Early life of cystic fibrosis transmembrane conductance regulator (CFTR) in the cell

Growing up to be a mature and stable ion channel appears to be especially difficult for the cystic fibrosis transmembrane conductance regulator (CFTR)1. Whereas other adenine nucleotide binding cassette (ABC) transporters including members of the same subfamily (ABCC) mature efficiently, only about 30% of wild-type nascent CFTR polypeptides mature and proceed beyond the endoplasmic reticulum (ER) in mammalian cells. The remaining 70% are polyubiquitinated and degraded by the proteasome2. The F508 present in 90% of cystic fibrosis (CF) patients and many different missense mutations completely preclude maturation. Although detectable at steady state, the immature core-glycosylated form of neither the wild-type nor F508 CFTR accumulate in the ER. There is some evidence from both mammalian cells and yeast that nascent CFTR may be "retrotranslocated" from the ER membrane enabling proteolysis by ER-associated proteasomes. Crucial questions that need to be answered to understand the biosynthetic lability of CFTR include (1) what are the features of the molecule that prevent it from maturing efficiently, and (2) how are molecules that can mature distinguished from those that cannot? CFTR most obviously differs from other ABCC family members by the presence of charged amino acids in membrane-spanning sequences, the large central R-domain rich in phosphorylation sites and an extended C-terminal domain beyond the second nucleotide-binding domain. Phosphorylation sites of the R-domain apparently play no crucial role in maturation3. C-terminal sequences were shown to impact maturation; however, a truncated CFTR molecule without C-terminal domain shows still a substantial degree of maturation4. Thus, neither the R-domain phosphorylation nor the C-terminal extension appears to be exclusively responsible for the failure to mature. Mutations identified in patients with cystic fibrosis are distributed all over the CFTR gene. Examination of more than 40 missense mutations indicated that a large proportion (60%) of those in cytoplasmic loops caused misprocessing, while none in the extracytoplasmic loops had this effect5. Hence, detection appears to occur on the cytoplasmic rather than the luminal side of the ER membrane. Consistent with this interpretation, although nascent CFTR normally binds calnexin, the conformation-sensing mechanism that this chaperone provides together with UDP-glucosyl glucose transferase, does not seem to be involved since prevention of glycosylation neither interferes with transport of wild-type CFTR to the cell surface nor enables F508 to do so. The cytoplasmic chaperones Hsp70 and Hsp90 bind nascent CFTR and inhibition of the latter interaction with ansamycin drugs further disrupts maturation and accelerates proteolysis6. As a Hsp90 substrate, nascent CFTR is typical of a small set of proteins that mature with difficulty. Because missense mutations at nearly all locations along the polypeptide may prevent maturation, it seems likely that achievement of a native overall tertiary structure is the criterion that must be met to enable export from the ER. Sequence changes elsewhere in the molecule can suppress or compensate for the null export phenotype of F508. An example is the combined substitutions R29K and R555K that rescue at least one third of F508 molecules7.