Cystic fibrosis (CF) is a progressive disease primarily affecting the intestines, lungs, and pancreas. The gene responsible for CF was identified in 1989 1 as coding for the cystic fibrosis transmembrane conductance regulator (cftr), a membrane chloride channel. CF is one of the most common autosomal recessive diseases in Caucasians with a carrier rate of 3–4% 2, and is characterized by recurrent infection and chronic inflammation. Recently it was found that infants with CF demonstrate changes in forced expiratory volume in 1 second (FEV₁), functional residual capacity (FRC), and other parameters of lung function prior to the onset of recurrent infection 345.

Soon after the CF gene was discovered, a knockout mouse was developed. This mouse demonstrates subtle changes in epithelial cell phenotype, including alterations in secretory glycoconjugates and changes in secretory vesicles 6. Monocytic infiltrates and altered lung mechanics have also been found 7. Unfortunately, the cftr knockout mouse does not develop overt lung disease, which has severely limited its usefulness. However, the availability of new methods for pulmonary testing in rodents 89 now presents the opportunity to re-examine the cftr knockout mouse for functional lung changes. In the present study, therefore, we examined pulmonary function in young adult cftr -/-, cftr +/-, and cftr+/+ S489x mice in an effort to establish a lung phenotype.

The usefulness of the cftr knockout mouse as a model of cystic fibrosis has been severely limited by its failure to demonstrate readily measurable lung disease, the primary cause of morbidity and mortality in humans 13. However, in the present study use of sophisticated measurements of lung function revealed a functional lung phenotype in the knockout mouse (Table 1); the complete absence of cftr in the lung of young adult animals resulted in decreased Cst and η and increased Raw, G and H as compared to normal littermate controls.

A particularly intriguing further observation was that Cst and hysteresis in cftr+/- mice was significantly higher than in cftr+/+ animals while G and H were decreased. As this was not associated with any pathology such as emphysema, we conclude that it represents a functionally different lung from that of the cftr+/+. Our data thus reveal a remarkable inverse correlation between the effect of one and two non-functional copies of the cftr gene.

What do these data mean in terms of lung structure? The knockout animal has cystic fibrosis by definition, and our data now show it to also have lung disease manifest as a reduced compliance and increased resistance. These changes could reflect changes in the intrinsic mechanical properties of the parenchyma, or simply a reduction in lung volume. The former effect could include alterations in the biophysical properties of the air-liquid interface in the lungs, and would be expected to result in a change in η14. Indeed, because cftr is a chloride channel and is thought to be involved in water balance, a change in surface tension in the lung, and consequently in η, might be expected. However, as shown in Figure 2 and Table 1, although G and H both increase, they do so in the same proportion so there is no significant change in η between the three cftr genotypes. On the other hand, lung weights were not different among the different groups of mice, so the decreased compliance and increased resistance of the cftr-/- animals was not simply due to their having smaller lungs than control animals. This suggests that the parenchymal structure in the lungs of the homozygous and heterozygous animals were organized differently, in a manner that affected G and H similarly.

As documented in numerous publications, the mouse strain used in the present study does not develop chronic inflammatory disease up to the age (30–60 days) used in this study (for review see 15). On the other hand, Broaches-Carter et al. 16 have shown that cftr levels are highest in the developing lung and decrease 75-fold at birth. In utero over-expression of cftr has also been shown to affect lung growth and development17, and the severity of disease in the knockout mouse has been shown to be influenced by genetic background 18. These data thus suggest that cftr may affect the early development of the lung in a manner that is affected by the interaction of other genes.

Are there any functional consequences for increased lung compliance in the heterozygous cftr animals? Interestingly, there is no decrement in lung function in human heterozygotes 192021. Also, the heterozygote frequency for CF in humans is higher than expected, likely reflecting a selective advantage because there is no evidence for genetic drift 2223. Indeed, selective advantage in CF has been proposed to reflect resistance to tuberculosis, influenza and cholera 24. When one looks in nature for other examples of heterozygous advantage, the sickle cell trait which confers resistance to malaria 25 is perhaps the only such recognized genetic trait in humans. In Norway rats, a single Mendelian gene controls resistance to Warfarin, an anticoagulant used to control rat populations; homozygous wild-type rats are killed by Warfarin and homozygous mutant allele animals are highly susceptible to vitamin K deficiency 26. The results of the present study indicate that a similar selective advantage may pertain to cftr, something we term a "Goldilocks Effect". That is, while two defective copies of the gene are detrimental and two normal copies are satisfactory, one normal and one defective gene may results in an optimal dosage for lung development.

Further studies of pulmonary mechanics in cftr knockout mice should reveal additional genetic loci that modulate the influence of cftr on lung growth and development. Corresponding studies in humans should be useful in evaluating the effect of therapies on reversing altered pulmonary function in the CF patient.

Using sophisticated techniques to a evaluate rodent pulmonary function; a distinct, readily quantifiable lung phenotype was identified in the cftr knockout mouse. In addition, the cftr+/- mouse had a distinguishable pulmonary function phenotype from that observed in either the homozygous normal or mutant genotype mice. These data are consistent with CFTR-dependent, physiologic changes in the structure and function of the lung.