IX. GENETICS AND HUMAN HEMOGLOBIN CHEMISTRY 439 



and Torbcrt (1958) with three hemoglobins. Hemoglobin Hopkins-2 has 

 been found to be altered in the a chain by dissociation and reassociation 

 experiments (Singer and Itano, 1959; see Section IV); the alteration of 

 the second abnormal hemoglobin, Hb-S, observed in the family is known 

 to reside in the /3 chain. Hb-S and Hb-Hopkins-2 have altered electro- 

 phoretic mobilities; Hb-S has a +2 charge and Hb-Hopkins-2 a 

 —2 charge with respect to Hb-A. The doubly abnormal hemoglobin 

 Q,^^Hopkins-2^^s happens to have the same charge as Hb-A. Itano and Rob- 

 inson (1960) have shown that this hemoglobin is present in the indi- 

 viduals with three hemoglobins of the family reported by Smith and 

 Torbert (1958). 



The occurrence of hemoglobin molecules abnormal in both peptide 

 chains allows one to make certain deductions about the mode of assem- 

 bly of the hemoglobin molecule. The persons with four hemoglobins are 

 doubly heterozygous, having the genotype a'^'/a^ /^Vj^"^; as suggested by 

 Atwater et al. (1960a) or aV«"°'"''"'"' )SVj8^ (Itano and Robinson, 1960). 

 Apparently each gene causes the manufacture of a peptide chain charac- 

 teristic of it, so that four types of peptide chains are produced in these 

 persons. It has to be assumed that each gene controls the synthesis of a 

 dimer, rather than that of single chains. Hemoglobin molecules having 

 two different a or [3 chains, such as a^ a^ fSn^ or a.'^ (^^ (3^, have never been 

 observed (Itano, 1956). The fact that each gene controls the synthesis 

 of a dimer may be a consequence of the way in which the hemoglobin 

 molecules are synthesized. A possible explanation is that the chains 

 have to dimerize in order to be released in solution. In this view each 

 microsomal particle has the information for only one type of peptide 

 chain or possibly several copies of the same peptide chain; the newly 

 synthesized chains are scarcely soluble and are released in solution only 

 upon dimerization. The different types of dimers present assemble in a 

 random fashion to complete hemoglobin molecules, leading to all the 

 possible combinations (see Fig. 9). 



An alternative explanation is that each chain associates with an 

 identical partner because of steric hindrances, which prevent association 

 of non-identical partners. It is rather difficult to hold such an explana- 

 tion in view of the rather limited alterations in three-dimensional struc- 

 ture that certain Hb-A variants seem to have. Hb-A and Hb-Dpunjab, for 

 instance, differ by only one amide group; a glutamic acid in Hb-A is 

 changed to glutamine in Hb-Dpimjab in position 121 of the ^ chain 

 (Baglioni, 1962b). It seems unlikely that this limited alteration could 

 prevent association of the /3'^ chains with the /J'^punjai. chains. 



According to the former explanation for the specificity of the hemo- 

 globin chain dimerization, the existence of a hemoglobin molecule having 

 two different types of « or ^ cliains should be possible. No such molecule 



