Truncated RanGAP Encoded by the Segregation Distorter Locus of Drosophila

See allHide authors and affiliations

Science  12 Mar 1999:
Vol. 283, Issue 5408, pp. 1742-1745
DOI: 10.1126/science.283.5408.1742


Segregation Distorter (SD) inDrosophila melanogaster is a naturally occurring meiotic drive system in which the SD chromosome is transmitted fromSD/SD+ males in vast excess over its homolog owing to the induced dysfunction of SD+-bearing spermatids. The Sd locus is the key distorting gene responsible for this phenotype. A genomic fragment from theSd region conferred full distorting activity when introduced into the appropriate genetic background by germline transformation. The only functional product encoded by this fragment is a truncated version of the RanGAP nuclear transport protein. These results demonstrate that this mutant RanGAP is the functional Sd product.

Examples of meiotic drive, in which a particular allele or chromosome of a heterozygous pair is preferentially transmitted to the offspring, have been described in natural populations of fungi, plants, insects, and mammals (1). This violation of the fundamental principle of Mendelian genetics can subvert the evolutionary process, which is contingent on the unbiased exposure of competing genes to selective forces. The molecular mechanisms of meiotic drive have remained elusive.

One of the best characterized meiotic drive systems isSegregation Distorter (SD) in Drosophila melanogaster (2). SD chromosomes are transmitted from SD/SD + males to more than 95% of the progeny; transmission from females is normal. Sd, a dominant gain-of-function mutation, is the primary gene onSD chromosomes required for distortion. Strong distortion also requires several linked modifier loci, includingEnhancer [E(SD)], Modifier[M(SD)], and Stabilizer [St(SD)] (2–4). The target of distortion is the Responder(Rsp) locus, which consists of an array of repeated satellite sequences whose copy number is correlated with sensitivity (4, 5). Chromosomes carrying Rsp s(sensitive) or Rsp ss (supersensitive) loci are subject to distortion, whereas SD chromosomes, which carryRsp i (insensitive), are resistant. Distortion ultimately involves sperm dysfunction, first visibly apparent as failed chromatin condensation in half of the developing spermatids (6).

The Sd locus was isolated by positional cloning and found to be associated with a tandem duplication that replaces the wild-type (Sd +) 6.5-kb Eco RI fragment with an 11.5-kb Eco RI fragment (7). When introduced into appropriate genetic backgrounds by germline transformation, the 11.5-kb fragment confers full distortion, which indicates that Sd activity is contained entirely within this fragment (8). Transformants that have lost either half of the duplication fail to cause distortion, demonstrating that Sd activity requires juxtaposition of particular sequences in both halves of the duplication.

To identify the particular gene or genes within the duplication that are responsible for distortion, we sequenced the 6.5-kbSd + and 11.5-kb Sd genomic clones as well as cDNAs derived from these chromosome segments. Two nested transcription units were identified in Sd +, both of which are represented twice in the duplication (Fig. 1, A and B). One of these genes (dHS2ST) encoded the Drosophila homolog of mammalian heparan-sulfate 2-sulfotransferase (9). The other,dRanGAP, was identified as the Drosophila homolog of mammalian RanGAP1, the guanosine triphosphatase (GTPase) activator for the Ras-related nuclear regulatory protein Ran (10,11).

Figure 1

Organization of dHS2ST anddRanGAP transcription units in SD +and SD genomic DNA. The upper line in each panel represents the genomic clones. The wild-type genomic sequence that is tandemly duplicated in SD is represented by the hatched bar. The alignment of cDNAs with the genomic sequence is indicated underneath the genomic fragment. Exons containing translated sequences are represented by boxes. Solid boxes represent dHS2ST exons and open boxes represent dRanGAP exons. A space between boxes indicates the presence of an intron. Introns in noncoding sequences are indicated by a V-shaped line. (A) Two cDNAs transcribed from opposite strands were identified in wild-type flies. ThedHS2ST cDNA is entirely contained within a large (2.7-kb) intron at the 5′ end of dRanGAP. (B) The junction of the tandem duplication is marked. Each half of the duplication contains both dHS2ST and dRanGAP transcription units. The distally encoded dRanGAP polypeptide is truncated at the junction site. The 5′ end of the distal dRanGAPtranscription unit is drawn so that it corresponds with that of the proximal dRanGAP transcript. However, cDNAs for this transcript are incomplete at the 5′ end, so the exact starting point is unknown. This uncertainty is indicated by the broken line representing the presumptive intron and the question mark at the 5′ end of the transcript. An aberrantly large dHS2ST transcript that initiates ectopically in the proximal part of the duplication and results in the production of a long 5′ untranslated leader is shown. (C) The 12A genomic transformation construct includes the entire distal half of the Sd duplication plus approximately 500 bp from the right half of the duplication. Two in-frame stop codons (asterisk) were introduced into thedHS2ST coding sequence to eliminate expression of this protein from the 12A transformation construct.

The organization of the two transcription units in the proximal half of the Sd duplication is essentially the same as inSd + (Fig. 1, A and B). Although the proximaldHS2ST transcript in some SD lines contained a premature stop codon, resulting in truncation of 46 amino acids, other strongly distorting lines lacked this single base polymorphism. The proximal dRanGAP transcript is identical with the wild type except for a 9–base pair (bp) deletion near the 3′ end. Because this deletion occurs outside the 11.5-kb fragment, this polymorphism is also unlikely to be responsible for the Sd phenotype.

The dHS2ST transcript in the distal half of the duplication encoded a wild-type protein. A cDNA representing a previously identified SD-specific 4.2-kb transcript (8, 9) initiated in the proximal half of the duplication and read through the distal copy of the dHS2ST gene, resulting in an aberrant mRNA that contained a wild-type dHS2ST coding region preceded by a long untranslated leader (Fig. 1B).

The remaining distal transcript encoded a mutant version of dRanGAP whose COOH-terminal portion differed from that of the wild type beginning at the duplication junction (Fig. 1B). Analysis ofSD cDNAs (10) indicated that the distaldRanGAP transcript extended about 300 to 400 bp beyond the junction into the proximal half of the duplication. The juxtaposition of sequences at the junction site introduced an in-frame stop codon immediately adjacent to this site. The resulting dRanGAPmRNA encoded a truncated polypeptide missing 234 amino acids at the COOH-terminus (12). Because this truncated dRanGAP was the only substantially altered protein encoded by the 11.5-kb fragment, and because its generation required the fusion of sequences from both halves of the duplication, it was the best candidate for theSd gene product.

To be capable of causing distortion, the postulated truncated dRanGAP must be stable and expressed in testes. The dRanGAPcDNA sequences predict proteins of 66 kD encoded both bySd + and by the proximal half of theSd duplication, as well as an Sd-specific 40-kD protein encoded by the distal half of the Sd duplication. Affinity-purified antiserum (13) detected the expected 66-kD protein in males of all genotypes on protein immunoblots of protein extracts from testes (Fig. 2, lanes 1 through 7). The 40-kD protein was detected in all native SDlines examined but not in the wild type (Fig. 2, lanes 1 and 2) (14, 15). Furthermore, three independent germline transformants for the 11.5-kb duplication that showed strong distorting activity expressed the truncated protein (Fig. 2, lanes 3 through 5), whereas another transformant that spontaneously lost part of the duplication and failed to distort did not express this protein (Fig. 2, lane 6). Thus, expression of an Sd-specific truncated dRanGAP correlated with distorting activity.

Figure 2

Protein immunoblot of testes proteins probed with Drosophila antibodies to RanGAP. Three pair of testes were dissected from newly eclosed flies, boiled in SDS sample buffer, and run on a 15% polyacrylamide gel. The proteins were transferred to nitrocellulose and probed with a polyclonal antiserum (1:500) raised against the amino-terminal region of wild-type dRanGAP. Horseradish peroxidase (HRP)–conjugated goat anti-rabbit secondary antibodies (Bio-Rad) (1:3000) were detected by chemiluminescence (Amersham). Lane 1,SD-Mad; lane 2, Canton-S (wild type); lanes 3 through 5, three independently generated transgenic lines containing the 11.5-kbSd duplication (8). These three transformant lines show high levels of distortion. Lane 6, a transgenic line transformed originally with the 11.5-kb Sd duplication that has undergone spontaneous loss of a significant portion of the duplication. No distortion is observed in this line. Lane 7, transgenic flies containing the P[(w+Sd)12A] transformation construct that specifically encodes the truncated form of dRanGAP. This transgenic line shows high levels of distortion.

To test directly whether expression of the truncated dRanGAP was responsible for distortion, we generated a transformation construct capable of expressing only this protein. Beginning with the intact 11.5-kb Eco RI fragment, all but 562 bp of the proximal half of the duplication was removed to eliminate both proximal-specific transcripts and the 4.2-kb version of the distal dHS2ST transcript (Fig. 1C). The coding potential of the remaining distal-specificdHS2ST transcript was abolished by introducing two in-frame stop codons. We used this modified genomic construct to generate a germline transformant, P[(w+Sd)12A]. This construct directs the production of the Sd-specific 40-kD dRanGAP protein although its level of expression appeared somewhat reduced (Fig. 2, lane 7) (15).

We introduced P[(w+Sd)12A] into different genetic backgrounds where the other components of theSD system were varied and tested its ability to cause distortion. The resulting segregation ratios are shown in Table 1. Distortion is indicated by the excess transmission of the Rsp i chromosome relative to the Rsp ss homolog. In the absence of any upwards modifiers of Sd, P[(w+Sd)12A] caused only a low level of distortion (Table 1, row 1). This result corresponds with results obtained for the intact 11.5-kb duplication as well as for nativeSd in the absence of other drive elements (8). To introduce the full constellation of modifier loci, we used derivatives of two different SD chromosomes,SD-5 Rev7 andSD-Mad Rev77, from which theSd locus had been deleted, leaving the other drive elements intact (4, 16). These reverted SDchromosomes have completely lost the ability to distort (Table 1, rows 2 and 4). However, when P[(w+Sd)12A] was introduced into these backgrounds, strong distortion of theRsp ss chromosome was observed in both cases (Table 1, rows 2 and 4). The strength of distortion caused by P[(w+Sd)12A] varied with the insertion site, as full distortion in both backgrounds was observed when the insert was remobilized to new locations (15). The complete distortion caused by P[(w+Sd)12A] in the appropriate background establishes its functional equivalence to Sd. To demonstrate that this distortion depended exclusively on the action of the SD system, comparable crosses were carried out in which both homologs carried a Rsp i allele. As expected, no distortion was observed in these backgrounds whether or not P[(w+Sd)12A] was present (Table 1, rows 3 and 5). Furthermore, distortion by P[(w+Sd)12A] was eliminated in the presence of SD-5 Rev16, a strong suppressor of distortion (4) (Table 1, row 6). Thus, P[(w+Sd)12A] reproduces the behavior of a native Sd in every respect. Therefore, we conclude that the truncated dRanGAP is the functional Sd product (17).

Table 1

Segregation distortion caused by the 12A construct expressing a truncated dRanGAP protein. The second chromosome genotype of the tested males, which includes the indicated components of the SD system, is shown in columns 2 and 3. At least 10 males of the indicated genotypes, either lacking (column 4) or carrying (column 5) the 12A construct inserted on the TM3third chromosome were individually crossed to cn bw females, and the progeny were scored for cn orcn bw eye color markers. lt and pk are eye color and bristle markers, respectively. The data (mean ± SD) are presented as the relative proportion of the total progeny that inherited the Rsp i homolog from the tested males. The total number of offspring (n) counted for each cross is indicated beneath the segregation ratios. The data have been corrected for small intrinsic viability differences associated with the segregating homologs as described in (8).

View this table:

Much recent work in yeast and mammalian cells has shown that nucleocytoplasmic transport is dependent on the small nuclear GTPase Ran (18). Genetic and biochemical data demonstrate that regulators such as the guanine nucleotide exchange factor (RanGEF) and the GTPase-activating protein RanGAP1, which cycle Ran between its GTP- and GDP-bound forms, are also critical (18). Ran and its cofactors have also been implicated in other functions, including cell cycle progression, RNA synthesis and processing, and maintenance of nuclear structure, but it remains unclear whether the effects on these processes are direct or secondary to effects on nuclear transport (18). In either case, the central role played by Ran and its regulators in coordinating key events of nuclear function place them in an ideal position to be subverted by a meiotic drive system. For example, Sd may preferentially impair nuclear transport inRsp s-bearing spermatids at a key stage in their development. The asymmetric effect on Rsp s- but not Rsp i-bearing spermatids could result either from biased distribution of the mutant dRanGAP to the affected spermatids or from enhanced sensitivity ofRsp s-bearing spermatids to impaired nuclear transport.

  • * Present address: Department of Molecular and Cellular Biology, Harvard University, 16 Divinity Avenue, Cambridge, MA 02138, USA.

  • To whom correspondence should be addressed. E-mail: ganetzky{at}


Stay Connected to Science

Navigate This Article