A Novel System of Fertility Rescue in Drosophila Hybrids Reveals a Link Between Hybrid Lethality and Female Sterility

Corresponding author: Daniel A. Barbash, Storer Hall, University of California, Davis, CA 95616., dabarbash{at}ucdavis.edu (E-mail)

Communicating editor: T. C. KAUFMAN

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Hybrid daughters of crosses between Drosophila melanogaster females and males from the D. simulans species clade are fully viable at low temperature but have agametic ovaries and are thus sterile. We report here that mutations in the D. melanogaster gene Hybrid male rescue (Hmr), along with unidentified polymorphic factors, rescue this agametic phenotype in both D. melanogaster/D. simulans and D. melanogaster/D. mauritiana F₁ female hybrids. These hybrids produced small numbers of progeny in backcrosses, their low fecundity being caused by incomplete rescue of oogenesis as well as by zygotic lethality. F₁ hybrid males from these crosses remained fully sterile. Hmr⁺ is the first Drosophila gene shown to cause hybrid female sterility. These results also suggest that, while there is some common genetic basis to hybrid lethality and female sterility in D. melanogaster, hybrid females are more sensitive to fertility defects than to lethality.

DROSOPHILA melanogaster can hybridize with the three related species of the simulans clade: D. simulans, D. mauritiana, and D. sechellia. Crosses of D. melanogaster females to simulans clade males produce only daughters, while the reciprocal cross produces viable sons and occasional daughters; all surviving F₁ hybrids are sterile (STURTEVANT 1920 ; LACHAISE et al. 1986 ). Much has been learned about the genetic basis of hybrid lethality by the analysis of alleles or mutations that suppress this lethality (reviewed in HUTTER 1997 ; SAWAMURA 2000 ). Otherwise inviable hybrid sons of D. melanogaster mothers are rescued by the D. simulans mutation Lethal hybrid rescue (Lhr; WATANABE 1979 ) and the D. melanogaster mutation Hybrid male rescue (Hmr; HUTTER and ASHBURNER 1987 ). Male rescue is also associated with the D. melanogaster inversion chromosome In(1)AB (HUTTER et al. 1990 ). Analysis of these and other rescue mutations has led to several models of hybrid lethality (HUTTER et al. 1990 ; SAWAMURA et al. 1993A ; BARBASH et al. 2000 ). In contrast, little is known about the genetic basis of sterility in D. melanogaster hybrids.

Two recent results led us to investigate both the causes and the possible rescue of sterility in hybrid daughters of D. melanogaster mothers. First DAVIS et al. 1996 discovered a D. simulans strain that produces fertile F₁ hybrids when crossed to D. melanogaster males. Although this strain did not rescue in the reciprocal cross and the genetic basis of rescue remains unknown, the rescue observed nevertheless demonstrates that female sterility of D. melanogaster/D. simulans hybrids is not insurmountable. Second a significant viability effect in hybrid females was recently described. While hybrid daughters of D. melanogaster are often simply described as viable, in fact they suffer from lethality at high temperature ( 25°). This lethality is due to the activity of Hmr⁺, because it is suppressed by Hmr^- mutations and deletions (BARBASH et al. 2000 ). The deleterious effect of Hmr⁺ is also manifest as various morphological defects that occur even when hybrid females are fully viable. We speculated that sterility might be an additional deleterious effect of Hmr⁺; if so, then Hmr^- mutations or deletions should suppress hybrid sterility. We now demonstrate that fertility rescue can indeed be achieved using Hmr mutations and use this novel rescue system to generate backcross flies of D. melanogaster hybrids with both D. simulans and D. mauritiana.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Fly stocks and nomenclature:
Stocks were obtained from the Drosophila Species Center (Tucson, AZ) or were described previously (DAVIS et al. 1996 ; BARBASH et al. 2000 ); the D. sechellia "Sy" stocks were obtained from Corbin Jones. The subscripts "sim," "mau," and "sec" are used to designate chromosomes from D. simulans, D. mauritiana, and D. sechellia, respectively; "sib" is used to refer to these three species collectively. Offspring of F₁ hybrid females and D. melanogaster males are referred to as BC1 (backcross 1) progeny.

Analysis of ovaries:
Ovaries were dissected from 5- to 7-day-old (25°) or 7- to 10-day-old (18°) females in PBS, squashed under a coverslip, and counted for number of mature eggs. Eggs were counted as mature if they were 75% of wild-type length. Some genotypes, particularly those with low egg counts, contained eggs with defective chorionic appendages; these eggs were counted, as they represent some level of rescue above the wild-type hybrid phenotype of completely agametic ovaries. All crosses were done at 25° unless otherwise noted.

Backcrosses:
Between 30 and 100 0- to 1-day-old virgin F₁ D. melanogaster/D. simulans females were mated in vials to 1.5–2 times as many 1- to 6-day-old virgin D. melanogaster males. Vials were changed every 1–2 days until the crosses produced no more eggs. Backcrosses to D. melanogaster/D. mauritiana females were done similarly except that 15–50 F₁ females were used per cross.

For quantitation of BC1 progeny and analysis of BC1 embryos, crosses were established for 24 hr in vials and then transferred to egg collection cups with 60-mm grape juice/agar plates supplemented with yeast paste. Eggs were collected for between 6.5 and 7 hr (daytime collections) or for 18 hr (overnight collections) over the course of 3–4 days; egg production dropped off significantly after this point. For quantification of backcross viability, groups of between 30 and 60 eggs from the daytime collections were arranged on grape juice/agar blocks and monitored daily to recover hatched first instar larvae. At the end of 3 days, unhatched eggs were briefly rinsed in 50% bleach and examined under a dissecting microscope. Eggs that were white and showed no morphological signs of development were scored as undeveloped. Brown eggs were scored as dead embryos. The number of dead larvae was inferred as the difference between the number of hatched larvae and the number of dead pupae, dead pharate adults, and eclosed adults.

For analysis of BC1 embryos, eggs from overnight collections were processed and fixed as described in PATEL 1996 and stored at -20° in methanol. Embryos were rehydrated for 10 min in 1:1 methanol:PTX (1x PBS + 0.1% Triton X-100) and for 10 min in PTX and then stained for 15 min in PTX with 1 µg/ml 4',6-diamidino-2-phenylindole (DAPI; Sigma, St. Louis). Following removal of the DAPI solution, embryos were washed two times for 15 min each in PTX, followed by clearing in 50% glycerol/1x PBS and then 70% glycerol/1x PBS.

Embryos were mounted in 70% glycerol/1x PBS and examined by differential interference contrast (DIC) and fluorescent microscopy. Embryos were scored as undeveloped if they contained no DAPI signal other than possibly from the pronucleus; eggs that were fertilized but failed to develop would be included in this class. Developed embryos were classified as normal unless they had gross abnormalities compared to intraspecific controls. Defective embryos were classified as early or late depending on whether the density of nuclei in regions of maximal density was less than or comparable to that in a normal blastoderm embryo.

DNA preparations and Southern analysis:
DNA preparations were done with a scaled-down protocol of E. J. Rehm (Berkeley Drosophila Genome Project; http://www.fruitfly.org/about/methods/inverse.pcr.html). Individual flies were ground with a micropestle in a 1.5-ml microfuge tube with 40 µl of solution A (100 mM Tris-HCl, pH 7.5, 20 mM EDTA, pH 8.0, 100 mM NaCl, 0.5% SDS), incubated at 65° for 30 min and then mixed with 160 µl of solution B (1.43 M KC₂H₃O₂, 4.29 M LiCl). Preparations were incubated for 10 min on ice and then spun for 15 min at maximum speed in a microcentrifuge. DNA was precipitated from the supernatant with 120 µl of isopropanol, washed with 70% ethanol, dried briefly (5–10 min) at room temperature and resuspended in 10 µl ddH₂O.

Genomic DNA was digested with BamHI and HindIII, whose sites are absent from the mariner transposon, fractionated on 0.8% agarose/TBE gels, and transferred to either Genescreen (Perkin-Elmer, Norwalk, CT) or Hybond-XL (Amersham, Piscataway, NJ) membranes by standard methods. The mariner probe (660 bp) was PCR amplified from D. simulans, using the primers 5'-CGACGACAAAGAGCACGGAAA-3' and 5'-AGGTCTGGTGGGTAAGCCGC-3' (60° annealing temperature) and radiolabeled with [³²P]dCTP by random priming. Membranes were hybridized overnight at 60° in 5x SSPE, 5x Denhardt's solution, 0.5% SDS, 100 µg/ml denatured salmon sperm DNA followed by washing at 60° for 15 min with 2x SSPE, 0.1% SDS, and for 15 min with 0.2x SSPE, 0.1% SDS.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Initial experiments revealed that some F₁ hybrid daughters from a D. melanogaster stock containing the hybrid rescuing X-linked inversion chromosome In(1)AB,w had mature ovaries with many eggs. We therefore sampled the sibling species by examining the In(1)AB,w/X_sib hybrid progeny from crosses of In(1)AB,w/FM6 females to males from various sibling species stocks (Table 1). Egg counts were used as a proxy for fertility in experiments to address the genetic basis of rescue; an analysis of the actual fertility of these F₁ hybrids is presented below (Table 4 Table 5 Table 6).

Rescue with D. simulans stocks:
High-level rescue was achieved with a D. simulans w stock, as the average number of eggs in In(1)AB,w/X_sim,w hybrids was comparable to those in intraspecific controls with the In(1)AB,w and D. simulans w stocks. The viability-rescued In(1)AB,w hybrid males from this cross remained sterile (n = 561); some of these males were dissected and all had rudimentary testes (n = 216).

Little or no rescue of oogenesis was observed with other D. simulans marker or wild-type stocks (except Lhr, see below), suggesting that D. simulans stocks are polymorphic for factors that allow rescue with D. melanogaster hybrids. Two results warrant further comment. First, while the v m stock produced only low-level rescue, two derivative stocks [v (F6i-w) and v m (F6i-w)] produced by repeated backcrossing to the high-level rescuing w stock had intermediate or high-level rescue. These findings suggest that the polymorphic factor(s) responsible for rescue in the w stock is unlikely to be closely linked to either the w or the v–m regions.

Second, no rescue was obtained with an inbred derivative of the D. simulans C167.4 strain that was previously shown to provide fertility rescue in the reciprocal cross of D. simulans females to D. melanogaster males (DAVIS et al. 1996 ). We confirmed that this stock does rescue fertility in the reciprocal cross: 26/27 hybrid daughters of C167.4 inbred females crossed to In(1)AB,w males contained a mean number of 20.4 ± 11.7 eggs. These data suggest that the genetic bases of fertility rescue in the two directions of crossing are at least partially distinct (see DISCUSSION).

Rescue with D. simulans Lhr:
Significant rescue was also obtained with a D. simulans stock that carries the mutation Lhr (Table 1). This stock suppresses hybrid female lethality in a manner analogous to In(1)AB, presumably due to the Lhr mutation (BARBASH et al. 2000 ). We therefore examined several D. melanogaster stocks that do not carry rescue alleles for fertility in hybrids with Lhr. Hybrids made with the marker stocks w¹¹¹⁸ and cn; bw and the wild-type stocks Oregon-R and Harare-2 showed little or no rescue, with a maximum of 1.1 ± 1.9 eggs observed (n = 32 to 65). Low-level rescue was observed in hybrids between y; ry⁵⁰⁶ and Lhr: 80.0% of F₁ females had eggs with a mean number of 3.2 ± 3.1 (n = 40). Low-level rescue with Lhr was also observed in other crosses (see Mapping of rescue below).

Rescue with D. mauritiana and D. sechellia stocks:
Because rescue depended, in part, on the D. melanogaster stock used (see below), we could also assay for rescue in hybrids with the other simulans clade species (Table 1). In(1)AB,w/X_mau females produced with the D. mauritiana iso 207 stock had eggs counts comparable to the intraspecific D. mauritiana control. Their hybrid male siblings, however, remained sterile (n = 639) and all males examined had rudimentary testes (n = 259). Hybrid females produced with other D. mauritiana stocks showed lower or essentially no rescue.

In a survey of 10 D. sechellia stocks, In(1)AB,w/X_sec hybrids showed little or no rescue, with a mean number of zero to two eggs and a maximum of eight eggs observed among all crosses.

Survey of Hmr mutant stocks:
This study was undertaken to explore the hypothesis that hybrid female sterility is a consequence of the Hmr- and Lhr-dependent hybrid lethality system. We therefore tested additional D. melanogaster stocks that are mutant for or lacking Hmr function to determine whether they also rescued at the same high level observed with In(1)AB,w (Table 2). All crosses were performed with the D. simulans w strain that produced the high-level rescue described above. We first examined three additional In(1)AB stocks that rescue hybrid male inviability but differ by the presence or absence of various X-linked markers. These stocks provided intermediate [In(1)AB], low [In(1)AB, v f], or essentially no [In(1)w^m4 + In(1)AB, y² w^m4] rescue, demonstrating that the hybrid rescue activity of In(1)AB cannot be the sole determinant of fertility rescue. We also note that In(1)w^m4 + In(1)AB, y² w^m4 failed to rescue regardless of whether the F₁ hybrids came from mothers that were heterozygous (Table 2, cross 3) or homozygous (cross 4) for the inversion.

We next examined a number of deficiencies known to be deleted for Hmr (BARBASH et al. 2000 ; D. A. BARBASH, unpublished work). Df(1)v-L11 and Df(1)ras59 rescued at an intermediate level, while other deficiencies showed little or no rescue. Finally, we tested three stocks carrying the mutant Hmr¹, an allele that is probably a hypomorph (HUTTER et al. 1990 ; BARBASH et al. 2000 ). Essentially no rescue was obtained with any of the three Hmr¹ stocks.

Mapping of rescue:
These varying results raise the question of whether Hmr mutations are actually required for fertility rescue. For those stocks where rescue was obtained we initially determined that the X chromosome, where Hmr maps, is required. We first compared In(1)AB, w/X_sim females to their FM6/X_sim sisters (Table 3, cross 1); FM6 is an Hmr⁺ balancer chromosome. Because Hmr⁺ hybrids have poor viability at high temperatures, this cross was performed at 18°, where Hmr⁺ female hybrids are fully viable (BARBASH et al. 2000 ). In(1)AB,w/X_sim females had lower egg counts than those at 25° (compare to Table 1), but >95% were fertile. It is interesting to note that male viability rescue by In(1)AB also appears to be reduced at low temperature (HUTTER et al. 1990 ; BARBASH et al. 2000 ). In contrast, no eggs were observed in the FM6/X_sim siblings. Similar results were observed with Df(1)v-L11 (Table 3, cross 2). Almost 90% of Df(1)v-L11/X_sim hybrids had eggs, while none were found in their FM6/X_sim sisters.

An alternative explanation for these results might be that the FM6 reference chromosome used is dominantly sterile in female hybrids. This explanation was excluded, however, by showing that FM6/X_sim; 2_mel/2_sim, Lhr females have a low level of fertility rescue; this effect was observed with two different Lhr stocks (Table 3, crosses 3 and 4). This rescue, in contrast to the complete sterility of FM6/X_sim, w hybrids observed, suggests that fertility rescue requires a hybrid rescue mutation (in this case Lhr) even under conditions of low temperature where females are fully viable.

Note, however, that in both crosses the In(1)AB,w/X_sim; 2_mel/2_sim, Lhr hybrids had much higher egg counts than those of their FM6/X_sim; 2_mel/2_sim, Lhr siblings. This finding suggests either that the combination of being doubly mutant for In(1)AB and Lhr provides higher rescue than Lhr alone does or that the In(1)AB chromosome contains additional factor(s) required for high-level fertility rescue.

Two schemes were used to map more precisely the effect of Hmr on fertility rescue. The Hmr locus cannot be mapped by recombination on In(1)AB rescuing chromosomes because it is near an inversion breakpoint. We therefore compared the fertility of In(1)AB,w/X_sim hybrid sibling females with and without the Hmr⁺ duplication Dp(1;2)v^+75d (Table 3, cross 5); this duplication suppresses the hybrid male rescuing activity of Hmr¹ (HUTTER et al. 1990 ) and In(1)AB (data not shown). Crosses were again done at 18° to obtain fully viable females of both classes. Fertility rescue was observed only in nonduplication hybrids, suggesting that rescue depends on the reduced activity of Hmr.

In the second scheme, we directly mapped fertility rescue relative to the presence or absence of the Hmr^– deficiency Df(1)v-L11. We crossed Df(1)v-L11 into two different genetic backgrounds and then compared the fertility of Df(1)v-L11/X_sim and +/X_sim hybrid siblings (Table 3, crosses 6 and 7). Df(1)v-L11/X_sim females had modest fertility rescue, with a mean number of four to five eggs, while no rescue was observed in 652 +/X_sim siblings. These data demonstrate that rescue is very tightly linked to Df(1)v-L11 and, we suggest, in fact requires this Hmr^- deficiency.

Variability and transfer of rescue:
If the above mapping results are correct, they leave open the question of why some In(1)AB and Hmr mutant stocks do not rescue fertility. One possibility is that additional rescuing factor(s) is required but is polymorphic among different strains. Evidence in favor of this interpretation was provided by crosses in which Hmr¹ or In(1)AB nonfertility rescuing alleles were crossed into fertility rescuing stocks. Female progeny of such crosses were then used to create hybrid sibling daughters that carried either of the two rescuing alleles. One such experiment used the In(1)w^m4 + In(1)AB, y² w^m4 chromosome from a stock shown in Table 2 to not rescue fertility. After crossing into the high rescuing In(1)AB,w stock, In(1)w^m4 + In(1)AB, y² w^m4/X_sim hybrids had egg counts approximately half that of their In(1)AB,w/X_sim siblings (Table 3, cross 8). This result demonstrates that In(1)w^m4 + In(1)AB, y² w^m4 can rescue fertility, if additional rescue factor(s) is "transferred" from the rescuing In(1)AB,w stock. Because the In(1)w^m4 inversion covers most of the X chromosome and only double recombinants with In(1)AB,w will be viable, these polymorphic rescuing factors are likely to be autosomal.

Low-level rescue was also obtained in Hmr¹/X_sim females after Hmr¹ was crossed into rescuing stocks (Table 3, crosses 9 and 10). No rescue was obtained with Df(1)ras203, however, after a similar crossing procedure, and in fact fertility of their In(1)AB,w/X_sim siblings was greatly reduced (Table 3, cross 11). These results suggest that D. melanogaster stocks may be polymorphic for factors that can either promote or inhibit hybrid fertility.

Production and analysis of BC1 hybrids:
To determine if backcross hybrids could be produced, F₁ D. melanogaster/D. simulans and D. melanogaster/D. mauritiana females derived from either the In(1)AB,w or the Df(1)v-L11 stocks were backcrossed to D. melanogaster males.

BC1 hybrids were obtained with various combinations of stocks, but only in low numbers. In eight different crosses the average number of progeny per F₁ hybrid female parent was well below one (Table 4). Eighty-two of these BC1 females were tested for fertility by mating to wild-type D. melanogaster males; 11 of them were fertile with a mean number of 81 progeny.

To determine whether the viable adults obtained were truly BC1 progeny we examined individual animals for the presence of the transposable element mariner, which is absent in D. melanogaster but present in D. simulans and D. mauritiana (Fig 1; MARUYAMA and HARTL 1991 ; DAVIS et al. 1996 ). The majority of BC1 flies contained mariner-hybridizing bands corresponding to those observed in the parental stocks, demonstrating that either D. simulans or D. mauritiana genetic material was segregating in a D. melanogaster background. This is the first demonstration that it may be possible to introduce D. mauritiana genetic material into D. melanogaster by repeated backcrossing. It is striking that many of the BC1 D. melanogaster/D. mauritiana hybrids contained most or all of the mariner markers found in the D. mauritiana parental stock (Fig 1B). Although we do not know the genomic distribution of these markers, this pattern does suggest that a substantial amount of the D. mauritiana genome may be segregating in these hybrids.

View larger version (104K): In this window In a new window Download PPT slide   Figure 1. Evidence for and analysis of BC1 hybrids. Southern blots of genomic DNA hybridized with a fragment of the transposable element mariner. (A) Blot of D. melanogaster/D. simulans hybrids. Lane 1, In(1)AB,w/FM6; lane 2, Df(1)v-L11/FM6; lane 3, Oregon-R (wild-type D. melanogaster); lane 4, D. simulans w; lane 5, In(1)AB,w/Xsim, w F1 females; lanes 6–13, single BC1 females from the cross of In(1)AB,w/Xsim, w F1 females to Oregon-R males; lanes 14–18, single BC1 females from the cross of Df(1)v-L11/Xsim, w F1 females to Oregon-R males. In a control experiment 50/50 single females from the D. simulans w stock contained the two strongly hybridizing bands shown in lane 4 (data not shown), suggesting that this stock is fixed for the detected mariner insertions. The bottom shows the same filter rehybridized with a probe for the gene rp49, demonstrating that the lanes with no mariner signal contain DNA. (B) Blot of BC1 D. melanogaster/D. mauritiana hybrids. Lane 1, In(1)AB,w/FM6; lane 2, D. mauritiana iso 207; lane 3, In(1)AB,w/Xmau F1 females; lanes 4–11, single BC1 females from the cross of In(1)AB,w/Xmau F1 females to Oregon-R males; lanes 12–15, single white-eyed BC1 males from the same cross; lanes 16–19, single wild-type (red-eyed) BC1 males from the same cross. In a control experiment 35/35 single females from the D. mauritiana iso 207 stock showed the identical pattern as found in lane 2 (data not shown), suggesting that this stock is fixed for the detected mariner insertions.

Some of the viable BC1 hybrids displayed a range of morphological defects, including rough eyes, missing bristles, and abdominal cuticle (Fig 2B) and wing venation defects (Fig 2D). Various morphological defects have also been observed previously in both F₁ and BC1 hybrids (STURTEVANT 1920 ; BARBASH et al. 2000 ; SAWAMURA et al. 2000 ).

View larger version (119K): In this window In a new window Download PPT slide   Figure 2. Morphological defects in BC1 hybrids. Thorax and notum (A) and wing (C) from D. simulans w females, showing the wild-type morphology. (B and D) Same tissues from BC1 female hybrids from the cross of In(1)AB,w/Xsim, w F1 females to Nguruman-4 (wild-type D. melanogaster) males. Note the defective abdominal cuticle and missing dorsocentral (arrowhead) and scutellar (arrow) bristles in B and the wing with ectopic vein material (arrowhead) in D.

To investigate the causes of the low fecundity of F₁ females we collected eggs from backcrosses and followed their developmental fate (Table 5). More than 96% showed no sign of development upon external examination, suggesting that most eggs either were unfertilized or arrested early in embryonic development. Among the eggs that hatched into larvae <20% reached adulthood (9/48).

To examine in more detail whether or not backcross eggs began embryogenesis we analyzed the external morphology as well as the number and distribution of nuclei in fixed eggs and embryos from these crosses (Table 6). In agreement with the above analysis, the majority of eggs (83%) appeared to be unfertilized (although our analysis cannot exclude the possibility that the hybrid eggs are fertilized but fail to begin embryogenesis). About half of these eggs had aberrant shapes, with the defects ranging between those displayed by the two eggs shown in Fig 3D.

View larger version (93K): In this window In a new window Download PPT slide   Figure 3. Aberrant morphologies in BC1 eggs and embryos. Embryos from the backcross of In(1)AB,w/Xsim,v females to Oregon-R males (see Table 5) were stained for DAPI (A–C) or examined under DIC (D). (A) An early arrest embryo with a small number of scattered nuclei. (B) An arrested embryo on the left side with an inhomogeneous distribution of nuclei as well as a large clump of nuclei (white arrow), next to an unfertilized egg. (C) An arrested embryo with a blastoderm-like density of nuclei in its left half. (D) Two undeveloped eggs of unusual shape. The scales of B and D are x0.72 and x0.9, respectively, relative to A and C.

Among developed embryos about half appeared to be normal. Among the remainder were embryos that arrested at various stages of development with many having an inhomogeneous density of nuclei (Fig 3, A–C). The number of "normal" embryos in both the experimental and the control crosses was higher than expected on the basis of the number of hatched first instar larvae from these same crosses (Table 5). For example, in the cross of In(1)AB,w/X_sim, v females and w¹¹¹⁸ males >9% of the embryos were classified as normal in morphology (Table 6), yet only 2.6% of the eggs from this cross hatched (Table 5). This discrepancy suggests that some of the embryos classified as normal had undetected defects and were in fact lethal. We emphasize though that both scoring criteria support the conclusion that the vast majority of BC1 eggs are inviable and/or unfertilized.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

We report here that fertile F₁ hybrid females were obtained from D. melanogaster mothers by using stocks containing viability rescue alleles or Hmr^- deletions. Previous attempts (DAVIS et al. 1996 ; HOLLOCHER et al. 2000 ) to achieve fertile hybrids from In(1)AB D. melanogaster females failed, presumably because the D. simulans strains used (C167.4, Tsimbazaza, Oxnard, and v) are not permissive for fertility rescue in this direction of crossing (Table 1).

Egg counts were used as a convenient measure of fertility rescue in our genetic assays, which is justified by the fact that unrescued hybrid females are completely agametic. Yet while we found that in some backgrounds rescued hybrids had egg counts equivalent to the intraspecific controls (Table 1), these hybrids produced small numbers of viable progeny in backcrosses (Table 4) and many of their eggs and embryos were aberrant (Table 5 and Table 6; Fig 3). These findings demonstrate that even in permissive genetic backgrounds Hmr⁺ is not the sole contributor to hybrid female sterility.

The low yield of backcross progeny has multiple causes. Zygotic lethality is clearly higher than that in pure species and occurs at all developmental stages (Table 5) including embryogenesis (Table 6). The predominant factor, however, appears to be that most eggs fail to develop. We observed that many apparently unfertilized eggs were structurally abnormal (Fig 3D), demonstrating that the rescue of oogenesis in the F₁ females is incomplete. In addition to these intrinsic defects, hybrid eggs might also fail to be fertilized due to mating isolation or incompatibilities between the hybrid mothers and the pure species sperm or seminal fluid. Finally we note that F₁ females also appeared to have a shorter duration of oogenesis compared to the pure species, although we did not quantitate this difference.

Comparison to rescue from D. simulans mothers:
DAVIS et al. 1996 discovered fertility rescue in F₁ hybrid daughters of D. simulans females and D. melanogaster males. The system of Davis et al. is distinct from that which we describe here, because the C167.4 D. simulans strain used previously does not rescue in the reciprocal cross to D. melanogaster females (Table 1), although Hmr may play a role in both systems (see The role of rescue alleles in the reciprocal cross below).

Rescue of oogenesis with both systems is incomplete. We have outlined above various defects observed in F₁ hybrid eggs and backcross embryos. Likewise, eggs and ovaries analyzed in hybrid daughters of C167.4 D. simulans females ranged from resembling the wild type to having severe defects in egg morphology and ovariole number (HOLLOCHER et al. 2000 ). To best compare the utility of these two systems for performing interspecific genetics will require knowing the distribution of genotypes present in fertile BC1 hybrids. The novel system reported here, however, does hold several potential advantages over the system of Davis et al. First, the fertile hybrids we have produced are progeny of D. melanogaster females and sibling species males. It has been long known that it is much easier to achieve this direction of hybridization than the reciprocal direction (STURTEVANT 1920 ). Second, hybrid female viability is greater among progeny of D. melanogaster mothers because they do not suffer from the embryonic lethality that occurs in the daughters of sibling species mothers (HADORN 1961 ; LACHAISE et al. 1986 ). For these reasons it will be much easier to obtain large numbers of F₁ females for backcrossing by using D. melanogaster mothers. Third, our system produces fertile hybrids with both D. simulans and D. mauritiana, which extends the possibilities for analyzing the genetics of species-specific differences using the many genetic resources available in D. melanogaster.

Hybrid sterility is caused by both Hmr⁺ and additional polymorphic factors:
Our results suggest that three conditions are required to rescue the agametic phenotype of D. melanogaster female hybrids: factors that are polymorphic in D. melanogaster stocks, factors that are polymorphic in the sibling species D. simulans and D. mauritiana, and mutations that rescue Hmr- or Lhr-dependent hybrid lethality. Evidence for the first two conditions is simply that some stocks of these species rescue and some do not; we have not attempted any further characterization. Rather we have focused on establishing whether Hmr⁺ does in fact cause female sterility. Doing so requires some caution if Hmr and In(1)AB stocks are themselves polymorphic for the hypothesized additional rescue factors. Correlative evidence for an Hmr requirement is that rescue was obtained only in the presence of hybrid rescue mutations. Stronger evidence was provided by the two independent mapping schemes described in Table 3. We first showed that rescue by In(1)AB is suppressed completely by an Hmr⁺ duplication (cross 5). In our second mapping scheme we demonstrated that rescue occurs only in hybrids containing the Hmr^- deletion Df(1)v-L11 after outcrossing it into two different genetic backgrounds (crosses 6 and 7).

Dose dependence of Hmr⁺ effects:
The lethal effects of Hmr⁺ are not absolute but rather are highly dependent on both gene dosage and temperature (BARBASH et al. 2000 ). We have found that lethality and female sterility share a common genetic basis in Drosophila hybrids because both phenotypes are rescued by Hmr mutations. But fertility is more sensitive to Hmr⁺ activity than viability, because female hybrids are fully viable at low temperature but remain sterile unless Hmr⁺ is removed. In other words, Hmr⁺ is a fully dominant hybrid sterile allele but ranges from a fully recessive to a fully dominant lethal allele depending on temperature.

The role of rescue alleles in the reciprocal cross:
As noted above, the fertility rescue observed in hybrid daughters of D. simulans by DAVIS et al. 1996 is at least partly distinct from that which we describe here. We suggest, however, that Hmr-dependent rescue may play a role in both systems of fertility rescue. Davis et al. concluded that fertility rescue does not depend on hybrid inviability rescue alleles. We reinterpret their findings in light of subsequent work that suggests that hybrid daughters of D. simulans mothers suffer from two distinct types of lethality (BARBASH et al. 2000 ): embryonic lethality, which is caused by D. melanogaster Zhr⁺ (SAWAMURA and YAMAMOTO 1993 ), and pupal/pharate lethality, which is caused by Hmr⁺ (BARBASH et al. 2000 ). This second form of lethality can be rescued by Hmr¹ or Hmr^- deletions [In(1)AB has not been tested for rescue in this assay].

We suggest therefore that Hmr⁺ may have the same phenotypic effect in both directions of crossing, namely that it causes dominant female sterility even under conditions of low temperature when it does not cause lethality. In support of this hypothesis we note that most reported cases of hybrid fertility rescue in crosses with D. simulans C167.4 females used In(1)AB D. melanogaster males (DAVIS et al. 1996 ; HOLLOCHER et al. 2000 ; SAWAMURA et al. 2000 ). Such males were used to rescue the embryonic lethality of hybrid daughters, because the In(1)AB chromosomes presumably also carried Zhr^- alleles (SAWAMURA et al. 1993B ). We suggest that the In(1)AB rescue allele itself may also have contributed to the fertility rescue by reducing Hmr⁺ activity. We must acknowledge, however, that one cross of D. simulans females to D. melanogaster males reported in DAVIS et al. 1996 is not consistent with this hypothesis, as fertility rescue was obtained in the absence of In(1)AB or any other known rescue allele.

Male vs. female hybrid sterility:
In male heterogametic species, male sterility is the most commonly observed form of hybrid incompatibility (WU and DAVIS 1993 ; LAURIE 1997 ), suggesting that hybrid male sterility may evolve more rapidly than either hybrid female sterility or lethality (WU and DAVIS 1993 ). Consistent with this proposal is the finding that Drosophila hybrids contain more male sterility alleles than either female sterility or (non-sex-specific) lethality alleles (HOLLOCHER and WU 1996 ; TRUE et al. 1996 ; SAWAMURA et al. 2000 ). In this context it is notable that we have found that Hmr alleles rescue both hybrid inviability and female sterility, but not male sterility. We suggest that the more polygenic nature of male sterility makes it unlikely that single gene rescue alleles will be found for this trait in D. melanogaster hybrids.

	ACKNOWLEDGMENTS

We thank C. Jones and the Drosophila Species Center (Tucson, AZ) for fly stocks, J. Roote and C. H. Langley for helpful discussions, and K. Sawamura and T. Cline for comments on earlier drafts of the manuscript. This work was supported by grants from the Medical Research Council and Leverhulme Trust to M.A. and from the National Science Foundation to D.A.B. and C. H. Langley.

Manuscript received July 9, 2002; Accepted for publication October 21, 2002.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

BARBASH, D. A., J. ROOTE, and M. ASHBURNER, 2000  The Drosophila melanogaster Hybrid male rescue gene causes inviability in male and female species hybrids. Genetics 154:1747-1771.[Abstract/Free Full Text]

DAVIS, A. W., J. ROOTE, T. MORLEY, K. SAWAMURA, and S. HERRMANN et al., 1996  Rescue of hybrid sterility in crosses between D. melanogaster and D. simulans.. Nature 380:157-159.[Medline]

HADORN, E., 1961  Zur Autonomie und Phasenspezifität der Letalität von Bastarden zwischen Drosophila melanogaster und Drosophila simulans.. Rev. Suisse Zool. 68:197-207.

HOLLOCHER, H. and C.-I WU, 1996  Genetics of reproductive isolation in the Drosophila simulans clade: X vs. autosomal effects and male vs. female effects. Genetics 143:1243-1255.[Abstract]

HOLLOCHER, H., K. AGOPIAN, J. WATERBURY, R. W. O'NEILL, and A. W. DAVIS, 2000  Characterization of defects in adult germline development and oogenesis of sterile and rescued female hybrids in crosses between Drosophila simulans and Drosophila melanogaster.. J. Exp. Zool. 288:205-218.[Medline]

HUTTER, P., 1997  Genetics of hybrid inviability in Drosophila.. Adv. Genet. 36:157-185.[Medline]

HUTTER, P. and M. ASHBURNER, 1987  Genetic rescue of inviable hybrids between Drosophila melanogaster and its sibling species. Nature 327:331-333.[Medline]

HUTTER, P., J. ROOTE, and M. ASHBURNER, 1990  A genetic basis for the inviability of hybrids between sibling species of Drosophila.. Genetics 124:909-920.[Abstract]

LACHAISE, D., J. R. DAVID, F. LEMEUNIER, L. TSACAS, and M. ASHBURNER, 1986  The reproductive relationships of Drosophila sechellia with Drosophila mauritiana, Drosophila simulans and Drosophila melanogaster from the Afrotropical region. Evolution 40:262-271.

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

MARUYAMA, K. and D. L. HARTL, 1991  Evolution of the transposable element mariner in Drosophila species. Genetics 128:319-329.[Abstract]

PATEL, N. H., 1996 In situ hybridization to whole-mount Drosophila embryos, pp. 357–369 in A Laboratory Guide to RNA: Isolation, Analysis, and Synthesis, edited by P. A. KRIEG. Wiley-Liss, New York.

SAWAMURA, K., 2000  Genetics of hybrid inviability and sterility in Drosophila: the Drosophila melanogaster-Drosophila simulans case. Plant Species Biol. 15:237-247.

SAWAMURA, K. and M.-T. YAMAMOTO, 1993  Cytogenetical localization of Zygotic hybrid rescue (Zhr), a Drosophila melanogaster gene that rescues interspecific hybrids from embryonic lethality. Mol. Gen. Genet. 239:441-449.[Medline]

SAWAMURA, K., T. K. WATANABE, and M.-T. YAMAMOTO, 1993a  Hybrid lethal systems in the Drosophila melanogaster species complex. Genetica 88:175-185.[Medline]

SAWAMURA, K., M.-T. YAMAMOTO, and T. K. WATANABE, 1993b  Hybrid lethal systems in the Drosophila melanogaster species complex. II. The Zygotic hybrid rescue (Zhr) gene of Drosophila melanogaster. Genetics 133:307-313.[Abstract]

SAWAMURA, K., A. W. DAVIS, and C.-I WU, 2000  Genetic analysis of speciation by means of introgression into Drosophila melanogaster.. Proc. Natl. Acad. Sci. USA 97:2652-2655.[Abstract/Free Full Text]

STURTEVANT, A. H., 1920  Genetic studies of Drosophila simulans. I. Introduction. Hybrids with Drosophila melanogaster. Genetics 5:488-500.[Free Full Text]

TRUE, J. R., B. S. WEIR, and C. C. LAURIE, 1996  A genome-wide survey of hybrid incompatibility factors by the introgression of marked segments of Drosophila mauritiana chromosomes into Drosophila simulans.. Genetics 142:819-837.[Abstract]

WATANABE, T. K., 1979  A gene that rescues the lethal hybrids between Drosophila melanogaster and Drosophila simulans.. Jpn. J. Genet. 54:325-331.

WU, C.-I and A. W. DAVIS, 1993  Evolution of postmating reproductive isolation: the composite nature of Haldane's rule and its genetic bases. Am. Nat. 142:187-212.

This article has been cited by other articles:

				B. J. Bolkan, R. Booker, M. L. Goldberg, and D. A. Barbash Developmental and Cell Cycle Progression Defects in Drosophila Hybrid Males Genetics, December 1, 2007; 177(4): 2233 - 2241. [Abstract] [Full Text] [PDF]

				D. A. Barbash Nup96-Dependent Hybrid Lethality Occurs in a Subset of Species From the simulans Clade of Drosophila Genetics, May 1, 2007; 176(1): 543 - 552. [Abstract] [Full Text] [PDF]

				D. C. Presgraves and W. Stephan Pervasive Adaptive Evolution among Interactors of the Drosophila Hybrid Inviability Gene, Nup96 Mol. Biol. Evol., January 1, 2007; 24(1): 306 - 314. [Abstract] [Full Text] [PDF]

				B. T. Sage and A. K. Csink Heterochromatic Self-Association, a Determinant of Nuclear Organization, Does Not Require Sequence Homology in Drosophila Genetics, November 1, 2003; 165(3): 1183 - 1193. [Abstract] [Full Text] [PDF]

				D. A. Barbash, D. F. Siino, A. M. Tarone, and J. Roote A rapidly evolving MYB-related protein causes species isolation in Drosophila PNAS, April 29, 2003; 100(9): 5302 - 5307. [Abstract] [Full Text] [PDF]