several hypotheses have been considered. Among 

 these is the proposal by Fritz and Jacobson 

 that the subbanding in mouse tissues is the 

 result of the differential binding of NAD by the 

 subunits (21). The possibility that a small mole- 

 cule such as NAD, by becoming attached to the 

 subunits, can change the net charge, and hence 

 the mobility of an isozyme, is certainly not un- 

 reasonable. However, this hypothesis was not 

 supported in identical experiments with rabbit 

 LDH. Another interpretation proposes that sub- 

 bands represent permutations of the tetrameric 

 combinations. This is supported by the observa- 

 tion that the mixing and recombination of rabbit 

 LDH-1 and LDH-5, which in themselves show no 

 subbanding, yields subbanding at the LDH- 3 posi- 

 tion. Kaplan and Costello have advanced the 

 hypothesis that the subbanding in mouse LDH re- 

 sults from the existence of two different A sub- 

 units each of which is under the control of a 

 separate gene (22). This interpretation is 

 strongly supported by numerous observations of 

 the existence of the subbanding in the pattern 

 predicted for two different A subunits. The 

 presence of such patterns in inbred strains of 

 mice rules out heterozygosity as an alternative 

 and suggests the existence of a fourth gene con- 

 trolling the synthesis of LDH polypeptides. 

 Clearly no single one of these interpretations 

 fits all of the data. Indeed, there may be no 

 all- encompassing explanation. 



The existence of the X-bands and subbands 

 tends to emphasize the fact that starch- gel 

 electrophoresis resolves mammalian LDH into 

 five major zones of activity. However, the iso- 

 zyme pattern can differ considerably among 

 mammals (Fig. 2) and among different vertebrate 

 classes, as shown in Fig. 10. It is of interest to 

 note that the net charge on the B subunit (the 

 more negatively charged subunit) of mamalian 

 LDH is apparently greater than that on the 

 homologous subunit of most other vertebrate 

 classes as reflected in the greater mobility of 

 mammalian LDH-1. However, the A and B sub- 

 units of the vertebrate classes, excepting some 

 fish, must be remarkably complementary in that 

 most can be combined to form functional hybrid 

 molecules (tetramers) of LDH by means of the 

 salt-freezing technique as illustrated in Fig. 11. 



PAPACONSTANTESfOU: Aren't there more 

 than five bands in the sixth column from the 

 left? Yet you started out with pure LDH. 



MASSARO: Yes, there are more than five 

 bands. I started out with pure beef LDH- 1 which 

 was hybridized with rattlesnake muscle LDH and 

 electrophoresed. Muscle of this species of 



rattlesnake contains several LDH isozymes, 

 seven, in fact. 



PAPACONSTANTINOU: Then are the com- 

 plementary LDH-l's combining? 



MASSARO: Yes, however when complemen- 

 tary LDH-l's are hybridized, if they have very 

 close mobilities, the hybrid isozymes do not 

 separate into distinct bands in our electro- 

 phoretic system. 



Let us divert for a minute to a fish story. 

 We have studied, to date, approximately thirty 

 species of fish and have found that they can be 

 placed conveniently into three categories 

 according to the number of isozymes of LDH 

 that they possess and the hybridization charac- 

 teristics of these isozymes. Those fish pos- 

 sessing a single band of LDH activity, as re- 

 vealed by starch gel electrophoresis, are placed 

 in one category. This group consists of the 

 fluke and related flatfish. In another category 

 are place those fish possessing either two or 

 three bands of LDH activity. There are some 

 twenty-plus species in this group, evenly dis- 

 tributed between the two and three banded varie- 

 ties. The third category consists of those fish 

 possessing more than three bands of LDH 

 activity. So far we have placed only three species 

 in this group, the herring (Alosa aestivalis) , the 

 shad (Alosa sapidissma), and the whiting (Mer- 

 luccius bilinearis). 



Under our conditions, any two of the LDH 

 isozymes of the herring will readily hybridize 

 with one another to form the expected auto- 

 genous hybrid molecules. This is also true for 

 the isozymes of the whiting, and the shad (i.e., 

 those fish possessing three or more bands of 

 LDH activity). The fluke, having only one band 

 of LDH activity, obviously does not show auto- 

 genous hydridization. All four of these species 

 will also form hybrid molecules with one another 

 and with mammalian LDH. Significantly, those 

 species possessing two or three bands of LDH 

 activity will not form autogenous hybrid mole- 

 cules although they will hybridize with mam- 

 malian LDH and LDH from the two other groups 

 of fish. The factors underlying the lack of auto- 

 genous hybridization within this group are under 

 investigation in our laboratory. 



Another interesting aspect of this study was 

 the discovery of a very rapidly migrating band 

 of LDH activity in the eye of many species of 

 fish. This band has a mobility greater than that 

 of mammalian LDH-1 and, like the C tetramers 

 of sperm LDH, may represent another type of 

 LDH isozyme. 



Further evidence of the remarkable com- 



84 



