Aequorin emits light in the presence of Ca2+, decomposing into apoaequorin, coelenteramide and CO2. Semisynthetic aequorins, produced by replacing the coelenterazine moiety in aequorin with analogues of coelenterazine, showed widely different sensitivities to Ca2+ as well as certain spectral variations. A group of semisynthetic aequorins, e-type aequorins, showed bimodal luminescence, with peaks at 400-405 nm and 440-475 nm in various intensity ratios, whereas all other aequorins luminesced with only one peak, in the range 440-475 nm. The cause of the spectral variation was studied by various experiments including: (1) comparison with the fluorescence of the spent solution and the luminescence of the spent solution produced by added coelenterazine; (2) luminescence in 2H2O; (3) the rate of conformational change of apoaequorin; (4) the rates of regeneration in the presence and absence of O2. The results suggested that the spectrum of Ca(2+)-triggered luminescence is strongly affected by the ionic charge on the amide N atom of the coelenteramide that is bound to apoaequorin. When the amide N atom is negatively charged, light is emitted with a 440-475 nm peak. In the case of e-type aequorins, the negative charge on the amide N atom is less because of the structure of e-coelenterazine involved, resulting in the emission of a 400-405 nm peak from the uncharged form of coelenteramide; the intensity ratio of 400-405 nm peak to 440-475 nm peak is determined by the amount of negative charge resting on the amide N atom of e-coelenteramide at the time of light emission. Most of the spectral variations in luminescence and fluorescence can be explained on the basis of ionic and hydrophobic interaction between a coelenteramide and apoaequorin.