Epithelial cells are key participants in the early and
late events that follow acute lung injury. A recent review summarizes abnormal
wound repair as a model for understanding IPF
ref.
Following epithelial injury, surviving and regenerating respiratory epithelium
contributes to abnormal antifibrinolytic activity at sites of injury and
also elaborates various fibrogenic cytokines/growth factors important in
recruitment and phenotypic modulation of fibroblasts and myofibroblasts
ref1,
ref2.
Fibroblasts, in turn, may amplify alveolar damage by inducing epithelial
cell apoptosis
ref
and modulating epithelial phenotype through the complex interactions attributed
to the "epithelial-mesenchymal trophic unit"
ref1,
ref2.
Despite the morphologic overlap, we now know that the fibroblast foci of
UIP/IPF differ from the
Masson
bodies
of
organizing pneumonia in several
important ways including fibroblast/myofibroblast phenotype, rates of fibroblast/myofibroblast
apoptosis, modulation of extracellular matrix deposition, and extent of
neovascularization
ref.
Differences in the character and extent of reepithelialization may play
a pivotal role in determining why "fibroblast foci" of UIP/IPF resolve
into a scar while the Masson bodies of organizing pneumonia disappear without
a trace. Disordered reepithelialization in UIP/IPF has previously been
linked to upregulation of epithelial cell apoptosis and impaired adhesion
to damaged basement membranes
ref1,
ref2,
ref3.
Pathogenesis : GM-CSF–initiated
signaling plays unique, nonredundant roles in alveolar macrophage function
and pulmonary homeostasis, including terminal differentiation and survival
of macrophages, intracellular lipid metabolism, surfactant catabolism and
recycling, expression of pathogen receptors, and phagocytosis and killing.
These functions are inhibited when neutralizing GM-CSF autoantibodies bind
to GM-CSF, block its binding to the chain of the GM-CSF receptor,
and prevent assembly of the GM-CSF–receptor complex. Thus, inhibition of
GM-CSF binding to its receptor by autoantibodies results in decreased clearance
of surfactant from the alveolar spaces, the hallmark of pulmonary alveolar
proteinosis. The hypothesis that pulmonary alveolar proteinosis is due
to ineffective signaling by GM-CSF receptors was first put forward in 1994,
when Stanley et al.
ref
and Dranoff et al.
ref
simultaneously reported on mice in which both alleles of the gene for GM-CSF
are disabled. These GM-CSF
–/– mice have striking pulmonary pathologic
characteristics that closely resemble those of patients with pulmonary
alveolar proteinosis, suggesting that the intracellular signaling initiated
by the binding of GM-CSF to its receptor is critical to pulmonary surfactant
homeostasis. Signaling initiated by GM-CSF receptors is mediated through
PU.1, a transcription factor modulating the expression of many genes that
are important in the terminal differentiation of alveolar macrophages.
Functions regulated by PU.1 include surfactant degradation, expression
of pathogen pattern-recognition receptors, toll-like–receptor signaling,
phagocytosis, and bacterial killing. Reconstitution of PU.1 in GM-CSF–deficient
alveolar macrophages restores most of these functions. GM-CSF signaling
also enhances the function of PPAR, another transcription factor that regulates
many cellular functions, including intracellular lipid metabolism. These
findings explain how inhibiting the binding of GM-CSF to its receptor causes
decreased clearance of surfactant from the alveolar spaces. GM-CSF is required
for the terminal differentiation and function of alveolar macrophages but
not for those of other tissue macrophages — which may explain why pulmonary
alveolar proteinosis is primarily a lung disease
ref.
The discovery that nearly all patients with primary pulmonary alveolar
proteinosis have high titers of GM-CSF antibodies has led to the development
of new therapies based on
G-CSF
supplementation. Previously, therapy consisted of whole-lung lavage, which
results in numerous complications and does not address the pathogenetic
mechanisms. Administration of exogenous GM-CSF appears to help many patients,
and its potential as a subcutaneous or aerosolized therapy is being evaluated.
Whether patients with pulmonary alveolar proteinosis have defects in host
defense and innate immunity is not clear, nor is the effect of such defects.
GM-CSF
–/– mice clearly have a defect in pulmonary defense against
pathogens, as well as extrapulmonary abnormalities and a decreased susceptibility
to experimentally induced autoimmune disorders. Patients with pulmonary
alveolar proteinosis do not have deficient expression of GM-CSF, but they
have neutralizing GM-CSF autoantibodies, which may account for differences
between host defense defects in such patients and defects in mice. Although
secondary infections have been described in these patients, our understanding
of host defense in this disorder is limited by its rarity, the variability
of its outcomes, and reporting biases. Patients may be at risk for secondary
infections, but bacteria that commonly cause respiratory infections are
not often the inciting agents
ref.
Opportunistic pathogens — most commonly
Nocardia, but occasionally
Cryptococcus,
Histoplasma,
Aspergillus,
or
Mycobacterium tuberculosis — have been reported in about 13%
of patients
ref,
with systemic infection documented in some cases. In a review of 65 cases
in which death was attributable to pulmonary alveolar proteinosis, respiratory
failure was the apparent cause of death in 47 cases (72%), whereas uncontrolled
infection was the cause in 12 (18%)
ref.
These findings suggest that clinically important lung infections occur
in some patients and that systemic infections are less common. Few researchers
have investigated whether GM-CSF autoantibodies cause defects in cells,
other than alveolar macrophages, that express GM-CSF receptors. Uchida
et al. report
ref
on the effects of GM-CSF autoantibodies on neutrophil functions. Their
observations show that neutrophils from patients with pulmonary alveolar
proteinosis have defects in both basal and GM-CSF–primed antimicrobial
functions. In particular, the phagocytic index and phagocytic capacity
of neutrophils isolated from these patients were approximately 90% and
30% lower, respectively, than those of neutrophils from healthy control
subjects. The basal capacity for adhesion to plastic, the oxidative burst
in whole blood, and the killing of
Staphylococcus aureus were also
reduced. Furthermore, GM-CSF priming in vitro was impaired. Similar defects
in basal neutrophil functions were observed in GM-CSF
–/– mice,
although no defect in GM-CSF priming was observed. The defect in affected
human neutrophils was mimicked by treating blood from healthy control subjects
with GM-CSF autoantibodies. Thus, neutrophils from patients with pulmonary
alveolar proteinosis have functional defects when tested
in vitro.
How the signaling of GM-CSF through its receptor on neutrophils augments
neutrophil function remains an important question
ref.
PU.1 is apparently critical in GM-CSF–initiated signaling in alveolar macrophages
but not in neutrophils, since Uchida et al. show that neutrophils from
patients with pulmonary alveolar proteinosis do not have lower PU.1 expression
than those from healthy control subjects. Furthermore, defects in neutrophil
function are apparent in vitro, but it is not clear how they contribute
to a defect in host defense in vivo. Humans and mutant mice with defects
in innate immunity, including abnormal neutrophil function, often have
increased blood levels of neutrophils, G-CSF, and cytokines; the failure
to destroy pathogens leads to the persistence of pathogen-induced stimuli.
In contrast, in the study by Uchida et al., patients with pulmonary alveolar
proteinosis who had defective neutrophil function in vitro did not have
increased neutrophil counts and serum G-CSF levels, suggesting that their
host defense and innate immunity were sufficient for managing daily exposures
to common pathogens and commensal organisms. Clinically apparent infections,
especially with opportunistic microbes, do occur in some patients; microbes
were identified at presentation in more than half the patients studied
by Uchida et al. Whether host defense mechanisms are able to compensate
for defects in neutrophil and macrophage function until the environment
delivers a particular pathogen or a large load of pathogens, or until a
combined assault is made on the immune system, remains to be elucidated.
The absence of repeated infections, particularly in the lungs, is surprising,
given the magnitude of the neutrophil defects seen
in vitro; perhaps
products of surfactant degradation have important antimicrobial effects.
Clearly, the study by Uchida et al. raises important, provocative questions
that beg to be pursued.