(Chest. 2002;121:166S-182S.)
© 2002
American College of Chest Physicians
Mucin Apoprotein Expression in COPD*
George D. Leikauf, PhD;
Michael T. Borchers, PhD;
Daniel R. Prows, PhD and
Leigh G. Simpson, PhD
* From the Departments of Environmental Health, University of Cincinnati, Cincinnati, OH.
Correspondence to: George Leikauf, PhD, Director, Molecular Toxicology Division, Department of Environmental Health, University of Cincinnati, PO Box 670056, Cincinnati, OH 45267-0056; e-mail: leikaugd{at}uc.edu
 |
Abstract
|
|---|
Mucins, which are complex glycoproteins that provide the viscoelastic properties of mucus that are essential for the protection of the airways, are characterized by a variable-number tandem repeats (VNTR) region that may undergo alternate splicing during transcription. Such transcripts may yield multiple proteins via diverse post-translational modifications involving glycosylation (within each VNTR). Fifteen distinct mucin genes have been identified, with several mapping to chromosomal clusters (ie, 7q22 and 11p15.5), possibly having evolved by gene duplication. The deduced protein sequences can be subdivided into both membrane-associated mucins and secreted mucins. Membrane-associated mucins consist of cytoplasmic, transmembrane, and extracellular domains. The membrane-associated mucins MUC1, MUC4, and MUC11 have been localized to the lung. In addition to VNTRs, secreted mucins possess repeated cysteine-rich D-domains (which are important in polymerization). Secreted mucins that are localized to the lung include MUC2 (in cells with and without secretory granules), MUC5AC (in surface and submucosal mucous cells), MUC5B and MUC8 (in submucosal mucous cells), and MUC7 (in submucosal serous cells). Currently, little is known about the regulation of mucins in COPD patients. Recent studies with acrolein and cigarette smoke have suggested that MUC5AC is inducible (accompanied by epidermal growth factor [EGF] ligand formation and the activation of EGF receptor- dependent pathways), whereas MUC5B is constitutively expressed (increasing through gland enlargement). Similarly, little is known about the genetic determinants that control mucus hypersecretion, but preliminary findings in animal models suggest that intrastrain differences in acrolein-induced mucin formation are amenable to genetic analysis. As our understanding of the functional genomics of mucin biology increases, further clinical targets and therapeutic strategies are likely to emerge.
Key Words: airway • COPD • gene regulation • genetics • mucins • mucus hypersecretion • strain differences
Mucus hypersecretion is a distinguishing feature of COPD, which is defined physiologically by limitations of airflow and pathologically by repetitive injury and inappropriate epithelial repair in airways (ie, bronchitis) and lung parenchyma (ie, emphysema).1
2
3
Initially experienced by patients as a mild annoyance on exertion, airflow obstruction can eventually become life-threatening even during rest, with onset occurring often in the fifth decade of life. Underlying the progressive loss of lung function is a complex remodeling process that alters the alveolar matrix and the airway epithelium. This process can be initiated by acute environmental exposures that result in epithelial injury, chemokine release, and leukocyte recruitment. Continued exposure, in turn, leads to persistent leukocyte activation with the release of elastinolytic activity. Many environmental exposures are associated with COPD as follows: mining; chemical manufacturing; metal refining; farming; cooking; and others that involve exposure to irritant gases and organic, metal, and mineral dusts. The strongest risk factor for COPD is exposure to cigarette smoke.
Early in the last century, COPD may have been underdiagnosed (especially in the United States) because physicians did not view the presenting bronchitis as a life-threatening illness.3
4
However, this view changed rapidly in the 1950s when fatalities occurred selectively among persons with COPD following several major air pollution episodes (eg, Donora, PA, and London, UK).5
6
Recognition of the involvement of environmental factors improved our understanding of COPD pathogenesis and led to better clinical definitions.7
8
The age-adjusted death rate due to COPD has increased continuously over the last 50 years to become the fourth leading cause of death in the United States.9
Of the leading causes of mortality in the United States, only COPD continues to rise, with death rates having increased by 22% in the past decade. In the United States, the number of patients with COPD doubled in the last 25 years, with the current estimated cost of medical care exceeding $14.5 billion per year. The magnitude of this epidemic is evident, especially when the data are presented as a percentage of the disease-specific, age-adjusted death rate for 1950 (Fig 1
). With cigarette sales and air pollution increasing throughout the World, the World Health Organization predicts that death due to COPD worldwide will continue to rise (possibly surpassing cerebrovascular disease by 2020).10
View larger version (23K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1. Trends in the leading causes of death in the United States over the last 50 years. Top: while marked decreases in heart and cerebrovascular diseases have occurred, beginning in the 1970s, the rate of COPD has continued to rise in the United States and throughout the world. Bottom: this trend becomes more evident when normalized to levels experienced in 1950. Deaths due to COPD increased > 550% from 1950 to 1998.
|
|
|
Inflammation, Cigarette Smoking, and COPD
|
|---|
Several lines of evidence suggest that inflammation and leukocyte infiltration, when combined with repetitive irritant exposures, may contribute to hypersecretion and epithelial remodeling in COPD patients. This process is likely to involve proteolytic enzymes released from chronically stimulated leukocytes. For example, neutrophil elastase induces the following: (1) alveolar enlargement and mucous cell differentiation11
12
; (2) mucus secretion from airway gland cells in tissue culture13
; and (3) mucin transcripts in cell culture.14
15
In addition, mucus secretion and gel thickness can be stimulated by macrophage-derived, lymphocyte-derived, and epithelium-derived inflammatory mediators, including interleukin (IL)-4, IL-5, IL-9, IL-13, prostaglandin E2 (PGE2), 12-hydroxyeicosatetraenoic acid (HETE), 15-HETE, tumor necrosis factor (TNF)-
, and macrophage-derived mucus secretagogue-68.16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Nonetheless, the relative roles of specific leukocytes in COPD are unclear. Emphysema is likely a result of increased elastinolytic activity within the alveolus due to the accumulation and activation of leukocytes in the lower respiratory tract.41
42
43
44
45
Because neutrophils contain 10-fold more elastinolytic activity per cell than macrophages,46
they have been proposed to be major pathogenic cells in emphysema in chronic smokers.43
47
48
49
50
Other findings, however, have focused attention on macrophages, which accumulate in the centriacinar regions where emphysema begins in smokers, as a source of elastinolytic activity.11
51
52
53
54
In contrast, neutrophils are uniformly distributed in the lung.55
Morphologic assessments have associated the extent of smoking-induced alveolar destruction and the numbers of macrophages in lung parenchyma.56
57
58
Hautamaki et al53
examined macrophage metalloelastase (MME; matrix metalloproteinase [MMP] 12)-deficient mice that were exposed to cigarette smoke. In such animals, macrophage numbers and alveolar disruption were reduced compared to wild-type control mice.59
60
Smoke-exposed MME-deficient (MME-/-) mice receiving monthly intratracheal instillations of monocyte chemoattractant protein-1 showed accumulations of alveolar macrophages but did not develop airspace enlargement. At least in the mouse, MME is critical to alveolar proteolysis. In humans, although MMP12 is associated with emphysema, additional factors (ie, MMP7 or MMP9) may be important.59
Studies by Ofulue and colleagues54
61
examined the elastinolytic potential of rat macrophages that were obtained from the lung during the onset and progression of cigarette smoke-induced emphysema. The time course of cigarette smoke-induced neutrophil and macrophage accumulations in the lung interstitium and BAL fluid was compared to that of the onset and progression of emphysema lesions. Neutrophil numbers increased rapidly, becoming maximal within 1 month of exposure but falling to control levels during months 2 to 6. In contrast, macrophages increased more slowly (maximal numbers noted at 2 months) with exposure and continued through month 6. The neutrophil elastinolytic activity from smoke-exposed rats was similar to that from air-exposed rats at 1 to 6 months of exposure. In contrast, the macrophage elastinolytic activity from these rats increased after 2 months of exposure and remained at that level throughout the exposure period. Excessive lung elastin breakdown in vivo was observed in the smoke-exposed rats after 2 to 6 months. These data suggest that, in rats, the time course of increased macrophage-derived elastinolytic activity, not that of neutrophils, is more closely associated with the development of cigarette smoke-induced emphysema.
A further study61
in rats treated with antibodies directed specifically at macrophages, but not neutrophils, reported the selective suppression of macrophage accumulation and macrophage-related elastinolytic activity in the lungs. Cigarette smoke-induced lung elastin breakdown and emphysema were not prevented in antineutrophil antibody-treated, smoke-exposed rats. In contrast, these responses were inhibited following antimacrophage antibody treatment, again implicating macrophages rather than neutrophils in the pathogenesis.
|
Epithelial Mucins
|
|---|
The airway epithelial surface is covered by mucus that is secreted by goblet cells (which are present in the luminal epithelium of the large diameter airways) and by mucous and serous cells (which are present in the submucosal gland of cartilaginous, large-diameter airways).3
6
62
63
64
65
66
Mucus protects the underlying epithelium from dehydration, pathogens, and chemical and particulate irritants. Mucus consists mainly of water (95%) combined with salts, lipids, proteins, and glycoproteins.67
68
69
70
Mucus properties vary throughout the airways, and even on the surface of each cell (which may consist of a few microns), to enable the wide range of properties needed for efficient mucociliary clearance.71
In the distal airways, for example, mucus is enriched with lower molecular weight and lower viscosity proteins (including surfactant and Clara cell secretory proteins). The secretions within submucosal glands also vary. The distal tubule regions contain serous cells, which are thought to produce lower molecular weight and less viscous secretions, whereas the proximal tubule regions contain more mucous cells, which are thought to produce secretions of higher molecular weight and greater viscosity.66
72
73
74
75
76
A particular subset of glycoproteins, the mucins, provides the viscoelastic properties of mucous and serous secretions. Several mucins recently have been identified and sequenced, and their tissue-specific expression has been assessed.
Mucins are large molecules (ie, > 100 kd) that are composed of 50 to 85% carbohydrates.70
77
78
79
80
Each mucin protein contains an apomucin core (typically 80% of the coding sequence) that is enriched in hydroxyamino acids, serine, and threonine.81
82
83
84
These amino acids contain O-glycosylation sites for oligosaccharides. The length and number of O-glycosylation domains vary between mucins, and each domain contains repeated series of sequences (ie, variable-number tandem repeats [VNTRs]).85
86
87
88
89
This permits a diverse set of messenger RNAs (mRNAs) to be formed from a single gene through alternate splicing during transcription. The resulting polydisperse signal often ranges from 0.4 to 20 kd and produces a characteristic smear in Northern blot analyses.
The amount of protein generated from each mucin transcript also can be varied by message stabilization. Mucin mRNA stabilization has been noted following the treatment of cells with cytokines.14
30
Because mucin transcripts are very large, message stabilization could allow for more efficient and dynamic translation, presumably involving protein binding to stem-loop structures within the 3'-untranslated region of mRNA-destabilizing sequences that are present in coding regions of the mucin transcript, thereby protecting transcripts from nucleolytic cleavage.
Each mucin transcript is capable of producing multiple peptides via a wide diversity of post-translational modifications. Mucin oligosaccharides are joined to the protein core through an initial -O-glycosidic linkage of N-acetylgalactosamine (GalNAc) to the hydroxyl region of serine or threonine.90
This linkage provides a starting site for branching oligosaccharide chains containing fucose, galactose, N-acetylglucosamine, and N-acetylneuraminic acid. Peripheral to a core of 1 to 20 of these five monosaccharides are sulfate or neuramic acids that produce a polyanionic (ie, acidic) molecule. Small amounts of mannose also are found in mucins.
Mucin O-glycosylation is accomplished by one of six uridine diphosphate-GalNAc-polypeptide--N-GalNAc transferases.89
91
92
93
Although respiratory tract mucins are primarily O-glycosylated, N-glycosylation is also possible at sites containing Asn-x-Ser/Thr motifs. Because one apomucin sequence can lead to numerous lengths of transcript, each transcript can vary in stability, and each peptide can have various combinations of several hundred different carbohydrate chains, this complex assembly scheme allows for wide mucin heterogeneity between and within individuals. This mechanism provides for huge diversity in an extracellular protein that the cell needs to manage a continuously altering environment (eg, the binding to adaptive pulmonary microorganisms with varied cell coats).94
95
96
97
|
Membrane-Associated Mucins
|
|---|
Fifteen primary mucin genes have been identified (although several have been sequenced only partially). Based on common structural motifs, mucins have been divided into the following two groups: membrane-associated mucins; and secreted mucins. (Other proteins sharing mucin-like domains that are not discussed further include the following: orosomucoid-1 and orosomucoid-2; uromodulin; epidermal growth factor (EGF)-like module containing, mucin-like, hormone receptor-like sequence-1 or hormone receptor-like sequence-2; biphenyl hydrolase-like mRNA; proteoglycan 4; CXC chemokine ligand-16; and mucosal vascular addressin cell molecule-1.)
Membrane-associated mucin genes include MUC1, MUC3A, MUC3B, MUC4, MUC9, MUC11, and MUC12, and Muc14 in rodents (Table 1
). Genes encoding these mucins are typically arranged with an extracellular domain containing different VNTRs and a transmembrane domain, followed by a cytoplasmic domain.80
88
98
99
100
101
102
These molecules are thought to have roles in cell-cell communication including cell adhesion, cell recognition, and signaling, and may be involved in tumor cell invasion and metastasis.103
104
105
106
107
108
109
The best-characterized member, MUC1, is contained on chromosome 1.80
98
99
100
101
102
103
104
105
106
107
108
109
110
Transcription, translation, and post-translational modifications produce a bimodal distribution of allele-length variants (range, 2.5 to 11 kilobases [kb]). The deduced MUC1 protein sequence includes an extracellular VNTR domain that is 20 amino acids in length. The transcripts produced by this gene can vary greatly mainly through differences in the number of VNTRs, ranging from 25 to > 125 repeats per allele. Once glycosylated and inserted into the membrane, the resulting proteins can be very large and can extend > 200 nm into the pericellular space. This distance of a highly charged moiety may lead to antiadhesive properties or to the inhibition of immune surveillance among human tumors, where MUC1 expression is aberrant.
One MUC1 variant (arising from variable splicing) lacks a cytoplasmic tail and is soluble. This suggests that the extracellular domain of MUC1 is sensitive to proteolytic cleavage in a nonglycosylated region near the transmembrane domain, thereby releasing a portion of the molecule from the cell surface. The sequence of transmembrane domain and cytoplasmic tail are conserved in mammals. The latter can interact with cytoskeletal actin filaments and contains a serine that is responsive to phosphorylation. MUC1 is expressed in almost all human glandular epithelial cells and diffusely throughout the lung extending to the distal bronchiolar epithelial cells. Studies in other organ systems suggest that MUC1 functions in membrane polarity and possibly in establishing developmental stage-specific recognition that is important to epithelial differentiation.
Less is known about the other members of this group. MUC3A, MUC3B, MUC11, and MUC12 are contained in a gene cluster on chromosome 7q22.111
112
113
Transcripts for MUC4 have numerous allele variants, resembling the multimodal MUC1 distribution but ranging from 7.5 to 21.5 kb.114
115
116
A sequence analysis of MUC4, MUC3A, and MUC12 indicates that each contains conserved, cysteine-rich, EGF-like domains and large areas of similarity when sequences are aligned.107
117
118
119
Each molecule has two extracellular EGF-like domains that are separated by an N-glycosylated domain of similar size in all three genes. The first EGF-like domain in MUC12 shows homology to a number of EGF receptor (EGFR)-binding ligands (ie, transforming growth factor [TGF]-, amphiregulin, heparin-binding EGF, and EGF itself). In addition, the rat orthologue of MUC4 contains two EGF-like domains (different from MUC12) that bind another growth factor receptor (ie, c-erbB-2).120
This action of rMuc4 initiates signaling that promotes mitogenesis.118
Of the 7q22 mucin subfamily, only MUC4 and MUC11 transcripts have been found consistently in the lung, MUC4 being localized intensely in the bronchial and bronchiolar epithelium and submucosal glands.66
111
112
121
122
123
124
In embryogenesis, lung MUC4 transcripts are expressed first, followed by MUC1 and MUC2, before the cytodifferentiation of epithelial cells into ciliated or secretory phenotypes.125
126
Although it has been investigated, MUC3 has not been found in normal lungs but has recently been found in lung tumors.127
Located on chromosome 1, MUC9 sequence analysis predicts a 654-amino acid polypeptide that contains four N-glycosylation sites, two N-myristoylation sites, and several potential phosphorylation sites.128
MUC9 has approximately 50% sequence similarity to members of a glycosyl hydrolase subfamily. We are unaware of any attempts to localize MUC9 (also known as oviductin) to the respiratory tract. Recently identified, Muc14 (National Center for Biotechnology Information accession number, NM 016885 [Mus musculus mucin 14, endothelial; Muc14-pending], mRNA) has been found in mouse endothelial cells and may have a role in leukocyte-endothelial cell adhesion.
|
Secreted Mucins
|
|---|
Mucins found in secretions include MUC2, MUC5AC, MUC5B, MUC7, and MUC8. Common to all members of this subgroup is a central region with several repeats of a D-domain. Also found in pre-pro-von Willebrand factor, which is a protein that is essential to its platelet-blood vessel wall function, D-domains are rich in cysteines. Although details are limited regarding how secreted mucins are processed in the cell, cysteines contained in the D-domains could provide for polypeptide crosslinking and packaging within secretory granules.
The most studied member of this subgroup is MUC2, with > 5,100 amino acids in its most common allelic form.110
129
130
131
132
133
134
135
136
137
138
At least nine MUC2 allele variants exist, each with acidic properties and three D-domains.139
Through analogy with von Willebrand factor, MUC2 is thought to polymerize end-to-end through disulfide bridges to form large, secreted, polymeric, gel-forming mucins. von Willebrand factor dimers form in the endoplasmic reticulum, and trimers and oligomers form in the Golgi cells shortly after synthesis. The primary function of MUC2 is to provide a protective barrier between the epithelial surfaces and the lumen, and, again by homology to von Willebrand factor, MUC2 may associate with the basement collagen. Initially identified in the intestine, the incidence of MUC2 decreases in colon cancer, and defective polymerization occurs in ulcerative colitis. Similarly, the expression of MUC2 has been found in the airways and in airway cells in culture. (Low levels have been detected in normal cells with little or no expression in tumor lines.)
MUC2 transcripts are present both in mucin-expressing and mucin nonexpressing epithelial cells, whereas lysozyme expression is limited to serous cells.122
This suggests that the restriction of mucin to mucous cells is controlled at a translational level.140
141
142
143
The exposure of normal human airway epithelial cells to retinoic acid, or of NCI-HL292 cells to TNF-, IL-4, or IL-1ß, increases MUC2 expression, the latter being inhibited by actinomycin-D (indicating transcriptional regulation). Budesonide has been reported to attenuate IL-1ß-mediated MUC2 expression, and this is reversed by a glucocorticoid receptor antagonist, RU-486. In situ hybridization detects a diffuse, weak MUC2 signal throughout the normal respiratory epithelium, with increased staining in inflamed airways. For example, Pseudomonas aeruginosa activates nuclear factor-
B in the MUC2 promoter through an Src-dependent Ras-mitogen-activated protein kinase-pp90/rsk pathway.
Interestingly, MUC2 maps to another gene cluster on chromosome 11p15.5.144
145
146
147
148
149
150
151
152
153
Three distinct mucin genes, MUC5AC, MUC5B, and MUC6, also map to chromosome 11p15.5 (between HRAS and IGF2 as tel-HRAS-MUC6-MUC2-MUC5AC-MUC5B-IGF2-cen), spanning a region of > 400 kb. In addition, MUC5AC and MUC5B are expressed in the lung and are major components of human airway secretions.
MUC5AC, due to its large size, was initially described as two genes. Two large partial clones (then named MUC5A and MUC5C) were subsequently found to be parts of the same gene. MUC5AC colocalizes with glycoprotein-positive goblet cells and submucosal gland mucous cells. Currently, in the approximately 4 kb of genomic DNA 5' to the TATA transcription start site, CACCC box and putative binding sites for nuclear factor, Sp 1, GRE, AP-2, and PEA3 have been identified.138
141
154
155
156
MUC5AC transcription is mediated through the c-Src and MAP kinase kinase (ie, mitogen-activated protein kinase/extracellular signal-regulated kinase kinase 1/2) pathways much like MUC2, but, unlike MUC2, MUC5AC is transcriptionally activated by protein kinase C-mediated events. In addition, because IL-9 effects are typically mediated through a JAK-STAT signaling pathway and because this cytokine increases MUC5AC expression, this is another likely alternative activation pathway.34
In contrast to MUC5AC, MUC5B is limited mainly to mucous cells in submucosal glands.114
151
157
158
159
160
161
162
Although less is known about the regulation of MUC5B transcription, it has been reported that human airway cells that are deprived of retinoic acid have diminished transcription. This can be reversed by prolonged readdition of retinoic acid to the culture medium.
Two other secreted mucins are MUC7 and MUC8. Little is known about the regulation of MUC7, which is localized to serous cells of submucosal glands.162
163
164
165
The predicted 377-amino acid sequence of MUC7 is relatively small for apomucin proteins with an apparent molecular mass of 39 kd. In contrast, MUC8, like MUC5B, is localized to mucous cells in the submucosal glands and is found in airway secretions.166
167
MUC5AC appears most frequently in studies of human secretions and is regulated by many mediators that are likely to be present in the airways of COPD patients.
|
Airway Mucus Hypersecretion
|
|---|
One mechanism involved in the pathogenesis of mucus hypersecretion is an increase in the numbers of mucus-secreting cells on the airway surface epithelium and in submucosal glands. Typically limited to the major airways, mucus traps foreign particles and microorganisms for clearance through coupling with the ciliated epithelium. Following irritation, the size of the submucosal glands in the large collagen-containing airways increases, and mucus-producing cells are found more frequently in the surface epithelium of the distal small airways. This condition can be induced experimentally in animals by exposing them to respiratory irritants that produce epithelial cell injury followed by proliferation. Irritants that can initiate mucus hypersecretion include tobacco smoke,168
169
170
ozone,171
172
173
sulfur dioxide,173
174
and acrolein.175
176
In rats and mice, increased Muc2177
178
and Muc5ac mRNA170
176
179
have been associated with the expression of EGFR.155
169
170
180
This concurs with studies linking hypersecretion with growth and differentiation of nonsecretory cells to those that secrete mucus.156
173
Similarly, increased Muc5ac (but not Muc2) accompanies morphologic changes of rat tracheal epithelial cells in culture.181
Increased Muc5ac (but not Muc2) mRNA levels also precede mucous cell differentiation in rats in vivo following acrolein exposure.176
In human cell culture studies, increased MUC2 and MUC5AC mRNA levels also accompany cell differentiation, with increased mucin secretion coinciding with altered morphology in human airway epithelial cells.182
Of the 15 mucins cloned, MUC4, MUC5AC, and MUC5B are most often detected in human airway secretions.107
|
Restitution of Epithelial Injury
|
|---|
Following injury by chemical irritants, the initial phase of mucosal wound repair consists of the rapid migration of epithelial cells to the wound site followed by spreading to re-establish the integrity of the luminal surface. This process has been studied primarily in the intestinal tract, but it also occurs in the airways following antigen challenge of sensitized rodents183
or mechanical damage to the airways.184
Epithelial restitution is mediated in part by a family of small, apically secreted peptides, the distinctive disulfide-bonded domain structure of which creates a trefoil configuration that confers resistance to degradation by proteases and extremes of pH. An intimate association between trefoil family factor (TFF) domain and mucin proteins is suggested by their colocalization in mucin-secreting cells, physical interaction via mucin von Willebrand factor domains,185
and evolutionary history (TFF domains are present in several of the frog mucins).186
To date, three TFF members have been identified. In humans, TFF3 has been localized to the secretory granules of mucous cells in the acini of bronchial submucosal glands and to mucous cells in the gland ducts,187
while TFF1 is found in airway surface goblet cells.188
TFF2 is rare or absent in the human lung.
When secreted into the airway lumen,189
EGF may synergize with TFF peptides to enhance cell migration and wound coverage.190
Successful mucosal repair requires the subsequent differentiation of migrated cells and the regeneration of the normal mucosal architecture. This second phase of repair is cytokine-dependent and growth factor-dependent, and it occurs at the basolateral surface. In the GI tract, cytokines (including IL-1, IL-2, IL-4, IL-15, and interferon-
) and growth factors (ie, EGF, hepatocyte growth factor, and fibroblast growth factors) stimulate repair that appears to proceed through a TGF-ß-dependent pathway.191
192
193
194
Restitution/repair may play a role in the airway remodeling that accompanies conditions such as asthma and chronic bronchitis in which repetitive injury occurs.
Mucosal protection and repair are overlapping functions, and repeated exposure to irritants causes adaptive changes in the mucosal epithelium that resemble the repair response. GI metaplasia exhibits increased expression of EGFR (c-erbB1), TFFs, and mucins, and evidence exists for cross-regulation within and between these families. For example, TFF3 transactivates the EGFR via tyrosine phosphorylation through a ras-dependent pathway.195
Conversely, EGFR ligands stimulate the increased expression of TFFs in gastric and breast cancer cell lines.196
197
In humans, the EGFR is detected on the lateral cell membranes of basal cells and occasionally in nonciliated columnar cells of the large bronchi, as well as peripherally in Clara cells and in some alveolar type II cells.189
Two other members of the EGFR family (ie, c-erbB2 and c-erbB3) have been detected in normal bronchial tissue and appear to be present in the same cells expressing EGFR.198
When multiple members of the erbB family are coexpressed in the same cell, changes in the expression level of any one receptor can alter ligand preference and the activation of downstream pathways.199
200
At least seven ligands for the EGFR have been identified. Of these, EGF, TGF-, amphiregulin, betacellulin, and heparin-binding EGF-like growth factor proteins have been detected in human serous acinar gland cells. Four of the five proteins have been detected as RNA in the lung.201
The synthesis of several ligands is observed in macrophages, platelets, eosinophils, and T cells, as well as in columnar bronchial epithelial cells, where basolateral release to EGFR-expressing basal cells may regulate their function.198
201
Each of the seven peptides is synthesized as a membrane-bound precursor that is released by proteolytic cleavage, resulting in an active growth factor that is either diffusible (ie, EGF and TGF-) or contains a heparin-binding domain that facilitates binding to the extracellular matrix (ie, amphiregulin and heparin-binding EGF-like growth factor).202
At sites of injury that have been initiated by chemical irritants, augmented protease activity may lead to the cleavage of precursor EGFR ligands.203
Ligand binding induces dimerization of the EGFRs, followed by the autophosphorylation of receptor tyrosines, the activation of ras, and the initiation of the mitogen-activated protein kinase cascade. The EGFR is also rapidly activated independently of the ligand by oxidative or osmotic stress.204
205
206
EGFR expression is typically low in the airways of pathogen-free rats but can be induced by proinflammatory stimuli. The subsequent instillation of EGFR ligands induces Muc5ac expression and goblet cell hyperplasia that is inhibited by pretreatment with an EGFR tyrosine kinase inhibitor.169
Mucous metaplasia induced by chemical,170
mechanical,207
or immune40
irritation also has been shown to be dependent on EGFR activation. These data suggest that, despite a paucity of mucous glands, the rat represents a valuable model of irritant-induced mucous metaplasia.
|
Tobacco Smoke Aldehydes: Acrolein
|
|---|
The major aldehydes (including formaldehyde, acetaldehyde, proprionaldehyde, and acrolein) have numerous indoor and outdoor sources.208
209
210
Of these, acrolein (CH2 = CHCHO) is a feedstock in chemical manufacturing that is formed in many occupational settings; results from the combustion of cooking oils, wood smoke, and diesel exhaust; and is one of the most potent irritants known.211
Present in high concentrations in tobacco smoke (up to 50 ppm in mainstream), acrolein also is a component of photochemical smog.210
212
213
214
Notably, second-hand tobacco smoke contains high levels of acrolein because aldehydes are enriched in sidestream smoke due to the lower combustion temperatures of smoldering cigarettes. Acrolein can penetrate the upper respiratory passages and forms a highly reactive zwitterion (+CH2CH = CHO-) through electron rearrangement of the ,ß-unsaturated bond.
Following acute acrolein exposure, inflammatory mediators and monocytes increase in BAL fluid.215
Animals exposed repeatedly to acrolein develop histologic changes including epithelial damage, mucus hypersecretion, and bronchiolitis that is marked by excessive macrophage accumulation in the airways.216
Moreover, Fisher 344 rats that were exposed to 4 ppm acrolein for 2 months developed the cardinal signs of emphysema.217
Importantly, these decrements persisted even after the cessation of exposure.
In rats, acrolein exposure (4 ppm for 6 h) produces emphysematous changes that develop over several weeks to months and persist for several weeks after the end of exposure,217
a time course that is similar to that produced by cigarette smoke.45
54
The dose used in these studies is relevant to that produced by cigarette smoke. This exposure regimen can produce an acrolein dose that is equivalent to that produced by smoking 1.5 packs of cigarettes per day, based on acrolein concentrations found in mainstream smoke.218
In addition, because acrolein exposure does not contain nicotine, it may be a useful tool with which to study the genetic determinants of COPD produced by cigarette smoke in rats. Furthermore, strain differences in animal species provide a powerful tool with which to uncover genetic factors controlling lung diseases. Coexposure to nicotine may confound genetic analyses, as genetic differences in nicotine tolerance and/or dependence between mouse219
220
and rat strains221
222
can alter behaviors such as ventilatory rate and sleep time, and, thereby, can alter the deposition of cigarette smoke.
Although other components of cigarette smoke may be used to cause emphysema, acrolein may be preferred for several reasons. Thus, nitrogen dioxide, another component of cigarette smoke, requires very high doses to produce the alveolar enlargement that is relevant to COPD patients. Sulfur dioxide is not a major component of cigarette smoke and is highly water-soluble, depositing mainly in the large conducting airways, and produces bronchitis but little emphysema. Unlike cigarette smoke and acrolein, most of the pulmonary effects of sulfur dioxide are rapidly reversible on the cessation of exposure.
Published literature suggests that rat strains vary in pulmonary responses to acrolein. For example, Kutzman and colleagues223
224
225
studied the pulmonary effects of acrolein on Dahl rat strains, which have been used to study the genetic determinants of high BP. One strain (Dahl-sensitive [DS]) develops salt-induced hypertension, while the other strain is resistant (ie, Dahl-resistant [DR]).226
227
These strains were found to have variable sensitivity to acrolein with normal dietary salt levels and independent of changes in systemic BP. Exposure to 4.0 ppm acrolein resulted in epithelial hyperplasia, alveolar elastin destruction, and monocytic infiltration, which differed between the DS and DR strains. Similarly, Costa et al228
reported that DS strains are more sensitive to ozone than DR rats, again independently of altered BP. In addition, it has been reported229
230
231
that Sprague-Dawley, Wistar, and Fisher 344 rats vary in pulmonary responses to instilled residual oil fly ash or inhaled ozone.
|
Acrolein-Induced Muc5ac Expression in Sprague-Dawley Rat Airways
|
|---|
To determine whether histologic signs of mucus hypersecretion are associated with steady-state Muc2 and Muc5ac mucin gene expression, Sprague-Dawley rats were exposed to acrolein (3 ppm, 6 h/d, 5 d/wk), and the trachea, mainstem bronchi, and intrapulmonary airways were examined.176
The temporal expression of acidic mucin glycoprotein, Muc2 and Muc5ac mRNA, and Muc5ac protein were determined.
Tracheobronchial Muc5ac mRNA increased within 2 days (Fig 2
) and was accompanied by an increase in Muc5ac immunostaining of the surface epithelium and submucosal glands. Compared to the trachea, increases in lung Muc5ac mRNA (Fig 3
) and mucin glycoprotein (Fig 4
) were delayed (increasing, respectively, on days 5 and 9). Glycoprotein-positive cells were enumerated, and the percentage of mucous cells per surface epithelial cell was calculated for the large-diameter intrapulmonary airways (ie, > 0.8 mm) and the small-diameter intrapulmonary airways (ie, 
0.8 mm). This cutoff corresponds to generations 11 to 16 in the rat (ie, airways that would be 0.2 mm in humans). Glycoprotein formation in the small-diameter airways was delayed and increased along with Muc5ac immunostaining in the airway lumen and lung epithelium on day 12. In contrast, Muc2 was not significantly changed in the lung or trachea. These findings indicate that acrolein-induced mucus hypersecretion is due, at least initially, to increases in Muc5ac gene expression rather than to Muc2 gene expression.
View larger version (28K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2. Time course of mucin mRNA in the trachea and mainstem bronchi of acrolein-exposed rats. Total RNA was isolated, mucin mRNA levels were determined by RT-PCR, and the results were normalized to ß-actin. Values are given as the mean ± SE (six animals).176
|
|
View larger version (26K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3. Time course of mucin mRNA in the lung of acrolein-exposed rats. mRNA was isolated from lung tissue (without trachea and mainstem bronchi), and mucin mRNA levels were determined by RT-PCR and normalized to ß-actin. Values are given as the mean ± SE (six animals).176
|
|
View larger version (26K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4. Time course of mucous cell metaplasia in acrolein-exposed rats (6 h/d, 5 d/wk). Sections of formalin-fixed and paraffin-embedded lungs were stained with Alcian blue and were viewed by light microscopy. The percentage of mucous cells per airway was calculated in large-diameter intrapulmonary airways (ie, > 0.8 mm) and small-diameter intrapulmonary airways (ie, 0.8 mm). Values are given as the mean ± SE (three animals per day).
|
|
|
Regulation of Human Airway Mucins by Acrolein and Inflammatory Mediators
|
|---|
Acrolein can act directly as an electrophilic compound to deplete protein sulfhydryls/glutathione in the respiratory epithelium232
233
or to inactivate metabolizing enzymes.234
235
Although little is known about the cis-activating and trans-activating elements that regulate Muc5ac expression, electrophiles can activate gene expression through a mechanism involving glutathione depletion and the production of reactive oxygen species.236
This function is mediated by a cis-regulatory element (ie, electrophile response element) that consists of adjacent activator protein-1-like binding sites that are activated by fos/jun heterodimeric transcription factors.237
In addition, other oxidants can stimulate mucus secretion and gene expression.22
172
238
239
Acrolein may also initiate mucus hypersecretion indirectly via inflammation. It causes acute and chronic pulmonary inflammation that is characterized, respectively, by neutrophils or macrophages and monocytes.175
215
224
Neutrophil and neutrophil elastase levels often are elevated in the sputum of patients with COPD.62
240
Neutrophil elastase can induce mucous cell metaplasia,12
is a potent mucus secretagogue,13
and can induce the transcription of IL-8,241
secretory leukocyte protease inhibitor,242
and 1-antitrypsin.243
While normally present in the lung, macrophage numbers increase markedly in COPD patients, becoming pronounced in the respiratory bronchioles (ie, 0.2 mm) where emphysematous changes are first manifested.244
Acrolein clearly affects a similar area in the rat lung (Fig 4)
. Macrophages are also a source of mediators that can initiate mucus secretions and include IL-1, eicosanoids (ie, PGE2 and 15-HETE), and macrophage-derived mucus secretagogue-68.17
18
26
To test whether acrolein increases mucin mRNA directly, human airway cells were exposed to acrolein in a cell culture.30
Dose-dependent increases in MUC5AC mRNA were observed in NCI-H292 cells after 4 h of exposure to 0.01, 0.1, and 1.0 nM acrolein (Fig 5
). In contrast, MUC5B mRNA levels did not increase. The finding that MUC5AC levels increase when MUC5B levels remain constant is consistent with the dual regulation of mucin genes that includes an inducible (or immediate) form and constitutive (or delayed) form. At concentrations of 100 nM acrolein, MUC5AC, MUC5B, glyceraldehyde-3-phosphate dehydrogenase, and total RNA levels were significantly less than control levels, suggesting that this dose is cytotoxic. These findings indicate that relatively low doses of acrolein (ie, 0.1 to 10 nM) induce mucin formation and that acrolein can act directly on epithelial cells.
View larger version (24K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5. Dose-response relationship of acrolein induction of mucin mRNA in airway epithelial cells. NCI-H292 cells were exposed for 4 h, and mRNA was assessed by RT-PCR. Top: mean (± SE) mucin densitometry levels (three tests). Bottom: duplicate determinations of mucin. Lane 1 = control; lane 2 = 0.01 nm acrolein; lane 3 = 0.1 nm acrolein; lane 4 = 1.0 nm acrolein; lane 5 = 10 nm acrolein; and lane 6 = 100 nm acrolein.30
|
|
To examine mucin gene regulation at the mRNA level, we also compared the effects of acrolein to those of eicosanoids and TNF- on MUC5AC and MUC5B
transcript levels in NCI-H292 cells (Fig 6
, 7
).30
Cells were exposed also to 15-HETE, PGE2, and phorbol 12-myristate 13-acetate (PMA) (Fig 7)
. Again, the levels of MUC5AC mRNA, but not MUC5C mRNA, increased. Consistent with these results, TNF- and PMA increased MUC5AC, by (mean ± SE) 2.3 ± 0.2-fold and 3.1 ± 0.2-fold, respectively, in A549 cells. In both cell lines, MUC5B could be detected under basal conditions with 10-fold less total mRNA than MUC5AC. The abundance of mRNA detected by reverse transcriptase-polymerase chain reaction (RT-PCR), and the lack of induction by the agonists of mucus secretion suggests that MUC5B is expressed constitutively.
We also examined the time course of MUC5AC mRNA induction by TNF-.30
Confluent NCI-H292 cells were exposed for up to 16 h to 1.0 nM TNF-, the total RNA was isolated, and the steady-state mRNA was analyzed by RT-PCR. Treatment induced a rapid, transient increase in MUC5AC mRNA that was maximal between 2 and 4 h without altering MUC5B and glyceraldehyde-3-phosphate dehydrogenase. In addition, because steady-state mRNA levels are controlled by transcription rate and message stability, we also examined the ability of TNF- to increase message half-life by pharmacologically inhibiting transcription after peak message induction at 2 h. Without TNF- stimulation, the half-life of MUC5AC mRNA was approximately 4 h (Fig 8
). After TNF- stimulation, the approximate half-life was increased to almost 10 h. This 2.5-fold increase in mRNA half-life could account for most of the threefold increase in MUC5AC steady-state mRNA levels observed after TNF- treatment (Fig 6)
. Similar results have been reported with neutrophil elastase.14
Therefore, acrolein can increase mucin gene expression directly and indirectly, through inflammatory mediators.
These studies with human airway epithelial cells in culture indicated that MUC5AC mRNA increased with exposure to 0.01 to 1.0 nmM acrolein, suggesting that acrolein can act directly on airway cells to increase MUC5AC transcription levels in low concentrations (ie, levels relevant to cigarette smoking). Several inflammatory mediators (ie, 15-HETE, PGE2, and TNF-) and a protein kinase C activator (ie, PMA) increased MUC5AC transcription levels, suggesting that acrolein also can induce mucus hypersecretion indirectly. In these studies, levels of MUC5B mRNA did not change significantly with any treatment. TNF- increases MUC5AC mRNA levels through message stabilization. Thus, during irritant exposure, induced MUC5AC also may increase quickly through transcript stabilization. The constitutive expression of MUC5B, which has been associated with submucosal glands, might reflect a longer term mechanism with which to increase mucus production through increased submucosal gland size and cell number to yield an increase in the amount of constitutively expressed mucus. The latter two mechanisms are consistent with the pathogenesis of mucin hypersecretion in COPD.
|
Mucin Expression in FVB/N Mice
|
|---|
To examine further the relationship between inflammation and mucin mRNA expression, FVB/N mice were exposed to 3.0 ppm acrolein 6 h/d, 5 d/wk for 0 to 3 weeks.179
Unlike rats, no change occurred through the first week of exposure. However, after 2 weeks, Muc5ac levels increased approximately fivefold and increased a further 10-fold at 3 weeks. No measurable Muc2 mRNA was detected in the control or exposed animals. In FVB/N mice, acrolein caused biphasic leukocyte infiltration, the first phase (day 1) being marked by a decrease in macrophages and an increase in neutrophils. In contrast, the second phase consisted of an increase in macrophages and a return of neutrophils to control levels (Fig 9
).
View larger version (14K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9. Inflammatory cells and neutrophil elastase activity in BAL fluid from acrolein-exposed FVB/N mice. Mice were exposed to 3.0 ppm acrolein 6 h/d, 5 d/wk for 0 to 19 days. The numbers of macrophages (top) and neutrophils (middle) in BAL fluid were determined from total cell counts and cell differential counts. Bottom: neutrophil elastase activity determined in BAL supernatants. Values are given as the mean ± SE (three animals per day).244
|
|
Through collaboration with Dr. Steven Shapiro, additional tests were conducted with C57BL/6-J MME- competent (MME+/+) and MME-/- mice. Following acrolein exposure, neutrophil levels were < 1%, whereas macrophage levels increased in C57BL/6-J and MME+/+ mice, but not in MME-/- mice (Fig 10
). C57BL/6-J mice responded similarly to FVB/N mice with an approximately fourfold macrophage increase. MME+/+ mice exhibited an approximately twofold increase, whereas knockout mice had no change in the levels of macrophages in BAL fluid. The increase in Muc5ac steady-state mRNA levels in C57BL/6-J mice after acrolein exposure was similar to that of FVB/N mice (Fig 10)
. Interestingly, increases in Muc5ac mRNA coincided with the increases in macrophage levels. FVB/N mice had the greatest increase in Muc5ac mRNA, followed by C57BL/6-J, MME+/+, and MME-/- animals.
View larger version (25K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10. Mouse strain differences in acrolein-induced macrophage accumulation and MUC5AC mRNA levels in mice. FVB/N, C57BL/6-J, MME-12-competent (MME+/+) and MME-12-deficient (MME-/-) mice were exposed to 3.0 ppm acrolein (6 h/d, 5 d/wk for 19 days). Top: The numbers of macrophages in acrolein-exposed mice (hatched bars) or mice breathing filtered air (open bars) determined in BAL fluid. Bottom: Muc5ac mRNA levels determined by RT-PCR and normalized to ß-actin. Values are given as the means ± SE (three to four animals.244
|
|
The increases in Muc5ac levels occurred after 12 to 20 days of acrolein exposure and, thus, were later in onset than in rats. Regardless, mice develop mucous cell hypersecretion following aldehyde exposure, and because inbred mouse strains are a powerful tool in genetic analyses, they may also be useful in future studies. Macrophages decreased and then increased slowly, which is a comparable time course to Muc5ac expression. Neutrophils increased transiently, and neutrophil elastase did not change significantly. Different mouse lines (ie, FVB/N, C57BL/6, MME+/+, and MME-/-) varied in the levels of macrophages in BAL fluid in proportion to the variation in Muc5ac transcript levels. Moreover, MME-/- mice did not develop a Muc5ac response, suggesting that macrophage-mediated proteolysis may have a role in augmenting mucus hypersecretion. Again, this supports the in vitro observations that acrolein can increase mucin gene expression through direct and indirect mechanisms.
|
Rat Strain Differences in Muc5ac mRNA and Growth Factor Induction
|
|---|
Most recently, we have initiated acrolein exposure of the following seven inbred rat strains and one outbred rat strain: Wistar-Kyoto; Dahl salt-resistant (Rapp); ACI; Brown-Norway; Dahl salt-sensitive (Rapp); Fisher 344; Lewis; and Sprague-Dawley (outbred).245
Preliminary studies indicated that filtered-air control levels of Muc5ac mRNA were low among all strains except Brown-Norway rats, a strain that is susceptible to parainfluenza type I (Sendai) virus. After 2 weeks of exposure, six strains (ie, Sprague-Dawley, ACI, Lewis, Dahl salt-sensitive, Brown-Norway, and Fisher 344 rats) 
TOP
Abstract
Inflammation, Cigarette Smoking,...
