| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Section of Pulmonary and Critical Care Medicine, Department of Medicine, University of Chicago, Chicago, Illinois
Correspondence and requests for reprints should be addressed to Joe G. N. Garcia, M.D., Chairman, Department of Medicine, Pritzker School of Medicine, University of Chicago, AMB W604, 5841 South Maryland Avenue MC 6092, Chicago, IL 60637. E-mail: jgarcia{at}medicine.bsd.uchicago.edu
ABSTRACT
Acute lung injury (ALI) is a common and often devastating illness characterized by acute hypoxemia, alveolar flooding, and an unacceptably high morbidity and mortality. Because only a fraction of the patients exposed to ALI-inciting events progress to development of the syndrome, there is significant interest in the identification of genetic factors potentially contributing to ALI susceptibility or prognosis. Two complementary strategies used to elucidate ALI genetics formulate the "candidate gene approach," whereby genes are identified by either global gene expression profiling in humans or animal models of ALI, often yielding highly conserved candidates across multiple species, or by related literature searches. Relevant variants or single nucleotide polymorphisms (single base pair substitutions) in these ALI candidate genes are tested for differences in allelic frequency for both ALI susceptibility and outcome between ALI cases and control patients at risk for ALI. This approach has yielded important variants in a number of genes (angiotensin converting enzyme, surfactant protein B, heat shock protein 70, pre–B-cell colony enhancing factor, myosin light chain kinase, and macrophage migration inhibitory factor) contributing toward an ALI phenotype. An alternative strategy not yet used in ALI genetic studies includes genomewide analyses to locate "hot" genomic segments harboring several hundred genes, with potential ALI candidate genes embedded within these segments. Overall, the detailing of specific ALI-associated polymorphisms will continue to provide new insights in the understanding of ALI pathogenesis, reveal novel molecular targets, and promote the development of individualized therapies to reduce morbidity and mortality from this devastating disease.
Key Words: microarray • genomics • translational research
Acute lung injury (ALI) describes the syndrome of respiratory distress characterized by acute onset, severe hypoxemia, and bilateral radiographic infiltrates in the absence of left atrial hypertension (1). ALI is both common and devastating, with an estimated 190,000 cases occurring annually in the United States and a mortality that exceeds 35% (2). Although our understanding of the pathogenesis of ALI and its most severe form, acute respiratory distress syndrome (ARDS), continues to evolve, we recognize that critical elements in the syndrome include profound inflammation, deranged vascular permeability, and flooding of alveoli with protein-rich fluid. Neither ALI nor ARDS are specific disease processes themselves, but rather are syndromes that occur in a fraction of patients exposed to at-risk conditions, such as sepsis, trauma, pneumonia, or aspiration (3).
The treatment of ALI-induced respiratory failure almost invariably necessitates mechanical ventilation; however, the ventilator itself can worsen preexisting lung injury and indeed may incite de novo lung injury (4). Ventilator-associated or ventilator-induced lung injury (VALI, VILI) can be minimized by restricting the ventilator-delivered tidal volume (VT) and maintaining plateau airway pressures below 30 cm of water, which reduces mortality and days spent on the ventilator (5). Lung injury caused by high-volume ventilation has also proven a very useful experimental model for investigators studying ALI.
Among the many unanswered questions regarding ALI is the role that genetics play in determining a patient's risk for or subsequent outcome after ALI. It is well recognized that of all the patients exposed to potential ALI-inciting insults, only a minority will develop the syndrome. In a prospective study in more than 130 patients who suffered from sepsis, aspiration, multiple transfusions, or trauma, sepsis was the condition most likely to result in ALI/ARDS, with almost 40% of septic patients developing severe lung injury (6). To better define the population at risk for ALI, recent attention has turned to genetic variation as a possible explanation for which patients develop the syndrome and which incur the greatest morbidity and mortality, and a number of studies implicate a genetic basis for ALI susceptibility and severity (7–12). This notion is also supported by the fact that mortality rate in ALI (as well as sepsis) is higher in African Americans than in other ethnic groups in the United States, and that the outcome in patients with similar clinical characteristics may vary from death to complete resolution (3, 13, 14). However, traditional genetic studies using family linkage mapping (discussed elsewhere in this issue) are not feasible given the sporadic nature of ALI and the necessity of an extreme environmental insult. In addition, genetic studies of ALI are challenging due to the tremendous phenotypic variance of critically ill patients, incomplete gene penetrance, complex gene–environment interactions, and a high likelihood for locus heterogeneity. Determining which environmental and genetic factors are relevant and how these factors interact with each other remains a significant challenge.
Despite these challenges and potential impediments, with the completion of the Human Genome Project the study of ALI genomics has greatly advanced, making the elucidation of the molecular basis of complex lung disorders such as ALI/ARDS a potentially achievable goal. The promise of this era of molecular medicine (enabled by rapid, high-throughput genotyping and gene expression profiling) is that we may rapidly ascertain a patient's genetic profile and target individualized therapies for specific risks. This review highlights currently available approaches to investigate potential genetic susceptibility factors, and presents several identified genes that impact susceptibility, protection, or severity of ALI (Figures 1 and 2).
|
|
There is substantial evidence that interactions between mechanical stress (overdistension of lung tissues with airway opening and closing) and inflammation are central to the pathogenesis of VALI/VILI. We recently reported mechanical stress–induced genomic responses throughout the heterogeneous injured lung in relation to local mechanical events in a canine model of VALI (15). Affymetrix (Sunnyvale, CA) microarray approaches were combined with functional computed tomography imaging to define regional mechanical stress with normal aeration in the control right lung (uninjured except for minimal dependent atelectasis) but severe loss of aeration throughout the injured left lung (except for apical regions), with the greatest loss of aeration occurring in the dependent base regions, consistent with human ALI (15). Gene expression profiles from regions experiencing differential mechanical stresses during mechanical ventilation revealed significant alterations in a large number of genes in dependent compared with nondependent regions; gene expression moved in the opposite direction in 63 genes, which was validated in five up-regulated genes with semiquantitative reverse transcriptase–polymerase chain reaction. Highly differentially regulated genes between apex and base regions included several genes commonly associated with ALI (vascular endothelial growth factor [VEGF], thrombospondin 1 [THBS1], plasminogen activator inhibitor 1 [PAI-1], transforming growth factor ß [TGF-ß]; see Table 1), as well as several novel genes not previously described in the context of ALI, such as pre–B-cell colony enhancing factor (PBEF) (16). Additional studies are required to more fully explore the influence of this gene expression on physiologic dysfunction in these mechanically challenged regions.
|
ACE as the First Candidate
One approach proven effective in the search to elucidate the genetic and molecular underpinnings of ALI/ARDS has been the "candidate gene approach," a variation on a case-control study model that examines the allelic frequencies of polymorphisms in genes with putative mechanistic roles in lung injury or inflammation. This candidate gene approach identified the first association of a restriction fragment length polymorphism (RFLP)—in this case, of the angiotensin converting enzyme (ACE) gene—with ALI (7). Interest in the ACE gene originated from hypotheses that activation of the pulmonary renin-angiotensin system might impact the pathogenesis of ARDS by altering vascular permeability, vascular tone, fibroblast activation, and altered endothelial and epithelial cell survival (7, 17–20). The human ACE gene sequence, located on chromosome 17q35, contains an intronic insertion (I) or deletion (D) of a 287-bp alu repeat sequence. This genetic variant was found to account for approximately half of the variance in serum ACE protein levels among healthy white volunteers (21, 22), whereby ACE gene expression of patients with the DD genotype is significantly higher than that of patients with the II genotype (23). Because the polymorphism resides in a noncoding intronic region, this RFLP variant may serve as a marker for another locus with more causative or functional significance (7, 24).
The ACE genotype of white patients with ARDS has been analyzed and allele frequencies and outcomes compared with ventilated intensive care unit patients without ARDS, patients after coronary artery bypass surgery, and healthy control subjects. The DD genotype and the D allele occurred with greater frequency in the ARDS population than all other control groups. Importantly, among patients with ARDS, the DD genotype had a markedly higher mortality (54%) compared with the II or ID genotypes (11 and 28%, respectively) (7). A subsequent study of Han Chinese patients in Taiwan confirmed a higher mortality for patients with ARDS and the DD or ID genotype as compared with the II genotype, although the D allele frequency in the Chinese population is significantly reduced relative to Western populations (25). In contrast to the association between ARDS outcome and ACE polymorphism, no association was noted in the Chinese population between the genetic variant and susceptibility for ARDS (25). Although no study has focused on ACE polymorphisms and lung injury in an African-American population, epidemiologic studies suggest a slightly increased rate of the D allele among African (Nigerian and African American) populations as compared with whites (26, 27), which may contribute to the observed racial disparity observed for patients with ALI/ARDS (13). In contrast, Mexican and Amerindian populations were found to have slightly lower allelic frequencies for the deletion (D) allele (26). As evidence mounts that the renin-angiotensin system plays an important role in ALI, a very interesting corollary is that blocking the renin-angiotensin system—for example, with ACE inhibitors or angiotensin receptor blocking drugs—may have a therapeutic effect, as suggested in an animal model of bleomycin-induced ALI (28).
Surfactant Peptides: Translating ALI Genetics from Pediatric Patients to Adults
Surfactant peptides (and the genes that encode them) have been a historical focus of interest in the pathogenesis of ARDS. Pulmonary surfactant, synthesized by type II respiratory epithelial cells, lowers the surface tension at the alveolar air–liquid interface, thereby preventing lung collapse (29, 30). Surfactant protein B (SP-B) seems particularly critical to normal lung function, because neonates born with congenital alveolar proteinosis, a rapidly fatal disease, were found to be deficient in SP-B due to a frameshift mutation in the SP-B gene (31, 32), and targeted deletion of SP-B in mice results in lethality (33). Experimental models of ALI indicate that serum levels of SP increase progressively, reflecting lung injury and increased permeability, and plasma levels in patients with ARDS correlate with a worse clinical outcome (34–37).
SP-B is encoded on a single gene on the short arm of chromosome 2, and its genomic and cDNA sequences are well characterized (38, 39). Insertion/deletion genetic variants of dinucleotide (CA) tandem repeats within intron 4 of SP-B were identified with a higher frequency among newborns with respiratory distress syndrome than among infants without respiratory distress syndrome (40), as well as among patients with ARDS compared with healthy control subjects (41). Although a case-control study failed to replicate the increased frequency of the intron 4 SP-B gene variant in patients with ARDS (12), allelic frequency differences did emerge when the ARDS cohort was stratified by sex, with women with the variant SP-B gene proving to be significantly more likely to develop ARDS and to be admitted to the intensive care unit with a direct pulmonary injury, such as pneumonia or aspiration (12). Given the intronic nature of this variant polymorphism, it must be considered that the variant may be closely linked to a functional locus on the SP-B gene or may have a regulatory effect, such as mRNA splicing, which might contribute to the observed phenotype (12, 42).
A possible link between SP-B polymorphisms and ALI pathogenesis is suggested by a pilot study using SP-B as a candidate gene. Researchers identified C-1580T, a functional single nucleotide polymorphism (SNP) on exon 4, which encodes a missense mutation that converts threonine to isoleucine and eliminates a potential N-linked glycosylation recognition site (9). The SP-B C allele has two functional N-linked glycosylation sites for asparagine, whereas the T allele has only one, (43) and N-linked glycosylation may affect protein processing or folding (44). Although in the control population, the two alleles were equally distributed (48% C and 52% T), among patients with ARDS the frequency of the C allele at 1,580 was significantly higher (62.5% C and 37.5% T, p = 0.019) (9). For patients with "idiopathic," or direct pulmonary risk for ARDS, rather than those patients with ARDS triggered by nonpulmonary sepsis or trauma, the C allele increased the odds of ARDS more than twofold relative to control subjects (9). Although this study mandates careful validation given that allelic frequency of the control group deviated from Hardy-Weinberg equilibrium (HWE) (45), the C-1580T polymorphism may yet provide an important insight to our understanding of the role of surfactant in the pathogenesis of ALI.
Heat Shock Protein 70: Preliminary Evidence for a Protective Genotype?
Among the most highly conserved biological processes across virtually all organisms is the heat shock response, involving the rapid induction of a set of genes that encode heat shock proteins (HSPs), which induce stress tolerance and protect against further injury (46). Serving as molecular chaperones during all aspects of protein synthesis, assembly, and transport, HSPs are involved in maintaining the structural and conformational integrity of intracellular proteins (47, 48). In addition, the HSPs are highly active within the immune system, inhibiting inflammatory cytokine production and mitigating the toxic effects of reactive oxidative species (47–49). The HSP70 family, the most prominent and best-studied eukaryotic class of HSPs, includes both a constitutive and an inducible form.
A Dutch group investigating the link between HSPs and lung injury found that animals exposed to endotoxin-induced ALI, followed by an injurious ventilation strategy using high VT with zero positive end-expiratory pressure (ZEEP), expressed significantly higher levels of HSP70 than either nonventilated control animals or animals with endotoxin-induced lung injury but ventilation with a lower VT (50). In addition, although animals from the ZEEP group had significantly higher expression of the inflammatory cytokine interleukin (IL)-1ß and a trend toward increased IL-6 expression (p = 0.08) relative to PEEP animals or controls, the ZEEP group also demonstrated a significant inverse correlation between inflammatory cytokines (IL-1ß, IL-6, and tumor necrosis factor [TNF]-
) and HSP70 expression (50). This negative correlation supports the concept that HSP70 has an antiinflammatory function in the model of ventilator-induced ALI, an observation further substantiated in HSP70 knockout mice with septic ALI, which demonstrated increased nuclear factor-
B activation, IL-6 and TNF- expression, and 5-d mortality compared with wild-type mice subjected to the same insult (51).
Three genes encode the HSP70 family proteins and reside in close proximity to the TNF locus within the human major histocompatibility complex on chromosome 6. Two of these genes, HSP70-2 (also called HSPA1B) and HSP70-Hom (also called HSPA1L), display well-characterized polymorphisms (52), with some evidence that suggests a possible influence of HSP genotype on immune efficiency in non-Hodgkin's lymphoma and breast cancer (53, 54). In contrast, no associations between the HSP70 SNPs and sepsis susceptibility or outcome were found among patients with severe sepsis (52). Although the specific examination of an association between HSP polymorphisms and ALI susceptibility is lacking, recent evidence in a small cohort of patients with major trauma suggests that the HSP genotype might influence clinical outcome. Over 80 trauma patients were genotyped with respect to the A-1538G polymorphism of HSP70-2 and the C-2437T polymorphism of HSP70-Hom (55). Genotypes were not compared with controls, thus limiting any inference on ALI genotype and trauma susceptibility. However, carriers of the HSP70-Hom genotype CT were three to four times more likely to develop multiple organ failure compared with those with the TT genotype (55). ARDS tended to be more common among patients with the CT genotype (with an odds ratio of 2.1), but this did not achieve statistical significance (55). The HSP70-2 polymorphism was not in HWE for the sample studied, and a significant association between A-1538G and outcome was not observed. Although this represents very preliminary evidence that the C-2437T polymorphism of HSP70-Hom may influence outcome from lung injury, the concept of a protective TT genotype at this SNP is intriguing, and warrants further investigation.
USING COMPARATIVE AND FUNCTIONAL GENOMIC APPROACHES TO INVESTIGATE SUSCEPTIBILITY TO SEARCH FOR NOVEL ALI CANDIDATES
A novel extension of the candidate gene approach to identify ALI candidate genes incorporates gene expression differences between control subjects and subjects with ALI, a strategy unencumbered by preexisting hypotheses, or known involvement of specific genes. We have used this approach, combining extensive gene expression profiling studies in both animal (murine, rat, and canine) and human models of ALI with novel bioinformatic methods linked to a Eukaryote Gene Orthologs database (15, 56–58). Given that orthologs have evolved from common ancestral genes and are presumed to retain a similar function despite speciation, overlapping responses across species to a common insult might serve as a powerful filter to yield information about conserved microarray-derived responses to injury. When orthologous genes exhibiting similar patterns of differential gene expression across all species were selected, the resultant candidate gene pool included genes already under investigation in ALI as well as several novel genes involved in processes and pathways not previously suspected to affect ALI pathogenesis (58). The power of this multispecies, microarray-driven, high-throughput functional genomic technology is thus readily apparent, and already yielding exciting results.
PBEF: An Unknown Candidate
An early success for the modified candidate gene approach was the identification of the cytokine PBEF in microarray analyses of saline lavage- (noted above), sepsis-, and high-ventilation–induced injury models of murine and canine ALI (16). PBEF was consistently demonstrated to exhibit very high levels of expression, with the lung tissue of ALI models having between two- and fivefold higher expression of PBEF relative to control lungs (16). Named for its effect on the maturation of B-cell precursors, PBEF was relatively unknown before the ALI experiments, with fewer than 10 citations in the literature. Although its function remained obscure, PBEF expression was known to be up-regulated in vitro in a human amniotic epithelial cell line exposed to mechanical force or inflammatory cytokines (59, 60). In addition, PBEF was found to be expressed in peripheral neutrophils of patients with sepsis, and it appeared to inhibit neutrophil apoptosis (61), an intriguing finding because sepsis is a frequent cause of ALI.
After detailing of its increased expression in ALI lung tissues, PBEF was studied as a potential biomarker in both bronchoalveolar lavage (BAL) fluid and serum. Both animals and humans with ALI had significantly increased levels of PBEF protein in their BAL samples—between two- and fivefold more than controls—and even serum levels of PBEF protein were doubled in dogs exposed to sepsis-induced ALI relative to control animals (16). In vitro experiments on human lung endothelial cells confirmed increased PBEF protein expression after challenge with endotoxin (LPS), mechanical stress, or inflammatory stimuli such as the cytokines TNF- or IL-Iß (16).
With mounting evidence implicating PBEF as a potential candidate gene for ALI and despite the fact that its molecular actions are incompletely understood, we next investigated whether common variants of the PBEF gene might affect susceptibility to ALI. SNP discovery carried out by direct DNA sequencing identified multiple PBEF SNPs, including a T-1001G transversion located in the PBEF immediate promoter with a high representation among subjects with ALI (16). The T-1001G SNP was in HWE, and the G allele was found with significantly higher frequency among patients with sepsis-induced ALI or with sepsis alone compared with healthy control subjects (16). A significant association was also noted between the GT or GG genotype and ALI, with carriers of the G allele incurring a 2.75-fold increased risk of ALI compared with control subjects, and the G allele remained an independent risk factor for ALI susceptibility in multiple logistic regression analysis controlling for clinical factors (16). Preliminary studies, however, did not suggest an association of the polymorphism with a difference in mortality.
A second SNP in the human PBEF gene, C-1543T, was also found to be in HWE and to have a significantly lower T allele frequency among patients with ALI than among healthy control subjects (i.e., a protective SNP) (16). A borderline association (p = 0.059) was observed between the C-1543T genotype and ALI, which led to a haplotype analysis of both SNPs, T-1001G and C-1543T. Four possible haplotypes were observed: GC, GT, TC, and TT. The GC haplotype was significantly more common in the ALI and sepsis samples, and univariate logistic regression analysis found that it conferred a 7.7-fold higher risk of ALI and a 4.8-fold higher risk of sepsis (16). Conversely, the TT haplotype appeared protective against ALI, occurring at half the frequency in patients with ALI compared with control subjects, and with a trend toward decreased risk on univariate analysis (16). Very recently, the influence of the T-1001G polymorphism and the GC haplotype was confirmed to associate with an increased risk of ARDS (odds ratio, 1.28–1.36) among a large cohort of patients at risk for ARDS (62). We are only beginning to understand the molecular actions of PBEF, but our recent work suggests a potential role in regulating endothelial cell calcium-dependent cytoskeletal rearrangement (63). PBEF may thus be critically involved in the vascular responses to inflammation and barrier regulation, processes of intense interest for ALI researchers (63).
Myosin Light Chain Kinase as a Candidate Gene
Complementing this global gene expression profiling approach in humans or animal models of ALI, an alternate strategy used to elucidate ALI genetics includes the identification of ALI genes via the interrogation of known pathways in conjunction with the published literature. This genomewide analytic approach identified the MYLK gene, which encodes myosin light chain kinase (MLCK), whose gene expression was only modestly altered in animal and human models of ALI (< twofold increase) and did not survive the rigorous filtering strategies we typically use. However, since its cloning by our laboratory (64), our structure and function studies have strongly suggested the MYLK gene product to be a robustly viable candidate to affect ALI susceptibility and outcome. The human MYLK gene resides on the short arm of chromosome 3q21 and encodes three proteins, including both nonmuscle and smooth muscle MLCK isoforms. The nonmuscle MLCK isoform is centrally involved in the cytoskeleton rearrangement regulating vascular barrier function and permeability, angiogenesis, endothelial cell apoptosis, and leukocyte diapedesis, suggesting a possible mechanistic role for MLCK in the elaboration of ALI (65–68). Endothelial cell MLCK knockout mice (which retain the smooth muscle MLCK isoform) are less susceptible to LPS-induced ALI as well as injury due to high VT mechanical ventilation (69). In addition, pretreatment with an inhibitor of MLCK attenuates VILI in a rat model (70).
Direct sequencing of MYLK by our group yielded 57 genetic variations, of which 36 SNPs were chosen for further linkage disequilibrium studies (10). Because there is significant variability in marker allele and haplotype frequencies across populations (71), study subjects, including patients with sepsis-associated ALI, patients with sepsis alone, and healthy control patients, were stratified by ethnicity, as either European American or African American (10). After selecting for SNPs in coding regions causing amino acid changes and confirming that the SNPs did not deviate significantly from HWE, genetic association testing was carried out on 28 SNP markers for European-American subjects and 25 SNP markers for African-American subjects.
By single-locus analyses, four MYLK SNPs were associated with the sepsis phenotype and one additional SNP was associated with an increased ALI risk in the European-American population (10). Genotype and haplotype analysis discovered additional regions of the MYLK gene that confer ALI susceptibility in both ethnic groups. In whites, the AC haplotype involving markers rs3845915 and MYLK_037 was associated with more than a fivefold increased risk of ALI and a sixfold increased risk of sepsis, whereas the haplotype GGT, composed of markers MYLK_021, MYLK_022, and MYLK_011, appeared to confer specific risk for ALI but not sepsis (10). In African-American subjects, the CAG haplotype from markers hcv1602689, MYLK_007, and rs11707609 was extremely overrepresented, occurring with a frequency of 11% in the ALI group versus only 1% in the sepsis group (10). The CAG haplotype did not occur at all in European Americans, highlighting a potential genetic contribution to observed differences between white and African-American patients with ALI/ARDS (13). Importantly, although a number of polymorphisms stratified with ALI susceptibility, there was insufficient information to determine any significant association with either ALI severity or outcome (10). Although the results are encouraging due to their replication in two different population groups, this association needs further exploration to unravel the functionally relevant variant(s). In addition, more study is warranted to decipher the mechanistic ramifications for the nonsynonymous SNPs conferring amino acid sequence changes that have the potential to induce major conformational change in the enzyme, thereby affecting either enzymatic activity or interactions with other regulatory proteins (10).
MIF: A Candidate Gene Authenticated by Genomewide Analysis
The genomewide approach can also be useful in confirming candidate genes already suspected to play a role in ALI pathogenesis. Such was the case for macrophage migration inhibitory factor (MIF), an inflammatory cytokine first implicated to potentiate ARDS in 1997 (72). Originally named for its ability to inhibit the random migration of macrophages, MIF was subsequently found to be a soluble product of activated T cells (73), and eventually proven to be expressed by a variety of cell types, such as monocytes/macrophages, pituitary cells, vascular endothelium, and respiratory epithelium (74, 75). MIF has been demonstrated to override or counterregulate the immunosuppressive effects of glucocorticoids, and may be an important regulator of the delicate cytokine balance between inflammation and immunity (75, 76).
Researchers found increased levels of MIF in the serum and BAL fluid of patients with ARDS, and performed ex vivo tissue culture studies to prove that treatment of alveolar macrophages from patients with ARDS with exogenous MIF caused increased expression of TNF- and IL-6 (72). The same cells, when treated with anti-MIF antibodies, had decreased expression of TNF- and IL-6 (72). Our understanding of MIF in lung inflammation was further extended by the demonstration of enhanced MIF protein expression in alveolar capillary endothelium and alveolar macrophages of patients with ARDS as compared with critically ill control patients, and that in an in vitro endothelial cell culture system, MIF up-regulates its own synthesis as well as that of TNF- (74). Pretreating cells with anti-MIF antibodies or glucocorticoid before delivering MIF, or before subjecting a murine model to LPS, blunted both the MIF and TNF- production and attenuated pulmonary pathology (74). In addition to its up-regulation of TNF-, MIF also induced a small increase in aquaporin 1 (AQP1) expression, and this effect was blocked by administration of anti-MIF or glucocorticoid (74). Because aquaporins are water channels expressed in alveolar endothelial and epithelial cells, they remain further candidates of interest for their potential role in ALI, perhaps modulating fluid movement between the airspace, interstitium, and capillary (77, 78).
We examined MIF gene and protein expression in murine and canine models of ALI (using high VT mechanical ventilation and endotoxin exposure) and in patients with either sepsis or sepsis-induced ALI. MIF gene expression and protein levels were significantly increased in each ALI model, with serum MIF levels significantly higher in patients with either sepsis or ALI compared with healthy control subjects of either African or European-American descent (J.G.N.G., unpublished data). In addition, MIF expression increased twofold in human lung endothelium exposed to 48 h of cyclic stretch (11).
Because these studies inferred a likely role for MIF in ALI and sepsis susceptibility, we next assessed MIF SNPs in ALI and sepsis. An SNP in the MIF gene at 173 bp (G-173C) has been associated with increased susceptibility to juvenile idiopathic arthritis (80, 81). C allele carriers were more likely to have idiopathic arthritis (odds ratio, 2.3) (80), and significantly higher levels of MIF in their serum and synovial fluid than G carriers (79). We have studied the association of eight MIF SNPs, including the G-173C SNP (all within a 9.7-kb interval on chromosome 22q11.23), with the development of sepsis and ALI in European and African-American populations. Genotyping in 506 DNA samples did not elucidate a significant association of any single SNP, including the G-173C SNP, but revealed several haplotypes located in the 3' end of the MIF gene to be strongly associated with sepsis and ALI in both populations (11). Thus, our combined functional genomic and genetic approaches suggest that MIF is a relevant molecular target in ALI.
FROM POLYMORPHISM TO FUNCTION: CROSSING THE CHASM
The study of genetic contribution to ALI pathogenesis, severity, and response to therapy remains a nascent field, albeit an exciting one with great promise. In addition to selecting the appropriate candidate genes to subject to further analysis, a great challenge in this arena continues to be the ability to define functional relationships to the observed variations. In this review, we have highlighted the utility of combining advanced bioinformatic techniques with multispecies gene expression profiling as a way to broaden our net for ALI-related genes. This genomewide approach is the ideal complement to more traditional, hypothesis-based testing of candidate genes, and indeed, we anticipate that the genomewide approach will fuel additional analysis of new candidates. In addition, the genomewide approach allows a wide-angle view of key biological processes, such as inflammation, coagulation, and cell proliferation, without being constrained by what may be our incomplete current understanding of how various factors interact.
In addition to furthering our understanding of pathogenesis, the great promise of functional genomic studies of ALI lies in our potential to develop and apply specific therapies to patients based on their individual risk. To date, we lack pharmacologic therapies for ALI and rely solely on safe ventilator and fluid management (5). In the future, however, it may be that the identification of, for example, an ACE-prolific or surfactant-deficient genotype may offer selective targets for therapy. In the interim, ongoing studies with well-phenotyped populations of patients with ALI/ARDS should continue our progress toward this lofty goal.
ACKNOWLEDGMENTS
The authors thank Shwu-Fan Ma for her critical review and assistance.
FOOTNOTES
Conflict of Interest Statement: Neither author has a financial relationship with a commercial entity that has an interest in the subject of this manuscript.
(Received in original form September 1, 2006; accepted in final form October 12, 2006)
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  J. E. Levitt, M. K. Gould, L. B. Ware, and M. A. Matthay Analytic Review: The Pathogenetic and Prognostic Value of Biologic Markers in Acute Lung Injury J Intensive Care Med, May 1, 2009; 24(3): 151 - 167. [Abstract] [PDF]
|
|
|
|
 |
|
 N. J. Meyer, Y. Huang, P. A. Singleton, S. Sammani, J. Moitra, C. L. Evenoski, A. N. Husain, S. Mitra, L. Moreno-Vinasco, J. R. Jacobson, et al. GADD45a is a novel candidate gene in inflammatory lung injury via influences on Akt signaling FASEB J, May 1, 2009; 23(5): 1325 - 1337. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. E. Parsons Genetics, Iron, and ALI: An Intriguing Relationship Chest, June 1, 2008; 133(6): 1295 - 1296. [Full Text] [PDF]
|
|
|
|
 |
|
 J. L. Koyner and P. T. Murray Mechanical Ventilation and Lung-Kidney Interactions Clin. J. Am. Soc. Nephrol., March 1, 2008; 3(2): 562 - 570. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
