MSU

LIBRARIES
”-

 

 

 

RETURNING MATERIALS:
5iace in booE drop to
remove this checkout from
your record. FINES wi11
be charged if book is
returned after the date
stamped be10w.

_____

 

 

 

 

 

 

CHROMOSOMAL SENSITIVITY TO MEIOTIC DRIVE IN DROSOPHILA MALES

By

Bruce David McKee

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Zoology and Genetics Program

1981

ABSTRACT

CHROMOSOMAL SENSITIVITY TO MEIOTIC DRIVE IN DROSOPHILA MALES

By

Bruce David McKee

In Drosophila melanogaster, males carrying sc43c8, an x chromosome

 

deficient for almost all of the basal heterochromatin, experience high
I§:X_nondisjunction and skewed sex chromosome segregation ratios. Z_and
‘gg classes are recovered poorly because of sperm dysfunction. In this
study, the nature of chromosomal sensitivity to this meiotic drive
system was addressed. The recovery of various normal and rearranged
chromosomes was assessed by crossing flies of appropriate genotypes.

Several conclusions emerge from these studies. 1) The recovery of
all chromosomes -- marked and unmarked X} , §_chromosomes including
euchromatic and heterochromatic deficiencies, and major autosomes -- i5"
disrupted by SC4SC8. 2) The recovery probability of a chromosome from
a sc4sc8 male is an inverse function of its length. 3) Autosomal and
sex chromosomal segregation ratios are independent of each other.
4) Drive levels are independent of the amount of sex chromatin in the
genome. 5) Heterochromatically duplicated g chromosomes do not induce
meiotic drive, implying that unpaired heterochromatin is not responsible
for the meiotic disruptions in scasc8 males. 6) Levels of drive and
nondisjunction in sc4sc8 males can be independently modified by Z_chro-
mosome or autosomal background.

These conclusions have several implications for understanding the
mechanism of sex chromosomal meiotic drive. The length dependence effect

could be explained by assuming that sc4sc8 disrupts production of a

Bruce David McKee
chromosome processing material, causing a shortage and leading to com-
petition among chromosomes. However, the lack of interaction between
sex chromosomes and autosomes and the failure of additional sex chro—
matin to enhance drive argue against the notion that chromosomes in
sc4sc8 males must compete for a scarce resource. An alternative ex-
planation is that mispairing of unequal sized homologs at meiosis I
causes a failure to inactivate the unpaired stretch of the larger

chromosome. This stretch is then an "armed bom "

which can destroy

any sperm which carry it. This hypothesis fails to account for auto-
somal sensitivity to scéscs-induced drive. It also predicts drive
induction by heterochromatically duplicated §fs, contrary to Observ-
ation. Furthermore, since large free pieces of g heterochromatin are
unable to restore X_recovery to normal, this hypothesis implies euchro-
matic participation in normal §:X_pairing, again contrary to observ—
ation. It is argued that meiotic drive is caused by separation of‘z

genes from a basal g controlling site, perhaps the same site implicated

in some cases of dominant male sterility.

I dedicate this dissertation to my parents, James and Alice McKee, and

to my wife Anne.

ACKNOWLEDGEMENTS

The author wishes to gratefully acknowledge the advice and guidance
of Dr. Leonard G. Robbins and the technical assistance of Ms. Nancy
Veenstra. Thanks are also due to my wife Anne for assisting with
editing and typing and to Ellen Swanson, Roger Denome, Dr. Thomas
Friedman, Dr. Peter Carlson, Dr. Loren Snyder, and Dr. Donald Beaver
for reading and commenting upon the manuscript. Special thanks are

due to Dr. Thomas Friedman for advice and encouragement.

iii

TABLE OF CONTENTS

INTRODUCTION . . . . . . . . . . . . . . . . . . .
Chromosomal Sensitivity to Meiotic Drive . . . . . . . .
Mbdifiers of Meiotic Drive . . . . . . . . . . . .
Mechanisms of Meiotic Drive . . . . . . . . . . . .
Tests of Models . . . . . . . . . . . . . . . .
Meiotic Drive and the Genetics of Spermiogenesis . . . . .
Meiotic Drive, Sterility, and §_Inactivation . . . . . .
Materials and Methods . . . . . . . . . . . . . .

CHAPTER 1 SENSITIVITY OF SEX CHROMOSOMES TO MEIOTIC DRIVE . .
RESULTS

X_Chromosome Sensitivity —- Crosses to Free-X Females . .

Crosses to Attached-X Females

Y Chromosome Sensitivity

sc4sc8/X[22_Males . . . .

Ir<

Chromosome Sensitivity

Euchromatin . . . . . .

Chromosome Sensitivity

Dc

Heterochromatin . . . . .

E Chromosome Sensitivity
DISCUSSION 0 O O O O O O O O O O O O 0 O O O 0

CHAPTER 2 GENETIC BACKGROUND EFFECTS ON LEVELS OF MEIOTIC DRIVE

RESULTS
Experimental Design . . . . . . . . . . . .
Z_Chromosome Modification in scasc8/3_Males . . . .

X_Chromosome Recovery in scascS/X/Qp Males

§_and X_Chromosome Recovery in scéscs/X/Dp_Males

iv

ll

16

18

19

22

23

27

29

31

33

34

38

41

44

44

47

50

Chromosome Recoveries from scasc8/X/Y_Males . . . . . 51
Modification by Autosomes . . . . . . . . . . . 51
DISCUSSION . . . . . . . . . . . . . . . . . . 57
CHAPTER 3 SENSITIVITY OF X_CHROMOSOME FRAGMENTS TO MEIOTIC DRIVE . 59
MATERIALS AND METHODS . . . . . . . . . . . . . . 62
RESULTS
Disjunction and Drive in sc43c8/YE/YE Males . . . . . 63
Relative Sensitivity of YE_and‘Y_ . . . . . . . . . 65

Relative Sensitivity of Yf_and YS-YS . . . . . . . 67

 

Drive Sensitivity of the Bobbed Locus . . . . . . . 69
Sensitivity of a Free g Duplication . . . . . . . . 69
§_Chromosome Recovery . . . . . . . . . . . . 72
sc45c8/BSY/YE/Dp(lgf)3 . . . . . . . . . . . . 73
DISCUSSION . . . . . . . . . . . . . . . . . . 77
CHAPTER 4 SENSITIVITY OF AUTOSOMES TO MEIOTIC DRIVE . . . . . 80
MATERIALS AND METHODS . . . . . . . . . . . . . . 81

RESULTS
Absence of Adjacent II Segregations . . . . . . . . 83

Sperm Recovery from T(2;3)bwv4 Males . . . . . . . 83

 

Autosomal Modification of Drive and Nondisjunction . . . 90
DISCUSSION . . . . . . . . . . . . . . . . . . 94
CHAPTER 5 .5 HETEROCHROMATIC DUPLICATIONS AND MEIOTIC DRIVE . . . 98
MATERIALS AND METHODS . . . . . . . . . . . . . . 100
RESULTS

4
Recovery of scSlsc . . . . . . . . . . . . . 103

8 4
Recovery of sc sc . . . . . . . . . . . . . 103

DISCUSSION . . . . . . . . . . . . . . . . . 107

CHAPTER 6 SUMMARY AND RECOMMENDATIONS
Summary . . . . . . . . . . . . . . . . . . . 111
Recommendations for Future Research . . . . . . . . . 115

BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . 121

vi

LIST OF TABLES

Table l Y_Chromosome Recovery from sc43c8[z Males
Table 2 Recovery of an Unmarked Y_from sc43c8 Males Using
Attached-X Females . . . . . . . . . . .
Table 3 Recovery of §_and‘Y_Chromosomes from sc43c8[Z/Dp(l;f)3
Males . . . . . . . . . . . . . . . .
Table 4 Recovery of Small Free g Duplications
Table 5 Comparison of Drive Sensitivities of YE_and 22(15f)3
Relative to BEY_
Table 6 Nondisjunction and Drive in sc43c8/Y Males
Table 7 Sperm Recovery from scasc8/Y/Dp(l;f)3 Males
Table 8 Chromosome Recovery from sc45c8/Y/EEY_Males . .
Table 9 Within Cross Correlations between §_and Y Recovery
Table 10 Autosomal Modification of Nondisjunction and Meiotic
Drive in ggfigngales . . . . . . . .
Table 11 Disjunction and Sperm Recovery from fi/YE/YE Males
L 3M S

Table 12 Comparison of Drive Sensitivities of Y z and B Y

in sc43c8 Males . . . . . . . . . .

Table 13 Comparison of Drive Sensitivities of Yf_and YS-YS
Relative to YLy3M . . . . . . . . . . .

Table 14 Comparison of Drive Sensitivities of YLbb+ and YLbb—

Relative to YE

vii

Table

Table

Table

Table

Table

Table

Table

Table

Table

15

16

17

18

19

20

21

22

23

Comparison of Drive Sensitivities of YE, YS°YS,

and Dp(1;f)3 Relative to BS . . . . . . . . . . 71

Comparison of Sibling scasca/Y/Y_and sc45c8/Y/Dp_Males . 74

Sperm Recovery from scasc8/BSY/Dp(l;f)3/YE_Males . . . 75

 

Results from Crosses between T(2;3)bwv4 Heterozygotes . . 87

 

Orthogonal Presentation of Results from

scasc8;T(2;3)waA Males . . . . . . . . . . . 89

 

Results from Control Crosses . . . . . . . . . . 92

Nondisjunction and Drive in sc4sc8 Males With and

Without T(2;3)bwv4 . . . . . . . . . . . . . 93

 

Recovery of scSISc4 . . . . . . . . . . . . 104
Recovery of sc8sc4 from Males Carrying Various

Scute Duplications . . . . . . . . . . . . . 105

viii

LIST OF FIGURES

Figure l Mating Scheme for Experiments 1 and 2 . . . . . . . 45

Figure 2 Pairing and Disjunction in T(2i3)bWVA Heterozygotes . . 82

Figure 3 Crosses to Generate sc4sc8;T(2;3)wa4 Males and Controls . 84

 

 

Figure 4A Diagram of Cross between Two Translocation Heterozygotes 85
Figure 4B Diagram of Control Cross . . . . . . . . . . . 85
Figure 5 Origin of Heterochromatically Duplicated and Deficienct

.K'S . . . . . . . . . . . . . . . . . . 101

ix

INTRODUCTION

In Drosophila melanogaster males, §_chromosomes deficient for the

4Lsc8R (scéscg), cause meiotic non—

 

basal heterochromatin, such as In(l)sc

 

disjunction of the §_and‘z_chromosomes and abnormal sex chromosome segre-
gation ratios (Gershenson 1933; Sandler and Braver 1954; Peacock and
Miklos 1973). Recovery of Y_sperm is depressed relative to §_sperm and
3: sperm are recovered very poorly relative to nullofig, nullon_sperm
(Sandler and Braver 1954). The absence in sc43c8 of most of the §_hetero—
chromatic "collochores" (discrete §:Y_pairing sites found at several loca-
tions in the heterochromatin of normal E and Y_chromosomes) may be respon—
sible for its failure to pair with the Y_in some spermatocytes (Cooper
1964). When pairing fails to occur, the §_and Y_move to the same pole,
thus generating g: and nullofig, nullon_meiotic products. There is no
evidence for meiotic loss; reciprocal secondary spermatocytes occur in
equal frequencies (Peacock 1965). Although it is difficult to rule out
post-meiotic chromosome loss, the recovery of the §.1n well over 50% of
the offspring in some crosses (Peacock, Miklos, and Goodchild 1975) argues
that chromosome loss can not be the whole story. The ratio of filto nullo-
.g, nullon_classes is substantially the same in secondary spermatocytes
and in the progeny (Peacock 1965; Peacock, Miklos, and Goodchild 1975),
implying that grbearing sperm are recovered as well as nullofg, nullon
sperm and that chromosomes are not lost post-meiotically (since that would

inflate the nullofg, nullofX class relative to the g class). Since the

egg hatch in matings involving scasc8 males is reasonably normal, zygotic
death can be ruled out (Peacock 1965). Electron microscopic examination
of testes from sc4sc8 males reveals aberrant spermatid development. The
most commonly observed abnormality is failure of individualization, with
syncytial spermatid breakdown and elimination in the waste sac (Peacock,
Miklos, and Goodchild 1975). The number of abnormal sperm per bundle
varies with the severity of drive and is reasonably close to the number
predicted from progeny counts. Thus, gamete dysfunction is evidently the
mechanism of sc4sc8-induced meiotic drive.

Similar developmental abnormalities are observed in spermatids from
males heterozygous for SQ, a naturally occurring second chromosome mutant
which causes sperm ratios heavily biased toward the §d_chromosome. In
some cases, over 99% of the progeny of an SQ heterozygous male receive the
‘SQ chromosome from their father (Sandler, Hiraizumi, and Sandler 1959;
Hartl and Hiraizumi 1976). As with scasc8 drive, sperm carrying the homo-
log of the driven element seem to self-destruct. In the testes of high
meiotic drive §d_males, half the developing sperm in each bundle of 64
appear defective (Peacock, Tokuyasu, and Hardy 1972; Tokuyasu, Peacock,
and Hardy 1977).

In SQ males, this self-destruct response can be traced to a specific
second chromosome locus called BEE (Responder). Normal chromosomes carry
a sensitive allele of R52. All naturally occurring §d_chromosomes carry
an insensitive variant (Hartl and Hiraizumi 1976). Current evidence favors
the idea that in §d_heterozygotes the SQ allele damages its homolog, ren-
dering it lethal to the developing sperm. An alternative view is that SQ
fails to do something essential for sperm development normally done to its

homolog by Sd+ (Hartl and Hiraizumi 1976). However, deficiency for SQ

3

mimics SQ:, and deficiency for BEE renders a chromosome insensitive to SQ
(Sandler and Carpenter 1972; Ganetzky 1977). Wild type chromosomes may
well carry no allele at the SS_locus. Although they do carry an BEE locus,
the function of that locus can not be required for sperm development after
anaphase since its deletion has no detectable effect except to render a
chromosome resistant to SQ,

No specific loci have been identified in the scasc8 case. All SS:
(bobbed minus, a deficiency for the £2Eé).§ deficiencies tested thus far
have induced both nondisjunction and drive (Lindsley and Grell 1968; Pea-
cock and Miklos 1973; Yamamoto and Miklos 1977). Thus, a bobbed locus
deficiency is a candidate for the role of drive inducer. However, an ex-
tra bobbed locus can be added to the sc43c8 genotype in the form of a
heterochromatic free S duplication (an S chromosome missing almost all the
euchromatin) without any apparent improvement in the recovery of the S
(Haemer 1978). Furthermore, the severity of both drive and nondisjunction
seems to vary with length of deficiency even though all the drive inducers
tested thus far are completely bobbed lethal. The implication is that
drive is due to deficiency for all or part of a large locus or a group of
loci in the centric heterochromatin near the bobbed gene. Since only a
few S heterochromatic deficiencies have been tested, nothing more definite

can be concluded.

Chromosomal Sensitivity £Q_Meiotic Drive

 

 

Very little is known at present about the nature of chromosomal sen-
sitivity to meiotic drive in scasc8 males. In fact, the literature is am-
biguous as to whether the sensitivity of X_chromosomes is a property of
the X_or of the translocated S genes used to mark X_chromosomes. Gershen-

son's original study (1933) using an unmarked X_chromosome reported poor

recovery of S: sperm but not of X_sperm. All subsequent studies have
demonstrated poor recovery of S_sperm as well, but all have used marked
st exclusively. Chapter 1 reports a series of experiments which resolves
this ambiguity by demonstrating poor z_recovery from sc4sc8 males carry-
ing unmarked S_chromosomes.

Uncertainty also exists as to the unit of sensitivity of the X_chro-
mosome. The sensitivity of the Z_might be determined by a single response
gene, as with BEE; or by a group of response genes. Or, conceivably,
sensitivity to drive could be a function of all Z_chromatin. Chapter 3
reports a series of experiments aimed at characterizing the "unit of sen-
sitivity" of the X chromosome. Centric fragments carrying various seg-
ments of the X_were tested for relative sensitivity in the presence of
scascs. The results are completely consistent with the view that all X
chromatin is drive-sensitive since the degree of sensitivity of a frag-
ment is a function of its length. These experiments eliminate the pos-
sibility of a single locus responder but do not rule out polygenic sen-
sitivity, since a relatively limited number of fragments are available
for testing.

Uncertainties also exist concerning §_chromosome sensitivity. Pea-
cock (1965) and Peacock and Miklos (1973) found similar ratios of §_to
nullojS, nullofiz_classes in secondary spermatocytes and in progeny of
sc48c8 males, implying that scasc8 does not distort its own recovery.
However, in high drive males, one can find sperm bundles in which more
than half of the sperm are defective (Peacock, Miklos, and Goodchild
1975) meaning that some S and/or nullojS, nullofiz_sperm are malfunction-
ing. When a third sex chromosome, either a second Z_(Sandler and Braver

1954) or a free §_duplication (Haemer 1978) is added to the genotype of

a sc4sc8 male, the result is suicide drive of the sc4sc8 S. It is recov-
ered in well under half, often less than a quarter, of the progeny. In
sc4sc8/Zfz_males, the two st pair and disjoin from one another leaving
the univalent S_to move randomly to one pole or the other. Despite its
unpaired condition, the §_is not lost during the meiotic divisions and
presumably is included in the nuclei of half the developing spermatids
(Cooper 1964). An alternative possibility that has not been rigorously
ruled out is post-meiotic loss of the S such as has been found in some
attachedf§Z_males where the attachedeszegregates into a micronucleus
which later degenerates (Hardy 1975). Micronuclei have not been reported
in sc43c8 males but since their occurrence was unknown at the time of the
light microscope studies of schsc8, they might have been overlooked. No
cytological examination of sc45c8/ZjSp males has been performed, but the
genetic data are similar to those from sc45c8/2/2_males: regular disjunc-
tion of the heterochromatic elements and random assortment and poor re-
covery of scasc8.

Genetic evidence for meiotic drive, rather than chromosome loss, as
the explanation for poor S recovery in sc4sc8/SfSp_males is presented in
Chapters 1 and 2. With different X_chromosomes, different levels of re-
covery of both S_and Z_chromosomes are observed. The correlation between
.S recovery and X_recovery is very strong in these males. This correlation
implies that the same forces and, presumably, the same mechanism are op-
erative in both cases. Since poor S_recovery must be caused by meiotic
drive and not by chromosome loss (as loss of the Z would produce nullojg,
nullofSp sperm which are not observed), poor S recovery is probably also
a case of meiotic drive. The definitive cytology still needs to be done

on these males, but until then this evidence argues strongly against

chromosome loss.

Further evidence for drive acting against the S as well as the X
has come from experiments with another sex chromosomal meiotic drive sys—
tem, T(l;4)BS. This is a reciprocal translocation between the S and the
tiny fourth chromosome with an.§ break point in the proximal euchromatin
(Lindsley and Grell 1968). The bulk of the S euchromatin is attached to
the centromere of a fourth chromosome, and the remaining §_base is capped
with a fourth chromosome tip. Disjunction is regular in T(l;4)BS males,
with the §_base disjoining from the X_and the fourth chromosome centro-
meres disjoining from each other. Reciprocal meiotic products are not
recovered equally. The longer element of each bivalent (the X_and gig?)
is recovered poorly. Recovery probabilities from the two bivalents are
independent of each other (Novitski and Sandler 1957). Evidently, via-
bility interactions, so common in zygotic lethals, do not occur in this
gametic lethal system.

In the drive systems just described, reduced recovery of the SLas
well as of the I is observed. In all these cases, drive is acting
against the euchromatic portion of the S, the portion in which almost
all of the genes are located. sc4sc8 is deficient for at least 90% of
the heterochromatin and is, therefore, an almost completely euchromatic
chromosome. The QESE element of T(l;4)BS consists of most of the tiny
fourth plus about four-fifths of the S euchromatin. No evidence exists
as to whether or not drive can act against the heterochromatic portion
of the S. When a free S duplication containing most of the heterochroma—
tin is added to a sc43c8 genotype, it disjoins from the X, Under these
circumstances, one can ask only whether Sp sperm are recovered better or

worse than S sperm. In fact, they are recovered far better, with two to

three 22 sperm recovered for every X_sperm (Haemer 1978). This does not
mean that drive acts against the X_only. It may be the case that drive
affects both X_and Sp but disrupts recovery of the larger S_more than
the smaller Sp.

Experiments described in Chapter 1 attempt to settle this point.
In one experiment, recoveries of two small free S duplications are mea-
sured in scésca/Sfyp_males. The duplications are small enough so that
they do not disjoin (except occasionally) from the §_or X, This makes it
possible to measure the relative recovery of otherwise identical sperm
classes with and without the duplication. The results are consistent
with but do not prove mild disruption of recovery of these rather small
chromosomes. The second approach is to compare the recoveries of two
reasonably large but different sized free duplications, one of S origin

and the other of X_origin, against a common standard in sc4sc8/BS

Y/Be

 

males. The result is that the EEZEQE recovery ratio is proportional to
the length of the duplication,implying that the absolute recovery of_Sp
sperm declines with increasing size of duplication. However, since one
can measure relative but not absolute recoveries, this interpretation is
not certain. An alternative is that drive acts only against the larger
of two paired homologs and only to the degree that the homologs differ
in size.

No information is available in the literature concerning autosomal
sensitivity to sc45c8-induced drive. A test for autosomal sensitivity
is described in Chapter 4. scl‘sc8 males heterozygous for a reciprocal
but asymmetric translocation between the second and third chromosomes are
crossed to females heterozygous for the same translocation. The break-

points of the translocation are such that sperm carrying three or five

major autosomal arms are generated in addition to sperm carrying the usu-
al four arms. Since the females produce complementary egg classes, the
experiment permits recovery of these unusual sperm types and determina-
tion of the numbers of each type in the presence and in the absence of
sc4sc8.

If scl’sc8 interferes with some aspect of chromosome processing which
involves the whole genome, then one expects the sperm with five autosomal
arms to have reduced recovery relative to sperm with three or four auto—
some arms. The results described in Chapter 4 demonstrate that autosomes
as well as sex chromosomes are sensitive to scascB-induced drive. The
recovery of autosomally duplicated sperm relative to euploid sperm is
considerably worse in scasc8[: than in control S/X_males. Furthermore,
recovery of the autosomal deficiency class is improved relative to either
of the euploid classes by the presence of sc43c8. These relationships
hold true for all sex-chromosomal sperm classes. The fact that the re-
covery ratio of deficiency sperm to euploid sperm is greater in the pres-
ence of sc4sc8 implies that both S and X_sperm with euploid autosomal
constitution suffer sperm dysfunction. Thus it now seems likely that S
sperm produced by sc45c8 males malfunction even if drive levels are not
extraordinarily high.

These data also provide evidence against sperm viability interactions
between sex chromosomes and autosomes in scésc8 males. The relative
frequencies of autosomally duplicated, euploid, and deficient sperm are
approximately the same in all sex chromosome classes. This result argues
against the notion, discussed in more detail later, that sex chromosomes
compete for a scarce material in scAsc8 males. If competition were oc-

curring, we would expect to find lower ratios of duplication to euploid

and euploid to deficiency classes among SS sperm than among X_or S sperm.
This is not the case; strict independence of sex chromosomes and auto-
somes seems to be the rule.

The conclusion of all these studies is that the probability of re-
covery of a sperm from a sc4sc8 male is an inverse function of its chro-
matin content. This may mean that sc45c8 disrupts chromatin processing
so that chromosomes become partial gametic lethals. Or, it may mean
that sperm physiology is altered in some other way that renders sperm
sensitive to the amount of perfectly normal chromatin. Since it is not
known at present whether or not chromatin is altered in any way by scasc8

no decision between these alternatives can be made.

Modifiers Q§_Meiotic Drive

 

It is often possible to gain important insights into the mechanism
of a phenomenon by studying its modifiers. Sex chromosome meiotic drive
systems are subject to modification by temperature at meiosis (Zimmering
1963; Zimmering and Perlman 1962; Zimmering and Green 1965), sOurce of Y
chromosome (Zimmering 1959; Zimmering 1960; Peacock and Miklos 1973),
number of X_chromosomes (Sandler and Braver 1954), source of autosomes
(Zimmering 1959; Zimmering 1960; Peacock and Miklos 1973), and amount of
heterochromatin (Haemer 1978). I

Lowered termperature tends to reduce meiotic abnormalities. When
scasc8 males are raised at 18 or 19 degrees instead of the usual 25 de-
grees, both nondisjunction and drive are greatly reduced (Zimmering 1963).
The reduction of nondisjunction is due partly to an increase in the prob-
ability of S:X_pairing and partly to random assortment of unpaired chro-
mosomes at 18 degrees (Peacock 1965; Peacock, Miklos, and Goodchild 1975).

Lowered temperature also moderates drive in T(l;4)BS males (Zimmering

10

1963; Zimmering and Perlman 1962) and reduces drive against the_§ in
scasc8/S/X_males (Zimmering and Green 1965).

Genetic background has also proved to be important in determining
the level of meiotic drive. The amount of autosomal heterochromatin is
a strong determinant of drive levels in scASc8 males. Deficiency for
second chromosome heterochromatin suppresses nondisjunction and drive
(Haemer 1978), but autosomal inversions have no substantial impact on
drive levels (Ramel 1968). SQ has been found to interact strongly with
sc43c8. In SQ: progeny of sc45c8/X; SQ/SQ: males, recovery of X_and S:
sperm is unusually high, almost equalling that ofig and nullofig, nullon
sperm, while in the SQ progeny the usual distorted ratios are seen (Mik-
los, Yanders, and Peacock 1972). It is as if a small fraction of gametes
are immune to both drive systems. The degree of drive in T(l;4)BS males
depends on the source of the X_chromosome and autosomes. Segregation
ratios are nearly normal when the autosomes and the X_are derived from a
"low drive" stock but quite abnormal when they come from a "high drive"
stock (Zimmering 1959; Zimmering 1960). Levels of drive and nondisjunc—
tion show a similar dependence on both X_and autosomal source in scl‘sc8
males. When individual males in one cross are ranked by frequency of
nondisjunction, it is found that high nondisjunction males are also char—
acterized by high meiotic drive and low fertility and that males low in
nondisjunction are low in drive and high in fertility. These correlations
imply that inter-male variation reflects segregation of modifiers, pre-
sumably autosomal (Miklos, Yanders, and Peacock 1972; Peacock and Miklos
1973). The experiments described in Chapter 2 test the effects of low
and high drive Eds and low and high drive autosomes on nondisjunction

and drive in scasc8/X, sc4sc8/Z[Sp, and scéscs/X/X_males. The result is

11

that both !_chromosomes and autosomes independently modify nondisjunction
frequency and meiotic drive level in sc4sc8 males. It is also found that
modifiers have parallel effects on Z_chromosome recovery in sc43c8/X_and
scésc8[X[Qp siblings, confirming the supposition that these two types of
males suffer from the same defect. It is also shown that there is no
necessary connection between sex chromosome nondisjunction and meiotic
drive. Despite completely regular disjunction of the Z_and QB in scascs/
Z[Qp_males, they are subject to the same modification of drive levels as
are sc48c8/Z_males in whom nondisjunction is frequent. Another result

is that S and X_chromosome recoveries covary in sc45c8[ [22 males. A
very strong correlation between recoveries of these two chromosomes was
found across a wide range of recovery coefficients. This implies that

‘S and X_chromosomes are recovered poorly for the same reason -- gamete

dysfunction.

Mechanisms Q§_Meiotic Drive

 

 

The parallel effects of both temperature and autosomal background

on disjunction and drive in scl‘sc8 males lend credence to the idea that
nondisjunction and drive are caused by the same fundamental problem --
mispairing 9f.§ and X. This hypothesis holds that unpaired and weakly
paired chromosomes tend to be improperly processed for spermiogenesis.
The consequence is developmental failure of the spermatids that carry
them. The absence of most of the collochores from sc4sc8 is supposed to
be responsible for its tendency to pair only weakly with the X, This
weak pairing leads to both nondisjunction and meiotic drive. Lowered
temperature strengthens pairing forces, thus decreasing both nondisjunc-
tion and drive. Support for this theory comes from observations on the

meiotic behavior of male-specific meiotic mutants. All twenty EMS-induced

12

mutants isolated in a mutant screen for lines with high S§X_nondisjunc—
tion also caused reduced X chromosome recovery (Baker and Carpenter 1972).
Two of these mutants were mapped roughly to the proximal euchromatin.

All the mutants proved unstable and were lost within a few months. Addi-
tional evidence for this interpretation comes from meiotic analysis of
scasc8[xfgp_and sc4sc8/Z/X_males. In these males the S essentially never
pairs. Although it is included in 50% of the spermatid nuclei, its re—
covery is poor (Sandler and Braver 1954; Cooper 1964; Haemer 1978).

There are a number of objections to this pairing model. Meiotic
pairing and disjunction is evidently inadequate to insure normal chromo-
some recovery. All X_sperm in sc4sc8/S males derive from spermatocytes
in which the S and Z_paired and disjoined from each other. Yet many of
these sperm fail to function. It may be that this pairing is weaker than
normal and, while adequate to insure disjunction, does not provide a
tight enough bond to insure normal chromosome processing. If so, why
doesn't the S experience recovery difficulties? A similar difficulty
arises in the analysis of meiosis in scasc8/Z[Sp_males. Here the hetero—
chromatic duplication, with its full complement of pairing sites, pairs
and disjoins regularly from the 3, However, the recovery of the z_is
poor (Haemer 1978). Similarly, disjunction is regular in T(1;4)BS
males, but segregation ratios are aberrant (Novitski and Sandler 1957).

Another version of the pairing model is based on Baker and Carpen-
ter's (1972) suggestion that chromosomes enter meiosis carrying "armed
bombs" destined to explode if not defused prior to the spermatid stage.
Normally, the §_and X defuse each other's bombs during meiotic pairing.
This mutual facilitation depends on size matching of the two chromosomes

so that each armed bomb on one chromosome can line up opposite a defusing

13

site on the other chromosome. The heterochromatic deficiency in scasc8
renders it less likely to defuse all of the Z_chromosome's bombs. The Y,
however, is long enough to defuse all of the S's bombs. This explains
why the I, but not the S, experiences recovery difficulties in scésc8[S
males. In schsc8/X71_males, the two Y's are the same size and so can
defuse each other's bombs. The absolute recovery of X_sperm should be
very good in this genotype, if it could be measured. The armed bombs of
unpaired chromosomes, such as scasc8 in scasc8/X/X_and sc4sc8/XjSp males,
can not be properly defused. Thus the recovery of unpaired chromosomes
is poor despite the absence of meiotic loss. Peacock and Miklos' pairing-
dysfunction model (1973) is similar in most respects to this armed bomb
model.

Although the armed bomb model is phrased in non-biological terms,
it is not difficult to imagine biological candidates for the armed bomb
role. For example, an armed bomb might be an actively transcribed gene.
Given the evidence (reviewed below) for transcriptional shut-off during
spermiogenesis, it is possible that persistence of gene activity past
the normal shut-off point is detrimental to sperm development. Normally,
.§ and X_chromosomes would help each other shut off transcription during
meiotic pairing but sc4sc8 disrupts this interaction. An alternative
armed bomb candidate would be a stretch of insufficiently condensed chro—
matin. Normal sperm heads contain highly condensed chromatin. This con-
formation is facilitated by the transition from lysine-rich to arginine-
rich histones during spermiogenesis (Das, Kaufmann, and Gay 1964). Per-
haps scasc8 interferes with the histone transition (as suggested by
Kettaneh and Hartl (1976) for SQ) by disrupting a necessary precondition

related to pairing. There is no evidence for either of these suggestions.

14

They are made merely to point out concrete biological models consistent
with the more abstract theory under discussion.

Although the armed bomb model explains much of the genetic data, it
suffers from one major drawback -- no one has ever observed pairing be—
tween the §_and X_at other than the restricted heterochromatic sites
known as collochores (Cooper 1964). The failure of a free S duplication
carrying a full complement of collochores to restore normal 1 recovery
implies that at least some of the pairing sites must be euchromatin. The
side by side, point for point pairing required for the S and 3 under the
armed bomb model has been observed with other chromosomes but never for
the S and X,

The armed bomb model and the pairing-dysfunction model share an em-
phasis on pairing as fundamental to proper chromosome processing. Other
models can be devised which explain most of the data, yet place less em—
phasis on pairing. For example, it is possible that the heterochromatic
deficiency disrupts production or distribution of some essential chromo-
some processing material (e.g., a sperm histone). The ensuing shortage
of the material would cause chromosomes to compete for adequate supplies
of it. Any chromosome unable to garner enough of it would become a gam-
etic lethal. Suppose that every binding site for the material has the
same probability of binding it and that each binding site on a chromosome
must be occupied for that chromosome to be non-lethal. Under shortage
conditions, the longer a chromosome, the less likely it is to have all
its binding sites occupied. The sensitivity of a chromosome to sc45c8-
induced drive would be proportional to its length. This model would
agree with the armed bomb model in predicting length dependence of chro-

mosome sensitivities, but for different reasons. Under the latter model

15

the longer of two homologs suffers recovery disruption (to the degree
that they differ in length) because the bombs on the overhanging piece
are not defused. Under the competition model, both homologs are affected
but the larger one is more sensitive than the shorter one. Pairing is
fundamental to the armed bomb model, because without it bombs can not

be defused. The role of pairing in the competitive model is not entirely
clear but certainly not fundamental. Perhaps chromosomes with less than
the normal amount of the scarce material also experience pairing diffi—
culties. Or perhaps the apparent connection between nondisjunction and
drive is a coincidence owing to the proximity of pairing sites and the
drive inducing site.

Another plausible explanation is that the heterochromatic deficiency
causes a shortage of time rather than material. Perhaps the length of
time available for a key meiotic process is dependent somehow on the
length of the S chromosome. For example, suppose that the signal to
finish one step in chromosome processing and move on to the next step
comes when the §_has finished the step. Suppose further that an unusually
short S finishes the step unusually early, before other, longer chromo-
somes have been able to finish. The severity of the consequences of this
failure would depend upon the length of material unprocessed and thus
upon the overall length of the affected chromosome. This time shortage
does not lead to competition between chromosomes since time can not be
sequestered. Thus the time shortage model differs from the material
shortage model in its consequences for chromosome interactions.

Several aspects of these models are amenable to experimental

testing.

16

‘IgggglgfigModels

One test of the armed bomb model is based on the relationship be-
tween size discrepancy of S and X chromosomes and drive against the
longer element. Size mismatch results in an unpaired region on the
longer chromosome which, according to the armed bomb model, should kill
developing sperm. In all the drive examples examined thus far, the X
has been the longer element of the Sex bivalent. What happens if the S
is longer than the 32 Tests for meiotic drive in males carrying hetero-
chromatically duplicated §_chromosomes are described in Chapter 5. These
unusually long S chromosomes have one dose of heterochromatin near the
centromere and one dose near the tip. Despite the considerable S:Z_size
discrepancy, no drive was observed. It seems, then, that the meiotic
disruptions associated with scl‘sc8 are due not to size mismatching per se
but either to the absence of some function normally located in the S het-
erochromatin or to its separation from the rest of the chromosome (as
in scascalzfgp males).

A second.test of the armed bomb model and of the competitive model
involves drive against the unpaired scasc8‘§_in scasCB/X/X_and scésca/Z/
'Sp_males. Accoring to the armed bomb model, poor recovery of this chro-
mosome is due to its unpaired, and therefore undefused, condition. There
is no reason to think that the level of interaction between the two het—
erochromatic chromosomes would affect the recovery of the S. There is
no place in the model for chromosomal interactions other than direct
pairing ones. The recovery of sc45c8 should be the same in scasca/sz
and scésc8/27 2 males when background genotype and environmental condi-
tions are the same.

Under the competitive model, recovery of the S chromosome should be

l7

inversely related to the amount of sex chromatin in the genome; the
more sex chromatin, the fiercer the competition and the smaller the aver-
age ration of the scarce material. Since a X_chromosome is considerably
larger than a free S duplication, drive levels should be higher (and S
chromosome recovery lower) in sc43c8[zfz_males than in scésc8[zfyp_males.

Some of the results in Chapter 3 are consistent with the prediction
of the competitive model. However, none of the results in which auto-
somal background is fully controlled are consistent with it. The best
evidence is that S_chromosome recovery is independent of amount of sex
chromatin in the genome, implying absence of competition.

Another test for competition involves adding a fourth sex chromo-
some to a scasc8 genome and monitoring its effect on §_recovery. Most
sc45c8 males with four sex chromosomes are sterile, but scasca/XE/SEZfSp

males do produce some offspring. .§ recovery in these males is no worse

than in scasc8/BSY[SE or sc43c8/BSY/XE siblings. Thus, increasing the

 

 

amount of sex chromatin in the genome does not exacerbate drive levels.

A third test for competition was described previously. To briefly
recapitulate, sc4sc8 males carrying a reciprocal asymmetric translocation
between the second and third chromosomes exhibit drive sensitivity of
autosomes as well as sex chromosomes. If chromosomes compete for a
scarce resource, one would expect the amount of autosomal chromatin to
influence the frequencies of the various sex chromosome classes and vice
versa. However, this is not the case. Autosomal and sex chromosome
recoveries are independent.

The absence of competition implies that the meiotic disruptions
induced by scl‘sc8 are not due to shortage of a chromosome processing ma-

terial. This result is consistent with but does not prove the armed bomb

18

model. It is also consistent with other shortage models which do not
imply competition (e.g., time shortage). However, no critical tests of

such models have been devised.

Meiotic Drive and the Genetics Q§_Spermi9genesis

 

 

The chief difficulty in explaining male meiotic drive systems in

Drosphila melanogaster is the evidence that post-meiotic germ cell devel-

 

opment is not dependent upon expression of any genetic functions carried
by those cells. If any such function existed, it would be possible to
find mutants that can not be transmitted by heterozygous males (gametic
lethals). None have been found in Drosophila (Lindsley and Lifschytz
1972). There is also no post-meiotic chromosomal requirement. Clearly
the S and X_are dispensible since each is absent from half the sperm.
That autosomes are also dispensible is proved by the recovery of sperm
deficient for one autosomal arm and duplicated for another as segregants
from whole arm reciprocal translocations (Muller and Settles 1927).
Males carrying compound second and third chromosomes routinely produce
nullofig and nullofS_sperm which function perfectly well in fertilization.
In fact, sperm carrying only the tiny fourth chromosome have been recov-
ered in progeny of such males (Lindsley and Grell 1969). Biochemical
studies (Olivieri and Olivieri 1965; Gould-Somero and Holland 1974) indi-
cate that transcription does not occur in spermatids or sperm of Droso-
phila, although some transcription clearly does occur in mammalian sper-
matids (Monesi 1965; MOore 1971).

The fact that no genetic functions are expressed post-meiotically
does not mean that post—meiotic development is free of genetic control.

Most male-sterile mutations in Drosophila interrupt development after

l9

meiosis (Linsley and Lifschytz 1972). Translational activity continues
unabated after meiosis despite the transcriptional shut-off (Brink
1968; Gould-Somero and Holland 1974). Apparently, messages synthe—
sized in the spermatocyte are stored for later use in spermatids.

Given the evidence for absence of post-meiotic gene expression,
explanations of meiotic drive can not invoke damage to essential
spermatid functions. Instead one is tempted by the opposite conclusion,
that genetic repression is essential for sperm development and that
meiotic drive is caused by inappropriate post-meiotic gene expression.
No direct evidence for this idea exists. A related explanation is that
meiotic drive is caused by a failure of spermatid chromosomes to achieve
adequate condensation. Kettaneh and Hartl (1976) have shown that
spermatids in SQ homozygotes do not undergo the histone transition from
lysine-rich to arginine-rich histones necessary for proper chromatin
condensation. Electron micrographs reveal inadequate chromatin con-
densation in many sperm of SQ males (Peacock, Tokuyasu and Hardy 1972;
Tokuyasu, Peacock and Hardy 1977; Kettaneh and Hartl 1980). Sex
chromosome meiotic drive also involves sperm dysfunction although it
is not clear whether or not chromatin condensation is normal (Peacock,

Miklos and Goodchild 1975).

Meiotic Drive, Sterility, aQQ §_Inactivation

 

 

In both SQ (Hartl, Hiraizumi and Grow 1967) and scasc8 (Peacock
and Miklos 1973) an inverse correlation between severity of drive
and degree of male fertility has been observed. Perhaps severe meiotic
drive could cause complete sterility. There are indications that cer-

tain types of chromosomally-based male sterility are related to meiotic

20

drive. In Drosophila and otherorganisms, many Sfautosome translocations
are male sterile. Addition of a duplication covering the region of
the §_breakpoint does not restore fertility to sterile T(X;A)'s, in-
dicating that the sterility is dominant. Many Sf: translocations are
also sterile, but an extra Z_usually rescues them, suggesting that the
X_breakpoint has interrupted a fertility factor. Almost all autosome-
autosome translocations are fertile in both sexes. The distribution
of breakpoints between the fertile and sterile Sfautosome transloca-
tions is informative. Fertile Sfautosome translocations either have both
breakpoints near the tips or have an‘S breakpoint in the proximal hetero-
chromatin and an autosomal break anywhere. In both cases the bulk of the
'§_euchromatin remains intact (Lifschytz and Lindsley 1972).

Lifschytz and Lindsley (1972) have explained these results by pro-
posing that S_genes are normally inactivated earlier than autosomal
genes in the primary spermatocyte and that Sfautosome translocations
interrupt this timing by separating S genes from a proximal controlling
site. Considerable evidence exists for early §_inactivation in a wide
variety of male heterogametic organisms.

It is interesting that at least two §;4 translocations with central
S breakpoints cause meiotic drive while §;4 translocations with terminal
S_breaks do not (Novitski and Sandler 1957; Zimmering 1960; Chapter 5,
below). Perhaps the fundamental lesion here is the same as that involved
in translocation male sterility; the difference between sterility and
meiotic drive being a matter of degree. Quantitatively, the difference
is not a large one; T(l:4)BS males are not very fertile (Novitski 1970).
If the largely heterochromatic fourth chromosome is closer to the S in

terms of inactivation cycle than are the two major autosomes, these

21

observations would imply that prevention of early inactivation of por-
tions of the S chromosome can lead either to meiotic drive or to
sterility, depending on the severity of the disruption.

Another parallel between sex chromosome meiotic drive and chromo-
somal sterility can be seen in the phenotypes of basal S deficiencies.
All heterochromatic deficiencies encompassing the bobbed locus which
have been examined thus far cause nondisjunction and meiotic drive.
Deficiencies encompassing both bobbed and suppressor of forked (§Q(§)),'
the most proximal known gene in the §_euchromatin, are male sterile,
at least in the presence of certain X_chromosomes (Lifschytz and
Lindsley 1972). Since these deficiencies are missing essential genes,
they are male lethal except in the presence of Z_chromosomes duplicated
for the proximal §_euchromatin, such as mgl:x_and SEX. There is a
negative correlation between the size of the duplication and male
fertility, an effect reminiscent of the impact of additional sex
chromatin on fertility in scl'sc8 males. The parallel would be stronger
if we knew that the fertile S£;22 combinations experienced drive,
but the appropriate tests have not yet been done.

Another interesting connection between sex chromosome meiotic drive
and male sterility has recently been observed (Lindsley, unpublished
data). Males carrying both a Zfautosome translocation and sc45c8 are
sterile even if the translocation is fertile in the presence of’ a
normal S, Attempts to map the S heterochromatic locus responsible for
this effect are inconclusive so far. It is not clear whether this
phenomenon is best seen as schsc8 modifying the phenotype of Zfautosome
translocations or vice versa. However, it does strengthen the view

that meiotic drive and chromosomal male sterility are closely related

22

phenomena.

Materials and Methods

 

The methods used in this study are those of classical genetics --
analysis of progeny counts from crosses of individuals with appropriate
genotypes. The materials are chromosome rearrangements; they are
described in the text as the experiments involving them are introduced.
All crosses were done on standard cornmeal-molasses-yeast-carragheenin

medium. Except where noted, crosses were done in single pairs.

o

abet

1'01

10:

.5

at

CHAPTER 1

 

SENSITIVITY QF_SEX CHROMOSOMES TS MEIOTIC DRIVE

 

 

Despite considerable research into the nature of scascS-induced
meiotic drive, (reviewed in the Introduction), some ambiguity remains
about the chromosomal targets against which meiotic drive acts. .This
study examines the sensitivity of §_and‘z_chromosomes to scascS-induced
drive. All investigators agree that in sc48c8/3_ma1es the recovery of
.XZ sperm is severly reduced relative to its reciprocal (nullofig, nullofz)
(Gershenson 1933; Sandler and Braver 1954; Peacock 1965). Some doubt
remains, however, about the recovery of S_sperm relative to S sperm.
When marked X chromosomes are used, 3 recovery is depressed (Sandler and
Braver 1954; Peacock and Miklos 1973). However, using an unmarked X}
Gershenson (1933) found equality of §_and Z_classes. Ambiguity also sur-
rounds the question of S_chromosome sensitivity. The absence of meiotic
loss points to gamete dysfunction as the mechanism responsible for poor
S recovery in sc4sc8[zfz_males (Sandler and Braver 1954; Cooper 1964).
However, post-meiotic loss has not been ruled out. If it can be estab—
ished that sc4sc8 is sensitive to drive in sc43c8[xfx_males, the sensi-
tivity of §_heterochromatin will remain in doubt because sc4sc8 is an
almost entirely euchromatic chromosome. It is necessary to measure the
sensitivity of S heterochromatic duplications to sc4sc8-induced drive.

Marked Z_chromosomes certainly are convenient for genetic studies,
but interpretation of results obtained with them is clouded by the pres-

ence of the translocated S genes. It is possible, especially given

23

24

Gershenson's results, that poor recovery of marked st from sc4sc8 males
is due to the translocation rather than to the Z_itself. It is known
that some translocations involving the S chromosome can induce meiotic
drive. Males carrying a translocation which moves about four-fifths of
the S euchromatin to the fourth chromosome (T§1;4)BS) produce distorted
gametic ratios with poor recovery of the 43§E_element (the translocation'
half containing the fourth chromosome centromere and the distal section
of the S) and of the X.(a marked Z) (Novitski and Sandler 1957). The
translocated pieces in marked st are certainly much smaller than in
T(l;4)BS and do not normally induce meiotic drive. But, in the presence
of drive inducers like scasc8 or T(l;4)BS, they might exhibit unusual
sensitivity to drive.

In order to determine whether the sensitivity of a marked X_is due
to the translocated S markers or to the X itself, a reexamination of
sc4sc8-induced drive was undertaken using unmarked Z_chromosomes. The
experiments were aimed at measuring recovery of the X_in sc4sc8/X and
sc4sc8/X/Sp_males. The latter genotype was tested because a previous
report (Haemer 1978) had indicated poor recovery of the X_despite regu-
lar disjunction from the free S duplication.

Uncertainties also remain about sensitivity of the S chromosome to
scqscs-induced drive. Peacock (1965) reported that in sc43c8/Z_males
the ratio of §_to nullojg, nullojx nuclei after the second meiotic divi-
sion is the same as the ratio of S’to nullojg, nullofz_classes among the
progeny. This suggesusthat no loss of Srbearing sperm (relative to
DUIIOfE, nullofiz_sperm) occurs. However, electron microscopy reveals
that in high drive males more than half of the sperm in some bundles are

defective, suggesting that some nonfz_sperm are malfunctioning (Peacock,

25

Miklos, and Goodchild 1975). Recovery of the §_is poor in scésc8[z[gp
(Haemer 1978) and sc43c8/X/S_(Sandler and Braver 1954) males. Cytologi-
cal studies (Cooper 1964) reveal that the S behaves as a univalent in
scasc8[xfz males, but is not lost during the meiotic divisions. Thus,
the evidence points to malfunction of Srbearing sperm in these males.

As a further test of this, S recovery was compared in high drive and in
low drive sc48c8[zfgp males. If the recovery of the S parallels that of
the Z_in these crosses, then the two chromosomes must be responding to
the same forces. The absence of nullofig, nullofz_sperm implies that
poor X_recovery is due to meiotic drive and not chromosome loss. Paral-
el behavior of the S and Z_would imply that the S313 also a victim of
meiotic drive.

The evidence demonstrating poor recovery of scasc8 from certain
types of males indicates that the S euchromatin is sensitive to meiotic
drive. However, since scl‘sc8 contains almost no heterochromatin, it is
not known whether or not the S heterochromatin is similarly sensitive.
When a free S duplication containing all the heterochromatin but very
little of the euchromatin is added to a scasc8 genotype, the males expe-
rience considerable meiotic drive (Haemer 1978). As the recovery of the
free duplication is the best of the three chromosomes, it is impossible
to say whether any duplication-bearing sperm malfunction. Two approaches
to this problem were taken. The first approach was to measure the sen-
sitivity of free S duplications so small that they do not disjoin from
the X, The recovery of the duplication can be determined under these
conditions because one can measure the recovery of otherwise identical
sperm with and without the free duplication. This can not be done with

the larger free duplications which disjoin from the 2; any sperm which

26

lack the free duplication carry the X_and are not otherwise identical to
duplication sperm, all of which lack the X.

In the second approach, chromosome recoveries were compared between
males carrying scascs, a‘X, and a large free S duplication and sc4sc8,
3‘1, and a smaller free X_duplication (but large enough to disjoin regu-
larly from the Z). Substantial differences in Z_recovery would indicate

that the size of the free duplication makes a difference, a result con-

sistent with the view that drive affects all sex chromosomes.

RESULTS

1 Chromosome Sensitivity -- Crosses t_o_ Free-X Females

 

 

The first set of experiments test for drive sensitivity of unmarked
X chromosomes in scascall males. The presence of a bobbed locus (31%)
on the X chromosome makes it possible to distinguish Iii-bearing from non-
_Y_-bearing progeny. Although normal S's also carry a bobbed locus, scl‘sc8
is completely deficient for it. In the first experiment, SC48C8, XE: _b_b_-_/
1 males were crossed to females of the genotype 1 Q fill E lag/mghffi,
y: 113:. These females are homozygous for a moderate bobbed allele, a
partial 5% deficiency. The bobbed locus on the free duplication covers
this deficiency. Half the disjunctional eggs are S and half are QR. In
the S eggs, progeny from all four sperm classes can be distinguished but
only two of them (the §Y_ and 1 classes) have normal viability. The other
two classes are bobbed and cannot be used to estimate sperm frequencies.
In the 2% eggs, all four sperm classes are viable but the X can not be
detected. A few simple calculations provide reasonable estimates of all
four sperm classes. First, to estimate the recovery of _X11_ and 1 sperm,
the numbers of 1 3:: females and y 3 males respectively are used. Second,
to estimate the recovery of S sperm, the y 15: females are subtracted
from the 3: females (because the _w: females include both _1_(_ and _XS classes
while the y 3‘: females come from E sperm only). Third, I _w males are
subtracted from 3 males to obtain an estimate of the recovery of nullo—S,
nullo-X sperm.

Table 1 presents the results for four unmarked X chromosomes and
two marked X's. Several points emerge from an examination of Table l.
The first, and most important, is that poor recovery of '_1'_ sperm is ob-
served whenever sc4sc8/l males are tested, no matter what kind of S

27

28

Table 1 Y Chromosome Recovery from scascajl Méles

Recoveries:

Y Chromosomes

 

 

 

Sperm Genotype Pgepy Phenotype Y1 Y2 Y3 Y 4

x y_ 1‘: lg Females 446 85 4 96

_X_Y_ ' x 3: Females 58 3o 7 31

x or a fi Females 1504 961 121 766

S y g _S Males 64 25 20 158

I y ! Males 982 715 65 471

_0_ or X 31 Males 1370 907 129 797
Estimates of Sperm Frequencies Y1 Y2 Y3 Y4 w+Y BSY

 

 

1446 931 114 685 149 1113

 

 

§Y_ 58 30 7 81 10 152
_Q 388 192 64 326 71 1287
X 982 715 65 471 70 688
Nondisjunction -- SI (X+0) . 21 . 17 . 36 . 32 . 32 . 54
Recovery Ratios -- 1:33 .68 .77 .57 .69 .47 .62
-- £:2 015 016 011 025 014 012

SC4SC8, y if: bb-/l males were crossed to y 1 1113/1 g SS/Dp(l;f)3,
y: bb+ females. The estimates of sperm class frequencies are calculated
as outlined in the text. For the marked X chromosomes, the sperm class

frequencies are actual numbers, not estimates.

29

chromosome (marked or unmarked) they carry. All four unmarked Eds and
both marked Zfs gave I S ratios well below 1, ranging from 0.47 to 0.77.
The second point is that nondisjunction is frequent in all cases, but is

highest for BSY and lowest for Y1 and Y2. The third point is that drive

is invariably more severe in the nondisjunctional class (S339) than in
the disjunctional class (Egg). The final point is that recovery ratios
show a rough inverse correlation with nondisjunction frequencies (al-
though there are some exceptions). Thus, these data reaffirm the oft-
remarked correlation between nondisjunction and drive (Peacock and

Miklos 1973).

Z_Chromosome Sensitivity -- Crosses SQ_Attached-X Females

 

 

To be sure that the poor recovery of the Z_in these crosses was not
an artifact of the experimental design, a second set of crosses involving
the same males but different females was undertaken. Table 2 presents
the results for crosses of sc48c8/X_males to three different kinds of

attachedjg females : (l) C(l)DX, bb-IBSY, (2) C(1)RM/BSY, and (3) C(l)RM[Q.

 

 

In the second cross, all four sperm types produce viable progeny, but the
Z_can not be detected. In cross (1) nullofg, nullofz sperm can not be
recovered because C(l)DX is 22:. All surviving females come from :7
bearing sperm. Since crosses (l) and (2) share a common class (scascsl

BS / or 9), that class can be used as a standard to derive a weighting

ratio. The estimated number of 2 sperm in cross (2) is the number of.z
sperm in cross (1) multiplied by the ratio of males in the two crosses.
The recovery of nullofS, nullofiX_sperm is total females in cross (2)
minus the estimated number of females derived from Z_sperm in that cross.

The only surviving males in cross (3) come from S: sperm since scasc8 is

bb-. Making use of the fact that crosses (2) and (3) share a common

30

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

coo mom mm Hum moaocosvopm mmmao Euoam vmumeflumm
0 mm o
l. I.|||IIII II! o
xloaaaz oHan>cH moan: wmmomwum Inn oHan>nH I
o\zmAHVo AmV
wmmfi o
I .I zmaavo
moamsmm >\zmfifivu no o\zmAHvo monEowmumz oHnmfl>cH
0 «cm
no I. um um
I Hm: w\wmm\w q wmm
xIOHasz oanma>aH uo wmm\womqom
wmm\zmaavo Ame
Hum o
1. n. zmaavu
monEom M\ZMAHWU uo o\zm-vo moneowmuoz manmfi>cH
a - Ima|
m um um III
I. a He: »\»mm\m a wmm
xloaasz manofi>cH no wmm\momqom
wmm\nas .xmaavo Adv
o o o
memEom .lll NOAHVU
w\onfimo Inn oHan>=H moHMEoMMuoz oHQmH>cH
» o _ w\ an .mumqum an awomqom mmzuocmu m uonou Hacksaw: mwouo
mmmao Euomm

 

monEwm xlvmsuouu< magma moan:

w c

um um Eoum w voxumECD so we huo>ooom

N oases

 

fr.

['4

31

class (C(l)RM/S or _0_), the estimated number of SS sperm in cross (2) is
the number of SS sperm recovered in cross (3) weighted by the ratio of
females in the two crosses. The recovery of S sperm is total males in
cross (2) minus the estimated number of males derived from g sperm. The
resulting estimates are listed in the last line of Table 2. These esti-
mates are not very different from those obtained using the same males

but normal S females (Table 1). In both cases, S recovery is substan-
tially worse than S recovery and SS recovery is very poor. There can be
no doubt that the sensitivity of S chromosomes to scl‘scB-induced meiotic
drive is a property of the S itself and not of the translocated S genes

used to mark many S chromosomes.

S Chromosome Sensitivity -- sc48c8/S/Sp Males

 

It has been reported that the addition of a largely heterochromatic
free-S duplication to a scl'sc8 genotype results in poor S recovery despite
regular disjunction of the S and Sp (Haemer 1978). To find out whether
poor S recovery in this genotype is a property of‘ the S itself or of
translocated S markers, several unmarked S's and two marked S's were
tested in crosses of sc4sc8, yfiSi/S/ngl'2f)3, i i males to yyflbi/
y 3 SS females. The females were chosen to permit easy detection of
3:112 nondisjunction by the occurrence of y la- SS (scascsly 1". b_b_) female
progeny. Very few such females were observed in any of the crosses (Ta-
ble 3). The reciprocal nondisjunctional class (Q) is detectable by
phenotype only in the marked S crosses, where it proved to be very rare.
To make sure that these nondisjunctional males are also rare in the un-
marked S crosses, the y: _w offspring males in line 1 were tested for fer-
tility. All were sterile, indicating that the S and the Sp disjoined

from each other regularly in these crosses as well.

32

mm.

mm.

#8.

mo.

no.

N

ma ceases mm» .H seas ea c was

ma caumu >uo>ooou.mmMM ogfi

+

mm.

Ho.

we.

.wcauammwo 0Hm8.HN.onu Ham «0 umou muwawuuom m no momma

.moaoaumoflmaom ma owuou huo>ooou.mmm.onh

.3m + gimp. + Mod

.H mason. a.“ an mono mama on“. who voummu PM “3m 2.5. .MM N. NW3 M». N

c

nn_HN .mNmmHme\N‘ no.mm N .mom on mama osu mo mmouo m mo muasmou onu muaomoun mafia comm

 

a sass man was ass mam
A mmm mmm mmm mmm w+3
. Comm mama sows mes s»
. mama mom omm mmm m»
. mmam smNN macs smug N»
o amen mus mam mm~ 1»
Nli .JmWI IMWI. IMwW. NMMI oEomoEouso w

 

mowumm muo>ooom

 

mommmao ahwmm

mo mofiuaosuoum

 

 

moan: mammavmn\M\womcom Eoum mosomoaounu.m.vco.mimm.wuo>ooom m manna

33

In all crosses, the recovery of the S_is worse than that of the
duplication, with 1322 recovery ratios ranging from 0.28 to 0.68. The
failure to recover nullojS, nullofiS sperm rules out S chromosome loss;
thus, poor S recovery must be due to meiotic drive. Since crosses in-
volving unmarked as well as marked S chromosomes exhibit this effect, it
can not be attributed to the influence of translocated S genes. Poor S
recovery from scésc8[S[22 males (just as with scasc8/S_males) is a prop-
erty of the S chromosome itself. This is not to say, however, that the
translocated S genes make no contribution to chromosome sensitivity.

Both in Table l and Table 3, SES shows the highest drive, suggesting that

the S duplication enhances its sensitivity.

S Chromosome Sensitivity -- Euchromatin

 

 

The data in Table 3 indicate that recovery of the S'in sc4sc8/Spr_
males is much lower than expected and is of the same order as that found
with sc48c8[SfS males (Sandler and Braver 1954). The failure to observe
meiotic loss of the unpaired S’in sc4sc8/S]S_males (Cooper 1964) suggests
that its poor recovery in these males and in scésc8[SfSp_males is due to
sperm dysfunction. Further evidence in favor of the idea that Srbearing
sperm malfunction in these males can be found in Table 3. Here we find
that with different st, the S322 and Stg_recovery ratios both vary over
quite a considerable range. In any one cross, however, the SSQ_ratio is
very close to (usually slightly below) the S322 ratio. This correlation
argues that the forces causing poor S_recovery are the same as the forces
causing poor S recovery. Knowing that the mechanism is sperm dysfunction
in the case of the S, we are justified in presuming that S sperm also

malfunction in these males. This argument is not watertight, but until

the definitive electron microscopy is done, it is a reasonable conclusion.

34

S Chromosome Sensitivity -- Heterochromatin

 

The evidence for S chromosome drive sensitivity in sc4sc8/SjS_and
scascB/Sfyp males reveals nothing about the sensitivity of S heterochro-
matin, because scl‘sc8 is an almost entirely euchromatic chromosome. In
crosses involving Qp(l;f)3, a largely heterochromatic S fragment, nothing
can be said about the recovery of the duplication except that it exceeds
that of the S, The difficulty is that the duplication disjoins from the
S, making it impossible to measure the effect of adding the free dupli-
cation to a sperm without changing any other aspect of the genome. All
'22 sperm lack the S_and all nonfigp sperm carry 31S, To circumvent this

problem, Dp(l;f)164 and Dp(l;f)1144, two small free duplications approx-

 

 

imately the size of the fourth chromosome, were tested. Males of the

genotypes scasc8/BSY/Dp(l;f)l64 and SC4808/BSY/DB(1;f)ll44 were generated

4sc8/BSY females to attachedfiSS/Dp(l;f)l64 and

 

4 8
from crosses of sc sc /sc

 

attachedfiSS/Dp(l;f)ll44 males and were crossed to yfy females. The re-

 

sults, presented in Table 4, indicate that the free duplication segre-
gates randomly in the vast majority of meioses. The results are equivo-
cal as to sensitivity of the free duplications. The recovery of Dpll44
is 0.94 which is significantly different from 1. The recovery 0f.221§3
is 0.98, not significantly different from 1. This may mean that the
duplications have different intrinsic sensitivities. An alternative
explanation is that the duplication disjoins from one of the other chro-
mosomes more frequently in the SS§_case than in the 1133 case. Occasional
disjunction would result in an excess of Sp_offspring and a shortage of
.XQE;.XQE; and SSSp_offspring relative to the nonfiQp classes. This dis-
tortion is evident in both crosses, but is more frequent in the SSA case

than in the 1144 case. Whenever the duplication disjoins from a

35

.moHMEom N\N ou mommouo mums

 

 

 

mam moans qofiawwavmo\mwrmonomuum mom qqfiHNmmvan\Mmrmonomuum ou monEom wmm\mom<um\ on on «o mommouo Eoum

 

 

 

 

 

m a
mouoaow no moon—IN. m m IN.omo no moon—IN. m 9 INN.om on
m o a a H + amass av n\»mm\use as m e m e a + asaaam av a\»mm\uss a m a
ma. mam sen omen mama as as was mmm smama\»mm\mumsum
so. «om mmm mafia owns an“ new mam mam seaama\wmm\messum
ma we was w an o mama ax max x masseuse assesses

muo>ooom

 

mommmao apomw

mo mowocosooum

 

maowumowamsa x mosh Hamam mo hmo>ooom q mapmH

36

chromosome, its recovery will be much greater than 1 since its disjunc-
tional homolog, either the S_or the S, is much larger than and therefore
more sensitive to drive than the Sp. The small fraction of meioses in
which the duplication disjoins will tend to inflate the overall recovery
of the duplication. The slight difference in apparent sensitivities of
the two duplications may be due to the slightly greater tendency of
‘SpSSfi_to disjoin from its homologs.

A second approach to the problem of‘S heterochromatic sensitivity
was to compare the relative sensitivities of Dp(1;f)3 (a large hetero-
chromatic S duplication) and S§_(a somewhat smaller S chromosome frag-

ment) against a common standard, BSY. Sibling males of the genotypes

4 8 4

sc sc /BSY[SE_and sc sc8/BSY/Dp(l;f)3 were generated from.a cross of

 

SC48C8/SC48C8/Dp(l;f)3 females to y_g/BSY/Sf_males. These males were

 

then crossed to y _w SS/y _w S13 females. The data are presented in Table
5. It is evident from the absence of y 3: fl; females among the progeny
of these crosses that both fragments disjoin regularly from SES.
However, recovery ratios in the two crosses are not the same. The

recovery of BSY relative to SE_is considerably worse than its recovery

relative to Dp(1;f)3.

37

mmouo a so: oououoaow who: moama +ob “N .QOUma\wmm\lnp .mfi N .w

«m. AN.

ow. III ma.

 

" m u a
o x a wmm my wmm

moauom mwo>ooom

 

wmm

omm

mam

.moamaom fl M N<Mm N N 3 oommouo 0.33 one moans mtwmann 3 N 3 monEow mmmfiummJomqoflwomeum mo

 

IIII fiom nu
mama III Hm

Illl. .NII .llll
m» ax wmmx

 

mommmao auomw

mo mofiocosooum

0D 0m

8

 

 

w m ou o>wumaom mammavma mam w «o mofiua>auwmcom o>wun wo oomwummaoo

m

m

cm 3.N . on on
e amm\mw\uns m

w c

 

.4. a um um
mam S Eamflm a

II. um um
wmm\m» m s

 

omhuoaoo Hmauoumm

m manma

a!

l
0
A]

F]

DISCUSSION

The experiments described in the first section of this chapter
demonstrate that poor recovery of S_chromosomes from sc4sc8 males is a
property of the S_chromosomes themselves and not of the translocated S
genes used as markers. When scasc8[S_males with unmarked S chromosomes
are crossed either to attachedfiS or to regular females, the recovery of
the S_is lower than that of the S, Similarly, sc43c8[SfSp_ma1es carry-
ing an unmarked S_show poor S recovery.

The question of S chromosome sensitivity is somewhat more problem-
atic, although a partial answer can be given. The evidence presented
here indicates that poor recovery of the S from scasc8/S/S_and scésc8/S/Sp_
males is most likely a consequence of dysfunction of Srbearing sperm.
The alternative, postdmeiotic chromosome loss, is unlikely because the
recovery of the S chromosome responds to the same modifying forces that
determine the level of S chromosome recovery, a phenomenon certainly not
due to chromosome loss. There remains some uncertainty about the recov-
ery of the S chromosome in sc4sc8/S_males. Light microscopic studies
(Peacock 1965; Peacock and Miklos 1973) argue for full S recovery, but
electron microscopy (Peacock, Miklos and Goodchild 1975) demonstrates
sperm dysfunction of nonfS_sperm in some instances. Uncertainty also
'remains concerning drive sensitivity of S heterochromatin. The results
of the small free duplication experiment are consistent with mild sensi-
tivity of both 221144 and SESSS -- a sensitivity partially masked by the
tendency of both duplications to disjoin occasionally from the other
chromosomes. The results are also consistent with other hypotheses.

For example, Dp1144 but not 221Q4_may be sensitive. Alternatively,
Dp1144 might be lost meiotically on occasion.

38.

39

The results obtained with large free duplications are consistent
with the view that all sex chromatin is sensitive to drive and that the
length of a chromosome determines its degree of sensitivity. However,
it is impossible to demonstrate by genetic means alone that the least
sensitive element of a genotype is susceptible to meiotic drive.

Three other points concerning this study deserve mention. One is
that S322 recovery ratios from scasc8[S[Qp_males (Table 3) are invariably
lower than S3S_recovery ratios from.scasc8[S_males (Table l) carrying
the same S, In light of suggestions that faulty S:S pairing is respon-
sible for drive in scasc8[S_males (Baker and Carpenter 1972; Peacock and
Miklos 1973), it is interesting that supplying the S_with a regular
pairing partner in the form of a free S duplication which carries a full
complement of pairing sites does nothing to enhance S_recovery. In fact,
it seems to make matters worse.

The second point is that recoveries of the S_and S from scésc8/S722
males are not independent of each other. In every cross the recovery of
.SS sperm is worse than expected on the basis of independence. (S? is
significant at the .05 level.) Evidently, S_and SLChromosomes interact
when present in the same sperm. This contradicts Haemer's (1978) finding
of independence in crosses of this sort. The reason for this difference
is unclear. Perhaps the overall higher drive levels in the present ex-
periments account for the occurrence of interaction in them but not in
Haemer's experiments.

Finally, the largely random disjunction of the unpaired small free
.S duplications Slflfi_and S§fi_(Table 4) in spermatocytes nondisjunctional
for S_and S is intriguing in light of Peacock's (1965) finding of nonran-

dom disjunction of‘S and S when they fail to pair. The forces causing

40

the unpaired S_and S_to migrate to the same pole do not affct the free

duplications. Perhaps they are too small to be included in this system.

CHAPTER 2

GENETIC BACKGROUND EFFECTS 91 LEVELS g1: MEIOTIC DRIVE

 

Sex chromosome meiotic drive systems in Drosophila melanogaster are

 

subject to modification by genetic background effects, including source
of S chromosome (Zimmering 1960; Peacock, Miklos, and Goodchild 1973),
source of autosomes (Zimmering 1960; Peacock, Miklos, and Yanders 1972),
amount of autosomal heterochromatin (Haemer 1978), and amount of S heter—
ochromatin (Haemer 1978). Males carrying T(l;4)BS (an S;_4_ translocation
with the S breakpoint in the proximal euchromatin) produce highly dis-
torted sperm ratios in some genetic backgrounds but not others (Novitski
and Sandler 1957; Zimmering 1960). With "A-type" autosomes and S chromo-
some, the recoveries of the S chromosome and the E element are im-
paired. With "E-type" autosomes and S chromosome, sperm ratios are nor-
mal. Mixtures of E and A type chromosomes produce intermediate levels of
distortion. In males carrying the S heterochromatic deficiency In(l)sc4L

sc8R (scasc8), the frequencies of S—S nondisjunction and of recovery dis—

 

ruption of SS and S sperm vary depending on the S chromosome used and
upon the segregation of uncontrolled, presumably autosomal, modifiers
(Peacock 1965; Peacock, Miklos, and Yanders 1972; Peacock and Miklos
1973). The addition of a S chromosome or a heterochromatic free S dupli-
cation to a scl'sc8 genotype enhances recovery disruption. In scésc8/S/S
males, the two S chromosomes pair and disjoin regularly (Cooper 1964).
The univalent S is recovered poorly (Sandler and Braver 1954) despite

absence of meiotic loss (C00per 1964). In scascg/S/Sp males (where pp

41

 

42

is one of several largely heterochromatic S fragments called free dupli-
cations) the S_and Sp_disjoin from each other, and recoveries of both
the S and S are depressed (Haemer 1978).

No systematic effort to sort out the effects of S_chromosomes and of
autosomes on the scl‘sc8 system has previously been made. Experiments
exhibiting S effects on nondisjunction and meiotic drive in scasc8 males
(Peacock and Miklos 1973) have not included controls on autosomal back- #25
ground. In the present study autosomes and S_chromosomes were varied
independently to obtain answers to the following questions concerning

drive in sc4sc8 males. 1) Can different S_chromosomes affect levels of

 

drive and nondisjunction in scl‘sc8 males independent of autosomal back-
ground? 2) Can different sets of autosomes affect levels of drive and
nondisjunction independent of sex chromosome content? 3) Does modifica-
tion by S.or autosomal background cause parallel changes in levels of
nondisjunction and drive in sc4sc8/S males? 4) Does modification byLS
chromosome or autosomal background cause parallel changes in S_chromosome
recovery from scasc8[S_and scasc8[S[Sp flies? 5) Does modification by
S_or autosomal background cause parallel changes in recovery of both S
and S chromosomes from sc4sc8[S[Qp flies? 6) Does modification by S_or
autosomal background cause parallel changes in.S recovery in scasce/ngp
and scasc8[S[S_males?

These questions are interesting because they bear upon several the-
oretical issues. One iSSue is the nature of the connection between pair-
ing and meiotic drive. Drive levels correlate with nondisjunction in
scasc8/S_males where S:S_pairing interactions are weak (Peacock 1965;
Peacock and Miklos 1973). This has been taken to imply a causal rela-

tionship between weak pairing and meiotic drive (Baker and Carpenter

43

1972; Peacock and Miklos 1973). If chromosome recoveries respond in
parallel fashion to modification in sc4sc8/S, sc4sc8/SfSp, and scésc8/S/S.
males, it is reasonable to conclude that the primary lesion in these
males is the same despite superficial phenotypic differences. It would
also be reasonable to conclude that meiotic drive levels can vary inde-
pendently of nondisjunction since there is no nondisjunction in sc43c8/
S[Sp_or scésc8/S/S_males.

A second issue is the reason for poor S recovery in scasc8[S7S_and
sc4sc8/Sfyp_males. Parallel modification of S_and S_chromosome recovery

in sc4sc8/SfSp_males would argue that S and S chromosomes are recovered

 

poorly for the same reason (i.e., sperm dysfunction, not chromosome loss).

‘4.

If modification by source of S_chromosome can affect S_recovery as well
as S recovery, that would argue that drive levels are influenced by in—
direct (non-pairing) interactions between chromosomes since the S and S
do not pair in these males.

The relative levels of S and S recovery from sc4sc8[S[Sp and sc4sc8/
S[S males are of interest because they have implications for the occur—
rence of competition between sex chromosomes. If the additional sex
chromatin in scasc8[S/S_males reduces recovery of the S, that implies
that chromosomes compete for a scarce resource in sc4sc8 males. Better
relative S recovery from sc4sc8/S/S_males than from sc43c8/S/Sp_males
implies either that closer size matching of sex chromosomes leads to
better recovery (presumably through pairing interactions) or that abso-
lute chromosome recovery is a function of chromosome length. Greater
similarity in size of two paired homologs may lead to greater relative
recovery of the larger element but not necessarily greater absolute

recovery. These issues are explored in more depth in the Discussion.

RESULTS

Experimental Desng_

 

Three experiments were undertaken. Experiments 1 and 2 (diagrammed
in Figure l) were very similar, differing only in choice of marked S
(w+Y or BSY) and in time of execution. In both experiments scésc8/S,

scéscslmarked S, sc43c8/S/marked S, scasc8[S/Dp(l;f)3, and scasc8/

 

:13
marked S/Dp(l;f)3 males were generated as siblings from a single cross.
In each experiment three such crosses were performed using S_chromosomes
and autosomes from different laboratory stocks. Thus in each experiment, . E
fifteen different types of males were generated and tested simultaneously. E7

The data from experiments 1 and 2 are presented in Tables 6-8. The third
experiment was a comparison of drive and nondisjunction levels in males

with identical sex chromosomes but different autosomes.

S Chromosome Modification SQ sc45c8/S_Males

 

Table 6 reveals that different Sfautosome sets do indeed cause dif-
ferent levels of nondisjunction and drive. In both experiments 1 and 2,
the chromosomes from stock 1 are associated with much lower nondisjunc-
tion and higher recovery values than the chromosomes from either stock 2
or stock 3. The latter two stocks have similar disjunction and recovery
coefficients.

Are the differences between stock 1 and the other two stocks attrib—
utable to the autosomes or to the S chromosomes? Two comparisons bear
on this question. The first is a comparison of crosses with the same Y
chromosome (either E:S_or SSS) but autosomes drawn from stocks 1, 2, or
3. Although simultaneous crosses involving the same S_and S.do not give

identical results, the similarities are certainly more striking than the

44

45

Experiment 1

 

22/2 E/il X W1),

 

Zfl/fi/Yi X sc4sc8/sc4sc8/Qp(1;QB

 

 

 

 

sc48c8/Yi

sc4sc8/Y—7wa X w bb/ w bb/ l'f)3

——-_i-— z__1__m_s__

8C48C8/W+Y

scascslw+Y/Dp(l;f)3 E
43 XEWflflafl ”’

sc sc /Yi/Qp(l;f)3

Experiment 2

 

zyrwflLXQfih

 

y;ng§S/Y1 X $04868/8C48C8/D9(13f)3

 

 

sc4sc8/Si
sc4sc8/Yi/SS_S x 1 _w 211/1 31 pS/ngl;f)3
sc43c8/BSY
sc43c8/BSY/Dp(l;f)3

4 8 X Ereyxafl.
sc sc /Y1/Qp(l;f)3

Figure 1 Mating Scheme for Experiments 1 and 2

eaaaz >\s

 

lib Ii

o\

 

PHI

I. 1 'l‘l‘ll’,l

C d,

94> u, Lfid

filed-(Q! :p... a .9

N»!uh.n-.fi(mu..§.- yF-A.vz

46

NH.

NH.

HH.

mm.

OH.

no.

mH.

dfi.

«H.

mH.

mg.

ca.

onyx

No.

00.

mm.

mm.

Hm.

mm.

mm.

mm.

or.

Nm.

om.

 

mowumm mmu>oomm

moans—ow MAECNAQNM N NEM. N N 9. mommouo 983 mmHmE M\wom om

 

q

n xeoam N
m xsosm a
N suosm N
N xeoem a
a seoem N
a aeosm a

 

 

cm. mama wmo and mafia
mm. cmw «Nod oofi mama
ow. mmH OHN Nu mwm
um. own due Hm mwo
on. mfiqfi com Ned mom
on. mmofi coma Hog «NAN
mm. mqfi owd mH mum
on. so no A egg
on. end“ 5mm 05H «mm
ma. 0mm omofi we momfi
mm. mm mm «H HNH
he. NmH was Om Hmm
~o+xv\o o » xx x
coauocow mommmao Euomw
Imaosoz
moan: w\womcoo cw o>wuo mom coauosnflmfionoz

oaomoEouso »

o maama

mosomous< uaoawuomxm
mo oouaom

 

47

differences. In experiment 1, there are no significant differences among

the three scasc8/w+Y crosses in terms of nondisjunction or meiotic drive.

 

In experiment 2, the differences in drive and nondisjunction are signif-
icant, mostly because of the abnormally high nullojS, nullojS_class in
the stock 2 cross. However, this difference does not account for the
major effect in Table 6 -- namely, the consistently low drive and nondis~

junction of the Y1

crosses. Those differences disappear when the S chro-
mosome is held constant. Evidently, they were caused by the S_chromosome
and not by the autosomes. This conclusion is confirmed by the second
comparison in which autosomes are held constant and S_chromosome is
allowed to vary. In Table 1, six pairwise comparisons involving differ-
ent S chromosomes but constant autosomes (e.g., line 1 with line 2, line
3 with line 4, etc.) can be made. Nondisjunction and recovery coeffi-
cients for members of a pair are not in general the same, and in several
instances are widely divergent. Clearly, the S chromosome can have a
drastic effect on levels of drive and nondisjunction when autosomes are
held constant.

These results confirm those of previous investigators (see Peacock
and Miklos 1973) in showing a strong correlation between nondisjunction
and drive. In Table 6, stock 1 shows consistently lower nondisjunction
and higher StS_recovery than the other two stocks. Similarly, SES_shows

both higher nondisjunction and lower S chromosome recovery than w+Y.

S_Chromosome Recovery SQ sc4sc8/S]22_Males

 

Another genotype in which it is possible to monitor S_chromosome
recovery is sc4sc8/S/Sp, Males of this genotype were generated as sib-
lings to the scasc8/S_males discussed above. Do these scésc8/S/Sp males

show the same inter-stock patterns of S_recovery as their sc4sc8[S

 

48

brothers. The data in Table 7 reveal that they do. Just as in Table 6,
recovery coefficientsare consistently higher for crosses in which S_and
autosomes have been derived from stock 1 than for crosses with chromo-
somes from stocks 2 or 3. Once again, stocks 2 and 3 are very close to
each other.

Are these differences attributable to the autosomes or to the S
chromosomes? The data in Table 7 permit comparison of simultaneous
crosses involving the same S chromosome (either E:S_or SES) but different
sets of autosomes. Again, the differences among these crosses are minor
compared to the differences observed when both the S_chromosomes and
autosomes vary. There are no significant differences in experiment 1.

In experiment 2, the S322_recovery ratios are significantly different,
but this difference is not repeated in the S;Q_recovery ratios, which
are the same. The autosomes alone do not contribute significantly to the

main effect -- the low drive in Y1 crosses. When autosomes are held con-

stant and pairwise comparisons between crosses involving the same auto-
somes (lines 1 and 2, lines 3 and 4, etc.) are made, the differences in
S recovery between members of a pair are generally substantial, sometimes
large. The S chromosome has a considerable impact on levels of meiotic
drive in these males as well as in scasc8/S_males.

4sc8/S/_Dp_males provide sup-

The parallel results for scasc8/S_and sc
port for the idea that the same defect is responsible for the meiotic
anomalies in both genotypes. Evidently, the defect is not simply absence
of an S heterochromatic function, since the duplication contains all the
heterochromatin missing from sc43c8. Rather, it must be the separation

of the heterochromatin from the bulk of S euchromatin that causes these

problems.

49

mm.

mm.

mm.

He.

mm.

cm.

on.

me.

Nm.

ow.

om.

mm.

em.

we.

as.

om.

mm.

me.

co.

om.

me.

 

mouumm zuo>ooom

Hum

dom

qu

HowN

coda

mama

Nod

onwfi

omm

mmm

qu

wmau

mm

«mm

me

ocm

nomfi

omm

qu

mu

wca

mom

mom

50H

omNN

W

own

mmm

mmm

coma

Hmc

HHQ

mu

cow

own

Nun

omfi

mmom

max

 

mommmao Buomw

oq

ow

«ma

nee

mm

NNH

Hm

mmm

we

find

ow

qwm~

.moamaom Mm N NW3: N N 3 oommouu mums monE m~m~.HMmo\M\womqom

m suosm N
m xeosm H
N seoam N
N seoam a
n seoom N
H aaosm a

 

 

oEomoaouco »

 

seam: mammavmn\w\mom

e

on Eoum hum>ooom Ehumm

n oaama

mofiomousm ucoafiuomxm
mo oousom

 

 

50

These results also demonstrate that it is possible to modify meiotic
drive levels in scASc8 males without at the same time modifying the sex
chromosome disjunctional pattern. In every sc4sc8/ngp cross performed,
the S_and the free duplication disjoined reliably from each other, and
scasc8 assorted randomly. Drive levels nevertheless varied over a con-

siderable range.

S and S Chromosome Recovery E scasc8/S/l_)p_ £133

The data in Table 7 show that S chromosome recovery is always de-
pressed in sc43c8/S/Sp_males. They also reveal a strong correlation
between recovery of the S_and the S. As pointed out in Chapter 1, the
same forces causing recovery disruption of the S are also acting on the
.S. These data supply more cases supporting this correlation. The data
in Table 7 also show that the recovery of S22 sperm is not always lower
than that of S sperm as it is in all the unmarked S cases. .EEX.1S more
sensitive than the S and Sp together, perhaps because of the S_genes on
B Y.

These data also permit another look-at the question of independence
of.S and S chromosome recovery. The scasc8/Sf22_experiments discussed
in Chapter 1 all revealed an interaction between S and S_chromosomes
such that SS sperm fared worse than expected. The new data in Table 7
(experiment 2) confirm this in all cases but one. Non-independence is
especially striking in the SES crosses which also show the highest drive
levels of any cross. This helps to confirm the earlier suggestion that
degree of interaction may be related to overall drive level. The one
independent case in Table 7 also helps to confirm this notion as it con-
tains the S chromosome and autosomes from stock 1, the "low drive"

stock.

51

Chromosome Recoveries from sc43c8/SjS_Ma1es

 

These experiments were designed to permit evaluation of sc4sc8/S/S_

males along with the other genotypes. However, scasc8/w+Y[S males were

completely sterile. scasc8/S/BSY males were fertile but did not produce

 

very many offspring. The data are displayed in Table 8. No distinctions
among stocks can be made on the basis of these data as the numbers are
too small. Still, two conclusions can be reached. One is that the re-
covery of S§S_relative to that of the unmarked S_chromosomes is consider-
ably better than its recovery relative to the free S duplication (Table
7). This suggests either that the sensitivity of sex chromosomes to
meiotic drive is a function of length or that the sensitivity of the larg-
er of two sex chromosome homologs is a function of their difference in
length. The experiments described in Chapter 3 confirm and extend the
notion of length dependence although they do not permit a decision be-
tween these two alternatives.

The second conclusion is that the degree of S chromosome recovery
in scasc8/S/S_males is in the same general range found for sc4sc8[S[Qp
males. However, the numbers in the sc48c8/S[S_cross are not large
enough to permit a decision as to whether there is some difference be—
tween the two genotypes. This is a question of considerable theoretical

import and is taken up again in Chapter 3.

ModificationISy_Autosomes

 

The differences between stock 1 and stocks 2 and 3 in terms of ef-
fects on disjunction and meiotic drive are mostly attributable to the S
chromosomes. One reason for the minor influence of autosomes may be
that the test males were generated by crossing males from different lab-

oratory stocks to identical females and then crossing their sons to

52

NN. om.
mm. SN.
ms. mm.
mmmm mmwmm

 

mowumm huo>oumm

.moamaow mfimmavmo\na 3 %\nn 3 % cu moomouo mums monE wmm\M\ um um

mm

mom

mod

 

qu om
oqm mm
om“ cm
.4. .I
mommmao Shown

mm

mm“

we

 

no on
Hm: rmm\w\w

c

on Scum muo>ooom osomoaouno

m a

m sooem

N xeoam

a suoum

 

mosomoaounu r can
moEomou=< mo mounom

m eases

53

identical females. Any between stock differences would be considerably
attenuated after the equivalent of two generations of backcrossing.

That autosomal modifiers are segregating in these crosses is sug-
gested by the sometimes considerable differences among males in a single
cross. These differences can not reflect sampling error alone since they
affect S and S chromosome recovery in a parallel fashion in sc43c8/ngp
crosses. In two such crosses, individual males were ranked in terms of
recovery of Syp_sperm relative to Sp sperm and were divided into cate-
gories based on this ranking (Table 9). Recovery of S sperm relative to
Sp sperm was then determined for males in each category. It was found
that the Styp ratio paralleled the S22322_ratio, suggesting that these
two ratios measure the same underlying meiotic disruption. These be-
tween-male differences might reflect segregation of autosomal modifiers.
Alternatively, uncontrolled non-genetic factors might be responsible for
the differences.

Convincing evidence for autosomal modification of sc4sc8-induced
meiotic drive comes from a set of crosses designed for another purpose.
The purpose of the experiments was to evaluate the drive sensitivity of

small free S duplications (Chapter 1). scascalscascB/BSY females were

 

crossed to attachedjSS/Qp(l;f)l64 and attachedfiSS/Qp(l;f)1l44 males.

 

Both of the free duplications are quite small and frequently fail to
disjoin from the attachedfiSS, Males carrying scl‘sc8 and SES_with or
without the free duplication were recovered from both crosses and were
mated to y/y_females. The results are displayed in Table 10. The dif-
ferences between the two scasc8/BSY crosses are considerable: the StS
ratio is .67 in one case and .49 in the other while the SStQ ratios are

.22 and .082, respectively. The nondisjunction fractions are .54 and

.58, respectively. These differences must be due to different autosomes

54

Table 9 Within Cross Correlations between X and Y Recovery

 

 

 

 

Cross 1
XDptgp Ratio Y:Dp Ratio
below 0.6 .59
.60 - .69 .72
.70 - .79 .72
.80 - .99 .92
above 1.0 1.17
Cross 2
QSQp:Dp Ratio Y:Dp Ratio
below 0.3 .38
.30 - .39 .38
.40 - .49 .46
above .50 .59

Males from two scésc8[S/Dp(l;f)3 crosses were ranked into the
categories in the XDp:Sp_column. Then the S;Sp_ratio was cal-
culated for each category. Crosses l and 2 are from lines 1

and 5 respectively of Table 7.

55

 

 

a mzouv cqfifimmmavma\mwroosomuum Ou moamamw wmm\womqom\wum

 

mo. as. me.
mo. me. mm.
Na. mm. mm.
mm. no. em.
onyx xx» Ao+x no
moauom :ofiuuaSH
>uo>ooom nmaoaoz

«N\N ouoa monEmw use

mum

moo

Imml

can

flow

mmo

N©-

.aeame as eee m

 

omcfi

mama

mm

mama

mcmm

ONHH

amom

o

q

«a

an

mnwx

 

mommmao Euomw

om mo

 

assoc salammavma\mwremeamsaa as see xN mes

mommouo aouw ooumuocow mums moans one

 

 

 

 

 

um on

so moo mmw qodmmxwmm\m q
III: um on

man mama Nmm\m s
m. on on

fiqm mom mmo quH n\wwm\w q
III: on on

sms smaa smm\m s
wk max x omhuocmu Hmaumumm

 

moan: um on ma o>wun owuowoz com cowuo:=nmaocoz mo cowumoNMHooz Hmaomou=< o“ oHBmH

w

c

56

coming from the attachedfiSS[Sp_stocks since the sex chromosome consti-
tution of the flies is identical. The presence of a free duplication in
these crosses has no effect on the drive and disjunction ratios; sibling

scasc8/BSY and sc4sc8/BSYISp_males give the same results. An interesting

 

feature of these data is that the drive ratios appear to be more sensi-
tive to modification than does the amount of nondisjunction. This con-
trasts with previous reports of modification (Zimmering 1963; Peacock and
Miklos 1973; also see Table 6 above) in which nondisjunction is the more

sensitive parameter.

DISCUSSION

The major results of this study are 1) both autosomes and S chro-
mosomes can influence levels of nondisjunction and meiotic drive in
sc sc8 males, 2) recovery of S chromosomes is modified in the same direc-
tion in sc43c8[S_and scésc8[S[Qp males, 3) recoveries of S and S chromo-
somes are modified in the same direction in sc4sc8[S[Sp males, and
4) replacing the free duplication with a second S improves the relative
recovery of SES,

These results have several implications for understanding the mech-
anisms responsible for meiotic drive in scl‘sc8 males. One is that the
lesion responsible for meiotic drive and nondisjunction in scasc8[S
males must be the same as the one responsible for drive in sc4sc8/ngp
males. This is not surprising since both genotypes include a hetero-
chromatically deleted S chromosome. It is interesting, however, that
restoring the normal amount of heterochromatin by adding a free S dupli-
cation to the genotype does nothing to improve S chromosome recovery.

A second implication is that meiotic drive levels can vary inde-
pendently of the disjunctional pattern. In sc4sc8/Sfyp_males, where the
S_and Sp disjoin regularly, drive levels are at least as sensitive to
'S_modification as in scésc8[S males where disjunction is irregular and
also subject to modification.

A third implication is that the S chromosome participates in the
determination of overall drive levels in the spermatocyte. S chromosome
modification of drive levels in sc4sc8[S/Sp_males does not affect only
the recovery of the S chromosome, but the recovery of the S_as well.
Since the S probably does not participate in meiotic pairing (judging

from its random segregation) in these males, it follows that recovery

57'

58

levels can be determined by other than direct pairing interactions.

A fourth implication is that sex chromosome meiotic drive sensi-
tivity appears to display length dependence. The relative recovery of
“SES is better when it is paired with a longer chromosome (a second S)
than when it is paired with the smaller free duplication. In the latter
cross, since one chromosome is Sfderived and the other is a.S, the dif-
ference may be due to the differing genetic contents of the two chromo-
somes rather than to their lengths. No firm conclusion concerning length
dependence can be drawn until more cases have been examined. In the
following chapter, the drive sensitivity of several S chromosome frag-

ments is examined.

[Eu

dr

C0

dr

IE

ta

t}

51

Qt

In:

(I)

CHAPTER 3

SENSITIVITY SE S CHROMOSOME FRAGMENTS SQ MEIOTIC DRIVE

 

 

 

All of the sex chromosome meiotic drive systems examined thus far

in male Drosophila melanogaster share at least one common feature: poor

 

recovery of the S chromosome. Sensitivity of S chromosomes to meiotic

drive has been demonstrated in T(1;4)BS (Novitski and Sandler 1957),

4L 8R 8
so

In(l)sc (SCASC ) (Gershenson 1933; Sandler and Braver 1954; Pea-

 

cock 1965), and Recovery Disrupter (Erickson 1965). The question ad—
dressed in the following study is: what property of the S_chromosome
renders it sensitive to meiotic drive? Perhaps the S chromosome con-
tains a discrete response function analogous to the Responder locus in
the SQ_system (Sandler and Hiraizumi 1960; Ganetzky 1977). Or several
such functions could contribute to the overall sensitivity of the S,
Or S chromatin in general may be sensitive to drive. To answer this
question, the sensitivities of several centric S fragments were assessed.
If a single response function exists it must be on one arm or the
other (SE or SE) of the submetacentric S chromosome. Therefore, either
.SE or SE but not both should prove sensitive when tested as separate
chromosomes. If both fragments prove sensitive, then there are at least
two response loci, one on each arm. By measuring the relative sensitiv-
ities of a series of S fragments, it was possible to test the idea that
sensitivity to meiotic drive is a diffuse function of S chromatin rather

than a property of a discrete response locus.

The experiments described below involve males carrying schsc8, an

59

 

di:
ju
pr
th

i:

[In

60

S deficient for the basal heterochromatin. This chromosome causes high
‘S:S_nondisjunction and depressed recovery of S_and SS sperm relative to
S and nullojS, nullojS sperm respectively (Gershenson 1933; Sandler and
Braver 1954). Cytology reveals that chromosomes are not lost meiotically
(Cooper 1964; Peacock 1965) and that many sperm fail to develop properly
(Peacock, Miklos, and Goodchild 1975). When scasc8 males carry two S
chromosomes (Sandler and Braver 1954; Cooper 1964), the two st pair and
disjoin from each other while sc4sc8 acts as a univalent. Regular dis-
junction of the S chromosomes in these males is thought to be due to the
presence of collochores (S:S_pairing sites) on both arms of the S, Given
that collochores are observed on both.Sf_and SE, it was expected that

in the scasc8/SEfSE_males and in the other three-sex-chromosome genotypes
tested in the present study, the heterochromatic elements would disjoin
from each other. This expectation proved correct. Deviations from the
expected one to one recovery ratios were used to calculate differences
between the fragments in sensitivity to meiotic drive. Pairwise compari-
sons of fragments permitted a ranking of the fragments in terms of sensi-
tivity. Fragment features such as length and genetic content were tested
for correlation with drive sensitivity.

The experiments also permitted measurement of S recovery. In
scasc8[S[S and in the scascS/S_fragment/S_fragment genotypes tested in
this study, recovery of the unpaired SLis quite low. Since the various
S chromosomes differ considerably in length, it was possible to ask
whether S recovery varies in a systematic fashion with the total amount
of sex chromatin in the genome. If sex chromosomes in spermatocytes of
scasc8 males must compete for adequate supplies of a scarce but essential

resource, the consequence would be an inverse relationship between total

def

61

sex chromatin content and S chromosome recovery. Thus, measurement of
S chromosome recovery permits a test of the idea that S_heterochromatic

deficiencies cause scarcity of a chromosome processing material.

 

dell

mes

dis

.3

th

ti

MATERIALS AND METHODS

4Lsc8R, an X

The chromosomes used in this study include: In(l)sc

deficient for the basal heterochromatin; BSY; YEy3M, a bb+

 

S_chromosome

 

missing SS, derived by recombination between the short arm of an intact

S and the distal heterochromatin of In(l)scSl and described by its cre-
ators (Crew and Lamy 1940) as an acrocentric rod the size of SE in mitotic
metaphase (but surely somewhat longer); YL-y+BZ, a SS: S_missing SS,
derived by recombination between the short arm of an intact S and the

distal heterochromatin of In(l)scsLENR; SE, a spontaneous derivative of

 

bw+Yy+ lacking SS and described as a small two-armed chromosome in mi-
totic metaphase (Baker and Spofford 1959); SESSE, a two-armedleshaped
chromosome with two doses of SS and measuring almost twice the length of
the short arm; and Dp(l;f)3, an Seray induced deletion of most of the

euchromatin of the X that carries bobbed and almost all the rest of the

heterochromatin. See Lindsley and Grell (1968) for further details.

62

 

at

01!

CE

b:

RESULTS

Disjunction and Drive in sc43c8/YS/YL Males

 

To find out whether S fragments disjoin from each other in the
presence of scascs, and to determine S fragment drive sensitivities,

scascslfi/SE males were mated to y g 132/1 _w fl females. Sibling control

males carrying a normal S, SE, and SE13_M were also tested. The results
are presented in Table 11. In the control all six types of sperm with
one or two sex chromosomes were recovered with reasonable frequencies

suggesting that all three chromosomes participate in a trivalent asso-
ciation as is usually found in males with three sex chromosomes (Cooper

1964). In the experimental cross (sc‘sc8/SijE), the failure to recover

sc4sc8/y 1 213 females or y E SS/Sgfllfi males implies that scl‘sc8 does
not disjoin from either of the S fragments and that SE_and thalways
disjoin from each other. Thus the pairing sites on SE_and SE, although
presumably evolved for SjS pairing, are perfectly capable of pairing
with each other when carried on separate chromosomes. The results here
are completely analogous to those obtained with scasc8/S/S_males and
suggest that no functional differences exist among the pairing sites

carried on the S, SE, and YL.

If the S chromosome contains a single discrete response function,
then either XE_°I.XE.PUt not both should be sensitive to scasca-induced
drive. If, on the other hand, response is a function of length, then
both S§_and SE should be sensitive, but SE should be more sensitive since
it is longer. The results in Table 1 indicate that SE is more sensitive
than SE; the recovery of SE_relative to Sf_is 0.43. This result is con-

L
sistent either with the presence of a discrete response locus on S__or

with the idea that chromosome sensitivity is a function of length.

63

64

.twcmwoun N N do no

Sm
zufiiuoum 93 Scam moosooo .33 :5on gym» mo monomnm oFH. .memEom fl N N\Nm N N x moans Smwaimlwu womqum
was 3:: 2833 mmouo Hmucoawuomxo o5. .Ewam wry 50.5 M nmfiowafiumfio 3 33.33:" mg mo omsmooo mound
Iooamo mos owumu Numerous.“ Maul». oz .moamamm ANNE N N no mleBln N NV N 60.3 moamfiom AMQwomqogfl N NV

MN N wawuomuunnm .3 mouse—Ems m9.» .5on N no woman: of 2N... N N\wom.~omv moamamm .nlm MN N mo muadnmfir noon

95 mo omsmoom .moamsom N\momqum cu commons 933 moans 2mwaw\w..m\mm. N N 3:: umufiwv Houuaoo on“. :H

 

 

 

. . N II um um
Na 2 o m? SON mmm 08 o 2m aimim s
as. 22 ems NNa m2 as N afiflfl m N

Em A m
oux mwqu mama a» m» ANN max x mmasoemu Hmeamsmm
moaumm Ndo>ooom mommmau ~5on

 

moan: Aw\mw\x Eoum zuo>oomm Epsom mam moauocafimfin Ha oHan

65

These data alone do not permit a decision between the diffuse and dis-
crete models.. They do imply that if there is a single response locus,

it must be on SE,

Relative Sensitivity QS SE and S

 

Since there is no direct way to determine the sensitivity of the SE
chromosome, a less direct way of measuring short arm sensitivity is
needed. One solution is to compare the sensitivities of SE_and a come

plete S, If SE_is insensitive, then all the sensitivity of the S-is con-

centrated in SE, SE:and a complete S should be equally sensitive. Table

12, line 1 presents data from a cross involving scasc8/YFyBM/BSY males.

 

The low fertility of these males is responsible for the relatively small
numbers. However, enough flies were recovered to demonstrate that YLXBM
is less sensitive than BSY; (S2 is significant at the .05 level).

A second approach, which bypasses the fertility problem, is to com—

pare the recoveries of YLXBM and BSY in two different crosses involving
a common standard, SE, scasc8/SE/YLy3M and scasc8/SE/BSY males were

generated as brothers and were crossed to y_g.SS/y_g 22 females. The
results, in lines 2 and 3 of Table 12, are that the recovery of SE rela-
tive to SE and the recovery of SES_relative to SE is 0.22. This differ-
ence is significant at the .01 level. There can be no doubt that SES_

is more sensitive than S2232.

Is it possible that the differences in S chromosome recovery observed
in these experiments are due not to length differences, but to different
overall drive levels? Since the males in Table 12 are all siblings and
since the experiments were done at the same temperature, any differences
in drive levels would have to be caused by the S chromosomes. However,

it is certainly the case that different S_chromosomes can cause different

66

. ms mu m u : IINIIJIII.
A NN u an a N Em A» w m

 

 

 

 

 

m
..m.ov mmouo sumo cu oEomoEou:o.M o>Huamaom mmoH onu ou o>fiuwmoom ouoe onu mo >uo>ooou mo oHumu o:u
ma 03mm muo>ooou N H: one .moamaow mm. N N\fl N N ou mommouo E3 a: wuss mam moans 2mg |N|u\aM\fl N N
on moaosou wmm\womqom\momqom mo mmouu m aouw mwawanwm mm moumumoow mums moama Hoououma ona
mm. mm. mam nun ommfi moH In: mmm wmm\mw\! on com
. . III III um on
Na as Nan NNsa ENN Has an a fi\ l:\ e
. . III: :III um um
on an HNH mod Hm mm 2m N:w\w m\w q
III N H III III III: III. III: III.
a u mam cam mayo m
0 N M W Mmm A? m% wmmx AWN mwx u 0 H u m
mofiumx huo>ooom mommmao Euomm

 

moan: momaom GH wmm mam 2mhqw mo moauw>fiuwmcom o>aun mo somfiummaoo NH wanmh

 

law

50

IE

to

67

levels of drive (see Chapter 2). Perhaps BSY causes more drive than SE

so that when these chromosomes are each paired with SE the relative
recovery of SES_is worse than that of SE. A test of this explanation is
to compare Srecovery in these crosses. The S segregates independently
of the S chromosomes yet responds to the same meiotic drive conditions
(Chapter 2). Thus the recovery of the S supplies a second, independent
measure of drive. The results in Table 12 provide strong evidence
against the idea that meiotic drive levels in each cross are different.
While the S chromosome recovery ratios range from .22 to .79, the S
chromosome recovery ratios range only from .38 to .46. Pairwise contin-
gency tests show that none of the S_chromosome recovery values is sig-
nificantly different from any of the others. Differences in S chromosome
recovery ratios can not be attributed to changes in overall drive levels.

They must be due to different sensitivities of SE_and BSY.

Relative Sensitivity_of YS and YS-YS

 

A reasonable explanation of this sensitivity difference is that de-
letion of the short arm reduces the drive sensitivity of a S chromosome.

However, the sensitivity difference could be due to different amounts of

translocated S material on BSY and YLy3M. Another way to assess the sen-

sitivity of the short arm is to compare the effects of Sf_and a chromosome

duplicated for the short arm (YS-YS) on the relative drive sensitivity

of a common standard, YLyBM. If SE_contains none of the sensitivity of

the S, the sensitivity of YLyBM relative to YS-YS will be the same as its

sensitivity relative to SE. Table 13 presents data from crosses involving

half-brother sc4sc8/SEIYLy3M and scasCS/YS-YS/YLy3M males. The recovery

 

of YLy3M relative to YS°YS is 0.61 while its recovery relative to SE is.

0.47, a highly significant difference. These experiments show that both

68

Cu

 

.monEom Mm N NNAM N N ou ommmouo mums mam monE mw.mw\aw.x mam “CATV“

 

moamsow h.w\ um om\wom om mo mommouo aouw muozuouolwam: mm moumuoaow ouoa moans Hmapoumm one

m

N

Em A m w c

as. nun Nmma ammN ---- Nam
In- Ne. mold ..... NNHN mNs
.m».as m».aw as m».m» m» ANN

 

mofiuom mwo>ooom

 

mommmao Summm

woo~

Nmofi

 

h.w ou o>fiumaom w. N van w mo moauq>fiufimdom o>fiun mo :omwummfioo mH

Zn A m m m

 

h. . um on
an a»\m» m»\m a

N II. um um
2m qs\m»\m a

 

omxuocmo Hmaumumm

mamas

69

SE_and SE are sensitive to sc4sc8-induced drive and that S£_is more sen-

S
sitive than S_, consistent with its greater length.

Drive Sensitivity_QS the Bobbed Locus

 

 

Since both SE and YLXBM carry rDNA genes, one wonders what contribu-
tion the rDNA makes to their relative sensitivities. One way to estimate
Lb+ L.-
this contribution is to compare the sensitivities of Y b and Y b
S Lb-l- Lb-
relative to S_. Since Y b is longer than Y b it should be more sen-
4 8 8 Lb +
sitive. The results of crosses involving half-sibling sc sc /S_/Y b
and scasc8[SE/YLbb- males, displayed in Table 14, bear out this expect—
ation. The sensitivity difference between YLbb+ and YLbb- is substantial;
Lb+ S L..-
the recovery of Y b relative to S__is only .45 while that of Y b rela-
tive t°.XE is .70. Deletion of the bobbed locus reduces the sensitivity
of a S chromosome. Does this result imply that the bobbed locus is the
S chromosome response function? If so, then SE (which is bb+) should be
more sensitive than YLbb-. This is clearly not the case. A more reason-

able interpretation is that the sensitivity difference between YLbb+ and

YLbb- reflects their length difference.

Sensitivity Q£.§ Free S DQplication

 

 

To determine how an.S chromosome fragment fits into this graded set

of S fragments, the sensitivity of Qp(l;f)3, an S chromosome deficient

for almost all the euchromatin was assessed versus both S§_and YS-YS

 

using BSY as standard. Four types of males were tested: sc4sc8/BSY[SE

S and schsca/BSY/

 

and scasc8/BSY/Dp(S;f)3 siblings and sc4sc8/BSY/YSéY

 

 

 

Dp(l;f)3 siblings. The results in Table 15 show that SE is less sen-

sitive than the Sp. The recovery of BSY relative to SE_is only .19 while

its recovery relative to Dp(l;f)3 is .27, a highly significant difference.

70

 

 

.moAmEmw.N\N.ou ommmouo mums mam mmAmE Nm+NfiAw\m>.x mam :\m w. x ou mmAmEow

 

 

 

 

 

 

Em A
m>\ momqum\womqom mo mommouo aouw muocuoublwAmn mm ooumuocom mama moAmE Amcumuma 65H
8. S. $2 23 ONE 3: MS .Nm+.a. \m [\wlll: as see
. . um um
ms no wmoA comm moo MONA +An. Em A w\mw\m q
oux mquw Aw w» wa mwx omNuocou Amcuoumm
mOAumm mwo>oomm mommon Summm

 

m» cu o>Auonm IaAA» mam +nnAw mo moAuA>AUAmcmm o>Auo mo comAummaou «A mAan

71

.moAmEom mm N N\fi. N N ou mama mommouo AA<

mocAA 6A mmonu cu muosuouonAm: mum N new A mooAA 6A moAma one

 

\womqom\ um um mo mmouo m scum oo>Auoo

w A

 

 

 

 

N.ms\smm\mzw.es meAaeem Nxmmflvmm

.moAma

w

 

.q mam m

.mwcAAAAm omAm mum q ocm m mocAA :A moAmE msfi .moAmE mw\>mm\3.N

 

m a um um
mam Av m\»mm\m s

VmM\mV. m7\w0m¢0m

 

 

 

w. m. um um
NAN Av m\wmm\m s

II. um um
Nmm\m»\m s
mmhuoomu Amcuoumm

 

ma manme

ou mmAmaow mAmwAvma\wum¢um\womeom mo wmouo m Eoum moumuooow .mwcAAnAm mum N mam A moawA :A mmAms osy
mm. mm. III SmOA mmm IIII IIII mom Ac IIII III
om. III mm. III mAq NNQA IIII III mm mwm III
om. mm. III mwm mwm IIII IIII Hem mm IIII III
om. IIII ¢A. IIII com IIII mqu III Am IIII mom
" u a m . mm .
o x mm Nmm m» smm a wmm m» m m» x wmmx » max max
mOAuom muo>oomm mommon Ehomm
wmm o» o>Auonm mAmMAVmQ mam «ww.m .mw mo moAuA>AuAmcom o>ANA mo somAummaoo

The

to

Cl:

SE

16

t'r

r1

72

The data also indicate that YS-YS and Dp(l;f)3 are equally sensitive.

The recovery of BSY relative to YS-YS is .28 and its recovery relative

to Dp(l;f)3 is .27. These data also provide further support for the
claim that YS'YS is more sensitive than Sf_since the two crosses of

scasc8/BSY/Dp(l;f)3 males give the same results, and the two crosses

 

involving S fragments do not.
The results presented above are consistent with the idea that the ”-

sensitivity of a sex chromosome in scl‘sc8 males is a function of its

length. In all comparisons of a shorter chromosome with a longer one,

the longer one proved to be more sensitive. This could reflect either

 

an even distribution of discrete response loci on the S_or a diffuse

response function.

S Chromosome Recovegy

 

In sc43c8/S/S_(Sandler and Braver 1954) males and in the three-sex-
chromosome genotypes tested in this study, the sex ratio is highly skewed
toward males. If the poor recovery of scasc8 (and of the S fragments) is
caused by shortage of an essential chromosome processing material, then
.S recovery should be inversely proportional to total sex chromatin in
the genome. The experiments displayed in Tables 12 and 15 provide tests
of this prediction. In Table 12, sex chromatin content ranges from one

L 3M 4 8 S L 3M
__JL_.

S and almost two st (scasc8/BSY/Y y ) to one S and one S_(sc sc /S_/Y ).

 

However, there are no significant differences in S recovery between any

two of the males. In Table 15, sc4sc8/SE/BSY and scésc8/YS-YS/BSY males

 

show identical S_recovery values despite their differences in sex chro—
matin amount. Tables 13 and 14 are not useful for this purpose because

autosomal background is not fully controlled.

73

To examine this question further, two additional experiments were
performed to compare S chromosome recovery in males with large differ-

ences in sex chromatin content. In the first experiment, scasc8/BSYKS

 

and scasc8[S/Dp(l;f)3 males were generated as brothers from a cross of

scasc8/BSY/Dp(l;f)3 males to C(l)DX[S_fema1es; In the second experiment,

 

sc4sc8/BSY[S, sc45c8[S/Dp(l;f)3, and scasc8/BSY/Dp(l;f)3 sibling males

 

 

 

were generated from a cross of sc43c8/sc43c8/Dp(l;f)3 females to 2.3/2!

 

BSY males. The results of both experiments are presented in Table 16.
In the first experiment, there is no significant difference between

scascs/BSY/S_and sc43c8/S/Dp(l;f)3 males in terms of S_recovery (.52 and

 

.49 respectively) although the differences in S recovery are large. In
the second experiment, S recovery is .43 from scasca/S/BSY, .41 from
scasc8/SfSp and .31 from scasc8/BSY/Sp males. The scascB/BSY[SE result

 

 

is significantly different from the other two but it is in the wrong
direction for demonstrating an inverse relationship between sex chromo-

some content and S_recovery.

scasc8/BSY[SE[Qp(l;f)3 Males

 

 

Another test of the competition hypothesis is to compare S recovery

in sibling males carrying either three or four sex chromosomes. When

scasc8/scasc8/Dp(l;f)3 females were crossed to y_g/BSY/Sf_males, the

 

offspring included scésc8/BSY/SE/Qp(l;f)3 males. They were crossed to

 

y g SIS/y g b_b females. The results, presented in Table 17, indicate that
no sperm carrying only sc4sc8 were recovered. The addition of a fourth
sex chromosome does not change the univalent behavior of sc45c8. Since
the S segregates randomly, its recovery can be used as an absolute measure

of meiotic drive. .K recovery from scAsc8/BSY/SE/Dp(l;f)3 males is .31,

 

not significantly different from either the .28 recovery observed for

74

onu

uA >£3 GAmAGKo hos AUAns mommouo 03u onu GA uGoquMAo mA N ooxumEG: one

.moomouo AAm GA

 

Am. mN.
As. as.
me. In-
as. me.
Nm. I---
onx mma»

An.

«N.A

raw m

m

 

mOAumm mwo>ooom

moo

cqu

«New

NO

mmo

Ace

ooMA

wwA

.N uGoEAHono GA wouuon asp A uGoeAuoaxo GA » m Gmnu omuo3 wouo>ooou

wAN

mmm

NNMA

 

momoon ahomm

0mm

New

mow

mAA

wx

m
.MM N NEE. N Nouos mvoaow

.N uGoEAuomxo GA owonu ouo mm mwGAAnAm own A uGoEAquxo GA moAmE AmGuouma one

 

 

moAmZ mn\>\wum

e

um ma >\w\mom

com wGAAAAm mo GomAuomaoo

. m, a um um
mam Av Q\>mm\w q

mammwvmmAMAmemsum N
III.I. om om
wmm\>\m q

NammAVNm\M\memsem
III.I. um um
wmm\»\m s

 

omNuOGoo AmGuouom uGoEAuomxm

0A oAan

75

.mvoEom Mm N NEH N N cu oommouo ouoa ooAmE +nn MN .mCmAvmo\mw\wmm\ AA m3 N .w

Am.

 

muo>ooom

aoEomoeouso x

 

 

oAA wNA w mm on me n
3 mflqu .AWII mINWN. m3 Mwflqn INNII
moAmz moAmEom

 

wumNDOGosm mdowoum

 

mvoz mw\mfiwmAvma\>mm\momqow Souw Nuo>ooom anomw NA oAnoH

Um Om

q

Si:

gL'.

U

76

scésca/BSY/Sf_siblings or the .34 figure observed for scésce/BSY/Dp(l;f)3

 

 

siblings (Table 15). The addition of sex chromatin to the genome of a
scésc8 male does not increase drive levels.

Although it is impossible to detect SE_in the presence of either
'SES or Dp(l;f)3, since both are SS:, some conclusions can be made about
recovery of chromosomes other than the S, First, Sf_has the best recovery
since the SE class exceeds the SQ_and SEER classes (which are indistin-
guishable) combined. Second, the recovery of SES_(43/310) is the worst.
These are exactly the relationships expected under the length dependence

hypothesis.

DISCUSSION

The major conclusion from this study is that the response of the S
chromosome to meiotic drive is not a property of a single, discrete re-
sponse function. The results are consistent with the idea that drive
response is a diffuse function of S chromatin so that the sensitivity of
a S_chromosome is proportional to its length. As a limited number of.S
fragments are available for testing, it is not possible to rule out the
idea that meiotic drive sensitivity is controlled by several discrete
loci.

These results are consistent with at least two hypotheses concerning
the effects of relative and absolute size of chromosomes on severity of
meiotic drive in scasc8 males. Under one hypothesis, only the longer of
two pairing partners is subject to recovery disruption. The shorter one
is protected by pairing along its length. The degree of recovery dis-
ruption of the longer chromosome is proportional to the length difference
between it and the shorter element. To borrow Baker and Carpenter's
(1972) colorful metaphor, each chromosome carries an array of "armed
bombs" (pairing sites) which must be defused by meiotic pairing. The
shorter of two homologs is able to defuse all its bombs because it is
fully paired. The longer homolog must be unpaired along some length and
can not be fully defused. This hypothesis implies that drive is much

milder in a genotype such as scascS/BSY/YLy3M where the two S elements

are close to the same size than in schsc8/BSY/Sf_where the two S elements

 

 

are vastly different in size. In either genotype, the unpaired S would
go through meiosis with unreacted pairing sites and would be recovered

poorly, although no worse in one genotype than the other.

77

78

4
An alternative hypothesis is that sc sc8 disrupts recovery of all
sex chromosomes, the degree of disruption being proportional to the

length of the chromosome. sc4sc8/BSY[SE_males exhibit a relatively high

 

S chromosome recovery ratio because BSY and SE are of similar sizes and

subject to similar levels of drive. The absolute recovery of BSY sperm

 

would be no better in these males than in scésc8/BSY/Sf_males.

Shortage of an essential chromosome processing material is one mech-
anism that could give rise to length-dependent chromosome recovery. If
it is assumed that each binding site for the material must be occupied
for a chromosome to be non-lethal, and if the number of binding sites is
a function of length, then the probability of at least one site being
unoccupied is proportional to length. Shortage of the material would
create competition between chromosomes. One consequence of competition
would be the more sex chromatin present in a spermatocyte, the less
processing material is available to each chromosome. Drive levels would
be proportional to the amount of sex chromatin in the genome.

Most of the data, and more important, the best data, are inconsis-

tent with this prediction. S recovery is the same in scésc8/BSY/SE,

scasca/SEfSE, and scascB/SE/BSY siblings (Table 12) even though total

 

sex chromatin content (and S chromosome recovery ratios) differ consid-

erably among the three genotypes. scasc8/S/BSY males have the same S re-

covery as scasCB/YSNS/BSY males (Table 15). These males are only half-

 

brothers but results from their scascB/BSY/Dp(l;f)3 full brothers showed

 

that no autosomal modifiers of any importance were segregating in these

4 8
crosses. In two experiments comparing sibling sc4sc8/S/BSY and sc sc [S/
Dp(l;f)3 males, the only significant difference in_S recovery was in the

wrong direction (Table 16). Finally, addition of a fourth sex chromosome

79

to the genotype to make sc43c8/BSY/SE/Dp(l;f)3 males does not affect S

 

chromosome recovery (Table 17). These results imply that sex chromosomes
do not compete for a scarce resource in sc4sc8 males. If they did, one
would expect to find an inverse relationship between sex chromatin con-
tent and meiotic drive levels. No such relationship exists.

These results are consistent with the armed bomb model. If the
poor recovery of the S chromosome in these males is due to its failure
to pair, one would not expect the amount of S_chromatin to matter. .§
recovery should be a constant, or should vary unsystematically with,
changes in S chromosome lengths. This is what is found. The major dif-
ficulty with this view is that it calls for pairing interactions between
parts of chromosomes that do not appear, under the microscope, to be
involved in meiotic pairing. The poor recovery of the S.in sc4sc8/Sfyp
males implies that the free duplication does not have enough pairing
sites to defuse all (or even most) of the S chromosome's bombs. Since
a normal S does, this implies that the remaining pairing sites must be
euchromatic. Yet no one has ever observed any involvement of the S

euchromatin in S:S meiotic pairing.

CHAPTER 4

SENSITIVITY 93 AUTOSOMES _'I_:_g MEIOTIC DRIVE

 

 

In Drosophila melanggaster males, the S_heterochromatic deficiency

In(l)sc4Lsc8R (scA

 

sc8) disrupts recovery of sex chromosomes by causing

 

sperm dysfunction (Gershenson 1933; Sandler and Braver 1954). Both the
S chromosome (in scasc8/S and scésc8/S/Sp) and the S (in sc45c8/S/Sp and
scASca/SfS) are subject to this recovery disruption (Gershenson 1933;
Sandler and Braver 1954; Peacock, Miklos, and Goodchild 1975; Haemer
1978). No evidence exists as to whether or not recovery of autosomes is
also disrupted. The question addressed in the following study is: does
scasc8 disrupt autosomal recovery?

Since all sperm normally have the same autosomal content, they would
be equally sensitive to length dependent recovery disruption in ordinary
scasc8 males. To detect autosomal sensitivity, a scasc8 genotype which
generates sperm with unequal amounts of autosomal chromatin was con—
structed. Males heterozygous for a reciprocal translocation between the
major autosomes with one break near the tip of chromosome S and the
other in the centric heterochromatin of chromosome_S generate sperm
carrying different amounts of autosomal chromatin. If autosomes are sen-
sitive to scascs-induced meiotic drive, recovery of sperm classes from
scl‘sc8 males carrying such a translocation should be inversely propor—
tional to their autosomal chromatin content. The results described below
indicate that this is the case. The implication is that S heterochromatic
deficiencies disrupt a developmental process affecting all chromosomes.

80 .

 

MATERIALS AND METHODS

T(2;3)bwV4 (Lindsley and Grell 1968) has one break near the tip of

 

chromosome S in the vicinity of the brown locus and a second break in the
centric heterochromatin of chromosome S. One half-translocation,
2L-2R3L, is unusually long, consisting of the equivalent of three auto-
some arms. The other half-translocation, SE, is unusually short. (For
simplicity the translocated tip of SE is neglected in this discussion --

it is too small to have any significant effect). Expected patterns of

chromosome segregation in T(2;3)bwv4 heterozygotes are illustrated in

 

Figure 1. Alternate segregations produce two types of euploid gametes,
either normal sequence or translocated. Adjacent I segregations produce
2L'2R3L;S (duplication) gametes and S;SE_(deficiency) gametes. Adjacent
II segregations, which should be relatively infrequent or perhaps non-
exiStent (Glass 1933; Glass 1935; Roberts 1976) generate 2L-2R3L;S and
‘SE;S gametes. The aneuploid gametes generated by adjacent segregations

lead to viable progeny only when they combine with reciprocal aneuploid

gametes from the other sex.

81

82

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mouowmnououom «>3aamwwva GA GOAquGnoAn on wGAuAmm N ouswAm
nAAGAGGAAAA. .. _ rHHHHmAAAAA GA _ A_AM_AMMMLL IAH
mI AA _ HHHHHHHHHHHHHHHHHHHHU HuuuuuuuuuuHu
HHHHHHHHHHHHU HHHHHHHHHHHHu HELLLELMAEELMMLLLuLLL
nHHHHHHHHHHHHHHHHHHHu w d. IL L GrI _
GOAumwouwom AA uGoomfimG GOAuomowmom A uGoomfio< GOAumwouwom oquuouA<
G. L
HHHHHHHHHHMH __ h?. a

 

 

GOAuouzwAmGou wGAuAmm

RESULTS

 

Absence 9S Adjacent SS Segregations

A preliminary question of some importance is whether or not adjacent

II segregants are recovered from males heterozygous for T(2;3)wa4.

 

Glass found very few or none (1933; 1935). To corroborate his finding,

T(2;3)bwv4 was marked with Star (S, an eye texture mutant), and males

 

heterozygous for T(2;3)bwv4, S and E(S) (Enhancer of Star, a dominant

 

second chromosome mutant that strongly enhances the Star phenotype) were

crossed to females heterozygous for the unmarked T(2;3)wa4. The

 

recovery of S[S£S) progeny (distinguished by very small, rough eyes)
would signal the occurrence of adjacent II segregation since they must
come from 2L'2R3L;S sperm (see Figure 1). No S/SSSl_progeny were found
among 345 offspring, implying that adjacent II segregation does not occur

in at least one sex. All aneuploid gametes derive from adjacent I

segregations.

Sperm Recovery from T(2;3)bwV4 Males

 

When males and females heterozygous for T(2;3)bwv4 are crossed, the

 

progeny will include individuals derived from sperm with three, four, or
five major autosomal arms. If sc45c8 disrupts autosomal recovery it
should bias recovery ratios in favor of sperm classes with the least
amount of autosomal material. Males carrying scasc8 or a normal sequence

S and T(2;3)bwV4 or non-translocated autosomes were generated as siblings

 

+
by the mating scheme diagrammed in Figure 3. This produced sc4sc8/y Y

 

T(2;3)wa4/SS_males and controls (y/y+Y T(2;3)wa4/SS, scasca/ny TM2[S£,

 

 

and y/y+Y TM2[S£) as siblings. These males were then crossed in single

pair matings to yfy T(2;3)wa4/SM1,. 1 females. Figure 4A is a diagram

 

+
of the cross for y/y+Y T(2;3)bwv4 males. A diagram for sc4sc8/y_Y

 

83

84

 

 

 

 

 

 

 

 

FM7 X y_ S l 2
+ + +
y Y y l
4 8 V V4
yfi X sc sc y SMl TM2 X :T(;;3)bw
yD BY y+Y + + +SM1;+
V V
sc sc8 .Sg X + T(2;3)bwv4
+ + +; TM2
y Y
V
sc43c8 T(2;3)bwv4
+ +; St
y Y
sc4sc8 TM2
+ St
3' Y v4
V4 X y_ T(2;3)bw
y T(2;3)bw y SMl; +
y+Y +; St
y TM2
+ 3!:
y Y

Figure 3 Crosses to Generate scl‘sc8;T(2;3)wa4 Males and Controls

85

Egg Classes

 

 

 

 

 

Sperm Classes y;SMl;+ y;2L'2R3L;+ y;SMl,’ 3R y;2L°2R3L;3R
y;:;S§ m Females Lb‘LSJi Females
1;2L-2R3L; _S_t >502wa: Females

l3i33_R .Lbfl Females

y;2 -2R3L;3_R 1%”. Females yb_w Females
11S;:;_5 SE Males SILSS Males
y+Y;2 '2R3L;fi Cywat Males

flsizi PE Males

£352 -2R3L;3_R gm Males b_w Males

y/y+Y;T(S;31wa4/:;SE males were crossed to y/y;T(2;3)bww‘/SM1;i
Females I

 

 

Figure 4A Diaggam of Gross between Two Translocation Heterozygotes

 

 

 

 

 

 

Egg7Classes
Sperm Classes y;SMl;+ y;2L'2R3L,°+ y;SMl,'3R y;2L'2R LQR
y_;i;S_M_S yCybe Females ybwax Females
y;:;_S_t_ ygySE Females m Females
fl;:;_TM_2 91% Males SILUSDS Males
13%;; SySE Males SESS Males

y/y+Y:TM2/_S_'t males were crossed to y/y;T(2;3)wa4/SM1 females.

 

Figure 4B Diagram of Control Cross

 

 

86

T(2;3)bwV4 males would be similar, but twice as large since SS_and nullogg

 

nullojS sperm are generated as well as X and Y sperm. The autosomal com-
binations are identical in the two crosses. A diagram of the control
cross in which the males carry SSS (instead of the translocation) and a
normal S appears in Figure 4B. Once again, the scasc8 cross is the same
except that SS and nullojS, nullofS_classes are present as well as S and S.

The data from these crosses appear in Tables 18 and 20. Table 18
presents the data for the translocation males, both scl‘sc8 and regular S,
and Table 20 presents the non-translocation male data.

Several conclusions are evident from the data in Table 18. The
first is that alternate segregants outnumber adjacent segregants in both
crosses; by 1925 to 458 in the normal S cross and by 1275 to 443 in the
sc4sc8 cross. The second is that in the normal S cross reciprocal prod-
ucts are recovered in the expected ratios for both types of segregation.
The expected ratio among the alternate (euploid) segregants is two normal
sequence sperm to one translocation sperm. 1925 normal sequence sperm
and 952 translocation sperm were recovered. Among the adjacent segre-
gants, the expected ratio is one 2L-2R3L;S to one S5SE sperm. The obser-
vations were 213 and 245 respectively.

If translocation homozygotes survive, they would be indistinguish-
able from the deficiency class in this cross. Since Glass (1933) found
homozygotes to be weakly viable, this possibility was examined in a
separate cross with appropriately marked autosomes. No translocation

homozygotes were recovered out ofJJQS flies. Thus the viability of

T(2;3)bwv4 homozygotes is zero in these crosses.

 

4 8 -
In the sc sc cross, reciprocal products of alternate segregation

were recovered in the expected two to one ratio: 1478 normal sequence

87

Table 18 Results from Crosses between T(2;3)bwv4 Heterozygotes

Paternal Genotype

 

 

 

 

 

 

 

 

Progeny
W Sperm 8.43.8 T(2;3)bw“ y T(2;3)bw“
Females Genotype y+Y +; St y+Y +; St

1 m it 5:1: _t 316 457
1 91 .31; §;::_t_ 279 456
1 2 9x 5:2 ~2R3L;_3_R_ 315 488
l 13.31 25:32:33 158 122
1 3E 91 fl £s2L-2R;__t_ 42 110
.12! §£ _Xl;:;_t_ 19 --—
91 it. 3:5; 31 -_-
.122 9): E;2L°2R3L;_3_E 27 ---
m and: 14 -—-
RE £31 £3 fl;2L-2R3L,g 5 -__
£3.13

1 Es S_t i;_t_ 243 ---
z 91 _t i;_t 256 ---
l b_w EX 2 .-2R3L;3_R 38 ---
3 la :;3_R 124 ---
1 bw 91 St 2 2R3L,_t 2 -__
1’! .311. X;:;S__ 156 516
91 .51 31;; 178 496
Is. 92 3.2L-2R3L;_3E . 161 464
*1? 131:3 85 123
S! 91 _S_t 32 ~2R3L;§_g 13 103

The females in both crosses were y/x;T(2;3)wa4/SM1, 91; :

 

 

88

sperm to 753 translocation sperm. But reciprocal products of adjacent I
segregation were not recovered equally. Summed over all sex chromosome
classes, 381 g,§§ and 62 2L-2R3L13 sperm were recovered. This six to one
ratio contrasts sharply with the one to one ratio obtained in the normal g
cross.

The ratios of deficiency to euploid classes and of duplication to
euploid classes are also affected by scasca. In the control these ratios f
are both approximately one to four in the offspring (using the euploid

translocation class as denominator), which implies a one to two ratio in

the gametes of both sexes (assuming equal nondisjunction in both sexes).

 

In the presence of scasc8, the deficiency to euploid ratio in the off-
spring is one to two which implies a one to one ratio in the sperm
(assuming that the egg ratio stays at one to two). The duplication to
euploid ratio in scl'sc8 is one to nine in the progeny (using only the K
and X_classes, for reasons described below) which corresponds to a one to
four or five sperm ratio. Relative to the euploid class, sc43c8
increases the viability of deficiency sperm and decreases the viability
of duplication sperm. This implies that in sc4sc8 males the probability
of recovery of a sperm is inversely proportional to the amount of auto-
somal chromatin it contains.

A particularly striking feature of these data is the absence of
interaction between autosomal and sex chromosomal meiotic drive. Despite
the gross violations of Mendelian ratios, autosomal and sex chromosomal
recovery ratios are mutually independent. This is illustrated in Table
19 which presents the data in an orthogonal array. Note that, with two
exceptions in the nullofifi, nullofiX_class (discussed below), the expected

number in each class (assuming independence) agrees remarkably well with

89

madam

mmma

mmm

moo

cm

QHHH

 

mESm

!. Iain-kl. ......

I... 3.15-5“

.uxmu onu ca vmmmsomfiv mEmHnoua hufiafinmfi> mcu mo wmsmomn

.Mroaasc .mroaasa asexuwz woumaaoamo mum Ammmmsuamuma Gav mm=Hm> vmuomaxo van meow

 

 

 

 

 

or 5mm mom amm
Aomv ma Amwv mm Aooav HOH Ammmv «mm

N «NH mm amc
Amy m Aqav «a Aamv RN Ammo on
Aan we Ammav me Aoamv mam Aqoov mom

9 . m Ila . m
um.nmmN AN mm + «m umMN am um +
moan: >BnAmm~vaum om Eoum.muasmmm mo cofiumuammmum Hmaowonuuo

c

w c

mE:m

 

 

mmxuocmo
msomosouso xmm
Ehoam

ma anmH

90

the observed number.

The two exceptions mentioned above are the nullojg, nullofz5
2L-2R3L;§E and nullojg, nullofiz52L-2R3L18 sperm which both show very
poor recovery. The few flies derived from these sperm which did survive
were late-hatching, thin-bristled, and tended to get stuck in the food.
A plausible explanation is partial dominant lethality due to variegation
of the paternally transmitted 2L-2R3L element in §9 males. The pheno-
type suggests Minute and there is a strong Minute locus at 282, just a
few bands proximal to the breakpoint. Variegation is implied by the
fact that the recovery of 2L°2R3L sperm is abnormal only in the g9
males. It is interesting that the same element shows normal recovery
when transmitted from the mother. This is an apparent example of the
parental source effect (discussed by Spofford 1976).

L R L

To test this explanation, nullojg, nullofz52 '2 3 males were gen-

erated by another route. scbsc8/y;T(2;3)wa4/SM1 females were crossed

 

to y/y+Y;T(2;3)bWVA[§t_males. E chromosome four-strand double exchanges

 

generate nullofg eggs which, when fertilized byig sperm, give rise to
§Q_males. Some of these males carry a paternal ZL-ZRSL element and
their recovery is depressed. No nullojg, nullofX_progeny with a pater-
nal 2L'2R3L chromosome were recovered out of 55 nullofg, nullofz_males.
In the other sex chromosome classes, recovery of the paternally trans-

mitted 2L-2R3L was normal.

Autosomal Modification of Drive and Nondisjunction

 

 

These experiments provide information relative to a second issue:
the modifying effect of different autosomes on nondisjunction and
meiotic drive in scl‘sc8 males. The raw data from the cross of

4 + ,
sc sc8/y Y;TM2/§£ are presented, along with normal_§ controls, in Table

 

91

20. From the control it can be seen that no genotype is associated with
particularly poor viability. The experimental cross is quite unremarka
able. Nondisjunction and drive are both relatively high. There is no
apparent tendency toward greater recovery of either TM; or its §£7
bearing homolog.

A comparison of sperm recovery from sc43c8/y+Y;T(2;3)/§£_and

 

scésc8/y+Y;TM2/§£_males is presented in Table 21. Nondisjunction is
higher and chromosome recovery values lower in the TM2/§£_ma1es. The
difference in nondisjunction values is exaggerated by the inviability of
nullofig, nullofz,2L'2R3L sperm. Assuming frequencies of these sperm
similar to their frequencies in the x and X_classes, nondisjunction in
T(253) males is .45, still substantially lower than the .57 value given
by the TME[§£ males.

Since these males have different second and third chromosomes, it
is unclear whether the differences in nondisjunction and drive should be
attributed to the translocation, to the balancer, to the unmarked second
chromosome, or to some combination. Ramel (1988) failed to find any
"interchromosomal effect" of the Curly inversions on scésc8-induced
drive or nondisjunction. It is possible that a third chromosome
balancer would behave differently. In any case, this is a clear example

of autosomal modification of meiotic drive.

-. .21.. HUFJIW'!‘

 

92

 

 

 

 

Table 20 Results from Control Crosses
Progeny Paternal Genotype
Phenotype sperm SCASCS TM2 _1L_ IE2.
Females Genotype y+Y St y+Y St

1 2‘1 S_t i, it; 69 132
x b_w Ea .11. 114.1 71 159
1 92 a; L at; 64 162
z 91 111).: L £2. 51 162
22 S_t LY; £2 5 ---
bw a a. 1112 o
91 .95 a; at. 5 ---
Ex _31 _XL 1“; 6 “-
Males

1 in; St §_§ lOl ---
1 fl be _Tfl 77 ---
y 91 St §£ 72 ---
y EX be _mg 87 ---
b! _S_t_ I: S_t 25 144
_b_w_ U_bx X; 1M3 40 128
El it 1, _S_t_ 20 146
£21 1b}; 1; T_M£ 40 140

The females in both crosses were y/X;T(2;3)wa4/SM1,

3 i

P

 

 

93

mo. me.

«H. mm.
onyx xuw
mowumm
>um>oumm

mm.
mm.
o+x uo

cowuocsn
unannoz

mNH

mmm

V

 

.+ m U .H2m\q>3nammmvfi «N\N wumS mummouo zuon CH mwamamw 0:8

 

 

 

MI. m h um um
amm 0H mmN m\~zs »+ \w s
ml. 3 m. m m um on
new om OHHH m\s> can Noe »+ \w s
o MN x mmxuocmu Hmcumumm

 

mommwao summw

 

q

>Bnamw~VH usosuwz cam pugs moan: um um ca m>wua can coauoaanmaccoz Hm manme

w e

DISCUSSION

The main conclusion of this study is that sperm recovery in sc4sc8
males is inversely proportional to autosomal chromatin content. The
ratio of autosomally deficient sperm to euploid sperm is twice as high
in sc43c8 as in controls. Similarly, the ratio of euploid to auto-
somally duplicated sperm is better than twice as high in sc4sc8. The
simplest explanation of these results is that scésc8 alters sperm via-
bility to favor recovery of sperm with less chromatin. One implication
of this result is that recovery of $5 autosomally euploid and of
nullojg, nullojz; autosomally euploid sperm from scasc8 males is less
than perfect.

The results from this study are also relevant to the question of
the role of pairing in the production of normal sperm. Following the
suggestions of Baker and Carpenter (1972) and Peacock and Miklos (1973),
sc4sc8 drive might result from mismatch of sex chromosome pairing part—
ners. The extra unreacted pairing sites on the longer pairing partner
act as "armed bombs" and destroy the sperm that contain them. This
hypothesis can account for length dependence of sex chromosome sensitiv-
ity (discussed in Chapter 3) but does not explain the failure of free x
duplications to improve X_chromosome recovery. It also does not account
for the data presented in this chapter. If sc4sc8 acts by interfering
with proper xi: pairing, there is no reason to expect autosomal involve-
ment at all. While a translocation between the second and third chromo-
somes might weaken local pairing and perhaps expose some autosomal pair—
ing sites, why should this effect manifest itself only in the presence
of scéscg? No hint of autosomal pairing difficulties can be seen in nor—

mal E males. These results imply that the X chromosome deficiency

94

 

95

affects autosomes in some fashion not related to pairing.

The evidence for independence of sex chromosome and autosome sen-
sitivity is interesting in light of the earlier discussion (Chapter 3)
of competitive chromosome interactions. If scl‘sc8 causes a shortage of
an essential chromosome processing material, then chromosomes must
compete for adequate supplies of it. Whether competition takes place
in the spermatocyte prior to anaphase or in spermatids after meiosis, one Q?
would expect that whenever both sex chromosomes succeed in garnering E
enough of the material, less would remain to be divided among the auto-

somes . Autosomally duplicated sperm would be less frequent among the

 

§X_class than among the §_or‘z_classes. Similarly, the frequency of .i
autosomally duplicated or even euploid sperm among the E or X_class
should be lower than among the nullofg, nullofX_class. Autosome recovery
frequencies should be conditional upon sex chromosome genotype and vice
versa. This is not the case. Autosomal recovery frequencies are the
same across all sex chromosome genotypes and sex chromosome recoveries
are the same across all autosomal genotypes.

The absence of competition among chromosomes in sc43c8 males implies
that sperm defects are not caused by shortage of a chromosome processing
material. An alternative is that sc4sc8 reduces the time available for
a key meiotic process. The likelihood of a chromosome completing the
step would be inversely proportional to its length. No competition would
ensue because the scarce quantity, time, can not be sequestered.

Frequent references in this chapter and the previous one have been
made to "chromosome sensitivity" as if it had been demonstrated that the
sc4sc8 deficiency actually alters the state of chromatin in some physical

way. This has not been demonstrated. It has been shown that the

96

recovery of sperm from sc43c8 males depends inversely on their chromatin
content, while sperm recovery from normal males is independent of chroma-
tin content. This implies either that the chromatin has been altered by
sc4sc8 to become sperm-lethal, or that some other aspect of sperm anatomy
or physiology has been altered to render the sperm sensitive to perfectly
normal chromatin. One example of the "insensitive sperm" type of expla—
nation would be reduced motility. Perhaps sperm from scésc8 males are '1}
unusually sluggish. Such sperm might be less sluggish when carrying a
lighter than normal nucleus and more sluggish when carrying extra nuclear

weight.

 

Finally, it is interesting to compare these results with the results Jr?
of investigations into the interactions between sex chromosomes and the
S2 autosomal meiotic drive systim. Sperm bearing the homolog of the SQ
chromosome in heterozygous males frequently fail to function so that as
many as 99% of the heterozygote's offspring inherit the SQ chromosome
(Sandler, Hiraizumi and Sandler 1959; Hartl, Hiraizumi and Grow 1967;
Tokuyasu, Peacock and Hardy 1977). Most of the SQ: survivors are females
(Hiraizumi and Nakazima 1967). An attachedfgz chromosome is recovered
much better than a Z_chromosome among SQ: offspring of attached-EZ/Xj
§dj§d: males. An even more pronounced effect is observed with attached-
'§X[gr§d[§d: males (Denell and Miklos 1971). It was concluded that dif-
ferent sex chromosomes can be ranked in order of recovery probability
among SQ: offspring of §d_heterozygotes. The order, from best to worst,
is attachedrzz, x, X, nullojg, nullofX. The amount of sex chromatin
affects the viability of these sperm; the more sex chromatin, the better
the viability. It is curious that in both drive systems the viability of

sperm should prove to depend dramatically on chromatin content, but in

97

opposite directions -- positively in S2, negatively in scasca.

 

CHAPTER 5

.E HETEROCHROMATIC DUPLICATIONS AND MEIOTIC DRIVE

In Drosophila melanogaster males, normal §:Z_pairing depends on

 

interactions between discrete pairing sites (collochores) located in the
g heterochromatin and in the centromeric regions of both arms of the Y
(COOper 1964). At metaphase, the §j1_bivalent appears to be held to-
gether by stringy material connecting one x site to one Z_site although
it is possible that prophase interactions involve more of the sites.
Deficiency for substantial amounts of §_heterochromatin causes frequent
£51 nondisjunction and dysfunction of both Z_and 5: sperm (Gershenson
1933; Sandler and Braver 1954; Peacock, Miklos, and Goodchild 1975).
Several investigators have suggested that sperm dysfunction is a direct
consequence of the pairing site deficiency and that normal §:X_pairing
is essential not only to insure disjunction but also to permit proper
chromosome processing for spermiogenesis (Baker and Carpenter 1972;
Peacock and Miklos 1973). For a sex chromosome to be correctly processed,
its pairing sites must interact with the pairing sites of its homolog
during meiosis. Non-interacted pairing sites become gametic lethals.
This hypothesis accounts for the recovery disruption of the X chro-
mosome from males carrying a heterochromatically deleted §_(such as

4L 8R 4 8
ac

In(l)sc (sc sc ) and two X_chromosomes. In these males, the two

 

Zfs pair and disjoin regularly from each other leaving the x without a
pairing partner. Although it has but few pairing sites left, the defi-
cient x evidently retains some pairing capacity since it can pair with a

98

99

single X_(Cooper 1964; Peacock 1965). Its complete failure to pair in
‘zzx_males should, then, have negative consequences for gfbearing sperm
viability. It does: .xg sperm are recovered less than half as frequently
as X_sperm (Sandler and Braver 1954).

If unreacted pairing sites are responsible for the skewed segre-
gation ratios in these deficiencyfig males, then other genotypes sharing
this pairing site asymmetry but not deficient for x heterochromatin
should also exhibit aberrant segregation. For example, an‘g or z_with
double the normal dose of pairing sites should complete meiosis with

unreacted pairing sites which would act as gametic lethals. The experi-

 

ments described in this report test that prediction by examining the

meiotic consequences of duplication for the g heterochromatin.

MATERIALS AND METHODS

Two g chromosomes duplicated for the pairing sites were selected
for these experiments. Both have one dose of heterochromatin in its
normal position adjacent to the centromere and a second dose near the

tip. One, In(l)scBLsc4R (scssc4) is the reciprocal product of the

 

recombination event that generated sc43c8, the heterochromatically
deficient §_that causes poor Z_recovery (Gershenson 1933; Sandler and

SlLscAR (SCSISCA), is of similar origin

Braver 1954). The other, In(l)sc
and structure. Figure 1 illustrates the structure and genesis of these

chromosomes.

 

The euchromatic breaks in In(l)sc8 and In(l)sc4 are to the left
and right, respectively, of the acute locus. The recombinant sc88c4
is deficient for that essential locus (Muller and Prokofyeva 1935)
and is, therefore, inviable in males unless a scute duplication is pres-
ent elsewhere in the genome. In one cross, the scute duplication was
carried on the fourth chromosome; in the others, it was part of a free
z duplication. Three different free §_duplications were tested: Dp(l;f)3,

Dp(l;f)856, and Dp(l;f)ll73. The euchromatic breaks in In(l)sc81

 

 

coincide, so males carrying scSlsc4 are viable without a duplication.

.To test for meiotic drive, males carrying one of the duplicated gfs,
a marked Z_(§EX), and, in the case of sc83c4, a scute duplication, were
crossed to normal females and the recovery of the §_was monitored. Two'
tests for zygotic lethality of scssca were performed. The first was an

egg hatch determination in a cross involving sc83c4 males. sc83c4/BSY/

 

Dp(1;f)3 and control Oregon R_ma1es were placed singly in vials with one
female each. Flies were transferred to fresh food every 12 hours and the

100

101

A. An §_chromosome with scute inversion breakpoints represented.

4

L. r.

31 S°31
SC SC

 

 

 

 

B. Pairing and Exchange between Insgl)s-c4 and sc8

In(l)sc4

{31 l0
1 Do

In(l)sc8

 

 

 

 

 

 

 

 

 

C. The Products of exchange between Ins(l)sc4 and sc8

In(l)sc8Lsc4R

 

 

 

 

 

 

 

 

 

 

 

{1L {10

In(l)sc4Lsc8R

 

The thin lines represent euchromatin and the wide areas represent
heterochromatin.

Figure 5 Origin of Heterochromatically Duplicated and Deficient X's

 

102

eggs in the old vials were counted. The egg hatch was the number of
adults divided by the number of eggs. In the second experiment, the
recovery of scasc4 relative to a normal §_was monitiored after trans-

mission from a female. sc83c4/y_females were crossed to y/BSY males.

RESULTS

81 4
Recovery gf_sc sc

If duplication for g heterochromatic pairing sites causes gametic
lethality, more sons than daughters should be recovered from males carry-
ing SCSISC4 and a normal X, When males carrying sc513c4 and BSY were

crossed to Oregon 3 or to y 3 females (Table 22) the sex ratio was nor-

 

5%-
mal and there were no indications of other meiotic anomalies. Therefore, 4
the §_heterochromatic duplication in scSlsc4 does not cause skewed §:X
segregation.
8 4 E
Recovery of SC SC

When males carrying scssca, BSY, and one of several scute dupli-

 

cations were crossed to XL! females, §_chromosome recovery was poor: the
sex ratio ranged from .45 to .63 (Table 23). Although a skewed sex ratio
might indicate segregation distortion, it might also indicate post-
fertilization lethality. The scute region deficiency in scasc4 could
cause dominant lethality. One feature of the data in Table 23 supports
the latter interpretation. In every cross, despite the poor recovery of
scssc4 and 8C88C4[X sperm relative to their reciprocal products, the
recovery of sc83c4/Dp_sperm equals that of §§X_sperm. The scute dupli—
cation suppresses the lethality of scascé. Since gamete phase lethality
suppression is unknown, the most plausible interpretation is that female
zygotes with one dose of the scute region frequently die. If this inter-
pretation is correct, then the equal recovery of sc8sc4[Qp sperm and
‘BEX sperm implies that sc83c4 does not cause segregation distortion.
Three tests of this hypothesis were performed. In one, egg hatch

measurements were made for crosses of sc83c4/BSY/Dp(l;f)3 males and

 

103

104

Table 22 Recovery of scSlsc4

 

 

 

 

Genotype of Progeny, Sex Ratio
Maternal Genotype sc513c4/X .§£§EX_ Females:Males
y Ely g 2153 1903 1.13
:l: 2205 2060 1.07
In(l)scSISCA/BSY males were crossed to the females listed in

 

the first column.

105

 

mo. moo
me. men
on. man
mq. How
moamzumoamsom mmwmm
ommmm1mmw

QwNH
moo

ooOH
mo

NNN mom
mam com
New Ham
cha «as
wwmu .mmm:

 

mommmao aboaw

.N N\N N 0.33 moamamw 93 .3330 :m cH

«ma

Nm

oNH

V mx

wNH

an

NNH

5mm

 

msowumowaadm ousom m:owum>.wdawuuwo moan: Baum

e

um um mo mum>oomm

 

I. on m m m um um
+\m aw av n wmm\q w

 

m 1m on on
mnaamm av n\wmm\q m

 

m m. um um
mmwAm Av Q\wmm\c m

 

m. m1, um on
«AM av o\»mm\e w

 

mmxuocmu HmGHMumm

MN manna

106

Oregon R controls to yjy_females. The results indicated significant
viability depression by sc8$c4. Egg hatch was 36% in the scssc4 cross
and 61% in the control cross; the difference is statistically signifi-
cant .

The second test was a measurement of scssc4 recovery following fe-
male transmission. sc83c4[y females were crossed to y[§EZ_males. sessc4
was recovered poorly: there were only 385 sc85c4ly daughters compared
to 1513 y/y daughters.

The third test was a determination of the efficacy of a female-
transmitted scute duplication in rescuing a male-transmitted sc83c4.
Since sc4sc8 is the reciprocal crossover product of sc85c4, it must

be duplicated for the scute region. Therefore, sc83c4/3c4sc8 females

 

have two doses of scute and should exhibit normal viability. When

sc83c4/BSY/Dp(l;f)3 males were crossed to sc4 4

recovery of sc88c4/sc43c8 daughters was normal. Thus sc8sc4 causes

sc8/sc sc8[z_females, the

 

 

 

partial post-fertilization dominant lethality because of the scute

region deficiency. It does not cause segregation distortion in males.

DISCUSSION

The major conclusion from these experiments is that despite the
presence of extra pairing sites, the heterochromatically duplicated é
chromosomes sc8sc4 and scSlsc4 do not induce meiotic drive or other
meiotic anomalies. This contradicts the notion that pairing site
symmetry of x and X chromosomes is necessary for proper meiotic chromo-
some processing. These results alone do not necessarily rule out the
idea that pairing is required for chromosome processing. One might argue
that the ability of x chromosomes carrying heterochromatin both distally
and proximally to form hunachromosomal loops (Cooper 1964) would enable
all the x pairing sites to participate in meiotic pairing. However, two
other findings argue against the idea that pairing site interactions are
crucial for chromosome processing. One is that addition of an §_chromo-
some deficient for almost all the euchromatin but carrying a full dose
of heterochromatin to a scésc8/X genotype does not mitigate the severity
of X chromosomal recovery disruption. The Z_disjoins regularly from the
free E duplication but is recovered less than half as frequently (Haemer
1978; chapter 1). Since the free duplication carries a full dose of
heterochromatin and since it pairs regularly with the Z, one would expect
normal pairing site interactions and normal Z_recovery. Second, the
probability of recovery of a sperm from a sc4308 male is inversely pro-
portional to the amount of chromatin it carries. Autosomal chromatin is
as damaging to sperm viability as is sex chromatin (Chapters 3 and 4).

It is difficult to see why the presence of unreacted Y chromosomal pairing
sites should affect the recovery of autosomes.

An alternative interpretation of these data is that the absence of

an x heterochromatic function from its normal position -- cis to the

107

 

108

euchromatic genes -- is responsible for the skewed segregation ratios.

in sc43c8 males. Meiotic drive can not be an ordinary deficiency pheno-
type since the presence of z heterochromatin in the form of a free §
duplication does not suppress it. Recovery disruptions are at least as
severe in scasc8/X/Dp_males as in sc4sc8[z_males. In both these males,
and in sc48c8[zfz_males, the §_heterochromatin is either absent alto-
gether or separated from the §_euchromatin. Perhaps an intact x chromo—
some is a prerequisite for normal spermiogenesis. If so, then other

types of rearrangements that violate the integrity of the §_shou1d dis-
rupt spermiogenesis. Many translocations that separate the bulk of the
euchromatin from the heterochromatin cause either meiotic drive or male
sterility. Two $13 translocations with proximal euchromatic g breakpoints
cause meiotic drive. Despite regular bivalent pairing and disjunction,
the longer member of each bivalent (the X_and fizz?) exhibits depressed
recovery (Novitski and Sandler 1957; Zimmering 1960). MOSt translocations
involving the g and one of the major autosomes (chromosomes 2_or 3) cause
sterility (Lifschytz and Lindsley 1972).

This sterility shares several features with the meiotic drive sys—
tems previously discussed. First, it affects only males. Second, it
involves production of nonfunctional sperm. Third, it is dominant:
addition of a duplication covering the region of the §_breakpoint does
not restore fertility. Fourth, male sterility, like meiotic drive, seems
to result from separation of the bulk of the §_euchromatic genes from a
heterochromatic locus. Translocations with both breakpoints near the
tips and translocations with the g breakpoint in the proximal part of
the heterochromatin do not cause sterility. All other translocations,

including those with X breaks in the distal part of the heterochromatin,

 

109

cause sterility. Fifth, sex chromosome meiotic drive systems all cause
partial sterility. T(1;4)BS (Novitski 1970) and sc4sc8[IfX_(Sandler and
Braver 1954; Chapter 2) males are only weakly fertile. scasc8[z_males
are moderately fertile but exhibit unusually early onset of sterility
(Peacock and Miklos 1973). Sixth, some g heterochromatic deficiencies
cause male sterility and others cause meiotic drive. Deficiencies en-
compassing both the heterochromatic bobbed locus and the proximal euchro-
matic suppressor of forked (gggfl) locus, are sterile even in the pres-
ence of a X_chromosome covering the deficient region (Lifschytz and
Lindsley 1972). Deficiencies encompassing bobbed but not suppressor of
forked, like sc48c8, cause meiotic drive even in the presence of a dupli-
cation covering the deficiency (Gershenson 1933; Peacock and Miklos 1973;
Yamamoto and Miklos 1977). There seems to be a complex locus or group

of loci in the distal section of the §_heterochromatin complete defi-
ciencies for which cause male sterility and partial deficiencies for
which cause meiotic drive. Finally, translocations between the X_and

one of the major autosomes are fertile in males carrying a normal §

but sterile in males carrying an §_deficient for a substantial section

of the basal heterochromatin (Lindsley, unpublished).

In light of the many similarities between sex chromosome meiotic
drive and dominant chromosomal male sterility, it is proposed here that
meiotic drive is a consequence of partial dominant chromosomal male
sterility. Perhaps partially defective sperm are sensitive to the
amount of chromatin they contain. Alternatively, sperm might malfunc-
tion in drive or in sterile genotypes because the chromosomes they con-
tain are partially or completely defective. Either explanation is con-

sistent with the observation that drive severity depends on chromatin

 

110

content.

Lifschytz and Lindsley (1972) have proposed that the requirement
for an intact §_chromosome reflects a fundamental regulatory role of the
glin spermatogenesis. They point out that a common feature of spermato-
genesis in heterogametic organisms is early (premeiotic) §_inactivation.
(Lifschytz 1972; Lifschytz and Lindsley 1972). They also cite the fact
that grautosome translocations are male sterile in many organisms, in-
cluding man. Perhaps severe disruptions of early §_inactivation such as
translocation of g genes to a major autosome or deletion of the entire

regulatory locus cause complete sterility. Milder disruptions, such as

 

translocation of §_genes to the fourth chromosome (which may be supposed a
to be closer to the x in inactivation cycle than are the major autosomes)

or deletion of part of the regulatory locus, cause meiotic drive.

CHAPTER 6

SUMMARY AND RECOMMENDATIONS

 

Summagy

Several conclusions emerge from these studies. 1) The recovery of
all chromosomes -- marked and unmarked Zfs, z chromosomes including

euchromatic and heterochromatic deficiencies, and major autosomes --

 

is disrupted by the §_heterochromatic deficiency, sc4sc8. 2) The

probability of recovery of a chromosome from a scl‘sc8 male is an invenxz

 

function of its length. The experimental discrimination is insufficient
to decide whether this means that all chromatin is drive-sensitive or
that there are many discrete response loci distributed along a chromo-
some. 3) If any interactions occur between autosomal and sex chromosomal
recovery disruption they are weak ones. Autosomal and sex chromosomal
recovery ratios are independent of each other. 4) Drive levels as
measured by.§ chromosome recovery are independent of the amount of sex
chromatin in the genome. 5) Heterochromatically duplicated g chromo-
somes do not induce meiotic drive, implying that unreacted pairing
sites are not responsible for meiotic disruption in scasc8 males. 6)
Levels of drive and nondisjunction in scasc8 males can be independently
modified by X chromosomes or autosomes.

These conclusions have several implications for understanding of
the mechanism of sex chromosome meiotic drive. The length dependence
effect could be explained by assuming that sc45c8 disrupts production

of a chromosome processing material, causing a shortage and leading to

111

112

competition among chromosomes. However, the lack of interaction
between sex chromosomes and autosomes and the failure of additional
sex chromatin to enhance drive argue against the notion that chromosomes
in scasc8 males must compete for a scarce resource. An alternative
explanation of the length effect is that mispairing of unequal-sized
homologs at meiosis I causes a failure to inactivate the unpaired
stretch of the larger chromosome. This stretch is then an "armed bomb"
which can destroy any sperm that carry it. This hypothesis implies
that unpaired chromosomes like scasc8 in three-sex-chromosome genotypes
experience poor recovery because their bombs can not be defused. The
recovery of an unpaired chromosome would be independent of the amount of
sex chromatin in the genotype. This is what is observed. This hypothe-
sis also accounts for the correlation between nondisjunction frequency
and drive level since it postulates that mispairing of gland X_is the
fundamental lesion which leads to both phenotypes. However, the armed
bomb model fails to account for autosomal sensitivity to sc4sc8-induced
drive. It also predicts drive induction by heterochromatically dupli-
cated gfs, contrary to observation. Furthermore, the finding that
regular X_chromosomal pairing with large z heterochromatic free dupli-
cations in scasc8 males fails to improve X_recovery implies that the
x defusing sites must be euchromatic. This seems far-fetched, given
the consistent observation that £7: pairing is limited to the g
heterochromatin.

Other, noncompetitive shortage models can be devised. For example,
scasc8 might reduce the time available for a key meiotic process.
Longer chromosomes would be less likely to finish the step than shorter

ones and would, more likely become gametic lethals. The evidence for

113

autosomal sensitivity is consistent with this idea, but no critical test
of it has been performed.
It is also quite possible that nothing at all is wrong with the

post-meiotic chromosomes of scl‘sc8 males. It could be that sc4sc8

damages some other aspect of sperm function in such a way as to leave
sperm sensitive to the amount of chromatin in them. Since there is no
evidence for defective chromatin, this is a possibility. ‘$g
Since meiotic drive is caused by deficiency for g heterochromatin if
but not by duplication for it, it is reasonable to suppose that the

meiotic disruptions are caused by the absence of some §_heterochromatic

 

1 ‘1:

function. However, addition of the §_heterochromatin in the form of a 1
free x duplication does not improve chromosome recovery, although it
regularizes disjunction. The implication is that the x heterochromatin
contains some function that must be present gig to the euchromatic
genes for normal spermatogenesis to occur.

A number of observations are consistent with this view. One is
the occurrence of meiotic drive in T(1;4)BS males. In these males,
most of the x euchromatin is separated from the basal heterochromatin
and attached to a fourth chromosome centromere. The consequence is
poor recovery of the Z_and of the EEK? element, the two longer elements
from each bivalent. Similar translocations between the §_and other
autosomes generally cause complete male sterility. The addition of a
duplication covering the region of the §_breakpoint does not rescue
fertility. This dominance is what one would expect if the sterility
is caused by separation of g genes from a gigfacting regulator.
Lifschytz and Lindsley (1972) have proposed that precocious g inacti—

vation, a common occurrence in male animal meiosis (Lifschytz 1972), is

114

essential for normal spermatogenesis and male fertility, and that separ-
ating §_genes from the g regulatory locus prevents early inactivation.
The regulatory locus is evidently located in the base of the §_judging
from the dominant sterility of deficiencies encompassing both bobbed
(in the heterochromatin) and suppressor of forked (sugf)) (in the
proximal euchromatin near the euchromatin-heterochromatin border)
(Lifschytz and Lindsley 1972).

It is suggested here that sex chromosome meiotic drive is a mild
form of dominant chromosomal male sterility, and that §_inactivation
is interrupted in the drive genotypes to a lesser degree than in the
sterile genotypes. Thus, T(154)BS would be understood as a milder case
of the sterile T(X;A)'s. Perhaps the fourth chromosome is closer to the
.§_in terms of inactivation cycle than are the other autosomes. The
drive-inducingԤ_heterochromatic deficiencies are similar in effect,
but less disruptive, than the larger, male-sterilizing deficiencies.
The implication is that the regulatory site is a large, complex locus
or a repeated gene cluster, partial deficiencies of which induce
varying degrees of sterility. With partial sterility, meiotic drive is
observed either because the partly disturbed sperm have become unusually
sensitive to chromatin content or because the chromatin has become
mildly spermrlethal. Another parallel between meiotic drive and male
sterility is the weak fertility of many drive genotypes: T(l;4)BS
(Novitski 1970), scasc8[ij, and to lesser degrees, sc4sc8/X_and
scasc8/nyp. Finally, there is a dominant sterile interaciton between
drive inducing g heterochromatic deficiencies and Ijziél's (Lindsley,
unpublished). The idea that sex chromosome meiotic drive and dominant

chromosomal male sterility are closely related phenomenon is amenable to

115

several experimental tests.

Recommendations for Future Research

 

 

The evidence for autosomal sensitivity to meiotic drive is so far
based on only one experiment. It is not certain that some uncontrolled,
extraneous factor is not responsible for the apparent sensitivity. It
would be valuable to demonstrate autosomal sensitivity with other experi-

mental set-ups. One experiment would be to construct sc4

 

sc8 males het-

erozygous for a large, heterochromatic deficiency such as Df(2)M-8210.

 

With appropriate markers the recovery of the deficiency and its homolog

can be monitored. The expectation is that the deficiency should be re-

 

ill’? ' '2 '
4

covered in excess of its homolog. A second solution would be to con—
struct sc4sc8 males with two normal second chromosomes plus a free het-
erochromatic chromosome duplication. One such chromosome, Dp(2§f)l
(Lindsley and Grell 1968), has been in existence for some time. Several
others have been constructed recently (Ganetzky, personal communication).
The expectation is that a free second chromosome duplication should show

8

reduced recovery in sc4sc males. Neither of these experiments is as

sensitive as the T(2;3)wa4 experiment of Chapter 4, since in that exper-

 

iment whole chromosomal arms were manipulated, whereas in this experiment
only the dose of basal heterochromatin is altered. However, the impact
of scl‘sc8 is strong enough that it should be observed even in these less
sensitive systems. With the deficiencies and duplications, interpreta—

4

tion of results is not complicated by possible effects of SC SC8 on auto-
somal nondisjunction.

The autosomal system can be used for another look at the impact of

additional sex chromatin on drive levels. The results in Chapter 3 indi-

116

cated that §_chromosome recovery is independent of the amount of sex
chromatin in the genotype. It would be valuable to test the effect of
sex chromatin amount on recovery of another, independently assorting
chromosome pair. It is not obvious in advance which autosomal system
(the translocation or the deficiencies and duplications) will prove most
useful for these tests but trial and error will settle the point. One
advantage of the deficiencies and duplications is that they permit a look
at the reciprocal question, namely, what is the effect of changing the
amount of autosomal chromatin on recovery of sex chromosomes?

Another way of measuring the impact of additional sex chromatin on
drive levels is to compare frequencies of sperm developmental abnormal-
ities between ggfi§£§[Z[X_and scasc8/X[QE_males in the electron microscope
to ascertain what happens to the recovery of 1 sperm when the relatively
small free duplication is replaced as pairing partner by a second X,
According to the armed bomb model, the recovery of X_sperm should improve
dramatically because the two st should be able to defuse each other's
bombs. Relative Z_recovery does improve dramatically (Chapters 2 and 3)
but this reveals nothing about absolute recovery of 2 sperm. Under
other, pairing-independent models of meiotic drive, absolute X_recovery
should stay the same or get worse as a result of the substitution of a
second X_for a free §_duplication. Progeny counts can not answer this
question because they measure only relative X chromosome recovery. While
the electron microscope can not reveal the genotype of abnormal sperm, it
does permit an estimate of the average number of dysfunctional sperm per
bundle. Under the armed bomb model, there should be fewer such sperm per
bundle in scasc8[z[x_than in §E:§E§[ijp_males. Under pairing-indepen-

ent models, there should be more bad sperm per bundle in scasc8[X[X_than

117

in scésc8[1[gp, since in the former genotype all sperm carry the rela—
tively large and, therefore, drive-prone X_chromosome, while in the lat-
ter genotype half the sperm get the small free g duplication. Although
the expected difference in mean dysfunctional sperm per bundle between
the two genotypes is fairly large and in opposite directions under these
two competing hypotheses, the results could be confounded by other fac-
tors. The biggest such factor would be autosomal recovery disruption.
If the severity of autosomal recovery disruption depends strongly upon
amount of sex chromatin in the genome, then it may be impossible to
detect stpecific effects on mean dysfunctional sperm per bundle. The
experiments outlined above should reveal whether or not there is an
effect of sex chromatin amount on autosomal recovery disruption.

Another reason for cytological examination of scasc8[zlz_and
§£f§2342422.m3133 is to look for post-meiotic X chromosome loss. '1
chromosomes are certainly not lost (Cooper 1964) and there are genetic
reasons (Chapter 1) for thinking that the x chromosome probably behaves
like the X, Post-meiotic cytology would be helpful in terms of confirm-
ing or refuting this genetic evidence.

Several investigators have suggested that meiotic drive in scasc8
males is caused by lfi-X'mispairin'g which in turn is due to the deletion of
g chromosomal pairing sites (Baker and Carpenter 1972; Peacock and
Miklos 1973). Although the results described above have not generally
supported this model, there is still a body of evidence suggesting a
close correlation between levels of nondisjunction and of drive
(Zimmering 1963; Baker and Carpenter 1972; Peacock and Miklos 1973).
Both drive and nondisjunction are deficiency phenotypes and both result

from deletion of basal g heterochromatin. A deficiency mapping experi-

118

ment should reveal whether or not the drive site(s) and the pairing sites
are identical. Such an experiment would also clarify the relationship
between dominant male sterility and dominant meiotic drive observed for
different deficiencies in that region. It would test the suggestion that
fertility is controlled by a tandemly repeated gene family, partial
deficiencies for which induce varying degrees of male sterility. A large
number of deficiencies for part of the basal heterochromatin are in exis-
tence or can be easily constructed including left-right combinations of
scute and white-mottled inversions (Lindsley and Grell 1968), X—ray in-
duced bobbed-lethal deficiencies (Lindsley, Edington and von Halls 1960),
and the spifl deficiencies implicated in male sterility (Lifschytz and
Lindsley 1972). More deficiencies can be generated by X-ray mutagenesis
as needed. By examining the phenotypes of these deficiencies with re-
spect to fertility, nondisjunction, and sex chromosomal and autosomal
recovery disruption, it should be possible to determine the locations of
the pairing sites, drive sites, and fertility sites to determine to what
degree these functions overlap.

A related question concerns the uniqueness of the x heterochromatic
function concerned with insuring normal segregation ratios. Is it also
found on the X_chromosome like other z heterochromatic functions-~bobbed
and the pairing sites? Or is it unique to the g2 A way of answering
this question is to replace the g heterochromatin with either XE_or Xi
and test the resulting chromosomes for ability to induce drive. Some
chromosomes appropriate for these experiments are already in existence,
and others can be constructed as needed.

The suggestion that dominant chromosomal male sterility and sex

chromosome meiotic drive are qualitatively similar rests on the numerous

119

parallels between the phenotypes. It seems reasonable to test for other
parallels. For example, §2(f)-bb deficiencies are partly fertile in the
presence of certain X_chromosomes. It would be valuable to know what
happens in terms of nondisjunction and segregation ratios in these males.
Meiotic drive and nondisjunction are subject to modification by a variety
of genetic and environmental factors (Chapter 2). Do these modifiers a
also interact with fertility levels in the appropriate genotypes?

The above experiments are reasonable extensions of the work de-
scribed in the previous chapters and would constitute the next few steps
in the attempt to understand the genetics of sex chromosome meiotic
drive. Two other experiments somewhat further afield from this work, but
perhaps equally interesting, will be described briefly. One is a survey
of T(154)'s to find out which of them induce drive. Very few have been
tested so far. Two with breakpoints in 162, including T(l;4)BS, are
drive inducers (Zimmering 1960). Another, T(154)scH, with a breakpoint
near the tip is not (Chapter 5). It is interesting that T(X;A)'s with
terminal g breakpoints tend to be fertile (at least when combined with
terminal autosomal breakpoints) while those wflflncentral §_breakpoints are
all sterile no matter where the autosomal breakpoint falls (Lindsley and
Lifschytz 1972). It would be valuable to know whether a similar pattern
holds for drive-inducing and non-drive-inducing T(Xg4)'s.

lluasecond experiment is a reisolation and reexamination of Baker and
Carpenter's (1972) male-specific X-linked meiotic mutants. All twenty of
their EMS-induced mutants, isolated on the basis of causing high §:X_non-
disjunction, were also found to cause skewed sex chromosome segregation
ratios. In fact, they all mimicked scésc8. Two were mapped to the

proximal g euchromatin. Before the others could be mapped, they all

 

 

«£35 v a:

 

120

reverted. It would be ineresting to reexamine mutants of this phenotype
in light of our current knowledge of scésca. For example, do these
mutants also induce autosomal drive? How do they behave in three-sex—
chromosome genotypes? Do chromosomes show length dependent recovery prob-
abilities in their presence? Another interesting aspect of these mutants
is their high frequency. Baker and Carpenter had little trouble isolat-
ing twenty of them. It would be very interesting to know where they map.
Are there a large number of x chromosomal functions involved in normal
zjz_disjunction and segregation? Or are there one or a few hot spots for
this type of mutation? The other interesting aspect of these mutants is
their instability. It is very odd that all of them should revert in such
a short time. Despite the difficluty of working with unstable mutations,
the potential benefit in terms of understanding the central processes of

spermatogenesis make them well worth a second look.

BIBLIOGRAPHY

BIBLIOGRAPHY

Baker, B.S. and A.T.C. Carpenter. Genetic analysis of sex chromosome
meiotic mutants in Drosophila melanogaster. Genetics 71:255-286;
1972.

 

Brink, N.G. Protein synthesis during spermatogenesis in Drosophila
melanogaster. Mutation Research 5: 192-194; 1968.

 

 

Cooper, K.W. Meiotic conjunctive elements not involving chiasmata.
Proceedings of the National Academy of Sciences, USA 52: 1248-1255;
1964.

Das, C.C., B.P. Kaufmann and H. Gay. Autoradiographic evidence of syn-
thesis of an arginine-rich histone during spermiogenesis in Dro-
sophila melanogaster. Nature 204: 1008-1009; 1964.

 

Denell, R.E. and G.L.G. Miklos. The relationship between first and sec-
ond chromosome segregation ratios from Drosophila melanogaster
males bearing Segregation Distorter. Molecular and General
Genetics 110: 167-177; 1971.

 

Erickson, J. Meiotic drive in Drosophila involving chromosome breakage.
Genetics 51: 555-571; 1965.

Ganetzky, B. On the components of segregation distortion in Drosophila
melanogaster. Genetics 86: 413-445; 1977.

 

Gershenson, S. Studies on the genetically inert region of the x chromo-
some of Drosophila. 1. Behavior of an §_chromosome deficient for a
part of the inert region. Journal of Genetics 28: 297-312; 1933.

Glass, H.B. A study of dominant mosaic eye-color mutants in Drosophila
melanogaster. II. Tests involving crossing-over and nondisjunc-
tion. Journal of Genetics 28: 69-112; 1933.

 

 

Glass, H.B. A study of factors influencing chromosome segregation in
translocations of Drosophila melanogaster. University of Missouri
Research Bulletin 231; 1935.

 

Gould-Somero, M. and L. Holland. The timing of RNA synthesis for sper-
miogenesis in organ cultures of Drosophila melanogaster testes.
Wilhelm Roux' Archives 174: 133-148; 1974.

 

Haemer, J. Functions of Heterochromatin. Ph.D. dissertation, Univer-
sity of Washington; 1978.

121

122

Hardy, R.W. The influence of chromosome content on the size and shape
of sperm heads in Drosophila melanogaster and the demonstration of
chromosome loss during spermiogenesis. Genetics 79:231-264; 1975.

 

Hartl, D.L. and Y. Hiraizumi. Segregation Distortion. Ashburner, M.
and E. Novitski, eds. The genetics and biology of Drosophila,
volume 13, London; Academic Press; 1976: 616-666.

Hartl, D.L., Y. Hiraizumi, and J.F. Crow. Evidence for sperm dysfunction
as the mechanism of segregation distortion in Drosophila melano—
gaster. Proceedings of the National Academy of Sciences USA 58:
2240-2245; 1967.

 

Hiraizumi, Y. and K. Nakazima. Deviant sex ratio associated with
Segregation Distorter in Drosophila melanogaster. Genetics 55:
681-687; 1967.

 

Kettaneh, N.P. and D.L. Hartl. Histone transition during spermiogenesis
is absent in Segregation Distorter males of Drosophila melanogaster.
Science 193: 1020-1021; 1976.

 

Kettaneh, N.P. and D.L. Hartl. Ultrastructural analysis of spermio-
genesis in Segregation Distorter males of Drosophila melanogaster:
the homozygotes. Genetics 96: 665-683; 1980.

 

Lifschytz, E. Ԥ_Chromosome inactivation: an essential feature of normal
spermiogenesis in male heterogametic organisms. Beatty, R.A. and
S. Gluecksohn-Waelsch, eds. Proceedings of the Edinburgh symposium
on the genetics of the spermatozoon. Edinburgh: University of
Edinburgh; 1972: 223-232.

Lifschytz, E. and D.L. Lindsley. The role of g-chromosome inactivation
during spermatogenesis. Proceedings of the National Academy of
Sciences USA 69: 182-186; 1972.

Lindsley, D.L. and E.H. Grell. Genetic variations of Drosophila melanogas-
ter. Oak Ridge, Tennessee: Carnegie Institution of Washington
Publication Number 627; 1968.

 

Lindsley, D.L. and E.H. Grell. Spermiogenesis without chromosomes in
Drosophila melanogaster. Genetics, Supplement 61: 69-77; 1969.

Lindsley, D.L. and E. Lifschytz. The genetic control of spermatogenesis
in Drosophila. Beatty, R.A. and S. Gluecksohn-Waelsch, eds.
Proceedings of the Edinburgh symposium on the genetics of the
spermatozoon. Edinburgh: University of Edinburgh; 1972: 203-222.

Lindsley, D.L. and L. Sandler. The meiotic behavior of grossly deleted
§_chromosomes in Drosophila melanogaster. Genetics 43: 547-563;
1958.

 

Lindsley, D.L., C.W. Edington, and B.S. vonHalle. Sex-linked recessive
lethals in Drosophila whose expression is suppressed by the X
chromosome. Genetics 45: 1649-1670; 1960.

123

Miklos, G.L.G., A.F. Yanders, and W.J. Peacock. Multiple meiotic drive
systems in the Drosophila melanogaster male. Genetics 72: 105-115;
1972.

 

Monesi, V. Synthetic activities during spermatogenesis in the mouse.
Experimental Cell Research 39: 197-224; 1965.

Moore, G.P.M. DNA-dependent RMA synthesis in fixed cells during sperm-
atogenesis in mouse. Experimental Cell Research 68: 462-465; 1971.

Muller, H.J. and A.A. Prokofyeva. The individual gene in relation to
the chromomere and the chromosome. Proceedings of the National
Academy of Sciences USA 21: 16-26; 1935. I

Muller, H.J. and F. Settles. The nonfunctioning of the genes in sperm-
atozoa. Z. Indukt. Abstamm.-Vererbungsl. 43: 285-312; 1927.

Novitski, E. The concept of gamete dysfunction. Drosophila Information
Service 45: 87-88; 1970.

 

Novitski, E. and I. Sandler. Are all products of spermatogenesis regu-
larly functional? Proceedings of the National Academy of Sciences
USA 43: 318-324; 1957.

Olivieri, G. and A. Olivieri. Autoradiographic study of nucleic acid
synthesis during spermatogenesis in Drosophila melanogaster.
Mutation Research 2: 366-380; 1965.

 

Peacock, W.J. Nonrandom segregation of chromosomes in Drosophila males.
Genetics 51: 573-583; 1965.

Peacock, W.J. and G.L.G. Miklos. Meiotic drive in Drosophila: new
interpretations of the Segregation Distorter and sex chromosome
systems. Advances in Genetics 17: 361-409; 1973.

Peacock, W.J., G.L.G. Miklos, and D.J. Goodchild. Sex chromosome meiotic
drive systems in Drosophila melanogaster.1. Abnormal spermatid
devglo ment in males with a heterochromatin deficient x chromosome
(sc sc ). Genetics 79: 613-634; 1975.

 

Peacock, W.J., K.T. Tokuyasu, and R.W. Hardy. Spermiogenesis and meiotic
drive in Drosophila. Beatty, R.A. and S. Gluecksohn—Waelsch, eds.
Proceedings of the Edinburgh symposium on the genetics of the sperm-
atozoon. Edinburgh: University of Edinburgh; 1972: 247-268.

Ramel, C. The effect of the curly inversions on meiosis in Drosophila
melanogaster.II. Interchromosomal effects on males carrying
heterochroatin deficient §_chromosome. Hereditas 60: 211-222; 1968.

 

 

Roberts, P.A. The genetics of chromosome aberration. Ashburner, M. and
E. Novitski, eds. The genetics and biology of Drosophila, volume 1A.
London: Academic Press; 1976: 68-184.

124

Sandler, L. and G. Braver. The meiotic loss of unpaired chromosomes in
Drosophila melanogaster. Genetics 39: 365-377; 1954.

 

Sandler, L. and A.T.C. Carpenter. A note on the chromosomal site of
action of SD in Drosophila melanogaster. Beatty, R.A. and S.
Gluecksohn-Waelsch, eds. Proceedings of the Edinburgh symposium
on the genetics of the spermatozoon. Edinburgh: University of
Edinburgh; 1972: 233-246.

 

Sandler, L. and Y. Hiraizumi. Meiotic drive in natural populations of
Drosophila melanogaster V. On the nature of the SD region.
Genetics 45: 1671-1679; 1972.

Sandler, L., Y. Hiraizumi, and I. Sandler. Meiotic drive in natural
populations of Drosophila melanogaster. I. The cytogenetic basis
of Segregation-Distorter. Genetics 44: 233-250; 1959.

 

Spofford, J.B. Position effect variegation. Ashburner, M. and E.
Novitski, eds. The genetics and biology of Drosophila, volume 1C.
London: Academic Press; 1976: 955.

Tokuyasu, K.T., W.J. Peacock, and R.W. Hardy. Dynamics of spermiogenesis
in Drosophila melanogaster. VII. Effects of Segregation-Distorter
(SQ) chromosomes. Journal of Ultrastructure Research 58: 96-107;
1977.

Yamamoto, M. and G.L.G. Miklos. Genetic dissection of heterochromatin
in Droshophila. The role of basal x heterochromatin in meiotic
sex chromosome behavior. Chromosoma 60: 283-296; 1977.

Zimmering, S. Modification of abnormal gametic ratios. Science 130:
1426; 1959.

Zimmering, S. Modification of abnormal gametic ratios in Drosophila. 1.
Evidence for an influence of Y_chromosomes and major autosomes on

gametic ratios from Bar-Stone translocation males. Genetics 45:
1253-1268; 1960.

Zimmering, S. The effect of temperature on meiotic loss of the Y
.chromosome in male Drosophila. Genetics 48: 133-138; 1963.

Zimmering, 8. Genetic and cytogenetic aspects of altered segregation
phenomena in Drosophila. Ashburner, M. and E. Novitski, eds.
The Genetics and Biology of Drosophila, volume 13. London: Academic
Press; 1976: 569-615.

Zimmering, S. and R.E. Green. Temperature dependent transmission rate
of a univalent §_chromosome in the male Drosophila melanogaster.
Canadian Journal of Genetics and Cytology 7: 453-456; 1965.

 

Zimmering, S. and M. Perlman. Modification of abnormal gametic ratios
in Drosophila. III. Probable time of the A-type effect in Bar-
Stone translocation males. Canadian Journal of Genetics and
Cytology 4: 333-336; 1962.

125

Zimmering, S., L. Sandler, and B. Nicoletti. Mechanisms of meiotic
drive. Annual Review of Genetics 4: 409-436; 1970.