(A) Cartoons showing putative helical structures of part of TM6 (341–354). Structure of the cytoplasmic part of TM6 was published by Serohijos et al. (2008). Six reactive sites identified by this study were represented as colored sticks. Position 349, albeit not reactive (indicated by parentheses), was included for clarity. Figures were prepared with PyMOL (V0.99; Schrödinger). (B) A continuous single-channel recording of cysless/I344C showing a dramatic increase of the spontaneous ATP-independent gating after MTSET modification. Before MTSET modification, negligible opening events were observed in the absence of ATP. After MTSET modification, the channel activity without ATP is nearly the same as unmodified channel in the presence of ATP. This phenotype can be reversed by DTT. For spontaneous ATP-independent openings, the mean open time is 0.38 ± 0.02 s (n = 3) before modification (because these openings are rare, we collected events in patches yielding macroscopic currents after ATP was removed as shown in Fig. 10 and performed statistical analysis) and 0.72 ± 0.06 s (n = 5) after modification. Similar results were obtained for cysless/M348C: 0.36 ± 0.03 s (n = 3) before and 0.55 ± 0.03 s (n = 3) after modification. (C) Homology model of CFTR using the structure of Sav1866 as a template (adopted from Serohijos et al., 2008). The coordinate of the homology model was published by Serohijos et al. (2008). Two horizontal lines depict the possible boundaries of the membrane. TMD1 and NBD1 were colored yellow, TMD2 and NBD2 were colored green, and the R domain was colored cyan. Part of TM3 (residues 199–217) was removed to show a clear view of TM6 (residues 330–352, blue). S341 and R352 (red) were represented as sticks.