Effect of connexin composition on pore structure of gap junction channels and their permeability to metabolites
MetadataShow full item record
Knowledge of actual signals and the rate at which they are transferred between cells by gap junctions has long stood as an impediment to developing a detailed understanding of the function of different connexins in various organs. To bridge this "gap", we have developed a method for measuring single channel flux rates of natural metabolites through gap junctions composed of different connexin isotypes expressed in Xenopus laevis oocyte pairs. Initially single channel flux rates of a non-hydrolysable variant of ATP (ATPγS) were measured for Cx26, 32 and 43 channels, and found to vary by over an order of magnitude (10 4 to 4.6x10 5 molecules/channel/sec, after normalization to a 1 mM concentration gradient), with Cx43 channels being the most permeable. These assays were then expanded to include radiolabelled AMP and cAMP. By reducing the temperature to 8°C, and including the phosphodiesterase inhibitor IBMX in the case of cAMP, metabolic conversion of these permeants could be largely prevented over the course of the experiment. A comparison of the fluxes of these three metabolites through Cx26, 32 and 43 channels, after normalization for concentration gradient and temperature, yielded several novel conclusions. Cx26 was consistently ∼2.5 fold more permeable than Cx32 for all metabolites examined, presumably reflecting the 2.5 fold difference in single channel conductance between these two connexins. ATP and AMP were 3-23 fold more permeable through Cx43 than either Cx26 or Cx32 channels. By contrast, cAMP was 2.5 fold more permeable through Cx26 than Cx43 or 32 channels. These results clearly demonstrate that different connexins form channels that show specific preferences for metabolites. The subtlety of this distinction is most evident in the 30 fold difference in permeability of AMP and cAMP through Cx43 channels, indicating that these channels can distinguish a single phosphodiester bond. We have shown that gap junction channels discriminate between permeants according to size and metabolite type. Understanding of the molecular basis of this selectivity has been hampered by a lack of information on comparative pore structures of different connexins. Using the substituted cysteine accessibility method (SCAM), we have previously mapped the pore lining residues of open Cx32 gap junction channels with a biotinylated maleimide reagent (MBB) [Skerrett et al., (2002)]. Reactivity was identified along one face of M3, and in the cytoplasmic half of M2. Reactivity in M1 was difficult to interpret, as cysteine substitution at many of these sites significantly modified channel gating, suggesting that the mutated protein was not in a wild type conformation. We have now completed a comparison with a second connexin, Cx50, with a much larger single channel conductance than Cx32 (220 pS, compared to 55 pS), yet a significantly lower size cut-off for dyes (∼600 compared to >760 MW). Cysteine substitutions in M1 of Cx50, unlike Cx32, induced no phenotypic gating changes, but also showed no reactivity, indicating that M1 contributes minimally to the pore lining of open gap junction channels. In contrast, six sites in M3 of Cx50 were reactive, with 5 corresponding to sites mapped in Cx32, and one towards the extracellular end of the pore being shifted by one residue from the pattern in Cx32. Four sites were also reactive on one face of the cytoplasmic end of the M2 helix in Cx50, but of these, only one corresponded to a site mapped in Cx32. The patterns of reactivity indicate that the M2 helix may line the pore at a different angle in the two connexins, consistent with the hypothesis, based on comparative permeability studies, that gap junction pores of different connexins may differ significantly in their topology. The final part of this dissertation focuses on the use of connexin mutants as tools for structure studies (Cx26M34A), to discriminate the functions of connexins in terms of their channel and non-channel physiological roles (Cx45.6E48K) and the use of a conserved dominant negative threonine mutant that can act as a suppressor of channel activity in cells (Cx26T135A, Cx43T154A, Cx50T157C). In these studies we test the functionality of these mutants electrically and with dye permeability assays to determine whether the mutated connexin renders the channel in a closed state, what type of closed state it is, and the relationship the connexin mutant has with its wild type variants. The goal here is to demonstrate the usefulness of connexin mutagenesis in understanding the function and roles of gap junction channels and their hemichannel counterparts.