11. REPLICATION OF DNA IN CHROMOSOMES 107 



strands (DNA double helices perhaps with some associated histone) 

 would have a molecular weight of 6-8X10" (Taylor, 1960b) and 

 presumably would represent the replicating unit in amoeba. 



The lampbrush chromosomes seen in meiosis may be a similar type 

 of structure except that the axial element has a much greater density to 

 the electron beam CMoses, 1956, 1960). Recently Bullivant (1962) has 

 demonstrated helically coiled filaments in metaphase chromosomes of 

 mammalian cells. The gyre diameter is about 75 A and these helices 

 occasionally appear to be coiled into a super helix comp.irable to those 

 found in the amoeba. The chromosomes in amphibian oocytes have loops, 

 the axis of which appear to be single DNA double helices (Miller, 1962). 

 These could be the pairs of replication units completely uncoiled which 

 have opened up into rings. When stretched along the axis they break 

 (Callan, 1956) at what would be the 3'R linkers in the model. 



The pairs of replication units in the model could also correspond to 

 the four-stranded DNA units reported by Cavalieri and Rosenberg 

 (1961a,b). If a chronuUid built according to the model is stripped of its 

 protein (including the H linkers) during Go, its weak points will be the 

 5' linkers. The separated units will consist of rings in which one 

 polynucleotide chain is interrui)ted with two free 5' ends. If one assumes 

 some type of biunial bonds the two double helices may remain paired. 

 In rapidly dividing cells with a veiy short G^ stage most of the DNA 

 might exist in this state. Upon heating the rings could open and allow 

 unwinding of the polynucleotide chains, with a decrease of the molecular 

 weight by a factor of two and the separation of chains of different 

 densities. Thus the observations of Meselson and Stahl (1958) on the 

 separation of hybrid molecules into single chains could apply to four- 

 stranded DNA as well as two-stranded DNA. 



The model has at certain stages some characteristics of a model 

 composed of tandemly linked rings, a model which has been suggested 

 l)y Stahl (1961) to explain recombination events in higlier organisms. 

 However, at other stages the model acts like a side-chain model for 

 purposes of genetic recombination (Taylor, 1962). 



Very little can be said at this time about the nature of the hypo- 

 thetical 3', 3'R, and 5' linkers. The H linkers have been assumed tenta- 

 tively to be peptides, but it now appears that the other linkers cannot be 

 peptides. At the chromatid level the best evidence comes from the 

 report (Callan and McGregor, 1958; (lall, 1958b) that proteases and 

 RNase do not sever the loops of the lampbrush chromosomes while 

 DNase quickly disrupts them. In addition, isolated deproteinized DNA 

 is not significantly degraded by proteases. A recent report by Bendich 

 and Roscnkranz (1962) indicates that sperm DNA contains about one 



