The Diazo Reaction of Bilirubin: Structure of the Yellow Products (studies on Plant Bile Pigments-14’)

The reaction of bilirubins with aromatic diazonium salts in alcoholic solvents leads to an equimolar mixture of two types of products. One is the well-known 9-azopyrromethenone. The other is a yellow product (A,,, z 420nm) identified as 9-alkoxymethylpyrromethenone. the alkoxy-substitucnt being derived from the solvent. Thus, reaction of the symmetrically substituted bilirubins llla (lc) and XIIIJ (lb) in methanol with diazotized sulfanilic acid yields one mole of the azoplgments (4b and 4a). respectively, and one mole of the corresponding 9-methoxymethylpyrromethcnones (2b and 2a). Bilirubin IXa (la) consequently yields a mixture of all four products. The two resulting 9-methoxymethylpyrromethenoncs were separated by chromatography and identified as 2a and 2b. They can react further with the dia7onium salts lo give the corresponding 9-aznpyrromethenones, but the rcactlon is much slower than that of bilirubin. which explains the observed product distribution. These results are d~scusscd in relation to earlier work. ALMOST a hundred years ago Ehrlich’ discovered the formation of a violet colour when a solution of bilirubin was treated with diazotized sulfanilic acid. Fischer and Haberland3 demonstrated that bilirubin was cleaved around the central methylene bridge by the aromatic diazonium salts thus forming a pair of isomeric azodipyrroles. It was soon recognized that this reaction was of practical use for the qualitative and quantitative determination of bilirubin in body fluids“ and today it is used widely for this purpose. It is also used as a research tool for the structural investigation of naturally occurring bilirubins and their conjugates.5 We recently used the diazo reaction as a novel degradation technique for biliproteins.6 In a two-step reaction the verdin type chromophore of denatured phycocyanin was reduced with NaBH, to the corresponding rubin and the “phycorubin” formed was subsequently treated with aromatic diazonium salt. The amount of azopigment found, however, never exceeded SO’>; of the theoretically expected value, if a complete conversion of the “phycorubin” to azopigments was assumed. Instead a second product with i.,,, = 400nm was obscrved.6 It was shown that this was not due to an unusual reactivity of the phycorubin chromophore since bilirubin IXa gave similar results. To establish the nature of the yellow products, we have now studied the reaction under different conditions. Spectroscopic studies. When bilirubin IXr (la, concentration range 14-23~tM) in methanol or ethanol containing a trace of dimethylsulfoxide was treated with diazotized sulfanilic acid (aqueous solution in 0.1 N HCI), the colour changed from yellow to orangered. The bilirubin band at 451 nm (MeOH) or 446 nm (EtOH) disappeared, and new bands arose instead at 518,415 and 330 nm (MeOH), or 524,420 and 330nm (EtOH). The 330nm band was only resolved in the presence of a small excess of diazo reagent. due to the interfering absorption of the latter at 268 and 297 nm.’ Isobestic points were observed at 487 and 381 nm (MeOH) or 486 and 398nm (EtOH). Identical absorption maxima were eventually obtained, irrespective of a I .2-, 4or 6-fold molar excess of diazonium salt over bilirubin being used. The reaction was accelerated, however. by an increasing excess of reagent. With only slight excess of diazonium reagent (1.2 molar ratio) it took approximately 30 min to reach the point where no further significant spectral changes could be observed. whereas this part of the reaction was already complete after mixing and recording the spectrum with a 4or 6-fold excess of diazonium reagent. These rapid spectral changes were followed by a much slower phase. during which the