Does Stellate Cause Meiotic Drive in Drosophila melanogaster?

Corresponding author: Leonard G. Robbins, Università di Siena, Via Aldo Moro 2, 53100 Siena, Italy., robbins{at}unisi.it (E-mail)

Communicating editor: K. GOLIC

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Drosophila melanogaster males deficient for the crystal (cry) locus of the Y chromosome that carry between 15 and 60 copies of the X-linked Stellate (Ste) gene are semisterile, have elevated levels of nondisjunction, produce distorted sperm genotype ratios (meiotic drive), and evince hyperactive transcription of Ste in the testes. Ste seems to be the active element in this system, and it has been proposed that the ancestral Ste gene was "selfish" and increased in frequency because it caused meiotic drive. This hypothetical evolutionary history is based on the idea that Ste overexpression, and not the lack of cry, causes the meiotic drive of cry^- males. To test whether this is true, we have constructed a Ste-deleted X chromosome and examined the phenotype of Ste^-/cry^- males. If hyperactivity of Ste were necessary for the transmission defects seen in cry^- males, cry^- males completely deficient for Ste would be normal. Although it is impossible to construct a completely Ste^- genotype, we find that Ste^-/cry^- males have exactly the same phenotype as Ste⁺/cry^- males. The deletion of all X chromosome Ste copies not only does not eliminate meiotic drive and nondisjunction, but it also does not even reduce them below the levels produced when the X carries 15 copies of Ste.

WHEN Sandler and Novitski first invented the term "meiotic drive" (SANDLER and NOVITSKI 1957 ) to refer to any alteration of the meiotic process that distorts the normal 1:1 ratio of reciprocal gametes, it was proposed as a novel, perhaps powerful, evolutionary force. Although numerous examples of meiotic drive are now known in Drosophila and other species (for reviews, see LYTTLE 1991 , LYTTLE 1993 ), demonstrating the evolutionary effect of any single meiotic drive element has proven difficult (HURST and POMIANKOWSKI 1991 ; LAURIE 1997 ; CARVALHO and VAZ 1999 ). It is even more difficult to show that meiotic drive of a particular genetic element has consequences for the host genome rather than for just that particular gene.

HURST 1992 , HURST 1996 has proposed that Stellate (Ste, located at salivary map position 11E1-2) is an evolutionarily important meiotic driver or, more exactly, that the Ste alleles in the modern genome are a "relict meiotic driver." Ste is a particularly attractive candidate because its sex linkage implies that Ste-induced meiotic drive is likely to have a profound effect on population structure and because its molecular biology suggests a plausible evolutionary history.

Ste is one of two repeated-sequence loci involved in this system; drive occurs only in males that are deleted for the crystal (cry) locus of the heterochromatic Y chromosome (HARDY et al. 1984 ; LIVAK 1984 ; PIMPINELLI et al. 1986 ). In cry^- males that have a low-copy-number Ste locus (termed Ste⁺ alleles), there is substantial nondisjunction and strongly distorted recovery of gamete classes. For the sex chromosomes, the result is an excess of X-bearing sperm over Y-bearing sperm, and an even more striking excess of nondisjunctional sperm that bear no sex chromosome over the reciprocal products that carry both the X and Y chromosomes. If the X chromosome carries many copies of the Ste sequence (a Ste allele), cry-deletion males are sterile, but males with a high Ste copy number and a partial deletion of cry also evince nondisjunction and drive. Hence, in a population of cry^- males, the frequency of chromosomes carrying one or a few copies of Ste would tend to increase, and the sex ratio would be seriously distorted.

Both of these loci are composed of tens to hundreds of copies of repeated sequences. The Ste sequence encodes an analog of the ß-subunit of casein kinase II (LIVAK 1990 ; BOZZETTI et al. 1995 ). The cry repeat is very similar to that of Ste, but it contains, in addition, a nearby copy of the hoppel transposable element and a cry-specific sequence outside of the coding region (BALAKIREVA et al. 1992 ). There is substantial variation among the copies of cry (MCKEE and SATTER 1996 ; KOGAN et al. 2000 ), and it has been proposed that cry suppresses Ste expression by producing an interfering RNA (ARAVIN et al. 2001 ). A few intact copies of Ste are also located in the pericentromeric heterochromatin of the X (SHEVELYOV 1992 ; TULIN et al. 1997 ). Ste transcripts are produced in males with a normal Y chromosome, as well as in females (SCHMIDT et al. 1999 ), but in cry^- males a distinct Ste transcript is produced and translated at high rates in the testes, and massive crystalline inclusions of the protein product are found in the spermatocytes (LIVAK 1984 ; BOZZETTI et al. 1995 ). The protein crystals are needle shaped if Ste copy number is low and star shaped if Ste copy number is high. Because the Stellate phenotype is absent in cry⁺ males, cry is also known as Su(Ste).

Because deletion of the Y chromosome of the sibling species Drosophila simulans does not provoke protein accumulation in spermatocytes, and because the cry repeat seems to have been invaded by a transposable element, it is parsimonious to assume that the ancestral D. melanogaster Y chromosome was cry^-. That, plus the observed correlation of Ste copy number with the level of meiotic drive (PALUMBO et al. 1994 ), would yield an evolutionary history of pressure toward increased Ste copy number, counterbalanced by accumulation of (degenerate) cry sequences in the Y chromosome that suppress the drive (HURST 1992 ). There are two critical elements in this hypothesis: (1) It is the hyperactivation of Ste, and not the absence of cry, that causes meiotic drive in cry^- males, and (2) Ste is the more ancestral of the two loci.

Although attractive, some earlier observations do not conform well to this evolutionary hypothesis. First, although meiotic drive of the sex chromosomes in Ste⁺ cry^- males would increase the frequency of Ste-bearing X chromosomes, that same genotype also produces autosomal nondisjunction and autosomal meiotic drive and severely reduces fertility (PALUMBO et al. 1994 ). Whether the frequency of a newly arisen Ste⁺ X chromosome would increase in a population, or would quickly disappear, thus depends on the balance between the fertility reduction and the excess production of X-bearing sperm. Second, although the level of meiotic drive is significantly correlated with Ste copy number, the slope of that correlation is shallow, explains only a small fraction of the variation in drive, and does not pass through the origin (PALUMBO et al. 1994 ; ROBBINS et al. 1996 ). In other words, those data suggest that deletion of cry would result in nearly the same level of meiotic drive even if the X chromosome did not carry a single copy of Ste.

To decide whether it is Ste activity or whether it is the deletion of cry that causes meiotic drive, we must know what happens if the X chromosome does not carry any Ste copies. If Ste is a meiotic driver, or even a relict meiotic driver, deleting it should also eliminate the drive, even in cry^- males. Palumbo has made several ingenious, large-scale attempts to generate a Ste-deleted X chromosome (PALUMBO et al. 1994 and personal communication) without success. He did, however, find one chromosome in a natural population that has no intact copies of the euchromatic Ste sequences (PALUMBO et al. 1994 ), and we have been able to use that as starting material for the construction of a Ste-deleted X chromosome. If Ste activity is indeed the cause of the meiotic drive seen in cry^- males, drive should be absent in males with the engineered Ste^- chromosome.

There is, however, a caveat to this simple prediction of complete absence of drive in Ste^- X/cry^- Y males. The only cry^- Y chromosome available is marked by a translocated B^S segment of the X chromosome that itself carries some Ste copies (PALUMBO et al. 1994 ), which may or may not be transcribed. Moreover, we cannot remove that piece of X chromosome because it also serves to cover the lethality of a deficiency needed for the construction of the Ste^- X chromosome. Hence, although we have succeeded in constructing a Ste-deleted X chromosome, we cannot construct a completely Ste-deficient genotype and there may be some residual Ste activity in the Ste^- X/cry^- Y males. The availability of a Ste-deleted X chromosome, however, has made direct measurement of the activity of the Ste copies in the B^S segment possible, and we show here that these copies are indeed active. Deleting all Ste copies from the X chromosome is therefore expected to reduce, rather than obliterate, meiotic drive, and this must be taken into account in the design of the genetic experiments.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Construction of a Ste-deleted X chromosome:
Except as indicated here, descriptions of the chromosomes and markers used may be found in LINDSLEY and ZIMM 1992 . As shown in Fig 1, three chromosomes were used as the starting materials for the construction. W12 is a wild-type X chromosome from a natural population that carries 15 copies of the Ste sequence. All these copies have the restriction pattern characteristic of the heterochromatic Ste block. Moreover, there is only an extremely weak in situ hybridization signal at the euchromatic Ste site of the W12 chromosome, indicating that only a partial copy of Ste is present there (PALUMBO et al. 1994 ). This euchromatic-deletion allele is denoted in Fig 1 as Ste^W12. The heterochromatic copies were then removed by selecting crossovers between the W12 chromosome and a Df(1)X1 chromosome. Df(1)X1 removes most of the basal heterochromatin, including all of the ribosomal RNA genes (rDNA), and extends a short distance into the euchromatin. It therefore entirely lacks the heterochromatic Ste block located between the heterochromatic-euchromatic junction and the rDNA. Df(1)X1 is inviable in homozygous females and hemizygous males because it is deleted for two essential euchromatic genes, but survives in males that bear a Y chromosome that covers the missing euchromatic segment (e.g., B^SY). The Ste^W12 Df(1)X1 chromosome is thus a synthetic deficiency for all of the X-linked Ste sequences. It is, however, also deficient for the rDNA, and the rDNA deficiency itself would result in random chromosome segregation and strong meiotic drive (MCKEE and LINDSLEY 1987 ). Thus, an rDNA array was introduced into these chromosomes by a second round of recombination. The rDNA array used, denoted Tp(1;1)NO, is an rDNA array stripped of all surrounding heterochromatin that has been transposed to a point adjacent to the w gene. It was generated by a Rex-induced mitotic reinversion of In(1)w^m51bLw^m4R. The reinverted chromosome, Dp(1;1)w^m51bLw^m4R, contains a naked rDNA array near w and a reconstituted, structurally normal, heterochromatic region at its base that contains another block of rDNA (ROBBINS and SWANSON 1988 ). The particular Tp(1;1)NO used is a large array that contains 390 copies of the rDNA, of which 100 are free of inserted transposons (CRAWLEY 1996 ). It is bb⁺, promotes completely normal chromosome pairing, and does not provoke any meiotic drive (L. ROBBINS and P. CRAWLEY, unpublished data). The result of the two sequential crossovers is therefore a Ste-deficient, rDNA⁺ chromosome. To permit assessment of background genotype effects on meiotic behavior, eight independently derived Ste^- chromosomes were used. Six independent crossovers between W12 and Df(1)X1 were selected, and for two of these, two independent crossovers that introduced the Tp(1;1)NO were used.

View larger version (24K): In this window In a new window Download PPT slide   Figure 1. Construction of Ste-deficient X chromosomes. Elements from three chromosomes were combined to create chromosomes carrying at most one euchromatic copy of Ste, a complete deficiency of the heterochromatic Ste block, and a transposed rDNA array. Ste and rDNA copy numbers are indicated where known.

Measuring Ste transcription:
Fifty pairs of adult testes from each genotype were hand dissected in modified Ringer's solution. The testes were homogenized in 0.9 ml of 0.1 M NaCl, 0.1 M Tris-base, 0.03 M Na2 EDTA, and 1% Sarkosyl and were extracted with phenol:chloroform:isoamyl alcohol (50:49:1). RNA was precipitated with 2.5 volumes of ethanol and stored at -20°. Samples were dissolved in 20 µl of a solution containing (in 1 ml) 500 µl formamide, 178 µl 37% formaldehyde, 100 µl 10x 3-(N-morpholino)propanesulfonic acid (MOPS) buffer, 60 µl bromo-phenol-blue, and 162 µl sterile water and heated at 65° for 15 min. RNA samples were separated on a 1% agarose-MOPS-formaldehyde gel and transferred overnight to a nylon membrane (Hybond N, Amersham Pharmacia). After washing, the filter was hybridized overnight at 42° with a labeled Stellate DNA probe (pUSTE; BOZZETTI et al. 1995 ) in 25 ml of 50% formamide, 5x SSC, 5x Denhardt's, 0.5x SDS, 10 mM EDTA (pH 8), and 100 µg/ml herring sperm DNA. After hybridization, the filter was washed at 65° for 15 min in 2x SSC and 0.1% SDS and then in 1x SSC and 0.1% SDS for 15 min. Autoradiography was performed on Alfa Curix 100 film.

Films were scanned with a flat-bed scanner and the 8-bit TIFF files imported into ImageQuant 5.0 (Molecular Dynamics, Sunnyvale, CA) for integration. Areas were identified using the Spot Finder protocol, and volumes were integrated using the histogram peak (of the periphery) for background correction. To check for film saturation, area integration was also done for each lane and truncated peaks noted on the plot.

Design of the experimental and control crosses:
To know if deleting Ste from the X chromosome completely abolishes meiotic drive, we would have to compare the meiotic behavior of only two genotypes: Ste^W12/cry^- Y and Tp(1;1)NO Ste^W12 Df(1)X1/cry^- Y males. However, because the B^S duplication of the cry^- Y (= B^Scry¹Y·y⁺) carries some X chromosome-derived Ste copies, it is possible that only drive would be diminished even if Ste activity directly causes the sperm dysfunction. Minor changes in level of drive might be masked or might erroneously be thought to exist because of zygotic viability differences provoked by the different structures and markers of these two X chromosomes. Ste^W12/B^Scry⁺Y·y⁺ and Tp(1;1)NO Ste^W12 Df(1)X1/B^Scry⁺Y·y⁺ males were therefore also tested to control for such marker effects.

To reduce the risk that background genotype differences might be misread as an effect of deleting Ste, eight independently derived Ste^- chromosomes were tested. To distinguish between casual and genotypic differences in phenotype, some of the crosses were replicated. In addition, because parental-source (imprinting) effects on heterochromatic phenomena are known (SPOFFORD 1976 ; GOLIC et al. 1998 ; LLOYD et al. 1999 ), in several of the replicates we separately tested males derived from the reciprocal crosses C(1)DX/B^SYy⁺ x Ste^-/Y and Ste^-/FM7 x X/B^SYy⁺.

Statistical analysis:
In addition to normal X-bearing and Y-bearing sperm, nondisjunction in cry^- males yields XY and nullo sperm, and meiotic drive results in unequal recoveries of the reciprocal types. Following MCKEE 1984 , these results can be described in terms of the frequency of disjunction (D), the survival (recovery) of sperm that carry an X chromosome (R_X), and the survival of sperm that carry a Y chromosome (R_Y). The frequencies of the four sperm classes among all sperm are , and and the solutions for the three parameters are , and . These estimators do not, however, take account of possible viability differences of the structurally normal Ste⁺ chromosome and the rearranged and differently marked Ste^- chromosome. To this end, the ratio of regular males to regular females in the control (cry⁺ Y) crosses was used to estimate relative survival (S), and the MLIKELY.PAS maximum-likely analysis program (ROBBINS 2000 , downloadable at http://www.unisi.it/ricerca/dip/bio_evol/sitomlikely/mlikely.html) was used to obtain estimates of D, R_X, and R_Y corrected for marker effects and to test for statistical significance. Maximum-likelihood analysis was also used to compare variation of the parameters among independently derived examples of the Ste^- chromosomes.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Ste transcription:
The Northern blot results are shown in Fig 2. There are several Ste transcripts (SCHMIDT et al. 1999 ). A 750-nt transcript is found only in the testes of cry^- males or in the testes of males homozygous for any one of several recently identified male-sterile mutants (SCHMIDT et al. 1999 ; MCKEE et al. 2000 ; TRITTO 2000 ; STAPLETON et al. 2001 ). Two other Ste transcripts are found in both males and females and in somatic tissues, whether cry is deleted or not. These are an 8-kb transcript (not followed in this experiment) and a 1400-nt product. In cry^- males with high-copy-number Ste alleles, the 750-nt testis-specific, cry-specific RNA is much more abundant than the ubiquitous 1400-nt transcript, but in fertile Ste⁺/cry^- males with low or moderate Ste copy numbers they are of similar abundance.

View larger version (68K): In this window In a new window Download PPT slide   Figure 2. Quantification of Northern blot analysis of Ste transcription. Transcript abundance was assessed in testes from six genotypes, including cry- males bearing two independently constructed Ste- X chromosomes. The scanned images for the Ste and rp49 hybridizations are shown along with the areas identified by the Spot Finder protocol of ImageQuant 5.0. The Ste copy numbers indicated are taken from PALUMBO et al. 1994 . The larger 1400-nt transcript is present in both cry+ and cry- testes (as well as in the soma and in females; SCHMIDT et al. 1999 ), and the ratio of this ubiquitous transcript to the rp49 control is nearly constant. The smaller 750-nt testes-specific transcript is completely absent in cry+ males and varies with Ste copy number in the other genotypes. Transcript levels relative to the values for SteW12/cry1Y are calculated as:

With respect to the question at hand, visual examination of the blot resolves the most important issue: Ste is transcribed from the transposed copies in Ste^-/cry^- males, but the level of the testes-specific 750-nt transcript is substantially lower than that in W12/cry^- males. Thus, if drive is caused by induction of Ste, we expect to find less drive in Ste^-/cry^- males than in W12/cry^- males. Quantification of the blots, however, is necessary to know how much amelioration to expect.

The quantity of the ubiquitous 1400-nt Ste transcript appears to be constant, and we can control for loading by comparing the level of the testes-specific transcript either with that transcript (eliminating potential differences between hybridizations, but assuming that the ubiquitous transcript is indeed constant) or with the rp49 transcript (introducing hybridization-to-hybridization variation, but making no assumption about the ubiquitous transcript). Both sets of estimates are similar.

Deleting the X-linked Ste copies results in a 40–50% reduction in the testes-specific transcript. These results also suggest that cry^--induced transcription of the heterochromatic copies, whether in their normal location in the X chromosome or transposed to the Y, is about one-third that of the euchromatic copies; W12/cry^-Y and y w f/0 have nearly the same total number of Ste copies, but noticeably different levels of the testes-specific transcript. That this is not caused by increased Ste activation in XO compared to cry¹ males is indicated by the high expression in y w f/cry¹ males (evident even though that level is underestimated because of film saturation).

Nondisjunction and meiotic drive:
The results of the crosses are summarized in Table 1. While it is quite obvious that meiotic drive is not obliterated when Ste is deleted from the X chromosome, the Ste⁺/cry^- and Ste^-/cry^- crosses are not homogeneous. If we do not take account of any viability differences between the Ste⁺ and Ste^- chromosomes, it appears that sperm survival (both R_X and R_Y) is somewhat higher when Ste is deleted, but nowhere near the 50% improvement implied by the 50% reduction in testes-specific Ste transcript. Closer examination of the data, however, indicates that even this slight difference is entirely artifactual. First, the apparent amelioration of spermatogenesis is not consistent; although sperm survival has improved, disjunction appears to be slightly less regular with the Ste^- chromosome. Although the differences are small, the maximum-likelihood analysis shown in the first line of Table 2 indicates that the apparent difference in disjunction and the apparent, but opposite, difference in drive are both statistically significant. Second, the proportions of regular males (S) in the cry⁺ control crosses, and the much lower fertility in both crosses involving the Ste^- chromosome, suggest that at least some of the difference may be unrelated to the Ste-cry interaction. Third, a look at the variation among independently derived Ste^- chromosomes reveals background genotype differences at least as large as the difference between the Ste⁺ and Ste^- cases. The second and third points are considered in the following analyses.

Taking account of viability differences:
The possibility of differences in marker effects was examined by a maximum-likelihood analysis, summarized in the second line of Table 2, that used both the experimental and control data. Viability effects can be described in several different ways. In both control crosses, viability of the y/B^SYy⁺ sons is depressed and, for the analysis shown in Table 2, it was assumed that only the viability of the y/B^SYy⁺ males is reduced. It is possible, however, that the B^SYy⁺ chromosomes have a dominant effect on viability, in which case the survival of the X/X/B^SYy⁺ nondisjunctional daughters would also be reduced. It is also conceivable that all males, and not only y/B^SYy⁺ males, have reduced viability, in which case the survival of the X/O nondisjunctional sons would also be reduced. Although the analysis of only one model is illustrated here, the other two models were also examined and gave results that differ only in detail. There is a significant difference in marker effects between the Ste⁺ and Ste^- crosses. Once that is taken into account, however, no significant difference in the level of meiotic drive remains. The analysis proceeds as follows:

The probabilities of survival of each sperm genotype among all sperm with the relative survival of y/B^SYy⁺ included are , and for the cry^- crosses and and for the cry⁺ crosses. Each cry⁺ cross has one independent observation, and each cry^- cross has three, giving a total of four independent observations for each experimental/control pair. The unique solutions for the four parameters are then for the cry⁺ crosses, and , and for the cry^- crosses.

If all four parameters differ between the Ste⁺ and Ste^- crosses, the maximum-likelihood estimates of the parameters are the unique solutions for the eight unknowns. If the description, including the assumed independence of viability effects and meiotic parameters, is not unreasonable, D, R_X, and R_Y should fall between 0 and 1 and this hypothesis should provide an exact fit to the observations. It does.

To test for differences in marker effects of the Ste⁺ and Ste^- chromosomes, we evaluated an hypothesis in which D, R_X, and R_Y differ, but in which the relative survival of y/B^SYy⁺ males is assumed to be the same for both Ste⁺/cry^- and Ste^-/cry^- males. We found a highly significant difference, confirming that a difference in survival occasioned by the use of a structurally modified and differently marked chromosome is a significant source of the difference between the Ste⁺/cry^- and Ste^-/cry^- crosses.

Similarly, to test for a difference in disjunction, D is assumed to be equal in both the Ste⁺/cry^- and Ste^-/cry^- crosses while the other parameters are allowed to vary, and to test for a difference in meiotic drive, R_X, and R_Y are assumed equal while D and S are allowed to vary. Although disjunction differs slightly, but significantly, between Ste⁺/cry^- and Ste^-/cry^-, there is no indication whatsoever of any difference in the level of meiotic drive once marker effect differences are accounted for.

The apparent absence of an effect of deleting Ste seen in the foregoing is meaningful, however, only if the sample size is large enough, and the statistical test is powerful enough, to detect a difference if there were one. Otherwise, we merely have the classic negative result. To assess this, empiric support intervals for R_X, R_Y and jointly for R_X and R_Y were computed as described in ROBBINS (2000). Drive improvements of 28% for the X alone, 18% for the Y alone, or 11% for both would have yielded a significant difference. An improvement of over 40% as predicted from the transcription data would have been significant at a level of P < 1 x 10^-9.

Analysis of variation and parental-source effects:
The foregoing analysis depends, in part, on the correctness of modeling viability differences. Eight independently constructed Ste^- chromosomes were used in these experiments, however, and examination of the variation among these chromosomes, and among the replicate crosses done for several of them, provides an alternative check for differences, if any, between Ste^- and Ste⁺ behavior. The parameter values calculated for each of the chromosomes, without correcting for viability differences, are graphed in Fig 3. Two properties of the data set are evident upon inspection: (1) For all of the parameters, the differences between many of the individual examples of the Ste^- construct are as large as the differences between Ste⁺ and Ste^-, and (2) except for fertility, which is generally higher for the simple Ste⁺ chromosome than for any of the rearranged Ste^- chromosomes, tests of individual Ste^- chromosomes yield values both above and below those for Ste⁺.

View larger version (42K): In this window In a new window Download PPT slide   Figure 3. Variation among independently constructed Ste-chromosomes. Disjunction, meiotic drive, and fertility were assessed in cry- males, and sex ratio and fertility were measured in cry+ males for eight independently derived Ste- chromosomes. Comparison of these results with those observed for Ste+ males indicates that variation among Ste- stocks is at least as large as the Ste+-Ste- difference. The discrete multivariate analysis shown in Table 3, which includes analysis of variation among replicate tests of several of the chromosomes, demonstrates that most of this is variation of background genotype rather than environmental variation.

The results of a more formal look at the variation of D, R_X, and R_Y in the entire data set are summarized in Table 3. These results are from a discrete multivariate analysis in which the control, cry⁺, data were not used to correct for viability differences, but an analysis performed using that correction yields nearly identical results. There is a significant difference among replicate tests of the same genotype (H3 vs. H1); hence these parameters are at least somewhat sensitive to environmental variation. That variation has, however, nowhere near the magnitude of the differences between different genotypes (H2 vs. H3)—compare the marked difference in value of G/d.f. Partitioning the genotypic variation between the difference between Ste⁺ and Ste^- (H4 vs. H2) and the differences among Ste^- stocks (H4 vs. H3) indicates that these are of about the same magnitude (the slight difference in G/d.f. is exaggerated by the small number of degrees of freedom in the H4 vs. H2 comparison), confirming the results of the simpler analysis and eyeball test.

In six pairs of replicate Ste^-/cry^- crosses, the Ste^- chromosome arrived from the mother in one case and from the father in the other. Although further partitioning of G (not shown) indicates that the differences among these replicates are significant and comparable in magnitude to the differences among other replicate crosses, the difference cannot be ascribed to parental origin. Disjunction was lower following paternal origin for four reciprocal pairs and higher for two; recovery of X-bearing sperm was lower following paternal origin in three cases; and recovery of Y-bearing sperm was lower in two. Of all 18 comparisons, the phenotype was more severe following paternal origin in 9 and better in the other 9. Hence, there is no imprinting effect on the phenotype of cry^- males.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Meiotic drive was proposed originally as an evolutionary mechanism, but it has been difficult to demonstrate that any example of drive serves this role. Hurst has suggested that the cry-Ste system could be an evolutionarily important case. Paraphrasing his reasoning:

1. Transcription of Ste is induced in the testes of cry^- males. Hence the element whose frequency would be increased by drive is the active element in the system, and presumably the one that causes the drive.

2. X/O males of the sister species D. simulans do not have protein accretions in their testes; hence D. simulans does not have a Ste-suppressing cry locus in its Y chromosome, leading to the natural presumption that Ste is the older of the two elements.

3. Because there is no recombination between the X and Y chromosomes, an increase in the population frequency of Ste would have a strong effect on the evolution of the entire chromosome.

This is an attractive hypothesis, but its correctness depends on the validity of the first two points. It is not obvious, however, that hyperactivity of Ste necessarily implies that it is the element that causes the drive, nor does the lack of crystals in spermatocytes of D. simulans XO males necessarily mean that cry is not present somewhere in that species' genome.

In the experiments reported here, we constructed a Ste-deleted X chromosome to test the first of these premises. We cannot find the slightest hint that Ste is necessary for the meiotic drive that occurs in cry^- males. It is the absence of cry, and not hyperactivation of Ste, that causes meiotic drive. To get around this conclusion would require that the Ste activity induced in the B^SY-born copies is sufficient to provoke drive and that there is no response to doubling its Ste copy number and testes-specific Ste transcription, even though there is a linear response to further increases in Ste copy number (PALUMBO et al. 1994 ; ROBBINS et al. 1996 ). For now, it seems preferable to think that deletion of cry triggers the downstream problems of sperm function, but that the downstream targets, if they include Ste at all, must include other elements as well. We do not lack for candidates: the cis-responsive element that MCKEE 1987 identified in translocation-provoked meiotic drive, recessive autosomal mutants that accumulate Ste protein (SCHMIDT et al. 1999 ; MCKEE et al. 2000 ; TRITTO 2000 ; STAPLETON et al. 2001 ), or the transposable elements that are reported to respond to cry deletion (ARAVIN et al. 2001 ).

If the role of cry⁺ is not merely the suppression of a parasitic Ste element, what does it do? Perhaps cry⁺ and the other genes whose mutants mimic cry^- form part of a general gene-silencing system (LIFSCHYTZ and LINDSLEY 1972 ) or part of a system aimed particularly at silencing repeated elements (BIRCHLER et al. 2000 ; ARAVIN et al. 2001 ) that must act before spermiogenesis can proceed. If so, meiotic drive is not the result of a normal checkpoint control that eliminates aneuploid products (MCKEE et al. 1998 ), but is instead one symptom of the dysgenesis produced by the testicular expression of genes (Ste and others) that should have been shut off by the end of meiosis. Working out the cellular pathway that, disrupted, leads to meiotic drive remains a problem yet to be resolved.

Construction of the Ste-deleted chromosomes also allowed us to directly measure the activity of the transposed heterochromatic block of Ste copies present in the B^SY chromosome. Testes-specific transcription of those copies is induced by deleting cry, but the heterochromatic copies appear to be substantially less active in the testes than are the euchromatic copies. While not directly relevant to the question that led to these experiments, the tightly regulated level of the 1400-nt ubiquitous transcript (independent of cry, independent of copy number, and independent of heterochromatic or euchromatic origin) needs to be considered as we attempt to understand the recent findings (ARAVIN et al. 2001 ) that implicate an iRNA in the regulation of Ste. How, in the very same tissue, is production of one Ste transcript inhibited by cry⁺, while the other is not only not inhibited, but is produced in a copy-number-compensated fashion?

At the same time as these experiments were undertaken, PALUMBO and BOZZETTI (personal communication) independently started a species survey of Ste and cry sequences. That project examines the second of the key assumptions in the evolutionary hypothesis, and their results are no more coherent with the selfish evolution of Ste than are those reported here. For example, although there is no crystal production in spermatocytes of D. simulans XO males, the D. simulans Y chromosome does carry Ste-like repeats.

Numerous situations in D. melanogaster cause distorted sperm recovery, including rDNA deletions, segregation of some translocations, segregation of univalents, compound autosomes, deletion of cry, and mutation of other genes that interact in the cry-Ste system. Hence, meiotic drive, in the sense of distorted sperm recovery rather than in the sense of an evolutionary force, seems to be a general downstream response to meiotic problems. Whether any situations that activate this process are of evolutionary importance remains to be seen.

	ACKNOWLEDGMENTS

We gratefully acknowledge a careful critique of the manuscript by Ellen Swanson and grants from the Università di Siena (PAR 1999) and from the Ministero dell'Università e della Ricerca Scientifica e Tecnologica (COFIN 1999).

Manuscript received September 10, 2001; Accepted for publication May 17, 2002.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ARAVIN, A. A., N. M. NAUMOVA, A. V. TULIN, V. V. VAGIN, and Y. M. ROSOVSKY et al., 2001  Double-stranded RNA-mediated silencing of genomic tandem repeats and transposable elements in the D. melanogaster germline. Curr. Biol. 11:1017-1027.[Medline]

BALAKIREVA, M. D., Y. Y. SHEVELIOV, D. I. NURMINSKY, K. LIVAK, and V. A. GVOZDEV, 1992  Structural organization and diversification of Y-linked sequences comprising Su(Ste) genes in Drosophila melanogaster.. Nucleic Acids Res. 20:3731-3736.[Abstract/Free Full Text]

BIRCHLER, J. A., M. P. BHADRA, and U. BADRA, 2000  Making noise about silence: repression of repeated genes in animals. Curr. Opin. Genet. Dev. 10:211-216.[Medline]

BOZZETTI, M. P., S. MASSARI, P. FINELLI, F. MEGGIO, and L. A. PINNA et al., 1995  The Ste locus, a component of the parasitic cry-Ste system of Drosophila melanogaster, encodes a protein that forms crystals in primary spermatocytes and mimics properties of the beta subunit of casein kinase 2. Proc. Natl. Acad. Sci. USA 92:6067-6071.[Abstract/Free Full Text]

CARVALHO, A. B. and C. C. VAZ, 1999  Are Drosophila SR drive chromosomes always balanced? Heredity 83:221-228.

CRAWLEY, P. G., 1996 The development of a Rex-induced spiral exchange technique for mapping the ribosomal DNA of D. melanogaster and further investigations of the genetic properties of Rex. Ph.D. Dissertation, Michigan State University, East Lansing, MI.

GOLIC, K. G., M. M. GOLIC, and S. PIMPINELLI, 1998  Imprinted control of gene activity in Drosophila. Curr. Biol. 8:1273-1276.[Medline]

HARDY, R. W., D. L. LINDSLEY, K. J. LIVAK, B. LEWIS, and L. SIVERSTEN et al., 1984  Cytogenetic analysis of a segment of the Y chromosome of Drosophila melanogaster. Genetics 107:591-610.[Abstract/Free Full Text]

HURST, L. D., 1992  Is Stellate a relict meiotic driver? Genetics 130:229-230.[Medline]

HURST, L. D., 1996  Further evidence consistent with Stellate's involvement in meiotic drive. Genetics 142:641-643.[Medline]

HURST, L. D. and A. POMIANKOWSKI, 1991  Causes of sex-ratio bias may account for unisexual sterility in hybrids: a new explanation of Haldane's rule and related phenomena. Genetics 128:841-858.[Abstract]

KOGAN, G. L., V. N. EPSTEIN, A. A. ARAVIN, and V. A. GVOZDEV, 2000  Molecular evolution of two paralogous tandemly repeated heterochromatic gene clusters linked to the X and Y chromosomes of Drosophila melanogaster.. Mol. Biol. Evol. 17:697-702.[Abstract/Free Full Text]

LAURIE, C. C., 1997  The weaker sex is heterogametic: 75 years of Haldane's rule. Genetics 147:937-951.[Medline]

LIFSCHYTZ, E. and D. L. LINDSLEY, 1972  The role of X-chromosome inactivation during spermatogenesis. Proc. Natl. Acad. Sci. USA 69:182-186.[Abstract/Free Full Text]

LINDSLEY, D. L., and G. G. ZIMM, 1992 The Genome of Drosophila melanogaster. Academic Press, San Diego.

LIVAK, K. J., 1984  Organization and mapping of a sequence on the Drosophila melanogaster X and Y chromosomes that is transcribed during spermatogenesis. Genetics 107:611-634.[Abstract/Free Full Text]

LIVAK, K. J., 1990  Detailed structure of the Drosophila melanogaster Stellate genes and their transcripts. Genetics 124:303-316.[Abstract]

LLOYD, V. K., D. A. SINCLAIR, and T. A. GRIGLIATTI, 1999  Genomic imprinting and position-effect variegation in Drosophila melanogaster.. Genetics 151:1503-1516.[Abstract/Free Full Text]

LYTTLE, T. W., 1991  Segregation distorters. Annu. Rev. Genet. 25:511-557.[Medline]

LYTTLE, T. W., 1993  Cheaters sometimes prosper: distortion of Mendelian segregation by meiotic drive. Trends Genet. 9:205-210.[Medline]

MCKEE, B. D., 1984  Sex chromosome meiotic drive in Drosophila melanogaster males. Genetics 106:403-422.[Abstract/Free Full Text]

MCKEE, B. D., 1987  X-4 translocations and meiotic drive in Drosophila melanogaster males: role of sex chromosome pairing. Genetics 116:409-413.[Abstract/Free Full Text]

MCKEE, B. and D. LINDSLEY, 1987  Inseparability of X-heterochromatic functions responsible for X:Y pairing, meiotic drive and male fertility in Drosophila melanogaster.. Genetics 116:399-407.[Abstract/Free Full Text]

MCKEE, B. and M. S. SATTER, 1996  Structure of the Y chromosomal Su(Ste) locus in Drosophila melanogaster and evidence for localized recombination among repeats. Genetics 142:149-161.[Abstract]

MCKEE, B. D., K. WILHELM, C. MERRILL, and X. REN, 1998  Male sterility and meiotic drive associated with sex chromosome rearrangements in Drosophila: role of X-Y pairing. Genetics 149:143-155.[Abstract/Free Full Text]

MCKEE, B. D., C. S. HONG, and S. DAS, 2000  On the roles of heterochromatin and euchromatin in meiosis in Drosophila: mapping chromosomal pairing sites and testing candidate mutations for effects on X-Y nondisjunction and meiotic drive in male meiosis. Genetica 109:77-93.[Medline]

PALUMBO, G., S. BONACCORSI, L. G. ROBBINS, and S. PIMPINELLI, 1994  Genetic analysis of Stellate elements of Drosophila melanogaster.. Genetics 138:1181-1197.[Abstract]

PIMPINELLI, S., S. BONACCORSI, M. GATTI, and L. SANDLER, 1986  The peculiar genetic organization of Drosophila heterochromatin. Trends Genet. 2:17-20.

ROBBINS, L. G., 2000  Do-it-yourself statistics: a computer-assisted likelihood approach to analysis of data from genetic crosses. Genetics 154:13-26.[Abstract/Free Full Text]

ROBBINS, L. G. and E. E. SWANSON, 1988  Rex-induced recombination implies bi-polar organization of the ribosomal RNA genes of Drosophila melanogaster.. Genetics 120:1053-1059.[Abstract/Free Full Text]

ROBBINS, L. G., G. PALUMBO, S. BONACCORSI, and S. PIMPINELLI, 1996  Measuring meiotic drive. Genetics 142:645-647.[Medline]

SANDLER, L. and E. NOVITSKI, 1957  Meiotic drive as an evolutionary force. Am. Nat. 41:105-110.

SCHMIDT, A., G. PALUMBO, M. P. BOZZETTI, P. TRITTO, and S. PIMPINELLI et al., 1999  Genetic and molecular characterization of sting, a gene involved in crystal formation and meiotic drive in the male germ line of Drosophila melanogaster.. Genetics 151:749-760.[Abstract/Free Full Text]

SHEVELYOV, Y. Y., 1992  Copies of a Stellate gene variant are located in the X heterochromatin of Drosophila melanogaster and are probably expressed. Genetics 132:1033-1037.[Abstract]

SPOFFORD, J. B., 1976 Position effect variegation in Drosophila, pp. 955–1018 in The Genetics and Biology of Drosophila, Vol. 1c, edited by M. ASHBURNER and E. NOVITSKI. Academic Press, London.

STAPLETON, W., S. DAS, and B. D. MCKEE, 2001  A role of the Drosophila homeless gene in repression of Stellate in male meiosis. Chromosoma 110:228-240.[Medline]

TRITTO, P., 2000 Charatterizzazione genetica e molecolare di mutazioni autosomiche che interferiscono con l'espressione delle sequenze Stellate in Drosophila melanogaster. Ph.D. Dissertation, University of Bari, Italy (available from http://www.biologia.uniba.it/dottorato/Tesi.html).

TULIN, A. V., G. L. KOGAN, D. FILIPP, M. D. BALAKIREVA, and V. A. GVOZDEV, 1997  Heterochromatic Stellate gene cluster in Drosophila melanogaster: structure and molecular evolution. Genetics 146:253-262.[Abstract]

This article has been cited by other articles:

				M. S. Klenov, S. A. Lavrov, A. D. Stolyarenko, S. S. Ryazansky, A. A. Aravin, T. Tuschl, and V. A. Gvozdev Repeat-associated siRNAs cause chromatin silencing of retrotransposons in the Drosophila melanogaster germline Nucleic Acids Res., August 15, 2007; (2007) gkm576v1. [Abstract] [Full Text] [PDF]

				L. N. Reynard, J. M.A. Turner, J. Cocquet, S. K. Mahadevaiah, A. Toure, C. Hoog, and P. S. Burgoyne Expression Analysis of the Mouse Multi-Copy X-Linked Gene Xlr-Related, Meiosis-Regulated (Xmr), Reveals That Xmr Encodes a Spermatid-Expressed Cytoplasmic Protein, SLX/XMR Biol Reprod, August 1, 2007; 77(2): 329 - 335. [Abstract] [Full Text] [PDF]

				M. Boschi, M. Belloni, and L. G. Robbins Genetic Evidence That Nonhomologous Disjunction and Meiotic Drive Are Properties of Wild-Type Drosophila melanogaster Male Meiosis Genetics, January 1, 2006; 172(1): 305 - 316. [Abstract] [Full Text] [PDF]