Pulmonary Perspective
Oxidative Stress in Pulmonary Fibrosis
A Possible Role for Redox Modulatory Therapy
Vuokko L. Kinnula, Cheryl L. Fattman, Roderick J. Tan, and Tim D. Oury
Division of Pulmonary Medicine, Department of Medicine, University of Helsinki, Helsinki, Finland; and Department of Pathology, University
of Pittsburgh, Pittsburgh, Pennsylvania
Idiopathic ulmonary fibrosis (histopathology of usual interstitial
pneumonia) is a progressive lung disease of unknown etiology. No
treatment has been shown to improve the prognosis of the patients
with this disease. Recent evidence, including the observations that
the patients with idiopathic pulmonary fibrosis have higher levels
of oxidant stress than control patients, and a recent multicenter
European study examining the effect of the antioxidantN-acetylcys-
teine on the progression of idiopathic pulmonary fibrosis suggest
that the cellular redox state may play a significant role in the pro-
gression of this disease. These complex mechanisms include activa-
tionofgrowth factors aswell as regulation ofmatrixmetalloprotein-
ases and protease inhibitors. Potential future approaches for the
therapy of interstitial pulmonary fibrosis may involve synthetic agents
able tomodulate cellular redox state. Investigation into therapeutic
approaches to inhibit oxidant-mediated reactions in the initiation
and progression of pulmonary fibrosis may provide hope for the
future treatment of this disease.
Keywords: antioxidant; idiopathic pulmonary fibrosis; oxidant; radical
Idiopathic pulmonary fibrosis (IPF) is a prototype of idiopathic
interstitial pneumonias, a chronic disease of the lung parenchyma
that leads to diffuse scarring and end-stage tissue fibrosis (1).
Typical features in this disease include dyspnea, diffuse intersti-
tial infiltrates, progressive lung fibrosis, and poor prognosis. The
pathologic changes in IPF include patchy fibrotic lesions that
vary both in age and activity, and only weak inflammation. The
focal zones of fibroblast proliferation are called “fibroblastic
foci” and appear to occur at sites of recent alveolar injury (1, 2).
The biochemical mechanisms in the pathogenesis of IPF are still
poorly understood and medical therapies have thus far offered
little if any benefit against the progression of this disease (1, 3, 4).
As the name implies, there is no known etiologic stimulus
that initiates this disease. Current evidence suggests that both
endogenous and exogenous stimuli may injure the alveolar epi-
thelium. This is followed by an abnormal repair process in indi-
viduals unable to effectively heal the damage. The majority of
the original articles and reviews on IPF have focused on the
(Received in original form January 5, 2005; accepted in final form May 3, 2005)
This and the original studies of V.L.K. have been partly funded by the Finnish
Antituberculosis Association Foundation, the Juselius Foundation, and the Founda-
tion of Helsinki University Hospital. This work is also supported by National Insti-
tutes of Health grants R01 HL63700, R01 HL73745, 1F30ES013621, and the
American Heart Association Established Investigator Award (T.D.O).
Correspondence and requests for reprints should be addressed to Vuokko Kinnula,
M.D., Ph.D., Department of Medicine, Division of Pulmonary Medicine, University
of Helsinki, P.O. Box 22 (Haartmaninkatu 4), FI-00014 Helsinki, Finland. E-mail:
vuokko.kinnula@helsinki.fi
Am J Respir Crit Care Med Vol 172. pp 417–422, 2005
Originally Published in Press as DOI: 10.1164/rccm.200501-017PP May 13, 2005
Internet address: www.atsjournals.org
fibrotic process, importance of the fibroblastic foci in disease
progression, and on the therapeutic regimens with antiinflam-
matory/antifibrotic drugs, such as corticosteroids, cytotoxic drugs,
and IFN-� (3, 5). However, there are a number of studies sug-
gesting that the cellular redox state and the balance of oxidants/
antioxidants play a significant role in the progression of pulmo-
nary fibrosis in animal models and also possibly in human IPF.
This article reviews the importance of oxidants and lung oxidant–
antioxidant balance in human lung fibrogenesis. Although exper-
imental models of lung fibrosis are difficult to extrapolate to
human IPF due to differences in inflammation and the occur-
rence of fibroblastic foci, many similarities exist between them.
Therefore, the key experimental studies examining potential
mechanisms by which oxidative stress contributes to pulmonary
fibrosis will also be emphasized.
MAJOR GENERATORS OF FREE RADICALS
IN HUMAN LUNG
The lung is exposed to higher oxygen tension than other tissues.
Exogenous oxidants and pollutants further increase oxidant pro-
duction and activate inflammatory cells to generate free radicals.
Many of these agents, including hyperoxia, cigarette smoke, as-
bestos fibers, drugs, and radiation, are also known to be associ-
ated with fibrotic interstitial lung reactions. Several important
reactive oxygen species (ROS) are generated endogenously in
these circumstances. They include the superoxide radical, hydro-
gen peroxide (H2O2), and the hydroxyl radical. The major en-
zymes/reaction pathways that can be activated to generate ROS in
human lung include nicotinamide adenine dinucleotide phosphate
oxidases, myeloperoxidase, eosinophil peroxidase, mitochondrial
electron transport chain, and possibly xanthine oxidase (reviewed
in References 6, 7). Microsomal and nuclear membrane cyto-
chromes can also generate ROS, and these redox-cycling reac-
tions may be linked to smoke- and drug-induced interstitial lung
diseases. A number of reactive nitrogen species (RNS), such as
peroxynitrite, can also be formed by reactions of superoxide
with nitric oxide (NO). The majority of NO is produced by the
inducible form of nitric oxide synthase (iNOS, NOS2), especially
during inflammatory states in the lung. Constitutive forms of
NOS are also expressed widely in human lung cells and have
been shown to regulate cell signaling in normal cell homeostasis.
Overall, a wide variety of oxidants are produced in response to
injuries leading to pulmonary fibrosis. These oxidants can acti-
vate several genes related to cell growth, cell death, and fibro-
blast proliferation.
PRIMARY ANTIOXIDANT PROTECTION OF THE LUNG
A balance between intracellular and extracellular oxidants and
antioxidants is a prerequisite for normal lung homeostasis. The
lung has highly specialized and compartmentalized antioxidant
418 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 172 2005
defenses to protect against ROS and RNS. These include the
following: (1) small-molecular-weight antioxidants (e.g., gluta-
thione, vitamins, uric acid), (2) mucins, (3) metal-binding pro-
teins (transferrin, lactoferrin, metallothionein, etc.), (4) super-
oxide dismutases (SODs; e.g., mitochondrial manganese SOD
[MnSOD], intracellular copper zinc SOD [CuZnSOD], and ex-
tracellular SOD [ECSOD]), (5) a group of enzymes that de-
composes H2O2 (numerous glutathione-associated enzymes and
catalase), (6) detoxification enzyme systems (e.g., glutathione-
S-transferases), and (7) other redox regulatory thiol proteins
(e.g., thioredoxin-peroxiredoxin system and glutaredoxins) (7–9).
These enzymes are localized in specific cell types, such as bron-
chial and alveolar epithelial cells and alveolar macrophages,
but can also exist extracellularly. Many of these enzymes are
regulated by a redox-sensitive Nrf2 transcription factor that is
essential to the induction of antioxidant enzymes and related
proteins in the lung. Nrf2 may be of particular importance in
IPF because Nrf2 deficiency significantly enhances bleomycin-
induced pulmonary fibrosis in mice (10). Induction of these anti-
oxidant enzymes and related proteins after pulmonary insults
may protect the lung and promote normal repair. Conversely,
impaired induction or inactivation/clearance of antioxidant en-
zymes may result in a sustained redox imbalance that may con-
tribute to the progression of pulmonary fibrosis.
OXIDANT–ANTIOXIDANT IMBALANCE IN IPF
Several studies suggest that oxidant–antioxidant imbalances in
the lower respiratory tract play a critical role in the pathogenesis
of IPF. For example, pulmonary inflammatory cells of patients
with IPF generate higher levels of oxidants than those in control
patients (11). Bronchoalveolar lavage fluid of patients with IPF
show elevated levels of myeloperoxidase and eosinophil cationic
protein, suggesting a pathophysiologic role for neutrophils and
possibly eosinophils in this disease (12). There is also evidence
that increased myeloperoxidase is associated with epithelial in-
jury in IPF (11). Mitochondrial generation of ROS has been
suggested to be associated not only with increased cellular oxida-
tive stress but also with apoptosis of alveolar epithelial cells
(13, 14). Bronchoalveolar lavage fluid of patients with IPF con-
tains higher levels of 8-isoprostane, a biomarker of oxidative
stress, than that of control subjects (15). Patients with IPF have
also been shown to have elevated levels of exhaledNO (reviewed
in Reference 16). In addition, lung specimens of the patients
with IPF show elevated expression of iNOS (17, 18). These
findings suggest that patients with IPF have increases in both
oxidative and nitrosative stress.
The presence of an oxidant–antioxidant imbalance in IPF is
also indicated by the finding of altered levels of antioxidants in
the lungs of affected patients. For example, patients with IPF
have decreased levels of reduced glutathione in the epithelial
lining fluid of their lungs (19–22). Several studies suggest that a
number of antioxidant and detoxification enzymes are elevated
in inflammatory and/or granulomatous interstitial lung diseases
and in areas of epithelial regeneration; however, they are low/
absent in the fibrotic lesions of IPF lungs. These enzymes include
MnSOD, catalase, glutamate cysteine ligase (�-glutamyl cysteine
synthetase, the rate-limiting enzyme in glutathione synthesis),
thioredoxin, glutaredoxin, and heme-oxygenase 1 (23–26). Al-
though a decrease in antioxidant capacity likely increases oxida-
tive stress, the observed elevations in antioxidants probably at-
tempt to compensate for increased oxidative stress in the lung.
It is, however, clear that there is a disruption of the normal
redox balance in the lungs of patients with IPF.
ROS AS MODULATORS OF PROTEASE–
ANTIPROTEASE BALANCE
Current evidence suggests that imbalance in proteases/antiprote-
ases contributes to the pathogenesis of IPF. The importance of
proteases in IPF was highlighted by the finding that matrilysin
(matrix metalloproteinase 7 [MMP-7]) is significantly associated
with pulmonary fibrosis in animal models, and is also overex-
pressed in human lung with IPF (27). Other MMPs are also
upregulated in IPF (28). Importantly, both ROS and RNS may
contribute to a protease–antiprotease imbalance because they
can activate MMPs and inactivate protease inhibitors.
MMPs share a commonmode of activation termed the “cyste-
ine switch” whereby disruption of the active site cysteine–zinc
bond leads to autocatalytic cleavage of the prodomain and activa-
tion of the latent enzyme (reviewed in Reference 29). Modifica-
tion of the cysteine switch by ROS or RNS can activate MMPs.
Hypochlorous acid oxidizes and activates MMP-7 (30). H2O2,
peroxynitrite, and oxidants produced by the xanthine/xanthine
oxidase system can activate both MMP-2 and MMP-9 (29). The
cysteine switch of MMP-9 can also be activated by NO (29). In
addition, ROS have also been shown to directly induce MMP
transcription (29). Importantly, ROS do not always activate pro-
teases, but can inactivate them as well (30). Therefore, it is likely
that local concentrations of ROS/RNS will determine whether
MMPs are activated or inactivated. It has been suggested that
the rapid activation and subsequent inactivation of MMPs by
oxidativemechanisms could regulate powerful “quantumbursts”
of proteolytic activity (31).
The primary antioxidant enzymes in the extracellular matrix
and alveolar lining fluid that may inhibit oxidative activation or
inactivation of MMPs are extracellular glutathione peroxidase
(32) and ECSOD (9). Augmentation of glutathione levels with
N-acetylcysteine (NAC) treatment has been shown to inhibit
MMP activation (33). Mice lacking ECSOD show increased
MMP activity in animal models of pulmonary fibrosis compared
with wild-type mice (unpublished observation). ECSOD binds
to the lung matrix via a matrix-binding domain that is itself
sensitive to proteolysis (9, 34). In fact, there is enhanced proteo-
lytic cleavage of ECSOD’s matrix-binding domain and clearance
of ECSOD from the matrix in various injury models that lead
to pulmonary fibrosis (35, 36). Overall, oxidative and proteolytic
processes can amplify each other and enhance injuries leading
to pulmonary fibrosis.
In addition to protease activation, both ROS and RNS can
inactivate protease inhibitors and further alter the protease–
antiprotease balance. Some examples include tissue inhibitor of
metalloproteinase (TIMP-1) inactivation by peroxynitrite (37)
and ROS-mediated inactivation of �2-macroglobulin (38). Secre-
tory leukoprotease inhibitor has been shown to ameliorate pul-
monary fibrosis, but notably can be inactivated by oxidants (39).
In addition, �1-proteinase inhibitor (�1PI) is sensitive to oxida-
tion and inactivation by peroxynitrite (40) and oxidants pro-
duced by myeloperoxidase and xanthine oxidase. The absence
of �1PI activity has been linked to the development of pulmonary
fibrosis, and mice treated with �1PI (41) are resistant to pulmo-
nary fibrosis. Collectively, this evidence suggests that oxidative
inactivation of�1PI contributes to the pathogenesis of pulmonary
fibrosis.
ROS AND TRANSFORMING GROWTH FACTOR �
Numerous cytokines and growth factors have been implicated
as mediators in the pathogenesis of pulmonary fibrosis. In addi-
tion to directly injuring lung cells and matrix, oxidants may also
contribute to the development of pulmonary fibrosis by their
Pulmonary Perspective 419
Figure 1. Potential roles of re-
active oxygen species (ROS) in
the pathogenesis of idiopathic
pulmonary fibrosis (IPF). Ex-
ogenous and endogenous irri-
tants in IPF create a redox
imbalance, resulting in the pro-
duction of ROS.Widespread ef-
fects on epithelium, myofibro-
blasts, growth factors (e.g.,
transforming growth factor �
[TGF-�]), inflammatory cells,
proteases (e.g., matrixmetallo-
proteinases [MMPs]), protease
inhibitors, and the extracellular
matrix (ECM) may ultimately
contribute to the development
of end-stage fibrosis. Shown
also are endogenous antioxi-
dants (green boxes) and the
steps at which they can protect
the lungs from the effects of
ROS. Processes outside the cell,
such as the activation of TGF-�
and MMPs, would be primarily
affected by the major extracellu-
lar antioxidants, including gluta-
thione (GSH), extracellular su-
peroxide dismutase (ECSOD),
and other small molecules, such
as ascorbate. Exogenous antioxidants, such as N-acetylcysteine (NAC) and SOD mimetics can augment antioxidant defenses and thus serve as
potential therapies for IPF. �1PI � �1-proteinase inhibitor; �2M � �2-macroglobulin; CuZnSOD � copper/zinc superoxide dismutase; ECSOD �
extracellular superoxide dismutase; eGPx � extracellular glutathione peroxidase; GPx � glutathione peroxidase; HO-1 � heme-oxygenase 1;
MnSOD � manganese superoxide dismutase; TIMP � tissue inhibitor of metalloproteinases.
direct effects on cytokines and growth factors. One such media-
tor, transforming growth factor � (TGF-�), is a key regulator
of both normalwound repair and the aberrant repairmechanisms
characteristic of many fibrotic diseases, including pulmonary fi-
brosis.
Previous studies have shown that TGF-�1, if released soon
after injury, acts primarily as a proinflammatory molecule be-
cause it has potent chemotactic effects on inflammatory cells
(42). Later, TGF-�1 function switches to resolution of inflam-
mation and initiation of repair. It has been hypothesized that
the persistence of chronic fibrosismay be due to unabated contin-
uation of repair processes after resolution of the inflammatory
response. TGF-� therefore may be an important mediator of
chronic but abnormal repair. For example, TGF-� is believed
to be a central regulator of the recruitment, activation, and
differentiation of myofibroblasts in early stages of tissue repair
(43). The persistence of the myofibroblastic phenotype in the
areas of active fibrosis is a characteristic finding in fibrotic lung
disease (44). TGF-� itself can also stimulate accumulation of
extracellular matrix through increased transcription of collagen
mRNA (45). Therefore, consistently elevated levels of TGF-�
in the lung may serve as a stimulus for myofibroblast activation
and production of extracellular matrix.
There are several potential interactions between TGF-� and
oxidants/antioxidants in the lung. For example, TGF-�–differen-
tiated myofibroblasts can themselves serve as a source of oxidant
production (46). Furthermore, in vitro studies have shown that
ROS increase the release of TGF-� from pulmonary epithelial
cells (47) and can directly activate TGF-� by disrupting its inter-
action with latency-associated peptide (48), suggesting a positive
feedback mechanism within the myofibroblast microenviron-
ment. TGF-� has been shown to activate NADPH oxidase in
human fibroblasts, leading to increased production of ROS (49).
Furthermore, TGF-� has been shown to downregulate glutamate
cysteine ligase mRNA synthesis, the rate-limiting enzyme in the
production of the antioxidant molecule glutathione (50), which
is known to be decreased in IPF (see above). Glutathione synthe-
sis is decreased in TGF-�–overexpressing mice, and these mice
are also more susceptible to oxidant-induced lung injury (51).
Thus, oxidants and TGF-� may interact to enhance the fibrotic
response in patients with IPF (Figure 1).
EXOGENOUS ANTIOXIDANTS AND ANTIOXIDANT
MIMETICS IN PULMONARY FIBROSIS
Experimental Models of Lung Fibrosis
It is well known that exogenous agents, including asbestos, radia-
tion, and drugs, can cause pulmonary fibrosis through production
of ROS/RNS in animal models (52). Studies examining these
models have shown not only increased oxidant burden on expo-
sure to these agents but also that exogenous treatment with
antioxidants can protect the lung in vivo. Themostwidely investi-
gated antioxidants in these models include glutathione, NAC,
and SODs.Glutathione, however, crosses cell membranes poorly
and can cause several side effects, including bronchoconstriction
(53, 54). As an alternative, NAC has been shown to improve
glutathione homeostasis by increasing cysteine levels, the rate-
limiting substrate in glutathione synthesis. Notably, NAC sig-
nificantly decreases primary inflammatory reactions, collagen
deposition, and the progression of bleomycin-induced lung fibrosis
420 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 172 2005
(55–57). NAC, however, also has prooxidant characteristics, and
there are studies suggesting that NAC does not necessarily im-
prove lung glutathione homeostasis (53, 54). Other compounds
with NAC-like activity, including glutathione esters and gluta-
thione precursors, have been tested in many laboratories, but
the effects of these compounds in fibrotic lung disorders is un-
known (54).
Another widely used group of antioxidants include the SODs
and their derivatives. Initially, SODs, encapsulated SODs, lipo-
somal SOD preparations, and recombinant MnSOD have been
shown to offer significant protection in animal models that lead
to fibrosis (reviewed in Reference 7). Because these compounds
were later shown to produce immunogenic complications, syn-
thetic small-molecular-weight SOD mimetics have been de-
veloped (58, 59). These agents include salen