Serum Vasoconstrictor (Serotonin)

In the fourth paper in this series (1) the isolation of the substance believed to be responsible for the marked vasoconstrictor activity of serum was reported. This substance was characterized as an indole derivative from its ultraviolet absorption spectrum and its behavior in various color and other chemical reactions. The analytical data agreed well with the empirical formula C14Hz103Nb.H2S04. The titration curve in aqueous solution revealed two titratable groups with PK’~ of 4.9 and pK’2 of 9.8. Both of these groups were believed to be basic in nature, and the pK’ value of the weaker suggested an acylated guanidine structure. Indeed, the value was very close to that reported in the literature for creatinine, while the formation of ammonia on treatment with alkali and the presence of one N-methyl grouping suggested that a creatinine residue might well be present in the molecule. The fact that only exceedingly small quantities of pure substance were available led to an investigation of this possibility by calorimetric means, and it was quickly determined that serotonin sulfate responded positively to both the Jaffe alkaline picrate reagent and the Benedict alkaline 3,5dinitrobenzoic acid reagent, with a color intensity equal (within an experimental error of about 10 per cent) to that of creatinine on a molar basis. Benedict’s study (2) of the specificity of the alkaline 3,5-dinitrobenzoic acid reagent with respect to creatinine showed that derivatives and closely related compounds such as glycocyamidine could be readily distinguished by measuring both the intensity of color and the rate of color development. Comparison of serotonin sulfate with creatinine in this fashion again revealed almost identical behavior. It then became necessary to investigate the question of whether a creatinine-like residue was actually part of the molecular structure of serotonin, or whether, as now seemed more likely, the entity isolated and designated as serotonin sulfate was not in reality a complex containing an indole base, creatinine, and sulfuric acid. The behavior of serotonin on a paper chromatogram developed with butanol showed that the latter was indeed the case, for the creatinine chromogen was completely separable from the indole chromogen.