequencing facility for sequencing with M13
(forward) and -21M13 (reverse) primers using Big Dye Terminator chemistry, and a 377
ABI Prism DNA Sequencer. Computer analysis of the sequences was performed using
SquebTM with the Wisconsin software package version 10.1TM (Genetics Computer
Group (GCG), Inc. GCG, Madison, Wisconsin).

RT-PCR for eqM U C2 and eqM U C5AC and testing of tissue specificity—Based on
the two consensus sequences (see results) primers speciﬁc for eqMUC2 and for
eqMUCSAC (Table 1), respectively, were designed using Primer3 software]. RT-PCR
(protocol as above; 36 cycles as determined in preliminary linear range-ﬁnding
experiments) was performed on airway (individual samples and complete RAO and
control pools of G 1, 5, 10, 15; S and P) stomach, and colon samples to detect steady-
state levels of the respective mucins in the tissues. To control for DNA contamination, a
RT- (reaction mix without MMLV reverse transcriptase, Promega) reaction was run for
all samples, and all reactions were performed in duplicate. Duplicate PCR reactions were
always run independently, but were based on the same RNA samples (i.e., the RNA was

not independently extracted in duplicates). PCR products were run on a 3% agarose gel

120

(NuSieve 3:1), and, after ethidium bromide staining, documented with a Bio-Rad
ChemiDoc image acquisition system and Quantity One (v4.0) quantitation software (Bio-
Rad, Hercules, CA), running on a Dell OptiPlex GXl computer. To conﬁrm the
speciﬁcity of the eqMUC2 and eqMUCSAC assays, PCR products of predicted size were
excised from the gel and puriﬁed using the Wizard® PCR Preps DNA puriﬁcation
system (Promega). The PCR product DNA (~20 ng in 12 uL of water) was directly
sequenced using big dye terminator reactions with eqMUC2 and eqMUCSAC forward
and reverse primers (at 30 picomoles concentration), respectively, and sequences were
compared to the eqMUC2 and eqMUCSAC consensus sequences using SquebTM
(GCG).

EqMUCSAC /Z0-1 ratio on pooled RAO and pooled control samples at different
airway generations-The RNA of the airway RAO and control pools was analyzed to
determine the steady-state levels of eqMUCSAC mRNA by semi-quantitative assessment of
the amount of eqMUCSAC cDNA produced by RT-PCR in comparison to the amount of
ZO-l cDNA produced from the same sample. ZO—l, a tight junction gene, has been
previously used as an internal standard to control for the proportion of epithelial cells in the
samples [210]. RT-PCR using the eqMUCSAC primers (Table 1) was performed as
described above. RT-PCR using a primer pair for ZO-l (Table 1) was performed using the
same reaction mixtures and protocols as described above, except for a different PCR proﬁle:
After the RT reaction, samples were immediately heated to 94°C for 3 min and then
cycled 33 times at 94°C for 1 min, 51°C for 1 min, and 72 °C for 1 min, after which an
additional ﬁnal extension step at 72 °C for 10 min was included. The abundance of cDNA

(proportional to the primer speciﬁc mRNA in the sample) was semi-quantitatively

121

assessed by densitometric analysis as described above. EqMUCSAC and ZO-l RT-PCR
products were all run on one gel, and band volume values (OD [optical density] * mm2
[surface of product band]) were calculated from duplicates. Band volumes of the
eqMUCSAC RT-PCR product were divided by the band volumes of the ZO-l RT-PCR
product bands. This eqMUCSAC / ZO-l ratio was used as a semiquantitative measure of the
steady-state level of eqMUCSAC mRNA in epithelium at each airway generation in control

and RAO horses.

Results

Identification of equine mucin genes eqMUC2 and eqMUCSA C—RT-PCR
ampliﬁcation using the panspecies MUC primers (Table 1) of total RNA isolated from
horse airways, stomach and colon samples yielded cDNA fragments of about 550bp. This
is in agreement with the size range of cDNA fragments produced by these primers [209]
from human, MUCSAC: 591 bp, MUC2: 540 bp, and from rat samples, rMuc2: 528 bp.
Sequencing of these cloned cDNA fragments revealed two distinct mucin-like sequences.
Colon sample clones showed a high degree of nucleotide sequence identity (73-77%)
with MUC2 of other species (Table 2), indicating an equine homologue of MUC2
(proposed name eqMUC2; colon isolate GeneBank AF345996). The sequences from the
stomach and airway samples showed a high percentage of nucleotide sequence identity
(72-80%) with MUCSAC of other species (Table 2), indicating an equine homologue of
MUCSAC (EdMUCSAC; GeneBank AF345995). A comparison of the eqMUC2 and

eqMUCSAC fragments showed 57% nucleotide sequence identity.

122

 

RT -PCR for eqMUC2 and eqMUC5AC and testing of tissue speciﬁcity—RT-PCR
product sizes using the primers developed for eqMUC2 and eqMUC5AC were about 190
and 250 bp, respectively (Fig. 1), as predicted from the primer design (Table l), and
direct sequencing conﬁrmed the identity of the products with their respective target
sequences. RT-PCR on airway, stomach and colon samples revealed that eqMUC5AC is
expressed in equine stomach (clearly detectable in < 1 ng total RNA) and in all airway
generations but is undetectable in colon samples up to the limit of speciﬁc (RT- up to 42
cycles) reactions. Fig. 1 shows the resulting bands from representative RT-PCR reactions at
standard conditions, i.e., 36 cycles and 100 ng total RNA. EqMUC2 mRNA was detected in
colon (strong band from <1 ng of total sample RNA) and very faintly in stomach (at
standard conditions), but not in any of the airway tissue samples up to the limit of speciﬁc
(RT— up to 40 cycles) reactions.

E qM U C5A C / Z 0-1 on pooled control and pooled RAO airway samples—The
product bands of eqMUC5AC and ZO-l at each airway generation are shown in Fig. 2
along with a graphic representation of the ratios of their volumes. RT-PCR of ZO-l
produced two bands (~190 and 430 bp; Fig.2), as expected because of the alternative
splicing of its mRNA transcripts [210]. Band volume values (OD*mm2) of eqMUC5AC
and ZO-l (both bands) RT-PCR products are presented in Table 3. These values as well
as the eqMUC5AC / ZO-l ratios in Fig. 2 show that eqMUC5AC steady-state mRNA
levels (both absolute and compared to ZO-l) were highest at airway generations G1, G10
and P in the pooled samples from healthy control horses as well as in RAO-affected
animals. However, eqMUC5AC values were higher and ZO—l values lower in RAO than

in control pools at all airway generations. The ratio therefore accentuates the differences

123

between RAO and controls, which is small at G1, G16 and P, but pronounced at G5, G10

and S.

Discussion

We identiﬁed the ﬁrst mucin gene sequences described in the horse: equine
homologue fragments of MUC2 and MUCSAC (proposed names: eqMUC2 and
eqMUC5AC, respectively). The interspecies nucleotide sequence identity (72-80%; Table 2)
between equine, human, rat and mouse MUC2, and equine, human, pig, hamster, rat and
mouse MUCSAC, respectively, was closer than between eqMUC2 and eqMUC5AC (57%).
Our results conﬁrm previous reports [209] that the cysteine-rich region in the terminal
domain, which we identiﬁed and partly sequenced, is highly conserved across species. For
instance, between the predicted equine MUCSAC and the known porcine MUCSAC amino-
acid sequence 18 of 19 cysteines were conserved (results not shown). This underlines the
functional signiﬁcance of this region and, in particular, of the cysteines essential for
oligomerization, i.e., gel-formation [211].

The detection of eqMUC2 mRNA steady-state levels in the colon samples (Fig. 1) is
in accordance with the tissue distribution of this mucin in humans and rodents [212]. While
there is evidence for a role of MUC2 in cystic ﬁbrosis [99], and its expression in human
airway cells in vitro can be induced by gram-negative bacteria, IL-9, as well as other agents
and mediators [96, 213], data on the presence or absence of this mucin in healthy human
airways is conﬂicting [91, 214]. We found no evidence for eqMUC2 expression in healthy
equine airways or for a role for this mucin in RAO. In contrast, eqMUC5AC mRNA steady-

state levels were detected in stomach and in each of the airway samples, but not in colon

124

(Fig. 1). This tissue distribution corresponds to that observed in humans and rodents [212].
Based on reactivity with a mix of polyclonal antibodies raised against human MUCSAC
other investigators have proposed that a homologue of MUCSAC is present in equine
airway secretions [101] and gastric surface mucus [100]. Our results support these ﬁndings,
and the present evidence suggests that eqMUC5AC, but not eqMUC2 may be a major
component of airway mucus in both healthy and diseased equine lungs. However, besides
MUCSAC the other major component of human mucus identiﬁed is MUC5B [90, 91, 206,
215]. None of the cloned panspecies MUC RT-PCR products we analyzed to date were
homologues of MUCSB (unpublished results). It is possible that this mucin was not
expressed in the samples we investigated, since it is predominantly produced by submucosal
glands, which are abundant in larger human bronchi, but very sparse in horse airways [70,
108]. However, based on polyclonal antibodies raised against human MUCSB, there is
evidence that a homologue of this mucin is present in mucus from horse airways [102]. It
seems therefore more likely that the panspecies MUC primers (Table 1) did not allow us to
identify equine MUCSB homologue mRNA, even though it may have been present in the
samples we studied. Further work will be needed to address the question of MUCSB
expression in equine airways.

In a next step, we investigated eqMUC5AC steady-state levels at six different
airway generations of pooled samples from control horses and from RAO-affected
individuals. The intensity of eqMUC5AC product bands differed markedly between samples
from different airway generations (Table 3; Fig. 2). Interestingly, a similar pattern of
eqMUC5AC expression at different airway generations was observed both in control and

RAO samples: generation 1 and 10 as well as the peripheral samples showed the strongest

125

expression of eqMUC5AC. At each airway generation, however, band volume values were
higher in RAO than in controls. We did control quality of total pooled sample RNA, but the
proportion of mRNA from inﬂammatory cells in the total sample RNA could have
inﬂuenced these values. ZO-l, a tight junction gene, has been previously used to control for
this variation [210]. We therefore compared the volume values of the eqMUC5AC RT-PCR
product band to that of the ZO-l RT-PCR product band at each airway generation pool
sample. The eqMUC5 / ZO—l ratio accentuated the differences between RAO and controls,
which was small at G1, G16 and P, but very marked at G5, G10 and S (Fig.2).

The limitations of our control vs. RAO comparison are clear: a semi-quantitative
assay was used to evaluate differences between pooled samples of two groups with a
considerable mean age difference. Future studies will have to address interindividual
variability and the inﬂuence of age on mucin expression with quantitative RT-PCR, and
the present results must be interpreted with caution. Also, the biological signiﬁcance of
the variation of MUCSAC expression in the different airway generations of both RAO
and controls is unclear at this point. However, the consistency of the eqMUC5AC
increase in RAO compared to control pools at all airway generations may indicate an
important role of eqMUC5AC up-regulation in RAO. The large difference between RAO
and controls in the peripheral airways (Table 3; Fig. 2) may be functionally important due
to the very large proportion these small airways contribute to the total surface area of the
airways.

In murine airways, Muc5AC mRNA is a marker of goblet cell metaplasia [216],
and higher eqMUC5AC mRNA levels could simply be due to higher numbers of goblet

cells per total epithelial cells. Speciﬁcally, an increase of goblet cells in small airways,

126

where we observed the largest difference between controls and RAO, has been reported
by Kaup et al. (1990 [70, 217]). However, none of the studies reporting mucus epithelial
hyper- and metaplasia in airways of RAO-affected horses (reviewed by Dixon, 1992 [2]
and Robinson et al. 1996 [3]) have employed rigorous morphometric methods, and a
better understanding of the histological correlate of the differences in mucin expression
must await further studies. On the molecular level, various allergic, toxic and infectious
stimuli, and the corresponding inﬂammatory mediators have now been shown to regulate
expression of MUC2, MUCSAC and MUCSB [95, 96, 136, 145, 166, 170, 218-222]. Of
particular comparative interest is the responsiveness of MUCSAC, which has been shown to
be up-regulated by neutrophil- (i.e., oxidative stress, epidermal growth factor activation
[151, 170], and allergen- (i.e., IL-4, IL-9 and IL-13; [144, 213, 216] induced mechanisms.
Similar mechanisms may increase mucin mRNA levels in equine RAO, for which a Th2-
type cytokine proﬁle [158], severe neutrophilic airway inﬂammation associated with
increased oxidative stress [152], as well as NF-KB activation [164] have been reported.
The RT-PCR assays we developed provide the ﬁrst tools to investigate expression
of speciﬁc mucin genes in horse airways, and the mucin gene sequences we identiﬁed
may become the basis for development of equine mucin speciﬁc immunological assays.
We conclude that of the two identiﬁed equine mucin genes, which show close similarities
in the cysteine-rich ﬂanking regions with their homologues in other species, eqMUC5AC
but not eqMUC2 is expressed at detectable levels in horse airways. We further propose
that increased eqMUC5AC production, in particular in the small pulmonary airways, is a

potential primary mechanism responsible for mucus hypersecretion in equine RAO.

127

Table 7-1: Primer sequences, product size and references.

 

Name Forward Reverse Size Reference

 

Panspecies 5’ GGCCAGTGCGG 5’ GCCCTCCGGACA ~540bp’ .1, (14)
MUC CACTTGCACCAAC 3’ GAAGCAGCCTTC 3’ ~590bp° c"urns

~550bp°‘° paper

 

EqMUC2 5’ ACTGTCCCCACTC 5’ GGTACCACACAG l93bpc This paper

CAGCTT 3 ’ CTGTCGAA 3 ’

 

EqMUCSA 5’ CTGTGTGTITGAC 5’ TTCTGTGATGGG 247bp° This paper

 

C CAGTGCC 3’ TCCAGCTT 3’
ZO-l 5’ CATAGAATAGACT 5’ CTGCTGGCTTGT ~190bpe (33)
CCCCTGG 3’ TTCT CT AC 3’ ~430bp°

 

 

 

 

 

 

 

RT-PCR product size in basepairs (bp) from respective primers on MUC2“, MUCSAC°,
rMucSAC°, eqMUC2“, eqMUC5AC° and the human and equine homologue of ZO-le (two

product bands due to alternative splicing).

128

Table 7-2: Nucleotide homology of (NS) of eqMUC2 and eqMUC5AC with other

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mucins.
Mucin cDNA GenBank accession number NS. identity and gaps
EqMU C2 fragment"
MUC2 (human) L21998 77% and 3%
Rmuc2 (rat) M81920 73% and 3%
Muc2 (mouse) AJ010437 73% and 5%
EqMUCSAC ﬁ’agment“

MUCSAC-like (pig) AF054584 80% and 1%
MUCSAC (human) 248314 72% and 9%#
MUCSAC-like Lhamster) AF 288468 72% and 3%
RmucSAC (rat) U83139 72% and 2%
Muc5AC (mouse) AJ010792 73% and 2%

 

 

 

*GenBank accession numbers: AF 345996 for eqMUC2 and AF 345995 for eqMUC5AC.
All gaps were consistent with gain or loss of complete codons; #gap percentage higher because of
18 amino-acid insertion in human MUCSAC sequence.

129

Table 7-3: EqMUCSAC and ZO-l on pooled RAO and pooled control airway

 

 

 

 

 

 

samples.
Airway generationi
G1 G5 G10 G16 S P
EqMUCSAC Control 77.8 1.5 43.3 3.9 0.2 66.1
*
RAO A 82.2 34.7 70.8 6.9 66.4 83.8
ZO-1* Control 75.6 13.5 84.3 53.7 17.3 69.6
RAO 40.1 4.1 19.4 8.1 10.4 54.6

 

 

 

 

 

 

 

 

 

 

*Band volumes (OD*rnm2) of eqMUC5AC and ZO-l (both bands) RT-PCR products, respectively
(run all on one gel and calculated from duplicates). IAirway generations (G) 1, 5, 10 and 15; small
airways (S) of approximately 1 mm diameter; and parenchyma (P) without macroscopically

visible airways.

130

Fig. 7-1: Tissue speciﬁcity eqMUC2 and eqMUC5AC mRNA detected by RT-PCR.
Representative samples from airway generation 1 (G1) from control and recurrent
airway obstruction (RAO)- affected horses, as well as from stomach and colon

tissue. Ethidium bromide stained PCR products were run on 3 % agarose gels.

MUC SAC

190 bp

control RAO stomach colon 100 bp control RAO stomach colon
G1 G1 ladder Gt G1

 

131

Fig. 7-2: EqMUCSAC and ZO-l RT-PCR cDNA products at each airway generation
and graphic representation of the ratios of the band volumes. All cDN A products
were run on the same gel (ethidium bromide stained 3 % agarose), lanes shown
were cut and pasted with graphic software. Pooled samples for control and
recurrent airway obstruction (RAO)- affected horses from airway generations (G) 1,
5, 10 and 15; from small airways (S) of approximately 1 mm diameter; and from

parenchyma (P) without macroscopically visible airways.

 

 

    
    

 

 

 

10
O 8 r
E T f I Control
2 6 _ ? Rao
8 2
° /
1t 4 /
m “-1
., r
a 2 _ t
/
¢
0 _ I.
G1 65 G10 G15 5 P
MUCSAC .‘ m “I
1 h :1 tr til-n:
ZO-‘l ‘
MUCSAC
ZO-‘l

 

 

 

 

132

CHAPTER 8
MUCSAC mRNA LEVELS, BUT NOT INTRAEPITHELIAL
MUCOSUBSTANCE, ARE ASSOCIATED WITH EQUINE CACCl-LIKE AND
EGFR mRNA LEVELS AND WITH INTRALUMINAL NEUTROPHILS IN
SMALL CARTILAGINOUS AIRWAYS OF RAO-AFFECTED AND

CLINICALLY HEALTHY HORSES

I showed that unfavorable rheological properties (chapter 5 and 6) can only partly
explain the increased airway mucus accumulation in RAO (chapter 2-4) and that
equine(eq)MUC5AC, but not eqMUC2 may play a role in recurrent airway obstruction
(chapter 7). Preliminary results indicated increased eqMUC5AC mRNA levels in RAO-
affected compared to control horses at six different airway generations. These results were
based on a semi-quantitative RT-PCR assay, however, and the airway generation samples
had been pooled from several individuals per RAO and control group. I therefore developed
a quantitative RT-PCR assay to accurately determine eqMUC5AC mRNA levels on
individual samples and test the hypothesis that eqMUC5AC expression is increased in
equine RAO.

Furthermore, I investigated the association of eqMUC5AC expression with stored
intraepithelial mucosubstrance and with the expression of key signaling elements involved
in the regulation of mucin expression and mucous cell metaplasia. MUCSAC is the main
“signature” mucin in experimental and natural models of airway disease. Many different,
Th-l and -2 type as well as innate immunity, pathways converge on up-regulation of

MUCSAC with subsequent mucous cell metaplasia and hypersecretion [136]. Epidermal

133

growth factor (EGF), its receptor, EGFR and homologues of the recently identiﬁed
human calcium-activated chloride channel 1 (hCACCl) are key signaling molecules

involved in these pathways, but have not been investigated in equine lung disease.

Abstract

We have previously found glycosylation alterations of mucin glycoproteins in
recurrent airway obstruction (RAO)-affected horses that may be associated with increased
secretion of equine(eq)MUC5AC. Epidermal grth factor (EGF), its receptor, EGFR and
CACCl are key signaling molecules involved in mucus cell meta- and hyperplasia and
mucin gene expression, but have not been investigated in equine lung disease. We
hypothesized that exposure to irritants and aeroallergens would lead to increased
eqMUC5AC expression and stored mucosubstance in the epithelium of RAO-affected
horses, associated with increased neutrophils and CACCl, EGFR and EGF mRNA levels.
We therefore identiﬁed equine homologues of human(h)CACCl and EGFR, and developed
real-time quantitative RT-PCR to determine mRNA levels of eqMUC5AC, equine
hCACCl-like, EGFR and EGF. These assays were performed on small cartilaginous
airways from cranial left and right and caudal left and right lung lobes of 5 clinically healthy
and 5 RAO-affected horses that had been exposed to indoor stable environment for 5 days
before euthanasia. Morphometric measurements of volume densities of intraepithelial
mucosubstances, as well as cytological differentiation of intraluminal inﬂammatory cells
from the same airways were also performed. Neutrophil percentages were increased and
macrophage percentages proportionally decreased in RAO-affected horses, but no
signiﬁcant differences of eqMUC5AC, equine hCACCl-like, EGFR and EGF mRNA

levels or stored intraepithelial mucosubstance were found between RAO-affected and

134

control horses. Thus, the present results did not support the hypothesis that eqMUC5AC
is up-regulated and intraepithelial mucosubstances are increased in RAO-affected
compared to clinically healthy control horses. However, equine hCACCl-like and EGF R
mRNA levels as well as neutrophil percentages were associated with eqMUC5AC
mRNA levels in airways of RAO-affected and of clinically healthy horses. This suggests
a role of these key signaling molecules in mucin regulation in airways of horses suffering

from RAO as well as of horses with milder degrees of airway inﬂammation.

Introduction

Irritants (e.g., endotoxin) and allergens (e.g., mold spores are abundant in the
enviromnent of stabled horses [6, 7, 223] and induce clinical disease characterized by
bronchospasm, intense neutrophilic airway inﬂammation and mucus accumulation in RAO-
affected animals [3, 182], and milder neutrophilic inﬂammation without overt clinical signs
in clinically normal horses [47 , 174].

Similar initants and allergens can increase mRNA steady state levels of speciﬁc
mucin genes, in particular MUCSAC, and concomitant mucous epithelial hyper- and
metaplasia in rodents [97, 222]. We found persistent alterations in glycosylation patterns of
mucin glycoproteins in RAO-affected horses [54] that may be associated with increased
secretion of MUCSAC. Therefore, we identiﬁed the equine homologues of MUCSAC and
MUC2 [224]. A semi-quantitative reverse transcriptase-polymerase chain reaction (RT-
PCR) assay showed no detectable mRNA levels of equine(eq)MUCZ, but robust

expression of eqMUC5AC throughout the bronchial tree of RAO-affected and clinically

135

healthy horses. Semi-quantitative comparison on pooled samples suggested increased
expression in RAO-affected animals.

MUCSAC can be up-regulated by neutrophil- [135, 175], and allergen— [118, 131]
induced mechanisms. Several cytokines that are increased in equine RAO are known to
up-regulate and activate epidermal growth factor receptor (EGFR) through epidermal
growth factor (EGF) and other ligands. Neutrophils can upregulate EGFR expression
through TNF-or and subsequently activate EGFR in a ligand-independent fashion through
the release of oxidative radicals [134, 135]. Recently characterized calcium-activated
chloride channels (CACC, in particular homologues of human(h)CACC1) are associated
with increased stored intraepithelial mucosubstance, a measure of mucus cell meta- and
hyperplasia, and increased MUCSAC expression in animal models of allergic airway
disease as well as in asthmatic subjects [131-133]. Thus, EGFR and CACC are key
signaling molecules involved in mucus cell meta- and hyperplasia and mucin gene
expression, but have not been identiﬁed in the horse.

We hypothesized that exposure to irritants and aeroallergens would lead to increased
eqMUC5AC expression and stored mucosubstance in the epithelium of RAO-affected
horses compared with clinically normal control animals. We further proposed that
eqMUC5AC mRNA levels and stored intraepithelial mucosubstance are associated with
neutrophils and CACC], EGFR and EGF mRNA levels.

The aims of this study were, therefore, to identify equine homologues of CACC and
EGFR, and develop real-time quantitative RT-PCR to determine mRNA levels of
eqMUC5AC, equine hCACCl-like, EGF R and EGF mRNA levels along with stored

intraepithelial mucosubstance in airways of clinically healthy and RAO-affected horses, and

136

to investigate regional distribution of gene expression and stored intraepithelial

mucosubstance.

Materials and methods

Animals, environmental exposure, euthanasia and sample collection-This study
was approved by the All-University Committee for Animal Use and Care of Michigan
State University. Five horses (2 mares and 3 geldings), 14 to 21 years old (mean i s.e.m.,
17.4 d: 3.4 years), with a previous diagnosis of RAO, served as the principal group. Five
horses (3 mares and 2 geldings), aged between 18 and 28 years (mean :t s.e.m., 22.8 :t
4.1), and without clinical evidence of respiratory tract disease (i.e., respiratory distress,
coughing), served as the control group. Horses were exposed to initants and aeroallergens
by stabling in stalls with straw bedding and feeding hay [6, 7, 223] for ﬁve days. Clinical
signs of airway obstruction, i.e., increased breathing effort, were numerically scored as
previously described [178]. In order to enter the study, control animals had to have a
clinical score of S 4 (3.1 i 0.9; range 2-4) and RAO horses had to develop a clinical score
of 2’. 5 (6.2 i 1.2; range 5-8).

Samples for gene expression studies, morphometry and cytology were collected
from the right cranial (a) and caudal (b) as well as the left cranial (c) and caudal (d) lung
lobes within 30 minutes after euthanasia by barbiturate overdose. In each lung lobe 1cm
length of a 2-3 mm diameter small cartilaginous conducting airway was excised for RNA
isolation and immediately submerged in 2-3 mL of TriReagent (Molecular Research
Center, Inc., Cincinnati, OH) and stored at —80 °C. For morphometric analysis an

adjacent cross-section of the same airway was removed and submerged in ﬁxative, and

137

for cytological analysis intraluminal secretions were sampled with a small cytology brush
from the same airway distal to the excised tissues.

Cytological analysis of intraluminal secretions—Intraluminal secretions sampled
by cytology brush were diluted in 500 pl of phosphate-buffered saline. Cell smears were
made from the diluted samples with a cytocentriﬁrge and stained with Diff-Quick.
Differential cell counts of macrophages, neutrophils, lymphocytes, eosinophils and mast
cells were performed by counting 200 cells per sample using standard morphologic
criteria under a light- microscope. The same person performed all cell scoring without
knowledge of diagnoses. Differentials are expressed as percent of total cells.

Tissue preparation and morphometry of stored intraepithelial muco-
substances—After ﬁxation for 40 hours in 1% paraforrnaldehyde and 0.1%
glutaraldehyde, the airway cross-sections were transferred to 30% EtOH and then
embedded in glycol methacrylate, and 1-2 pm ultrathin sections were cut from the
anterior surface. Sections were stained with Alcian Blue (pH 2.5)/Periodic Acid-Schiff
(AB/PAS) to detect intraepithelial mucosubstances. To estimate the amount of the
intraepithelial mucosubstances in the epithelium lining the axial airways, the volume
density of AB/PAS-stained mucosubstances was quantiﬁed using computerized image
analysis and standard morphometric techniques. The area of AB/PAS stained
mucosubstance was calculated from the automatically circumscribed perimeter of stained
material using a Dell Optiplex GX260 computer and the public domain NIH Image
program (written by Wayne Rasband, US. National Institutes of Health and available on
the Internet at http://rsb.info.nih.gov/nih-image/). The length of the basal lamina

underlying the surface epithelium was calculated from the contour length of the digitized

138

image of the basal lamina. The volume of stored mucosubstances per unit of surface area
of epithelial basal lamina was estimated using a method described previously in detail
[225]. The volume density of intraepithelial mucosubstances is expressed as nanoliters of
intraepithelial mucosubstances per mm2 of basal lamina. The same person performed the
morphometrical analysis without knowledge of diagnoses.

Processing of RNA samples and reverse transcription of total RNA—Total RNA
was isolated from homogenized samples based on the method of Chomczynski and
Sacchi [208]. Isolated RNA pellets were resuspended in nuclease-free water and
incubated with DNase solution [100 units rRNasin (Promega, Madison, WI), 100 mM
DTT (Life Sciences Technology Inc., Grand Island, NY), and 10 units DNase I
(Boehringer Mannheim, Indianapolis, IN) in 5X transcription buffer (Promega)] for 45
min. at 37°C. The RNA was extracted sequentially with equal volumes of
phenol/chloroform/isoamyl alcohol (25:24:1) and chloroform/isoamyl alcohol (24:1), and
precipitated with 10 M ammonium acetate and isopropanol. The pellet was washed with
75% ethanol, air dried, and resuspended in nuclease-free water containing rRNasin (40
units/100111). Total RNA concentrations were determined using a PowerWavex
spectrophotometer (Bio-Tek Instruments, Inc.). Reverse transcriptase reaction was
performed in a volume of 20 [11 containing 2 pl 10x buffer, 5 mM each dNTP, 1 unit! [.11
rRNasin (SUPERase'Inm; Arnbion, Inc., 2130 Woodward Austin, TX 78744-1832,
USA), 10 uM random hexamer primers, 2 ug total RNA, and 4 units of Omniscript
reverse transcriptase (Qiagen, Inc., 28159 Avenue Stanford, Valencia, CA 91355, USA).
All RNA samples in the 20 u] reaction mix were then incubated at 42°C for 60 min,

followed by an incubation at 93°C for 5 min to inactivate the reverse transcriptase. To

139

control for DNA contamination, a RT- (reaction mix without reverse transcriptase)
reaction was run for all samples. Preliminary quantitative PCR (protocol see below)
showed DNA contamination in 5 of the samples. 2ug RNA of each of these 5 samples
was again DNase-treated with DNA-freeTM (Ambion) according to the manufacturer’s
instructions. Two of these 5 samples subsequently still showed DNA contamination and
were eliminated ﬁ‘om the study. Six other samples were eliminated from the study
because insufﬁcient amounts of total RNA were available.

Identiﬁcation of equine homologous partial sequences of CA C C1 and
EGFR—Cross-species conserved regions were identiﬁed by comparison of human and
mouse CACCl (GeneBank [GenBank Online. NCBI. http://www.ncbi.nlm.nih.gov)]
accession numbers gi4585468 and gi3721911, respectively) and EGFR (giZ9725608 and
gi23271839, respectively) homologue nucleotide sequences, using the basic local
alignment search tool (BLAST) 2 Sequences comparison [226] available through the
NCBI web site (http://www.ncbi.nhn.nih.gov/BLAST/). Based on the identiﬁed
conserved regions, cross-species consensus primers were developed. Primers were
designed over less degenerate or mutable amino acids within regions of greatest
nucleotide conservation. In most cases, the 3' end of each primer was designed to match
the second codon position of an amino acid with low mutability to minimize the
possibility of a mismatch at this position [227]. The second codon position for most
amino acids is known to be less mutable than the ﬁrst or third positions [228] (P.J. Venta
personal communication). Melting temperatures of candidate primers were determined
using algorithms available at the Virtual Genome Center website:

(http://alces.med.umn.edu)

140

Candidate primer pairs were tested by amplifying equine airway cDNA and
sequencing PCR products. A PCR master-mix consisting of 25 ul 2x PCR buffer
(including HotStarTaq DNA Polymerase and QuantiTect SYBR Green; Qiagen), 0.4 uM
each of forward and reverse primers, and the cDNA samples (equivalent to 25 ng starting
total RNA) for a ﬁnal volume of 50 ul was heated to 95°C for 15 min and then cycled 50
times at 92 °C for 305, variable annealing temperature based on primer melting
temperatures for 605, and 70 °C for 608 in a iCycler thermal cycler with iCycler iQ real-
time PCR detection system (Bio-Rad Laboratories, 1000 Alfred Nobel Drive Hercules,
CA 94547). At the end of the PCR, the temperature was increased from 60 to 95 °C at a
rate of 2 °C/min, and the ﬂuorescence was measured every 15 s to construct the PCR
product dissociation (melt) curve. A nontemplate control (NTC) was run with every
assay, and all determinations were performed at least in duplicates to achieve
reproducibility. When speciﬁc (single peak melt curve, primer-dimer melting
temperature) ampliﬁcation was observed, PCR samples were further analyzed.
Electrophoresis of PCR products (10 ul) was performed on 3 % agarose gels (NuSieve
3:1; FMC Bioproducts) that were stained with ethidium bromide. PCR products were
excised from the gel and puriﬁed using the Wizard® PCR Preps DNA puriﬁcation
system (Promega) according to the manufacturer’s instructions, and then submitted for
sequencing with speciﬁc forward or reverse primers using Big Dye Terminator
chemistry, MJ Research DNA engine-PTC-ZOO for extension reactions and 377 ABI
Prism DNA Sequencer to run gels after ethanol precipitation. Analysis of the sequences
was performed using the standard nucleotide-nucleotide BLAST available through the

NCBI web site (http://www.ncbi.nlm.nih.gov). The primer pair 5’-

141

ATTGGACAAAGAATTGTGTG-3’ / 5’-ACAATTTCAGATCC ATCAGTT-3’ (anneal-
ing temperature 52 °C) ampliﬁed a partial cDNA sequence of an equine hCACCl-like
homologue and the primer pair 5’-GAGAGGAGAACTGCCAGAA-3’ / 5’-
GTAGCATTTATGGAGAGTG-3’ (annealing temperature 47 °C) ampliﬁed a partial
cDNA sequence of an equine EGFR homologue.

Selection of primers for quantitative RT -PCR—Using Primer3 software (Steve
Rozen, Helen J. Skaletsky (1998) Primer3. Online. Code available at http://www-
genomewi.mit.edu/genome_software), primers with PCR products < 150 bp suitable for
quantitative real time PCR , were designed based on an eqMUC5AC partial sequence
(GeneBank accession number gi:19070191) we had previously identiﬁed [224], a
published [229] complete cDNA sequence of equine EGF (gi:688071), a published [230]
partial sequence of the equine 18s rRNA gene (gi218369730) and the newly identiﬁed
partial equine homologue sequences of EGFR and hCACCl (see results). Quantitative
real-time PCR (standardized protocol as above; annealing temperature 57 °C for all
primers) was performed on airway (individual and pooled samples) cDNA. Primers were
selected based on linear performance over a 5 log dilution range, 100 ng to 10pg starting
RNA template per reaction. Speciﬁcity of ampliﬁed PCR products was monitored by
melt curve analysis and conﬁrmed by identiﬁcation of a single band on gel
electrophoresis, excision, puriﬁcation and sequencing of RCR product as described
above. Primers selected for real-time PCR are listed in Table 1. To assess if a primer pair
ampliﬁes genomic DNA, and what size the PCR product is, genomic DNA templates
were also tested. To control for DNA contamination, RT- reactions were run for all

samples and to detect primer-dimer formation negative (no template) reactions were

142

performed. Primers selected for real-time PCR based on product speciﬁcity, linear,
efﬁcient performance and absence of primer-dimer artifacts are listed in Table 1.

Relative quantitation of mRNA levels-Real-time PCR product accumulation was
monitored using the intercalating dye, SYBR® Green I, which exhibits a higher
ﬂuorescence upon binding of double-stranded DNA. Relative gene expression was
calculated using conditions at the early stages of PCR, when ampliﬁcation was
logarithmic and, thus, could be correlated to initial copy number of gene transcripts. The
ﬁrst 5-10 initial cycles were considered as a baseline or background in which no changes
in ﬂuorescence intensity occur. The level above this baseline, at which increments in
ﬂuorescence become detectable, is termed the threshold. For each target gene, the
threshold was set at a predetermined value > 10 times the standard deviation of the mean
baseline emission calculated for the baseline cycles. The iCycler iQ real-time PCR
detection system (Bio-Rad Laboratories) software determined the cycle number when a
reaction reached the threshold. This value, termed the 'cycle threshold' (Ct), appears
during the exponential phase of the PCR and is inversely proportional to the initial
number of template molecules in the sample. Standard curves (100ng, 50ng, lOng, Sng
and lng starting RNA/reaction) of cDNA pooled from 5 samples were run in triplicate on
the same plate as the unknown samples for each target gene. These standard curves
displayed a linear relationship between Ct values and the logarithm of the starting RNA
concentrations calculated using linear regression analysis. All unknown samples were
within the C, range of the standard curves for all target genes. Amount of target in each
unknown sample was determined by interpolation from the standard curve of the target

gene relative to the amount of ribosomal RNA in the same sample calculated from the

143

standard curve for 18s rRNA. This unit-less ratio was then expressed as an x-fold
difference compared to the sample that showed the lowest levels for all target genes.
Coefﬁcients of variance of within-run and between-run assays were less than 5% for all
assays. The correlation coefﬁcient of the RNA concentrations and Ct values was >0.95
for all standard curves. The same person performed all relative quantitative RT-PCR
without knowledge of diagnoses.

Statistical analysis—All statistical analyses were performed using a statistical
software program (SAS2 for Windows95, version 8.02). Dependent measures were
analyzed by repeated measures ANOVA using the general linear model. Percentages of
macrophages, neutrophils, lymphocytes, eosinophils and mast cells; volume density of
intraepithelial mucosubstances; and eqMUC5AC, equine hCACCl-like, EGFR and EGF
mRNA levels were the dependent measures of interest. Independent factors in the model
included two levels (RAO-affected and control horses) of the between subjects factor,
group, and four levels (right cranial lobe {a}, left cranial lobe {b}, right caudal lobe {c}
and left caudal lobe {d}) of the repeated measure, location. Partial correlation analysis
was used to probe the relationships of the two endpoint measures, eqMUC5AC mRNA
levels and volume density of intraepithelial mucosubstances with each other, with
neutrophil percentages and mRNA levels of equine hCACCl-like, EGFR and EGF.
Partial variables included the 2 levels of group and 10 levels of individuals. For all
analyses, signiﬁcant differences were declared if P < 0.05. All data are herein expressed

as the means i SD.

144

Results

Cytological analysis of intraluminal secretions—A complete set of 40 airway
secretion samples was available for cytological evaluation. Summary data are presented
in Table 3. There was a signiﬁcant effect of group on macrophage and neutrophil
percentages. Neutrophils were more than double (66 i 21 %) in RAO-affected horses
than in control animals (29 i 29 %). Macrophages (20 i 12 % vs. 44 i 23 %) and all
other cell types were proportionally decreased in RAO horses compared to the control
group. For lymphocyte, eosinophil, and mast cell counts, there were no main or
interaction effects. Localization (cranial vs. caudal, left vs. right lobes) had no effect on
any of the cytological variables.

Morph ometry of stored intraepithelial mucosubstances ~Thirty—seven airways
could be morphometrically evaluated, in three airways volume density could not be
measured due to excessive artefacts on the histological slides. Thevolume density of
intraepithelial mucosubstances showed no signiﬁcant differences between groups or
between lung lobes. Means were similar between RAO-affected and control horses
(Table 1), but individual measurements varied considerable between and also within
horses (Fig. 1A). Within the same horse, for example horse V, Fig. 1A) some airways
may appear to store mucosubstance in rounded goblet cells (lobe c, Fig. 1A and B), while
other airways showed goblet cells expressing mucus (lobe a, Fig. 1A and C), or were
almost completely devoid of intraepithelial mucosubstance (lobe b, Fig. 1A). The volume
density of intraepithelial mucosubstances was not signiﬁcantly correlated with

neutrophils or mRNA levels of eqMUC5AC, equine hCACCl-like, EGFR or EGF.

145

Identiﬁcation of equine hCA C C 1 -like and E GFR homologues—RT-PCR
ampliﬁcation using the cross-species CACCl primer pair (see methods) yielded a cDNA
fragment of about 293bp. Sequencing of this cDNA fragment revealed a high degree of
nucleotide sequence identity (257/298bp nucleotides; 86%) with human CACCl
(=hCLCA1; giz4502864), indicating an equine CACCl-like gene cDNA fragment. Other
similarities were found with a porcine chloride channel protein (194/234bp, 82%;
gi:6002645), with mouse(m)CLCA3 (=gob-5, =mCACC1; 85/96bp, 88%; giz8567335
and gi:3721911) and with rat mCACC2/3-like (144/ l76bp, 81%; gi:27729602). RT-PCR
ampliﬁcation using the cross-species EGFR primer pair (see methods) yielded a cDNA
fragment of about 406bp. Sequencing of this cDNA fragment revealed a high degree of
nucleotide sequence identity (366/4l3bp, 88%) with human EGFR (gi:29725608), mouse
EGFR (342/391bp, 87%; giz458123), rat EGFR (337/385bp, 87%; gi:25742616), porcine
EGFR (343/394bp, 87%; gi:21913l75), rabbit EGFR (158/176bp, 89%; gi:l3l73350)
and chicken EGFR (288/358bp, 80%; gi:211740), indicating partial equine EGFR
homologue cDNA.

Relative quantitative RT -PCR—Thirty-four airway RNA samples were available
for quantitative RT-PCR of eqMUC5AC, equine hCACCl-like, EGFR and EGF. None of
these target genes showed signiﬁcant differences between groups (RAO-affected and
control horses) or between lung lobes (a, b, c and d). EqMUCSAC showed very large
variation between horses but relative good agreement between lung lobe samples within
horses (Fig. 3A). Mean EqMUCSAC mRNA levels were higher in the control group
(Table 1), although not signiﬁcantly, due mainly to the caudal lobe samples in one

control horse (ii, b and (1). Mean equine hCACCl-like mRNA levels were higher in the

146

RAO group (Table 1), but again not signiﬁcantly because of large variations between
horses with relatively consistent levels within individuals (Fig. 38). Similarly, for EGFR
mRNA levels it was again horse I of the RAO-group that had the highest expression
levels (Fig. 3C), but overall there was less difference between horses than for equine
hCACCl-like. EGF showed somewhat less consistency within horses than the other gene
targets, but the least apparent differences between individuals (Fig. 3D). Pearson
correlation analysis on means of all lobes per individual showed signiﬁcant correlations
between eqMUC5AC mRNA levels and equine hCACCl-like (partial correlation
coefﬁcient r = 0.47, P= 0.002; Fig. 3A) and EGFR (r = 0.3471, F = 0.028; Fig. 3B)
mRNA levels as well as neutrophil percentages (r = 0.34, P = 0.032; Fig. 3C), but no
signiﬁcant association was found with volume density of intraepithelial mucosubstance

and mRNA levels of EGF.

Discussion

In the present study, mRNA levels of eqMUC5AC, but not intraepithelial
mucosubstance were correlated with equine hCACCl-like and EGFR mRNA levels and
with neutrophil percentages in small cartilaginous airways of RAO-affected and control
horses. No signiﬁcant differences of eqMUC5AC, equine hCACCl-like, EGFR and EGF
mRNA levels or stored intraepithelial mucosubstance were found between RAO-affected
and control horses, however.

Based on our previous studies [51, 179, 182, 231] using endoscopic observation to
quantify visible mucus accumulations in the airways, we have estimated that 2-8 ml per

hour of endoscopically visible mucus are transported along the trachea in clinically

147

healthy horses. Compared with healthy animals, in RAO-affected animals these volumes
can be extrapolated to 2-3 times more per hour at pasture and 3-10 times more after
exposure to the dust-laden stable environment. These increased mucus accumulations are
associated with the neutrophilic inﬂammation in the airways [51, 179].

Overall, mucus accumulation in RAO-affected horses is likely due to a
combination of factors, which all may be related to airway neutrophilia, a hallmark of
RAO. Neutrophils can potentially cause to both mucus hypersecretion and decreased
clearance. Since we found that decreased clearability of mucus can only partly explain
the increased accumulations in RAO [169], we hypothesized that up-regulation of
production at the level of mucin-gene expression may contribute to the mucus
accumulation. This would be in accordance with the often described, but never proven
epithelial mucus cell meta- and hyperplasia in RAO [2, 3]. Indeed, we found persistent
alterations in glycosylation patterns of mucin glycoproteins in RAO-affected horses [54]
that may be associated with increased secretion of MUCSAC. Therefore, we identiﬁed
equine homologues of MUCSAC and MUC2 [224]. A semi-quantitative reverse
transcriptase-polymerase chain reaction (RT-PCR) assay showed no detectable mRNA
levels of equine(eq)MUCZ, but robust expression of eqMUC5AC throughout the bronchial
tree of RAO-affected and clinically healthy horses. Semi-quantitative comparison on pooled
samples suggested increased expression in RAO-affected animals.

In this study, we had therefore hypothesized that exposure to irritants and
aeroallergens would lead to increased eqMUC5AC expression and stored mucosubstance in

the epithelium of RAO-affected horses compared with clinically normal control animals,

148

and further proposed that eqMUC5AC mRNA levels and stored intraepithelial
mucosubstance are associated with neutrophils and CACCl , EGFR and EGF mRNA levels.
The ﬁrst part of this hypothesis was clearly not supported by our present ﬁndings.
Only neutrophil and macrophage percentages were signiﬁcantly different between
disease groups. This is not surprising since neutrophils invade the airways of RAO-
affected horses within 6 hours of exposure [14] to high dust loads, in particular endotoxin
and mould allergens from hay feeding and straw bedding [6, 8, 156, 232]. The
experimental protocol of ﬁve days of exposure was chosen because we expected maximal
differences at this time point. Within a day of exposure endoscopically visible airway
mucus accumulation approximately doubles in RAO-affected animals, but remains
unchanged in clinically healthy controls [182]. Increased accumulation scores are not
speciﬁc for RAO, however, and can be observed in various equine lower airway diseases
[5]. Even in clinically healthy sport horses that perform well there is large variation and
markedly increased airway mucus accumulation is frequently observed [231].
Furthermore, even though maximal clinical signs, mainly reﬂective of
bronchospasm [3, 178], can be consistently reproduced in RAO-affected horses exposed
for several days [3], clinical scores in this study were not very different between RAO-
affected (6.2 i 1.2; range 5-8; cut-off RAO: 5 or more [178]) and clinically healthy
horses (3.1 i 0.9; range 2-4; cut off clinically normal: up to 4 [178]). However, animals
with normal clinical scores can have subclinical airway obstruction [178]. This and the
fact that mean (Table 2) and individual (results not shown) neutrophil percentages in
control horses were well above proposed normal values in small airway secretions of

horses [12] indicate that after the 5 days of exposure there was subclinical, but signiﬁcant

149

inﬂammatory airway disease among the animals in the control group. This is in
accordance with the previously observed high incidence and large variability of
neutrophilic inﬂammation and mucus accrunulation in airways of stabled, clinically
healthy horses [47, 182, 231] and may explain the lack of signiﬁcant differences of
intraepithelial mucosubstance and eqMUC5AC mRNA levels between the disease groups
in this study. Methodological differences could also have contributed to the discrepancy
between our earlier preliminary [224] and the present results. In our earlier report we had
used a semi-quantitative RT-PCR to test samples that were pooled by disease group. The
present results show that large (> 100x) differences in eqMUC5AC mRNA levels
between individuals (Fig. 1A) could have disproportionately inﬂuenced the previous
results based on the semi-quantitative comparison of the pooled samples.

Mucus epithelial hyperplasia and metaplasia in airways of RAO-affected horses
have been reported in numerous studies (reviewed by Dixon [7] and Robinson et al. [29]).
An increase in goblet cells speciﬁcally in small conducting airways, which we investigated
in this study, has been reported [70, 217]. None of these reports are based on quantitative
measurements, however. In contrast, the present study is the ﬁrst to employ rigorous
morphometric methods. There are two explanations for the lack of a difference observed in
this study. RAO-affected horses may have low, (i.e., normal) amounts of stored
intraepithelial mucosubstance. Alternatively, the clinically healthy control horses, which had
evidence of signiﬁcant airway inﬂammation, had mucus epithelial hyperplasia and
metaplasia. Comparison with volume density of intraepithelial mucosubstance obtained by
the same morphometric method in proximal and distal axial airways of rats shows that

means in horses of both disease groups are more than 10-fold higher than in untreated

150

rats, and are in the range of values found in rats exposed to ozone and endotoxin. Even
though such cross-species comparisons, especially with rodents, must be interpreted very
cautiously, this supports the alternative explanation, that clinically healthy control horses
have signiﬁcant mucus epithelial hyperplasia and metaplasia. Ideally, we should have
included a control group of clinically healthy horses that have never been exposed to stable
environment, but such age-matched controls are virtually impossible to ﬁnd in the Northern
hemisphere.

Considerable variation of intraepithelial mucosubstances was noted not only
between, but also within horses (Fig. 1A). Statistical analysis showed neither regularity
(e. g., cranial vs. caudal lobes) of this variation, nor any association with neutrophil
percentages, eqMUCSAC, equine hCACCl-like, EGFR or EGF mRNA levels. Subjective
observations suggested that within the same horse different airways appeared to be in
different stages of mucosubstance storage (Fig. 18) or collective release (Fig. 1C). We have
not quantiﬁed this observation and can only speculate that consistency within airways, but
heterogeneity within the lung of these storage and release stages may explain the variations
in intraepithelial mucosubstances. In absence of a better biological explanation, artefactual
variation, i.e., release of mucosubstance post mortem due to traumatic handling, must also
be considered. We can only stress that all samples were dissected and collected with the
utmost care and immediately placed in ﬁxative.

In contrast to intraepithelial mucosubstances, the relative, quantitative assessment
of eqMUC5AC, equine hCACCl-like, EGFR or EGF mRNA levels showed remarkable
consistency within individuals (Fig. 2A-D), indicating that the underlying pathogenesis is

a diffuse process involving the entire organ. Conversely, differences between horses were

151

considerable. These differences reﬂect the variability in the degrees of mucus
accumulation observed in RAO-affected and particularly also in clinically healthy horses
[182, 231]. While the present results cannot explain the increased mucus in RAO
airways, they may indicate that in some horses, RAO-affected, but also clinically healthy
with milder degrees of airway disease, eqMUC5AC mRNA up-regulation may contribute
to mucus hypersecretion. An important link, which has been demonstrated in other
species, but not in the horse, is the demonstration that mucin mRNA levels are associated
with mucin protein translation and ﬁnally secretion.

Perhaps the most exciting results of this study were associations of eqMUC5AC
mRNA levels with equine hCACCl-like and EGFR mRNA levels and with neutrophil
percentages, but not with EGF mRNA levels. Mindﬁil of the dangers of drawing
conclusions from multiple correlations in a limited data set, these associations connect
mucin gene expression with key signaling elements in the pathogenesis of airway mucus
cell hyperplasia and hypersecretion. '

MUCSAC can be up—regulated by neutrophil- [135, 175], and allergen- [118, 131]
induced mechanisms. Several cytokines that are increased in equine RAO are known to
up-regulate and activate EGFR through EGF and other ligands. Neutrophils can
upregulate EGFR expression through TNF—or and subsequently activate EGFR in a
ligand-independent fashion through the release of oxidative radicals [134, 135].
Interestingly, residual lung function deﬁcits and persistent neutrophilic inﬂammation in
RAO remission and are highly correlated with increased NF-KB activity and ICAM-1

expression of epithelial cells [164]. NF-KB is an important intracellular signaling element

that is involved in mucin gene up-regulation in vitro [96]. Sustained NF-KB activation in

152

equine RAO is driven by granulocytes and mediated by IL-1 [3 and TNF-or [165], both of
these cytokines increase MUCSAC mRNA levels in vitro [166]. The recently
characterized calcium-activated chloride channel (CACCI) is associated with increased
stored intraepithelial mucosubstance, a measure of mucus cell meta— and hyperplasia, and
increased MUCSAC expression in animal models of allergic airway disease and in
asthmatic subjects [131-133]. Our ﬁndings indicate that these pathways may be
implicated in the pathogenesis of airway mucus accumulation in equine RAO and in
milder forms of lower airway disease. Due to the large variations between individuals,
larger numbers of horses need to be investigated.

In conclusion, the present results did not support the hypothesis that eqMUC5AC
is up-regulated and intraepithelial mucosubstances are increased in RAO-affected
compared to clinically healthy control horses. However, we have identiﬁed equine
homologues EGFR and hCACCl, which important regulators of mucin gene expression
and mucus cell changes in the airways. Equine hCACCl-like and EGFR mRNA levels as
well as neutrophil percentages were associated with eqMUC5AC mRNA levels in
airways of RAO-affected and of clinically healthy horses. This suggests a role of these
key signaling molecules in combination with neutrophilic inﬂammation in mucin
regulation in airways of horses suffering from RAO as well as of horses with milder

degrees of airway inﬂammation.

153

Table 8-1: Real-time PCR primer sequences (5’ to 3’) and predicted product size in

 

 

 

basepairs
Name Forward Reverse - Size
MUCSAC CT GCCT CTT GCCCACCTT GTGGCACTGGTCAAACACAC 120bp
EGF CCCAGTCCTATGACGGGTACT CT CAAGTCCT GGTGC'I’ GACA 124bp

 

CACCl CCGCC'I'I‘ AAACGACT GACT C TCCCTATCAGTGCCACCTTT l44bp

 

EGFR GATGGACGTCAACCCAGATG TTCCT CCACCT CGTAGCT GT 133bp

 

1 SS rRNA GCAATTA'ITCCCCATGAACG GGCCTCACTAAACCATCCAA 123bp

 

 

 

 

 

 

154

Table 8—2: Airway secretion cytology, intraepithelial mucosubstance and gene
expression in RAO-affected and control horses: Number of observations, mean
values 1 SD for percentages (%) of macrophages, neutrophils, lymphocytes, mast
cells and eosinophils, volume density (Vs, nl / mm2 basal lamina) of intraepithelial

mucosubstance (IMS), and relative mRNA levels of eqMUC5AC, equine hCACCl-

like, EGFR and EGF.

 

RAO-affected horses (n = 5)

Control horses (11 = 5)

 

Number of Mean i SD Number of Mean i SD
Observations Observations

Macrophages 20 19.8 i 11.9* 20 44.4 :t 22.5
Neutrophils 20 65.7 :1: 21.3“ 20 28.7 i 28.7
Lymphocytes 20 12.1 a: 10.2 20 16.4 i 10.6
Mast Cells 20 0.9 :t 1.7 20 2.6 d: 3.9
Eosinophils 20 1.3 i 2.7 20 3.6 :t 7.0
VsofIMS l9 1.7i1.1 18 1.53:1.3
eqMUC5AC 17 46.9 :t 42.9 17 64.4 i 76.5
CACCl 17 115.8 :t 84.9 17 47.9 :1: 36.7
EGFR 17 17.6 i 16.8 17 10.1 i 8.2
EGF 17 12.4 :t 7.7 17 9.9 i 5.4

 

*signiﬁcantly different (P < 0.05) from control group.

155

Fig. 8-1: Volume density of intraepithelial mucosubstance (A) in lung lobes a, b, c
and d of 5 RAO-affected (I-V; black bars) and 5 clinically healthy control (i—v; white
bars) horses. *In three airways volume density could not be measured due to
excessive artefacts. Volume densities did not differ between groups or between lung
lobes. Within a horse some airways could appear to store mucosubstance in rounded
goblet cells (B; horse V, lobe c), while others showed goblet cells expressing mucus
(C; horse V, lobe a), or were almost completely devoid of mucosubstance (horse V,
lobe b).

 

N on 4:- or
I I I l
>

Intraepithelial Mucosubstance

(nl / mm2 Basal Lamina)

 

 

 

.. I...“

abcd abcd abcd ab'd abcd abcd abcd abcd ab'd ab'd
I II III IV V i ii iii iv v

 

 

 

RAO-affected horses clinically healthy control horses

 

 

 

156

Fig. 8-2: Distribution of relative mRNA levels of eqMUC5AC(A), CACC1(B), EGFR
(C) and EGF (D) in the 4 lung lobes (a, b, c and d) of 5 RAO-affected (black bars)
and 5 clinically healthy control (white bars) horses. All results are expressed as
relative x-fold levels of mRNA compared with the sample that was lowest for all
target genes, Ivc, which is set as l-fold. *Six RNA samples were not available for
measurements. ‘

 

 

 

 

 

 

 

 

 

 

   

 

 

 

 

0 300
g -
o A
3 250,
2
0'
,2 200. “
o
E
g 150-
2
‘2‘ Ioo~ .
(I
E
s 50’ ”l
1'6
3 0_ n_n_ , 1
abc* abc* abcd a*cd abcd abcd abcd *bcd ab” abcd
I II III IV V i ii iii iv v
RAO-affected horses clinically healthy control horses
350
300~ B

250 ~

200 .

150 -

100-~

50~

 

 

     

 

 

 

.mi “1|th

 

Relative mRNA levels of CACC1-like

 

 

O -
abc‘ abc‘ abcd a‘cd abcd abcd abcd *bcd ab” abcd
I II III IV V i ii iii iv
RAO-affected horses clinically healthy control horses

157

Fig. 8-2 (con’t)

Relative mRNA levels of EGF R

Relative mRNA levels of EGF

70

 

60‘

4o-

30-

 

 

 

 

Irintlll

 

 

abc‘ abc* abcd a‘cd abcd abcd abcd *bcd ab" abcd

 

 

 

I II III IV V i ii iii iv v
RAO-affected horses clinically healthy control horses

35

30 - D

25 -

1
20 -
' I

 

 

 

 

 

 

 

 

 

 

 

 

ill 1

 

 

abc* abc‘ abcd a‘cd abcd abcd abcd *bcd ab” abcd
I II III IV V i ii iii iv v

RAO-affected horses clinically healthy control horses

 

 

158

Figure 8-3: Partial correlation analysis (r =
significant correlations between eqMUC5AC mRNA levels and equine hCACCl-like
(A) and EGFR (B) mRNA levels as well as neutrophil percentages (C). Fig. D
illustrates the loglinear relationship between mean (a, b, c, d lobes) eqMUC5AC
mRNA levels and equine hCACCl-like.

Relative mRNA levels of eqMUCSAC

Relative mRNA levels of eqMUCSAC

100‘

300‘

100.
.00 ‘ .

 

 

 

 

A
o
o
o

. O
. . 0 o

8 o .0

r = 0.47, p= 0.002

0 100 200 300

Relative mRNA levels of CACC1-Iike

 

C r = 0.34, P = 0.032
O

 

:."~'-‘-

 

0 10 20 30 40 50

Neutrophil percentages (%)

159

Relative mRNA levels of eqMUC5AC

2

Log mRNA levels of eqMUCSAC

100 -

1.0

100.0 a

10.0 .

partial correlation coefficient):

 

 

    

o
"5 ﬁ 0
o =

.. r 0.34-1.1.9 0.023

10 20 30 40 50

 

Relative mRNA levels of EGFR

 

 

 

 

D
o
O
. o
O o .
O
o
1.0 10.0 100.0 1000.0

Log mRNA levels of CACC1-Iike

CHAPTER 9

SUMMARY, CONCLUSIONS AND OUTLOOK

I developed and tested three major hypotheses during my thesis project. In the
following, I will discuss the results in the context of previous ﬁndings in equine lower

airway disease and relevant comparative aspects of the literature.

Background

A literature review showed that increased airway mucus accumulation is a
clinically signiﬁcant problem in equine lower airway disease, both in recurrent airway
obstruction (RAO) and, particularly in racehorses, in the milder inﬂammatory airway
disease (IAD). However, there is no critically validated tool to quantify mucus
accumulation in equine RAO and IAD, and, furthermore, the causes of airway mucus
accumulation are largely speculative. There is evidence that mucus clearance is reduced
in RAO. Findings in human airway diseases indicate that unfavorable alterations of
mucus Viscoelasticity can cause mucus stasis in the airways. Neutrophils, the
predominant inﬂammatory cell type in RAO, can cause both unfavorable rheological
alterations by neutrophil breakdown products like DNA and f-actin. Several authors have
also suggested that mucus cell hyper- and metaplasia play a role in equine RAO, but this
claim has so far not been substantiated by morphometric measurements. Since
submucosal glands are sparse in horse airways, changes in epithelial goblet cells can be
expected to account for most increased production and secretion of mucins in equine

lower airway disease. Gel-forming mucins (MUC2, MUCSAC, MUCSB) have been

160

identiﬁed in several other species, and their expression and secretion is regulated within
complex immunological networks. Neutrophil products as well as Th1 and Th2 cytokines
that are involved in equine RAO can inﬂuence the expression of mucins, particularly of
MUCSAC, in airway diseases of other species. Recent evidence from experimental
models of airway disease further suggests that MUCSAC up-regulation converges on
EGFR and CACC] pathways in the airway epithelium.

Neither mucus accumulation nor mucus Viscoelasticity nor mucin production and
storage have been critically studied in equine RAO. The purpose of my work was
therefore to quantify airway mucus accumulation and investigate its causes in RAO-

affected horses and clinically healthy controls.

Hypotheses

1) Increased mucus accumulation is a function of disease status (RAO) and age of
the horse as well as the environment. Further, increased mucus accumulation is
associated with neutrophilic airway inﬂammation.

2) Increased mucus accumulation is a result of and associated with unfavorable
alterations of mucus physical properties, i.e., increased Viscoelasticity and decreased

predicted clearability and/or

3) increased production of mucus, i.e., increased mucin gene mRNA levels and

storage of mucosubstance in the airway epithelium.

161

Airway mucus accumulation

To develop the tool to test the ﬁrst hypothesis, I validated an endoscopic scoring
system. Endoscopic scoring has been previously used to assess changes in both mucus
quantity and quality in horse airways. The validity of such scoring systems, however, has
not been investigated. In order to determine if endoscopic scoring could be used to
investigate airway mucus in horses, I investigated observer and horse variance of
endoscopic scores and their relationship with measured variables.

Mucus accumulation scoring showed excellent interobserver agreement and
moderate horse-related variance, was related to measured volumes of “artiﬁcial mucus”,
and correlated well with neutrophilic airway inﬂammation. Scores of mucus viscosity,
color and localization, on the other hand, showed high observer-related variance, and did
not correlate with measured mucus Viscoelasticity. I concluded that apparent viscosity,
localization and color scores should be interpreted with caution. In contrast, endoscopic
scoring of mucus accumulation is a reliable clinical and research tool.

I then used endoscopic scoring of mucus accumulation to test the hypothesis that
airway mucus accumulation is a function of disease status and/or environment. I
compared mucus accumulation scores in RAO-affected and clinically healthy control
horses before (when they were at pasture) and after (6, 24 and 48 hours) stabling on straw
bedding and hay feeding.

I found that large amounts of mucus were speciﬁc for RAO-affected horses in this
study. Variation among controls was considerable, however, and intermediate grades
were non-speciﬁc, showing complete overlap between the two groups. During

exacerbation, mucus accumulations increased 2-fold in RAO-affected horses, but

162

exposure to stable environment had no signiﬁcant effect on mucus accumulations in
controls, even though controls also showed a moderate increase of broncho-alveolar
lavage ﬂuid neutrophils. In conclusion, mucus accumulations, before and especially after
exposure to dust and allergens, are increased in RAO-affected horses compared to
controls. Healthy controls show considerable variability in mucus accumulation, but no
increase of mucus accumulation after exposure to hay dust.

To investigate if airway mucus accumulation is also a function of age, I compared
mucus accumulation scores in two age groups of clinically healthy well-performing sport
horses that lived in the same controlled and constant stall environment. Several
alternative outcomes were possible: no difference between age-groups; increased mucus
in the older horses, as suggested by research in human smokers that produce more
sputum with increasing pack-years; or, increased mucus in younger horses, as observed in
racehorses entering training facilities.

I found that airway mucus accumulation did not differ between age groups.
Clinically healthy older horses, having been exposed 10 more years to conventional
stable environment without developing clinical signs of lower airway disease, do not
show increased subclinical airway inﬂammation or mucus accumulation. Overall,
however, large interindividual variations were observed in mucus quantity scores as well
as broncho-alveolar lavage inﬂammatory cells. As previously observed in our research
control horses, the majority of the asymptomatic well-performing sport horses were
suffering from subclinical IAD. I propose that mild to moderate degrees of airway
inﬂammation and mucus accumulation are a normal (i.e., observed in the majority of

individuals) reaction to conventional stable environment, and that the signiﬁcance of such

163

ﬁndings should be interpreted according to the level of respiratory performance expected
from an individual.

The combination of these studies and reports on mucociliary clearance rates from
other groups allowed for an estimate of the mucus volumes that are transported along the
trachea in healthy and RAO-affected horses. I estimate that 2-8 ml per hour of
endoscopically visible mucus are transported along the trachea in clinically healthy
horses. In RAO-affected animals the volumes can be extrapolated to 6-14 ml per hour at
pasture and 15-30 ml per hour when stabled. These estimates only account for mucus
volumes cleared by mucociliary action, but not by coughing. A mean of 5 ml per hour
corresponds to 120 ml/day for normal horses, while a mean of 22 thr corresponds to
540 ml/day for the stabled RAO-affected animals. Given the size of a horse, the 120
ml/day normal value would be consistent with an estimate of about 10-15 ml/day in adult
healthy humans, while the 540ml/day is proportionally less than the estimated 200-300
ml/day in humans with exacerbation of chronic bronchitis.

In contrast to the results in clinically healthy horses alone, mucus accumulation
scores correlated well with tracheo-bronchial secretion neutrophil percentages and
broncho-alveolar lavage ﬂuid neutrophil numbers in horses that evenly represented the
severity spectrum of non-infectious equine lower airway disease. Neutrophilic
inﬂammation may cause mucus accumulation in horse airways through increased
production and secretion of mucins and/or by unfavourably altering physical properties

and clearability of the secretions.

164

Role of mucus rheology

I used the microrheometric method to test the second hypothesis that increased
mucus accumulation is a result of and associated with unfavorable alterations of mucus
physical properties, i.e., increased Viscoelasticity and decreased predicted clearability. I
compared mucus Viscoelasticity and predicted clearability in RAO-affected and clinically
healthy control horses before and after stabling on straw bedding and hay feeding.

The central ﬁndings were rheological changes in airway mucus, which occurred
over time in RAO-affected animals, but not in controls. Mucus rheology was similar in
both groups at baseline, but Viscoelasticity increased about 3-fold on a linear scale in
RAO—affected horses within 24 hours after environmental challenge, and was
accompanied by signiﬁcant decreases in predicted mucociliary and cough clearability.
Rheological values did not correlate with the hydration of mucus or broncho-alveolar
lavage ﬂuid cytology. I concluded that viscoelastic properties of tracheal mucus samples
from RAO-horses in remission do not differ from those of norrnal horses. However,
environmental challenge causes acute exacerbation of airway obstruction and a
concurrent increase in mucus Viscoelasticity only in RAO-horses. Therefore, I inferred
that unfavorable changes in mucus rheology may contribute to stasis and accumulation of
mucus in RAO-horses in exacerbation, but not in clinical remission.

Although rheological values did not signiﬁcantly correlate with broncho-alveolar
lavage ﬂuid cytology, the increase of mucus Viscoelasticity coincided with the dramatic
inﬂux of neutrophils in the airways of RAO-affected horses. Neutrophils in the airways
degenerate and release breakdown products, speciﬁcally DNA and ﬁlamentous (F)-actin

that can unfavorably alter mucus rheological properties.

165

In order to test the hypothesis that neutrophil breakdown products cause
unfavorable rheological changes, i.e., the observed increased mucus Viscoelasticity
during RAO exacerbation, I investigated the effects of DNase, which degrades DNA, and
gelsolin, which degrades F-actin. I compared mucolytic effects on airway mucus from
RAO-affected horses in exacerbation with effects on control mucus samples in a
preliminary mucolytic study.

Gelsolin had a signiﬁcant mucolytic effect on mucus of RAO horses in
exacerbation, effectively reducing Viscoelasticity almost 3-fold to normal levels, and
indicating that F-actin plays a role in unfavorable rheological changes observed in mucus
of RAO-affected horses during exacerbation. In contrast, DNase, at concentrations
effective on cystic ﬁbrosis sputum, did not alter Viscoelasticity in RAO or control mucus.
Thus, DNA ﬁbers either do not play a similarly important rheological role in RAO mucus
as in CF mucus, or, alternatively, they may be protected to a larger degree from
mucolytic action. F-actin ﬁbers have been shown to form an entangled network with
DNA ﬁbers, not only causing increases in viscoelasticity, but also protecting DNA ﬁbers
from mucolytic action by thNase. Further studies are needed to elucidate the role of

DNA and F-actin ﬁber interactions in RAO mucus.

Role of mucin production

Unfavorable rheological properties leading to decreased mucus clearance can only
partly explain the increased mucus accumulation in RAO, however. Particularly during
remission, but likely also in exacerbation, increased production of mucus must also play a

role. Based on ﬁndings in other species, any or all of the secreted, gel-forming mucins,

166

MUC2, MUCSAC and/or MUCSB, could be expressed in horse airways and may be
involved in increased mucin production in RAO.

In order to test the hypothesis that increased production of mucus is a cause of the
observed increased mucus accrunulation, I identiﬁed equine homologues of gel-forming
mucins and investigated the regional distribution of mRNA-levels of equine mucin gene
homologues at different airway generations.

I identiﬁed equine (eq) homologues of MUC2 and MUCSAC, and found robust
expression of eqMUC5AC in RAO and control horse airways, but any expression of
MUC2 was undetectable and unlikely to be of physiological consequence. A semi-
quantitative reverse transcriptase-polymerase chain reaction (RT-PCR) assay on pooled
samples indicated increased levels of eqMUC5AC at large and small airway generations
of RAO-affected horses. I consequently hypothesized that eqMUC5AC up-regulation
may be a primary mechanism responsible for mucus hypersecretion and accumulation in
RAO. MUCSAC is the main “signature” mucin in experimental and natural models of
airway disease. Many different, Th-l and -2 type as well as innate immunity, pathways
converge on up-regulation of MUCSAC with subsequent mucous cell metaplasia and
hypersecretion. Epidermal growth factor (EGF), its receptor, EGFR and homologues of
the recently identiﬁed human calcium-activated chloride channel 1 (hCACCl) are key
signaling molecules involved in these pathways, but have not been investigated in equine
lung disease.

I therefore identiﬁed equine homologues of hCACCl and EGFR and developed
quantitative RT-PCR assays for eqMUC5AC, equine hCACCl-like, EGFR and EGF. This

allowed me to accurately determine mRNA levels in individual samples and investigate the

167

 

association of eqMUC5AC expression with the expression of these key signaling elements,
with stored intraepithelial mucosubstance and with neutrophilic inﬂammation. I studied
samples from four lung lobes each in ﬁve RAO-affected and ﬁve clinically healthy
control horses after the animals had been exposed to dusty stable environment for ﬁve
days.

My ﬁndings did not support the hypothesis that eqMUC5AC is up—regulated and
intraepithelial mucosubstances are increased in RAO-affected compared to clinically
healthy control horses. Comparison with voltnne density of intraepithelial mucosubstance
obtained by the same morphometric method in proximal and distal axial airways of rats,
however, showed that means in horses of both disease groups were more than 10-fold
higher than in untreated rats, and are in the range of values found in rats exposed to ozone
and endotoxin. The relative, quantitative assessment of eqMUC5AC, equine hCACCl-
like, EGFR or EGF mRNA levels showed remarkable consistency across different lung
lobes within individuals, indicating that the underlying pathogenesis is a diffuse process
involving the entire organ. Conversely, differences between horses were considerable.
These differences reﬂect the variability in the degrees of mucus accumulation observed in
RAO-affected and particularly also in clinically healthy horses. While the present results
Cannot explain the increased mucus in RAO airways, they may indicate that in some

horses—bothRAO-affected and clinically healthy horses with milder degrees of airway

disease—eqMUC5AC mRNA up-regulation may contribute to mucus hypersecretion.
Furthermore, equine hCACCl-like and EGFR mRNA levels as well as neutrophil
percentages were associated with eqMUC5AC mRNA levels in both groups. I therefore

Propose that these key signaling molecules play an important role in mucin regulation in

168

airways of horses suffering ﬁ‘om RAO as well as of horses with milder degrees of airway

inﬂammation.

Interpretation and outlook

Unfavourable rheological changes associated with F-actin can at least partly
explain increased mucus accumulation during RAO exacerbation, but not in remission.
An important unanswered question is whether therapeutic improvement of these
unfavourable rheological changes is of any clinical beneﬁt in RAO. It would be
interesting to further study possible interactions and rheological effects of F-actin and
DNA ﬁbers. From a clinical perspective, however, peptide mucolytics such as gelsolin
are at present cost prohibitive in the horse, and more practical solutions are needed. Since
I have also shown good mucolytic in vitro efﬁcacy of nacystelyn, it would be rational to
test the clinical efﬁcacy of an inexpensive equivalent of nacystelyn, such as N-acetyl-
cysteine.

I could not support my hypothesis that increased mucin production, evidenced by
eqMUC5AC expression, is a cause of the increased airway mucus in RAO. My results
also cast doubt on the often cited subjective observations of mucus cell meta- and
hyperplasia in RAO, since measurements of intraepithelial mucosubstance showed no
difference between RAO-affected and control horses. It is important to note the high
variability of mucus accumulation scores, neutrophils counts, stored intraepithelial
mucosubstance and eqMUC5AC in RAO-affected horses as well as in control horses.
Furthermore, compared with untreated laboratory rats, clinically healthy horses have

highly increased mucus accumulation. Such cross-species comparisons can be

169

treacherous, however, and it would be a very worthwhile project to investigate the lungs
and airways of horses that have never been exposed to indoor stable environments, for
instance Mustangs.

Based on our present data, it seems that bronchospasm and not mucus
accumulation and neutrophilic inﬂammation is the main deﬁning pathophysiological
characteristic of RAO. I propose that IAD, manifest with increased airway mucus and
neutrophils, stored mucosubstances and mucin gene upregulation is very common in
clinically healthy research horses. My results indicate that the clinical signiﬁcance of
these ﬁndings needs to be further determined with speciﬁc emphasis on the use of the
animal, i.e., sport horse vs. racehorse.

Other areas that should be further explored are airway mucin biology and the
causes of persistent mucus accumulation in RAO remission. In the former area, the
equine homologue of MUCSB needs to be identiﬁed and mRNA levels of all gel-forming
mucins must be assessed in combination with the identiﬁcation and, if possible,
quantiﬁcation of their respective proteins products in airway secretions. The latter area is
currently being addressed at the Pulmonary Laboratory. In addition to the investigation of
the role of B-cell lymphoma 2 (Bel-2) protein in the pathogenesis of persistent mucus
accumulation in RAO remission, it would be very interesting to explore the link to the

reported persistently increased levels of NF-KB and TNF-or. I plan to determine TNF-a,

IFN-y and, if possible, il-l3 mRNA levels in the tissues from horses in acute exacerbation
(chapter 8). Pending these results, the expression of this Th1 vs. Th2 cytokine balance
may be particularly interesting in the context of resolution vs. persistence of mucus cell

hyper- and metaplasia during RAO remission.

170

Two inherent limitations of the horse as a model for allergic lower airway disease
have marked my thesis research: large interindividual variability due to underlying
(subclinical) airway disease and difﬁculties of mechanistic interventions, i.e., lack of
speciﬁc molecular blocking agents and in vitro systems. I use this as a legitimate excuse
for the near absence of hard “mechanistic” facts, the prevalence of descriptive ﬁndings
and the consequently mostly deductive reasoning that characterize my thesis research.
None the wiser, I will continue to engage the challenges of “horse research”.

I am intrigued by my observation that eqMUC5AC levels were associated with
equine hCACCl-like, EGFR and neutrophils. Since the stable environment is similar for
most of these horses, I propose that genetic differences in key signaling elements such as
CACCl and EGFR underlie the observed variability in clinical signs, airway mucus
accumulation and mucin gene expression.

I plan to pursue this aspect. In the short term I will further investigate the genetic
identity and single nucleotide polymorphisms of the equine hCACCl-like fragment,
EGFR and later possibly eqMUC5AC and selected cytokines. I expect that any further
investigations will have to contend with the noted interindividual variability in the horse.
Two approaches are possible. Disease groups can be more rigorously deﬁned so as to
maximize differences between RAO-affected and control horses. Conversely, as a
clinician, I plan to embrace the messy reality and investigate larger numbers of horses of
varying degrees of non-infectious lower airway disease. My conceptual model is that of a
continuous clinical spectrum due to overlapping, genetically determined gene expression
proﬁles and resulting responses to environmental allergens and irritants. My long-term

goal is to uncover some of the genetic background of the immunological and pathogenic

171

 

mechanisms that lead to non-infectious equine lower airway disease. At the University of
Beme I will have access to a large number of horses of varying degrees of non-infectious
lower airway disease to investigate the associations between clinical signs,
pathophysiological abnormalities, gene expression data and genetic differences on a
sound statistical basis. I also plan to explore the possibilities of using speciﬁc blockers,
for instance for CACCl, EGFR and Th-2 type cytokines, in order to test more

mechanistic hypotheses in the horse.

172

10.

11.

12.

13.

BIBLIOGRAPHY

Gerber, H., Chronic pulmonary disease in the horse. Equine Veterinary Journal, 1973. 5:
p. 26-33.

Dixon, P.M., Respiratory mucociliary clearance in the horse in health and disease, and
its pharmaceutical modification. Vet Rec, 1992. 131(11): p. 229-35.

Robinson, N.E., et al., The pathogenesis of chronic obstructive pulmonary disease of
horses. Br Vet J, 1996. 152(3): p. 283-306.

Matthews, L.M., et al., Studies on pulmonary secretions. Am. Rev. Respir. Dis., 1963.
88: p. 199-204.

Dixon, P.M., D.I. Railton, and BC. McGorum, Equine pulmonary disease: a case control
study of 300 referred cases. Part 2.‘ details of animals and historical and clinical
ﬁndings. Equine Veterinary Journal, 1995. 27: p. 422-427.

Woods, P.S., et al., Airborne dust and aeroallergen concentration in a horse stable under
two diﬁrerent management systems. Equine Veterinary Journal, 1993. 25(3): p. 208-13.

Pirie, R.S., Role of inhaled endotoxin in the pathogenesis of equine COPD, in
Proceedings of Workshop on Allergic Diseases of the Horse, J .L. Watson, Editor. 1998:
Lipica, Slovenia. p. 10.

McGorum, B.C., J. Ellison, and RT. Cullen, Total and respirable airborne dust

endotoxin concentrations in three equine management systems. Equine Veterinary
Journal, 1998. 30(5): p. 430-434.

Bracher, V., et al., An investigation of the incidence of chronic obstructive pulmonary
disease (COPD) in random populations of Swiss horses. Equine Veterinary Journal,
1991. 23: p. 136-141.

Sasse, H.H.L., S. Boerrna, and FAA. Smolders, The relationship between pulmonary
function tests and other parameters. Results of a research project into the aetiology of
COPD in horses, in Lung Function and Respiratory Diseases in the Horse, RE. Beadle,
Editor. 1985, Hippiatrika: Hannover, Germany. p. 47-48.

Anon, International Workshop on Equine Chronic Airway Disease. Proceedings of
Workshop at Michigan State University, June 2000. Eq. Vet. J ., 2001. 33: p. 5-19.

Anon, International Workshop on Inﬂammatory Airway Disease: Deﬁning the Syndrome.
Proceedings of Workshop in Boston, October 2002. Eq. Vet. Educ, 2002. 15: p. 61-63.

McPherson, EA, et al., Chronic obstructive pulmonary disease (COPD): Identiﬁcation
of aﬂected horses. Equine Veterinary Journal, 1978. 10: p. 47-53.

173

14.

15.

l6.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

Fairbairn, S.M., et al., Early neutrophil but not eosinophil or platelet recruitment to the
lungs of allergic horses following antigen exposure. Clinical and Experimental Allergy,
1993. 23: p. 821-828.

Derksen, F .J ., et al., Bronchoalveolar lavage in ponies with recurrent airway obstruction
(heaves). American Review of Respiratory Disease, 1985. 132: p. 1066-1070.

Robinson, N.E., et al., Determinants of maximal changes in pleural pressure in horses
with heaves. Proceedings of the 13th Veterinary Respiratory Symposium, 1994: p. Al 1.

Robinson, N.E., et al., The airway response of horses with recurrent airway obstruction
(heaves) to aerosol administration of ipratropium bromide. Equine Veterinary Journal,

. 1993. 25: p. 299-303.

Dixon, P.M., D.I. Railton, and BC. McGorum, Equine pulmonary disease: a case control
study of 300 referred cases. Part 1: Examination techniques, diagnostic criteria and
diagnoses. Equine Veterinary Journal, 1995. 27: p. 416-421.

Fairbairn, S.M., et al., Duration of antigen-induced hyperresponsiveness in horses with
allergic respiratory disease and possible links with early airway obstruction. Jom'nal of
Veterinary Pharmacology and Therapeutics, 1993. 16: p. 469-476.

Robinson, N.E., Chairperson's introduction: International Workshop on Equine Chronic
Airway Disease, Michigan State University, 16—18 June 2000. Equine Veterinary Journal,
2001.33: p. 5-19.

Schatzmann, U., H. Buergi, and R. Straub, Der Einﬂuﬂ von Bisolvon auf das
T racheobronchialsekret des chronisch lungenkranken Pferdes [The inﬂuence of Bisolvon

on tracheobronchial secretions of horses with chronic lung disease]. Deutsche
Tieréirztliche Wochenschrift, 1973. 80: p. 97-100.

Gerber, V., et al., Airway mucus and inﬂammation-short-term response to environmental
challenge in clinically healthy control and in RAO-affected horses. Journal of Veterinary
Internal Medicine, 2003: p. submitted.

Gerber, H., Nebenwirkungen von kortikosteroid-behandlungen beim grosstier. Schweizer
Archiv ﬁir Tierheilkunde, 1970. 112: p. 51-57.

Robinson, N.E., et al., Eﬁ’icacy of three corticosteroids for the treatment of heaves.
Equine Veterinary Journal, 2002. 34(1): p. 17-22.

Murphy, J.R., E.A. McPherson, and RM. Dixon, Chronic obstructive pulmonary disease
(COPD): eﬂects of bronchodilator drugs on normal and aﬂected horses. Equine
Veterinary Journal, 1980. 12(1): p. 10-4.

Soma, L.R., J. Beech, and NH. Gerber, Eﬂects of cromolyn in horses with chronic
obstructive pulmonary disease. Veterinary Research Communications, 1987. 11(4): p.
339-351.

Jackson, CA, et al., Environment and prednisone interactions in the treatment of
recurrent airway obstruction (heaves). Equine Veterinary Journal, 2000. 32: p. 432-438.

174

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

Pearson, E.G., and T.W. Riebold, Comparison of bronchodilators in alleviating clinical
signs in horses with chronic obstructive pulmonary disease. Journal of the American
Veterinary Medical Association, 1989. 194(9): p. 1287-91.

International Workshop on Inﬂammatory Airway Disease: Deﬁning the Syndrome.
Proceedings of Workshop in Boston, October 2002. . Eq. Vet. Edu., 2003. in press.

Hoffman, A.M., M.R. Mazan, and S. Ellenberg, Association between bronchoalveolar
lavage cytologic features and airway reactivity in horses with a history of exercise
intolerance. Am J Vet Res, 1998. 59(2): p. 176-81.

Rose, R.J., et al., Indices of exercise capacity in horses presented for poor performance.
Eq. Vet. J., 1995. 18: p. 418-421.

Burrell, M.H., Endoscopic and virological observations on respiratory disease in a group
of young Thoroughbred horses in training. Equine Veterinary Journal, 1985. 17(2): p. 99-
103.

MacNamara, B., S. Bauer, and J. Iafe, Endoscopic evaluation of exercise-induced
pulmonary hemorrhage and chronic obstructive pulmonary disease in association with
poor performance in racing Standardbreds. J Am Vet Med Assoc, 1990. 196(3): p. 443-
5.

Sweeney, C.R., K.A. Humber, and K.A. Roby, Cytologic ﬁndings of tracheobronchial
aspirates ﬁ'om 66 thoroughbred racehorses. Am J Vet Res, 1992. 53(7): p. 1172-5.

Martin, B.B., Jr., J. Beech, and E.J. Parente, Cytologic examination of specimens
obtained by means of tracheal washes performed before and after high-speed treadmill
exercise in horses with a history of poor performance. Jotn’nal of the American
Veterinary Medical Association, 1999. 214(5): p. 673-677.

Morris, EA. and HI. Seeherman, Clinical evaluation of poor performance in the
racehorse: the results of 2 75 evaluations. Equine Vet J, 1991. 23(3): p. 169-74.

Martin, B.B., Jr., et al., Causes of poor performance of horses during training, racing, or
showing: 348 cases (1992-1996). J Am Vet Med Assoc, 2000. 216(4): p. 554-8.

Moreno, RH, J.C. Hogg, and RD. Pare, Mechanics of airway narrowing. American
Review of Respiratory Disease, 1986. 133: p. 1171-1180.

Kim, W.D., Lung mucus: a clinician's view. European Respiratory Journal, 1997. 10: p.
1914-1917.

Sherman, C.B., et al., Longitudinal lung function decline in subjects with respiratory
symptoms. Am Rev Respir Dis, 1992. 146(4): p. 855-9.

Annesi, I. and F. Kauffmann, Is respiratory mucus hypersecretion really an innocent
disorder? A 22-year mortality survey of 1,061 working men. Am Rev Respir Dis, 1986.
134(4): p. 688-93.

175

 

42.

43.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

Broadstone, R.V., et al., Effects of atropine in ponies with recurrent airway obstruction.
Journal of Applied Physiology, 1988. 65: p. 2720-2725.

Derksen, F.J., N.E. Robinson, and CE. Bemey, Aerosol pirbuterol: bronchodilator
activity and side eﬂects in ponies with recurrent airway obstruction (heaves). Equine
Veterinary Journal, 1992. 24: p. 107-112.

Rush, B.R., et al., Pulmonary distribution of aerosolized technetium T c 99m pentetate
after administration of a single dose of aerosolized albuterol sulfate in horses with
recurrent airway obstruction. American Journal of Veterinary Research, 1999. 60: p.
764-769.

Schatzmann, U., R. Straub, and H. Gerber, [Bronchial secretion aspiration in the horse].
Schweiz Arch Tierheilkd, 1972. 114(8): p. 395-403.

Dieclmrann, M.P., Zur Wirksamkeit von Ambroxolhydrochlorid (Mukovent) bei
lungenkranken Iferden - klinische, F unktionelle und zytologische Untersuchungen, in
Tieraerztliche Hochschule. 1987: Hannover, Germany.

Holcombe, S.J., et al., Stabling is associated with airway inﬂammation in young Arabian
horses. Equine Vet J, 2001. 33(3): p. 244-9.

Luft, J ., Der Einﬂuss koerperlicher Belastung auf die Zytologie des
Tracheobronchialsekretes beim Pferd [the effect of exercise of the composition of
tracheobronchial secretions in the horse]. Dissertation, 1987. Tieraerztliche
Hochschule Hannover, Germany.

Dixon, P.M., D.I. Railton, and BC. McGorum, Equine pulmonary disease: a case control
study of 300 referred cases. Part 3.‘ Ancillary diagnostic ﬁndings. Equine Veterinary
Journal, 1995. 27: p. 428-435.

Christley, R.M., et al., Coughing in thoroughbred racehorses: risk factors and tracheal
endoscopic and cytological findings. Vet Rec, 2001. 148(4): p. 99-104.

Robinson, N.E., et al., Cough, mucus, airway obstruction, and inﬂammation in heaves-
aﬁ'ected and control horses. American Journal of Veterinary Research, 2003. 64(5): p.
550-557.

Raidal, S.L., D.N. Love, and GD. Bailey, Inflammation and increased numbers of
bacteria in the lower respiratory tract of horses within 6 to 12 hours of confinement with
the head elevated. Aust Vet J, 1995. 72(2): p. 45-50.

Chapman, P.S., et al., Retrospective study of the relationships between age, inﬂammation
and the isolation of bacteria from the lower respiratory tract of thoroughbred horses. Vet
Rec, 2000. 146(4): p. 91-5.

Jefcoat, A.M., et al., Persistent mucin glycoprotein alterations in equine recurrent airway
obstruction. Am J Physiol Lung Cell Mol Physiol, 2001. 281(3): p. L704-l2.

Sweeney, OR, I. Beech, and K.A. Roby, Bacterial isolates from tracheobronchial
aspirates of healthy horses. Am J Vet Res, 1985. 46(12): p. 2562-5.

176

56.

57.

58.

59.

60.

61.

62.

63.

65.

66.

67.

68.

69.

70.

Houtmeyers, E., et al., Regulation of mucociliary clearance in health and disease. Eur
Respir J, 1999. 13(5): p. 1177-88.

Maxson, A.D., et al., Eﬂects of furosemide, exercise, and atropine on tracheal mucus
transport rate in horses. Am J Vet Res, 1995. 56(7): p. 908-12.

Turgut, K. and H.H. Sasse, Inﬂuence of clenbuterol on mucociliary transport in healthy
horses and horses with chronic obstructive pulmonary disease. Vet Rec, 1989. 125(21):
p. 526-30.

Sweeney, C .R., et al., Effect of water vapor-saturated air therapy on bronchoalveolar
lavage and tracheal mucus transport rate in clinically normal horses. Am J Vet Res,
1989. 50(2): p. 276-9.

Raidal, S.L., The incidence and consequences of failure of passive transfer of immunity
on a thoroughbred breeding farm. Aust Vet J, 1996. 73(6): p. 201-6.

Gerber, V., et al., Inﬂuence of gravity on tracheal mucus velocity in excised horse
tracheas. J Respir Crit Care Med, 1996. 153(4): p. A483.

Schlesinger, R.B., Clearance from the respiratory tract. Fund. and Appl. Tox., 1985. 5:
p. 435-50.

Del Donno, M., et al., The eﬂect of inﬂammation on mucociliary clearance in asthma: an
overview. Chest, 2000. 118(4): p. 1142-9.

Coombs, S.L. and RM. Webbon, Observations on tracheal mucociliary clearance in
horses. Tierarztl Prax Suppl, 1987. 2: p. 5-9.

Willoughby, R.A., et al., Use of scintigraphy for the determination of mucociliary
clearance rates in normal, sedated, diseased and exercised horses. Canadian Journal of
Veterinary Research, 1991. 55: p. 315-320.

Willoughby, R., et al., The effects of equine rhinovirus, inﬂuenza virus and herpesvirus
infection on tracheal clearance rate in horses. Can J Vet Res, 1992. 56(2): p. 115-21.

Sweeney, C.R., Tracheal mucus transport rate in healthy horses. Am J Vet Res, 1989.
50(12): p. 2135-7.

Camner, P., K. Philipson, and L. Friberg, Tracheobronchial clearance in twins. Arch.
Environ. Health, 1971. 23: p. 82-7.

King, M. and BK. Rubin, Rheology of airway mucus: relationship with clearance
function, in Airway Secretion: Physiological Bases for the Control of Mucous
Hypersecretion, S. Shimura, Editor. 1994, Marcel Dekker, Inc.: New York. p. 283-314.

Kaup, F.J., W. Drommer, and E. Deegen, Ultrastructuralfindings in horses with chronic

obstructive pulmonary disease (COPD). I: Alterations of the larger conducting airways.
Equine Vet J, 1990. 22(5): p. 343-8.

177

71.

72.

73.

74.

75.

76.

77.

78.

79.

80.

81.

82.

83.

84.

Galati, P., et al., Atypical cilia in the tracheal epithelium of healthy horses. J Comp
Pathol, 1991. 105(2): p. 185-90.

Gerber, V., et al., Microscopic investigation of mucus qualities, mucus transport velocity
and ciliary beat frequency on horse tracheal epithelium. Respiration Physiology, 1997.
107: p. 67-74. '

Jeanneret-Grosjean, A., et al., Sampling technique and rheology of human
tracheobronchial mucus. Am Rev Respir Dis, 1988. 137(3): p. 707-10.

Girod, S., et al., Phospholipid composition and surface-active properties of
tracheobronchial secretions from patients with cystic ﬁbrosis and chronic obstructive
pulmonary diseases. Pediatr Pulmonol, 1992. 13(1): p. 22-7.

Gerber, V., Untersuchungen iiber die mukoziliiire Clearance in der Luftrohre des Pferdes
[Investigation of mucociliary clearance in equine trachea]. Dissertation, 1995.
University of Berne, Switzerland.

Im Hof, V., et al., Exogenous surfactant accelerates tracheal mucus velocity in
spontanously breathing non-anaesthetized horses. J Respir Crit Care Med, 1996. 153(4):
p. A109.

Im Hof, V., et al., In vivo determination of surface tension in the horse trachea and in
vitro model studies. Respir Physiol, 1997. 109(1): p. 81-93.

Dorwald, M., G. Vanden Bosche, and A. Gerull, Zur
Surfactantphospholipidzusammensetzung im T racheobronchialsekret des Pferdes und
ihrer klinischen relevanz fuer die Beurteilung des Lungenstatus bei chronisch
lungenkranken Pferden. Wien Tieraerztl. Mschr., 1991.78: p. 118-26.

Hobo, S., et al., Surfactant proteins in bronchoalveolar lavage ﬂuid of horses: assay
technique and changes following road transport. Vet Rec, 2001. 148(3): p. 74-80.

Boat, T.F., R.-W. Cheng, and M. Warren Leigh, Biochemistry of mucus, in Airway
secretion: Physiological bases for the control of mucous hypersecretion, S. Shimura,
Editor. 1995, Marcel Dekker, Inc.: New York. p. 217-82.

McGorum, B.C., et al., Evaluation of urea and albumen as endogenous markers of
dilution of equine bronchoalveolar lavage ﬂuid. Res Vet Sci, 1993. 55(1): p. 52-6.

Marcy, T.W., eta1., Limitations of using urea to quantify epithelial lining ﬂuid recovered
by bronchoalveolar lavage. Am Rev Respir Dis, 1987. 135(6): p. 1276-80.

Persson, C.G., Airway Mucosal Exudation of Plasma, in Airway secretion: Physiological
bases for the control of mucous hypersecretion, S. Shimura, Editor. 1995, Marcel Dekker,
Inc.: New York. p. 451-67.

Winder, NC. and R. von Fellenberg, Immunoﬂuorescent evaluation of the lower

respiratory tract of healthy horses and of horses with chronic bronchiolitis. American
Journal of Veterinary Research, 1986. 47: p. 1271-1274.

178

 

85.

86.

87.

88.

89.

90.

91.

92.

93.

94.

95.

96.

97.

98.

99.

Rose, M.C., Mucins.‘ structure, function, and role in pulmonary diseases. American
Journal of Physiology: Lung Cellular and Molecular Physiology, 1992. 263: p. L413-
L429.

Gendler, SJ. and AP. Spicer, Epithelial mucin genes. Annual Review of Physiology,
1995. 57: p. 607-634.

Jeffery, P.K. and D. Li, Airway mucosa: secretory cells, mucus, and mucin genes.
European Respiratory Journal, 1997. 10: p. 1655-1662.

Perez-Vilar, J., J.K. Sheehan, and SH. Randell, Making More M UCS. Am J Respir Cell
Mol Biol, 2003. 28(3): p. 267-70.

Lundgren, JD. and J.H. Shelhamer, Pathogenesis of airway mucus hypersecretion. J
Allergy Clin Irnrnunol, 1990. 85(2): p. 399-417.

Thornton, D.J., et al., Identiﬁcation of two glycoforms of the MUCSB mucin in human
respiratory mucus. Evidence for a cysteine-rich sequence repeated within the molecule. J
Biol Chem, 1997. 272(14): p. 9561-6.

Kirkham, S., et al., Heterogeneity of airways mucus: variations in the amounts and
glycoforms of the major oligomeric mucins M UC5AC and MUC5B. Biochem J, 2002.
361(Pt 3): p. 537-46.

Thornton, D.J., et al., Respiratory mucins: identiﬁcation of core proteins and glycoforms.
Biochem J, 1996. 316 ( Pt 3): p. 967-75.

Leikauf, G.D., et al., Mucin apoprotein expression in COPD. Chest, 2002. 121(5 Suppl):
p. 1668-182S.

Chen, Y., Y.H. Zhao, and R. Wu, In silico cloning of mouse Muc5b gene and
upregulation of its expression in mouse asthma model. Am J Respir Crit Care Med, 2001.
164(6): p. 1059-66.

Borchers, M.T., M.P. Carty, and GD. Leikauf, Regulation of human airway mucins by
acrolein and inﬂammatory mediators. Am J Physiol, 1999. 276(4 Pt 1): p. L549-55.

Basbaum, C., et al., Control of mucin transcription by diverse injury-induced signaling
pathways. Am J Respir Crit Care Med, 1999. 160(5 Pt 2): p. S44-8.

Harkema, J .R. and LG. Wagner, Non—allergic models of mucous cell metaplasia and
mucus hypersecretion in rat nasal and pulmonary airways. Novartis Found Symp, 2002.
248: p. 181-97; discussion 197—200, 277-82.

Jany, B., [Molecular cloning of human respiratory tract mucin and studies of mucin gene
expression in tracheobronchial epithelium of the rat]. Pneumologie, 1993. 47(8): p. 479-
87.

Li, D., et al., Localization and up-regulation of mucin (M UC2) gene expression in human
nasal biopsies of patients with cystic fibrosis. J Pathol, 1997. 181(3): p. 305-10.

179

 

100.

101.

102.

103.

104.

105.

106.

107.

108.

109.

110.

111.

112.

113.

114.

115.

116.

Bullimore, S.R., et al., Surface mucus in the non-glandular region of the equine stomach.
Res Vet Sci, 2001. 70(2): p. 149-55.

Kemmish, A.R., et al. Biochemical comparison of respiratory mucins from normal horses
and horses with chronic inﬂammatory small airway disease. in British Equine Veterinary

Association. 1998. Birmingham.

Walley, E.A., et al. Characterization of equine respiratory tract mucins. in World Equine
Airway Symposium. 2001. Edinburgh.

Scharfrnan, A., et al., Recognition of Lewis x derivatives present on mucins by ﬂagellar
components of Pseudomonas aeruginosa. Infect Immun, 2001. 69(9): p. 5243-8.

Trivier, D., et al., The binding of surface proteins from Staphylococcus aureus to human
bronchial mucins. Eur Respir J, 1997. 10(4): p. 804-10.

Majima, Y., et al., Effect of biochemical components on rheologic properties of nasal
mucus in chronic sinusitis. Am J Respir Crit Care Med, 1999. 160(2): p. 421-6.

Kurashima, K., et al., Ejfect of thromboxane A2 synthetase inhibitor, OKY-046, on
sputum in chronic bronchitis and dijfuse panbronchiolitis. Eur Respir J, 1995. 8(10): p.
1705-11.

Groneberg, D.A., et al., Distribution of respiratory mucin proteins in human nasal
mucosa. Laryngoscope, 2003. 113(3): p. 520-4.

Widdicombe, J.H. and LS. Pecson, Distribution and numbers of mucous glands in the
horse trachea. Equine Vet J, 2002. 34(6): p. 630-3.

Reid, L.M., Pathology of chronic bronchitis. Lancet, 1954. l: p. 275-276.
Dunnill, M.D., M. GR, and J .A. Anderson, A comparison of the quantitative anatomy of
the bronchi in normal subjects, in status asthmaticus, in chronic bronchitis, and in

emphysema. Thorax, 1969. 24: p. 176-179.

Aikawa, T., et al., Marked goblet cell hyperplasia with mucus accumulation in the
airways of patients who died of severe acute asthma attack. Chest, 1992. 101: p. 916-921.

Hotchkiss, J.A., Guelph, Canada. Airway mucous synthesis, storage, and secretion. in
World Equine Airways Symposium. 1998. Guelph, Canada.

Haschek, W.M., Respiratory system, in T oxicologic Pathology, C.G. Rousseaux, Editor.
1998, Academic Press: San Diego. p. 108-22.

Stricter, R.M., J.A. Belperio, and MP. Keane, Cytokines in innate host defense in the
lung. J Clin Invest, 2002. 109(6): p. 699-705.

Wills-Karp, M., IL-lZ/IL-I3 axis in allergic asthma. J Allergy Clin Immunol, 2001.
107(1): p. 9-18.

Romagnani, S., The 7721/7712 paradigm. Immunol Today, 1997. 18(6): p. 263-6.

180

117.

118.

119.

120.

121.

122.

123.

124.

125.

126.

127.

128.

129.

130.

131.

132.

Holgate, ST, The immunopharmacologz of mild asthma. J Allergy Clin Immunol, 1996.
98(5 Pt 2): p. S7-16; discussion $33-40.

Longphre, M., et al., Allergen-induced IL-9 directly stimulates mucin transcription in
respiratory epithelial cells. J Clin Invest, 1999. 104(10): p. 1375-82.

Soussi-Gounni, A., M. Kontolemos, and Q. Hamid, Role of IL-9 in the pathophysiology
of allergic diseases. J Allergy Clin Immunol, 2001. 107(4): p. 575-82.

Grunig, G., et al., Roles of interleukin-I3 and interferon-gamma in lung inﬂammation.
Chest, 2002. 121(3 Suppl): p. 88S.

Morokata, T., J. Ishikawa, and T. Yamada, Antigen dose deﬁnes T helper I and T helper
2 responses in the lungs of C5 7BL/6 and BALB/c mice independently of splenic
responses. Immunol Lett, 2000. 72(2): p. 119-26.

Dearman, R.J. and I. Kimber, Cytokine proﬁling and chemical allergy. Toxicol Appl
Pharmacol, 2002. 185(3): p. 228-9; author reply 229-30.

Dearman, R.J. and I. Kimber, Cytolcine ﬁngerprinting: characterization of chemical
allergens. Methods, 1999. 19(1): p. 56-63.

Tulic, M.K., P.G. Holt, and RD. Sly, Modiﬁcation of acute and late-phase allergic
responses to ovalbumin with lipopolysaccharide. Int Arch Allergy Immunol, 2002.
129(2): p. 119-28.

Anderson, G.P., The immunobiology of early asthma. Med J Aust, 2002. 177 Suppl: p.
S47-9.

Jeffery, P.K., Lymphocytes, chronic bronchitis and chronic obstructive pulmonary
disease. Novartis Found Symp, 2001. 234: p. 149-61; discussion 161-8.

Wills-Karp, M. and M. Chiaramonte, Interleukin-I3 in asthma. Curr Opin Pulm Med,
2003. 9(1): p. 21-7.

Riffo-Vasquez, Y. and D. Spina, Role of cytokines and chemokines in bronchial
hyperresponsiveness and airway inﬂammation. Pharmacol Ther, 2002. 94(3): p. 185-211.

McMillan, S.J., et al., The absence of interleukin 9 does not aﬂect the development of

allergen-induced pulmonary inﬂammation nor airway hyperreactivity. J Exp Med, 2002.
195(1): p. 51-7.

Grunig, G., et al., Requirement for IL-13 independently of IL-4 in experimental asthma.
Science, 1998. 282(5397): p. 2261-3.

Nakanishi, A., et al., Role of gob-5 in mucus overproduction and airway
hyperresponsiveness in asthma. Proc Natl Acad Sci U S A, 2001. 98(9): p. 5175-80.

Toda, M., et al., A calcium-activated chloride channel (HCLCA I) is strongly related to

IL-9 expression and mucus production in bronchial epithelium of patients with asthma. J
Allergy Clin Immunol, 2002. 109(2): p. 246-50.

181

133.

134.

135.

136.

137.

138.

139.

140.

I41.

142.

143.

144.

145.

146.

Zhou, Y., et al., Characterization of a calcium-activated chloride channel as a shared
target of Th2 cytokine pathways and its potential involvement in asthma. Am J Respir
Cell Mol Biol, 2001. 25(4): p. 486-91.

Shim, J.J., et al., IL-13 induces mucin production by stimulating epidermal growth factor
receptors and by activating neutrophils. Am J Physiol Lung Cell Mol Physiol, 2001.
280(1): p. L134-40.

Takeyama, K., et al., Oxidative stress causes mucin synthesis via transactivation of
epidermal growth factor receptor: role of neutrophils. Journal of Immunology, 2000.
164(3): p. 1546-52.

Fahy, J.V., Goblet cell and mucin gene abnormalities in asthma. Chest, 2002. 122(6
Suppl): p. 3208-3268.

Becker, S., M.J. Fenton, and J.M. Soukup, Involvement of microbial components and tall-
like receptors 2 and 4 in cytokine responses to air pollution particles. Am J Respir Cell
Mol Biol, 2002. 27(5): p. 611-8.

Wang, X., et al., Toll-like receptor 4 mediates innate immune responses to Haemophilus
inﬂuenzae infection in mouse lung. J Immunol, 2002. 168(2): p. 810-5.

Shi, Z.O., et al., IFN-gamma, but not Fas, mediates reduction of allergen-induced
mucous cell metaplasia by inducing apoptosis. J Immunol, 2002. 168(9): p. 4764-71.

Tesfaigzi, Y., et al., Bel-2 in LPS— and allergen-induced hyperplastic mucous cells in
airway epithelia of Brown Norway rats. American Journal of Physiology: Lung Cellular
and Molecular Physiology, 2000. 279(6): p. L1210-7.

Pack, R.J., L.H. Al-Ugaily, and G. Morris, The cells of the tracheobronchial epithelium
of the mouse: a quantitative light and electron microscope study. J Anat, 1981. 132(Pt 1):
p. 71-84.

Mercer, R., et al., Cell number and distribution in human and rat airways. Am. J. Respir.
Cell Mol. Biol., 1994. 10(6): p. 613-624.

Zhu, Z., et al., Pulmonary expression of interleukin-13 causes inﬂammation, mucus
hypersecretion, subepithelial ﬁbrosis, physiologic abnormalities, and eotaxin production.
J. Clin. Invest., 1999. 103(6): p. 779-788.

Temann, U.A., et al., A novel role for murine IL-4 in vivo: induction of MUC5AC gene
expression and mucin hypersecretion. Am J Respir Cell Mol Biol, 1997. 16(4): p. 471-8.

Kim, C.H., et al., IL-13 suppresses MUC5AC gene expression and mucin secretion in
nasal epithelial cells. Acta Otolaryngol, 2002. 122(6): p. 638-43.

Seong, J.K., et al., Upregulation of MUC8 and downregulation of MUC5AC by

inﬂammatory mediators in human nasal polyps and cultured nasal epithelium. Acta
Otolaryngol, 2002. 122(4): p. 401-7.

182

147.

148.

149.
150.

151.

152.

153.

154.
155.
156.
157.
158.
159.

160.

Jayawickreme, S.P., et al., Regulation of 15-lipoxygenase expression and mucus secretion
by IL-4 in human bronchial epithelial cells. Am J Physiol, 1999. 276(4 Pt 1): p. L596-
603.

Sonea, I.M., RM. Bowker, and NE. Robinson, Distribution of substance P binding sites
in equine airways. Equine Vet J, 1999. 31(3): p. 238-42.

Olszewski, M.A., et al., Mediators of anaphylaxis but not activated neutrophils augment
cholinergic responses of equine small airways. American Journal of Physiology, 1999.
276: p. L522-29.

Takeyama, K., et al., Neutrophil-dependent goblet cell degranulation.‘ role of membrane-
bound elastase and adhesion molecules. American Jom‘nal of Physiology: Lung Cellular
and Molecular Physiology, 1998. 275(2 Pt 1): p. L294-302.

Takeyama, K., et al., Epidermal growth factor system regulates mucin production in
airways. Proceedings of the National Academy of Sciences, USA, 1999. 96(6): p. 3081-6.

Art, T., et al., Indices ofoxidative stress in blood and pulmonary epithelium lining ﬂuid
in horses suﬂering from recurrent airway obstruction. Equine Veterinary Journal, 1999.
31: p. 397-401.

Lindberg, A., et al., Evaluation of leukotriene biosynthetic capacity in lung tissues from
horses with recurrent airway obstruction. Am J Vet Res, 2002. 63(6): p. 794-8.

Halliwell, R.B., et al., Local and systemic antibody production in horses aﬂected with
chronic obstructive pulmonary disease. Vet Immunol Irnmunopathol, 1993. 38(3-4): p.
201-15.

Ainsworth, D.M., et al., IgG antibody responses to an inhaled antigen in horses with
”heaves" (recurrent airway obstruction). Vet Immunol Irnmunopathol, 2002. 84(3-4): p.
169-80.

Pirie, R.S., P.M. Dixon, and BC. McGorum, Evaluation of nebulised hay dust
suspensions (HDS) for the diagnosis and investigation of heaves. 3.‘ Effect of
fractionation of HDS. Equine Vet J, 2002. 34(4): p. 343-7.

Pirie, R.S., et al., Evaluation of nebulised hay dust suspensions (HDS) for the diagnosis
and investigation of heaves. 2: Ejfects of inhaled HDS on control and heaves horses.
Equine Vet J, 2002. 34(4): p. 337-42.

Lavoie, JP, et al., Neutrophilic Airway Inﬂammation in Horses with Heaves Is
Characterized by a Th2-type Cytokine Proﬁle. American Journal of Respiratory Critical
Care Medicine, 2001. 164(8): p. 1410-3.

Beadle, R.B., D.W. Horohov, and SD. Gaunt, Interleukin-4 and interferon-gamma gene
expression in summer pasture-associated obstructive pulmonary disease affected horses.

Equine Vet J, 2002. 34(4): p. 389-94.

Bowles, K.S., et al., A novel model for equine recurrent airway obstruction. Vet Immunol
Irnmunopathol, 2002. 87(3-4): p. 385-9.

183

161.

162.

163.

164.

165.

166.

167.

168.

169.

170.

171.

172.

173.

Giguere, S., et al., Cytokine induction in pulmonary airways of horses with heaves and
effect of therapy with inhaled ﬂuticasone propionate. Vet Immunol Irnmunopathol, 2002.
85(3-4): p. 147-58.

Fanucchi, M.V., J.A. Hotchkiss, and J .R. Harkema, Endotoxin potentiates ozone-induced
mucous cell metaplasia in rat nasal epithelium. Toxicol Appl Pharmacol, 1998. 152(1):
p. 1-9.

Gordon, T., et al., One-month exposure to inhaled endotoxin produces a dose-dependent
increase in stored mucosubstances in rat intrapulmonary airways. Exp Lung Res, 1996.
22(5): p. 509-23.

Bureau, F., et al., Correlation between nuclear factor-kappaB activity in bronchial
brushing samples and lung dysfunction in an animal model of asthma. American Journal
of Respiratory Critical Care Medicine, 2000. 161(4 Pt 1): p. 1314-21.

Bureau, F., et al., An autocrine mechanism maintains persistent NF-B activation in the
bronchi of an animal model of asthma. American Journal of Respiratory Critical Care
Medicine, 2000. 161: p. A330.

Koo, J.S., et al., Overexpression of mucin genes induced by interleukin-1 beta, tumor
necrosis factor-alpha, lipopolysaccharide, and neutrophil elastase is inhibited by a
retinoic acid receptor alpha antagonist. Exp Lung Res, 2002. 28(4): p. 315-32.

Herholz, C., et al., Relationship between clinical signs and pulmonary function estimated
by the single breath diagram for CO(2) (SBD-CO(2)) in horses with chronic obstructive
pulmonary disease. Vet J, 2002. 163(2): p. 187-95.

King, M., Magnetic microrheometer, in Methods in Bronchial Mucology, L. Allegra,
Editor. 1988, Raven Press, Ltd.: New York. p. 73-83.

Gerber, V., et al., T racheobronchial mucus viscoelasticity during environmental
challenge in horses with recurrent airway obstruction. Equine Vet J, 2000. 32(5): p. 411-
7.

Fischer, B. and J. Voynow, Neutrophil elastase induces MUC5AC messenger RNA
expression by an oxidant-dependent mechanism. Chest, 2000. 117(5 Suppl 1): p. 3178-
20S.

Armstrong, J. and J. White, Liquefaction of viscous purulent exudates by
deoxyribonuclease. Lancet, 1950. 2: p. 739-742.

Pietra, M., et al., In vitro mucolytic activity of recombinant human deoxyribonuclease on
equine tracheobronchial mucus. Vet Rec, 2000. 147(22): p. 627-9.

Keller, H., et al., Klinische Studie zur Wirksamkeit und Vertrc'iglichkeit von N-
Acetylcystein bei der Behandlung der COB/COPD des Pferdes [eﬂicacy and tolerance
study of acetylcysteine in the COB/COPD therapy in horses]. Prakt. Tierarzt, 2001. 82: p.
108-117.

184

174.

175.

176.

177.

178.

179.

180.

181.

182.

183.

184.

185.

186.

187.

188.

Tremblay, G.M., et al., Eﬂect of stabling on bronchoalveolar cells obtained from normal
and COPD horses. Equine Vet J, 1993. 25(3): p. 194-7.

Voynow, J .A., et al., Neutrophil elastase increases MUC5AC mRNA and protein
expression in respiratory epithelial cells. American Journal of Physiology, 1999. 276(5
Pt 1): p. L835-43.

Wagner, J .G., et al., Endotoxin enhancement of ozone-induced mucous cell metaplasia is
neutrophil-dependent in rat nasal epithelium. Toxicol Sci, 2001. 60(2): p. 338-47.

Rubin, B.K., The pharmacologic approach to airway clearance: mucoactive agents.
Respir Care, 2002. 47(7): p. 818-22.

Robinson, N.E., et al., Relationship between clinical signs and lung function in horses
with recurrent airway obstruction (heaves) during a bronchodilator trial. Equine
Veterinary Journal, 2000.32: p. 393-400.

Gerber, V., et al., Endoscopic scoring of mucus: Reproducibility, and relation to mucus
viscosity and airway inﬂammation. EVJ, 2003. in preparation.

Tremblay, G.M., et al., Ejfect of stabling on bronchoalveolar cells obtained from normal
and COPD horses. Equine Veterinary Journal, 1993. 25(3): p. 194-197.

Bracher, V., et al., An investigation of the incidence of chronic obstructive pulmonary
disease (COPD) in random populations of Swiss horses. Equine Vet J, 1991. 23(2): p.
136-41.

Gerber, V., et al., Airway mucus in RAO - short-term response to environmental
challenge. Journal of Veterinary Internal Medicine, 2003: p. accepted.

Zayas, J .G., G.C. Man, and M. King, Tracheal mucus rheology in patients undergoing
diagnostic bronchoscopy. Interrelations with smoking and cancer. Am Rev Respir Dis,
1990. 141(5 Pt 1): p. 1107-13.

Viel, L. and J. Hewson, BAL cytology in horses with exercise intolerance: what does it
tell us? Second World Equine Airways Symposium (WEAS) proceedings, Edinburgh,
Scotland, 2001.

Sutton, G.A., et al., Pathogenesis and clinical signs of equine herpesvirus-1 in
experimentally infected ponies in vivo. Can J Vet Res, 1998. 62(1): p. 49-55.

Derksen, E.J., et al., Comparison of transtracheal aspirate and bronchoalveolar lavage
cytology in 50 horses with chronic lung disease. Equine Veterinary Journal, 1989. 21: p.
23-26.

Hoffman, A. Airway reactivity in the assessment of lung function. in Second World
Equine Airways Symposium (WEAS). 2001. Edinburgh, Scottland: WEAS.

Gerber, V., et al., Mucus quality on horse tracheal epithelium: microscopic grading
based on transparency. Respir Physiol, 1997. 107(1): p. 67-74.

185

 

189.

190.

191.

192.

193.

194.

195.

196.

197.

198.

199.

200.

201.

202.

203.

204.

App, E.M., J.G. Zayas, and M. King, Rheology of mucus and transepithelial potential
dijference.‘ small airways versus trachea. Eur Respir J, 1993. 6(1): p. 67-75.

Shih, C.K., et al., Ejfect of nondialyzable solids concentration and viscoelasticity on
ciliary transport of tracheal mucus. American Review of Respiratory Disease, 1977. 115:
p. 989-995.

Picot, R., 1. Das, and L. Reid, Pus, deoxyribonucleic acid, and sputum viscosity. Thorax,
1978. 33(2): p. 235-42.

Vasconcellos, C., et al., Reduction in viscosity of cystic ﬁbrosis sputum in vitro by
gelsolin. Science, 1994. 263: p. 969-971.

Tomkiewicz, RB, et al., Species differences in the physical and transport properties of
airway secretions. Can J Physiol Pharmacol, 1995. 73(2): p. 165-71.

Dasgupta, B., N.E. Brown, and M. King, Effects of sputum oscillations and thNase in
vitro: (1 combined approach to treat cystic ﬁbrosis lung disease. Pediatr Pulmonol, 1998.
26(4): p. 250-5.

Gerber, V., et al. Rheology of mucus in horses with moderate COPD. in World Equine
Airways Symposium. 1998. Guelph, Canada: WEAS.

King, M., Relationship between mucus viscoelasticity and ciliary transport in guaran
gel/frog palate model system. Biorheology, 1980. 17(3): p. 249-54.

King, M., Role of mucus viscoelasticity in clearance by cough. Eur J Respir Dis Suppl,
1987. 153: p. 165-72.

King, M., M. Agarwal, and J .B. Shukla, A planar model for mucociliary transport: effect
of mucus Viscoelasticity. Biorheology, 1993. 30(1): p. 49-61.

Dasgupta, B., et al., Rheological properties in cystic ﬁbrosis airway secretions with
combined thNase and gelsolin treatment, in Advances in Physiological Fluid
Dynamics, V.P. Saxena, Editor. 1996, Narosa: New Delhi, India. p. 74-78.

Dasgupta, B. and M. King, Reduction in Viscoelasticity in cystic ﬁbrosis sputum in vitro
using combined treatment with nacystelyn and thNase. Pediatr Pulmonol, 1996. 22(3):
p. 161-6.

Sheils, C.A., et al., Actin ﬁlaments mediate DNA ﬁber formation in chronic inﬂammatory
airway disease. Am J Pathol, 1996. 148(3): p. 919-27.

Zahm, J.M., et al., Improved activity of an actin-resistant DNase I variant on the cystic
ﬁbrosis airway secretions. Am J Respir Crit Care Med, 2001. 163(5): p. 1153-7.

Buisine, M.P., et al., Developmental mucin gene expression in the human respiratory
tract. Am J Respir Cell Mol Biol, 1999. 20(2): p. 209-18.

Davies, J.R., et al., Identiﬁcation of MUC5B, MUC5AC and small amounts of MUC2
mucins in cystic ﬁbrosis airway secretions. Biochem J, 1999. 344 Pt 2: p. 321-30.

186

205.

206.

207.

208.

209.

210.

211.

212.

213.

214.

215.

216.

217.

218.

219.

Groneberg, D.A., et al., Expression of respiratory mucins in fatal status asthmaticus and
mild asthma. Histopathology, 2002. 40(4): p. 367-73.

Thornton, D.J., et al., Characterization of mucins from cultured normal human
tracheobronchial epithelial cells. Am J Physiol Lung Cell Mol Physiol, 2000. 278(6): p.
L1118-28.

Voynow, J .A., What does mucin have to do with lung disease? Paediatr Respir Rev,
2002. 3(2): p. 98-103.

Chomczynski, P. and N. Sacchi, Single-step method of RNA isolation by acid
guanidinium thiocyanate-phenol—chloroform extraction. Anal Biochem, 1987. 162(1): p.
156-9.

Guzman, K., T. Bader, and P. Nettesheim, Regulation of MUC5 and MUCI gene
expression: correlation with airway mucous diﬂerentiation. Am J Physiol, 1996. 270(5 Pt
1): p. L846-53.

Steiger, D., et al., Use of mucin antibodies and cDNA probes to quantify hypersecretion
in vivo in human airways. Am J Respir Cell Mol Biol, 1994. 10(5): p. 538-45.

Perez-Vilar, J. and R.L. Hill, Identification of the half-cystine residues in porcine
submaxillary mucin critical for multimerization through the D-domains. Roles of the
CGLCG motif in the D1 - and D3-domains. J Biol Chem, 1998. 273(51): p. 34527-34.

Van Klinken, B.J., et al., Mucin gene structure and expression: protection vs. adhesion.
Am J Physiol, 1995. 269(5 Pt I): p. G613-27.

Louahed, J., et al., Interleukin-9 upregulates mucus expression in the airways. Am J
Respir Cell Mol Biol, 2000. 22(6): p. 649-56.

Davies, J.R., et al., Mucins in airway secretions from healthy and chronic bronchitic
subjects. Biochem J, 1996. 313 (Pt 2): p. 431-9.

Sheehan, J .K., et al., Physical characterization of the M UC5AC mucin: a highly
oligomeric glycoprotein whether isolated from cell culture or in vivo from respiratory
mucous secretions. Biochem J, 2000. 347 Pt 1: p. 37-44.

Zuhdi Alimam, M., et al., Muc-5/5ac mucin messenger RNA and protein expression is a
marker of goblet cell metaplasia in murine airways. Am J Respir Cell Mol Biol, 2000.
22(3): p. 253-60.

Kaup, F .J ., et al., Ultrastructural ﬁndings in horses with chronic obstructive pulmonary
disease (COPD). II: Pathomorphological changes of the terminal airways and the
alveolar region. Equine Vet J, 1990. 22(5): p. 349-55.

Adler, K.B., et al., Platelet activating factor stimulates secretion of mucin by explants of
rodent airways in organ culture. Experimental Lung Research, 1987. 13(1): p. 25-43.

Dabbagh, K., et al., IL-4 induces mucin gene expression and goblet cell metaplasia in
vitro and in vivo. J Immunol, 1999. 162(10): p. 6233-7.

187

 

220.

221.

222.

223.

224.

225.

226.

227.

228.

229.

230.

231.

232.

Harkema, J .R. and J .A. Hotchkiss, Ozone- and endotoxin-induced mucous cell
metaplasias in rat airway epithelium: novel animal models to study toxicant-induced
epithelial transformation in airways. Toxicol Lett, 1993. 68(1-2): p. 251-63.

Hotchkiss, J .A., M.A. Stam, and J .R. Harkema, Platelet activating factor stimulates rapid

mucin secretion in rat nasal airways in vivo. Experimental Lung Research, 1993. 19(5):
p. 545-57.

Rose, M.C., et al., Model systems for investigating mucin gene expression in airway
diseases. J Aerosol Med, 2000. 13(3): p. 245-61.

Vandenput, S., et al., Airborne dust and aeroallergen concentrations in diﬂerent sources
of feed and bedding for horses. Veterinary Quarterly, 1997. 19(4): p. 154-158.

Gerber, V., et al., Mucin genes in horse airways: M UC5AC, but not MUC2, may play a
role in recurrent airway obstruction. Equine Vet J, 2003. 35(3): p. 252-7.

Harkema, J .R., et al., Eﬂects of an ambient level of ozone on primate nasal epithelial
mucosubstances. Quantitative histochemistry. American Journal of Pathology, 1987.
127(1): p. 90-6.

Tatusova, TA. and TL. Madden, BLAST 2 Sequences, a new tool for comparing protein
and nucleotide sequences. FEMS Microbiol Lett, 1999. 174(2): p. 247-50.

Collins, D.W. and TH. Jukes, Rates of transition and transversion in coding sequences
since the human-rodent divergence. Genomics, 1994. 20(3): p. 386-96.

Fitch, W.M., The molecular evolution of cytochrome c in eukaryotes. J Mol Evol, 1976.
8(1): p. 13-40. '

Stewart, F., et al., Identiﬁcation of the horse epidermal growth factor (EGF) coding
sequence and its use in monitoring EGF gene expression in the endometrium of the
pregnant mare. J Mol Endocrinol, 1994. 12(3): p. 341-50.

Janke, A., et al., Phylogenetic analysis of I85 rRNA and the mitochondrial genomes of
the wombat, Vombatus ursinus, and the spiny anteater, T achyglossus aculeatus:
increased support for the Marsupionta hypothesis. J Mol Evol, 2002. 54(1): p. 71-80.

Gerber, V., et al., Airway inﬂammation and mucus in two age groups of asymptomatic
well-performing sport horses. Eq. Vet. J ., 2003. 35(5): p. 491-495.

Kirschvink, N., et al., The use of cardboard bedding material as part of an environmental
control regime for heaves-aﬂected horses: in vitro assessment of airborne dust and

aeroallergen concentration and in vivo effects on lung function. Vet J, 2002. 163(3): p.
319-25.

188

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

MWWWWW