We investigated the folding, stability, and specificity of dimerization of the neck-hinge region (residues 356-416) of the kinesin-like protein Kif3A. We showed that the predicted coiled-coil on its own (residues 356-377) will fold autonomously in solution. We then explored the ability of oppositely charged regions to specify heterodimer formation in coiled-coils by synthesizing analogs of the neck coiled-coil region with and without various negatively and positively charged extensions to the C-terminus of the neck coiled-coil and characterizing these analogs by circular dichroism spectroscopy. The charged region alone (residues 378-416) adopted a random-coil structure and this region remained unfolded in the presence of the coiled-coil: Redox experiments demonstrated that oppositely charged regions specified the formation of a hetero-two-stranded coiled-coil. Denaturation studies with urea demonstrated a decrease in coiled-coil stability with the addition of negatively charged residues in the homostranded coiled-coil; conversely, the addition of the positively charged region (residues 403-416) of Kif3A C-terminally to the neck coiled-coil did not affect coiled-coil stability. Overall, our results suggest that electrostatic attractions drive the specificity of heterodimerization of the coiled-coil, not the removal of positive or negative charge-charge repulsions, while maintaining the stability of the heterodimer compared to that of the stablest homodimer. (C) 2002 Elsevier Science (USA).