The Nature of Genetic Recombination near the Third Chromosome Centromere of Drosophzla Melanogaster'

Previous studies have indicated that recombination near the third chromosome centromere is associated with negative chromosome interference, a phenomenon for which GREEN (1975) and SINCLAIR (1975) suggested gene conversion as a possible mechanism. In this report, we demonstrate that negative chromosome interference is still observed when deficiencies or translocation breakpoints are scored as the middle markers in recombination experiments and the rate of recombination is increased by interchromosomal effect. We argue that these chromosomal rearrangement breakpoints are not subject to conversion. Since neither successive premeiotic and meiotic exchanges, nor negative chromatid interference, can by themselves account for the negative chromosome interference, we conclude that a greater than expected frequency of multiple exchanges actually occurs. We further suggest that negative chromosome interference may be characteristic of all chromosomal regions normally showing very little exchange in relation to physical length. T H E proximal portion of chromosome 3 of Drosophila melanogaster is characterized by a low frequency of genetic recombination in proportion to physical chromosome length. The gene radius incompletus ( r i ) is located in the left arm at 47.0 on the standard recombination map and in region 77 (or possibly 78) on the polytene chromosome map (ARAJARVI and HANNA-ALAVA 1969). Only 1 % recombination normally separates ri from pink (p), which maps in the right arm at 48.0 and has been localized to 85A (DUNCAN and KAUFMAN 1975; ALEXANDER 1975). Thus, the ri to p interval represents about 15% of the euchromatic portion of the third chromosome (estimated from polytene chromosome numbered units), but is associated with slightly less than 1% of the recombinational map. This region also contains the third chromosome centric heterochromatin, which probably shows little if any recombination (BAKER 1958; ROBERTS 1965). Reciprocal exchange in Drosophila and other eukaryotes is characterized by positive chromosome interference, and multiple exchanges normally are not observed at all in intervals of 10 CM or less. MORGAN, BRIDGES and STURTEVANT 1This work was supported by National Science Foundation Grants PCM76-11750 and PCM78-07710, and D. 0. KEPPY was supported by Public Health Service training grant HD00422. Present address: Department of Genetics; Iowa State University; Ames, Iowa 50011. Genetics 93: 117-130 September, 1979. 118 R. E. DENELL A N D D. 0. KEPPY (1 925) reported that the chromosome 3 centromeric region represents an exception to this generalization; doubles were observed within an interval of only 6 CM long, and a coincidence value of 1.3 was observed. This early observation of negative interference in the chromosome 3 centromeric region has been confirmed recently by GREEN (1975) and SINCLAIR (1975). Both investigators found apparent multiple exchanges within short intervals, and coincidence values were consistently greater than one. Both GREEN and SINCLAIR suggested that negative interference exists because of the occurrence of gene conversions, which may give products genetically equivalent to those arising from double crossovers. They noted that earlier studies showed that there is no interference across the centromere in Drosophila ( GRAUBARD 1934; STEVENS 1936), so that the expectation is that the coefficient of coincidence for classical double exchanges flanking the centromere should be one. The degree to which coincidence was greater than one was considered to reflect the occurrence of conversion events. As other alternatives to a high frequency of real double exchanges, SINCLAIR (1975) suggested that premeiotic exchange and negative chromatid interference could also be responsible for the high coincidence values observed. We have tested the hypothesis that apparent double exchanges in this region are generated by gene conversion by utilizing genetic markers that should not be subject to conversion. Thus, deficiencies and Y-autosome translocations have been used as middle markers in threeand four-point recombination experiments. MATERIALS A N D METHODS The genetic variants utilized in the experiments are described in LINDSLEY and GRELL (1968) or when first mentioned below. Crosses were routinely performed in half-pint bottles, with about ten females and 15 males per bottle. Parents were subcultured after three to five days, and in some cases subcultured one additional time. Cultures were maintained at 25 & 1 O on standard corn meal-yeast-molasses-agar medium. Recombination frequencies and coefficients of coincidence, and their standard deviations, were calculated using the method of STEVENS (1936). Recombination with Tpl deficiencies: A uniquely dose-sensitive region, the Triplo-lethal (Tp l ) locus, is located in proximal 3R at 83D-E on the polytene chromosome map (DENELL 1976). The locus is associated with lethality when present either in three doses or in one dose in an otherwise diploid individual. In an analysis of the mutational properties of this locus, KEPPY and DENELL (1979) recently generated a series of deficiencies and Y-linked duplications including the locus. Two of these deficiencies, Df(3R)29c76 (=Df(3 R)83D;84A4,5) and O f (3R)3Oc76 (=Of( 3R)83Ci,2;84Bl,2), were used in recombination experiments. (The genotype of the females tested is diagrammed in Figure 1, and the resulting data are presented in Table 1.) When normal disjunction occurs, the only viable offspring that receive a TpZchromosome are males that also receive Dp(3;Y) l la76 [=Op(3;Y)830; 84F4+99El;fOOF], marked with Bs; the sequentially normal maternal third chromosome is recovered only in females. Thus, by scoring progeny phenotypes with respect to n', BS, pp, and Sb, we can recognize apparent exchanges in each of the intervals shown in Figure 1. For two reasons, the resulting crossover frequencies and coefficients of coincidence presented in Table 1 are calculated only from the female progeny. (1) There is a consistent excess of female (with respect to male) offspring among the nonrecombinant classes; we interpret this skewed sex ratio to be due to RECOMBINATION NEAR THE CENTROMERE 119 reduced viability of males receiving a Tplthird chromosome and D p (3;Y)iIa76. ( 2 ) A potential ambiguity associated with these experiments should be noted. The parental females arose from a cross of putative C ( I ) M 3 , f bb; Dp(3R)21173/ri Tplp” females and YsX-YL, Zn(l)EN, y /Dp(3 ;Y) l la76 , B S ; Sb/Df(3R)3g74 males. Dp(3R)21173 and Df(3R)3g74 are respectively, a small reverse duplication and a deficiency of the Tpl region (KEPPY and DENELL 1979). A degree of uncertainty is associated with the presence of ri and p” on the TpZchromosomes; in spite of the low frequency of crossing over in this region, which should be reduced still further by the presence of Dp(?;R)23273, some of the Tplchromosomes may have lost either the ri or p p marker by an exchange with the duplication-bearing chromosome. With respect to the basis of negative interference in these experiments, the class of greatest significance has apparent double crossovers in regions 1 (ri to Df) and 2 (Df to p ” ) . The frequency of recovery in this class and calculation of coincidence values are little affected by uncertainties associated with the presence of ri and p” as flanking markers. The rate of crossing over in region 1 is calculated from female progeny that have received ri, and thus must have arisen from mothers bearing this allele. Similarly, the crossover frequency of region 2 is calculated from female offspring that receive p P . Thus, we can recognize these recombinant progeny unambiguously and reasonably estimate coincidence values. This would not be true if we used data from male offspring as well, since females homozygous for ri+ transmit noncrossover chromosomes to their sons, which are scored as having arisen by an exchange in region 1. In other experiments, we have occasionally observed the loss of the BS marker from the Dp(3; Y)lla76 chromosomal element, presumably due to spontaneous heterochromatic exchanges. Thus, in the present experiments, it is possible that non-Bar females are recovered rarely that have the Tplthird chromosome and an unmarked free duplication; conversely, males phenotypically Bar could have two Tpl-normal third chromosomes and a Y-fragment marked with B S . In the experiment using 30~76 , all offspring that were putative double recombinants in regions 1 and 2 were progeny tested to determine if their Tpl genotype was consistent with respect to the expression of non-Bar or Bar; no case of a breakdown of Dp(3;Y)l la76 was found. This result suggests that the recovery of breakdown products did not significantly affect the data in the experiment with 29~76 . In the experiment using 29~76 , crossover progeny were counted and then all offspring were retained and the total number was estimated by the dry weight method of DORN and BURDICK (1962). In order to calculate frequencies based on female data, it is necessary to estimate the number of female noncrossover progeny. Based on the experiment with 3 0 ~ 7 6 presented here, and an additional recombination experiment using a sequentially normal Tplbearing third chromosome, we can estimate that 70% of the nonrecombinant progeny, or 3787, were females. Recombination with an Antennapedia deficiency: Two additional recombination experiments were performed with Dj(3R)AntpNs++RI7 (=Df(3R)84B1,2;84011,12). This deficiency fails to complement roughened eye (DUNCAN and KAUFMAN 1975), and the segregation of the deficiency in heterozygous females was followed by crossing to males homozygous for roe. Recombination with Y-autosome translocations: Additional recombination experiments were performed using two Y-autosome translocations: T(Y;3)J139 and T(Y;3)B155, which have third chromosome breakpoints at 80 and 82C, respectively (LINDSLEY et al. 1972). These crosses follow the transmission of four chromosomal elements: the compound X chromosome, the sequentially normal third chromosome, and the two translocation elements. In each cross, approximately one quarter of the progeny arose from a meiocyte in which three of these elements passed to one pole and one element to the other at the first meiotic division (3:l disjunctions) ; in the remaining cases, two chromosomes passed to each pole (2:2 disjunctions). Since an examination of the results with a contingency x2 test showed no differences in the frequencies of crossover progeny arising from maternal 2:2 and 3: 1 disjunctions, all data were pooled to calculate the crossover frequencies and coincidence values. 120 R. E. DENELL AND D. 0. KEPPY Ki is known to be localized in chromosome 3R (MERRIAM and GARCIA-BELLIDO 1969) proximal to Antennapedia (GREEN 1975), and our crosses show that Ki is clearly distal to the chromosome 3 breakpoint of T(Y;3)B155 (see RESULTS). In one bottle of the T(Y;3)B155 cross (denoted mating 6), a large number of phenotypically 9 n Ki flies were recovered. This class is expected to arise from an exchange in region 3, but occurred in about six times the expected frequency. The reciprocal class (Bs p p progeny) was recovered in the proportion expected. It seems likely that most yz ri Ki flies arose from an exceptional event, possibly a premeiotic exchange in a parental female. Thus, the results from mating 6 have been excluded from those of the remainder of the T(Y;3)B155 cross. In the crosses involving Y-autosome translocations, we intended to progeny test all male flies recovered from apparent multiple exchanges to confirm their genotypes. Because of the Y-chromosome translocation breakpoints, males bearing either of the translocations are sterile in the absence of additional Y chromosomes. The cross presented in Figure 3 was designed to insure male fertility by the presence of a compound-XY chromosome. Nevertheless, all males tested were sterile, presumably because of a genetic breakdown of YsX-YL in the stock yielding parental males.