Scientia Sinica

106 SCIENTIA. SINICA Vol. V

ratios near or greater than 20; (4) the relative stability to organic solvents and to dilute acid; and (5) definite electrophoretic mobility.

The ratio of a/b and molecular weight differs slightly from one tropomyosin to another; this may be a reflection of minor differences in the polypeptide chain folding. Prawn tropomyosin seemed to suffer a change of configuration when the ionic environment was altered. However, it is difficult to agree with Wassermann'" to the extent that rabbit tropomyosin molecules are flexible chains capable of marked changes in configuration. Such a concept, drawn from theories of simple synthetic polymers, is probably applicable to proteins in an unfolded, denatured condition, and is difficult to reconcile with the crystallizability and extraordinary stability!) of the tropomyosin molecule.

Extensive, systematic and comparative studies of the molecular dimensions of any one kind of crystalline proteins are lacking. The few scattered. investigations pertain to proteins isolated from sources which are very closely related in species.

Relatively more attention, however, has been paid to haemoglobin. There exists a definite integral relationship among different haemoglobins and their subunits, a situation well summarized by K. O. Pedersen!” in his restatement _ of the Svedberg multiple concept: “.. .similar proteins, i.e., proteins that occur in the same way or have the same function, will either have about the same molecular weight or they will usually belong to the same multiple system.” With the appearance of the extensive X-ray crystallographic work of Perutz'! and Porter and Sanger’s elegant terminal analysis of haemoglobin", the above-mentioned relationship becomes immediately understandable in terms of a limited number of polypeptide chains arranged in the forms of primitive laminae or rodlets in the larger globular protein molecules. The structural principles of asymmetric protein molecules are far less known. There appears to be no simple multiple relationship among the molecular weights of tropomyosins from different sources. It is conceivable that these different tropomyosins are structurally quite similar, but that they are elaborated to different lengths along the long axis.

According to Ellenbogen and Olson", the molecular weight of dogheart myosin is 100,000. This is only about 1/8 of that of the same protein in rabbit skeletal muscle and is smaller than the subunit weight, 165,000, deduced by Tsao!!. On the other hand, their value is identical with that of tryptic or chymotryptic L-meromyosin?® 7! and is not far off from the molecular weight of pig-heart tropomyosin, 90,000, deduced in the present investigation. For rabbit skeletal muscle, the size of the myosin molecule is some 16 times that of tropomyosin monomer. It is this disparity in size and partial similarity in amino acid composition and physico-chemical properties that make it not illogical to think of tropomyosin as a possible subunit,