Chapter 24 

 CHROMOSOME MAPS 



TETRAD ANALYSIS AND CENTROMERE LINKAGE 



By tetrad analysis it is possible to determine how closely two 

 different freely assorting genes are linked to their respective cen- 

 tromeres. We have shown in Neurospora that the centromere under- 

 goes reduction at the first division and, therefore, a gene close to 

 the centromere usually undergoes reduction at the first division 

 (Chapter 23). Amatingof aG x ag produces three kinds of tetrads: 

 (I)aG, aG, ag, ag, (II) aG, aG, ag, ag, (IH) aG, ag, ag, aG. Type I 

 contains only parental combination spores, type n contains only 

 recombination spores, and type III contains all four possible types. 

 If types I and II are equal, the total number of original and recom- 

 bination spores will be equal, proving that the genes are not linked. 

 The "checkerboard" shown in the Table 24-1 proves that the random 

 expectation of 1:11:111 equals 1:1:4. Fig. 23-1 shows that 6 kinds of 

 tetrads are produced when the genes are so far from their respective 

 centromeres that the random expectation of first- to second-division 

 segregation occurs and when orientation in the ascus is determin- 

 able and the spindles do not overlap. Although order is not d ter- 

 minable in Saccharomyces because the ascus is a small oval cell, 

 the checkerboard serves to calculate the expected frequencies of 

 the 36 different types of tetrads and from this, the 3 distinguishable 

 types can be calculated. The first row at the top of Table 24-1 

 represents the 6 types of asci with the haploid nuclei in the 6 possi- 

 ble arrangements in the ascus (fig. 23-1). The column at the left 

 side represents the arrangements possible for the second gene pair. 

 Thus 12 arrangements combine to produce 36 total kinds of asci. 

 Of these, 6 are type I, 6 are type II, and 24 are type III, so the ran- 

 dom expectation of the 3 types is 1:1:4. Many tetrad analyses of 

 two gene pairs yield the 1:1:4 ratio, but the aG x ag heterozygotes 

 produced a total of I:n:m = 79:81:167. The equivalence of type I 

 and II prove that the a/a and G/g loci are not linked but the deviation 

 of tetrad types from random expectation prove that some mechanism 

 has prevented random distribution. We have interpreted this to mean 

 that both loci are relatively close to their respective centromeres 

 causing them to segregate reductionally at Meiosis I with higher than 

 random frequency. 



An examination of Table 24-1 reveals that the combination of 

 first-division segregation of A/a and B/b produce only tetrads of 



24-1 



