..._.——-. fif—‘V-r-r'u .-. “‘.'?_A vW-. . ,1,1 ‘ 755% ' IMXE1IIII1,,I ”I. ‘ 1, 1." :11 1 q“... - 5 .. 52;“? 11' 3.11.1... 511313;, 5255 5'15] , .- 5,, 119',“ 11:11 ’E’Ir .I .155": {.5551 124;}.‘1I‘”"f Qw‘fi. 551:1, 55:33: 1513"""11 “11;“ $1,555, 1’5”" 111‘ “3’5 “’13“? ‘ l W’ 11‘; 1:1,, 5““; ‘5: M ""1“ (‘33:? "viii. ‘2‘" "" : 111V‘1L‘ ’ \,1‘ 'l‘ ‘ 54 ‘1‘1‘i:. “' 113 I . 1.1.91" “ .II “III: 11““ .. 11111115. “‘1 ‘ “531:1? .21“ 52.5211953511 "133%; 11:. ’1'“ ,;.,I}15"!I;5}.,.,1.1513 .5611 “55$ 11:11, ,I 11;: 11,551,{%€,‘5:‘:551- '1‘I'21 11.55 ,~ 1:1‘. fighmgT “1“, ‘7..- a grit“: ~ "555, I “n’ ‘5’ qu : .WHW "lg"??? 521%,}, A 4 .555" Ki; Elm-"11‘: “111% c 1. "4% 1. In! luv-"1?: III':’. 1 1%:11‘1 .II .. . . .. 1.1.‘111 ”'1, 1‘13: II. ”Imp “II .51} . . .1 . 4511.311? ~ 11,511.15, . I :1 .5; I: 111511 #155; 1”. 125.115 yfififiu '9'... . 11“" 1'1‘1,‘: '"1 111 h?- g. “a, ”1%” 1 11 I; I 1’11?“ E 1111"". 1* , .I ) I4 ‘ I 5:3. 1:11 .\. :é‘ . I; 1. "“111515’1m1‘1 35'” $5., 1111:3351“; "it"; “‘1 11““ I5, 7111‘ 1, ”5111111???" ail 3."; .-.':~£:- 1 “319:!" Si?“ ‘3" ‘1; I" "'21 I": 1‘5““.1‘1‘23 . Lfl ‘ _ TM 1:31;: 51"“ 1. . m '1 1.1111. II 1.1.1. I E "" 7:17. .35.;‘51 MR 211.351,, {0. «$1, :. a... 2.}; .vr ‘1. 2,3 .1, 5115.121. . (1‘1. 1 ’1'I- 113111 “‘1’; 15‘5“ - ' “‘1 11I414‘{1,1:1kl;1‘l.,1 [11111.4 ‘15,; 5 “‘1’ JL..11,5,5,13'41 "“1 URI", 511.5? ‘ 5.11215“? £15291: 3 WAT! V? I 3.61% ‘k‘ 13‘ . 3:1‘1célIfiII.... w’i. “35‘1“.1‘ I- . I. 51. 5.) a .. , "8-; "11 “1921‘le "’2; 1. :51: "WWW" 5.151;“ 5% 5235 111! w- ‘I ’5 41""‘I' ' ’51 .11 If ,5: I II‘S; 5133151" 5555),, ”' ,. "I1. {:3},qu 5,515 51?} (5t ’, kgfiiIa 1 .1111 1111111111111 . .- III .I (3", 5,91” ""2? "’ 5555‘“ "115,11" 555111,: ‘ 1.5%,? 11:5 1‘11 L "(“hirfif’zzfl‘gk‘ " .43. ' ’1 ‘ {E5113 ”1111’ I? 121"" “£55 118%, ’ ‘ ' ‘ 1 1“ "'“ 1111‘ 1;! 51“? 51.15555; » I.I .. . 111111.11. . 1, , I, 11.1.... ,» ,"I 51555315., 351,1511555. . “"'1"£1”“11“ WKLM‘H . ,5 I.I, AI, “L $5115“? I I I . ‘1. . r1. ,1, 11.11.35 $13551H2’1‘fif‘fifigv . . ,1, IL, 1: 1.5L .. . 59‘ ‘ 1” . 1: I .-'J, 1115,51HM .I 5515 . 5,555 R 1.1., 5555;“ 1113‘: . "‘5 "15.5“ I : -1 f ' I5 5 §EQ hi '1 1 1 "‘ W1 131;" I51? fig’a}? ““53?th ,5; ‘1! ,~ 313“!" 1511113,,5‘55 1155:3555“ ‘ .. . . 1'- 11”“,2 “III 11.1 “1112. 1‘13“:- 1 “1’:§1"“1‘C‘“ 5:11} .. “:1 ,Q1111I.....,j}“ ’ 1 111.: '“I I 11311. I111“ - '14 .. ..: .w, ‘ ,_. ‘¥. “‘1 5:“??? "1, 1,1553%’ $5,155 ‘13 "1,5, § 115:“! ._‘_. —:~.r. {$11.1 I IS. . 1: £031 ‘3le 3‘1““: I‘d: 1H. .25... 92.15::- —.. f r I ‘_g .1. . 5 j .. .I _I , f . ’ I > A l~a_*‘ ‘V‘ ‘ IV» D _ A :. 3-; ~ :2 1:: :1” . .f... a“... w , we 3. 4'17; 1"”, . - S: 1:: ‘WI' > 3.5.... -‘ ‘ 4:? . "IL "Ft— ‘5‘. _ I 3...; #‘u‘W‘Tfi u--- ..x._. ._ .. , . -25.- " ' " ~ A - .__. ‘ 42-!» I ’35:: 4:5; “3E: seas: $32M "‘1 t: ‘ I ‘f;‘,‘;" fl“ ‘4“ ‘1_,‘1‘.I "413%?! 213511.711 1,4 ‘41 ‘1‘! I. - .. arr- ,. {$11 '2“ 4:...- . 1v ‘1 “ ’1: 3‘ 11:11:31 9.15 . 1’3"”; ‘ I 11“. , l’. Ih’ ..I 1 I" . V, ‘.I 11'? 1‘. ’5‘ . n 5’11": 1:1" 1 . ,. 111. '11 I: I __.':, via v,‘ “i? «’- Date May 4, 0-7639 " * V 4 Itlllllllllllllll'lllllllllll i' W T “ 3 1293 00659 3454 $12,163» 4‘ C 11". 2.: ~, i Mi”: " " “ “51 it 0:: life‘?“ 04» - 31;? a, m ,, - m This is to certify that the thesis entitled Parameters Of The Rex Phenotype In Drosophila meZanogaster presented by Ellen E. Swanson has been accepted towards fulfillment oftherequhenunnsfor Ph. D. degree in Genetics Zoology W474. MW Major professor Leonard G. Robbins 1984 MSU is an Affirmative Action/Equal Opportunity Institution MSU LlBRARlES _——. .— RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. I 3321);?» ite—- as necEP€F1fihl fi PARAMETERS OF THE REX PHENOTYPE IN DRUSUPHILA MTLANDGASTfR By Ellen E. Swanson A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Genetics Program and Department of Zoology 198A ABSTRACT PARAMETERS or THE REX PHENOTYPE IN DRUSUPHILA MTLANDGASTER By Ellen E. Swanson Rex is a dominant, maternal-effect locus located in the heterochromatin of the X chromosome of Drosophila melanogaster. It causes an early mitotic exchange between heterochromatic regions of YSX,YL, y y f B,y+in X/XY embryos of Rex mothers. This exchange leads to the production of males carrying free Y chromosomes and gynandromorphs that are part X/XY female and part X/Y male. The exchange event occurs near or within the bb locus, the site of the ribosomal genes. In this study, crosses and progeny counts were used to uncover and clarify further aspects of the Hex phenotype. Data are presented that illuminate several characteristics. l) The site of Rex action in the responding chromosomes is the rDNA; possession of a duplication for the rDNA is necessary and sufficient for a chromosome to be sensitive to Rex. 2) Responding chromosomes can pair and exchange in two conformations, a spiral and a hairpin. 3) Some ribosomal cistrons are lost at each exchange event. A) Rex acts on both maternally- and paternally-transmitted chromosomes. 5) Rex is a neomorph, and a + . Hex allele may not exust. These results suggest avenues for further investigation of the Hex effect, including the use of Rex to synthesize unusual chromosomes, the further delineation of the criteria for responsive chromosomes, and the search for flex alleles in other stocks and natural populations. In addition, the relationship of Rex to known phenomena is examined. Rex is a heterochromatic gene which affects mitotic chromosome behavior at the bb cistrons. The phenotype of Rex implies the potential involvement of this locus in controlling chromosome stability or the regulation of ribosomal cistron copy number. The study of Rex may also present an opportunity to better understand the action of heterochromatic genes in general. To Lenny ACKNOWLEDGEMENTS I am grateful for the patient instruction, perceptive guidance and ceaseless encouragement of Dr. Leonard G. Robbins. I also thank Dr. Thomas Friedman, Dr. P. T. Magee and Dr. Barbara Sears for their critical reading of the manuscript, and Adli Karadsheh for his help with one of the experiments. I am grateful to Dr. w. Baker, Dr. D. Lindsley and especially to the Mid-America Drosophila Stock Center of Bowling Green State University. Their promptness and accuracy in sending requested stocks greatly facilitated the research presented here. Special thanks are also due Linda Smith for her support and empathy. TABLE OF CONTENTS LIST OF TABLES .................................................... vi LIST OF FIGURES ................................................... vii INTRODUCTION ...................................................... I Methods and Materials ........................................ 7 CHAPTER I. THE RESPONDING SITE TO Rex ACTION ..................... IO CROSSES AND RESULTS .......................................... 16 Response of attached-XV chromosomes to Rex ................... 16 Response of sc inversions to flex ............................. 19 Response of wm inversions to flex ............................. 21 DISCUSSION ................................................... 23 CHAPTER 2. THE Rex EXCHANGE EVENT ................................ 26 CROSSES AND RESULTS .......................................... 30 Test for hairpin pairing ..................................... 30 Effect of Rex on maternally-transmitted chromosomes ..... ..... 33 Production of bb deletions from SCSISCA ....... . .......... . . 33 Reciprocity of the Rex event ......... . ............ . ...... .... 35 DISCUSSION ................................................... 44 CHAPTER 3. THE NATURE OF THE Rex LOCUS ........................... 48 RESULTS ...................................................... 53 DISCUSSION ................................................... 55 CHAPTER A. SIGNIFICANCE AND RECOMMENDATIONS ...................... 58 The bobbed Locus ............................................. 62 Heterochromatin .............................................. 77 Mitotic Chromosome Instability ............................... 81 APPENDIX. SUPPRESSOR OF Rex ...................................... 86 LIST OF REFERENCES ................................................ 99 Table Table Table Table Table Table Table Table Table Table Table Table Table LIST OF TABLES l. A comparison of Rex sensitivities of five attached-XY's. . l7 2. Rex sensitivities of two 55 inversions. ........ . ......... 20 3. Rex sensitivities of three wm inversions. ................ 22 A. Test for hairpin pairing. ................................ 32 5. Bobbed variability of y+ fragments. ...................... 36 6. Retest of “reinversions” (Dp/f;1/scS]sch). ............... 38 7. Bobbed variability of proximal end of reinverted chromosomes. ............................................ 39 8. Bobbed variability of distal end of reinverted chromosomes. ...................................... ...... 42 9. Comparison of bb variability of two ends of reinverted chromosomes. ................................. 43 IO. Results of dosage experiments. .......................... 54 ll. Suppressing stocks. ........ ......... ............ . ....... 88 12. Mapping Su/Rex/ from y cv v f can stock: results. ....... 92 13- Mapping Su/Rex/ from y ; T/Z;3/e/$M1;TM2 ,' spam] stock: results. ....... . ............. . ................ .. 94 vi Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure LIST OF FIGURES 1. Synthesis of YSX,YL. ....... . ............................ 2 2. Spiral exchange in YSX,YL. .............................. 4 3. 35 inversions of the X chromosome. ...................... 12 A. Spiral exchange in 35$]sch. ............................. 13 5. wm inversions of the X chromosomes. ..................... 15 6. Hairpin exchange in SCSISCA ............................ 28 7. Mating scheme to test for hairpin exchange. ............. 31 8. Mating scheme to collect y+ fragments. .................. 34 9. Mating scheme for distal end of reinversions. ........... 40 IO. Mating scheme for control and reduced-dose tests. ...... 51 11. Mating scheme for increased-dose tests. .............. . 52 12. Mating scheme for detection of suppressors. ............ 87 13. Loss of Su/Rex/ effect. ................................ 90 1h. Mapping SulRex/ from y cv v f can stock: mating scheme. 91 15. Mapping Su/Rex/ from y ; Tl2;3/e/SM1;TM2 ,' spapo] stock: mating scheme. ......................... ...... . 93 16. Mating scheme to test for maternal suppression effect. . 96 17. Suppression by FM7. ..... ... ....... . ..... . ..... . . ..... 97 vii INTRODUCTION In 1981, Robbins reported the discovery of a genetic element with a number of interesting properties. This element (technically a variant, as opposed to a mutant, since it was found in a pre-existing stock) is located in the X heterochromatin, and it has a dominant maternal effect. It seems to cause a mitotic exchange in an otherwise stable 7 attached-XY, producing free Y chromosomes. Because this exchange event occurs predominantly between the ribosomal cistrons of the. X and Y, the element was named Rex, for ”dominant inducer of ribosomal exchange”. The attached-XY chromosome in which the Rex-induced detachment occurs is YSX.YL (Lindsley and Novitski, 1959). The structure and origin of this chromosome is shown in Figure I. In this and all subsequent figures, broad lines designate heterochromatin and narrow lines designate euchromatin. In/7/[N is a complete inversion of the X euchromatin. The short arm of the Y (YS) is attached to the distal X euchromatin, and the long arm of the Y (YL) is attached to the centromere. YL is marked with a small transposition of the X euchromatin carrying the X marker, y+. The detachment event is seen in crosses of y/y cv v f Rex females and YSX,YL, y y f R,y+/0 males. Along with the regular and non-disjunctional progeny of this cross, y+ male offspring are produced. These males carry an X chromosome from their mothers and a I L .qx.km& +0 m_moc+c>m ._ oczm_u 23:7: zmHZZH 4» free Y chromosome marked with a y+ transposition of the X chromosome (y+Y). Gynandromorphs, flies which are part male and part female, are also produced. The male parts of the gynandromorphs have phenotypes which indicate that these tissues also carry a maternal X and a y+Y. The y+Y's can only have been derived from the attached-XY. Since the attached-XY breakdown event is seen in the progeny of Rex mothers, it must occur post-fertilization in what would have been X/XY zygotes. Thus, Rex causes a maternal effect and is dominant. Rex segregates from the X chromosome balancer FM7 (Merriam, 1968), and meiotic recombination mapping places Rex proximal to carnation, in or near the centric heterochromatin of the X. Robbins' (1981) hypothesis for the mode of action of Rex is shown in Figure 2. The Y detachment is envisioned as an exchange-like event between the X centric heterochromatin and Y5, resulting in loss of the X euchromatin in an acentric ring. Since the event occurs during post-fertilization mitoses, early events will produce males which have detachment Y chromosomes in all their body tissues (”whole-body” detachment males), whereas later events will produce gynandromorphs. Figure 2 shows detachment at a two-strand stage (i,e,, in GI of the cell cycle). If the detachment occurs at the first mitotic division, a whole-body y+ male will result. A second division detachment will produce a half X/XY (female), half X/y+Y (male) gynandromorph (X/XY:X/y+Y). Exchange at a four-strand stage produces only gynandromorphs, either X/XY:X/y+Y or X/XX:X/y+Y, depending on which two strands exchange. Since whole-body males occur more frequently than gynandromorphs, it is unlikely that four-strand events occur. Any .NM.NMM c_ omcmcoxo _mc_am .N oC3m_u XIosz onHzmoc event occurring later than second division will produce a gynandromorph with predominantly female tissue. Robbins (1981) scored five body parts in a sample of gynandromorphs and found that each structure was male about half the time. Those data support the notion that the Rex event occurs during CI of (usually) the first or (sometimes) the second mitotic division. Meiotic recombination in the euchromatin and heterochromatin was also examined, and was found to be unaffected by Rex (Robbins, 1981). The site of action of Rex was studied by analyzing phenotypes of males bearing the y+Y chromosomes produced. The most obvious region of homology between the X heterochromatin and YS is the bobbed (bb) locus. The phenotype of homozygous bb flies is short, thin thoracic bristles, delayed development, and, in some backgrounds, etched abdominal cuticle (Lindsley and Grell, 1968). The bb locus is the site of the 18$ and 28$ ribosomal RNA genes in DrOSOphila and also corresponds with a cytological marker, the nucleolus organizer region (NOR) (Ritossa et al,, 1966). The terms ”bobbed locus”, ”rDNA”, “ribosomal genes" and "nucleolus organizer region” are used interchangeably in the literature and will be so used here. The genes are clustered together, and there are 130-300 copies/locus (Ritossa et al,, 1971). Two observations from the progeny of Rex mothers make it likely that the detachment event occurs in or near the bb locus. First, there are no ring-X chromosomes recovered. These would be expected if YS ever paired and exchanged with YL. Second, the y+Y chromosomes carry all the YS fertility factors. Thus the exchanges do not occur distal to these, even though the first Ys factor is just distal to bb (Kennison, 1981). However, Robbins' (1981) investigation does not rule out the possibility that some of the events may occur in the heterochromatin surrounding bb. Robbins (1981) examined the bb constitution of 172 y+Y chromosomes. Males bearing these chromosomes showed wide variation in bb phenotype. This was not due to pre-existing bb variation in the attached-XY, and it could be concluded that Rex causes exchange at many points within the ribosomal cluster. However, the phenotypic variation seen in the y+Y's does not equally represent the full range of exchange possibilities. If the two regions overlapped at random, at least 50% of the products should be bb+, because they received a normal or greater than normal number of rRNA genes. Actually this number should be a little greater than 50%, since some small deletions should also have a bb+ phenotype. Fewer than 10% of the detachment chromosomes tested, however, were bb+. The paucity of bb+ chromosomes can be explained by three different models. First, one of the two ribosomal complexes in the attached4XY may be much smaller than the other, so that even when the overlap occurs randomly, deficiencies might be much more common than duplications. Second, overlap may not be, random, and exchange may occur at preferred sites. This might result in the majority of the rDNA going to the acentric ring-X, and the remainder, not enough to produce a bb+ phenotype, going to the detachment Y. Third, though the products seem to indicate an exchange, it might not be a reciprocal exchange event, and ribosomal cistrons may always be lost in the process. Rex is a Mendelian variant located in the X heterochromatin, and it has a dominant maternal effect. It seems to cause mitotic exchange in the ribosomal genes of certain chromosomes very early in development The experiments that follow were designed to further explore the nature of the Rex locus and its action. They are designed to answer the following questions: What is the specific chromosomal site that responds to Rex action? What conformation(s) do the responding chromosomes assume, and what happens to the genes at the site of the exchange event? What is the functional nature of the Rex locus? The experiments show that: (l) the Rex-induced event does not depend on the presence of Y Chromosome material; (2) the sensitive site in the chromosome is the bb locus and only the bb locus; (3) responding chromosomes can pair and exchange in two conformations, a spiral and a hairpin; (A) some ribosomal cistrons are lost at each exchange event; (5) Rex can act on maternally- as well as paternally-transmitted chromosomes; and (6) Rex acts as a neomorph, and a Rex+ allele may not exist. Methods and Materials All flies for stocks and crosses were raised on a standard Drosophila medium of cornmeal, molasses and brewer's yeast. Mass matings were done in polyethelene bottles, 10-15 males with 10-15 females. The matings for counting offspring were done in glass shell vials, one female with three males or one male with three females. All matings were done at 25°. Unless otherwise noted, a description of all mutants and chromosomes used can be found in Lindsley and Grell (1968). This reference also provides a description of Drosophila nomenclature for those not familiar with its conventions. Except for Rex and bb, all loci mentioned here serve only to mark regions of chromosomes and their phenotypes per 3e are of no importance. A note on bobbed nomenclature: The bobbed (bb) locus was first described by Sturtevant in 1920, but it was not recognized as a redundant gene until 1966. This situation, the large number of workers involved with the locus, and the lack of a good system for naming genes present on both the X and Y chromosomes, have combined to produce a seemingly unending list of descriptors added as superscripts to the designation ”bb”. The following is not a complete list, but is designed to help the reader with those bb mutants mentioned in this work. The reader should keep in mind that all bb mutants are ribosomal cistron deletions of various sizes. (I) bbo, bbNO-, #0-, bb-z These designations indicate a complete deletion (as well as can be measured) of the ribosomal cistrons. (2) bb]: These mutants are lethal when homozygous or in combination with a complete deletion; unlike the mutants in (1), they are additive with other bbl's or bb's to produce less severe bb phenotypes or bb+. (3) bb: When homozygous or in combination with a deletion, these mutants have a bb phenotype; with other bb mutants they may be more or less an, or bb+. Each of the above can also be used as a superscript to the X or Y NO- ybb- chromosome symbol (as: X , ) to designate which chromosome carries the mutant. CHAPTER I THE RESPONDING SITE T_Q m ACTION The hypothesis which has been developed to explain the chromosomes generated by Rex (see Introduction) postulates an exchange-like event between heterochromatic segments during early somatic divisions. This exchange is believed to occur between the X and Y copies of the ribosomal genes. A number of questions can be asked about the pairing and exchange of chromosomes in response to Rex. Is this response a peculiarity of the attached-XY chromosome in which it was first discovered (YSX,YL, y y f B,y+), or does it occur in any similarly constructed chromosome? Is the response to Rex dependent on the presence of Y heterochromatin, or will chromosomes containing only X heterochromatin also respond? Is the orientation of the heterochromatin and/or of the ribosomal cistrons important? Is a duplication of the ribosomal cistrons alone enough to produce a response, or are other heterochromatic elements necessary? Rex was first discovered because of its effect on YSX,YL, y y f B,y+. A number of similar attached-XY chromosomes are available. A collection of these was tested to determine whether the response to Rex is a general characteristic of chromosomes with this structure. All of these chromosomes are derived from the original YSX,YL, Inlf/[N synthesized by Lindsley and Novitski (1959). This chromosome was made by two consecutive exchanges between an inverted X chromosome, In(1/£N, and a Y chromosome (Figure 1). All the attached-XY chromosomes tested with Rex have the same structure with 10 II regard to the X and Y segments. Most have the X euchromatin in inverted order, as it was in the original Lindsley and Novitski chromosome, though the chromosome with which Rex was discovered has the euchromatin in normal order. (This is believed to have been a spontaneous reinversion.) They do not all have the same X chromosome markers, and one does not have the y+ transposition at the tip of the YL arm. Whether the rDNA of the attached-XY chromosomes is derived from the X, the Y, or both is not evident. There are other chromosomes that have duplications of the ribosomal cistrons but do not involve the Y SlLschR (=sc515ch). The chromosome. One such chromosome is In/7/sc scute inversions of the X chromosome have left breakpoints near the euchromatic locus scute (3c) and right breakpoints somewhere in the heterochromatin (Figure 3). They all invert most of the X euchromatin. scSISCAwas synthesized by an exchange between two scute inversions with different right and left breakpoints. It has a duplication, near the tip, of a large part of the centric heterochromatin, including all the ribosomal cistrons (Baker, 1971; Hilliker and Appels, 1982). Figure A depicts the pairing of 355155“ and the products expected if it responds as does the attached-XY. Another chromosome of interest is 111/1]ch2 (=scvz) (Figure 3). Its right breakpoint is in the middle of the ribosomal cistrons and it moves the distal heterochromatin and half the rDNA to the tip. (Lindsley et 3],, 1982). Thus, it has ribosomal cistrons at both ends, but there is no duplication of the rDNA region. 12 V2.8 sc BREFIKPDINTS V2 8 NDRMRL X 4i)? T:_L_+> yfsc ND 4,81 4 SI chZ +1 ; Y’NO 3c ND 8C43C8 -+—I-Ir F.) Y’sc at 0T sc8 D E.) Y* NO as 0 SC4 -l--I—I 1:. b SCSI ++-:::I a.) yrsc NO I. 5551584 ~I—I-::- III—D y* 30 ND ND Figure 3. so inversions of The X chromosome. 13 oz XIosz Smezmoc w 0m. NMOm :_ omcmcoxo _mc_am .q oczm_d ezmzommc.s oz +> IIH|+I L ... 14 mA mA ). InH/wmslb (=wm51b). and Inlf/meIwamhR (=wm51bwmh). As shown in m51b Three other chromosomes were also tested: In/1/w (=w Figure 5, the breakpoints of wmA and w define the proximal and distal ends of the X ribosomal region (Hilliker and Appels, 1982). In/1/wmh moves the heterochromatin distal to the rDNA to a point near the w locus at the tip of the euchromatin; it leaves the bulk of the ribosomal cistrons near the centromere. In/f/wmslb moves most of the rDNA to the tip but leaves a very small portion near the centromere. Intt/wm5‘wam“R is produced by an exchange between these two chromosomes and is duplicated for only the rDNA (Figure 5). It should be noted that all the chromosomes used to test the Rex response have been kept in stocks for years with no sign of breakdown. I have observed each of these stocks for at least twelve generations and have never seen any products of spontaneous breakdown. For example, in one experiment YSX,YL, y y f B,y+ males were crossed to FM7/y females and no male progeny bearing y+Y chromosomes were seen in l0,000 total offspring. Similar results have been obtained with In/1/scSILschR and In/7/wmh. All the chromosomes which respond to Rex are generally regarded as extremely stable elements when Rex is not present. 15 um BREFIKPDINTS NDRMFIL X .4 4 1” H ND Nm4 ‘ 5:.) ND NmSlb 4:. I ND V meIbwm4 a :-) N0 N0 RND Nm4Nm51b + 1 Figure 5. mm inversions of The X chromosome. CROSSES AND RESULTS Response 91 attached-XY chromosomes £9 Rex Four attached-XY'S were obtained to test for response to the Rex chromosome. These were all described as YSX,YL, In(7/[N. They had four different marker combinations: y y f R,y+, y,y+. y V f,y+, and y 3 (no y+ transposition). All were tested in females with an uninverted X homolog for the production of single crossover progeny. No single crossovers were produced, so all the attached-XY's still have the In/1/EN inversion. They were then tested for the order of the Y arms by separating the two arms by a single exchange with another inverted X (In/7/scthc8R). The recombinants were tested for the ability to confer fertility on males that had received a Y8 or a YL arm from their mothers. All were found to have the Y arms in the order described. Males carrying each of these four attached-XY's and the original y V f B,y+ attached-XY were mated to homozygous y w cap Rex females. The results are shown in Table l. Detachment offspring bearing a maternal X chromosome and a free Y from the attached-XY were recovered as y+ w males or gynandromorphs in all crosses except for those with y B fathers. In the latter case, both regular-disjunction males and detachment males were y w; however, detachment males had a free Y and were fertile, whereas regular-disjunction males were X/0 and sterile. To determine the number of detachment males produced in this cross, all y w males were collected and mated to test for fertility. It can be seen that detachments were recovered from all the attached-XY's. Sensitivity to Rex is thus not a property inherent solely in the first we. «.7 l7 .Q\N ocm mo_mE Lm_:mom .mo_me o__+co+ mm oo+oo+oo ** Lo+mLoE:c + mo_mso+ cm_:moC .00_ x mcaeoeocvcmc>u + mo_ms +coEcom+on u +coEcom+oe R * .m.mxioo;om++m eo+mo_vc_ oc+ mc_>ccmo mo_mE 0+ vommoLo 0L0: mo_mso+ Rom.&oo 3 m\8mm.&oo 3 m 8.0 I- ....m o 0 RR 3.2 m s Seams Ra _ a a N 29 :2 3% s s .EEE. ea; _ m: m o Rom mom: Es SEES. Ra e t a 0 seem mmmm Krm .5 s s ..EESA 3.0 m D m 0 seem mzm Bums s s +C®E£Um+®U mLQLOEOLD m0 _ mE m®_ ME MG _ mE®+ mm _ 0E m0 _ ME®+ QEOmOEOLLU * +coocod Icmc>o +coELOm+oo mc_Cam++o mc_cam++o MNIuOcoo++< co_+0c:wm_ococtk co_+o:3wm_eicm_:mom .m.MNIUOcom++m o>_+ +0 mo_+_>_+_mcom afim +0 c0m_cmaeoo < ._ o_nmk attached-XY tested. Two points need to be noted about the data presented in Table 1. First, the frequency of detachment in these experiments is variable but lower than expected. The original attached-XY (YSX.YL, y v f B.y+) normally detaches at a rate of 1 to 3%; 2% is common and the rate has been as high as nearly 10% (Robbins, 1981). The data in Table 1 are significantly different from a rate of 2%. (A typical previous experiment had 1520 regular-disjunction females and 32 detachment products; in a test of homogeneity, X2 = A3.9A with 5 d.f., p < 0.005.) At the time these experiments were done, the y w can Rex chromosome was being kept in a homozygous stock; that is, y w car Rex/y w car Rex females crossed to y w car Rex/Y males. A stock with a maternal effect variant present in the females is likely to accumulate modifiers by natural selection. Indeed, eventually the y w can Rex stock lost all Rex phenotype. (Suppressors of Rex have been found in a number of stocks; see Appendix. All Rex chromosomes are now stocked only in males, balanced by crossing to attached-X-bearing females. Rex chromosomes from these stocks induce the higher (1 to 3%) rate of detachment.) Despite the lower frequency of detachment seen in Table 1, the reduced Rex activity has no effect on the conclusion drawn from this experiment (i.e.. that all these attached-XY's are sensitive to Rex action). Since a contemporaneous experiment was run using the original attached-XY (first line of Table I), all frequencies can be compared to this standard. .IE Second, the frequency of detachment of YSX,YL, In/1/ffl, y B is quite low compared to the others. The set of data in Table I is not homogeneous (X2 = 10.05 with A d.f., p = 0.0Al), but becomes so when the y 8 results are removed (X2 = 5.11 with 3 d.f., p = 0.22). This chromosome has no y+ marker on the YL arm, making the detection of detachment males difficult, and the detection of gynandromorphs impossible. Detachment males were detected by collecting all y w males and testing them for fertility. Single y w males were put with two females in each well of a plastic titer plate. The tops of the plates were filled with Drosophila medium and the wells were inverted onto them. This method was not completely satisfactory. since some of the males got stuck to or trapped under parts of the titer wells and died. It is therefore quite possible that not all fertile males were detected, and of course this method would also fail to detect detachment males that were sterile, an event that occurs about 15% of the time (Robbins, 1981). Response 9: 5; inversions £9 Rex In the next set of experiments, Rex females were crossed to Intf/scSILschR, cv v B/Y males or to In/1/ch2, scvz/Y males. As Figure A shows, the exchange event should produce heterochromatic fragments marked with y+. Though this event is not really a detachment, since nothing was originally ”attached“, for continuity the measure of the event will still be called a ”frequency of detachment”. . . . + Fragment-bearing males In these experIments were y and had none of their fathers' X chromosome markers. The data are shown in Table 2. 20 .co_+m:m_axo c0+ +xo+ oom * .m200comoeoz >__mo_+m_+m+m ocm m can < m+coE_ ILoQXm ._ o.pm+ c_ mm no+m_:o_mo mm: +coEcom+On a .m.» ooc+ new moeomoeocco N oo+oo_nc_ 05+ c+_z mo_me 0+ commoLo ocoz mo_mEO+ Ao new m m+coE_Loaxov eomyk a no m\8omqk o :0 m L0 A< +coE_c0axov m\8om.% a so m 1.01m...” m me am a $2 8: 8V meow .msofldem mo.P m m_ m o NNo_ moo? Amy oo.m ©_ ow _ N mmw_ moom Ao mc_cmop mc_cam++o mc_cam++o _mcco+md +cosmmcd co_+0c:wm_ccocix co_+0c:wm_oILo_:mom .mco_mco>c_ om 03+ +0 mo_+_>_+_mcom 8mm .N o_nmh 21 The frequency of detachment of the scv2 chromosome is difficult to calculate because the chromosome is not well marked. scvz itself is a poor marker, variable and weak in expression. Therefore, some of the wild-type males classified as X-nondisjunction products are probably y/y+ fragment males which are either non-crossover, or crossover distal to cv or proximal to f. The lower of the two frequencies of detachment shown for scv2 has been calculated with only the known fragment-bearing males (y+ and/or cv and/or v and/0r f) and the gynandromorphs; the upper limit has all the + males (classified as nondisjunction in Table 2) added in. The true frequency is probably somewhere between these two. In any case, both SCSISCL and chZ respond well to Rex action. m. . Response 9i y InverSIons £2 Rex The last set of experiments was done by crossing Rex females to [alt/Wm“, wmh/Y maIes. Intt/wm5‘b, me‘b/Y males or In/7/w mSIbLWmAR mA m51b w , which leaves the rDNA at the distal end, and w , CV/Y males. The results are shown in Table 3. Both , which moves . . m51b mA the rDNA to the prOXImal end, yield few detachments. Only w w , with a full duplication of the ribosomal region, responds to Rex. In all cases, therefore, it is necessary that a chromosome have ribosomal cistrons at both ends for it to respond to Rex. 22 ._ o_nok :_ mm oo+m_:o_mo mm; +coscom+on & .moeomoeoCLO N vo+mo_ec_ oc+ c+_3 mo_mE 0+ commoco ocoz mo_meo+ m\8&m.x o no m med e E a M £8 5mm 8 .sesflwssafi o o o a _ mtm Rom fines .fiwesEfi no.0 _ o m a $2 RE 2.3 «53:5. +coEcom+oo mcacoeocn mo_ms mo_mE mo_mEO+ mm_ms mo_mEo+ osomoeoLco N +c00cod Icmc>o mc_cmoo mc_cam++o mc_eam++o .mcco+md +coEmde co_+o::wm_ococik c0_+0c:+m_eico_:mom .mco_mco>ch EB mocc+ +0 mo_+_>_+_mcom 8mm .m o_nmh 23 DISCUSSION The effect of Rex is not due to some peculiarity present in the YSX.YL, y V f B.y+ chromosome. Other similar attached-XY's respond to Rex as well as the original. It is also apparent that the orientation and markers in the euchromatin do not affect the response, nor does the euchromatic transposition used to mark YL. Because of the nature of their construction, these attached-XY chromosomes are not good candidates for a detailed investigation of the responding site. In/1/[N inverts the entire X chromosome; its right breakpoint is in the rDNA and it carries a bb allele at either end (Lindsley and Grell, 1968). Thus, either of the two ribosomal regions in the attached-XY might be wholly or partially derived from X or Y cistrons (see Figure 1). This is important because the X and Y rDNA cistrons have a number of differences (Tartof and Dawid, 1976). It is also known that the ribosomal region in YS of the original attached-XY is partially deficient (Robbins, 1981). For these reasons other less complex chromosomes were used to examine the responding site. In/1/5551LschR was chosen because much of the X heterochromatin is duplicated in this chromosome. A number of inferences can be made from its positive response to Rex. First, elements of the Y chromosome need not be present for detachment to occur. Second, the orientation of the two ribosomal regions does not seem to be important. If the attached-XY has been made as in Figure 1, both ribosomal regions should be in their normal distal to proximal order. When the attached-XY chromosome pairs as in Figure 2, the regions will pair distal with 24 distal and proximal with proximal. In SCSISCA, however, the distal ribosomal region has been inverted. When this chromosome pairs in a spiral, the two ribosomal regions are in opposite order to each other. This apparently does not affect their ability to “exchange”. It may be that the ribosomal cistrons are not all in tandem, but are oriented in both directions. (For further evidence that homologous order is not important, see Chapter 2.) The experiment done with scvz indicates that a duplication, in the strict sense, is not even necessary for the Rex event. Nothing has been duplicated in scvz; the ribosomal region has simply been divided and half has been inverted and moved distally. The responsive region (the rDNA) is therefore divisible, and both halves of the rDNA are as competent as a full region would be. In fact, considering the high frequency of detachment, they may be more competent. The only identical regions proximally and distally in this chromosome are the rDNA genes. However, it is possible that Rex could work nonespecifically on heterochromatin. In fact, one of the molecular characteristics of heterochromatin is that it contains much repetitive material (Kram et 3],, 1972; Brutlag et 8].. 1977); similar repeats could be dispersed on either side of the bb locus so that scv2 actually does have regions of homology at either end. Rex, then, could work on these ”duplicated” regions. If this were the case, then it should .be possible to recover y+ fragments from In/1/wmh or In/7/wm51b. Like scvz, these chromosomes do not have any duplicated material. Some of the heterochromatin, ( m51b) including w or excluding (wmh) the bb locus, has simply been 25 moved to a distal location. Therefore, they should be functionally identical to scvz except that they have the rDNA in only one place. As Table 3 shows, however, these chromosomes do not respond to Rex. 0n 1 A m5 wam R WmSImeA the other hand, In/1/w is responsive to Rex. is single crossover product made from wmh and wm51b’ and contains the same heterochromatin blocks along with an rDNA duplication (Figure 5). Hence, possession of two sets of ribosomal cistrons is both necessary and sufficient in a chromosome to provide a site of response to Rex. CHAPTER 2 _TH_E m EXCHANGE EVENT By analyzing the products of the Rex event, Robbins (1981) made a number of conclusions about the event itself. Figure 2 illustrates the current hypothesis for the Rex-induced event. When the responding chromosome is an attached-XY, this event leads to the production of a male carrying a y+Y or a gynandromorph which is X/XY:X/y+Y. From the variation in bb phenotype of the y+Y males, Robbins determined that the event is not site specific within the ribosomal cistrons. Analysis of the gynandromorphs set the time of occurrence of the event at or before G1 of the second mitotic division (see Introduction). Comparison of the structures of the chromosomes produced bY Rex with the structures of the parental chromosomes should allow inferences to be drawn about the mechanism involved. To make this comparison, it is necessary to know the structure of the chromosomes before the event occurred. As was discussed above, there are ambiguities surrounding the formation of the attached-XY which preclude a full knowledge of its structure. Therefore, the 555155“ chromosome, whose structure is well-defined, was used in the present investigation. It was the purpose of the experiments described below to answer four questions about the Rex event. The first two questions were: (1) Are there alternative pairing conformations (other than the spiral illustrated in Figure 2) in which the Rex exchange can take place?; and (2) Can Rex act on a chromosome transmitted from the mother? Figure A shows 5551551+ in the spiral configuration. From this kind of pairing 26 27 and exchange two offspring classes are possible: males carrying a y+ heterochromatic fragment and X/X:X/y+ fragment gynandromorphs. Both of these products have been observed. Another possible pairing configuration is shown in Figure 6. An exchange in this case would result in a chromosome in which the euchromatin has been reinverted. The possibility of this exchange was investigated by mating SCSIsch males to Rex females, collecting heterozygous female progeny, and looking at their offspring for crossover classes from recombination between the putative reinverted SCSISCA and the normal-sequence, Rex-bearing X. If there were also males carrying the y+ fragment among the progeny of these scSlsch/Rex females, this would show that Rex could act on a responsive chromosome when it is transmitted to the embryo by the mother. The third and fourth questions were answered by examining the products of the hairpin and spiral types of exchange. (3) Does the spiral pairing and exchange of sc515ch generate the same preponderance of deleted (bb) chromosomes from SCSISCA as were produced from the attached-XY? Robbins' (1981) investigation of the y+Y chromosomes resulting from detachment of the attached-XY showed that fewer than 10% of the y+Y chromosomes were bb+. This indicated that the two ribosomal regions did not overlap and exchange at random; if they had, at least 50% of the products should have been bb+. The excess of deletions could have occurred if the two rDNA regions in the attached-XY were so unequal in size that a duplication was infrequent. The discovery that the YS portion of the attached-XY is bb does not seem to indicate an extreme enough difference in size to make this explanation likely. To 28 .vomwmom c_ omcmzoxo c_ac_mI .o 0L:m_d +> um 02 >0 > m 02 m 02 um ...x II + llllllflll. > n r 0 >0 OZ 29 resolve this question, males bearing y+ fragments from the .scmscl+ pairing and exchange were collected and used to determine the bb constitution of these fragments. Since sc515ch has two complete rDNA regions, examination of the y+ fragments should test whether the production of over 50% bb chromosomes is a feature of all Rex exchange or is due to a peculiarity of the rDNA of the attached-XY. If the SCSISCA does produce an overabundance of bb chromosomes, then (A) Why does Rex-induced recombination result in an excess of deletions? There are two possible explanations. Pairing may not be at random and exchange, though reciprocal, may occur at preferred sites. Perhaps only the recoverable products from spiral exchange (y+ fragments or y+Y’s) are always deleted, and the duplication products are in the acentric rings. Alternatively, the exchange may not be reciprocal, so that cistrons are always lost. The products of hairpin pairing and exchange were examined to test these hypotheses. Both products of the exchange were recovered by separating the two ends of the reinverted SCSISCh. The ends were tested for bb phenotype to examine the reciprocity of the Rex "exchange” event. If the event is reciprocal, when one end is a deletion, (bb or bbl), the other end should be a duplication. If there are cistrons lost in the process, both ends may carry deletions. 3O CROSSES AND RESULTS Test for hairpin pairing The mating scheme to test for hairpin-type pairing is shown in Figure 7. Homozygous Rex females were mated to InH/scsn'scAR ,cva males in bottles (generation (1)). If hairpin pairing and exchange can occur, some of the female offspring of this cross should carry the Rex chromosome and a homolog with a heterochromatic duplication and the euchromatin in normal, rather than inverted, sequence. Henceforth this new chromosome will be designated 0p/1;1/scS]sch. One thousand virgin female offspring were collected and mated singly with y- males (generation (2)). The Drosophila X chromosome is over 60 cM long. If the scSlscA chromosome has re-inverted, at least 50% of the offspring of a heterozygous female in generation (2) should be single crossovers. If the scSISCL has not re-inverted, there should be only a few offspring that look like single crossovers, and those will actually be double crossovers with one crossover outside the marked regions. Offspring were scored as y+ B, y 8+ (non-crossovers), y B, and y+ 3+ (crossovers). Six generation (2) females were found to have Dp/1;7/scS]sch chromosomes (frequency of reinversion = 0.6%). Table A shows a sample of the results of the progeny test, including the six reinversions and six chromosomes which remained inverted. 31 y 0v v f’B+ Rex Inf])scSlsc4, y+ 0v U B (I) X y cv v f'B+ Rex i Y + (2) youva Rex X _y_ Dp(1;1)8031304 OR In(1)3051304, y+ 00 U B Y (3) 9 9+ 8 y B 9+ 3+ DOD and NCO SCO offspring offspring Figure 7. Mating scheme to test for hairpin exchange. 32 Table 4. Test for hairpin pairing NCO and DCO SCO Maternal X chromosome Line # y y+ B y B 9* B+ 53A 32 17 28 16 108A 38 35 28 16 128A 46 27 42 7 Dpf151)3031$c4 230 38 26 28 15 708 6 4 5 6 744 9 11 11 10 107 25 27 I I 325 27 26 1 O 462 31 25 O 1 In(1)scSJse4 619 18 21 4 O 726 15 9 O 1 918 32 16 O O Heterozygous females of y 00 v fWRex and the chromosomes indicated were crossed to y/Y males. Apparent SCO's in the inversion cromosomes are DCO's with one event outside the marked regions. 33 Effect 9: Rex 93 maternally-transmitted chromosomes Males carrying a y+ fragment were also found among the progeny in generation (3) of Figure 7; they are the only y+ males that are not also cv. These males arose from spiral exchange of 55$]schin the embryos of the Rex-bearing mothers of generation (2). Thus, they indicate that Rex can act on chromosomes transmitted to the embryo from the mother as well as from the father. An accurate estimate of the frequency of the y+ males is difficult because it is so low. The progeny of 345 sc515cb/y cv v f Rex females was summed to 21,3IA. Since the offspring were not scored with regard to sex, this number was halved to estimate the number of daughters. There were 13 fragment-bearing males and 2 gynandromorphs, a frequency of detachment of 15/10672 X 100 = O.lAl%. Apparently, maternally-transmitted chromosomes are sensitive to Rex, though they may be less sensitive than paternally-transmitted chromosomes. The low frequency of detachment is considered further in the Discussion section below. Production gj fig deletions from eeSIQQh The production of 90% bb deleted chromosomes from YSX,YL, y y f R,y+(R0bbins, 1981) may be due to a peculiarity of this attached-XY. To determine whether SCSIschspiral exchange also produces a high frequency of bb deletions, y+fragments were examined. The usual mating which produces y+ fragment-bearing males is y/y cv v f Rex SILSCAR, cy y B/Y males. These y+ fragments cannot be females X In/1/SC studied further, however, because the males which bear them lack a Y chromosome and are sterile. Figure 8 shows the mating scheme used to 34 (J) y cv v f'Rex X In(1)3051364, cv U B FM7 / y (2) y 01) v f Rex X ySX-YL, In(1)EIV, y B In(1)scszsc4 ‘1 0 (3) y+ B sons = YSX-YL, In(1)EN, y B/y+ fragment Figure 8. Mating scheme to collect y+ fragments. 35 collect y+ fragments in fertile males. This scheme takes advantage of the result from the previous experiment, that Rex can act on maternally-transmitted chromosomes. Since the frequency of this event is so low, generation (2) was done in bottle matings of 15 females with 15 males. From these bottles, approximately 25 y+ 8 males were collected. Each of these males was mated to three C(7/RM; y v/0 females, and y+ progeny were selected to set up a stock. C/7/RM is a bb+ attached-X. Only eight of these males were fertile. Male progeny from these eight stocks were then crossed to C(7/DX, y f/Y females to test for bb penetrance. [If/0X is an attached-X totally lacking the rDNA. Progeny of this cross were scored for the presence and bb phenotypes of y+ f females. The results are shown in Table 5. The phenotypes range from Db] to bb. As was the case with the y+Y chromosomes recovered from the attached-XY (Robbins, 1981), there are no duplications recovered from spiral exchange in SCSISCA. Reciprocity g: the Rex event The products of hairpin pairing and exchange were also examined to determine the reason for the high frequency of bb deletions produced. Three types of males were collected from among the progeny of the six females carrying Dp/1;1/scS]sch chromosomes. The three phenotypes were y+ CV V B. the non-crossover progeny; y 8, single crossover progeny bearing the proximal end of Dp/1;1/scS]sch; and y+ cv v f 8+, single crossover progeny bearing the distal end of 0p/7;7/SCS]sch. Each male was used to set up a separate stock by mating him with three C/1/DX, y f females. Only the y 8 males from line #7AA were fertile, 36 .0a>+0500 +508005+ +0\M\NQ«~00 0 0>05 0:0 .50_+05:mm_0505 _0550+0E +0 +_:m05 05+ >_505oca 050 m0_0E0+ 0005kx m0_ms0+ .05+0_ + m0_0s0+ _0+0+ 00.0E0+ I m0_mE n m0_0E0+ _05+0_ . m0_ms0+ _05+0_ + m0.0£0+ 00 u 00:mc+0c0a 00 m0_0s\m0_0e0+ u >50>000L 0_0E00 .00_0s0+ xxx 0 «mem00 0+ 0000050 0503 m+505005+ +0 00+00_ec_ 05+ 050 MNI00500++0 m m 0 05_5005 m0_0z A000 00.0 0__._ _w_ FmN mmm 0 A_05+0_0 0._ 000.0 0 0 +0m w A_05+0_0 0._ N00.0 0 *— 0_m 0 A000 00.0 005.0 0v_ 0m 0mm m A_05+0_0 0._ 000.0 0 *N mmo v A_05+0_0 0.+ m00.0 .0 *_ mvm m A000 mm.0 mum._ mmm 00_ com m 0mmm+mv 0._ 0m_.0 0w 0 nmm _ 0050L+0500 >L0>000C 00.0500 00 % m0_0e00 k. m0_0z m m % +5050050 00 0.0800 m+c0emmc+ +0 +0 >+___50_50> 005500 .m 0_50+ 37 so no further testing could be done on this line. For the remaining five 0p/7;7/SCS]sch lines, one of each of the stocks (that is, the progeny from one of each class of males recovered in generation (3) of Figure 7) was used for testing. The non-crossover chromosomes from each of the five lines were tested to make sure these chromosomes were actually in normal sequence. Heterozygous y/y+ cv v 3 females were crossed to y males. The results are shown in Table 6. All of the five lines carry normal sequence chromosomes. Meiotic exchange in generation (2) of Figure 7 separates the two ends of the reinverted X chromosome. Thus, each class of single crossover males collected carries one of the two products of hairpin pairing and exchange. Both ends can be examined for their bb constitutions to determine if the exchange event is reciprocal. The y 8 males, carrying the proximal product from each of the five lines, were tested for Oh penetrance by crossing them to C/1/RM; y V/0- Sons from this mating were X/0 and their only source of rDNA was from the y B chromosome. The results are shown in Table 7. The phenotypes for this experiment range from bbI to bb. Testing the distal end, carried by the y+ 3+ males, required more manipulation. Before this end could be tested, the proximal rDNA which had come from the y chromosome had to be removed. The mating scheme to accomplish this is shown in Figure 9. The proximal rDNA from the y chromosome was replaced with the rDNA of y bblth. This chromosome had been previously tested and shown to be bb‘ and to have no other lA52 X-linked lethals. (A cross of y bb /FM7 X 38 Table 6. Retest of "reinversions" (Dp(1;1)SCSJSC4). NCO and DCO SCO Line # y y+ B y B y+ 8* 53A 225 101 143 89 108A 218 III 151 81 128A 394 230 256 143 230 133 58 91 46 708 109 53 96 62 y/y+ av v B females from the putative reinversion lines were crossed to y/Y males and the indicated phenotypes of their offspring were scored. 39 .00>+0c00 0\N\m m 0 0>05 050 50.+05000.0coc .0550+00 505+ +_:005 >.505o50 00.05 00055 * I I . 00.05 _05+0. + 00.05 _0+o+ I 00.05 00.050+ I 00.05 .05+0_ . 00.05 .05+0. + 00.05 0011.1 00505+0500 00 00.050+\00.05 I >50>0005 0.02 .00_050+ 0\o m .:m«~00 0+ 0000050 0503 005.. 50.050>5.05 00+00.0c. 05+ 505+ 00.05 M\m m +505.550000 0mmm+3 0.. 00... 00m 0 00. 00+ :28: 0.. 000.0 N ... 03 0mm 0.55%..8 .00 0.0.0 mm 00 .04 <00. 35. E0 0R0 RN 0: NR <00. :58: 0.. 000.0 _ .1 0:. (mm 00505+0c00 >50>0005 00.02 00 m m 00.02 m 0 00.0500 0 m 0 05.5 00 0.02 .00500050550 00+50>c.05 +0 050 .05.x050 +0 >+._.50.50> 005500 .5 0.505 4O (1) y+ 0” v f'B+ (500 from reinverted X) X y be453 Y FM? (2) 37+ 01) v f 8* X + (Oregon—R) + + + Z452 (3) y at) 1) it B bb x C(1)DX, y f (Stock) Y it Figure 9. Mating scheme for distal end of reinversions. 41 Inl1/schL358R, y/y+YBS produced no y 8+ females and many non-fM7 + + males.) Males with single crossover y+ CV y+ f 3 beASZ mAL mSIbR W 9 chromosomes were mated to females carrying In/1/w which is deficient for rDNA (Hilliker and Appels, 1982; see Figure 5). The results are shown in Table 8. Again, a range from [217] to bb is present. The data for the two ends is summarized in Table 9. It appears that the exchange event is not reciprocal, since in no case did one and receive a duplication of ribosomal cistrons when the other end received a deletion. .00>+0500 N\mmvN00 no «0000\0N053053 0 0>05 050 50.+05:w0.0505 .0550+00 505+ +_:005 5.505050 00.050+ 00055 * 00.050+ .05+0_ + 00.050+ +m .0+o+ 00.050+ _05+0_ + 00.050+ 00 +m H 00505+0500 00 00.05 +m\00.050+ +m u >50>0005 0.0500 .00_050+ m mm q5zm\100.5 0 «05053053 0+ 0000050 0503 005.. 50.050>5.05 00+00.05. 05+ 505+ 00.05 M\mwm~00 00 «0000 +505.550000 A_05+0.0 00.0 500.0 0 *0 0+0 M00 m5_ 005 A00 0._ 00._ wow 0 _mm 00m .m_ m 0505+00 0 N .00 . . 0505+00 0 _ .00 0 0mm 0 00m mmm 55 <0N_ A_05+0_0 0.. 000.0 0 0 0_N 0mm 00 <00— Aaa . . 0505+00 00 0 N00 0 mmm *m N5N 00m 00 50>0005 00.0500 00 +m 00.0500 +m 00.0500 m 00.02 53 00.02 m mm 0 05.5 00 0.0500 .00500050550 00+50>5.05 +0 050 _0+0.0 +0 >+...50.50> 005500 .0 0.505 43 .w Ucm N mm_nm._. EOL+ U®N.L®EE:M mLm m+mo 55 0565+0 .05+0. 0.. 00.0 00... 500.0 005 .05+0. 55 0565+0 0.. 0.. 000.0 00.. 0mm 55 0565+0 55 0565+0 .0.0 0.. 5.5.0 .00.0 <05. 55 .05+0. 55.0 0.. 055.0 000.0 <00. .05+0. 55 0565+0 0.. 00.0 000.0 500.0 <50 .ms.xo55 _m+0.0 _05.on5 _m+0.0 _mE.x655 _m+m.0 5 05.. 00>+05050 00505+0500 >50>0000 .00500050550 00+50>5.05 +0 0050 03+ +0 >+...50.50> 00 +0 500.500500 .0 0.505 44 DISCUSSION Results from the experiments above show that hairpin pairing and exchange can occur. In this configuration, the two ribosomal regions are paired in identical register (proximal to proximal and distal to distal) because the distal heterochromatic block has been inverted. This does not, however, increase the frequency of exchange; in fact, it would appear from the frequency of reinversion that the spiral configuration is favored: fragments are produced more frequently than reinversions. Apparently strict chromosome homology is not necessary for the Hex event. This is contrary to the conclusions from the evidence presented by Palumbo et a], (l973), who found that the rRNA genes had a, definite polarity for exchange. However, these data support the conclusions of Maddern (l98l), whose experiments showed that exchange could occur with the rDNA in either orientation. This certainly raises the possibility that, on the molecular level, the cistrons are in alternating direction. Laird and Chooi (l976) have reported a polarity of rRNA synthesis and a tandem gene arrangement of the ribosomal cistrons. Their evidence is based on electron micrographs of nascent rRNA fibers being synthesized from ribosomal DNA transcription units. However, even their longest DNA strands show only a few rDNA genes at a time, while each locus contains l30-300 genes (Ritossa ét 3],, l97l). From the present results, it seems likely that the genes are arranged in blocks, with the genes within each block being in tandem, but with adjacent blocks being in reverse order to each other. 45 The experiment which shows hairpin pairing also shows that Rex can act on chromosomes transmitted maternally. The frequency at which this occurs appears to be very low (O.lhl%); however, there are reasons why the number calculated may not be a completely accurate estimate. The way the experiment was conducted provides two of these reasons. (l) The offspring were not scored with regard to sex, and patroclinous (y) males, produced by X-nondisjunction or four-strand double exchanges with inverted chromosomes, were included with the other progeny. Thus, half the total progeny is an inflated estimate of the number of regular females. (2) Many of these matings were only counted once and discarded, since very few flies are needed to establish a 50% single crossover rate. Gynandromorphs are likely to be late hatching; detachment males may also have this problem. Thus, counting only once may have deflated the number of detachment products. However, another experiment done earlier did not have either of these flaws and still showed a very low frequency of detachment. The experiment was essentially the same except that it used the YSX,YL, y y f B,y+ chromosome instead of Inlf/scS‘LschR. Progeny were scored with regard to sex and patroclinous males were scored separately. The frequency of detachment was 0.062% (l/l+l6lh X lOO). Another possible explanation for the low frequency is genetic. The InH/scs’uscl'R stock may, like many other stocks, carry a suppressor of flex (see Appendix). Alternatively, the actual frequency of detachment of maternally-derived chromosomes may be low. This is the more interesting explanation, since it implies that the transcribed and/or translated product(s) of the Hex gene in the egg cytoplasm may 46 distinguish between maternally- and paternally-derived chromosomes. An experiment with the maternal and paternal nuclei carrying differently-marked responsive chromosomes could discriminate between these alternatives. In that case, both chromosomes would be subject to the same effects of any suppressor, and differences in their response would necessarily reflect a distinction between maternally- and paternally-transmitted chromosomes. The range of bb phenotypes among the y+ fragments is comparable to the range of bb seen in the y+Y chromosomes collected by Robbins. Eight y+ fragments is a small sample and the low fertility of the y+ fragment-bearing males (8/25) is perplexing. Effects of chromosomal content on male fertility are known (Lifschytz and Lindsley, l972), but sterility in those studies was caused by either a loss or translocation of part of the X. This situation is believed to keep the X from condensing early in spermatogenesis as it should. It is possible that, in the present case, the y+ fragment either fails to condense or causes the X to fail to condense. This particular kind of sterility effect, caused by a heterochromatic duplication, has not been previously reported. There is enough variation in this sample, small as it is, to say that exchange does not occur at only one site. There also is a lack of bb duplications, if their expected frequency is 50%. However, because of the unexplained sterility of nearly 70% of the males, there is doubt about whether these eight are a representative sample of the fragments produced. Larger fragments containing bb duplications may have been produced, but they may have been those which caused sterility. 47 The products of reinversion are more revealing, since both exchange products can be studied. A comparison of the two ends of the reinversion chromosomes (Table 9) shows a striking result. Though InH/scsu'schR carries two complete bb loci, in all cases reinversion results in both regions being deleted for rRNA cistrons. It must be emphasized that each line of 0p/7;7/SCS]sch chromosomes was derived from a single female, so that the two chromosome ends represent the two products of a single event. It is obvious that the reinversion does not involve a reciprocal exchange; some portion of the rDNA repeats is always deleted. There also appears to be a trend toward loss of the same size portion each time, since a bb lethal at one end is often accompanied by a bb at the other. Neither the number of observations nor the precision of measurement of phenotype is sufficient to push this idea too far, but the event caused by Rex is obviously more complex than simply pairing and exchange. Perhaps the rDNA regions pair in a loop, and when exchange occurs, the looped-out section is lost. If so, there may be times when the looped sections are excised but the breaks are not resolved as a crossover; this should result in a chromosome which remains inverted but loses rDNA at both ends. This hypothesis can be tested with an experiment like that shown in Figure 7, but in this case separating and testing the ends of a sample of Sls h Rex-exposed but still-inverted sc c '5 would be needed. CHAPTER 3 THE NATURE g: Ifli Rgx LOCUS One way to get an understanding of the function or action of a locus is to study its phenotype while varying the dose of the locus. The wild-type allele of a gene expresses a normal function. According to a classification scheme first proposed by Muller (l932), a mutant can differ in expression from the normal state in five ways. The mutant can be an amorph, which has no expression of the normal function; a hypomorph, whose expression is less than normal; 3 hypermorph, whose expression is greater than normal; a neomorph, which expresses a function not present in the wild-type state; or an antimorph, which expresses a function exactly the opposite of normal. Of these five, an antimorph is the most difficult to demonstrate conclusively. The proof requires the ability to add copies of the mutant allele and the normal allele to the genome, to show that the effects of one cancel the effects of the other. In Drosophila, this requires construction of small free duplications, since flies will not tolerate extensive aneuploidy. These duplications must contain the gene in question, must be adequately marked and must contain as little extraneous material as possible. They are difficult to produce and maintain, especially in multiple copies. With current techniques for transferring plasmid-like elements into the genome, testing for antimorphs may soon become more routine in Drosophila. Though this study does not directly test Rex for antimorphy, the other tests that are done nevertheless rule out antimorphy as a possibility. 48 49 Many dominant mutants are hypermorphs or neomorphs; however, a dominant could also be a hypomorph or amorph if there is a threshold level of expression necessary for the normal phenotype. These four types of expression can be distinguished by a series of dosage tests. To perform these tests, it would be ideal to have an inactive deficiency and a fully active duplication for the region where the mutant is located. If the mutant were a hypomorph or amorph, mutant individuals would have a phenotype similar to individuals carrying the deficiency. For a hypomorph, increasing the number of copies (the dose) of the mutant allele would bring the phenotype closer to the wild-type state; a true amorph would not show this response. The difference between a severe hypomorph and an amorph can be difficult to demonstrate. If the mutant were a hypermorph, individuals carrying it would be similar in phenotype to those carrying a duplication. The phenotype of individuals carrying a hypermorphic mutant would approach the normal phenotype when the mutant is heterozygous with a deficiency. If the mutant is a neomorph, there would be no clear dose response; neither the phenotype of the duplication nor that of the deficiency would resemble the mutant phenotype, nor would the duplication nor the deficiency suppress the mutant. The presence or absence of the mutant phenotype would depend only on the presence or absence of the mutant. The chromosome which carries Rex (y cv v f Rex) is also bb, which raises the possibility that Rex is caused by a loss of function in the ribosomal cistrons. Since Rex maps to the centric X heterochromatin (Robbins, l98l), duplications and deletions of this region were used to construct a dose series. Two chromosomes are of use here, 5O lL 4R Sl 4 8R (=schsc8) and In/1/scS sc (=35 sc ). The former is a In/1/scthc deletion of most of the heterochromatin, and the latter is a duplication. Each was constructed from two scute inversions (Figure 3). The mating schemes used to construct the relevant females to test for the Hex phenotype are shown in Figures l0 and ll. These schemes produced females with a series of different doses of the X heterochromatin. The females were mated to In/1/scSILschR , cv v B . + + males and scored for the production of y 8 sons and gynandromorphs. If the Hex allele is an amorph or hypomorph, it will be mimicked by h 8 Sl h . . . 3c sc and suppressed by sc 3c ; experiments l, 2, 3 and A will give similar positive results, and experiments 5, 6 and 7 will show no Rex activity. Conversely, the Re; allele is a hypermorph, the duplication will mimic and the deficiency suppress the response; experiments l, h, 5, 6 and 7 will show positive results while 2 and 3 will show no activity. If the Hex allele is a neomorph, only those experiments where the female parent carries Rex, i,e,, l, 3, h and 6, will show any activity, and those genotypes lacking Rex, i,e,, 2, 5 and 7, will be I negative. 5i .m+mo+ omOUiuousomc 0cm _0L+cou L0+ 050:0m mc_+mz .o. 0L:m_u 0.0030053 00.0.5503. 0Q 0853855 50 5 m mo:mwow%mm\8mm 5m. 00 a. m o no wommmom mommwow%mQ\+ «m. {80.5 2. 80m .... o 0.0 0 0. «0.00000 0 8055.... 0. on 0. 53 8055.5 0 no 0 : 0 ”500000. 50 03E 5.. 80.55.305.00 05 ammuaooom 05 000000.000. x 5.. 80.55.595.005: \lilill’llk mam 5555/ .0 \ m 53E N ammkooom. 05 fl 52 N .m+m0+ mmOniummmoL0c_ L05 msocom mc_+mz ... 0L:m_m :0550055030\20550055530 55. 5 0 00 .0005550 5 amm\50550555005 50. o m «vomwmom +\50550055000 55. o m amomNMUm o m «vomNmom 55. 80mg% 9 no 0 555 5 505 5 0 o 0 «wommmom N o 0 «vowmmoo x 80m.% 9 on 0 mam x: a m N .N III 52% o m «vomwmow m 53 RESULTS Table l0 summarizes the results of all seven experiments. Detachment frequencies in experiment 3 required special calculation. senses/fragment males will not survive unless the y+ heterochromatic fragment generated has enough ribosomal cistrons to cover their complete absence in the X. The first of the two detachment frequencies shown for experiment 3 is calculated on the assumption that all schsc8-bearing detachment males survive. For the second frequency, the number of detachment males and gynandromorphs is doubled in the numerator and the denominator on the assumption that no sch358 detachments survive. Given that a minority of the y+ fragments probably are [2b+ (Robbins, l98l and Chapter 2 of this work), the true frequency should be an intermediate value, somewhat closer to the second figure. Table l0 shows that neither 35$]sch, the duplication, nor schSCB, the deficiency, mimics nor eliminates the effect of Rex. Experiments 1, 3, h and 6, where Rex is present in the mothers, all produce Rex-induced detachments and gynandromorphs among the progeny. (The somewhat low activity seen in experiment 6 may indicate the presence of 51Lsch, y y stock; see Appendix) a suppressor in the Inl1/sc Experiments 2, 5 and 7, where Rex is not present in the mothers, show no detachment activity. These results are consistent with the hypothesis that Rex is a neomorph; that is, the Rex allele is expressing a new function. 54 ._ 0_000 c_ 00 00+0_:0_00 +c0scom+0o a .co_+mcm_0x0 L0+ +x0+ 00m * ... 000 o. 00L:m_0 00m .0mc.+mE L00 00.0 . 0 . 0 050. 005. 5555\05 .5. 00.0 5 .5 05. 5 0505 ..00 emm\05 .0. 0.0 0 0 05 . .50. 55.5 {05 .0. 5.. 0 0. 0 0 550. 000. 80.58005 .0. *5..-00.0 0 0. 0.. . 0005 5505 550000. .0. 0.0 0 0 50 0 0000 0000 .55}. .5. 0.5 0. 00 . 5 05.5 0005 +080. ... +c0Ecom+00 050L050L0 00_0E 00.0E m0_0E0+ m0_0E 00.0E0+ m0_mE0+ +c00c00 icmc>o +c0ELUm+0Q _m+c0cma c0_+oc:mm_0cocix mc_cam++0 cm_:m0m +0 0000 .m+002_c00x0 0m0000 +0 0+_:mom .o. 0_000 55 DISCUSSION The results in Table l0 show clearly that Rex is most likely a neomorph. There are two assumptions made in the design of this experiment which, if incorrect, could alter this conclusi0n. First, it is assumed that since Rex maps proximal to car (Robbins, l98l), it is located in the heterochromatin. car is, however, about 3 cM from the locus generally used as the most proximal euchromatic marker, suppressor of forked (sulf/; Lindsley and Grell, l968). Rex might be located in these 3 cM of euchromatin. Second, it is assumed that Rex is located in the portion of the heterochromatin deficient in scbsc8 and duplicated in SCSISCA. As Figure 3 shows, the heterochromatic breakpoints of these two chromosomes are not at the very boundaries of the heterochromatin. A small amount of heterochromatin is present in schsc8, and is in only, one copy in scS]sch. Rex might be located in this area of the heterochromatin. The only experimental evidence bearing on this point comes from the original mapping of Rex: 75 of 76 crossovers between can and the centromere also separate car and Rex. This puts Rex 75/76 of the way from car to the centromere, and is fairly good evidence that Rex is in the heterochromatin. A stronger case will have to await more precise mapping. A neomorph presents a difficult problem in analysis. Mutants are often used to deduce normal gene function; by studying the phenotype and dose response of a mutant one can make deductions about the wild-type gene product. The assumptions necessary for thes deductions 56 may not be valid with a neomorph. Rex could be a lesion in a wild-type gene which causes it to function in a totally new and different way. For example, perhaps Rex+ is a gene which works at a particular time in development, and its functioning at a different time produces a wholly different effect. On the other hand, there may be no entity which corresponds to a Rex+ allele. The Rex phenotype may be caused by a wholly new element introduced into the genome. As such, it could work in one of two ways; it could either control the normal cellular “equipment'I for a new purpose, or it could bring in its own ”equipment”. The former would imply that there are other genes that make transcription and/or translation products capable of performing at least some of the sequence of events induced by Rex (Gl pairing between two 00 loci in the X and subsequent exchange). The latter would imply that the Rex element is large and complex enough to produce all the transcription and/or translation products necessary to perform these events. More information on the structure of Rex, most likely on a molecular level, is needed to choose between these possibilities. The chromosome which carries Rex (y cvrv f Rex) also carries a bb mutation. This raised the possibility that Rex was caused by a loss of function in the ribosomal cistrons. The experiments with schsc8 above clearly show that this is not the case. That chromosome is completely devoid of rDNA, but does not have any Rex-like activity. However, it may be that the Rex element is an insertion into the rDNA that causes a loss of function of some genes. This opens the possibility that Rex can be cloned, a normally impossible task with a heterochromatic gene having an unknown product because of the repeated nature of heterochromatin. 57 CHAPTER h SIGNIFICANCE Aflg RECOMMENDATIONS Rex is a dominant, neomorphic, maternal*effect variant causing a pairing and exchange-like event in the ribosomal cistrons which is always accompanied by a loss of some of these cistrons. This event occurs in Gl of the first or second mitotic division of the offspring of Rex mothers. The pairing can occur in two configurations, resulting in either the loss of the X euchromatin between the two rDNA regions or its inversion. Both maternally- and paternally-transmitted chromosomes are affected, though there may be some preference for paternally-transmitted chromosomes. All of the information to date indicates that Rex is an intriguing locus which has a rather bizarre effect on early mitotic chromosome behavior. It is interesting enough to warrant study on its own merits. The studies presented here have suggested some lines of future research. Rex can act on a number of different chromosomes; the only criterion so far discovered is that the chromosome must carry two rDNA regions. The range of chromosomes on which Rex can act has not been fully explored. Many of the problems are technical: chromosomes and/or genotypes must be devised to enable the detection of the product of the Rex event. For example, one might like to ask how far apart in the chromosome the two rDNA regions need to be for Rex to act on them. This will probably be difficult to answer. Because of Drosophila's sex determination system, mitotic loss of a whole X (which Rex causes) simply converts female cells to male cells. If the rDNA were moved to 58 59 the center of the euchromatin instead of the distal end, though, Rex would cause loss of only half an X. This would result in cells that were either extremely hypoploid female or extremely hyperploid male, and neither is likely to survive. It might be possible to answer this question using the hairpin-type exchange, if a chromosome with an rDNA duplication in the center of the euchromatin can be constructed. An easier question to answer about the range of Rex is whether Rex can cause interchromosomal events. Since most Rex-induced events occur at first Gl before pronuclear fusion, it will be necessary to transmit both chromosomes through one gamete. One of the ways to test this would be to set up a cross of y/y cv v f Rex females to y/Y/y+YRS males. One of these Y's has its short arm marked with y+ and its long arm marked with RS. Since the two Y's of an X/Y/Y male tend to separate at meiosis I (Lindsley and Grell, l968). half the sperm will be y/y+YRS. If an interchromosomal exchange between the X and Y5 occurs in this pronucleus, the result will be an X with YLBS attached and a free y+YS. These y/X,YLBS/y+Ys females will look like their y/y/y+YBS sisters, and progeny tests will be required to determine if Rex has caused an interchromosomal exchange. If it has, the females with the y/X,YLRS/y+yS genotype will produce RS daughters and y+ sons among their progeny, whereas the y/y/y+YBS females will produce y+ BS sons. Another possible interchromosomal event would be the production of attached-X chromosomes. This could be accomplished by having an exchange event between the rDNA of a normal sequence X and an X with the rDNA moved distally, like InH/scsl (Figure 3). This event would 60 form a compound reverse acrocentric chromosome and a free y+ fragment. Again, since this is a mitotic event, a progeny test would be required to find those females that produced only y matroclinous daughters, indicating that they were carrying an attached-X which had lost the y+ marker. The normal-sequence X in the original cross of the above experiment would have to carry Rex. This raises another question which would have to be answered before that experiment could be done. Can Rex act on the chromosome it is in? Since there is some suspicion that the site of the Rex locus may be the bb cistrons, it may be that these cistrons are altered so that they can't participate in the Rex event. A test of this possibility is straightforward. Rex could be put into Dp/1;1/scS]sch by meiotic recombination, and the resulting 0p/1;1/scS]sch, cv v 8 Rex chromosome could be tested for detachment production. The results in Chapter 2 have shown that Rex can act on maternally-transmitted chromosomes, albeit at a lower frequency than on paternally-transmitted chromosomes. This raises the possibility that Rex has a preference for paternally-transmitted chromosomes, and that some part of the Rex system can distinguish between paternal and maternal chromosomes in the embryo. An experiment with differentially-marked responsive chromosomes brought into the embryo simultaneously in the maternal and paternal nuclei could test this hypothesis. In this case any suppressing background effects on Rex should be equal for the two sensitive chromosomes, since they would both be exposed to egg cytoplasm from the same mother. A reciprocal 61 experiment, which reversed which chromosome came from which parent, would correct for any possible difference in sensitivity of the two chromosomes. The suggestion was made in Chapter 3 that Rex may be located in the bb locus. Because of the repetitive nature of heterochromatin, it is impossible to “walk” into this region to clone a gene located there. The ability to clone heterochromatic genes generally depends on having a known gene product, which is lacking for Rex. The ribosomal cistrons, however, have been cloned, and if Rex is an element inserted into these cistrons, it should be possible to find it by looking for differences in the restriction fragments of the ribosomal cistrons from the Rex chromosome. If Rex can be cloned, the molecular structure of the Rex locus could provide some insight into its function. In particular, it would be interesting to know just how large the Rex element is. If Rex is large, it might be producing all the transcription and/or translation product(s) necessary to express its phenotype; if it is small and simple, it is more likely to be changing or controlling normal cellular products to its own ends. A clone of Rex could also be used to search for Rex loci in other stocks or in natural populations. Since Rex was not discovered as the direct result of a mutagenesis, but was found in a pre-existing stock, it is possible that its occurrence is widespread. This would have implications for the function of Rex and its evolutionary significance. There are also a number of aspects of the Rex phenotype that suggest it may be connected to other phenomena which have been studied. The potential role of Rex in three of these will be examined below. 62 The first is the control of gene copy number at the bb locus, the site of action for Rex. The second is the functioning of genes in the heterochromatin, where Rex is located. The third is early mitotic chromosome stability, which Rex affects and which Rex+, if a + . Rex eXIsts, may control. The fig locus Since the site of action of Rex is the bb locus, it is worthwhile to review the work that has been done on these genes. The bb locus has been the subject of many studies, both genetic and molecular, and it has a number of interesting properties. This locus and its regulators are the only X-linked genes known in Drosophila which have homologous genes on the Y. They are also the only essential genes known in the X heterochromatin or on the Y (Lindsley et 3],, I960). The phenotype of homozygous bb flies is short, thin thoracic bristles, delayed development and, in some backgrounds, etched abdominal cuticle (Lindsley and Grell, l968). Ritossa et a], (l966) first showed that the bb locus was the site of the DNA coding for ribosomal RNA. It is now believed through many sources of evidence (Ritossa and Spiegelman, I965; and for reviews see Atwood, I969; Birnstiel et a],, l97l; Long and Dawid, l980) that this locus also exactly corresponds to the cytological locus of the nucleolus organizer region (NUR). The bb locus consists of a series of repeats of the genes for the l8$ and 285 rRNA's (Wellauer and Dawid, l977; White and Hogness, l977; Glover and Hogness, I977). Each repeat consists of an external transcribed spacer, the l8$ gene, a short internal transcribed spacer 63 which probably contains the 5.85 and ZS rRNA genes (Pellegrini et 3],, I977), and the 28S gene (Long and Dawid, l980). The foregoing is transcribed into one long rRNA precursor, which is then processed. There follows on the DNA a non-transcribed spacer region of variable length (A*ZO kb; lndik and Tartof, I980). The genes on the X and Y both follow this basic pattern. RNA fingerprinting techniques have shown that the rRNA from the X and Y are identical. However, there are a number of differences in the DNA. The IBS and 28S and the internal and external spacers of the X are homologous to the same regions of the Y. The non-transcribed spacer varies in size and has internal repeats (Wellauer and Dawid, l978; Long and Dawid, l979b); the distribution of size classes of this spacer is not the same in the X and Y. There are also insertion sequences in some of the 285 genes. The first to be described were the type I sequences; these are not found in the Y (Tartof and Dawid, I976; Dawid et 3],, l978). Most type I insertions have a length of 5 kb; there are also some of l.0 and 0.5 kb. The smaller insertions are homologous to parts of the 5 kb insertion (Dawid et 37,, I978). As well as being dispersed throughout the 285 genes of the X chromosome (Tartof and Dawid, I976), these sequences are found in the heterochromatin of all four arms of the second and third chromosomes and at a euchromatic site on the fourth chromosome (Peacock et 3],, l98l). From 49% (Wellauer et 8]., 1978) to 602 (Kidd and Glover, l98l) of the 285 genes on the X contain type I inserts. Another class of inserts has also been described; these type 2 inserts have no homology to type I (Dawid et 3],, I978). They have lengths of l.5 to 4 kb (Wellauer and Dawid, l978; Wellauer et 37.. l978) and 64 insert at a slightly different place in the 285 gene (Dawid and Rebbert, l98l; Kidd and Glover, l98l). They are found in the rDNA of both the X and Y chromosomes (Wellauer et 3],, l978), and I63 of the 285 genes of both have type 2 inserts (Wellauer and Dawid, l978; Kidd and Glover, l98l). It is thought that the genes with 285 inserts do not contribute to the active cellular rRNA (Long and Dawid, l980). Transcripts with type I inserts are extremely rare (Long and Dawid, l979a; Long et 3],, l98l). Type 2 transcripts are found at somewhat higher levels, but they are not long enough to contain a complete 285 gene (Kidd and Glover, l98l). If insertion-containing genes are truly inactive and the distribution figures for these sequences quoted above are accurate, then 65% to 75% of the X rDNA sequences are nonfunctional. The number of total copies of the rDNA repeat varies within and among strains, and a wild-type locus contains l30-300 genes (Ritossa 8t 37.. l97l). If the total number of copies in a fly is below I30, the bb phenotype is expressed. Regulation of the bb locus is complicated and not completely understood. An in-depth study of the literature uncovers much conflict, due not only to the complexity of this locus but also the personal disagreements of the investigators. There are two major phenomena involved. These are compensation and magnification/reduction. Both are changes in the copy number of the rDNA cistrons. (Selective amplification, another type of copy number change seen in the oocytes of some species, in which extra copies of the rDNA are produced extrachromosomally in a cascade fashion, has not 65 been found in Dr050phila. Birnstiel et 3],, l97l; Long and Dawid, l980). Compensation was first described by Tartof in l97l. Usually, the amount of rDNA is directly proportional to the number of NUR's in the genome (Ritossa and Spiegelman, I965). However, in X/0 males and X/XNO- females there are on average I50 more rRNA genes per X as compared to normal X/X females or X/Y males (Tartof, I97l). This increase in rDNA is somatic and is not inherited; X/X daughters of X/XNO- mothers do not get the extra I50 genes. Copy number in all experiments was determined by RNA/DNA saturation hybridization. The X in X/Ybb— males also compensates (Tartof, l973a). There has been much controversy about the nature of the Ybb- chromosome. Ritossa (l968a) reported several stocks with Ybb- chromosomes which had nearly wild-type amounts of rDNA (about I00 genes). However, Xbb/Ybb- flies have a bb phenotype and XNO-/Ybb- and XNO-leb-leb- are inviable. Ritossa explained this apparent contradiction by concluding that these Ybb- chromosomes contained nonfunctional ribosomal genes. Tartof (I973a) disputed this conclusion, and reported that the Ybb- chromosome carried only about Al genes; the phenotypes of flies with this chromosome can be explained by assuming deletion of ribosomal genes, therefore, and not a loss of function. The "error” in Ritossa's results, Tartof explained, was that Ritossa measured Ybb- in X/Ybb- flies, where the X is compensating. However, a later report (Henderson and Ritossa, I970) confirms Ritossa's earlier finding, in this case in attached-XY/Ybb- flies, where there is no compensation. Endow (l982b) attempted to resolve this controversy by measuring rDNA 66 content in a different way. She ran Southern transfer blots of ybb' DNA and probed it with 32P—IabeIIed cloned rDNA. The Ybb' DNA did hybridize with the cloned DNA, indicating that this Y carries some ribosomal genes. Endow then probed the Y DNA with cloned type 2 insert. She concluded that the insert does hybridize, the Ybb- does carry the type 2 insert, and therefore its genes are largely nonfunctional (Long and Dawid, I980; Franz and Kunz, l98l). It is difficult to evaluate these differing results. Endow's measurements are not quantitative. The high error inherent in saturation hybridization determinations (discussed below) hinders a direct comparison of Ritossa's and Tartof's data. A simple resolution to the conflict has been proposed (R. S. Hawley, personal communication). It is possible that Tartof and Ritossa have been examining Ybb- chromosomes of different origin. If this is the case, it may be that Tartof‘s Ybb- is a simple deletion, whereas Ritossa's carries nonfunctional genes. Two mechanisms for the regulation of compensation have been proposed. In I973, Spear and Gall published a report comparing rDNA levels in diploid and polytene tissues of X/0 and X/X flies. They found that, in diploid tissues, the number of rRNA genes is proportional to the number of NDR’S present (i,e,, one in X/0 and two in X/X). However, in polytene tissues, the number of rRNA genes is constant regardless of the number of NUR'S. In D, melanogastep different regions of the chromosome show independent polytenization (Tartof, I975). Euchromatic regions go through about ten rounds of DNA replication with no cell division. Centric heterochromatic regions, 67 along with most of the Y chromosome, replicate little, if at all. The rDNA is intermediate, going through seven or eight rounds of replication (Spear and Gall, I973). Spear and Gall explained compensation with the hypothesis that the higher level of rDNA per X chromosome in X/0, X/Ybb-, and X/XNO- comes from looking at a mixture of diploid and polytene tissues in adult flies. The consequence of looking at diploid tissue, with its ribosomal gene ratio of 2:I in X/X v3, X/0, and polytene tissue, with its higher copy number but a ratio of Izl, is that overall X/0 flies appear to have more than half the rDNA of X/X flies, and the X seems to be compensating. There are a number of problems with this hypothesis. The proportion of adult tissue that is polytene and the extent of its polytenization is unknown. It is also not known whether all polytene tissues underreplicate the rDNA; this has only been examined in salivary glands. Spear and Gall's hypothesis does not account for the observation that diploid tissues can show compensation (Spear, l97A). The 55 RNA genes, which are wholly euchromatic, also show compensation (Tartof, I975; Long and Dawid, l980); they have never been shown to undergo independent polytenization. 0n the other hand, the Y chromosome rDNA does undergo independent polytenization, but it does not show compensation. Endow (l982a) later offered an extension and further explanation of Spear and Gall's hypothesis. She showed that the EcoRI rDNA restriction patterns of six different chromosomes, the X's and Y's from Canton-S, Oregon-R and OK-l strains, are all unique. Combination of any two of these chromosomes in diploid tissues gives a restriction 68 pattern which looks like a mixture of the two chromosomes. In polytene tissues, however, one pattern predominates. She was able to construct a dominance hierarchy among the six chromosomes. She believes that only one ”HR is active in polytene tissues. This would explain Spear and Gall's results, since both X/0 and X/X flies would only have one active ”HR in polytene tissues. Compensation, then, would not be the result of turning up the rDNA copy number in X/0 flies but of turning it down in X/X. Though more detailed, Endow's hypothesis suffers from the same faults as that of Spear and Gall (see above). In addition, restriction patterns on her gels are not always unambiguous. She makes no attempt to map the locus responsible for inactivating one NUR. Autosomal background effects were not controlled, and inasmuch as they may be a factor, the dominance hierarchy constructed may be altered. A different sort of explanation for compensation has been offered by Procunier and Tartof (I977). Working under the hypothesis that the observations of compensation imply the existence of a locus which senses a deficiency in its homolog and directs compensation, they set out to find it. Since it is logical that this locus would be near the rDNA, they used a set of heterochromatic deficiencies and duplications to define the locus. They found such a locus, which they named compensatory response (er) in the distal penultimate one-eighth of the X heterochromatin. This gene has two functions. It works in trans to sense the presence or absence of its homolog, and in cis to control the disproportionate replication of the rDNA if its homolog is absent. The cis function is dependent on contiguity of cp+ with the rDNA; . . . . + chromosomes with InverSIons thch move or away from the rDNA can be 69 sensed by their homologs as CP+ (i,e,, they do not induce compensation in their homologs) but do not themselves compensate when their homologs are cp—. Procunier and Tartof posit that the Y carries a cp+ locus too far away from its rDNA to control disproportionate replication, because Y chromosomes behave like X inversions. However, they did not attempt to map cp in the Y, and it seems equally likely that the Y cr may simply have lost its cis function. There is a later report (Krider et 3],, I979) that the Canton-S X chromosome also fails to compensate. Tartof’s model would attribute this to loss of cp+. However, in all these studies, no attempt has been made test or control for autosomal effects on compensation. Would the Canton-S X compensate in an Oregon-R background, where cp was originally found? What happens to an Oregon-R cp+ X in a Canton-S background? The explanations of Procunier and Tartof, and Endow and Spear and Gall are not mutually exclusive, and it is possible some melding of the models, perhaps with an controlling independent polytenization of the rDNA, more accurately describesthe mechanism of compensation. However, since all these studies fail to control for background effects on the system, it is likely that the actual control of compensation is either much different or much more complex than what has been envisioned. The other major phenomenon associated with the rDNA genes in Drosophila is magnification/reduction. When bb mutations are kept in homozygous stocks, so that the flies are phenotypically bb, the bb phenotype has a tendency to disappear. This was generally ascribed to the accumulation of modifiers (Lindsley and Grell, l968). In addition, bb mutations of unknown origin recur in many laboratory stocks. 7O Normally one would not expect to see high rates of mutation in a repeated gene, since the effect of any single point mutation will be swamped out. In order to see a mutant phenotype, one would need many simultaneous point mutations or a large deletion. However, bb mutants are frequently discovered in stocks. In I968, Ritossa explained both these occurences as increases and decreases in the number of rRNA cistrons. His experiments showed that an Xbb carried in a stock over a Ybb- has a bb+ phenotype after three generations. If this X is now put into X/0 males (with half the . + autosomes having been replaced) those males are stIll bb . If the new + + . . Xbb+ is now carried over Ybb or Xbb , It can revert once agaIn to bb. Ritossa called the process of increasing the rDNA from bb to bb+ I'magnification". Early reports on magnification, especially those involving measurement of rDNA levels by saturation hybridization, are confusing and at times invalid. Since compensation was not discovered until l97l. flies with genotypes likely to be compensating were often used to measure rDNA levels. In those experiments, it is impossible to separate compensation from magnification effects (for example see Ritossa et 3],, I97l). There is also the difficulty, described above, of assigning the correct number of genes to Ybb- when this chromosome is used as a reference. Furthermore, Spear (l97h) has shown that great .variation can occur between separate saturation hybridization experiments, even among those done under identical conditions in the same laboratory. An analysis of variance shows these to be true differences in hybridization. Unless two experiments are run 7i simultaneously in the same rRNA solution, a comparison of their results may be meaningless. Despite these problems, most of the observations of magnification are clear. In addition to the original observations described above (Ritossa, l968b; Ritossa and Scala, I969) it was shown (Boncinelli et 37,, I972) that Ybb chromosomes can also magnify when they occur in a male with either an Xbb] or an XNO-. Two very diferent theories have been offered to account for magnification. Ritossa (I972) has proposed that magnification is due to extrachromosomal rDNA synthesis into small circles followed by integration. The synthesis occurs in all cells of the male, but integration can only occur in the germ line. (There have been no reports of magnification in the female.) These extrachromosomal circles are transcribed, but they do not contribute to the cell's mature rRNA pool (Ritossa et 3],, I97l, I973). The integration occurs during meiosis (Ritossa, I973; Ritossa et 3],, I973). In experiments designed to detect exchange between the X and Y, Ritossa (I973) found a 20- to 30-fold increase in this exchange among flies undergoing magnification. This would be expected if, during integration, it were optional which rDNA region the extrachromosomal genes will be integrated into. Simultaneous integration of a large circle into both the X and the Y would sometimes lead to an exchange between them. The production of extrachromosomal circles could explain the observations that magnification occurs as a stepwise phenomenon and that newly magnified bb chromosomes are unstable. A number of studies have shown that it takes about 3 generations for a bb chromosome to become 72 completely bb+ (Ritossa, l968b; Ritossa and Scala, I969; Henderson and Ritossa, I970; Ritossa et 3].. l97l; Boncinelli et 3],, I972). There is also evidence that after these chromosomes have been under magnifying conditions for about four generations, there is less tendency to revert to bb than in earlier generations (Ritossa, l968b; Ritossa and Scala, I969; Henderson and Ritossa, I970; Boncinelli et 3],, I972). Details of the mechanism with which Ritossa et 3], (I973) propose to connect this instability and the production of extrachromosomal copies of rDNA are quite vague. Tartof (l973b, I974) has presented evidence that magnification occurs by a wholly different mechanism. He proposes that unequal sister chromatid exchange (SCE) during mitosis in the male germ line is responsible. This idea was originally proposed by Atwood (I969). Tartof presents a number of lines of evidence to support this model. A fluctuation test shows that magnified bb+ chromosomes tend to be produced in clusters. This indicates a premeiotic origin for these chromosomes. These new bb+ chromosomes are stably inherited. A loss of ribosomal cistrons occurs simultaneously, so that some of the progeny are [2])1 while others are 33+. Tartof has named this process of loss ”reduction”, and it would be expected to occur under his hypothesis since one chromatid will receive more rDNA and its sister less. These reduced chromosomes can be re-magnified in further generations. Somatic mosaics are also produced, and Xbb/Ybb- males with mosaic cells show a h- to 5-fold increase in magnification over unselected males. No magnification is seen in ring-Xbb chromosomes. This would also be expected from the proposed mechanism, since SCE 73 would lead to dicentric bridges and interlocked rings, which might be broken or lost in later mitotic or meiotic divisions. Deciding between these two conflicting models is quite difficult. Ritossa's hypothesis of synthesis of extrachromosomal rDNA has a parallel in the process of amplification, which is not known for Drosophila ribosomal cistrons but does occur in Drosophila histone genes (Long and Dawid, I980). There are some reports of extrachromosomal circles seen in male germ cells undergoing magnification (Long and Dawid, l980) and of unintegrated ribosomal genes in some tissues (Zuchowski and Harford, l976a, l976b). The extrachromosomal synthesis mechanism would explain the observed increase in X-Y exchange. Double exchange that might occur could explain how the X and Y have maintained extremely homologous rRNA genes through evolution (Maden and Tartof, I97A; Long and Dawid, I979). However, there is recent evidence (Tartof and Dawid, I976; Long and Dawid, I980; Peacock et 3],, I98I) that the X and Y have a number of sequence differences in the ribosomal cistrons. If X-Y exchanges have occurred with any frequency, those differences should have been eliminated. Ritossa's evidence for meiotic timing of synthesis and integration (Ritossa et 3]., I973) is very weak. It is based on the notion that the NUR of the Y is inverted with respect to the NOR of the X (Palumbo et 3],, I973). That idea was later refuted by Maddern (l98l) who showed that exchange between the two regions can occur in either order. The results presented in Chapter 2 of the present work support Maddern's conclusion. The extrachromosomal synthesis hypothesis is very complex, requires a meiotic exchange-like event in 74 male 0, melanogastep where meiotic exchange does not usually occur, and requires a cellular mechanism which can differentiate between rRNA transcribed from extrachromosomal rDNA and rRNA transcribed from chromosomal rDNA. It can explain the rectification of bb mutants after they occur, but does not explain their high rate of formation. Tartof's hypothesis of sister chromatid exchange would appear to be a simpler and more complete model. It can explain both the production and loss of bb mutations. The two phenomena are seen as the results of a regular occurrence that keeps the number of rRNA cistrons in flux around some mean, with large deletions and duplications being selectively eliminated. There is evidence that unequal SCE in the rDNA occurs during mitotic divisions in yeast (Szostak and Wu, l980). In addition, regularly occurring SCE could account for the maintenance of intraspecies homogeneity by ”horizontal“ evolution of this tandem gene array. Computer simulation studies (Smith, I973; Tartof, l973b) have shown that reasonable levels of SCE can lead to crossover fixation of a single sequence in a short time. However, Tartof's arguments are not completely convincing. Like male meiotic recombination, SCE does not generally occur in D. melanogastep, either meiotically (Novitski, I952) or mitotically (Gatti et 3],, I979). Thus magnification would require a special mechanism for the ribosomal cistrons to overcome this absence. The SCE hypothesis offers no explanation for early instability of magnified chromosomes as seen by Ritossa. Tartof tried to refute Ritossa's evidence of instability, but the test he used was inappropriate. Ritossa maintained the magnified bb chromosomes over a 00+, then tested for production of bb. Tartof maintained a magnified 75 bb over bbo (SCLSCB) thus selecting for maintenance of bb+ and guaranteeing stability. There is always the problem of a difference in stocks, and it is quite possible that the instability that Ritossa has seen is not a general phenomenon. These stock and background differences could also account for the more gradual accumulation of rRNA cistrons seen by Ritossa (three to four generations to 33+) as opposed to Tartof's observations (usually only one generation to bb+). Tartof's evidence for somatic mosaics is also questionable. Small spots of a variably expressed mutant like bb must be extremely difficult to unambiguously identify. These spots could also be explained by somatic nondisjunction, especially since they occur with equal frequency in magnifying and nonmagnifying males (Tartof, l973b, I97A). In addition, one of Tartof's major arguments for SCE is that a ring-Xbb does not magnify. However, Ybb- also fails to magnify although, even by Tartof's measurements (Tartof I973a), Ybb- is not a complete deletion and should be capable of SCE. Other chromosomes with bb mutations which fail to magnify have also been discovered (R. S. Hawley, personal communication). The factors influencing the change in copy number of ribosomal genes are far from completely understood. Since the site of action of Rex is the ribosomal cistrons, it is possible that Rex may be involved in the regulation of this process. Some preliminary studies on how Rex affects magnification have been done (R. S. Hawley, personal communication). Unlike many other bb mutants, the bb mutant present on the Rex chromosome (Chapter 3) does not magnify. However, Rex can induce magnification in other bb mutants, and this magnification can 76 occur in females, a heretofore unreported event. Rex also shows a maternal effect on the _magnification of paternally-transmitted chromosomes. Some of the XN0-/bb daughters from a cross of Rex/XNO- X bb/Y have a bb+phenotype. Further testing has shown that the bb chromosome has magnified to bb+. This appears to be magnification in early mitoses of female embryos. The meaning of these results is not yet clear. No work has been done to discover if Rex affects compensation. A number of lines of enquiry can be suggested. Does the Rex chromosome have an active cr+ locus? Does Rex have a maternal effect on compensation? If so, is it the maternally- or paternally-transmitted chromsome which is affected? The problem with proceeding with any of these experiments is the uncertainty about the validity of the studies and the merits of the conclusions on magnification and compensation to date. The mechanism of magnification is in doubt; however, the existence of the phenomenon of magnification has not been questioned. Studies of how Rex affects that phenomenon are therefore possible, and may eventually shed some light on its mechanism. In the case of compensation, however, it is possible that the phenomenon itself may be simply an artifact of the technique used to measure gene copy number. Even if compensation does exist, there remain many questions about background effects. Any test of Rex's effect on compensation would introduce yet another set of background effects from the Rex stock which would be impossible to separate from the consequences of Rex itself. Until these doubts about compensation are answered, it seems better to leave the Rex experiments undone. 77 Heterochromatin Rex is probably located in the centric X heterochromatin. Though the heterochromatin of Drosophila has been extensively studied, only a little is understood about its structure and even ,less is understood about its function. It is known to be condensed when the euchromatin is not, chiefly during mitotic interphase (Heitz, I928; Shah et 3],, I973). It is therefore believed to be inactive during much of the cell cycle. The heterochromatin is located around the centromeres in Drosophila, and much of it is composed of repeated sequences (Kram et 3],, I972; Brutlag et 3],, I977). These areas lack exchange (Baker, I959). and conventional mutation analysis has uncovered very few genes in them, considering their physical size. Some of the difficulty in finding genes in the heterochromatin may stem from a difficulty in defining the exact boundaries of these regions. There is some evidence (Schalet and Lefevre, I973; Lifschytz, l978) that there is no actual junction of euchromatin and heterochomatin, but there is an alternation of euchromatic and heterochromatic elements, gradually becoming more euchromatic distally and more heterochromatic proximally. Though not exactly ”genes”, in the commom understanding of the word, the elements controlling a number of phenomena have been shown to reside in the heterochromatin. The pairing sites for the X and Y in males are located in the heterochromatin of these chromosomes (Lindsley and Sandler, I958; Cooper, I964). Heterochromatin is also responsible for position-effect variegation, a phenomenon in which euchromatic genes moved in or near broken heterochromatin fail to function uniformly. There are also apparently some factors in centric 78 heterochromatin responsible for the kinetic strength of centromeres during division (Lindsley and Novitski, I958). In addition, there are a few more ordinary genes which have been mapped to the heterochromatin. These genes do, however, have some unusual properties which distinguish them from euchromatic genes. Most of the systematic studies of heterochromatin have focussed on the X and Y. Nevertheless, there are a few genes known in the autosomal heterochromatin. Hilliker (I976) analyzed the heterochromatin of the second chromosome, and uncovered nine complementation groups of lethal alleles. They were mapped by deficiencies, are late larval and pupal lethals and are apparently non-repetitive. This is the first indication of lethal genes located in Drosophila heterochromatin, and it shatters the notion that this material is largely dispensible. The other genes known to be in the autosomal heterochromatin are associated with the segregation distortion (50) system (Ganetsky, I977). 50 causes meiotic drive: SDISD+ fails to produce 50+ sperm. This is due to a gametic failure causing 50+ sperm to “suicide“, possibly because of a failure of the 50+ chromosomes to condense normally during spermatogenesis. Three (components of the 50 system have been identified. The 3d locus itself is euchromatic, located on the second chromosome. The responder (Rsp) locus is in the sens heterochromatin of 2R. Chromosomes containing Sd+ Rsp suicide in S the presence of an 5d RspIn homolog. A third part of the system is an enhancer locus ( [(Sd/ ) which is in the heterochromatin of 2]. 79 The Y chromosome, which is wholly heterochromatic, has six male fertility genes (Kennison, l98l). Four of these are located on YL and two are on YS. All of them function during spermatogenesis (Williamson, I976) and all are necessary for the production of motile sperm. The Y chromosome also contains a bb locus. This is the only locus shared by both the X and the Y and the only essential heterochromatic locus known for either of these chromosomes (Lindsley et 3],, I960). There is also a partially active cp locus on the Y, (Procunier and Tartof, I977) as was discussed above. A few genes have also been discovered in the X heterochromatin. In addition to a bb locus and a fully active cp locus, there are also apparently a number of loci located distally which affect early development and which interact with a group of closely linked genes on the second chromosome. These second chromosome genes are daughterless (d3; Bell, l95h), abnormal oocyte (aha; Sandler et 6].. I968), da-abo-like, wavoid-like, and hold-up (dal, wd], and flap; Sandler, I977). These five genes are euchromatic and are located within 2.5 cM of each other on the left arm of the second chromosome. They are all recessive, hypomorphic, maternal-effect mutants which cause unequal recovery of the sexes in their progeny. All produce zygotic mortality before egg hatch, and all interact with the sex-chromosome heterochromatin. In most cases, an increase in X or Y heterochromatin in either the mothers or their offspring serves to rescue all or part of this mortality. In the case of abo. the region of this rescue effect has been mapped to the penultimate one-eighth of the X heterochromatin (Parry and Sandler, I974). Sandler (I977) has 80 postulated that the other second chromosome loci also have corresponding, though different, sites of interaction in the heterochromatin. There is already some evidence of this for da (Sandler, I972, I977). These heterochromatic loci can apparently substitute for the euchromatic loci early in development, if the egg is deficient for the maternally-packaged d3 or abo products. It does not seem unreasonable, then, to assume that these are actually heterochromatic copies of the euchromatic genes, and that they function early in development before the euchromatic genes are active. Except for the ribosomal cistrons, these heterochromatic genes all have one thing in common. They are all active at times when the rest of the genome is not. For 50 and the Y fertility factors, this is during spermatogenesis, when the rest of the chromatin is condensed. The second chromosome genes which affect development function in the very early embryo, which normally depends on maternal gene products (Robbins, I980, I983). It is possible that heterochromatin's function is to contain certain genes meant to be ”on” at times when the euchromatic genes are ”off”. Rex also functions at one of these times -- in the early embryo before the pronuclei have even fused (Sonnenblick, I950; Davring and Sunner, I973). A number of experiments could clarify the possible develoomental role of Rex and its interaction with the rest of the genome. A mutagenesis experiment to look for enhancers of Rex is currently underway. The Rex phenotype is difficult to use for precise mapping; at this time only one to two percent of what were originally female embryos show the effects of Rex when their mothers carry Rex. Boosting 8i this frequency to ten to twenty percent would drastically reduce the numbers of progeny needed to do a more precise mapping experiment -- to map Rex between su/f/ and bb, for example. Meiotic-recombination and deletion mapping would assure that Rex was indeed located within the heterochromatin. Assuming that the heterochromatic location were confirmed, experiments could then proceed to see if Rex has any interaction with aha or the other developmental genes near it. If the cloning of Rex (mentioned above) could be accomplished, in situ hybridization of the cloned Rex DNA to Drosophila chromosomes would show if Rex has a euchromatic sister locus. (Interestingly enough, 3b0 also has an interaction with the rDNA (Krider and Levine, I975; Krider et 3],, I979) lending support to the notion that Rex may be related to abo and its nearby loci.) Mitotic Chromosome Instability Genetic elements which affect mitotic chromosome behavior have been previously identified. Some of these have effects similar to Rex, though all have differences. It is intriguing to speculate that Rex is part of a system, with these elements, that controls early chromosome behavior. An examination of the effects and actions of these elements is therefore worthwhile. One of the first described instances of mitotic chromosome instability was the loss of ring-X chromosomes (reviewed by Leigh, I976). These rings tend to be lost in early cleavage divisions, producing gynandromorphs. X/0 males are also produced; these were assumed to come from meiotic loss of rings, but there is no evidence 82 that these are not losses very early in mitosis. The loss of rings has several similarities to Rex action. The instability disappears over several generations in culture, which might be due to the accumulation of modifiers (see Chapter I and Appendix). There is evidence of factors in the oocyte which influence the stability of rings and can cause stable rings to become unstable. This suggests that there might be a maternal effect involved. Finally, the locus causing ring instability was mapped to the centric X heterochromatin. Taken together these facts suggest that there might be a factor similar to Rex in unstable ring systems. Rex causes a mitotic recombination event. Mitotic recombination has been studied extensively both for itself and for its use in determining developmental pathways. Mitotic exchange occurs spontaneously in Drosophila melanogastep at a low rate (Garcia-Bellido, I972). Unlike the meiotic process, mitotic recombination is prOportional to physical chromosome length; that is, the heterochromatin exchanges as readily as the euchromatin. Mitotic recombination requires homology (Ripoll and Garcia-Bellido, l978), and it increases in frequency when cells are treated with X-rays. Therefore, its spontaneous occurrence may be due to chromosome breaks. A number of meiotic recombination-defective loci which also have mitotic effects have been identified (Baker et 3]., I978). In addition to increasing mitotic recombination, these loci cause breakage and loss. They are believed to be defective in cellular DNA repair functions. One of these, mei-41, has recently been shown to induce interchanges between the X and the Y which involve the X rDNA (Hawley 83 and Tartof, I983). Rex has a similar effect, and might also be involved in this recombination-repair system. There are three other mutations affecting mitotic chromosome behavior. The first is mitotic induced loss (mit; Gelbart, I974). mit is a recessive, maternal-effect mutant which causes loss of the X or fourth chromosomes in mit or mit+ progeny. The X or 4 lost can be of maternal or paternal origin. The loss occurs during the third or fourth mitotic division. Loss of the second and third chromosomes was not tested, but mit does not cause loss of a Y or an attached-XY. mit maps in the proximal X euchromatin. Modifiers tend to accumulate in mit stocks, but these modifiers are readily removed by outcrossing. The second mutant of interest is paternal induced loss (pal; Baker, I975). This mutant is located on the left arm of 2. When the progeny of homozygous p3] males are examined, there are many that have lost one or more paternal chromosomes. Baker assumes that this is meiotic loss; however, there is no way to decide if an embryo never received a chromosome or if it was lost during GI of the first embryonic division. Moreover, there is a high frequency of gynandromorphs and mosaics in the progeny as well. It seems simpler to attribute both of these results to early mitotic chromosome loss. The effect is seen equally in pa] and p3]+ progeny of both sexes, but in all cases only the paternally-derived chromosomes are ever lost. Offspring missing one chromosome are more likely to be missing two than would be expected if the two events were independent. There is also an interaction between p3] and the X heterochromatin: some X chromosomes are less sensitive to loss induced by p3] and this insensitivity .maps 84 to the X heterochromatin. The third mutant is claret-nondisjunctional (cand; Davis, I969). It causes meiosis | nondisjunction and maternal chromosome loss; again the loss probably occurs at early mitotic divisions in the zygote. Exchange is normal in meiosis, so the nondisjunction is not due to a failure of pairing or a disruption of recombination. The loss is a maternal effect. All of these mutants are similar in causing chromosome loss during mitotic divisions. The last two are similar to Rex in that they may be active as early as Gl of the first mitotic division. A meiotic nondisjunction phenotype similar to cand has been seen in Rex crosses; Tables l-3 show an excess of exceptional male offspring. In other experiments involving Rex (S. Keller and L. Robbins, personal communication), exceptional female offspring, many of which are recombinant and homozygous for distal markers, have been seen. These observations taken together resemble the products from meiosis I nondisjunction (Bridges, l9l6; Merriam and Frost, l96h): X-diplo exceptions homozygous for distal markers and, at a higher frequency, X-nullo exceptions. Some of the excess of patroclinous males may also come from early mitotic chromosome loss. The similarities in phenotype of mit, p3], 53nd, and Rex suggest that Rex may be part of a group of genes controlling early chromosome behavior. It would be interesting to study their interactions. For example, does the Rex locus correspond to the X heterochromatic locus which is insensitive to pal? If so, other Rex alleles should be present in either the pal-sensitive or the pal-insensitive stocks Baker 85 (I975) identified. In addition, the enhancer now being sought would provide a frequency of Rex products high enough that a cytological examination of the early mitotic divisions of Rex-affected embryos might be feasible, following the technique of Zalokar, et 3], (I980). This examination would provide further insight into the Rex event. The enhancer would also enable further investigation of some of the rarer products associated with Rex. These include the putative meiosis I nondisjunction products mentioned above. In addition, occasionally flies are produced from Rex crosses which can only be explained by postulating that an early mitotic exchange followed by a meiosis | (reductionaI)-Iike segregation has occurred. These products are recovered at frequencies an order of magnitude lower than are detachment products. To date it has not been clear that these are definitely products of Rex; an increase in their frequency with an enhancer would back the notion that they are, make feasible experiments to map them unequivocally to Rex, and allow a more complete study of them. APPENDIX APPENDIX Suppressor of flex In the course of experiments done to investigate Rex, another element of the system has been uncovered. This element acts as a suppressor of the Rex effect, and has been designated Su/Rexl. Because of the suppressor's detrimental effect on some experiments and its possible utility in others, a number of experiments have been done to discover something about its mode of action and location. Although many aspects of this element or elements remain unclear, the following has been shown: the suppressor acts as a dominant; it is not a maternally-transmitted cytoplasmic element; and it is present in different chromosomal locations in different stocks. It may be present in more than one chromosome location in some stocks. Figure l2 shows a general mating scheme similar to the one that originally uncovered the suppression and which has been used since to test stocks for the suppressor. The suppressor acts in the heterozygote and is therefore a dominant. Table II presents the results of tests of a number of X chromosome stocks for the presence of a suppressor. 86 87 (1) y ...... Rex X PuTaTive suppressing sTock Y YSX-YL U B- +/0 (2) y _______ Rex . suppressor X ’ y f, y ’ 4 "‘ OR Ih(1)scSlsc4, cv v B/O (3) DeTachmenT producTs Figure 12. Mating scheme for detection of suppressor(s). 88 .. 0_000 0. mm 00+0_:0_00 +00Ecom+0o & .N. 0L:m_0 000 .mmc.+me L00 8055050000.. 560.% o 06 m5 Nom.o m 000 soo.% a 00 mm mxommmosom 80%.... o no 0.. 0+0. 0 055.0 0 500.. 5555 i0- 05 55% 80mg% 0 so 0 M New 5: o x . .5 mm. o m 00. 0 52% mam $3 N 52M 80% 50 mm «~0535N>RQ 500.0 0 no 5 o A J .00 o o mmm 0. 560.% o 00 m 030m>NOEOI +00Ecom+00 0+0300L0 m0_mE0+ L0_:m0m o0+m0+ 00_me0+ x00+m +0 +c0oc00 +c0scom+0o +0 00>+000w 00500050Lco x0m .mx00+m mc_mm0c00:m ... 0_000 89 An attempt was made to map the location of the suppressor found in the y cv v f cap stock. From previous experiments (Robbins, unpublished data) it was known that the suppressor did not segregate with the y cv v f cap X chromosome or any of the recombinant X chromosomes from a y cv v f cap/y heterozygote (for mating scheme see Figure I3). The stock y ; T]2;3]e/SM7;IM2 ; spapo} was used to replace the autosomes of the y cv v f can stock both singly and in combination. SM” and [M2 are, respectively, second and third chromosome balancers. The mating scheme for this experiment is shown in Figure IA, and the data are presented in Table l2. Eight different genotypes are produced, and each was tested for Rex activity with YSX,YL, y y f B,y+. All the data are statistically homogeneous, and each of the classes shows evidence of suppression. These data can be explained by at least three possibilities. The autosome balancer stock itself might carry a suppressor, or the suppressor might be a non-chromosomal element which is transmitted through maternal cytoplasm, or there might be multiple suppressing elements in the y cv v f cap stock. To test the first of these, mapping experiments were set up using the autosome-balancer stock as the source of the suppressor, as shown in Figure I5. The data are presented in Table I3. Once again, the data show no particular trends. Most of the classes (I, 2, 3, 5, and 6) have detachment frequencies that do not indicate suppression. A contingency test shows that all classes are homogeneous. Of tests of the relevant combinations of genotypes, spapol vs. spapoH, 3M1 vs. SM”, ”12 vs. ”12+, only the last was 2 significant (X = 7.309 with I d.f., 0.0l>p>0.005) and this is in the 9O ycvvfcar Xi Y y y 0?) v f car X DfYUwI'Jl, y2 sn Rex y .Y 030(1)er1, yz sn Rex X ySXq/‘L.9 y I) fB°y+ . 8papal 9 (av 72 1“ car) recombinant 0 ’ Spapol no suppression Figure 13. Loss of Su(Rex) effect. y on v f’car Sn? y cv v f’car X 9i y T(2;3)e spaPOZ y on v f’ca:// Y SMZ;TM2 spapoZ Sn? Sn? y (recombinant) SM] TMZ spaPOZ (1) Dle)er1, yg sn Rex (2) (3) (4) (5) (6) (7) (8) Figure I4. Mapping Sn(Rex) from y cv v f'car y (recombinant) H H N H I! I! H S . L . (l)-(8) x YXY’yvay+- LfiYl)wrdj, y2 en Rex (1) - (8) 8ND TM2 + + Sn? TM2 + + SM] Sn? + + Sn? Sn? + + SWE flM2 + + 5n? TM2 + + SWO Sn? + + Sn? Sn? + + 0 detachment products +++ w+Y +++ Spapol + Spapol + spapoz spaPOZ Sn? Sn? Sn? Sn? Spaged ’ spapoZ stock: mating scheme. 92 Table 12. Mapping Sn(Rex) from y on v f'car stock: results. Regular Detachment Percent Genotype females products detachment (I) SM] TM2 poZ 430 0 0 (2) Sn? TM2 pol 1063 I 0.188 (3) SMU Sn? poZ 791 0 O (4) Sn? Sn? pol 1185 0 0 (5) 8M0 TM? Sn? 397 I 0.504 (6) Sn? TM? Sn? 1661 I 0.120 (7) SM? Sn? Sn? 1037 3 0.579 (8) Sn? Sn? Sn? 1693 1 0.118 Totals 8257 7 0.170 YSX'YL, y n f’B-y+/0 ; spapaZ/spapoZ males were mated to females with the indicated genotypes (for mating see Figure 14). % detachment calculated as in Table l. 93 y T(2;3)e spapoZ y X y l y SMJ;TM2 spaPOZ ) y SM] (Sn?) TMZ (Sn? spaPOZ (Sn?) X Df(1)er1, yz sn Rex y + + + wty (1) - (8) (1) DflzmrJl, y2 sn Rex SM] (Sn?) TM? (Sn?) Ispapoz (Su?) y + + + SM] (Sn?) TM? (Sn?) + (2) n + + + (3) n SM] (Sn?) + spa?"Z (Sn?) + + + '7 (4) " 5M1 (Sn.) .i. ..i. + + + + TM2 (Sn?) spapoz (Sn?) + + + (5) H + TM2 (Sn?) + + ' + + 01 (7) H :— -+-'_ spap (SH?) + + + (8) ,, __+_. _+- _.+_ + + + YSXoYL, y 7) fBoy+ . spaPOZ (z) — (8) X , 0 spapoZ de tacbment products , pol Figure 15. Mapping Sn(}?ex) from y;%;% stock: mating scheme. 94 T(2;3)e , §pai90Z Table 13. Mapping Su(Rex) from y ; SM1°TM2 , spapaz stock: results Regular Detachment Percent Genotype females products detachment (I) SME TM? pol 469 5 2.09 (2) SMQ TM? + 187 2 2.09 (3) 5M0 + poZ 547 4 1.44 (4) SMQ + + 905 3 0.65 (5) + TMZ pol 374 3 1.58 (6) + TMZ + 742 7 1.85 (7) + + poZ 821 2 0.48 (8) + + + 1087 3 0.55 Totals 5132 29 1.12 YSX-YZ, y v f'B-yt/O ; spaPOZ/spapoZ males were mated to females with the indicated genotypes (for mating see Figure 15). % detachment calculated as In Table l. 95 wrong direction: TM? showed less suppression than TMQ+. These data do not support the hypothesis that the autosome-balancer stock carries a suppressor. The second possibility, that Su/Rex/ is a maternally-transmitted cytoplasmic element in the y cy v f cap stock, was tested using the mating scheme shown in Figure I6. Genotypically identical F1 females are derived from reciprocal crosses. If SulRex/ is maternally transmitted, the F1 females from cross (I) should produce fewer detachment products than those from cross (2). However, this result was not obtained. Females from cross (I) produced 2860 regular daughters and 7 detachment offspring, while females from cross (2) produced 2142 regular daughters and 9 detachment offspring. The data are homogeneous and both detachment frequencies (O.l7% and 0.42% respectively) show evidence of suppression. Thus, neither a suppressor carried by the autosomal balancer stock nor a maternally-transmitted cytoplasmic element can explain the results of the mapping experiment in Table 12. This leaves the possibility of multiple suppressing elements. A contrary result, however, is obtained in experiments with stocks of the X chromosome balancer FM7 (Merriam, 1968). As was noted above, previous experiments had shown that all crossover and noncrossover daughters from y cv v f cap/y mothers and Rex fathers no longer show any suppression (Figure 13). However, a very different result is obtained when FM7 is used instead of y cv v f cap, as diagrammed in Figure 17. In the first generation, FM7 is transmitted from a male to eliminate any maternal cytoplasmic elements. In the next two 96 .+o0++0 c0_000000:0 .mcc0+mE L0+ +00+ 0+ 020500 mc_+02 .0. 0L:m_0 mnoxfiosm 5503500500 5 80m 500 a 5 m» 55. 500.% 0 00 m 0 80m 500 3 m M 580m530 500.0 0 00 m N 80m 500 3 0 80m 500 3 m N 580m530 500 5.0 00 m 97 EM7, y2 ma v sn B Y 9 FM7 y .... X .... y 1y” mainTained as a selecfion sfock for Two generafions FM7 X y 02) v f Rex y Y FM7 X YSX°YL, y v f B-y+ ; spaPOZ y av 22 f Rex l 0 spapoZ suppression Figure 17. Suppression by EM7. 98 generations, it is again outcrossed to a non-suppressing stock and transmitted through females to permit free re-assortment of the autosomes. The data for the last mating in Figure l7 are shown on the \ third line of Table ll. fM7 does not lose the ability to suppress after passing through the fM7/y heterozygote. In fact, the data in lines 2 and 3 of Table ll are homogeneous. In the FM7 stocks, then, a single suppressor is present on the X chromosome. Because FM7 is a very effective crossover suppressor, however, it would be quite difficult to map the position of Su/Rex/ within this chromosome. It may therefore be concluded that more than one genetic locus can function as a suppressor of Re; action. A suppressor definitely resides on the X chromosome in the fM7 stock, and the y cv V f cap stock definitely lacks a suppressor on the X. The attempts to map SulRex/ in y cv v f car may have failed either because there are actually a number of sites on different chromosomes in that stock, each functioning as suppressors, or because the autosome-balancer stock actually does have a suppressor, but it is located on the X, a notion that would also account for the intermediate detachment frequencies found for all the autosome-balancer stock's derivatives. LIST OF REFERENCES LIST OF REFERENCES Atwood, K. C. I969. Some aspects of the bobbed problem in Drosophila. Genetics 6I, Suppl. I: 3l9-327. Baker, 8. S. I975. Paternal loss (pal): a meiotic mutant in DPOSObela melanogaster causing loss of paternal chromosomes. Genetics 80: 267-296. Baker, B. S., A. T. C. Carpenter and P. Ripoll. I978. The utilization during mitotic cell division of loci controlling meiotic recombination and disjunction in Drosophila melanogastep. Genetics 90: 53I-578. Baker, w. K. I959. Crossing-over in heterochromatin. Am. Natl. 92: 59-60. Baker, W. K. l97l. Evidence for position-effect suppression of the ribosomal RNA cistrons in Drosophila melanogastep. PNAS USA 68: 2h72-2h76. Bell, A. E. 195A. A gene in Drosophila melanogaster that produces all male progeny. Genetics 39: 958-959. Birnstiel, M. L., M. Chipcase and J. Speirs. l97l. The ribosomal RNA cistrons. Prog. Nucl. Acid Res. Mol. Biol. ll: 35l-389. Boncinelli, E., F. Graziani, L. Polito, C. Malva and F. Ritossa. I972. rDNA magnification at the bobbed locus of the Y chromosome in Urosophila melanogaster. Cell Differentiation 1: l33-lh2. Bridges, C. B. I9I6. Nondisjunction as a proof of the chromosome theory of heredity. Genetics I: I-52, l07-l63. Brutlag, 0., R. Appels, E. S. Dennis and w. J. Peacock. I977. Highly repeated DNA in Drosophila melanogaster. J. Mol. Biol. II2: 31-97. Cooper, K. w. I956. Phenotypic effects of Y chromosome hyperploidy in Drosophila melanogastep and their relation to variegation. Genetics hi: 2h2-26h. Cooper, K. w. I959. Cytogenetic analysis of the major heterochromatic elements (especially Xh and Y) in Drosophila melanogastep and the theory of ”heterochromatin". Chromosoma IO: 535-588. 99 IOO Cooper, K. W. I96h. Meiotic conjunctive elements not involving chiasmata. PNAS 52: l2h8-I255. Davis, D. G. I969. Chromosome behavior under the influence of claret-nondisjunctional in Drosophila melanogastep. Genetics 61: 577-59“- Davring, L. and M. Sunner. I973. Female meiosis and embryonic mitosis in Drosophila melanogaster I. Meiosis and fertilization. Hereditas 73: 5l-6h. Dawid, I. B. and M. L. Rebbert. I981. Nucleotide sequences at the boundaries between gene and insertion regions in the rDNA of Drosophila melanogaster. Nucl. Acids Res. 9: 50ll-5020. Dawid, I. B., P. K. Wellauer and E. 0. Long. I978. Ribosomal DNA in Drosophila melanogastep l. Isolation and characterization of cloned fragments. J. Mol. Biol. l26: 7h9-768. Endow, S. A. I982a. Polytenization of the ribosomal genes on the X and Y chromosomes of Drosophila melanogastep. Genetics IOO: 375-385. Endow, S. A. I982b. Molecular characterization of ribosomal genes on the Ybb- chromosome of Drosophila melanogastep. Genetics I02: 91-99. Franz, G. and w. Kunz. l98l. lntervening sequences in ribosomal RNA genes and bobbed phenotype in Drosophila hydei. Nature 292: 638-6h0. Ganetsky, 8. I977. 0n the components of segregation distortion in Drosophila melanogastep. Genetics 86: 321-355. Garcia-Bellido, A. I972. Some parameters of mitotic recombination in Drosophila melanogaster. Mol. Gen. Genet. 115: Sh-72- Gatti, M., G. Santini, S. Pimpinelli and G. Olivieri. I979. Lack of spontaneous sister chromatid exchanges in somatic cells of Drosophila melanogaster. Genetics 9l: 255-27“. Gelbart, w. M. I97h. A new mutant controlling mitotic chromosome disjunction in Drosophila melanogastep. Genetics 76: 5I-63. Glover, D. M. and D. S. Hogness. I977. A novel arrangement of the I85 and 285 sequences in a repeating unit of Dr030phila melanogaster rDNA. Cell l0: l67-l76. Graziani, F., R. Caiizzi and S. Gargano. I977. Circular ribosomal DNA during ribosomal magnification in DPOSOpfiiIa melanogastep. J. Mol. Biol. ll2: h9-63. IOI Hannah, A. l95l. Localization and function of heterochromatin in Drosophila melanogaster. Adv. Genet. A: 87-125. Hawley, R. S. and K. D. Tartof. I983. The effect of mei-41 on rDNA redundancy in Drosophila melanogaster. Genetics IOA: 63-80. Heitz, E. I928. Das Heterochromatin der Moose. Jahr. Wiss. Botan. 69: 762-8I8. Henderson, A. and F. Ritossa. I970. 0n the inheritance of rDNA of magnified bobbed loci in Drosophila melanogaster. Genetics 66: h63-h73. Hilliker, A. J. I976. Genetic analysis of the centromeric heterochromatin of chromosome 2 in Drosophila melanogaster: deficiency mapping of EMS-induced lethal complementation groups. Genetics 83: 765-782. Hilliker, A. J. and R. Appels. I982. PIeiotropic effects associated with the deletion of heterochromatin surrounding the rDNA on the X-chromosome of Drosophila. Chromosoma 86: h69-h90. Hilliker, A. J., R. AppeIs and J. Schalet. I980. The genetic analysis of D. melanogaster heterochromatin. Cell 21: 607-6I9. Indik, Z. K. and K. D. Tartof. I980. Long spacers among ribosomal genes of Drosophila melanogaster. Nature 28h: h77-h79- Kennison, J. A. I981. The genetic and cytological organization of the Y chromosome of Drosophila melanogaster. Genetics 98: 529-5h8. Kidd, S. J. and D. M. Glover. I98l. Drosophila melanogaster ribosomal DNA containing Type II insertions is variably transcribed in different strains and tissues. J. Mol. Biol. l5I: 6h5-662. Kram. R., M. Botcham and J. E. Hearst. I972. Arrangement of highly reiterated DNA sequences in the centric heterochromatin of Drosophila melanogaster: evidence for interspersed spacer DNA. J. Mol. Biol. 6h: IO3-II7. Krider, H. M. and B. I. Levine. I975. Studies on the mutation abnormal oocyte and its interaction with the ribosomal DNA of Drosophila melanogaster. Genetics 8I: 50l-5I3. Krider, H. M., B. Yedvobnick and B. I. Levine. I979. The effect of abo phenotypic expression on ribosomal DNA instabilities in Drosophila melanogaster. Genetics 92: 879-839- Leigh, B. I976. Ring chromosomes and radiation induced chromosome loss. Chapter 12. In: The Genetics and Biology of Drosophila, Vol. 1b. Edited by M. Ashburner and E. Novitski. Academic Press, London. IOZ Lifschytz, E. I978. Fine-structure analysis and genetic organization at the base of the X chromosome in Drosophila melanogaster. Genetics 88: A57-A67. Lifschytz, E. and D. L. Lindsley. I972. The role of X chromosome inactivation during spermatogenesis. PNAS USA 69: l82-I86. Lindsley, D. L. I955. Spermatogonial exchange between the X and Y chromosomes of Drosophila melanogaster. Genetics A0: 2A-AA. Lindsley, D. L., R. Appels and A. J. Hilliker. I982. The right breakpoint of Inlf/chZ subdivides the ribosomal DNA. DIS 58: 99-I00. Lindsley, D. L., C. u. Edington and E. S. von Halle. I960. Sex linked recessive lethals in Drosophila whose expression is suppressed by the Y chromosome. Genetics A5: I6A9-I670. Lindsley, D. L. and E. H. Grell. I968. Genetic variations of Drosophila melanogaster. Carnegie Inst. Wash. Publ. 627. Lindsley, D. L. and E. Novitski. I958. Localization of the genetic factors responsible for the kinetic activity of X chromosomes of Drosophila melanagaster. Genetics A3: 790-798- Lindsley, D. L. and E. Novitski. I959. Compound chromosomes involving the X and Y chromosomes of Drosophila melanogaster. Genetics AA: I87-196. Lindsley, D. L. and L. Sandler. I958. The meiotic behavior of grossly deleted X chromosomes in Drosophila melanogaster. Genetics A3: 5A8'563. Long, E. 0. and M. Collins, B. I. Kiefer and I. B. Dawid. I98l. Expression of the ribosomal DNA insertions in bobbed mutants of Drosophila melanogaster. Mol. Gen. Genet. l82: 377‘383- Long, E. 0. and I. B. Dawid. l979a. Expression of ribosomal DNA insertions in Drosophila melanogaster. Cell I8: 1185-1196. Long, E. 0. and I. B. Dawid. l979b. Restriction analysis of spacers in ribosomal DNA of Drosophila melanogaster. Nucl. Acids Res. 7: 205-2I5. Long, E. O. and I. B. Dawid. I980. Repeated genes in eukaryotes. Ann. Rev. Biochem. A9: 727-76A. Maddern, R. H. l98l. Exchange between the ribosomal RNA genes of X and Y chromosomes in DPOSOphiIa melanogaster males. Genet. Res. 38: I-7. 103 Maden, B. E. H. and K. Tartof. l97A. Nature of the ribosomal RNA transcribed fron the X and Y chromosomes of Drosophila melanogaster. J. Mol. Biol. 90: 51-6A. Merriam, J. R. and J. Frost. l96A. Exchange and disjunction of the X chromosomes in Drosophila melanogaster. Genetics A9: lO9-I22. Muller, H. J. I932. Further studies on the nature and causes of gene mutations. Proc. 6th Intl. Cong. Genet. I: 2I3-255. Novitski, E. I952. The genetic consequences of anaphase bridge formation in Drosophila. Genetics 37: 270-287. Palumbo, G., R. Caizzi and F. Ritossa. I973. Relative orientation with respect to the centromere of ribosomal RNA genes of the X and Y chromosomes of Drosophila melanogaster. PNAS USA 70: I883-I885. Parry, D. M. and L. Sandler. l97A. The genetic identification of a heterochromatic segment on the X chromosome of Drosophila melanogaster. Genetics 77: 535-539- Peacock, w. J., R. Appels, S. Endow, and D. Glover. I98l. Chromosomal distribution of the major insert in Drosophila melanogaster 28$ rRNA genes. Genet. Res. 37: 209-2IA. Pellegrini, M., J. Manning and N. Davidson. I977. Sequence arrangement of the rDNA of Drosophila melanogaster. Cell 10: 213-22A. Procunier, J. D. and K. D. Tartof. I977. A genetic locus having trans and contiguous cis functions that control the disproportionate replication of ribosomal RNA genes in Drosophila melanogaster. Genetics 88: 67-79. Ripoll. P. and A. Garcia-Bellido. I978. Mitotic recombination in the heterochromatin of the sex chromosomes of Drosophila melanogaster. Genetics 90: 93-IOA. Ritossa, F. M. l968a. Nonoperative DNA complementary to ribosomal RNA. PNAS USA 59: IIZA-II3I. Ritossa, F. M. I968b. Unstable redundancy of genes for ribosomal RNA. PNAS USA 60: 509-5I6. Ritossa, F. I972. A procedure for magnification of lethal deletions of genes for ribosomal RNA. Nature New Biol. 2A0: IO9-lll. Ritossa, F. I973. Crossing-over between X and Y chromosomes during ribosomal DNA magnification in Drosophila melanogaster. PNAS USA 70: l950-l95A. IO4 Ritossa, F. M., K. C. Atwood and S. Spiegelman. I966. A molecular explanation of the bobbed mutants of Drosophila as partial deficiencies of ”ribosomal” DNA. Genetics 5A: 8l9-83A. Ritossa, F., C. Malva, E. Boncinelli, F. Graziani and L. Polito. l97l. The first steps of magnification of DNA complementary to ribosomal RNA in Drosophila melanogaster. PNAS USA 68: 1580-158A. Ritossa, F. M., and G. Scala. I969. Equilibrium variations in the redundancy of rDNA in Drosophila melanogaster. Genetics 6i, suppl. I: 305-3l7. Ritossa, F., F. Scalenghe, N. di Turi and A. M. Contini. I973. 0n the cell stage of X—Y recombination during rDNA magnification in Drosophila. Cold Spr. Harb. Symp. Quant. Biol. 38: A83-A90. Ritossa, F. M. and S. Spiegelman. I965. Localization of DNA complementary to ribosomal RNA in the nucleolus organizer region of Drosophila melanogaster. PNAS USA 53: 737-7A5- Robbins, L. G. I980. Maternal-zygotic lethal interactions in Drosophila melanogaster: the effects of deficiencies in the zeste-white region of the X chromosome. Genetics 96: I87-200. Robbins, L. G. l98l. Genetically induced mitotic exchange in the heterochromatin of Drosophila melanogaster. Genetics 99: AA3-A59. Robbins, L. G. I983. Maternal-zygotic lethal interactions in Drosophila melanogaster: zeste-white region single cistron mutations. Genetics l03: 633-6A8. Sandler, L. I970. The regulation of sex chromosomes heterochromatic activity by an autosomal gene in Drosophila melanogaster. Genetics 6A: A8l-A93. Sandler, L. I972. 0n the genetic control of genes in the sex-chromosome heterochroamtin of Drosophila melanogaster. Genetics 70: 26I-27A. Sandler, L. I975. Studies on the genetic control of heterochromatin in Drosophila melanogaster. Israel J. Med. Sci. ll: ll2A-II3A. Sandler, L. I977. Evidence for a set of closely linked autosomal genes that interact with sex chromosome heterochromatin in Drosophila melanogaster. Genetics 86: 567-582. Sandler, L., D. L. Lindsley, B. Nicoletti and G. Trippa. I968. Mutants affecting meiosis in natural populations of Drosophila melanogaster. Genetics 60: 525-558- 105 Schalet. A. and G. Lefevre. I973. The localization of ”ordinary“ sex-linked genes in section 20 of the polytene X chromosome of DPOSOphila melanogaster. Chromosoma AA: l83-202. Shah, V. C., S. C. Lakotia and S. R. V. Rao. I973. Nature of heterochromatin. J. Sci. Indus. Res. 32: A67-A80. Smith, G. P. I973. Unequal crossover and evolution of multigene families. Cold Spr. Harb. Symp. Quant. Biol. 38: 507-5l8. Sonnenblick, B. P. I950. The early embryology of Drosophila melanogaster. Chapter 2. In: Biology of Drosophila. Edited by M. Demerec. John Wiley and Sons, Inc., New York. Spear, B. B. l97A. The genes for ribosomal RNA in diploid and polytene chromosomes of Drosophila melanogaster. Chromosoma A8: I59-I79. Spear, B. B. and J. G. Gall. I973. Independent control of ribosomal gene replication in polytene chromosomes of Drosophila melanogaster. PNAS USA 70: l359-l363- Szostak, J. w. and R. Wu. I980. Unequal crossing over in the ribosomal DNA of Saccharomyces cerevisiae. Nature 28A: A26-A30. Tartof, K. D. I97l. Increasing the multiplicity of ribosomal RNA genes in Drosophila melanogaster. Science l7l=‘29A-297- Tartof, K. D. I973a. Regulation of ribosomal RNA gene multiplicity in Drosophila melanogaster. Genetics 73: 57-71. Tartof, K. D. l973b. Unequal mitotic sister chromatid exchange and disproportionate replication as mechanisms regulating ribosomal RNA gene redundancy. Cold Spring Harb. Symp. Quant. Biol. 38: A9l-500. Tartof, K. D. I97A. Unequal mitotic sister chromatid exchange as the mechanism of ribosomal RNA gene magnification. PNAS USA 7I: I272-l276. Tartof, K. D. I975. Redundant genes. Ann. Rev. Genet. 9: 355-385. Tartof, K. D. and I. B. Dawid. I976. Similarities and differences in the structure of X and Y chromosome rRNA genes of Drosophila. Nature 263: 27-30. Wellauer, P. K. and I. B. Dawid. I977. The structural organization of ribosomal DNA in Drosophila melanogaster. Cell l0: I93-2l2. Wellauer, P. K. and I. B. Dawid. I978. Ribosomal DNA in Drosophila melanogaster II. Heteroduplex mapping of cloned and uncloned rDNA. J. Mol. Biol. l26: 769-782. 106 Wellauer, P. K., I. B. Dawid and K. D. Tartof. I978. X and Y chromosomal ribosomal DNA of Drosophila: comparison of spacers and insertions. Cell IA: 269-278. Williamson, J. H. I976. The genetics of the Y chromosome. Chapter l6. In: The Genetics and Biology of Drosophila, vol. 1b. Edited bY M. Ashburner and E. Novitski. Academic Press, London. White, R. L. and D. S. Hogness. I977. R-loop mapping of the IBS and 285 sequences in the long and short repeating units of Drosophila melanogaster rDNA. Cell l0: l77-I92. Yamamoto, M. and G. L. G. Miklos. I978. Genetic studies on heterochromatin in Drosophila melanogaster and their implications for the functions of satellite DNA. Chromosoma 66: 7I-98. Zalokar, M., l. Erk and P. Santamaria. I980. Distribution of ring-X chromosomes in the blastoderm of gynandromorphic D, melanogaster, Cell I9: I33-IAI. Zuchowski, C. I. and A. G. Harford. l976a. Ribosomal genes not integrated into chromosomal DNA in Drosophila melanogaster. Chromosoma 58: 2l9-23A. Zuchowski. C. I. and A. G. Harford. l976b. Unintegrated ribosomal genes in diploid and polytene tissues of Drosophila melanogaster. Chromosoma 58: 235-2A6. "lilllllll.lllllllllllI“