© 2004 American Thoracic Society doi: 10.1164/rccm.2408007
Smoke and MirrorsMouse Models as a Reflection of Human Chronic Obstructive Pulmonary DiseaseBrigham and Women's Hospital, Harvard Medical School Boston, Massachusetts In this issue of the Journal (pp. 974–980), Guerassimov and colleagues exposed five strains of mice to cigarette smoke and found impressive variability in the development of emphysema between strains (1). Questions that some readers of the Journal might be asking are: Why are mice "smoking," and why am I reading about mice in the Blue journal? Why the mouse? The mouse has been an invaluable tool to dissect disease pathways in humans. Mice are becoming the animal of choice for many types of research because: (1) like us, they are mammals; (2) we have great understanding of mouse biology and many molecular probes to study the mouse; (3) breeding is rapid; (4) genetic homology maps are available to allow translation of genetic findings in laboratory mice to the human genome; and (5) we can manipulate the murine genome eliminating ("knock-outs") or enhancing (transgenics) the expression of individual gene products or introducing specific variants within a gene of interest ("knock-ins"). Thus, controlled genetic experiments in mammals can be performed. Disadvantages of the mouse include their small size, which precludes some surgical models. Their small size also makes physiologic assessment more difficult, although much progress has been made in this regard. In fact, the phenotyping in the Guerassimov manuscript, which includes quantitative measurements of morphometry and physiology, is a major strength. The biggest criticism of the mouse, of course, is that mice are not (wo)men (note the whiskers). Mouse biologists would argue that basic biological processes are usually conserved in mammals. Overall, information derived from studies in mice is best used to guide studies in humans, many of which have been incredibly informative, including important insights into obesity and cancer (2, 3). The marked strain-to-strain variability in mice exposed to the same environmental conditions strongly suggests that genetic differences between the strains influence the differential susceptibility to develop emphysema. These results provide further evidence that genetic factors can predispose to emphysema, and more importantly, that these genetic factors are amenable to genetic dissection by careful study of the histologic and physiologic quantitative phenotypes examined by Guerassimov and colleagues. This type of detailed phenotyping is crucial in human chronic obstructive pulmonary disease (COPD) studies as well, because phenotypic heterogeneity is a potential contributor to the inconsistent results of previously reported human COPD candidate gene association studies (4). Although lung histology is not usually readily available from patients with emphysema, high-resolution chest CT scans may provide a useful surrogate. It is unclear how many genes influence the development of emphysema in these mouse strains, where they are located, and what their individual effect size is. The next steps will likely involve localization of broad genomic regions that influence these quantitative phenotypes; this can be performed using crosses between susceptible and nonsusceptible strains and assessment of the relationship in subsequent generations between the phenotype and genetic variants across the genome that differ between the parental strains—traditional quantitative trait locus (QTL) mapping. However, the investigation of complex traits in mice has been accelerated with the development of new analytic methods and computer software, and in silico gene mapping using comparative SNP maps between inbred mouse strains is now an important addition to QTL mapping (5). Using a bioinformatic approach to relate known genetic variation between strains to phenotypic variation may speed the discovery of disease genes in COPD, as it eliminates some of the time and cost involved in classical QTL approaches. In either case, the size of the genomic region identified will depend on the magnitude of the experiment. In classical QTL analysis, phenotyping and genotyping a larger number of mice in crosses between strains will increase the number of recombination events and lead to a narrower genomic region of interest. Including a larger number of strains for bioinformatic comparisons will similarly lead to a more refined estimate of QTL localization. The next step gets really hard. The process of fine mapping and gene identification for a QTL, in mouse or in man, remains quite challenging. Direct positional cloning efforts, by successively narrowing the key region using repeated crosses in mice, are time-consuming and expensive. Searching for reasonable pathophysiologic candidate genes within the regions of linkage or assessing expression differences in RNA from relevant tissues to select candidate genes are additional approaches to identify the key susceptibility variant or set of variants within a linked region. The recent development of novel approaches, such as chromosome substitution strains in mice, in which strains of mice are created that contain a single chromosome of one strain and the remainder of their chromosomes of a second strain, will likely make it easier to identify QTL signals and to localize them. Research continues to move rapidly in this area (6). During the process of fine mapping a QTL, the ability to compare murine genetic results to human genetic association analysis and linkage analysis results can be very useful. As shown in Table 1, analogous processes in human and murine studies are involved in demonstrating that genes likely influence a trait, using linkage approaches to achieve broad genomic localization of the susceptibility gene, and performing fine mapping and gene identification. If a potentially functional variant is identified, murine studies are especially critical, although investigation of function using in vitro methods can also be considered. Ideally, the human and mouse studies should be synergistic, rather than isolated pathways of investigation. Human studies of candidate gene variants within regions of murine linkage can provide strong evidence that a positional candidate gene is the QTL, whereas human linkage studies that show evidence for linkage in chromosomal regions syntenic to murine linkage signals suggest that the murine linkages are relevant to human disease. To date, this intersection of mouse and human results has not been applied in COPD, but efforts to link mouse QTL studies to human genetic association studies are ongoing in asthma (7, 8).
COPD is a complex disease, influenced by genetic factors, environmental determinants (mainly cigarette smoking), and gene-by-environment interactions. Few nonsmokers in the United States develop COPD, but only a minority of smokers ever develop the disease, suggesting variation between humans for COPD susceptibility similar to the variable susceptibility observed between different mouse strains. Thus, exposure to a noxious agent, invariably cigarette smoke in developed countries, combined with likely multiple predisposing genes, results in COPD. Several genomic regions have shown suggestive or significant linkage to pulmonary function phenotypes in families with severe, early-onset COPD, and a region of significant linkage on chromosome 2 to FEV1/FVC in these families with early-onset COPD has been shown to influence FEV1/FVC in families from the general population (9, 10). The intersection of these human linkage results with mouse strain QTL analysis will be essential. In this regard, the genetic loci in cigarette smoke–susceptible mice may mirror the susceptibility loci in people who develop COPD. Additional linkage studies of COPD in humans are required, and QTL mapping in mice together with fine mapping of linked regions in both humans and mice has the potential to identify novel COPD susceptibility genes. Improved understanding of COPD pathophysiology would result, potentially with new opportunities for treatment. FOOTNOTES Conflict of Interest Statement: S.D.S. has participated on the Advisory Board for Wyeth and receives a fixed stipend from the American Thoracic Society for editorial responsibilities for the American Journal of Respiratory Cell and Molecular Biology; D.L.D. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; E.K.S. has received grant support and honoraria from GlaxoSmithKline for a study of COPD genetics. REFERENCES
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
