The Sex-Ratio Trait in Drosophila simulans: Genetic Analysis of Distortion and Suppression

(1)

Copyright 0 1997 by the Genetics Society of America

The Sa-Ratio Trait

in

Drosophila simulans:

Genetic Analysis of Distortion

and

Suppression

Michel Cazemajor, Claudie

Landre and Catherine MontchampMoreau

Znstitut Jacques Monod, Laboratoire Dynamique du Ginome et Evolution, CNRS-Universite‘ Paris 7, 75251 Paris Cedex 05, France

Manuscript received December 7 , 1995 Accepted for publication June 18, 1997

ABSTRACT

The sex-ratio trait described in several Drosophila species is a type of naturally occurring X-linked meiotic drive that causes males bearing a sex-ratio X chromosome to produce progenies with a large

excess of females. We have previously reported the occurrence of sex-ratio X chromosomes in Drosophila simulans. In this species, because of the co-occurrence of drive suppressors, the natural populations and the derived laboratory strains show an equal sex-ratio even when sex-ratio X chromosomes are present at a high frequency. The presence of sex-ratio X chromosomes is established via crosses with a standard strain that is devoid of drive suppressors. In this article, we show first that the sex-ratio trait in D. simulans

results from the action of several X-linked loci. Second we describe drive suppressors on each major autosome as well as on the Y chromosome. The Y-linked factors suppress the drive partially whereas the

autosomal suppression can be complete.

T

HE sex-ratio trait is a type of meiotic drive due to X-linked factors that appears widespread among

Drosophila species. Males that bear a sex-ratio Xchromo-

some (sexratio males) transmit mainly X-bearing sperm

and therefore produce only or mostly female progeny.

Driver sex-ratio X chromosomes were first described

within the subgenus Sophophora, in seven species be-

longing to the D. obscura species group (MORGAN et al.

1925; GERSHENSON 1928; STURTEVANT and DOBZHANSKY

1936; JUNGEN 1968). They have been also found in spe-

cies belonging to four different groups of the subgenus

Drosophila: D. paramelanica (STALKER 1961), D. medio-

punctata (DE CARVALHO et al. 1989), D. testacea

UAMES

and JAENIKE 1990), D. guinam’a and D. recens (JAENIKE

1996). Since they should be transmitted by males at

higher rates than a normal X, sex-ratio X are expected

to invade the populations and cause their extinction, once fixed, if the bias toward females is complete (GER- SHENSON 1928; HAMILTON 1967). However sex-ratio X

chromosomes are usually maintained at low or moder-

ate frequency in natural populations (STURTEVANT and

DOBZHANSKY 1936; DOBZHANSKY and EPLING 1944; DOE ZHANSKY 1958; JUNGEN 1968; DE CARVALHO et al. 1989; HAUSCHTECK~UNCEN 1990; JAENIKE 1996). Two mecha- nisms are put forward that may prevent their spread through populations: countervailing selection on indi-

viduals bearing sex-ratio X (reviewed in JAENIKE 1996)

and selection on autosomal and Y-linked suppressors of

drive (THOMSON and FELDMAN 1975; WU 1983a; CLARK

1987). Drive suppressors have been found in D. affinis

Corresponding author: Catherine MontchampMoreau, Laboratoire Dynamique du C4nome et Evolution, Tour 424, Institut Jacques Mo- nod, 2 Place Jussieu, 75251 Paris Cedex 05, France.

E-mail: [email protected]

Genrtirs 147: 63.5-642 (October, 1997)

(VOELKER 1972),

D.

paramelmica (STALKER 1961), D.

mediopunctata (DE CARVALHO and KLACZKO 1993, 1994)

and possibly occur in D. subobscura (HAUSCHTECK-

JUNGEN 1990). Nevertheless, in all these species, sex- ratio males are observed within populations. Thus, the

drive suppression is incomplete at the population level.

The situation appears rather different in D. simulans,

a species that belongs to the D. melanogaster group. A

first report of female-biased sex-ratio, in a stock derived from a natural population of California, has been ac- counted for by a recessive autosomal drive factor (FAUL HABER 1967). A second report (DE MAGALHAES et al.

1985) also refers to a recessive autosomal factor, in Bra- zilian populations, that might correspond to the same

system as FAULHABER’S. Autosomal sex-ratio factors have

never been reported in other Drosophila species and theoretically, an autosomal driver acting on sex-chro-

mosomes cannot be selected for (HAMILTON 1966).

Thus, a possible interpretation of FAULHABER’S data is

that a “hidden” sex-ratio X chromosome, whose action

was suppressed by an autosomal factor, was present in the Californian population and that the few males ex- pressing the drive were those that bore an inefficient

allele at the suppressor locus (HURST and POMIANKOW-

SKI 1991). In fact, the presence of hidden sex-ratio X

chromosomes in D. simulans has been recently reported

in our laboratory (MERCOT et al. 1995). Such chromo-

somes have been characterized in stocks originating

from many locations over the world (ATLAN et al. 1997).

By contrast with other species, sex-ratio X chromosomes

may reach high frequencies (up to 60%) in natural

populations of D. simulans (ATLAN et al. 1997). However,

they rarely if ever express their drive ability within the

(2)

636 M. Cazemajor, C. Landre and C. MontchampMoreau

because of the systematic co-occurrence of drive suppressors (ATLAN et al. 1997). The presence of sex-ratio

X chromosomes is revealed via crosses with a standard strain (ST) that is devoid of drive suppressors. F1 hybrid males that bear a sex-ratio X and a standard Yare typical sex-ratio males: they exhibit spermiogenic failure

(MONTCHAMP-MOREAU and JOLY 1997) and usually pro-

duce between 70% and 100% females in their progeny. The bias toward females is not associated with postzy- gotic mortality nor d o e s i t depend on the type of female the males are mated with (MERCOT et al. 1995).

The aim of the present work is to characterize the genetic components of the sex-ratio system i n

D.

sim- ulans, using a laboratory stock

(SR)

in which sex-ratio X became spontaneously virtually fixed. First we show that several X-linked factors are responsible for the sex-ratio trait, second that recessive autosomal drive suppressors on the second and the third chromosomes, as well as Y-linked ones, together are responsible for a complete supression of the sex-ratio trait in the SR strain.

MATERIAL AND METHODS

Strains: ST is a reference standard strain, free of sex-ratio distorters and of drive suppressors (MERCOT et al. 1995). The distribution of frequency of females among progeny of 53 ST males individually crossed with ST females is shown on Figure 1A. SR is our reference strain for the sex-ratio system (ATLAN et al. 1997). It shows an equal sex-ratio (Figure 1B) and the distribution of female percentage among individual progeny of 55 SR males does not differ significantly from that of the ST males

(x2

= 9.733, 5 d.f., P = 0.14). However, it carries driver X chromosomes, resulting in a large excess of females in most of the individual progenies of the F, hybrid males bearing an X chromosome from the SR strain

( y R )

and a standard Y chromosome

(p')

(Figure 1C). The SR5 strain carries a standard enetic background together with driver X chromosomes ( x ' ) whose driving factors come from a unique Xchromosome of the SR strain. As a result, SR5 males express the sex-ratio trait. We assume that SR5 got rid of autosomal drive suppressors present in the SR strain because no drive resistance has evolved in this strain for 2 years. This strain was obtained as follows. (1) A parental cross between one ST female and one SR male was followed by alternate backcrosses with ST females (even generations) or ST males (odd generations). At each even generation (F2 to FI2), only males producing more than 80% females were retained, and their fernale progenies (odd generations) were used in the subsequent cross. (2) One pair of F13 male and female, which had produced only sex-ratio F14 males, was used as founder of the SR5 strain. The net e strain bears the recessive mutants net (net) and ebony ( e ) on the second and third chromosomes, respectively. np lz v / C ( I ) R M , y w is a strain with attached X

females bearing yellow (y) and white ( w ) mutations and males with an X chromosome carrying nipped-wing (np), lozenge ( l z ) and vermilion ( v ) mutations. This strain was obtained after recombination of X chromosomes from the stocks (lz[sp]/ C(l)RM, y w ) , np and v m from the Bloomington Stock Cen- ter, at Indiana University. The vermilion and lozenge mutations are homologous with the D. melanogaster mutations of the same name. Our tests gave a localization of lozenge 6.0 2 1.4 cM distal to vermilion and of nipped-wing 15.6 2 2.2 cM distal

to lozenge. Both net eand np lz v/C(l)RM strains are assumed

k

to be free of distorters and suppressors, for they behave like the ST strain in crosses with the SR strain.

General procedure for sex-ratio estimate: Males under test were individually crossed with virgin ST females. They were transferred to new vials approximately every 3 days. The adults emerging from one or more vials were sexed and counted until vials were exhausted. Unless otherwise stated, only crosses producing 50 or more flies were considered. All of

the experiments were carried out at 25".

Localization of SR on the chromosome: Twenty SR5

females were mass-crossed with 30 np lz v males, then 40 F1 females ( PR5/Yp

'

") were individually crossed with ST males to obtain recombinant X chromosomes in F2 flies. Distortion ability of recombinant X chromosomes against

P'

was measured by individually crossing F2 males of each genotype with ST females. Because of variable expressivity of np, the genotype of F2 males with a np+ phenotype was checked on F4 males through Fs sib-mating. As a control, sex-ratios in the progenies of males bearing "parental" combinations of sex chromosomes (

pRs/ry7'

and Fp

'

" / P T )

were also measured:

( 1 ) 30 SR5 females were mass-crossed with 40 ST males, and sex-ratios in the progenies of XSR5/PTFI males were measured by individually crossing them with ST females, (2) 30 np lz v /

C(l)RM, y w females were mass-crossed with 40 ST males, 30 p'/C(l)RM, y w FI females were mass crossed with 40 males

of the np lz v strain and sex-ratios in the individual progenies

of np lz

v / p T

F2 males were measured as above.

Test for Y-linked drive suppressors in the SR strain: Forty

SR males were mass-crossed with 40 net e females. Then, 40 FI males were mass-crossed with 40 net e females. F2 males, bearing Y" chromosomes together with X , second, and third chromosomes from the net e stock, were masscrossed with

SR5 females. Sensitivity of

PR

chromosomes to distortion was measured using individual crosses of 52 F, males

(pH/

pR;

ZpR5/ZFt; ZZpR5/ZZT) with ST females. The same protocol, starting with 40 ST males, was used to test P7'sensitivity a ainst X'Rs distortion. In this case, 49 F3 males ( X ' R 5 / r"; IF R F /ZP;

ZZfH/ZZT) were tested. Only crosses producing at least 100 flies were taken into account.

Test for autosomal drive suppressors in the SR strain: Eight SR females were individually crossed with net e males. Three to eight F1 males from each progeny were then individually crossed with SR females. F2 males, together with Y chromosomes from the net e stock, could have four different autosomal combinations: heterozygous for Z f R and

UpR,

homozygous for either I f R ,

UpR,

or both (see Figure 4). The genotype of these males as well as their distorting ability were checked by crossing them with net e females.

RESULTS

Segregation of the sex-ratio trait in

PT/P

fe-

males: The aim of this experiment was to d e t e r m i n e

whether one or several regions on t h e

PR

chromosomes were responsible for the sex-ratio trait.

SR

females were

mass-crossed with ST males and six F1 females

(pR/p")

were individually crossed with ST males. Between 25 a n d 39 F2 males were recovered from each F1 female a n d individually tested for their drive ability. All t h e six females were found to p r o d u c e b o t h sex-ratio a n d n o n - sex-ratio males. The distributions of the percentages of females among individual progenies of the FZ males were statistically homogeneous among the six FI females

(x2

= 16.9, 15 d.f., P = 0.33), and the pooled

(3)

Sex-Ratio in D. simuluns 637

",

6

J I L

'i

0 -4-i-

4 12 20 28 36 44 52 60 68 76 84 92 100

OD C

2

40,

= I

Q) 35 1

SWST

hybrid

males

3o

1 mean91.4%

cc

a

25

0 20 -

a

1 5 -

E

lo

z

;L

t + - t i - I ~ + t i " r _ - L A " i

4 12 20 28 38 44 52 60 68 76 84 92 100

D

'

sons of SWST

hybrids

females

U

4 12 20 28 38 44 52 60 68 76 84 92 100

Female percent in progeny

FIGURE 1.-Expression of the sex-rutio trait. Distribution of female percentage among individual progenies of males from

ST strain (A), SR strain (B), F1 hybrids obtained by crosses between SR females and ST males (C), F2 males, sons of F1 hybrids females, obtained by crosses between SR females and

ST males (D). Mean female percentages are indicated in the corresponding histogram. To allow for a statistical compari- son of the distributions, each individual progeny consists of

50 individuals randomly sampled (sampling without replace- ment among the whole individual progeny scored).

tested, two main classes can be clearly distinguished:

117

were non-sex-ratio (from 40 to 60% females in their progeny), and

78

were sex-rutio (from 76 to 100% females), as expected of a monogenic trait. However, six males, whose female percentage in progeny lay within the range 64-7496, were not in accordance with this hypothesis, unless they could correspond to either of the following cases. (1) A variable expressivity of the drive factor(s), as already observed among F1

x""/r"'

males (Figure 1C). This would be consistent with the fact that a given sex-ratio male showed variations in female percentage between progenies of successive mat- ings that sometimes encompassed the sampling effect (data not shown).

(2)

A dominant effect of some SR autosomal drive suppressors (about a quarter of F2 autosomal genes were from the SR strain). Alternatively, according to a model in which several tightly linked loci are responsible for the drive ability of the

p n

chromosomes, the F2 males with a moderate drive ability could possess rare recombined Xchromosomes lacking one or several of the

Pn

drive factors. A multilocus model could also account for the significant excess of non-sex-ratio among F2 males (117 vs. 84,

x2

= 5.4, 1 d.f.,

P

<

0.05), which might include males bearing inefficient combinations of drive factors.

Localization of distorter(s) on the

F

chromo-

some: To eliminate any potential interaction with au-

tosomal suppressors when looking for linkage relation- ships of the drive factor(s), we used the SR5 strain that was supressor-free and whose driving factors came from a unique Xchromosome of the

SR

strain (see MATERIALS AND METHODS). The distortion ability of parental F p

'

and

Ym

chromosomes (from the np lz v and SR5 strains, respectively) against

P"

is shown in Figure

2,

A

and B. The 94 Fp ' I ' chromosomes we tested did not

induce distortion and led to equal mean sex-proportions among progenies (from 31% to 69% females). By contrast, the

Ym

chromosomes were found to induce a variable but always significant excess of females (from

67%

to 100%) among the individual progenies of the 126 tested males. These males produced a mean of 91.6% females, which was similar to the value obtained with hybrid

pn

/P7'

males (91.4%, Figure 1C). This strongly suggests that the

pK5

chromosomes retained most if not all the drive factors present on the

p n

chromosomes.

The distortion ability of six classes of Xchromosomes obtained among the male progeny of F1 hybrid females

x s R s / x " p " is shown in Figure

2, B-H.

Double recombi-

nant chromosomes np

+

v and

+

lz

+ were rarely

ob-

tained and seemed to behave like np

+ + and

+

lz v,

respectively (data not shown). It can be inferred from Figure

2,

C-F that the sex-rutiofactor(s) is tightly linked to the np

lz

region. (1) All but 11 of the 238 np Zz v F2 males (Figure 2C) and all of the 51 np lz

+ F2 males

(4)

638 M. Cazemajor, C. LandrC and C. Montchamp-Moreau

i

I

r + i b + ,

-I

5 15 25

A+-

35 45 55 65 75

C

F2 np lz v

mean 5

1.8%

- i i -+ +

,

5 15 25 35 45 55 65

0

L++-+-

t I"+ t - 4 - I -i"r-

5 15 25 35 45 55 65 75 85 95

75 85 95 5 15 25 35 45 55 65 75 85 95

it

30

3 5 ~ F i

2 5 1 F 2 + + v 20

1

mean 91.4% 15

1

lo ,

5 4

"+~+---H 0

c+

I""i++-A"

75 85 95 5 15 25 35 45 55 65 75 85 95

25 I

H

.-

8

25

G

gJ

20 ~

2

1 0 ; 5 1

C I

I

F 2 + I z v

AlL

e l

y . 1

15 i mean 58.8%

Q)

€ 1

S !

0 I 1 I t l - - i ~ i - - t ~

5 15 25 35 45 55 65 75 85 95 5 15 25 35 45 55 65 75 85 95

Female percent

in progeny Female percent

in progeny

FIGURE 2.-Linkage of sex-ratio factors with lz and np loci. Distribution of the progenies of males with parental Xchromosomes

(A and B) and F9 males (C-H). Male genotype and mean female percentage are indicated in the corresponding histogram. Abscissa, percentage females in progenies; ordinate, number of progenies.

70% females). Figure 2, A and B, showed that the two

parental distributions partially overlapped over the

range 65-70%. Thus the distortion ability of the five

np lz v F2 chromosomes lying within this range was

checked on the np lz u F4 male progenies, following

Fs sib-mating. None of them were found to possess a heritable driving effect [since the sex-ratio factors (s) are tightly linked with the np 1 . region, the probability for

one of them to be present in a np Zz u F2 male but not

in its np lz u F4 male progeny is very low]. Only 11 np

lz u F2 males (4.6%), which produced >70% females,

were classified as sex-ratio. ( 2 ) Similarly, the sex-ratio trait

(criterion >70% females) appears expressed by all of

the

+

v F, males and all but five of the

+ + +

ones.

A strong distortion ability was recovered among F4 male

progeny from four out of these five. Thus, at most 0.4%

of the

+ + +

F2 males appeared to lack an heritable

drive ability. In addition, there were significantly- more

moderate or weak distorters (<85% females) among

+

F2 chromosomes than among

Yrn

ones (Fisher

exact test P = 3.10-4), which suggests a partial lost of drive ability by a few of the former.

The effect of a recombination between np and lz is

(5)

Sex-Ratio in D. simulans 639

2G is bimodal and the proportion of sex-ratio among

+

lz v F2 males (criterion >70% females) was significantly

higher than that observed among the np lz v males

(x2

=

11.77,

1 d.f., P

<

0.001)

,

as expected from a monogenic trait with a drive factor located between the np and lz loci. However, it must be pointed out that the right tail of the 35-75% group included sex-ratio males:

the three

+

lz v F2 males that had led to between 65%

and 75% females were found to possess an heritable

drive effect (tests on Zz F4 males). Interestingly, the dis-

tribution of the percentage of females among individ-

ual progenies of np

+

males was not bimodal (Figure

2G): more than one third of these progenies showed a moderate bias toward females (60-85%). Such values

were rarely observed with both parental Xchromosome.

The distortion ability of the 26 recombinant np

+ +

chromosomes that resulted in between 55% and 70% females was checked on the corresponding np F4 male progenies. Most of them (22) were found to possess an heritable drive ability. This supports the hypothesis that

at least two loci, in the np-1% region, are responsible for

the sex-ratio trait. Because moderate distorters (60-85%

females) seldom occur among

+

lz v F2 males when

compared to np

+ +

ones, more than two loci may be

involved and/or these may have quantitatively different effects on the trait.

This model is doubtless overly simple, because the

recovery of 4.6% of distorters (some of them strong

ones) among np lz v F2 males and the excess of weak

distorters among the

+

F2 ones (Figure

2,

C and

D) can hardly be explained by double recombination

events within a 16 cM region. Although major roles are played by loci in the np-1% region, additional gene(s) outside this region may be involved in the expression of the sex-ratio trait.

Test €or Clinked suppressors of drive: In this experi-

ment,

pm

chromosomes were found to induce a mean

of 66.4% females in the progeny of pR-bearing males

(Figure 3A). Meanwhile, control

p m / p T

males pro-

duced a mean of 81.2% females (Figure 3B). These

means were significantly different (Mann-Whitney rank

sum test uObs = 6.9, P

< 10"'). Thus, the

pR

chromo-

somes are less affected by drive than the P'ones:

assuming, after LYITLE (1979), that the frequency of

nonfunctional "Fspermatids" is z = (2k - 1 ) / k (where

k is the observed frequency of females in the progeny),

then

77%

of P7-spermatids but only 50% of pR-sperma-

tids were assumed to degenerate during the spermio- genesis.

Both distributions of sex-ratios of

pm/PT

and

pm/

p R

males did not fit binomial distributions

(x2

= 42.1,

3 d.f., P

<

lo-' and

x2

= 66.3, 5 d.f., P

<

lo"',

respectively). In fact they presented an extra-binomial vari-

ance. SHARP and HILLIKER (1990) have shown that seg-

regation ratios within Segregation Distderlines of D. mela-

nogastercan conform to beta-binomial distributions, this means that the segregation ratio shows stochastic varia-

4 12 20 28 36 44 52 60 68 76 84 92 100

B 6 j 5 4

8 2 1

z'

0 - I ,- 1 - , " , - - - 7 - ~ , - l ~ i

4 12 20 28 36 44 52 60 68 76 84 92 100

Female percent in progeny

FIGURE 3.-Suppressor effect of

PR

chromosomes. Male genotype and mean female percentage are indicated in the corresponding histogram. Abscissa, percentage females in progenies; ordinate, number of progenies.

tions, following a beta distribution, between males of the same genotype, and that the sampling adds a binomial variability. Using the same methods with our data,

we found that a beta-binomial distribution can be fitted

to each of the observed distributions of sex-ratios (best

fit for

pm/pr:

p = 0.812, g 2 = 0.007,

x'

= 5.58, 5 d.f.,

P = 0.13 and for

pm/r"FK:

p = 0.664, u2 = 0.005,

x 2

=

5.42, 4 d.f., P = 0.25). Therefore, both the ST and SR

strains were assumed to be monomorphic for the Y

chromosomes regarding their drive sensitivity.

Test for autosomal drive suppressors in the

SR

strain: Figure 4 presents the distributions of the per-

centages of females among

PR/

Ft ' males bearing dif-

ferent combinations of autosomes from the

SR

strain.

Heterozygous IfR/ZPet IIfR/ZZT males produced a mean

of 81,3% females, males homozygous only for the IFR

or the IZfR chromosomes produce, respectively, 66.4%

and 60.7% females and males double homozygous IFR/

I f K IIpR/IIfR show an equal sex-ratio (52.6% females). Therefore both the IFR and the I I f R chromosomes carry autosomal drive suppressors. When compared by a

Mann-Whitney test (rank sum test) the ZIPR chromo-

somes present more powerful suppressors of drive than

the IFR chromosomes (uObr = 2.249, P = 0.025); either

of the two autosomes can restore a normal sex-ratio in the homozygous state in some cases. The combination of I f K and I I f R suppressors (Figure 4D) restores an even sex-ratio in all but two cases. In fact, for each of the four chromosomal combinations tested, there is a great variability for drive ability among males, which

suggests a polymorphism of

ZFR

and ZZFR for their sup-

pression ability. Alternatively, this variability may be due

to a polymorphism of

p R

sensitivity to the autosomal

(6)

640 M. Cazemajor, C. Landre and C. MontchampMoreau

4 12 20 28 36 44 52 60 68 76 84 92 100

$

287B

4 12 20 28 36 44 52 60 68 76 84 92 100

L

4 12 20 28 36 44 52 60 68 76 84 92 100

4 12 20 28 36 44 52 !6 68 76 84 92 100

Female percent In progeny

FIGURE 4.-Suppressor effect of SR autosomes. Male genotype and mean female percentage are indicated in the corresponding histogram. Abscissa, percentage females in progenies; ordinate, number of progenies.

DISCUSSION

An X-linked complex of smratio genes probably oc-

curs in D. simulans: Our results clearly show that several

X-linked genes are involved in the sex-ratio trait ex-

pressed by the

pm

chromosomes. A major role is played

by loci in the nplr region that is

-

16 cM long. Neverthe- less, additional gene(s) outside of this region may be

involved. Data on

pm

chromosomes may reflect the

general structure of sex-ratio X chromosomes of the SR

strain, since similar results were obtained using

p R

chromosomes instead of

ym

ones (data not shown).

Because several loci appear to be implicated in distor-

tion, we cannot exclude that some of them were polymorphic among

p R

and also among

x""5

(x"R5 driving

factors come from a unique X chromosome of the SR

strain, but extensive recombination was allowed with

xs'

chromosomes). Such a polymorphism could result

in a minor drive effect and explain the variations in drive ability of these chromosomes between the experi-

ments (Figure IC us. 4A for

pR,

and Figure 2B us.

3A for

pm).

However it must be pointed out that in

experiments using the SR5 strain, a strong drive ability

can be recurrently recovered among the second genera-

tion male progeny of weakly distorting males (com-

ments of Figure 2 and data not shown). In addition,

some

pn

chromosomes isolated in lines with attached-

Xfemales and a standard genetic background exhibit variable drive ability (data not shown). Thus, variation

in drive ability of

pn

and

pm

chromosomes may be

due as well to variable expressivity of the driving factors.

Enhancers of drive are expected to evolve on sex-ratio

X chromosomes since any increase in distortion will be

favored unless associated with fertility loss (THOMSON

and FELDMAN 1975). Therefore, if several loci acting on

drive are polymorphic in a population, no matter where they are located on the Xchromosome, the allelic com-

bination inducing the greatest distortion will be se-

lected for, unless fitness is affected. The effect of linkage strength between X-linked driving loci on their fate within populations has not yet been analyzed. However, intuitively, when a new allelic combination occurs that results in a stronger driver X chromosome, the more linked the responsible loci are the more easily this combination will be selected for, since it will be split up less frequently by crossing-over events. This may lead

to complexes of sex-ratio genes as we suppose to be the

case in

D.

simulans. In Drosophila species where inver-

sions prevent crossing-over along the major part of the Xchromosome, more than one region of this chromo-

some is generally involved in the sex-ratio trait ( STALKER

1961; Wu and BECKENBACH 1983). The lack of inversion

polymorphism in

D.

simuluns probably explains that a

single and small region on the Xchromosome contains most of the factors responsible for the drive.

A multifactorial system restores an equal sex-ratio

in

the

SR

s t r a i m We showed that x"m drive is weaker

against

p R

chromosomes than against

P''

chromo-

somes. However, it is not possible by now to determine

whether this difference results from differential sensitiv-

ity of the target sequences or from the presence of

suppressors on I*iR chromosomes. We arbitrarily de-

cided to use the terminology of Flinked suppressors of

drive as in other Drosophila species (STALKER 1961; DE

CARVALHO and KLACZKO 1994). Drive resistant Y chro-

mosomes are not unexpected since they are subject to

strong selection in natural populations carrying sex-ratio

Xchromosomes (THOMSON and FELDMAN 1975). CLARK

(1987) has shown, through theoretical models, that

conditions for a stable Y polymorphism are very un-

likely. Stable polymorphism would only occur in the presence of X polymorphism for meiotic drive ability

associated with differences in males and females fit-

(7)

Sex-Ratio in D. simulans 641

the parameters allowing a stable polymorphism are very

restrictive. Polymorphism of X and Y have been de-

scribed in D. paramelanica (STALKER 1961), but it is not

clear whether or not this is an intrapopulation polymor-

phism, given that the different types of Y and X seem

to only partially overlap in their geographical distribu-

tions. DE CARVALHO and KLACZKO (1994) described an

intrapopulation polymorphism for Y-linked sex-ratio

suppressors but did not know whether the polymor-

phism was stable or not. Moreover, in both D. parama-

lanica and D. mediopunctata, Y-linked suppressors can restore an equal sex-ratio and the described polymorphism was Yuwessor and Y’onruwsso7

.

In D. simulans the

pR

chromosomes were found to partially suppress

PR5

drive. Sex-ratio distortion among individual progenies

of FR5/pR males fits a beta-binomial distribution, which

suggests there could be no polymorphism of pR regard-

ing their suppressor effect against

pm

but that sex-ratio

expression varies stochastically between males as shown

for Segregation Distorter in D. melanogaster (SHARP and HILLIER 1990). However, we cannot exclude that the

Figure 3B shows a bimodal distribution (two peaks

around 62 and 7296, respectively), resulting from two

kinds of EdR that would present a very slight difference

in their suppressor effect. More

p R

chromosomes have

to be tested to look for such a polymorphism. Alterna-

tively, many uncontrolled factors resulting from the S T

or SR5 background may explain the extra-binomial vari-

ability of sex-ratio shown by both

p R 5 / p T

and PR5/pR

males.

On the other hand, we showed that autosomal drive suppressors, possibly polymorphic, occur on the two major autosomes in the SR strain. Such suppressors are expected to evolve in response to the sex-ratio bias in

populations (Fisher’s principle in FISHER 1930). Condi-

tions for their polymorphism do not seem stringent

(WU 1983a). Autosomal suppressors have been re-

ported in D. mediopunctata and D. paramelanica, but they

appear to have a restricted effect in these species when

compared to Flinked suppressors. In D. mediopunctata,

partially dominant autosomal modifiers, described on all major autosomes, together can lead to complete autosomal suppression of drive as shown in a line devoid of Y-linked suppressors that had been selected for non-

distorting males (DE CARVALHO and KLACZKO 1993).

However in the wild most of the males bearing the

chromosomal arrangement associated with the sex-ratio

trait seem to express it (DE CARVALHO et al. 1989). Simi-

larly, the autosomal suppression in D. paramelanica con-

sists of a so-called “minor suppression system” that

seems efficient in the wild, at least in some males, but

only partially suppresses drive (STALKER 1961). Unlike

the previously described autosomal suppressors, those in D. simulans seem recessive and appear responsible for a “major effect”: the sex-ratio trait is well expressed

(90% females) when SR autosomes are in a heterozy-

gous state and a complete SR autosomal sets usually

restores an equal sex-ratio. Probably because in other

Drosophila species studies only led to the recovery of

dominant autosomal suppressors of drive (DE

CAR-

VALHO and KLACZKO 1993), theoretical studies on the invasion of recessive autosomal suppressors were never performed. Such suppressors would have to be homozygous to be selected, which may occur in small or struc- tured populations. In our case it is not clear whether autosomal suppressors are completely recessive or not. In fact the proportions of sexes vary among progenies

of SR/ST hybrid males (Figure 1C us. 4A). FP hybrid

males studied on Figure

4A

are sons of F1 sex-ratio males,

and the more the sex-ratio in the progeny of a F, male

was skewed toward females, the more his F2 sons where

sterile (MERCOT et al. 1995 and unpublished results).

Therefore it is possible that we selected combinations of autosomes with dominant suppressors.

In the SR strain, we found more suppressors than

necessary: on the one hand autosomal suppressors are sufficient to restore an equal sex-ratio in the progeny

of

PR

males, on the other hand, Ychromosomes them-

selves partially resist the drive.

By

now, it is not possible to determine how such a complex regulation system of

sex-ratio drive evolved in D. simulans. Assuming that more potential modifier sites occur on autosomes, autosomal suppressors of drive are more likely to appear than Y-linked ones. Alternatively, drive insensitivity of

Y chromosomes can result from modifications of the

target. Since the Y chromosome is mainly heterochro-

matic, an analogy can be made with the SD system,

where the sensitivity of the heterochromatic target locus

(Rsp)

correlates with repeat copy number (WU et al.

1988). If so, Ychromosome insensitivity should be more likely to evolve than autosomal suppression. When drive is rare, selection on resistant Ychromosomes is stronger

than on autosomal suppressors (HURST and POMIA-

NKOVSKI 1991), which could give the former an advan-

tage. Then it is probable that drive-resistant

pR

chromo-

somes were fixed before autosomal suppression was

able to completely repress the drive.

The common pattern among Drosophila species is

that sex-ratio X chromosomes are at low frequencies in

natural populations whereas drive is not or only partially suppressed, as discussed above. This implies that the spread of these chromosomes has been halted by countervailing selection on individuals that carry them.

A reduced fertility of sex-ratio males is commonly ob-

served when there is a high rate of mating (WALLACE

1948;

BECKENBACH

1981; Wu 1983b,c; JAENIKE 1996).

In addition, in species where the trait is associated with

chromosomal inversions, females homozygous for the

sex-ratio arrangement suffer reduction in fitness (CURT-

(8)

642 M. Cazemajor, C. Landri. and C. Montchamp-Moreau

curs within these populations and strains (MERCOT et

al. 1995; ATLAN et al. 1997). This strongly suggests that the fitness of sex-ratio X bearing flies is little or not at all affected in D. simulans. Given the lack of inversions in D. simulans, this supports the hypothesis that in other species, the low fitness of flies carrying sex-ratio factors tied up within chromosomal inversions is due to linked lethal alleles and not to the driver loci themselves.

We thank C A R O I , ~ E ENGEL., DENIS POINSOT, and an anonymous referee for helpful comments on the manuscript. This work was sup- ported by a grant from Ministere de I’Enseignement Supkrieur et de la Recherche, ACC SV3 no. 9503017.

LITERATURE CITED

ATLAN, A,, H. MERCOT, C. LAND@ and C. MONTCHAMP-MOREAU, 1997 The sex-ratio trait in Drosophila simulans: geographical distribution of distortion and resistance. Evolution (in press).

BECKENBACH, A. T., 1981 The sex-ratio trait in Drosophila pseudoobsc-

ura: fertility relations of males and meiotic drive. Am. Nat. 112:

97-117.

C L m , A. G., 1987 Natural selection and Y-linked polymorphism. Genetics 115: 569-577.

CURTSINGER, J. W., and M. W. FELDMAN, 1980 Experimental and theoretical analysis of the sex-ratio polymorphism in Drosophilapseudo- obscura. Genetics 94: 445-466.

IIE CARVAL.HO, A. B., A. A. P E I X ~ T O and L. B. KIACZKO, 1989 Sex-

ratio in Drosophila mediopunctata. Heredity 62: 425-428.

DE CARVALHO, A. B., and L. B. KLACZKO, 1993 Autosomal suppressors of sex-ratio in Drosophila mediopunctata. Heredity 71: 546-551.

DE CARVALHO, A. B., and L. B. KLACZKO, 1994 Flinked suppressors of sexratio in Drosophila mediopunctata. Heredity 73: 573-579.

DE MAGAI.H.&S, L. E., I. M. ROVERONI and S. H. A. CAMPOS, 1985 Oc- currence of the sex-ratio trait in natural populations of Drosophila simulans in Brazil. Rev. Brasil. Genet. 8: 449-456.

DOBZHANSW, T., and C. EPLING, 1944 Contributions to the gmetics

taxonomy and ecology of Drosophila pseudoobscura and its relatives.

Carnegie Inst. Washington Pub. 554.

DOBZHANSKY, T., 1958 The genetics changes in populations of Dro- sophila pseudoobscura in the American South-West. Evolution 12:

385-401.

FAULHABER, S. H., 1967 An abnormal sex-ratio in Drosophila sim-

ulanr. Genetics 56: 189-213.

FISHER, R. A,, 1930 The Genetical T h e q of Natural Selection. Oxford University Press, Oxford.

GERSHENSON, S., 1928 A new sex-ratio abnormality in Drosophila obsc-

ura. Genetics 13: 488-507.

HAMILTON, W. D., 1967 Extraordinary sex-ratios. Science 156: 477- 488.

HAIJSCHTECK-JUNGEN, E., 1990 Postmating reproductive isolation and modification of the sex-ratio trait in Drosophila subobscura in- duced by the sex chromosome gene arrangement Ay+3+5+7. Ge- netica 83: 31-44.

HURST, L. D., and A. POMLANKOWSKY, 1991 Causes of sex-ratio bias may account for unisexual sterility in hybrids: a new explanation of HAIDANE’S rule and related phenomena. Genetics 128: 841-

858.

JAENIKE J., 1996 Sex-ratio meiotic drive in the Drosophila quinaria

group. Am. Nat. 148: 237-254.

JAMES, A. C., and J. JAENIKF., 1990 Sex-ratio in Drosophila testacea. Genet- ics 126: 651-656.

JUNGEN, H., 1968 Sexratio in naturlichen populationen von Drosoph-

ila subobscura. Arch. Jul Klaus-Stift. 43: 52-57.

LYTTLE, T. W., 1979 Experimental population genetics of meiotic drive systems. I1 Accumulation of genetic modifiers of Segregation

Distorter (SD) in laboratory populations. Genetics 91: 339-357. MERCOT, H., A. ATIAN, M. JACQUES and C. MONrCHAMP-MOREALI,

1995 Sex-ratio distortion in Drosophila simu1an.r: co-occurrence of a meiotic drive and a suppressor of drive. J. Evol. Biol. 8: 283- 300.

M O N r ( : w P - M o m u , C., and D. JOLY, 1997 Abnormal spermiogen- esis is associated with the X-linked sex-ratio trait in D. simulans.

Heredity (in press).

MORGAN, T. H., C. B. BRIDGES and T. STURTEVANT, 1925 The genetics of Drosophila. Bibliogr. Genet. 2: 1-262.

SHARP, C. B., and A. J. HILLIKER, 1990 Segregation ratios within Seg-

regation Distorter lines of Drosophila melanogasterconform to a beta- binomial distribution. Genet. Res. Camb. 55: 165-170.

STALKER, H. D., 1961 The genetic systems modifying meiotic drive in Drosophila paramelanica. Genetics 46: 177-202.

STIJRTEVANT, A. H., and T. DOBHANSKY, 1936 Geographical distribution and cytology of sex-ratio in Drosophila pseudoobscura and related species. Genetics 21: 473-490.

THOMSON, G. J,, and M. W. FEI.DMAN, 1975 Population genetics of modifiers of meiotic drive. I V . On the evolution of sex-ratio distortion. Theoret. Popul. Biol. 8: 202-211.

VOELKER, R. A,, 1972 Preliminary characterization of sex-ratio and rediscovery and reinterpretation of male sex-ratio in Drosophila afjnis. Genetics 71: 597-606.

WAI.IAC:E, B., 1948 Studies on sexratio in Drosophila pseudoobscura. I. Selection and sex-ratio. Evolution 2: 189-217.

WU, C.-I., 1983a The fate of autosomal modifiers of the stx-ratio

trait in Drosophila and other sex-linked meiotic drive systems. Theoret. Popul. Biol. 2 4 107-120.

WU, C.-l,, 19831, Virility deficiency and the Sex-Ratio trait in Drosoph- ila pseudoobscura. I. Sperm displacement and sexual selection. Genetics 105: 651-662.

Wu, C.-I., 1983c Virility deficiency and the Sex-Ratio trait in Drosoph- zlapseudoobscura. 11. Multiple mating and overall virility selection. Genetics 105: 663-679.

Wu, C.-I., and A. T. BECKENBACH, 1983 Evidence for extensive genetic differentiation between the Sex-Ratio and the Standard arrangement of Drosophila pseudoobscura and Drosophila p m i m i s

and identification of hybrid sterility factors. Genetics 105: 71- 89.

Wu, C . 4 , T. W. LYITLE, M.-L. Wu and G.-F. LIN, 1988 Association between a satellite DNA sequence and the Responderof Segregation Distorler in Drosophila melanogaster. Cell 5 4 179-189.