Download Analysis of Spot Formation in Flies: Somatic Crossing Over and Twin Spots and more Study notes Accounting in PDF only on Docsity!
SOMATIC CROSSING OVER AND SEGREGATION I N
DROSOPHILA MELANOGASTER*
CURT STERN University of Rochester, Rochester, N. Y Received May 6, 1936
TABLE O F CONTENTS Page
............................................................... 626 627 METHODS....................................................................... 628
TS............................... ,..............................
THEACTION OF hfINUTE FACTORS................................................. (^629)
Blond-Minute........ ,... ,.......... ,......... , , ,.. ,. ,....
Autosomal Minutes and sex-linked spots......................
The specificity of the effects of sex-linked and autosomal Minutes....................... 634
63 j Various hypotheses............................................. (^63 ) SomaticSegregation......................................................... _. 636 y/sn3 flies; preliminary discussion.............................................. 636 Minute-n and Blond-Minute “elimination” as somatic segregation......... (^)....... 638 Further analysis of somatic segregation in y/sn3 flies............. (^)............... 639 Somatic segregation and crossing over.............................................. 644 Experiments involving y sn3/ + flies....... ,......... (^).......................... 644 Experiments involving Blond..................... (^)......................... 648 Experiments involving y , sn3, and M n.. ,............................... 6 j I Experiments involving y , sn3 and “Theta”..... ,........ , ,.......... ,.......... 6j Experiments involving bobbed as a means of determining the rightmost crossover re- gion.................................................................
THEMECHAMSMOF MOSAICFORMATION...........................................
664 The number of X chromosomes i n cells of spots........................ 667 A comprehensive experiment involving y , sn3, M n , and Theta................
Somatic autosomal crossing over and segregation, ,.. ,.... ,........... The occurrence of autosomal crossing over i n females and males,. ,.....
AUTOSOMALMOSAICS., ,. ,.... ,... ,.......... ,........ ,.................
.... ,. ....... MOSAICAREASIN FLIESHETEROZYGOUSFOR X CHROMOSOMEINVERSIONS.......... Experiments involving y , sn3, bbDi; no Y chromosome present....................... Experiments involving y , ma,bbDi; a Y chromosome present... ,.... ,........ ,...... A n exceptional case of segregation i n experiments involving y , sn3, bbDi and an extra Y chromosome................................................................. Experiments involving y , sn3, bbDi, and Theta; no Y chromosome present............... Experiments involving y , sn3, bbDi, and Theta; a Y chromosome present................ Experiments involving y, sn3, M n , bbDi, and Theta................. ,................
Experiments involving the dl-49 Inversion.................. ,....... ,.... ,.........
THEINFLUENCEOF A Y CHROMOSOMEON MOSAICFORMATIONIN FEMALES NOT CARRYING A bbDf INVERSION................... ,........ , ,.... ,... ,.. ,.... ,.. , ,.. ,...... MOSAICSIN FLIESHETEROZYGOUSFOR A RING-SHAPEDX CHROMOSOME................
- A part of the cost of the accompanying tables and figures is paid by the Galton and Mendel Memorial Fund.
GENETICS21: 625 Nov. 1936
626 CURT STERN
SEX-LINKEDMOSAICAREASIN SUPERFEMALESAND IN MALES......................... 718 RELATIONSBETWEEN SOMATICCROSSINGOVER AND THE ONTOGENETICPATTERN.,........ 719 DISCUSSION..................................................................... 723 SUMMARY....................................................................... 726 LITERATURECITED.............................................................. 7 2 8
INTRODUCTION N 1925 BRIDGESfound that females of Drosophila melanogaster contain- I ing the dominant factor Minute-n in one X chromosome and some re-
cessive genes in the other X chromosome often exhibit a mosaic condition.
While the main surface area of these flies showed the effect of the dominant Minute-n (I, 62.7) without the effect of the recessive genes, as was to be expected, smaller areas, in different regions of the body and of varying size, were not Minute-n, phenotypically, but displayed the effects of the recessive genes. BRIDGES’interpretation was this : the Minute-n factor has
the property to eliminate occasionally the X chromosome in which it
itself is located. The cells of mosaic spots are descended from one common ancestral cell in which such elimination had taken place. They possess, therefore, only one X^ chromosome and show the phenotype produced by its genes. Minute-n is only one of a group of factors which are very similar in their phenotypical expression. The “Minutes” behave as dominants whose most striking phenotypic effect is a reduction in bristle size; in addition there is a strong retardation in development, tendency to rough eyes, etc. The homozygous Minute condition is lethal. Some Minutes have been shown to be deficiencies (SCHULTZ1929). Many Minute factors have been found in different loci of all chromosomes. They are distinguished by adding
different letters or numbers to the symbol M.
Following BRIDGES’discovery of mosaics with respect to sex-linked factors, the appearance of mosaic spots which exhibit autosomal characters was described (STERN1927b). Such spots appear on flies which originally had a constitution heterozygous for genes determining the characters. These mosaics occurred in crosses in which autosomal Minute factors were present and the facts seemed to agree with the interpretation that the spots were due to an elimination of that arm or part of an autosomal chromosome which carried the Minute. The present investigation was originally designed to attack the prob- lem: How is a Minute factor able to eliminate the chromosome or that part of a chromosome in which it itself is located? At the same time the solution of another problem was sought. The fact that small mosaic spots showed the phenotypic effect of certain genes con- tained in their cells whereas the remainder of the individual showed an- other phenotype was proof of the autonomous development of these char-
628 CURT STERN
For suggestions concerning the manuscript thanks are due to members of the Biological Laboratories in Pasadena and to my wife, EVELYNSTERN.
METHODS The methods were similar to those used by BRIDGES.Flies were made heterozygous for recessive genes whose phenotypic effects were of such a kind as to be exhibited by very small areas, preferably even single setae. The setae of Drosophila are divided into macrochaetae and microchaetae, the former generally called bristles, the latter, hairs. As far as the purpose of the present study is concerned the distinction is of no intrinsic im- portance. Genes mainly used were: (a) yellow body-color (y, I , o.o), pro- ducing an effect which can be distinguished in a single hair, making it yellowish-brownish as opposed to the black not-yellow condition (the general coloring effect of y on the hypodermis is often not very distinct in spots (STURTEVANT 1932)) and (b) singed-3 (sa3, I, 21.0), producing a thickened, curved or crooked condition of the setae, which generally can be distinguished in single hairs also. However, doubts occasionally’remain as to whether a single hair on a heterozygous +/sn3 fly is genotypically singed or whether it is normal but slightly more bent than usual. With spots of two or more hairs such doubts hardly ever occur. 1 Following BRIDGES,the flies were inspected originally for spots only on the head and thorax. I n later experiments, however, inspection of the ab- domen was included. I n order to discover even the smallest spots the flies were scrutinized under a binocular magnification of 37X (Bausch & Lomb objective 3.7, eyepieces roX).^ The use of^ a^ simple device made by the Bausch & Lomb Optical Company which allows for the fine adjustment to be made by foot movements and leaves both hands free for manipula- tion, proved to be of great value (for more detailed description see Droso- phila Information Service 6 :60). As the study of spots was practically confined to setae-bearing regions, a table of the number of setae on different parts of the body of average sized females was computed. Generally only the dorsal and lateral parts of the thorax and only the tergites of the abdomen were inspected and, except in special cases, no effort was made to remove the wings in order to uncover completely the median part of the abdominal tergites. This partly covered region excluded about 20 per cent of the abdominal setae from inspection (table I). A separate record was kept for each spot consisting of an outline drawing in case of head and thorax mosaics or of a notation of number of macro- and microchaetae and position in case of abdominal mosaics. Spots on the dorsal side of the thorax nearly always form one single clearly defined mosaic area, while spots on abdominal tergites are fre-
SOMATIC CROSSING OVER 629 TABLEI Number of setae on different body regions. ABDOMINAL TERGITE HEAD TEORAX** ___ I A N D 2 3 4 5 6 7
Mean num- berof setae 34 238 108 I 2 0 1‘7 I 23 I 2 0 61
Not inspected (estimate) % o 0 50 40 2 0 0 0 0
Inspected 34 238 54 72 94 1 23 1 2 0^61
% of total in- spected setae 4 30 66 ~ _ _ _ ~ _ _ - _ _ _ _ - _ _ Inspected ab- dominal setae (%of total) I O 13 17 24 24 1 2
Total setae inspected: ca. 800. Total abdominal setae inspected: ca. 525.
quently broken up into two or more separate parts. Apparently the growth processes in the imaginal discs of thorax and abdomen are somewhat different.
THE ACTION O F MINUTE FACTORS
Blond- Minute
As stated in the introduction, a number of Minute determiners have been found to be deficiencies for short regions. The hypothesis suggested itself that the apparent tendency of Minutes to eliminate the chromosomes or chromosomal parts in which they are located is due to some mechanical disturbance of chromosome division which itself is caused by the material defect in the deficiency chromosome. In order to test this hypothesis use was made of the “Blond-transloca- tion” (Bld) which represents a reciprocal translocation between the left
end of the X chromosome and the right end of the second chromosome
(BURKARTand STERN 1932). When all chromosomes of an individual are balanced with respect to the translocation, normal sized Blond bristles are produced. However, in females if one X chromosome lacks its extreme left end without being compensated for by the presence of the translocated piece on the second chromosome, then the individuals possess bristles of Minute, Blond character. BURKARThad found that these Minute females
SOMATIC CROSSING OVER 631 stable when its phenotypic effect was suppressed by the normal allele for that region translocated to the second chromosome?
Autosomal Minutes and sex-linked spots When the ability of autosomal Minutes to produce “autosomal” mosaics had been discovered, the following cross was made in order to detect a possible influence of such Minutes on the appearance of sex-linked mosaics : sn3/sa3 9 by + s n ; M y / + 8 ( M y , Minute-y, 111,40.4). While 1020 +MF females had 3 single-bristle singed spots, 811 M y F1 sisters exhibited 16 spots (13 single-bristle, 3 larger ones, table 2a, first row). One of the spots in a + M fly occurred on the abdomen and is not included in the table. This result showed (I) that mosaic spots appear as rare occurrences even in not-Minute flies (BRIDGESin MORGAN,STURTEVANT,BRIDGES1929, has encountered four such cases in his experiments) and (2) that an autosomal Minute increases the frequency of these occurrences, that is,
is able to influence the fate of an X chromosome.
Later work confirmed this finding and has led to the use of autosomal Minutes as tools in the study of sex-linked mosaics. Besides M y three other autosomal Minutes have been tested for their effect on the X chromosome behavior, namely Mw (Minute-w, 111, 8 0 _ f ) , M33j (Minute-ggj, 111, 40.4) and M/? (Minute-/?, 111, 85.4). Table 2a shows the results of some tests. It should be pointed out that the con- stitution of the X chromosomes varied in these experiments, so that the frequencies of spots in different experiments are not comparable. I n some experiments only the head and thorax were inspected for spots. In all later tests the abdomen was inspected also. Due to the comparatively small size of the head and the low number of setae the number of head spots has been added to the number of thorax spots under one grouping. An inspec- tion of tables 2a, b shows, among other results: (I) The frequency of sex-linked spots in not-Minute flies varied from 0.0 to 6.0 on the head-thorax region and from 4.6 to 20.0 per cent on the abdomen. The frequency in Minute flies varied from 0.0 to 22.3 in the head-thorax region and from 8.0 to 36.6 per cent on the abdomen. (2) A positive correlation is indicated between percentage frequency of head-thorax and abdominal spots: (a) In not-Minute flies experiment M w ( 5 ) MP M33j(3) M33i(2) Mw(4) M33jW frequency on head-
frequency on abdo-
thorax (^) 0.0 0. 7 (^) 0.s 1. 4 4. 5 6. 0
men (^) 15.0 4.6 (^) 8. 2 9. 2 11.7 2 0. 0
6 3 2 CURT STERN
(b) In Minute flies
irequency on head-
- 7 16.8 22. frequency on abdo- men 2 7. 1 8.0 28.1 33.4 36.6 15. Two of the discrepancies in these series seem to be based on a special con- dition which hindered the appearance of all head-thorax spots. This ex- periment ( M w ( 5 ) ) therefore is not comparable with the rest. ( 3 ) The relative increase of spots in Minute as compared to not-Minute flies varies from 1.0 to 15.6 times in the head-thorax and from 1.7 to 4. times in the abdominal region. (4) The average increase is distinctly lower in the abdominal region. ( 5 ) As far as the data are significant, no correlation seems to exist be-
experiment M 4 5 ) MP M d 4 ) M33j(3) M33J.b) M33jb)
thorax 0.0 3.1 4.
tween amount of increase of spots in the two different body regions :
increase on head-thorax I .o 2.8 4.7 7.2 15.
experiment M 4 4 ) M33i(I) MP M33A3) M33.W
increase on abdomen 2.4 I .8 1. 7 4.1 1. 7 The data of tables 2a,b had shown the effect of autosomal Minutes on the frequencies of occurrence of sex-linked spots. Is there also an influence
TABLE za Frequency of sex-linked spots i n Minute and not-Minute females. HEAD-TRORAX SPOTS ABDOMlNAL SPOTS ___ MINUTE USED INDIVIDUALS ____ AND NO. OF INSPECTED (^) ____NO. (^) ____% % (^) NO. (^) -% % EXP.
- M + M + M M : + + M + M M : +
M Y t IOZO 811 2 16. Z^ 2.0^ 10.5^ -^ - - -^ -
Mw(1)t 964 813 8 23 .8^ 2. 8^ 3.4^ -^ -^ -^ - ( 2 ) t 377 284 7 17 1.9^ 6.0 3.2^ - - - - (3)t 1 5 1 119^3^20 2.0^ 16.8^ 8. 5^ - - - - (4) 154 1r4 7 5 4.5 4.4^ 1.0^18 32 11.7^ 28.1^ 2. ( 5 )$ 432 247^0 0 0 0 -^65 67 15.0^ 2 7. 1^ 1. M33j(1) 135 131 8 2 2^ 6.0^ 16.8^ 2. 8^^27 48 20.0^ 36.6^ 1. ( 2 ) 349 251 5 56. 1.4^ 22.3^ 15.6^32 40 9.2^ 15.9^ 1. 7 (3)* 377 296 3 17_. 8_ 5.7 7.2 31 99 8. 2^ 33.4^ 4. M p (^307) 2 2 7 2 7 .7 3. 1 4.7 14 18 4.6^ 8.0^ 1.
~ _ _ _ - _ _ _ _ _ - ~ - - - t Abdomen not inspected. $ No head-thorax spots present.
on time of occurrence in development of the process which leads to appear- ance of a mosaic spot? The earlier this time is, the larger the spot should be. Accordingly, in table 3 the spots are divided into three groups, those
634 CURT STERN single seta spots is lower in the abdomen than in the head-thorax region. (4) I n most experiments no striking influence of the Minute on the time of occurrence of the spot-producing process can be found. The ontogenetic meaning of some of these findings will be dealt with later.
The specificity of the ejects of sex-linked and autosomal Minutes
I n view of the effect of autosomal Minutes on the X chromosome it
seemed desirable to test for a possible influence of an X^ chromosome
Minute on the behavior of autosomes. The Blond Minute is known t u be one of the most potent factors for the production of sex-linked spots and accordingly was chosen for the test in regard to autosomal spots. Females of the constitution h st cu sr es ca (located over most of the length of chromosome 111) were mated to Blond males and their Minute daughters were inspected for a condition mosaic for h, st, sr, es or ca. Although the 413 FI females exhibited on head and thorax 94 mosaic conditions for Blond, none was found to possess an autosomal spot. This shows that the influence of Blond-Minute on the production of autosomal mosaics must be very slight, if it exists a t all. This fact is significant. For if one makes a general “physiological Minute condition” responsible for the occurrences of spot-producing processes one might expect to find a corresponding seriation of different Minutes in respect to their potencies to produce both sex-linked and autosomal spots. However, a seriation in respect to frequency of sex-linked mosaics would show roughly: Blond-Minute > M n >M33j, M w , M y while the same Minutes ( M n excluded) with respect to frequency of autosomal mosaics probably would have to be arranged as: Ma, M y > M33j, Blond-Minute The two seriations are given after an analysis of the data presented in different tables of this paper. Too great reliance cannot be attributed to details of these arrangements, as the experiments were carried out over a period of years and under different genetic and environmental conditions. Therefore no special table has been made up from these data, as the quantitative results might indicate a higher degree of accuracy than they really represent. However, there seems no doubt as to the validity of the main result, namely that the effect of Minutes on the frequencies of sex- linked and of autosomal spots varies independently so that one Minute factor may affect strongly the number of sex-linked but only slightly the number of autosomal spots and vice versa.
SOMATIC CROSSING OVER 63 5
An even more striking correlation between certain autosomal Minutes and areas mosaic for definite regions of the same autosome will be pre- sented in the chapter on “Autosomal spots.”
T H E MECHANISM O F MOSAIC FORMATION Vurious hypotheses In the foregoing pages the effect of Minutes on the process of mosaic formation was discussed in general terms. The following part will contain an analysis of the process itself.
BRIDGES’work with females carrying M n in one X and recessive genes
in the other seemed to have established (I)^ that the cells of a spot^ do^ not
e t a t
a b C FIGUREI a<. Three possibilities to account for elimination of an X chromosome.
contain the M n chromosome and (2) that these cells are male in constitu-
tion, containing only one X with the recessive genes. Three main possi-
bilities suggested themselves as mechanisms for the elimination of the M n
chromosome from the cells of the spot: (a) during a somatic division the M n X chromosome does not divide and consequently passes into one daughter nucleus, leaving the other one in possession of only a division product of the not-Mn. X chromosome; (b) the M n X chromosome divides into two halves, but only one half passes into one of the daughter nuclei. The other half lags behind, is not included into a daughter nucleus, and degenerates in the cytoplasm; (c) the Mn X chromosome divides into two halves. These halves do not disjoin but pass together into one daughter nucleus. The constitution of the daughter nuclei according to the three hypothe- ses is pictured in figure I. No way of distinguishing between the mecha- nisms (a) and (b) was found, but a test between (a) and (b) on one side and (c) on the other seemed possible. The sister cell of the “elimination
SOMATIC CROSSING OVER 63 7
of spots in N 8 was 3.6 per cent, in controls 3.9 per cent. There was no
interaction of the Notch deficiency with the autosomal Minute-w. The nine spots in the N8/sn3 flies were recognized by singed setae oc- curring presumably as a consequence of elimination of the N s chromosome.
I n cases where the sn3 containing X chromosome would have been elimi-
nated, no visible spot would have been produced, for the N 8 chromosome
contained no recessive genetic marker which would have expressed itself in a spot (provided that a cell containing only the Notch-deficient X chromosome is able to reproduce sufficiently to give rise to a large enough
cell-patch).
TABLE 4 N 8 / y H w dl-49 b y sn3; M w / f. Head and thorax spots only.
-^ Na^ +N +" Mw +" Mw
y spots -^ - I I sn3 spots 4 5 0 2 sn3 '-1^ twin^ spots^ -^ -^2^9
Total spots 4 5 ' 3 I
The situation was different in the y H w dl-49/sn3 control flies. If there was no preference which of the two X chromosomes would be eliminated, then two different types of spots might be expected to occur in about equal
numbers: (I)^ y^ spots in case^ of^ elimination of the^ sn3^ X^ chromosome and
(2) m3 spots in case of elimination of the y H w dl-49 X chromosome. The
15 spots found consisted of ( I ) two y spots and ( 2 ) two sn3 spots while
(3) the remaining 11 spots showed an unexpected structure: they were twin-spots formed by a yellow not-singed area adjacent to a singed not-yellow area. The obvious explanation is that, during a somatic division of one of the cells of these 11 y H w dl-49/sn3 females, a segregation had taken place whereby one daughter cell obtained the yellow gene carried originally by
one of the X chromosomes while the other cell obtained the singed gene
carried originally by the other X. A process similar to gametic segregation
of genes lying in opposite members of a pair of chromosomes had occurred in a somatic cell. The further division and normal somatic differentiation of the two daughter cells finally gave rise to mosaic twin areas. These findings of somatic segregation suggested that the so-called chromosome elimination in certain cells leading to the appearance of mosaic spots was in all or most cases the consequence of somatic segrega-
638. CURT STERN
tion. This theory is substantiated by three facts. (I) Further experiments demonstrated the general occurrence of twin spots in flies of suitable con- stitutions. ( 2 ) In appropriate experiments it could be shown that nearly all mosaic spots exhibit the results of somatic crossing over. This makes simple elimination hypotheses improbable. (3) The theory solves the diffi- culties encountered by the assumption that the sex-linked Minutes elim- inate only their own chromosomes. We shall first discuss point (3). The more important statements (I) and ( 2 ) will be dealt with in later sections.
Minute-n and Blond-Minute “elimination” as somatic segregation
Somatic segregation of the X chromosomes in a female carrying M n in
one X^ and^ a^ recessive gene in the other will lead to two daughter cells, one
containing only M n , the other only the recessive. M n is known to be lethal
to a male or a homozygous M n female zygote. If we assume M n in such a condition to be lethal to a somatic cell also we shall expect the one daughter cell to die while the other one, containing only the recessive, will give rise
to the observed spot. (It might be argued that a cell containing only M n
is viable but phenotypically not different from the non-segregated sur- rounding tissue. This, however, is excluded by having a recessive gene to-
gether with M n in the one X chromosome but not in the other X. Segrega-
tion without lethal effect of the M n segregation product should exhibit M n spots which also show the recessive gene effect. In the experiments discussed on p. 635 the^ M n^ chromosome contained a white or cherry gene, but no spots showing the respective eye colors were found. Other experi- ments of similar nature are described in later sections of this paper.) Somatic segregation can account also for the single spots in “Bld- Minute” flies. In heterozygous Bld-Minute females it will lead to a “not Bld-Minute” and a “Bld-Minute” cell. The former will give rise to a spot with not “Bld-Minute” phenotype while the latter is^ expected to die, that is, if one assumes a cell not to be viable in case it contains a completely uncovered X chromosome deficiency. Male zygotes containing the un- covered Bld-deficiency X chromosome or female zygotes homozygous for such a condition are known to be not viable. The reader might be inclined to strengthen these arguments by pointing to DEMEREC’Sstudies (1934) on cell lethals. This, however, would not be justified as DEMEREC’Sinterpretation is based on the acceptance of the theory of somatic segregation and therefore cannot be used to prove this theory. The influence of a Minute condition on the occurrence of mosaic spots thus consists in an increase of the tendency to somatic segregation. The
640 CURT STERN Tables 5 and 6 contain data pertinent to this question. Four different but fundamentally similar groups of experiments are summarized. I n ex- periment I the females were of the constitution y H w dl-49/sn3 with or without Mw. They include, together with others, the flies discussed a t the beginning of the present chapter. It is seen (table 5 ) that 23 out of 38 TABLE 5 K i n d s and sizes of spots in experiments involving primarily y and sn when located in opposite chromosomes. y SPOTS Sl13 SPOTS 1j-sn3TWIN SPOTS EXP. CONSTITUTION INDIVID. SPOTS -- NO. OF SETAE NO. O F SETAE NO, OF SETAE I 2 > 2 I 2 > 2 2 > 2 (I) y Hw dl-49/sn3 551 38t 2 - 2 8 1 2 4 ' (2) y w (orw)/sd 376 Z I Z $ 40 9 1 2 62 1 2 IO 7 60 (3) y g2bb/sn3bbz (a) 635 61 - - 5 - - (b) 635 157: 2 2 8 6 60 I4 9 9 29
g2 bb1/sn3bbx (a) 214 1st I - - 5 4 8 I
(b) 214 73* 16 4 I 1 2 7 15 3 '
I -^ _^ _
(4) y bi cv ctg v
7 Head and thorax spots only. 1 Head, thorax and abdominal spots.
spots were twins and that IO out of the 15 single spots were so small as to
include only one seta, which obviously makes it^ impossible for them to show a twin condition. Of the 28 spots covering two or more setae the great majority, namely 23, were twins. Experiment I then corresponds closely to our expectation. I n experiment 2 the constitution of the flies was y w/sn3 or y we/sn3 (w and w e will be disregarded here). In addition, all individuals contained M33j. There were 67 twin spots out of a total of 212, and 102 out of the 145 not-twin spots were single seta spots. Of the I IO spots covering two or more setae 67 exhibited the twin condition, and 43 did not. But 21 of these 43 were so small as to include only two setae, thus still making it probable that the supposed twin area did not happen to cover a setae-forming region. Although the results in experiment 2 did not come as near to expectation as in I the agreement can be regarded as sufficient. The results will be further discussed after a description of ex- periments 3 and 4. I n these a high frequency of spotting was induced by making the flies homozygous for recessive, mutant alleles of bobbed. I n 3 one allele was the standard bb, the other one either the same or a very similar one; in 4 one was the lethal allele bb' the other was as in 3. Bobbed
SOMATIC CROSSING OVER 641
can be called a recessive Minute gene, so that the flies in these experiments were under the influence of a “physiological Minute condition,” which caused the high spotting frequency. Again the presence of other genes besides y and sn3 will be disregarded a t this point. I n 3 only 38 out of 163 spots were twins, but a consideration of the different sizes of spots again shows that 87 out of the^ 1 2 5^ single spots were^ so^ small as to include only one seta. Of the remaining 38 single spots, 2 2 were so small as to include only two setae, but of the 76 spots covering two or more setae, there were 38 twin spots. Experiment 3 then, although showing a general agreement with expectation seems to deviate more from it than the two experiments I and 2. Before discussing this we shall consider experiment 4. Here 19 out of 92 spots were twins, and 34 out of the 73 single spots included only
one seta. Of the remaining 39 single spots, 1 5 were so small as to include
only two setae. Out of the total of 58 spots covering two or more setae, 19 exhibited the twin condition. As in experiment 3 these results seem to be in general, but not very close, agreement with the theory of simple somatic segregation as the cause of spotting.
TABLE 6 Further data on the size of spots in experiments 1-4 of tuble 5.
(1) (2) (3) (4)
Total sn3 setae in single spots “3 I45 I 2 1 183
Total y setae in single spots 137 182 69 26
Total y setae in twins Total sn3^ setae in twins Average y setae in twins Average sn3 setae in twins
I 2 1 230 64 27 80 208 86 45 5.3 3.4 1. 7 1. 3.5 3.1 2.3 2. 4 Average (y+sn3) setae in twins 8.8 6.5 4.0 3.
Table 6 summarizes certain facts which help to explain the apparent discrepancies. The last horizontal line shows that the average size of the twin spots in experiment I was more than double that of the twin spots in 3 and 4 while it was intermediate in 2 , and that in 3 and 4 the average size of one of the twin areas was only of the order of magnitude of two setae. It follows that there was a considerably higher chance for the small twin areas in 3 and 4 to appear phenotypically only as single spots than there was for the large areas in I and 2. Part of the deviations from our expecta- tion are further cleared up by the following considerations. If single spots are really parts of twin areas in which only one area had the opportunity to exhibit its phenotype, then if chance alone determined which of the two areas covered a seta-forming region one should expect an equal frequency of sn3 and of y single spots. An inspection of table 5, however, shows that,
SOMATIC CROSSING OVER 643 from the second part of table 6. Here the frequencies of sn3 and y setae are listed for all twin spots, that is, for all cases in which there is no doubt as to the occurrence of somatic segregation. The relative frequencies of the two types of setae in twin spots vary in the same direction as the total frequencies in single spots. Obviously, in these experiments, the chances for survival and reproduction of the two daughter cells from a segregating division were not equal. The vitality of y cells was lower than that of sn cells in 3 and 4 but higher in I and 2. The greater the inequality in survival value, the higher the proportion of spots even of larger sizes which should appear only as single spots, the twin area having died or been kept small. This is roughly borne out by the data. What the causes of lowered via- bility were in these experiments cannot be determined accurately now.
The presence of different alleles of bobbed in the two X chromosomes of 4
and possibly 3 has nothing to do with it, as will be shown later. However,
it is suggestive that the cells in 4 which are pure for y are also pure for bi,
ct6, cv, 8, and g2. As it is known that this multiple mutant condition con- siderably lowers the viability of a whole individual, the assumption seems justified that a similar effect may be found also in mosaic parts. Table 7 has been added to give a more complete representation of all twin spots with more than 2 setae which occurred in these experiments.
It is interesting that such great inequalities of the two twin areas were
observed as the case in which y^ area covers^ 5 0 ;^ sn3 area covers^ I^ ;^ or^ y^ area has 24 setae and sn3 area 3 ; or y area I and sn3 area I 5 seta. Some of these very unequal twin spots might possibly be regarded as two single spots of independent origin, lying next to each other, but this must be very exceptional. The frequency of spots in most experiments is low enough as to make rare the occurrence of more than one spot on an individual, although the incidence of flies with two or more spots is higher than according to chance (BRIDGESin MORGAN,STURTEVANTand BRIDGES 1929; also numerous data of the author). If several spots occur on one fly they have no tendency to be neighbors. Summing up, it seems clear that somatic segregation does not only account for the occurrence of twin spots but also for that of single spots. But while the evidence in the cases of twin spots is direct, in the cases of single spots it is of such a nature as to leave open the possibility that not all single spots can be regarded as vestigial twin spots whose region did not happen to affect a seta. A numerical treatment of the data which theoretically should be able to give a final decision is not feasible on account of the many variable factors involved. The next section, however, will show that a certain proportion of single spots are to be expected which have never been partners of an original phenotypic twin group.
644 CURT STERN
Somatic segregation and crossing over Experiments involving y sn3/ + flies In the experiments summarized in table 5, y and sn3 were in opposite chromosomes. When both mutants were in the same chromosome, some new results were obtained (table 8): Three kinds of spots appeared with
TABLE 8 Spots ita Pies of the basic constittdioiz y s n 3 / +.
E X P. CONST. IND. SPOTS -^ 1/^ 8n3^ Y^^8113 OTHER 1 2 > 2 I 2 > 2 I 2 ,2^ SPOTS
(I) y sn3bb/car bb 83 34 IO 4 6 3 3 5 2 - I - (2) ysn3bb/car bb 495 7 2 25 15 IO^ 1 2^2^5^2 -^ I^ - ( 3 ) Y sn3bb/+i^508^35 10 11 4^5 4 - (4) YSn3/+t 3 2 1 2 1 8 3 4 I I 2 - - - **2 ***
I - - -
53 33 24 21 1 0 1 2 5 - 2
- ~ _ _ - _ Totals 1407 162 I10 43 7 2
t Partly Mm/+.
* +U+’”; $-colored.
different frequencies, namely I I O y sn3, 43 y and 7 sn3 spots. The finding of y sn3 spots was expected, for segregation in y sn3/+ females (disregard- ing the presence of bb and car) should give rise to y sn3 and + cells, which would be visible as y sn3 spots. The occurrence of y and of sn3 spots needs an additional interpretation. Somatic crossing over between y and sn would separate these two genes from each other and thus afford an ex- planation. If the crossover process occurred during a two strand stage, the resulting strands would be y+an and +usn3, and if segregation ensued a y and a sn3 twin spot would be produced. No twin spots were found, making the two strand crossing over assumption invalid. If, however, somatic crossing over occurred a t a four strand stage between two of the four strands, and segregation two strands by two strands followed, then the facts can be explained (fig. 2 ). It is seen that following single crossing over,
different types of chromatid segregation, namely x and y, give rise to
either y or sn3 single spots. (Throughout this paper the term “chromatid” is used in reference to the strands which constitute a multivalent chromo- some group during prophase and metaphase, as well as in reference to those chromosomes of anaphase and telophase which originated from a multivalent .) If crossing over occurs a t the four strand stage the subsequent segrega- tion process will be expected to lead to either one of two results. (I) A sep- aration of the four chromatids into two daughter cells will occur, followed
