A series of recombinant peptides, each including the sequence proposed to be the first nucleotide-binding fold of cystic fibrosis transmembrane conductance regulator (CFTR), has been produced in an attempt to find a model peptide that would autologously fold into a soluble structure with native-like properties. The peptide NBDIF, which contains the 267-amino acid sequence of CFTR from 384 to 650, meets these requirements. The peptide was produced with a high expression bacterial plasmid pRSET, purified from inclusion bodies following solubilization with 6 M guanidine-HCl and refolded from 8 M urea. Competitive displacement of trinitrophenol-ATP by nucleotides reveals binding of ATP and related nucleotides with KDs in the low micromolar range; the KD for ATP gamma S is 1.0 +/- 0.4 microM and for ADP 8.8 +/- 3.1 microM. The native-like character of the model peptide's structure is further supported by the findings that the KD for the ATP analog, 5'-adenylimidodiphosphate, is fourfold lower than the KD for the methylene analog, 5'-adenylmethylenediphosphonate, and that ATP binding slows the trypsin proteolysis of NBDIF. The CD spectra of NBDIF and the parallel peptide containing the most common cystic fibrosis mutation, deletion of Phe 508, are essentially indistinguishable, both spectra indicating 28% alpha-helix and 23% beta-sheet, with insignificant differences in the amounts of beta-turns and random structure. Extensive investigation using multiple conditions with highly purified preparations of the model peptides demonstrates that they do not support ATP hydrolysis. These large recombinant peptides offer practical models for the investigation of the first nucleotide-binding domain of CFTR.