Gene modifiers in cystic fibrosis

Studies of modifier genes in cystic fibrosis (CF) have often been performed in small or narrowly defined populations, leading to conflicting results. In this issue of the JCI, Dorfman et al. demonstrate in a large, population-based study that two previously studied modifier genes, coding for mannose-binding lectin 2 and TGF-β1, influence pulmonary outcome in pediatric CF patients (see the related article beginning on page 1040). They further show gene-gene interaction between the two, underscoring the complexity of CF lung disease. Their findings provide further impetus to study these molecules and associated signaling pathways in CF. In addition, these findings argue strongly for collecting genotypes of known modifiers prospectively in CF clinical trials as well as in longitudinal studies of infants identified through newborn screening, where the full impact of such modifiers can be defined more precisely.

Cystic fibrosis (CF) is a complex condition affecting a number of organs, including the exocrine pancreas, intestine, sweat glands, and lung; the effects of the latter are the major cause of disease morbidity and early mortality. Identification of the CF transmembrane conductance regulator (CFTR) gene in 1989 (1) was followed by the discovery of hundreds of mutant CFTR alleles and by attempts at genotype-phenotype correlations (2). The associations between CFTR genotype and exocrine pancreatic and sweat gland phenotypes were soon found to be greater than the association between CFTR genotype and the pulmonary phenotype (2, 3). Pulmonary phenotype is widely variable, even among patients with the same CFTR genotype. This variability could arise from some combination of environmental, stochastic, and modifier gene effects. Studies of twins and of siblings, which control for CF genotype and environment, have demonstrated high degrees of heritability of pulmonary outcomes (4), intestinal obstruction at birth (5), and early exocrine pancreatic dysfunction (6), supporting an important role for modifiers in explaining clinical variability of this disease.

Identification of modifier genes could be useful in several ways. Modifiers could provide clues to disease pathogenesis that might lead to development of new treatments. In addition, knowledge of modifiers might improve stratification for clinical trials and clinical prognostication. For these reasons, several groups began modifier studies. At least one validated modifier of lung disease in CF has been identified. A large study, primarily in adults, examining extremes of pulmonary CF phenotypes found that a genetic variant leading to higher production of TGF-β is associated with worse pulmonary outcome (7). More than 20 other candidate modifiers have been examined, often in relatively small studies and frequently without validating populations, leading to conflicting results (8).

In this issue of the JCI, Dorfman et al. demonstrate in their large, prospective study that mannose-binding lectin 2 (MBL2), a protein key to the innate immune response, is a modifier of CF lung disease in children and adolescents: individuals possessing genetic variants that result in low MBL2 production had an increased rate of decline of lung function (9). This was first proposed in 1999 (10), but a large subsequent study focusing on older CF patients was unable to confirm the findings (7). In addition, Dorfman et al. found that MBL2 is associated with earlier Pseudomonas aeruginosa colonization. They further demonstrate that TGF-β1, when considered alone, is a modifier of lung function decline in CF but not of age at P. aeruginosa colonization in CF. To make matters more interesting, they show statistically significant gene-gene interaction between MBL2 and TGFB1 in decline in lung function and age at P. aeruginosa colonization. This is believed to be the first demonstration of modifier gene-gene interaction in CF.

Several aspects of their experimental approach and additional findings are noteworthy (9). The Canadian CF Modifier Study, from which the patients investigated by Dorfman et al. were recruited, offers the largest number of patients yet for this area of research. Because many clinics and research centers participate, thus providing a patient population representative of the overall CF population, it is likely that the findings are robust. The investigators restricted their study population to individuals with insufficiency of the exocrine pancreas — a condition, manifest in 85% of patients with CF, in which the individual cannot properly digest food due to the absence of digestive enzymes made by the pancreas. In general, these patients have worse outcome than patients with sufficiency of the exocrine pancreas, emphasizing the need for careful investigation of this insufficient group. MBL2 protein levels were determined to confirm the effects of genetic variation — yet another strength of the article. However, TGF-β1 protein levels were not determined, possibly because of difficulties with this assay. Two important pulmonary end points were chosen: age at detection of first P. aeruginosa infection (the most common respiratory pathogen in CF patients) and rate of decline in lung function. The latter was assessed through forced expiratory volume in 1 second, which is the best predictor of pulmonary outcome in CF. The choice of end points is meaningful because delaying P. aeruginosa infection and slowing the rate of decline in lung function are currently the key goals of CF treatment (11, 12). The magnitude of the modifier effects, namely the delayed age of onset of P. aeruginosa infection by several years and differences in decline of lung function of 1%–2% per year, would be considered clinically very meaningful and likely would translate into differences in survival. A very interesting additional finding in the study of Dorfman et al. is the remarkable similarity in clinical characteristics between patients homozygous for the ΔF508 CFTR mutation in CF and patients who carry 2 different exocrine pancreatic-insufficient mutations with or without ΔF508 (9). Based on in vitro and in vivo studies, it is currently believed that exocrine pancreatic-insufficient patients have 1% or less of CFTR function compared with their wild-type counterparts (13). The findings of Dorfman et al. (9) tell us that the pathophysiology accompanying exocrine pancreatic insufficient mutations is essentially the same, regardless of the type of mutation.

Current models of CF airway pathophysiology include elements of impaired ion transport leading to abnormal bacterial clearance, persistent infection, intense inflammation, and structural injury. Both MBL2 and TGF-β1 could easily contribute to this cascade. Patients in other clinical settings who have genetic variants of MBL2 that result in low MBL2 protein levels are at risk for a variety of bacterial, viral, and fungal infections, including pathogens commonly encountered in CF such as P. aeruginosa, Staphylococcus aureus, and Hemophilus influenzae. As has been pointed out, low MBL levels in CF could make it just that much easier for offending bacteria to take hold in an airway that is already compromised (10). There is also some evidence that low MBL–producing genetic variants are associated with early mortality in CF (14, 15). Studies of MBL supplementation in CF may be justified, perhaps not only in patients who have low MBL levels.

The TGF-β signaling pathway is important in numerous human disease states, including more than a dozen hereditary conditions and almost as many multifactorial diseases (15). Many of these entities involve connective tissue formation or fibrosis in some manner. Very little is known about the mechanisms that lead to the prominent fibrosis and bronchiectasis that characterizes the structural injury to the CF airway. It is tempting to speculate that high TGF-β expression is associated with worse pulmonary outcome in CF because of accelerated airway scarring. TGF-β can also have both pro- and antiinflammatory activities. However, this is dependent on the site, cells involved, and stimuli (15). Given the important role of inflammation in CF, TGF-β could also be acting in this limb of the pathophysiologic schema.

Some notion of whether there are remaining modifiers to be discovered can be gained by comparing the effects of the known modifiers to estimated heritability. The magnitude of the effects of TGF-β1 and MBL2 observed by Dorfman et al. (9) is modest in terms of the number of patients affected. Roughly 20% of patients have either an MBL2 variant or a TGF-β1 variant that contributes to poorer outcome (9). Given the high degree of heritability of pulmonary outcomes, which approached a correlation coefficient of 0.8 in identical twins (4), it is almost certain that important modifiers remain unidentified. Future studies are therefore likely to identify other genes that play a role in modulating disease severity — some independently of others, but many involving gene-gene interactions. Each modifier is likely to have a modest effect, and as the most promising genes are validated or refuted, the remaining candidates may demonstrate an even smaller overall effect or may be influential in a smaller number of patients. To test each candidate and the interactions between them will require many more patients. These large-scale investigations will not be able to be carried out by any one of the existing gene modifier studies.

In a sense, the search for gene modifiers in CF is similar to the search for complex traits in polygenic diseases such as diabetes or hypertension. The effects of individual modifiers in monogenic conditions, as with individual traits in polygenic diseases, may be relatively small, but the overall picture becomes clearer as gene upon gene is identified through even larger collaborative efforts.

CF care and research are changing. Many more therapeutic agents are undergoing clinical trials (16). Response to a given treatment could rely on modifier status, which suggests that gathering this information should be part of clinical trial design. In addition, early treatment of P. aeruginosa infection is changing the prevalence of this microbe (12). Will current modifier effects defined in terms of P. aeruginosa infection persist if the prevalence of this organism changes? It is also likely that in the future, parents will demand all the prognostic information available. Are we at the point where parents should be given information on the modifier status of their child? Dorfman et al. (9) advise caution with respect to use of their results for stratification in clinical trials or in counseling parents. They believe that careful quantitation of sensitivity and specificity of modifier effects should be obtained by prospective studies of infants identified through newborn screening before we change current trial stratification or parental counseling approaches. Indeed, most communities are moving rapidly toward newborn screening and early diagnosis because of demonstrable clinical benefit (17). The collective work of Dorfman et al. and others in the modifier field should inform the longitudinal studies of early treatment of CF that are almost certain to follow adoption of universal newborn CF screening.

This work was supported in part by Cystic Fibrosis Foundation grant C017 and National Heart, Lung, and Blood Institute grant 1 U01 HL081335-01.

Address correspondence to: Frank J. Accurso, Mike McMorris Cystic Fibrosis Care and Research Center, Children’s Hospital, University of Colorado Denver, 13123 E. 16th Avenue, B395, Aurora, Colorado 80045, USA. Phone: (720) 777-6181; Fax: (720) 777-7284; E-mail: FAccurso@tchden.org.

Nonstandard abbreviations used: CF, cystic fibrosis; CFTR, CF transmembrane conductance regulator; MBL, mannose-binding lectin.

Conflict of interest: The authors have declared that no conflict of interest exists.

Reference information: J. Clin. Invest.118:839–841 (2008). doi:10.1172/JCI35138.

See the related article at Complex two-gene modulation of lung disease severity in children with cystic fibrosis.

1. Kerem, B., et al. 1989. Identification of the cystic fibrosis gene: genetic analysis. Science. 245:1073-1080.
2. Kerem, E., et al. 1990. The relation between genotype and phenotype in cystic fibrosis — analysis of the most common mutation (delta F508). N. Engl. J. Med. 323:1517-1522.
3. Kristidis, P., et al. 1992. Genetic determination of exocrine pancreatic function in cystic fibrosis. Am. J. Hum. Genet. 50:1178-1184.
4. Vanscoy, L.L., et al. 2007. Heritability of lung disease severity in cystic fibrosis. Am. J. Respir. Crit. Care Med. 175:1036-1043.
5. Blackman, S.M., et al. 2006. Relative contribution of genetic and nongenetic modifiers to intestinal obstruction in cystic fibrosis. Gastroenterology. 131:1030-1039.
6. Sontag, M.K., et al. 2006. Genetic and physiologic correlates of longitudinal immunoreactive trypsinogen decline in infants with cystic fibrosis identified through newborn screening. J. Pediatr. 149:650-657.
7. Drumm, M.L., et al. 2005. Genetic modifiers of lung disease in cystic fibrosis. N. Engl. J. Med. 353:1443-1453.
8. Buscher, R., Grasemann, H. 2006. Disease modifying genes in cystic fibrosis: therapeutic option or one-way road? Naunyn. Schmiedebergs Arch. Pharmacol. 374:65-77.
9. Dorfman, R., et al. 2008. Complex two-gene modulation of lung disease severity in children with cystic fibrosis. J. Clin. Invest. 118:1040-1049.
   View this article via: JCI Google Scholar
10. Garred, P., et al. 1999. Association of mannose-binding lectin gene heterogeneity with severity of lung disease and survival in cystic fibrosis. J. Clin. Invest. 104:431-437.
    View this article via: JCI Google Scholar
11. Konstan, M.W., et al. 2007. Risk factors for rate of decline in forced expiratory volume in one second in children and adolescents with cystic fibrosis. J. Pediatr. 151:134-139.
12. Frederiksen, B., Koch, C., Hoiby, N. 1997. Antibiotic treatment of initial colonization with Pseudomonas aeruginosa postpones chronic infection and prevents deterioration of pulmonary function in cystic fibrosis. Pediatr. Pulmonol. 23:330-335.
13. Rowe, S.M., Accurso, F., Clancy, J.P. 2007. Detection of cystic fibrosis transmembrane conductance regulator activity in early-phase clinical trials. Proc. Am. Thorac. Soc. 4:387-398.
14. Buranawuti, K., et al. 2006. Variants in mannose-binding lectin and tumor necrosis factor {alpha} affect survival in cystic fibrosis. J. Med. Genet. 44:209-214.
15. Harradine, K.A., Akhurst, R.J. 2006. Mutations of TGFbeta signaling molecules in human disease. Ann. Med. 38:403-414.
16. Mayer-Hamblett, N., Ramsey, B.W., Kronmal, R.A. 2007. Advancing outcome measures for the new era of drug development in cystic fibrosis. Proc. Am. Thorac. Soc. 4:370-377.
17. Farrell, P.M., et al. 2001. Early diagnosis of cystic fibrosis through neonatal screening prevents severe malnutrition and improves long-term growth. Pediatrics. 107:1-13.