With an autosomal recessive pattern of inheritance, CF was long considered a monogenic disease with 1 mutant allele inherited from each parent. While CF neonatal screening is offered in several states of the U.S., counseling of families has been difficult, because CF genotyping does not easily predict onset and severity of pulmonary complications [14
]. Strategies to identify modifier genes for the CF phenotype are important for defining disease prognosis and developing new strategies to prevent progression of the disease.
There are several chloride conductances, which have been characterized in the mammalian lung: the cAMP-dependent cystic fibrosis transmembrane conductance regulator (CFTR) [15
], the Ca++
-dependent chloride channel (CaCC) [16
], the outwardly rectifying chloride channel (ORCC) [19
], the purinergic receptor-mediated chloride channel [20
], and the voltage- and volume-regulated, ClC family of chloride channels [3
One or more of the chloride channels present in the respiratory epithelium may be able to partially compensate for defects in another. For example, there was no lung pathology in the first CF knock-out mouse models, where there is enhanced activity of a Ca++
-dependent chloride channel [25
], however lung disease is present when alternative chloride channels are absent [26
]. The CF mouse, however, develops severe intestinal disease leading to premature death, which has been attributed to inadequate secretion via alternative chloride channels. Ca++
-dependent chloride conductance is low in the intestine of the CF knock-out mouse. To take advantage of alternative chloride channels in the lung, UTP analogues have been used to stimulate chloride secretion in CF individuals via the purinergic receptor-mediated chloride channels [20
One member of the ClC family of chloride channels may also be an alternative chloride conductance in the airway epithelium. We have demonstrated that CLC-2 mRNA and protein are abundantly expressed in the fetal lung [1
] and that acidic pH can activate chloride secretion [3
]. CLC-2 mRNA and protein are much higher in brain and kidney compared to tissues that are more severely affected by defective CFTR (lung, intestine, liver) [1
] suggesting that CLC-2 expression may protect against disease manifestations in certain tissues. CLC-2 immunolocalizes to the apical surface of the respiratory epithelium [2
], consistent with the potential to function as a chloride channel in a secretory organ. In this study, we have shown that several CF subjects do express CLC-2 protein as adults (figure ), unlike in rats [2
]. In single channel recordings, overexpression of CLC-2 in a CF bronchial epithelial cell line demonstrated that chloride secretion can be enhanced [3
]. While the CLC-2 knock-out mouse has degeneration of the retina and testes [28
], loss of CLC-2 function has not been associated with lung disease. To date overexpression of CLC-2 has not been described in an animal model to determine if this channel can be upregulated and serve as a potential therapeutic target for CF.
In this study of 31 CF subjects, we identified 5 single nucleotide polymorphisms that have not previously been described for human CLC-2. One of these is -693 relative to the ATG start site of hCLC-2 (Genbank S7770). The -693 G/A polymorphism is a putative AP-2 binding site, predicted by TESS and MATINSPECTOR [12
] and may be important for regulation of the gene. This polymorphism was no more frequent in the CF subjects with mild lung disease compared with the subjects with severe lung disease.
In the rat, SP-1 sites are important for gene regulation [5
]. ClC-2 expression in the lung is developmentally downregulated at birth [2
] and is dependent on Sp binding to GC boxes in the ClC-2 promotor [5
]. These GC boxes are highly conserved in human and rat, suggesting they are important sites for gene regulation. Phosphorylation of Sp-1 decreases its DNA binding activity and coincides with the downregulation of CLC-2 expression [5
]. SNPs in the conserved GC boxes were not identified in the subjects of this study.
We also identified 4 polymorphisms in hCLC-2 intron1. These did not appear more or less frequently in the mild CF subjects. Two of the polymorphisms were in complete linkage disequilibrium. The polymorphisms were not identified in areas that were highly conserved in rat or mouse. While splice variants of exons may be affected by intron/exon boundaries, we did not find any polymorphisms in the region of the rat exon 20 splice variant [6
These findings suggest that CLC-2 is not regulated differently at the genomic level in relatively healthy CF adults. Lack of an association in this study does not exclude the possibility that CLC-2 plays a role in modifying the CF phenotype as might be suggested by the variability of CLC-2 protein expression in primary respiratory epithelial cells from CF subjects in this study. Although we were limited by inadequate power with a small sample size and because phenotypic contrast was low, our data suggest that gene regulation of CLC-2 in relation to polymorphisms in regulatory domains does not play a major role in protection against CF lung disease. Studies which rely on recruitment of small numbers of patients have been shown to detect a difference when a strong relationship is present [31
]. Another limitation may be in the selection of FEV1 at a single point in time, rather than using rate of decline of FEV1. Other studies of CF modifier genes have similarly found difficulty in confirming a candidate gene, which also relied on FEV1 at one time point. In addition, the effect on lung phenotype may occur at an earlier stage of CF lung disease, and examination of adults only as in this study may have limited our ability to detect a difference. The polymorphisms identified in this report should facilitate further investigation of CLC-2 regulation. While we did examine subjects with the same CF genotype (namely delF508 homozygous), measures of ion transport (sweat chloride, nasal potential difference), time to colonization with Pseudomonas, and frequency of pneumonia, should be taken into account in future studies.
The identification of candidate genes, which may modify CF lung disease is important so that new therapies may be developed. Multi drug resistant genes have recently been identified that provide some "protection" to the CF lung phenotype [32
]. Ion transport dysfunction of CFTR and the channels it regulates, however, may not be the only determinant of disease severity. Many have suggested that inflammatory mechanisms may also impact disease progression and survival in CF individuals [33
]. Other classes of candidate genes possibly related to CF phenotype include tumor necrosis factor alpha (TNF-α), nitric oxide synthase (NOS), alpha 1-antitrypsin, mannose-binding lectin [34
], and other ion channels such as the basolateral K+ channels [17
Lastly, gene expression and function may be independent of genomic polymorphisms, as suggested by our data demonstrating variable expression of hClC-2 protein in CF nasal polyps and must also be considered as a mechanism whereby, CLC-2 could alter the course of CF disease. Haug et al. recently identified mutations in the CLC-2 coding region that are associated with idiopathic generalized seizures in humans [39
]. No lung disease was reported from loss of function of CLC-2, so presumably CLC-2 is not critical for the function of mature respiratory epithelium when CFTR is present. A ClC-2 knock-out mouse shows severe degeneration of the retina and testes, but no evident lung disease [28
]. While there has been no report of ClC-2 lung abnormality in these mice, they do not replicate the human seizure disorder and mouse models do not exclude the possibility of a role in airway epithelial ion transport. For example, initial studies of CF knock out mice also suggested no discernible lung disease that mimics CF in humans [41
]. The activation of CLC-2 currents by acidic pH, suggests that alterations of key regulatory domains of the channel may affect function. There is disagreement about whether or not a specific region of the N-terminus of CLC-2 is the sensor for acid and voltage regulation [43
This study provides important information about the human CLCN2 genomic organization. Several polymorphisms of key regulatory domains of CLCN2 were identified in a cohort of subjects with cystic fibrosis, who carry the same CF genotype. While we have found no significant association of CLC-2 polymorphisms with FEV1 % predicted in adulthood, further study of potential polymorphisms in CF subjects at an earlier age and investigation of potential mutations in the coding region of CLC-2 that would lead to enhanced transepithelial chloride transport would be necessary to determine if CLC-2 can modify CF.