Article Text


Chronic obstructive pulmonary disease • 3: Experimental animal models of pulmonary emphysema
  1. R Mahadeva1,2,
  2. S D Shapiro
  1. 1Respiratory Medicine Unit, Department of Medicine, University of Cambridge, Cambridge Institute for Medical Research, Cambridge CB2 2XY, UK
  2. 2Division of Pulmonary and Critical Care, Department of Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, 02116, USA
  1. Dr R Mahadeva, Respiratory Medicine Unit, Department of Medicine, University of Cambridge, Cambridge Institute for Medical Research, Cambridge CB2 2XY, UK;


The use of genetically manipulated mice together with traditional animal studies are steadily increasing our knowledge of the factors important in determining alveolar formation and destruction in emphysema. A review of the animal models used to study emphysema is presented.

  • pulmonary emphysema
  • chronic obstructive pulmonary disease animal models

Statistics from

Chronic obstructive pulmonary disease (COPD) is one of the commonest reasons for ill health worldwide, with 16 million individuals affected in the USA alone. It is ranked as the fourth and fifth highest cause of death in the USA and UK, respectively, with a mortality rate at least 14 times that of asthma.1,2 The single most important factor in the development of emphysema is cigarette smoke. Inhalation of cigarette smoke causes a chronic pulmonary inflammatory infiltrate of macrophages, neutrophils and CD8+ cells that persists long after smoking cessation. In susceptible individuals this ultimately leads to irreversible destruction and dilatation of the terminal airspaces of the lung, chronic disability due to respiratory failure, and premature death.

Animal studies have been critical in shaping contemporary views regarding the pathogenesis of emphysema. Almost 40 years ago Gross reported that intratracheal instillation of the plant proteinase papain into rats resulted in emphysema. This, combined with the clinical observation by Eriksson that deficiency of the antiproteinase α1-antitrypsin was associated with early onset panlobular emphysema in humans,3,4 fuelled the concept that an imbalance within the lung favouring proteinases over their inhibitors resulted in emphysema. As more sophisticated animal models have evolved, they have continued to develop fresh insights into lung biology and enabled the testing of novel treatments for emphysema. While the initial association of emphysema with α1-antitrypsin deficiency suggested that neutrophil derived proteinases were critical to the development of emphysema, recent studies—largely in animal models—have broadened the scope of cells and proteinases that may cause emphysema. Particular attention has been given to macrophage derived matrix metalloproteinases (MMPs).5 Furthermore, stemming from the study of alveogenesis in experimental animals, retinoic acid treatment has been shown to regenerate alveoli in rats both during development and following experimental emphysema.6 It is now clear that complex interacting pathways are involved in the initiation, progression, and the failure of correct repair in emphysema

While hypotheses and mechanisms of emphysema can be addressed in the test tube and cell culture, ultimately these ideas need to be tested in mammals to elucidate the complex pathophysiology of COPD. Rodents, dogs, guinea pigs, monkeys, and sheep have all been used to study emphysema. With the advent of genetic engineering in mice, this species offers the greatest ability to dissect pathogenetic pathways in mammals. In addition, more is known about its genome than any other animal and cDNA and antibody probes are abundant. They also offer the practical benefits of short breeding times, large litters, and relatively cheap housing. The main issue confounding the use of mice and other animals as a template for human disease is that we do not know how accurately they reflect human biology and pathology. Mouse and man clearly share many basic physiological processes, but the details of how gas exchange is achieved will determine how closely findings in mice can be applied to man. Each animal model should be viewed as one component of the process for studying human disease and not viewed in isolation or extrapolated directly to humans.

Emphysema can be modelled in many ways. Exogenous administration of proteinases, chemicals, particulates, and exposure to cigarette smoke result in features characteristic of human COPD. Genetic manipulation itself can result in airspace enlargement during development and throughout life. These different approaches each have their own merits and limitations and require individualised interpretation. The methods also often complement each other so their usefulness can be enhanced by using a combination of approaches to study the disease. The field has also not infrequently been stimulated by the inadvertent generation of emphysema in animals where experiments were initially undertaken for another reason.


Intrapulmonary challenge with injurious proteins, chemicals, particulates, and other compounds into lungs of animals has been used to cause emphysema directly. Compounds have also been administered that inhibit protein function (loss of function models), resulting in airspace enlargement.

A single intrapulmonary challenge with proteinases including porcine pancreatic elastase (PPE), papain, and human neutrophil elastase causes panacinar emphysema.7 Their effectiveness was directly related to their elastolytic activity, while instillation of bacterial collagenases did not cause emphysema. PPE mediated emphysema was accompanied by secretory cell metaplasia and abnormalities of pulmonary function, hypoxaemia, and right ventricular hypertrophy that are characteristic of human COPD. Following an intratracheal bolus of PPE, there is an initial loss of elastin and collagen. Over time, elastin and glycosaminoglycans return to normal and collagen is enhanced, yet intraparenchymal extracellular matrix (ECM) remains diminished and distorted and the architecture of the lung is grossly and permanently abnormal.8 Airspace enlargement develops immediately after elastolytic injury, followed by inflammation. The subsequent progression of emphysema over the next few months is probably caused by the destructive effect of host inflammatory proteinases. Although repair mechanisms fail to establish the normal lung structure, impairment of elastin and collagen cross linking with the addition of the lathyrogen β-aminopropionile (BAPN) worsens the emphysema,9 suggesting that some repair occurs. Instillation of elastases remains a useful tool, particularly to study effects downstream of proteolytic injury. Yet inflammation and host injury following instillation suggests that this model might also provide insights into upstream events such as inflammation and endogenous proteinases. However, extrapolating these findings to the slowly developing smoking induced disease in humans is not straightforward as additional mediators are likely to be involved.

Various other agents have also been used in an attempt to recreate inflammation and lung injury. For example, repeated endotoxin administration recruits neutrophils and activates macrophages with resultant airspace enlargement.10 Administration of oxidants such as nitrogen dioxide and ozone, the two major airborne pollutants, causes lung injury. Repeated long term administration of nitrogen dioxide results in mild focal emphysema, while ozone results in fibrosis.11 Administration of cadmium chloride, a constituent of cigarette smoke, also results in primarily interstitial fibrosis with tethering open of airspaces simulating emphysema. While this mechanism differs from airspace enlargement secondary to matrix destruction that characterises emphysema, we now appreciate that airway fibrosis is a significant factor in centrilobular emphysema seen in human smokers.12 Interestingly, a combination of cadmium and BAPN enhanced the emphysematous changes.13

Coal dust and silica result in focal emphysema, and animal models have uncovered complex inflammation and oxidant injury with connective tissue breakdown that was attributed to neutrophil mediated injury.14

Intravascular administration of a vascular endothelial cell growth factor receptor-2 (VEGFR-2) blocker has recently been used to generate a model of non-inflammatory emphysema.15 VEGF is a trophic factor required for endothelial cell survival and normal blood vessel development, and its absence results in endothelial cell apoptosis.16,17 There is increased septal cell death in human emphysematous lungs, which is associated with reduced lung expression of VEGF and VEGFR-2 (KDR/Flk-1).18 Chronic VEGFR-2 blockade with SU5416 caused alveolar septal cell apoptosis, pruning of the pulmonary arterial tree, and airspace enlargement. This mechanism of airspace enlargement revisits the previous “vascular” hypothesis of the disappearing alveolar cells in emphysema.19 Importantly, this also raises the concept that initial loss of epithelial cells can lead to matrix destruction and airspace enlargement, as opposed to the traditional view that inflammatory cell proteinases degrade matrix with subsequent loss of attachment and death of structural cells.


Various animals have been exposed to cigarette smoke using smoking chambers. The first unequivocal animal model of smoking related emphysema was reported in 1990 by Wright and colleagues.20 Exposure of guinea pigs to the smoke of 10 cigarettes each day, 5 days per week, for 1, 3, 6, and 12 months resulted in progressive emphysema and pulmonary function abnormalities similar to those seen in humans with cigarette smoke induced COPD. Cessation of smoke exposure stabilised—but did not reverse—airspace enlargement. In response to long term cigarette smoke exposure guinea pigs exhibit marked neutrophilia and accumulation of macrophages and CD4+ cells.21,22 Latent adenoviral infection combined with cigarette smoke exposure potentiates the emphysema in guinea pigs and, interestingly, this infection independently increased neutrophils, macrophages, and CD8+ cells.21

Mice are able to tolerate at least two cigarettes per day for a year with non-toxic carboxyhaemoglobin levels and minimal effects on body weight. Despite the fact that they are obligate nose breathers, they do not have extensive cilia and inefficiently filter tobacco smoke. Other anatomical differences from man include a paucity of submucosal glands (limited to the trachea) and fewer Clara cells. They also have less extensive airway branching and lack respiratory bronchioles. The development of emphysema in mice is dependent on the strain of mouse. Neutrophil recruitment occurs following the first cigarette and is followed by a more gradual accumulation of macrophages. Early neutrophil influx in the lungs is accompanied by a measurable increase in collagen and elastin degradation products. At this very early stage both the BAL neutrophilia and the connective tissue breakdown is preventable by pretreating the mice with an antibody to neutrophils or α1-antitrypsin.23 Epithelial changes including loss of cilia are observed after a couple of months of cigarette exposure and Clara cell hyperplasia after about 6 months (SD Shapiro, unpublished observations). Increased alveolar duct area and enlarged alveolar spaces are clearly seen after 3–6 months of exposure to cigarette smoke in susceptible strains,24 and macrophage mediated destruction appears to be prominent at these time points.


Several inbred strains of mice develop emphysema spontaneously due to genetic abnormalities (table 1). A number of mutations, many of which are not obviously related to lung structure, result in multisystem effects including abnormal airspace size. The effect of fibrillin on elastogenesis and lung development is obvious in tight skin mice, yet patients with Marfan’s syndrome do not appear to have emphysema as a major problem. While the gene defects in blotchy, pallid, and beige mice have been determined, the pathophysiological basis for abnormal lung development and increasing airspace enlargement with age is not clearly understood. Emphysema in some of these strains may in part be attributed to reduced α1-antitrypsin levels.

Table 1

Naturally occurring mice strains that develop airspace enlargement and emphysema


Transgenic and gene targeted mice provide powerful techniques to determine protein function in vivo. Gain of function models may be achieved by overexpression of proteins in transgenic mice and loss of function models achieved by targeted mutagenesis.25 Techniques to achieve germline transmission of genetic material in mice have drastically changed our ability to approach biological questions by allowing investigators to change single variables and perform controlled experiments in mammals. These models may help to determine both physiological functions of proteins as well as dissect mechanisms of disease.

Transgenic mice are derived by injecting DNA constructs—often with a tissue specific promoter driving the gene of interest—into the pronucleus of individual egg cells soon after fertilisation. Inducible transgenic systems, often using tetracycline, have been generated whereby the gene of interest may be expressed at specific times. Null mutant or gene targeted mice (“knockout mice”) are generated by electroporation and homologous recombination of a targeted construct in embryonic stem (ES) cells. One can essentially place (“knock in”) any gene under the control of any other using ES technology.


Emphysema and airspace enlargement in gene targeted mice can occur as a result of abnormal lung development or can develop spontaneously with age and following challenge with cigarette smoke (table 2).

Table 2

Examples of mouse models where “knockout” of a protein leads to the development of airspace enlargement due to a failure of alveogenesis or emphysema

Gene targeted mice with abnormal alveogenesis

A dynamic coordinated interaction between transcription factors, growth factors, extracellular matrix components, and integrin signalling pathways directs cell migration and lineage determination during lung morphogenesis.26 Lung embryogenesis begins in mice at embryonic day (E) 9.5 with the outpouching of two epithelial buds from the ventral foregut into the surrounding mesoderm. Subsequent airway development is divided into several well characterised stages culminating with alveogenesis that occurs largely by septation of the saccules from postnatal day 4 through to 14 (in the mouse).26 Developmental airspace enlargement with gene targeting is elucidating pathways of alveogenesis (table 2).

Transcription factor null mutant mice have resulted in various developmental abnormalities described in detail elsewhere.26,27 Disruption of elastic fibres would be expected to impair the structural integrity of the lung. In support of this concept, a deficiency of elastin, platelet derived growth factor A (PDGF-A), and fibulin-5 (a microfibrillar component) leads to airspace enlargement.28,29 PDGF-A null mice lack myofibroblasts which is a key source of tropoelastin and is required for alveolar septation.30,31

Members of the large family of fibroblast growth factors (FGF) are essential for several stages of mammalian lung development. Lungs of gene targeted mice lacking receptors for both FGFR-3 and FGFR-4 (but not single knockouts) have markedly impaired alveogenesis with increased synthesis of collagen.32 Retinoic acid signalling is essential for normal embryonic development and deletions of retinoic acid receptor isoforms (nuclear transcription factors) have varied effects on morphogenesis.33 Retinoic acid receptor-γ deficient mice display airspace enlargement that is potentiated by co-deletion of retinoid X receptor-α.34 Null mice for the transcription factors POD-1 and Forkhead Box F1 (Foxf1) and the transmembrane metalloprotease-disintegrin tumour necrosis factor-α converting enzyme (TACE/ADAM-17) all cause more generalised abnormalities of lung development which include impaired alveogenesis.35–37 Developmental airspace enlargement should be distinguished from emphysema defined as destruction of mature alveoli. Nevertheless, understanding the process of normal lung development will be crucial for devising strategies to restore normal lung architecture and function in emphysema.

Spontaneous emphysema in gene targeted mice

The gradual appearance of emphysema with age has been described following targeted ablation of the surfactant protein D (SP-D) gene and in mice lacking tissue inhibitor of metalloproteinase-3 (TIMP-3) (but not in null mice for TIMP-1 or 2) In both cases matrix metalloproteinases were implicated as mediators of tissue destruction. The SP-D null mice had progressive pulmonary emphysema from 3 weeks of age associated with pulmonary infiltration by lipid laden alveolar macrophages. SP-D–/– macrophages have increased reactive oxidant species activating NF-kappa-B with consequent MMP expression.38 TIMP-3–/– mice develop progressive airspace enlargement first detected at 2 weeks of age. In addition to likely developmental abnormalities, these mice display inflammation and increased MMP activity with further airspace enlargement over time.39 It is interesting to speculate that in humans subtly abnormal alveogenesis followed by years of mechanical stress may be an important mechanism of emphysema.

Emphysema in gene targeted mice following exogenous challenge

The effect of exposing developmentally normal mice lacking a specific protein to cigarette smoke allows one to probe disease pathogenesis. Macrophage elastase (MMP-12) is expressed in macrophages of human smokers and in patients with emphysema. MMP-12 null mutant mice develop normally. In contrast to wild type mice, the lung structure of MMP-12–/– mice remains normal following long term exposure to cigarette smoke. MMP-12–/– mice also failed to develop macrophage accumulation in response to cigarette smoke.5 This effect may be related to MMP-12 generation of elastin fragments that are chemotactic for monocytes.40,41 Neutrophil elastase deficient mice are partially protected against smoking induced emphysema, while gelatinase B (MMP-9) and uPA (an MMP activator) null mutants had no demonstrable effect (unpublished observations).24 MMPs can also degrade α1-antitrypsin, and neutrophil elastase can degrade TIMPs. These findings reinforce the concept that many cells and proteinases are likely to be important in the development of emphysema. Subsequent studies have supported MMP involvement in emphysema (tables 2 and 3).

Table 3

Examples of mouse models where overexpression of a protein leads to emphysema

It has recently been shown that mice deficient in both IL-1β type 1 receptor and types 1 and 2 TNF-α receptors are protected from the development of emphysema following intratracheal challenge with PPE. They exhibited reduced inflammation and increased apoptosis. However, there was no such protection in the individual null mutant strains.42


Data from transgenic mice must be interpreted cautiously. Firstly, because of the random nature of transgene integration, several founder lines should be shown to have similar phenotypes. Secondly, aberrant expression of a gene may have the capacity to cause a phenotype yet not be involved in actual disease. This is particularly important in the interpretation of airspace enlargement, since introduction of a protein not naturally associated with lung development could alter this but have no biological relevance. Use of inducible transgenic mice diminishes this concern.

Lung specific overexpression of platelet derived growth factor B (PDGF-B), transforming growth factor α, TNF-α, interleukin (IL)-11 and IL-6 all interfere with normal lung development.43–47 The impact of altering the timing of protein production has been elegantly illustrated with IL-11 where inducible overexpression showed that delaying IL-11 expression until adulthood abrogated the effects on lung structure.47

In contrast, inducible overexpression in adult mice of either the Th2 cytokine IL-13 or Th1 cytokine interferon γ (IFN-γ) produced airspace enlargement consistent with destruction of mature lung characteristic of pulmonary emphysema. With both mice transgene expression caused inflammation, activation of metallo- and cysteine proteinases. Emphysema associated with IL-13 was partially prevented by treatment with either MMP or cysteine proteinase inhibitors.48,49 The important relationship between collagen turnover and emphysema was supported by transgenic mice that constitutively overexpressed human interstitial collagenase (MMP-1) resulting in airspace enlargement.50 Use of inducible systems would strengthen interpretation of this finding being the result of purely collagen degradation in mature lung tissue.

The finding of airspace enlargement in “Klotho” mice highlights the importance of transgene integration effects. The Klotho mouse was developed to be a model of hypertension by overexpressing a sodium channel. The mice were normotensive but displayed abnormal premature aging in several organs and also had airspace enlargement. Further analysis revealed insertion of several copies of the transgene in the 5′ region of a membrane protein with sequence homology to β-glucosidase enzymes, with disruption of function of this gene termed “Klotho”. It is not settled whether the mice have abnormal development and/or premature aging, but clearly this gene affects lung structure.51


Intratracheal instillation of elastase is a relatively simple and quick method for generating emphysema, and so lends itself well to assessing the effect of any intervention on emphysema. Early studies showed that chloromethylketone elastase inhibitors, eglin-C from the medicinal leech Hirudo medicinalis, and heparin fragments (inhibits neutrophil elastase) could prevent elastase induced emphysema.7,52 Cantor et al53 have shown that a single instillation of hyaluronidase into hamster lungs given with PPE results in a decrease in airspace size compared with elastase treated controls. Since hyaluronic acid is not an elastase inhibitor, an effect on “shielding” elastin from elastases has been proposed and supported experimentally.54

More recently, the elastase model has been used to demonstrate the capacity of retinoic acid to augment septation and alveolar regeneration. It has been noted that, when septa are being formed to subdivide the primitive lung saccules, fibroblasts in the alveolar wall contain storage granules of vitamin A. Furthermore, retinoid binding proteins and nuclear retinoic acid receptor isoforms are upregulated in alveolar septal regions.55 Administration of all-trans retinoic acid (tRA) to rats abrogated PPE induced emphysema by promoting the formation of new alveoli.6 tRA also partially rescued the failure of septation seen in “tight skin” mice.56 Subsequent to this study, tRA treatment using a different protocol failed to reverse cigarette smoke induced emphysema in the guinea pig.21 tRA binds to two forms of nuclear receptors which have multiple isoforms (retinoic acid receptors (RARs) α, β and γ, and retinoid X receptors (RXRs)). RARs seem to have disparate effects. RARβ prevents septation while RARα enhances it.34,57 Thus, RARγ specific agonists might have a greater effect than the general activator RA.


Use of genetically manipulated mice in concert with traditional animal studies are steadily increasing our knowledge of the factors that are important in determining alveolar formation and destruction. Further progress will require addressing fundamental questions such as: how does smoking cause pulmonary inflammation, why do only a minority of cigarette smokers get emphysema, what are the factors that control normal lung development, and how can we restore normal alveolar tissue? We now have the exciting opportunity to combine proteomics and gene microarray technology with animal models. Therefore, rather than focusing solely on individual mediators, we can begin to dissect both the cellular interactions that lead to emphysema and the factors that govern normal alveolar development. It is hoped that this will identify novel strategies to achieve the ultimate goal of a cure for this devastating disease.


RM is a Wellcome Trust Advanced Clinical Fellow, National Institute of Heart, Lung, and Blood Disease (SDS) and Lung Biology Center Brigham and Women’s Hospital.


View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.