834 LIGHT AND LIFE 



the tryptophan residues in pepsin are in contact with water or a 

 similar polar solvent, whereas in chymotrypsin they are mainly in 

 contact with a nonpolar solvent. In human and bovine albumins, 

 which have the lowest non-zero content of tryptophan, no treatment 

 has revealed any contribution of tyrosine fluorescence (Teale) . In 

 fact, direct evidence of the loss of the quanta absorbed by tyrosine 

 was found, in that between 280 m^u, and 300 m^ the absolute yield 

 increases as the relative proportion of quanta absorbed by tryptophan 

 increases, while above 300 ni/A, where only tryptophan absorbs, the 

 yield is constant. The cause of the quenching of tyrosine fluorescence 

 in these proteins remains obscure. The changes in quantum yield 

 with pH for bovine serum albumin (2 tryptophan residues) and 

 human serum albumin (1 tryptophan residue) are interesting. Fluo- 

 rescence of the former is quenched in two steps, at pH 3.5 and pH 

 0.5; of the latter in one step, at pH 0.5 (A. White) . One tryptophan 

 residue in the bovine albumin molecule must therefore be close to 

 a proton donor (glutamic acid or aspartic acid carboxyl group) ; the 

 other, like that in human albumin, must be more distant from such 

 groups. Lysozyme has steps in the quenching of fluorescence on both 

 acid and alkaline sides of pH range. Such observations clearly indi- 

 cate differences in the accessibility of different tryptophan residues to 

 protons, and exclude the possibility that protons migrate freely over 

 the protein surface. 



The polarization spectra of proteins containing tryptophan are 

 essentially like that of tryptophan itself, but with a proportionately 

 higher maximum at 305 ni/x than at 270 ni/x. Minor differences exist 

 in the relative magnitude of polarization at 305/270 m^, the presence 

 or absence of a band at 295 m^, as in N-glycyltryptophan, and the 

 fall in polarization above 310 ni/x. These distinctions are largely 

 wiped out in 8 M urea — reversibly in bovine and human serum al- 

 bumin, irreversibly in rabbit muscle lactic dehydrogenase. A high 

 305/270 mix polarization ratio occurs in human serum albumin, pos- 

 sessing only a solitary tryptophan residue, and hence cannot be taken 

 to signify a transfer of the excited state between tryptophan residues. 

 It is best explained as a result of interaction between the tryptophan 

 residue and the protein backbone. Since the fluorescence of each 

 tryptophan residue can be quenched independently by pH in bovine 

 serum albumin and lysozyme, no transfer of excited state occurs be- 

 tween the tryptophan residues in these proteins. 



