J Appl Physiol 110: 1716–1722, 2011.
First published December 23, 2010; doi:10.1152/japplphysiol.01310.2010.
Review
HIGHLIGHTED TOPIC
Emergent Behavior in Lung Structure and Function
Complex systems in pulmonary medicine: a systems biology approach to lung
disease
David A. Kaminsky,1 Charles G. Irvin,1 and Peter J. Sterk2
1
Vermont Lung Center, Department of Medicine, Division of Pulmonary and Critical Care Medicine, University of Vermont
College of Medicine, Burlington, Vermont; and 2Department of Respiratory Medicine, Academic Medical Center, University
of Amsterdam, Amsterdam, The Netherlands
Submitted 5 November 2010; accepted in final form 22 December 2010
emergence; complexity; biological network; physiome; clinical phenotype
THE LUNG is a remarkably complex organ whose range of
function varies from the molecular (e.g., gas exchange) to the
whole organ level (e.g., bulk flow of air). An understanding of
lung health and disease must therefore necessarily incorporate
information from across this broad scale of organ structure and
function (Fig. 1). This information has to come from integrative physiology and is currently best obtained by a quantitative
systems biology approach (6, 70).
In systems biology, the approach to understanding a biological system involves integrating information from all levels of
structure and function of the system (1, 36, 37). This is, in fact,
what respiratory scientists do when they describe lung function
in terms of the forced expiratory volume in 1 s (FEV1), which
is a global, robust measure of function, such as in chronic
obstructive pulmonary disease (COPD) (67). This process
differs from the traditional “reductionist” approach, which
involves taking a system apart in order to identify and understand its component parts. For example, consider the complexities involved in the study of the transcription factor nuclear
factor kappa B (NF␬B) and its role in inflammation (68). While
this reductionist process is critical to understanding basic
Address for reprint requests and other correspondence: D. A. Kaminsky,
Pulmonary and Critical Care Medicine, Given D-213, 89 Beaumont Ave.,
Burlington, VT 05405 (e-mail: [email protected]).
1716
scientific and biological principles, complex biological systems
reveal a functional behavior that cannot be predicted just by the
sum of their parts. Such a property is known as “emergence”
and is the key principle in systems biology (1, 37, 40). Other
important principles that are cornerstones of this approach to
the complex, network structure of biological systems include
“robustness” and “modularity” (1, 37). Robustness confers
stability to a system in response to perturbation and is generally
characterized by the presence of feedback loops, redundancy,
and structural stability (3). Modularity relates to how multiple
components of a system function in a such a manner as to lead
to similar outcomes, providing robustness for the system and
allowing it to evolve (3). Together, these properties of emergence, robustness, and modularity comprise the system structure and function, and intrinsic methods for control and design,
that are needed to allow a complex biological system to adapt,
function, and evolve in response to variations and perturbations
in its environment (37).
Given the lung’s highly complex structure and function, a
systems biology approach to understanding the lung is an ideal
application of the method. Applying systems biology to the
study of lung disease involves the use of many resources (59).
These include a multidisciplinary team of scientists, from
biologists to mathematicians and clinicians, who can apply
their unique knowledge and skills and communicate with one
another across disciplines in order to acquire data sets (fre-
8750-7587/11 Copyright © 2011 the American Physiological Society
http://www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.33.4 on September 18, 2016
Kaminsky DA, Irvin CG, Sterk PJ. Complex systems in pulmonary medicine:
a systems biology approach to lung disease. J Appl Physiol 110: 1716 –1722, 2011.
First published December 23, 2010; doi:10.1152/japplphysiol.01310.2010.—The
lung is a highly complex organ that can only be understood by integrating the many
aspects of its structure and function into a comprehensive view. Such a view is
provided by a systems biology approach, whereby the many layers of complexity,
from the molecular genetic, to the cellular, to the tissue, to the whole organ, and
finally to the whole body, are synthesized into a working model of understanding.
The systems biology approach therefore relies on the expertise of many disciplines,
including genomics, proteomics, metabolomics, physiomics, and, ultimately, clinical medicine. The overall structure and functioning of the lung cannot be predicted
from studying any one of these systems in isolation, and so this approach highlights
the importance of emergence as the fundamental feature of systems biology. In this
paper, we will provide an overview of a systems biology approach to lung disease
by briefly reviewing the advances made at many of these levels, with special
emphasis on recent work done in the realm of pulmonary physiology and the
analysis of clinical phenotypes.
Review
SYSTEMS BIOLOGY IN LUNG DISEASE
1717
quently large in size) and discover links between the various
parts of a complex system; the application of an iterative
approach to discovery and problem solving, by using mathematical models based on experimental data, and then refining
the models as they are tested and new data are obtained; and
utilization of advanced mathematical and computational methods to handle large quantities of data, statistically analyze the
data, and apply an iterative modeling approach (Fig. 2).
Most papers dealing with systems biology focus on the
utilization of high-throughput genomic data to understand
patterns of gene expression in relation to phenotypic features of
various diseases. Other “-omic” approaches also include proteomics and metabolomics, related to protein expression and
metabolic biomarkers, respectively. While these approaches
are obviously important components of systems biology, the
physiological (physiomic) and clinical aspects of the disease
are often not emphasized. Indeed, the IUPS human physiome
project was specifically designed to model the human body
through advanced computational modeling that integrates data
from the cell, tissue, organ, and organ system (33). This more
comprehensive approach has been dubbed the “renaissance of
physiology,” as traditional physiological approaches must now
J Appl Physiol • VOL
incorporate data and principles from the many fields of “-omics” that are used in the systems biology approach (58).
Applying this integrative strategy to medicine in general defines what has become known as “systems medicine” (7). The
purpose of the present paper is to describe a systems biology
approach to lung disease that incorporates not only genomic
information but also physiological and clinical data that contribute to the phenotypic expressions of disease.
SYSTEMS BIOLOGY APPROACH USING HIGH-THROUGHPUT
DATA
The key steps involved in a systems biology approach
include data acquisition, model construction, disease simulation, model perturbation, theory formation and prediction, and
experimental validation of theory (37). According to Wu and
Kaminski, however, “ . . . the majority of the studies in
chronic lung diseases is still in the first step of systems biology
research” (70); i.e., data acquisition. Most data acquisition
occurs through the utilization of high-throughput methods
involving microarrays of gene expression or more recently
RNA sequencing techniques. These data reveal information
110 • JUNE 2011 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.33.4 on September 18, 2016
Fig. 1. Overview of a systems biology approach to lung disease, with emphasis on the different scales of investigation. Starting with the whole person (A), lung
disease can be understood through clinical symptoms as well as various tools such as imaging (B), physiology (C), and analysis of exhaled breath (D). At the
whole organ level (E), the lung can be examined at the level of tissues (F), cells (G), and cellular and molecular components (H). The lung can finally be analyzed
at the level of gene arrays (I) and the DNA itself (J). Ultimately, the entire process relates back to the human condition (A), but only if the data sets from each
of the different scales are integrated.
Review
1718
SYSTEMS BIOLOGY IN LUNG DISEASE
about the complex composition of mRNAs expressed in the
tissue or cell of interest. The systems approach to analyzing
these data is to look for transcriptional profiles of different
phenotypes of tissue in order to associate these profiles with the
phenotype. Therefore, this strategy is essentially unbiased,
meant to discover previously unknown associations and
thereby generate novel hypotheses. This approach has been
taken for a number of lung diseases, including idiopathic
pulmonary fibrosis (IPF), asthma, COPD, and pulmonary hypertension (70). As an example, a systems approach has lead to
important new insights in IPF (70). Traditionally thought of as
an inflammatory disease, clinicians know far too well that
anti-inflammatory therapy has little effect. Microarrays have
shown, however, that genes involved in wound healing and
excessive deposition of extracellular matrix are upregulated in
this disease. In particular, these gene profiles are distinct from
the proinflammatory genes that are active in a disease like
hypersensitivity pneumonitis, which clearly involves inflammation and responds to anti-inflammatory therapy (70). Similarly, microarrays of tissue from patients with asthma reveal
upregulation of genes involved in the immune response, such
as antigen processing, cytokine production, and T-cell regulation (70). In COPD, genes involved in stress-related activation
have been identified (70), and recently, genes associated with
degradation around small airways have been found (27). Functional genomic profiling has been used extensively in the area
of non-small cell lung cancer. A recent review (56) highlights
many of the findings in this area, among which has been the
recognition of alterations in epidermal growth factor receptor
(EGFR) and subsequent development of targeted therapy with
tyrosine kinase inhibitors. Recently, Goto and colleagues (28)
discovered a unique epigenetic profile of genes in malignant
pleural mesothelioma, which was distinct enough that it may
serve as a diagnostic marker of disease. Other examples include the upregulation of genes involved in inflammation and
stress response in pulmonary hypertension (13), and genes
J Appl Physiol • VOL
associated with various aspects of lung structure and inflammation associated with susceptibility to acute lung injury (20).
Thus high-throughput analyses of gene expression have
yielded unique insights into the patterns of biological function
associated with distinct lung diseases.
Regulation of posttranscriptional gene expression is also
governed by microRNAs (miRNA). Since these molecules will
alter ultimate gene expression, molecular methods to characterize patterns of miRNAs are also involved in a systems
approach to lung disease. For example, Tan and colleagues
(60) have shown that miRNAs are specifically targeted to
HLA-G in patients at risk of asthma. Other aspects of epigenetics may also be involved in determining risk and severity of
asthma and therefore be useful in classifying phenotype and
designing targeted therapies (32).
The data from microarrays can also be used in genomewide
association studies (31), where a common approach is to search
for single-nucleotide polymorphisms (SNPs) within a population. This approach is particularly used to elucidate pharmacogenetic factors associated with treatment responses, such as,
in asthma, the bronchodilator response (17, 69) or the response
to anti-leukotriene therapy (46). However, focusing on all the
polymorphisms associated with a functional phenotype, rather
than just isolated ones, is a more powerful method that a
systems biology approach can bring to genomewide association studies (66). A recent example of this approach was
described by Moffatt and colleagues (44), who demonstrated
that a number of polymorphisms are associated with asthma
and, as a whole, implicate an interactive role between epithelial
damage, adaptive immunity, and airway inflammation in the
pathogenesis of the disease.
Protein expression profiles, or proteomics, are also used in a
systems biology approach to lung disease. Recent examples
include IPF, asthma, and pulmonary hypertension (70). Since
proteins are measurable products in blood and other body
fluids, this approach is key to searching for relevant biomarkers
110 • JUNE 2011 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.33.4 on September 18, 2016
Fig. 2. The integrative process of systems biology. The systems biology approach takes an
unresolved disease problem and analyzes it at
the many levels at which it exists as a complex,
integrated network, from the molecular to the
microscopic to the whole organ and clinical
realms. These complex sets of data are integrated into a conceptual, computational model
used to explain the various aspects of the disease. Experiments are performed to perturb the
system in order to analyze the response and
further refine the model based on the data.
Ultimately a best-fit model is generated that
serves to not only explain phenomena but also
generate new hypotheses. The entire process is
highly complex and requires rigorous standards
of quality assurance, secure data storage and
integrity, and open source access for scrutiny
and use by the entire scientific community.
Review
SYSTEMS BIOLOGY IN LUNG DISEASE
SYSTEMS BIOLOGY APPROACH THROUGH LUNG
MECHANICAL BEHAVIOR: TIME SERIES
An essential characteristic of biological systems is their temporal behavior. In-built feedback loops are creating a dynamic
homeostasis in physiology, which is better labeled as “homeokinesis” (40). Thus far the dynamics of the respiratory system have
not been studied using molecular profiles, but there is increasing
evidence of highly characteristic dynamic behavior in relation to
disease from measures of respiratory microscopic structure and in
vitro and in vivo airway function (42).
As the lung is a highly mechanical structure, it makes sense
that searching for emergent mechanical properties would be an
important systems biology strategy in lung disease. This is
especially true in a disease like asthma that is characterized by
temporal variability, which might explain why the reductionist
approach to understanding this disease has been limited. A
recent example of the failure of the reductionist approach in
lung disease is found in the work by LaPrad and colleagues
(38), who showed that, contrary to expectations, oscillations of
an airway smooth muscle while intact within the airway wall
did not attenuate its ability to constrict in response to acetylcholine, unlike the behavior of isolated smooth muscle strips.
As pointed out by Bates, “ . . . we still cannot tell a priori how
behavior will cross length scales in biological systems” (8).
Bates and colleagues (9) have recently published an elegant
example of how emergent properties govern the behavior of
biological networks. Using an elastic network model of the
lung parenchyma, these authors were able to demonstrate the
J Appl Physiol • VOL
process of percolation, an emergent property known to be
involved in many natural processes, such as forest fires and
infectious disease epidemics. Percolation describes the process
by which physical events are transmitted across a network of
elements. In the lung, these physical events might include the
laying down of collagen in the case of IPF, or the destruction
of the lung parenchyma in the case of emphysema. Initially,
these events might progress in a gradual manner without much
overall change in the bulk mechanical properties of the lung.
However, at some point, known as the percolation threshold,
these events link together across the entire lung to abruptly
change its bulk mechanical properties. Clinically, this process
may be manifested as gradual progression of disease, followed
by acceleration of the disease process when a percolation
threshold is reached. Indeed, in IPF an initial gradual progression of disease may be followed by an acute exacerbation of
IPF (35). Similarly, in emphysema, radiographic evidence of
disease may exist before significant changes in lung function
occur (49), but progressive emphysema and accelerated loss of
lung function typically follow after a period of time (19, 43). In
both cases, radiographic and pathologic data support the concept of percolation as it is predicted to occur in these two
diseases (8). Although greatly simplified, this modeling concept may have important clinical implications for the way in
which progression of lung disease occurs and how it can be
monitored by measuring various aspects of structure by imaging or biopsy.
Another example of applying a systems biology approach to
lung disease is in predicting asthma exacerbations based on
underlying fluctuation of airway function. Asthma is a complex, chronic disease (5, 64). The complexity of asthma arises
from the nonlinear interaction of inflammatory, immunological, physiological, mechanical, and environmental factors, all
within a context of genetic background and unique, personality
traits and behavior (30, 63). It is no wonder that no one
molecule, and no one therapy, can “cure” asthma or apply
equally across individuals. Likewise, it is difficult to isolate
clinical factors that may predict poor asthma control (41).
However, asthma can be approached as any complex system by
using a systems biology approach to understand the interaction
of these many factors (6, 21).
One of the manifestations of these complex interactions is
the fractal nature of lung function, which no doubt arises, in
part, from the fractal structure of the airway tree (10). Such a
structure is characterized by anatomic similarities at different
length and time scales, a so-called scale-free network (4). A
unique feature of such a network is that, while it remains robust
to small perturbations, it can suddenly decompensate with an
“avalanche” of heterogeneous airway narrowing that leads to
reductions in airway function and a loss of asthma control (4).
However, because the fractal nature of lung function is scale
free, it can be statistically analyzed at different time scales,
such that previous events contain information that leads to
prediction of future events (22, 23).
One technique that applies this concept is long-range correlation analysis (detrended fluctuation analysis), wherein complex systems with fractal structure show correlations during
long time periods (22, 23). Using this method, Frey and
colleagues examined the day-to-day fluctuations in lung function (peak expiratory flow, PEF) in patients with chronic
asthma to determine whether bronchodilator therapy affected
110 • JUNE 2011 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.33.4 on September 18, 2016
of disease. Notably, serum protein expression profiles have
recently been shown to be highly successful in discriminating
the chronic airway diseases asthma, COPD and cystic fibrosis
(26). In addition, the analysis of global metabolic profiles,
so-called metabolomics, of substances in exhaled breath condensate by NMR spectroscopy, reveals airway biochemical
“fingerprints” (14) of oxidized and acetylated compounds in
patients with asthma. Furthermore, “breathprints” of volatile
organic compounds by electronic noses and/or mass spectrometry can distinguish between patients with asthma and those
with COPD (18), and between patients with lung cancer and
patients with other malignancies (51). Even though these
findings require external validation according to the STARD
Guidelines (Standards for the Reporting of Diagnostic Accuracy Studies, www.stard-statement.org), they do suggest that
proteomic and metabolomic assessments can be extremely
helpful in discriminating patients with lung diseases.
While all of these “-omic” approaches to understanding lung
disease are necessary, they are not sufficient. Only through
integration of this information through unbiased statistical
methods can we derive a complex systems view of lung disease
and produce insights that will enhance our understanding of
disease classification and pathophysiology (39). For example,
from mouse models of asthma, a “functional and regulatory
map” constructed using a synthesis of microarray data sets has
revealed distinct patterns of response to treatment and led to
new insight into the role of IL-13 and transforming growth
factor-beta (TGF-␤) as key regulators of asthma (48). However, at this time, further integration of molecular “-omics”
data into a comprehensive, systems view of human lung
disease is still in its infancy (70).
1719
Review
1720
SYSTEMS BIOLOGY IN LUNG DISEASE
SYSTEMS BIOLOGY APPROACH THROUGH CLINICAL
PHENOTYPES
Another scale at which to apply a systems biology approach
to lung disease is at the clinical level (6, 7). Examining
differing patterns of clinical expression of disease can be done
by complex statistical approaches such as factor analysis (55)
or cluster analysis (45). For example, cluster analysis is a
mathematical method that quantifies the similarity between
individuals in a population based on specified variables, and
groups individuals together such that individuals within the
same cluster are highly similar, and individuals in different
clusters are very different. Two recent examples applied to
asthma illustrate this principle. In the study by Haldar and
colleagues (30), 371 patients with asthma were found to group
into five phenotypic types that corresponded to specific responses to therapy. These included patients characterized by
varying degrees of eosinophilic inflammation, symptoms, and
body mass index. Similarly, Moore and colleagues (45) found
five clusters of asthma phenotypes that were distinguished by
lung function, atopy, age of onset, and use of corticosteroids.
Cluster analysis has also been applied to 175 random subjects with respiratory symptoms or airflow obstruction by
spirometry and, interestingly, also found five distinct phenotypes, in this case related to variability of airflow obstruction,
atopy, eosinophilic inflammation, emphysema, and chronic
bronchitis (63). A similar method, latent class analysis, was
able to distinguish five phenotypes of children with wheeze and
cough (57). In COPD, Paoletti and colleagues (50) recently
J Appl Physiol • VOL
applied multiple advanced methods to patients and found
phenotype clusters corresponding to the classic types involving
bronchitis vs. emphysema. Hurst and colleagues (34) demonstrated that exacerbations of COPD cluster in time, revealing a
high-risk period for recurrent exacerbation in the 8-wk period
following the initial exacerbation.
Why should we reorganize our thinking about the phenotypic expression of lung disease? The answer lies in the
recognition that traditional classifications of phenotype do not
appear to relate to mechanisms of disease or to prognostic
information or therapeutic outcomes. For example, using FEV1
to classify severity of disease in either COPD or asthma is
inadequate, as shown by the wide distribution of symptoms and
comorbidities across GOLD stages in the ECLIPSE cohort of
patients with COPD (2), or the variable distribution of clinical
attributes across degrees of lung function abnormality in a
cohort of patients with severe asthma (45). Indeed, in many
cases, the use of FEV1 and other traditional clinical and
physiological features fails to even distinguish asthma from
COPD (24). This inability to better phenotype patients may be
one of the reasons that clinical trials fail to show improvement
in unselected groups of patients, but can and do demonstrate
efficacy in specific subgroups. Identifying clinical phenotypes
based on statistical methods such as cluster analysis and factor
analysis will allow us to reclassify and improve our understanding of the clinical presentations of asthma and COPD,
well beyond the traditional subjective categories of intrinsic or
extrinsic asthma, or pink puffers and blue bloaters. This approach to classification can then be related to specific molecular mechanisms suggested by analysis of biospecimens, and
functional abnormalities assessed by physiological changes.
An example of this approach was recently taken using bronchoalveolar lavage specimens from patients with asthma and
found that panels of cytokines could be grouped by cluster
analysis into four molecular phenotypes, one of which strongly
corresponded to patients with severe asthma (11). Using systems methods, continued iteration will be necessary to further
refine phenotypes based on molecular, cellular, and physiological profiles that relate to significant clinical outcomes. This
strategy has been successful in some areas, such as brain
cancer (29), where genetic profiling has improved classification of gliomas based on molecular subtypes that correspond to
survival better than histological classification.
This systems biology approach will undoubtedly also generate novel hypotheses that can then be further tested in clinical
trials. By isolating more clinically relevant phenotypes based
on underlying disease mechanism, the systems approach may
radically change the way in which these clinical trials are
performed. In order to deal with inherent patient heterogeneity,
phase III drug trials currently must be highly selective in their
inclusion criteria, and yet still involve large numbers of patients treated and monitored over long periods of time. Thus
many patients may be excluded from study, and these trials are
costly and time consuming. A better strategy is to keep initial
clinical trials large and simple, allowing broad inclusion of
subjects and, through careful phenotyping, isolation of distinct
patient subtypes that respond to the intervention being tested.
Subsequent, more definitive trials can then focus on specific
patient phenotypes and monitor changes in relevant biomarkers
and other outcomes via fast, high-throughput methods, allow-
110 • JUNE 2011 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.33.4 on September 18, 2016
the correlation, and whether future risk of exacerbation could
be predicted from fluctuations in PEF. The authors found that,
based on computer modeling, the regular use of short-acting
beta-agonists, but not long-acting beta-agonists, would result in
a more random, and hence less predictable, fluctuation in lung
function and an increased risk of future exacerbations (22, 23).
In addition, they found that their model allowed baseline
fluctuation in PEF to predict the probability of future exacerbations. Such computer modeling was applied in a subsequent
study (61) that demonstrated the utility of baseline fluctuation
in PEF to predict future asthma symptoms and treatment
response to beta-agonists. These studies certainly shed new
light on the importance of monitoring of asthma by PEF,
which, despite recommendations by international guidelines
(25), is notoriously underutilized (62). Interestingly, a recent
study involving user-friendly, electronic monitoring in the
context of strong clinical support was able to demonstrate
markedly improved adherence to PEF monitoring by patients
with asthma (53). Such electronic devices could potentially be
adapted to collect and analyze time series data.
Another view on temporal behavior of the lung is provided
by the short-term fluctuations of respiratory impedance over
minutes, rather than days, as measured by the forced oscillation
technique. Even though the first fluctuation analyses of respiratory impedance were not consistent in discriminating asthmatics from controls (16, 52), a recent analysis using a nonlinear, distance-based deterministic model (attractor analysis)
did show adequate distinction between patients with asthma
and COPD (47). This suggests that the dynamics of the respiratory system provides relevant information in relation to
airways disease.
Review
SYSTEMS BIOLOGY IN LUNG DISEASE
ing the sample size and length of such trials to be reduced, and
bringing to practice sooner new therapeutic strategies (54).
CONCLUSIONS
ACKNOWLEDGMENTS
The authors thank Gary J. Nelson for creating the illustration in Fig. 1.
DISCLOSURES
No conflicts of interest, financial or otherwise, are declared by the author(s).
REFERENCES
1. Aderem A. Systems biology: its practice and challenges. Cell 121:
511–513, 2005.
J Appl Physiol • VOL
2. Agusti A, Calverley P, Celli B, Coxson H, Edwards L, Lomas D,
MacNee W, Miller B, Rennard S, Silverman E, Tal-Singer R, Wouters
E, Yates J, Vestbo J. Characterization of COPD heterogeneity in the
ECLIPSE cohort. Respir Res 10: 122, 2010.
3. Ahn A, Tewari M, Poon CS, Phillips R. The clinical applications of a
systems approach. PLoS Med 3: 956 –960, 2006.
4. Albert R. Scale-free networks in cell biology. J Cell Sci 118: 4947–4957,
2005.
5. Anderson G. Endotyping asthma: new insights into key pathogenic
mechanisms in a complex, heterogeneous disease. Lancet 372: 1107–
1119, 2008.
6. Auffray C, Adcock I, Chung K, Djukanovic R, Pison C, Sterk P. An
integrative systems biology approach for understanding of pulmonary
diseases. Chest 137: 1410 –1416, 2010.
7. Auffray C, Chen Z, Hood L. Systems medicine: the future of medical
genomics and healthcare. Genome Med 1: 2.1–2.11, 2009.
8. Bates J. The multiscale manifestations of airway smooth muscle contraction in the lung. J Appl Physiol 109: 269 –270, 2010.
9. Bates J, Davis G, Majumdar A, Butnor K, Suki B. Linking parenchymal disease progression to changes in lung mechanical function by
percolation. Am J Respir Crit Care Med 176: 617–623, 2007.
10. Boser S, Park H, Perry S, Menache M, Green FH. Fractal geometry of
airway remodeling in human asthma. Am J Respir Crit Care Med 172:
817–823, 2005.
11. Brasier A, Victor S, Boetticher G, Ju H, Lee C, Bleecker E, Castro M,
Busse W, Calhoun W. Molecular phenotyping of severe asthma using
pattern recognition of bronchoalveolar lavage-derived cytokines. J Allergy
Clin Immunol 121: 30 –37, e6, 2008.
12. Broadhurst D, Kell D. Statistical strategies for avoiding false discoveries
in metabolomics and related experiments. Metabolomics 2: 171–196,
2006.
13. Bull T, Coldren C, Geraci M, Voelkel N. Gene expression profiling in
pulmonary hypertension. Proc Am Thorac Soc 4: 117–120, 2007.
14. Carraro S, Rezzi S, Reniero F, Keberger K, Giordano G, Zanconato
S, Guillou C, Baraldi E. Metabolomics applied to exhaled breath condensate in childhood asthma. Am J Respir Crit Care Med 175: 986 –990,
2007.
15. Clermont G, Auffray C, Moreau Y, Rocke D, Dalevi D, Dubhashi D,
Marshall D, Raasch P, Dehne F, Provero P, Tegner J, Aronow B,
Langston M, Benson M. Bridging the gap between systems biology and
medicine. Genome Med 1: 88.1–88.6, 2009.
16. Diba C, Salome C, Reddel H, Thorpe C, Toelle B, King G. Short-term
variability of airway caliber—a marker of asthma? J Appl Physiol 103:
296 –304, 2007.
17. Duan Q, Gaume B, Hawkins G, Himes B, Bleecker ER, Kanderman B,
Irvin C, Peters SP, Meyers DA, Hanrahan J, Lima J, Litonjua A,
Tantisira K, Liggett S. Regulatory haplotypes in ARG1 are associated
with altered bronchodilator response. Am J Respir Crit Care Med 183:
449 –454, 2011.
18. Fens N, Zwinderman A, van der Schee M, de Nijs S, Dijkers E,
Roldaan A, Cheung D, Bel E, Sterk P. Exhaled breath profiling enables
discrimination of chronic obstructive pulmonary disease and asthma. Am
J Respir Crit Care Med 180: 1076 –1082, 2009.
19. Fletcher C, Peto R. The natural history of chronic airflow obstruction.
BMJ 1: 1645–1648, 1977.
20. Frank J, Erle D. Progress toward a systems biology approach to acute
lung injury. Am J Physiol Lung Cell Mol Physiol 293: L290 –L291, 2007.
21. Frey U. Asthma as a non-linear complex dynamic system: a novel
approach to understand temporal behavior of chronic asthma and its
response to beta-agonists. Eur Respir J 17: 67–69, 2008.
22. Frey U, Brodbeck T, Majumdar A, Taylor D, Town G, Silverman M,
Suki B. Risk of severe asthma episodes predicted from fluctuation analysis
of airway function. Nature 438: 667–670, 2005.
23. Frey U, Suki B. Complexity of chronic asthma and chronic obstructive
pulmonary disease: implications for risk assessment, and disease progression and control. Lancet 372: 1088 –1099, 2008.
24. Gibson P, Simpson J. The overlap syndrome of asthma and COPD: what
are its features and how important is it? Thorax 64: 728 –735, 2009.
25. Global Initiative for Asthma. www.ginasthma.org, 2009.
26. Gomes-Alves P, Imrie M, Gray R, Nogueira P, Ciordia S, Pachedo P,
Azevedo P, Lopes C, de Almeida A, Guardiano M, Porteous D, Albar
J, Boyd A, Penque D. SELDI-TOF biomarker signatures for cystic
fibrosis, asthma and chronic obstructive pulmonary disease. Clin Biochem
43: 168 –177, 2010.
110 • JUNE 2011 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.33.4 on September 18, 2016
Applying a systems biology approach to lung disease makes
sense, because the lung is a highly complex biological organ,
which will easily evade understanding by using only a traditional reductionist research approach. This has been revealed
repeatedly by the many failed clinical attempts at treating
asthma, COPD, acute lung injury, and other diseases by targeting only one molecule or mediator. However, application of
systems biology to lung disease is truly an “emerging” field.
Progress has been made in acquiring high-throughput data on
molecular and genetic profiles of disease, in measuring and
developing analytic methods to understand functional dynamics, and in applying unbiased statistical methods to discover
and redefine clinical phenotypes of disease. The challenge
remains in integrating this information into a comprehensive
model that can be tested and refined, and ultimately related to
clinical outcomes. This process will lead to better diagnosis
and individualized treatment. In fact, the promise of systems
biology is a new approach to medicine that has been described
as “predictive, personalized, preventive and participatory”
(65). It will be predictive in the sense of allowing better
prognostic information based on a more individual understanding of disease; personalized because individual phenotypes will
be better characterized and related to treatment outcomes;
preventive because of the predictive nature and subsequent
better understanding of mechanism and consequent intervention to prevent disease; and participatory because in the true
sense of a complex network, individual behavior, stress and
other lifestyle factors, as well as more global public health
issues, will also play a role in disease pathogenesis (7). Of
course, as with any new paradigm, barriers to successful
application of the systems approach must be addressed. These
include cost-effective methods to gather large bodies of data,
and careful techniques to analyze these data in order to avoid
false discovery that may contribute to confusion, not understanding (3, 12). Integrating systems biology into medical
research and practice will require sufficient funding, infrastructure, academic recognition, and facilitation of the transdisciplinary approach required (15). A major issue will be one of
effective communication among scientists, clinicians, and policy makers across many disciplines, both in academia and in
industry (7). Common, open-access databases, analytic tools,
and tailored knowledge management systems, including the
language of systems biology with pathway and network analyses, are already in use and under further development in an
effort to facilitate the growth and development of the systems
biology approach in medicine (7).
1721
Review
1722
SYSTEMS BIOLOGY IN LUNG DISEASE
J Appl Physiol • VOL
48. Novershtern N, Itzhaki Z, Manor O, Friedman N, Kaminski N. A
functional and regulatory map of asthma. Am J Respir Cell Mol Biol 38:
324 –336, 2008.
49. Omori H, Fujimoto K, Katoh T. Computed tomography findings of
emphysema: correlation with spirometric values. Curr Opin Pulm Med 14:
110 –114, 2008.
50. Paolleti M, Camiciottoli G, Meoni E, Bigazzi F, Cestelli L, Pistolesi M,
Marchesi C. Explorative data analysis techniques and unsupervised clustering methods to support clinical assessment of chronic obstructive
pulmonary disease (COPD) phenotypes. J Biomed Inform 42: 1013–1021,
2009.
51. Peng G, Hakim M, Broza Y, Billan S, Abdah-Bortnyak R, Kuten A,
Tisch U, Haick H. Detection of lung, breast, colorectal and prostate
cancers from exhaled breath using a single array of nanosensors. Br J
Cancer 103: 542–551, 2010.
52. Que C, Kenyon C, Olivenstein R, Macklem P, Maksym G. Homeokinesis and short-term variability of human airway caliber. J Appl Physiol
91: 1131–1141, 2001.
53. Reddel H, Toelle B, Marks G, Ware S, Jenkins C, Woolcock A.
Analysis of adherence to peak flow monitoring when recording of data is
electronic. BMJ 324: 146 –147, 2002.
54. Richens J, Urbanowicz R, Lunt E, Metcalf R, Corne J, Fairclough L,
O’Shea P. Systems biology coupled with label-free high-throughput
detection as a novel approach for diagnosis of chronic obstructive pulmonary disease. Respir Res 10: 29, 2009.
55. Riekert K, Eakin M. Factor analysis: a primer for asthma researchers. J
Allergy Clin Immunol 121: 1181–1183, 2008.
56. Singhal S, Miller D, Ramalingam S, Sun S. Gene expression profiling of
non-small cell lung cancer. Lung Cancer 60: 313–324, 2008.
57. Spycher B, Silverman M, Brooke A, Minder C, Kuehni C. Distinguishing phenotypes of childhood wheeze and cough using latent class analysis.
Eur Respir J 31: 974 –981, 2008.
58. Strange K. The end of “naive reductionism”: rise of systems biology or
renaissance of physiology? Am J Physiol Cell Physiol 288: C968 –C974,
2005.
59. Studer S, Kaminski N. Towards systems biology of human pulmonary
fibrosis. Am J Respir Crit Care Med 4: 85–91, 2007.
60. Tan Z, Randall G, Fan J, Camoretti-Mercado B, Brockman-Schneider R, Pan L, Solway J, Gern J, Lemanske R, Nicolae D, Ober C.
Allele-specific targeting of microRNAs to HLA-G and risk of asthma. Am
J Hum Genet 81: 829 –834, 2007.
61. Thamrin C, Stern G, Strippoli M, Kuehni C, Suki B, Taylor D, Frey
U. Fluctuation analysis of lung function as a predictor of long-term
response to beta-agonists. Eur Respir J 33: 486 –493, 2009.
62. Verschelden P, Cartier A, L’Archeveque J, Trudeau C, Malo JL.
Compliance with and accuracy of daily self-assessment of peak expiratory
flows (PEF) in asthmatic subjects over a three month period. Eur Respir J
9: 880 –885, 1996.
63. Weatherall M, Travers J, Shirtcliffe P, Marsh S, Williams M, Nowitz
M, Aldington S, Beasley R. Distinct clinical phenotypes of airways
disease defined by cluster analysis. Eur Respir J 34: 812–818, 2009.
64. Wenzel S. Asthma: defining the persistent adult phenotypes. Lancet 368:
804 –813, 2006.
65. Weston A, Hood L. Systems biology, proteomics, and the future of health
care: toward predictive, preventative, and personalized medicine. J Proteome Res 3: 179 –196, 2004.
66. Whitcomb D, Aoun E, Vodovotz Y, Clermont G, Barmada M. Evaluating disorders with a complex genetic basis. Dig Dis Sci 50: 2195–2202,
2005.
67. Wise R. The value of the forced expiratory volume in 1 second decline in
the assessment of chronic obstructive pulmonary disease progression. Am
J Med 119: 4 –11, 2006.
68. Wright J, Christman J. The role of nuclear factor kappa B in the
pathogenesis of pulmonary diseases: implications for therapy. Am J Respir
Med 2: 211–219, 2003.
69. Wu A, Himes B, Lasky-Su J, Litonjua A, Li L, Lange C, Lima J, Irvin
C, Weiss S. Development of a pharmacogenetic predictive test in asthma:
proof of concept. Pharmacogenet Genomics 20: 86 –93, 2010.
70. Wu W, Kaminski N. Chronic lung diseases. WIREs Syst Biol Med 1:
298 –308, 2009.
110 • JUNE 2011 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.33.4 on September 18, 2016
27. Gosselink J, Hayashi S, Elliot W, Xing L, Chan B, Yang L, Wright C,
Sin D, Pare P, Pierce J, Pierce R, Patterson A, Cooper J, Hogg J.
Differential expression of tissue repair genes in the pathogenesis of
chronic obstructive pulmonary disease. Am J Respir Crit Care Med 181:
1329 –1335, 2010.
28. Goto Y, Shinjo K, Kondo Y, Shen L, Toyota M, Suzuki H, Gao W, An
B, Fujii M, Murakami H, Osada H, Taniguchi T, Usami N, Kondo M,
Hasegawa Y, Shimokata K, Matsuo K, Hida T, Fujimoto N,
Kishimoto T, Issa J, Sekido Y. Epigenetic profiles distinguish malignant
pleural mesothelioma from lung adenocarcinoma. Cancer Res 69: 9073–
9082, 2009.
29. Gravendeel L, Kouwenhoven M, Gevaert O, de Rooi J, Stubbs A,
Duijm J, Daemen A, Bleeker F, Bralten L, Kloosterhof N, De Moor B,
Eilers P, van der Spek P, Kros J, Sillevis Smitt P, van den Bent M,
French P. Intrinsic gene expression profiles of gliomas are a better
predictor of survival than histology. Cancer Res 69: 9065–9072, 2009.
30. Haldar P, Pavord I, Shaw D, Berry M, Thomas M, Brightling C,
Wardlaw A, Green R. Cluster analysis and clinical asthma phenotypes.
Am J Respir Crit Care Med 178: 218 –224, 2008.
31. Hardy J, Singleton A. Genomewide association studies and human
disease. N Engl J Med 360: 1759 –1768, 2009.
32. Ho S. Environmental epigenetics of asthma: an update. J Allergy Clin
Immunol 126: 453–465, 2010.
33. Hunter P, Robbins P, Noble D. The IUPS human physiome project.
Pflügers Arch 445: 1–9, 2002.
34. Hurst J, Donaldson G, Quint J, Goldring J, Baghai-Ravary R, Wedzicha J. Temporal clustering of exacerbations of chronic obstructive pulmonary disease. Am J Respir Crit Care Med 179: 369 –374, 2009.
35. Hyzy R, Huang S, Myers J, Flaherty K, Martinez F. Acute exacerbations of idiopathic pulmonary fibrosis. Chest 132: 1652–1658, 2007.
36. Kirschner M. The meaning of systems biology. Cell 121: 503–504, 2005.
37. Kitano H. Systems biology: a brief overview. Science 295: 1662–1664,
2002.
38. LaPrad A, Szabo T, Suki B, Lutchen K. Tidal stretches do not modulate
responsiveness of intact airways in-vitro. J Appl Physiol 109: 295–304,
2010.
39. Loscalzo J, Kohane I, Barabasi AL. Human disease classification in the
postgenomic era: a complex systems approach to human pathobiology.
Mol Syst Biol 3: 124, 2007.
40. Macklem P. Emergent phenomena and the secrets of life. J Appl Physiol
104: 1844 –1846, 2008.
41. McCoy K, Shade D, Irvin C, Mastronade J, Hanania N, Castro M,
Anthonisen N. Predicting episodes of poor asthma control in treated
patients with asthma. J Allergy Clin Immunol 118: 1226 –1233, 2006.
42. Mertens M, Tabuchi A, Meissner S, Krueger A, Schirrmann K,
Kertzsxher U, Pries A, Slutsky A, Koch E, Kuebler W. Alveolar
dynamics in acute lung injury: hetergeneous distension rather than cyclic
opening and collapse. Crit Care Med 37: 2604 –2611, 2009.
43. Miller M, Cho J, Pham A, Friedman P, Ramsdell J, Brodie D.
Persistent airway inflammation and emphysema progression on CT in
ex-smokers observed for 4 years. Chest 21 October, 2010 [EPub ahead of
print].
44. Moffatt M, Gut I, Demenais F, Strachan D, Bouzigon E, Heath S, von
Mutius E, Farrall M, Lathrop G, Cookson W. A large-scale, consortium-based genomewide association study of asthma. N Engl J Med 363:
1211–1221, 2010.
45. Moore W, Meyers D, Wenzel S, Teague W, Li H, Li X, D’Agostino R,
Castro M, Curran-Everett D, Fitzpatrick A, Gaston B, Jarjour N,
Sorkness R, Calhoun W, Chung K, Comhair S, Dweik R, Israel E,
Peters S, Busse W, Erzurum S, Bleecker E. Identification of asthma
phenotypes using cluster analysis in the Severe Asthma Research Program. Am J Respir Crit Care Med 181: 315–323, 2010.
46. Mougey E, Feng H, Castro M, Irvin C, Lima J. Absorption of montelukast is transporter mediated: a common variant of OATP2B1 is associated with reduced plasma concentrations and poor response. Pharmacogenet Genomics 19: 129 –138, 2009.
47. Muskulus M, Slats A, Sterk P, Verduyn-Lunel S. Fluctuations and
determinism of respiratory impedance in asthma and chronic obstructive
pulmonary disease. J Appl Physiol 109: 1582–1591, 2010.