Evidence that Na+/H+ exchanger isoforms NHE1 and NHE3 exist as stable dimers in membranes with a high degree of specificity for homodimers.

In this study, we have investigated whether members of the Na+/H+ exchanger (NHE) family are oligomers and whether such oligomeric structure is required for function. Fibroblasts overexpressing NHE1 were treated briefly at 0 degrees C with the cross-linker disuccinimidyl suberate, then membranes were prepared and proteins analyzed by SDS-polyacrylamide gel electrophoresis. Disuccinimidyl suberate treatment converted quantitatively the immunoreactive monomeric form of NHE1 (110 kDa) to a putative dimeric form (210 kDa). Utilization of NHE1 mutant deleted of the cytoplasmic domain (delta 515TH) demonstrates that the transmembrane domain of the antiporter is sufficient for dimerization. Moreover, coimmunoprecipitation of NHE1 and delta 515TH, coexpressed in the same cell, formally proved the existence of dimers. This dimerization was also shown to take place with the epithelial and apically expressed NHE3 isoform, suggesting that oligomerization is a common feature of these transporters. However, coexpression of NHE1 and NHE3 in the same cells did not lead to the formation of heterodimers demonstrating an isoform specificity for the subunit interaction. The domain(s) involved in the isoform-specific dimerization is (are) likely to be confined within the transmembrane segments, as deletion of the 300 amino acids of the cytoplasmic domain did not disrupt dimerization. Exploiting the dimeric properties of the receptor tyrosine kinases and the fact that dimerization triggers kinase activity, we constructed a NHE1/insulin receptor chimera to probe NHE1 dimerization in vivo. When transfected into hamster fibroblasts, this chimera containing the N-terminal transmembrane domain of NHE1 and the cytoplasmic beta-subunit domain of the insulin receptor generates a functional transporter that is autophosphorylated on tyrosine and that presents properties of a constitutively active insulin receptor. These findings support the notion that NHE1 exists in an oligomeric state in intact cells. Finally, to test whether individual subunits of NHE1 are the minimum functional unit for Na+/H+ exchange, we coexpressed a truncated form of NHE1 (delta 515) together with an inactive mutant of NHE1 (E262I). In spite of good expression of the inactive transporter and its capacity to dimerize with active NHE1, no dominant negative effect was observed on amiloride-sensitive 22Na+ flux. This observation would suggest that individual subunits of NHE1 function independently within the oligomeric state.