SIDNEY I'. I'ELICK 129 



Protein subimlts or domains of interaction. When the fluorescence 

 of a protein with nuiUipIe binding sites is quenched by addition of 

 Hgands there is no a priori reason for assuming that the amount of 

 quenching should be Hnearly proportional to the number of ligands 

 added. For examjile, the first DPNH molecule to become attached 

 to a protein might in principle capture the major portion of the 

 protein excitation energy. Subsequent molecules of DPNH might 

 bind in statistically normal fashion but quench a disproportionately 

 small amount of the protein emission. If one plotted a titration 

 curve obtained under such conditions the apparent stoichiometry 

 would be low and the binding sites would not appear to be equivalent 

 and independent. In practice this has not happened (35) . The titra- 

 tion curves of enzyme with DPNH, using either the quenching of 

 protein fluorescence or the directly excited nucleotide emission as 

 indicators, are strictly concordant with respect both to stoichiometry 

 and to apparent intrinsic dissociation constants. It must be concluded 

 therefore that each molecule of nucleotide quenches only the trypto- 

 phan within its own domain and that there is negligible overlap be- 

 tween domains. The proteins behave as though they possessed n-fold 

 symmetry corresponding to n binding sites, with n equivalent subunits, 

 real or virtual. 



Although the proteins that have been examined in the present 

 studies may indeed be specific aggregates of subunits, no direct evi- 

 dence for subunits has been provided, and if they occur it is unlikely 

 that they dissociate under the conditions of the fluorometric measure- 

 ments. GPD, for example, shows no sign of dissociation by light- 

 scattering measurements at concentrations down to 100 ^ug per ml. 

 Glutamic dehydrogenase is a rather large aggregate, of molecular 

 weight 1,000,000, which dissociates at high dilution; and Frieden (10) 

 has found that this dissociation is promoted by the binding of DPNH 

 at specific inhibitory sites. The subunits in this case are, however, 

 of large molecular weight (about 250,000) and still contain 3 to 4 

 coenzyme binding sites each. The occurrence of domains in GPD 

 poses no serious problem because the maximal quenching of the 

 protein by DPNH is only 40 per cent. There are about 15 tryptophans 

 in GPD or 5 per binding site, and 40 per cent quenching corresponds 

 to the capture of all the excitation energy from only two of them 

 by a molecule of DPNH. The domains in this case could easily be 

 determined by proximity of the tryptophans to the binding sites. 

 The behavior of LDH-APNH is more unusual. In this complex the 

 quenching is 80 per cent. Each molecule of the bound coenzyme 



