The Genetic Constitution of Tandem Duplications

Methods for genetically mapping the end points of tandem -1plicaticEns c the rII region of bacteriophage T4D are presented. Analysis of ten duplications indicates (1) that the position of duplication end points and therefore the length of the duplicated segment differ in strains of independent origin; (2) that there is a direct relationship between segregation frequency and length; (3) that segregation is more frequent than expected on the basis of standard genetic mapping; and (4) that while duplications frequently include non-rII genetic material, frequently they do not include the entire rII region. The duplications studied range from less than two to about five cistrons in length. a cross between the complementary, overlapping rII deletions rJlOl and '?589, exceptional progeny arise which are able to grow on bacteria lysogenic for phage X (WEIL, TERZAGHI and CRASEMANN 1965). The principal characteristics of these exceptional phages, reported by WEIL, TERZAGHI and CRASEMANN (1965) are (1) each exceptional phage produces three types of plaque-forming progeny: exceptional phage like itself, rJIOl progeny and 7-1589 progeny; (2) under defined conditions each independently isolated line produces the three types of progeny with a characteristic frequency; (3) as a rule, the r progeny ( r segregants) are hemizygous rather than homozygous; and (4) in single cycle growth, the kinetics of r segregation and the r segreganl clone size distribution resemble those of recombination. To account for these observations, we have proposed (PARMA and INGRAHAM 1970; PARMA, INGRAHAM and SNYDER 1972) that the exceptional phage contains a tandem duplication for a portion of the T4 genome that includes the rII region and that the length of the duplication varies from one isolate to another. By analogy to higher organisms, the tandem duplications were presumed to undergo both symmetrical and asymmetrical pairing. A similar hypothesis has recently been proposed by SYMONDS et al. ( 1972). An asymmetrically paired tandem duplication of the rII region is presented in Figure 1. A phage carrying such a duplication is in a general sense partially diploid. The recombinations depicted by dashed lines x and y produce hemizygous rJIOl and r1589 segregants, respectively. The r segregants are produced by recombination and therefore have recombinant kinetics and clonality. The relative and absolute frequency of rJlOl and A589 segregants is a function of the size of Genetics 73: 161-183 February, 1973. 162 D. H. PARMA A N D M. SNYDER LEFT RIGHT B A B A 8262 ~ 4 l r1589 rJlOl s2 E51 I I la I I nl I t I I I I Y ---------......k...(................. . . 4 . . . . . . . . . . . . . . . * . x ----. E L--\ \ \ I . . ..+. . \ r-. --............ FIGURE 1 .-An asymmetrically paired tandem duplication of the rII region of bacteriophage T4D. The duplication's break point is represented by the squiggly vertical line. The left end point is between m41 and the B cistron and the right between the A cistron and S2. In this and other figures, the order r1589.rJ101 is assumed because it is more frequent among our duplications. the bracketed regions x and y. In a test of the hypothesis, three predictions were verified (PARMA, INGRAHAM and SNYDER 1972). First, duplications segregate triplications (Figure 1, dotted lines). Second, triplications occur with the same frequency as structurally normal chromosomes among the progeny of duplications. Third, the r mutation in the left half segregates by recombination to its right and the r mutation in the right half by recombination to its left. It should be noted that because of the restriction on the amount of DNA that can be packaged into a T4 head, phage particles containing a triplication segregated from a duplication lack plaque-forming ability, but they can complement amber mutants in a helper phage assay and can thus be titered. It is possible to generate viable (plaque-forming) triplications, which carry three distinguishable overlapping deletions, by recombination between triplications segregated from duplications and a long acriflavin-resistant r l l deletion which has neither A nor B cistron function (PARMA, INGRAHAM and SNYDER 1972 and Figure 3 ) . These recombinant triplications are selected by their ability to form plaques on a A lysogen in the presence of acnflavin and are termed acriflavin-resistant (acr) triplications. This communication presents methods for genetically mapping the end points of duplicated segments and the results of a critical test, based on these techniques, of the assumption that the length of the duplicated segment differs in strains of independent origin. The results show (1) that the position of duplication end T A N D E M DUPLICATIONS 163 points and therefore the length of the duplicated segment differ in strains of independent origin; (2) that there is a direct relationship between segregation frequency and length; and (3) that while duplications frequently include nonrII genetic material, frequently they do not include the entire rII region. E26 E52 E566 E556 E1204 NG604 E205 S2 MATERIALS A N D METHODS Phage strains: Mutations of bacteriophage T4 used in these experiments are given in Table I. Their relative map positions are indicated in Figure 2. Unless noted, all mutants originated in