Sterility of Drosophila with Mutations in the Bloom Syndrome Gene--Complementation by Ku70

See allHide authors and affiliations

Science  30 Mar 2001:
Vol. 291, Issue 5513, pp. 2600-2602
DOI: 10.1126/science.291.5513.2600


The Drosophila Dmblm locus is a homolog of the human Bloom syndrome gene, which encodes a helicase of the RECQ family. We show that Dmblm is identical tomus309, a locus originally identified in a mutagen-sensitivity screen. One mus309 allele, which carries a stop codon between two of the helicase motifs, causes partial male sterility and complete female sterility. Mutant males produce an excess of XY sperm and nullo sperm, consistent with a high frequency of nondisjunction and/or chromosome loss. These phenotypes ofmus309 suggest that Dmblm functions in DNA double-strand break repair. The mutant Dmblm phenotypes were partially rescued by an extra copy of the DNA repair geneKu70, indicating that the two genes functionally interact in vivo.

The human recessive disorder, Bloom syndrome, is characterized by an elevated risk for a wide variety of cancers, as well as immunodeficiency, slow growth, and partial sterility (1). Cells from Bloom syndrome patients exhibit greatly enhanced rates of sister chromatid exchange. The causative mutations are in BLM (2), which encodes a RECQ helicase. Mutations in two other RECQ helicase genes, WRNand RecQ4, are responsible for Werner syndrome and Rothmund-Thomson syndrome, which are both associated with cancer predisposition and premature aging (3,4).

In mice, one BLM mutant line is viable and exhibits Bloom syndrome phenotypes (5). Our previous work showed that theDrosophila Dmblm gene is closely related to BLM(6). Here, we show that Dmblm corresponds to mus309, which had been originally identified in a mutagen-sensitivity screen (7).

In an earlier study, Beall and Rio (8) reported thatmus309 corresponds to a Drosophila homolog ofKu70, whose gene product binds to the ends of double-stranded DNA breaks. Their conclusion was based on low-resolution mapping information and a preliminary indication of partial complementation by a Ku70 transgene of the mutagen sensitivity of mus309. However, Szabad (9) found that deficiency Df(3R)T7, abbreviated Df T7, failed to complement the sterility ofmus309 mutations. This deficiency lies in polytene chromosome region 86F1-2 (10, 11) as opposed to 86E2-3 where DmKu70 had been placed (11). These considerations suggested that Dmblm, which maps to 86F1-4 (6), is a better candidate for mus309than DmKu70.

To test this possibility, we used P element–induced male recombination (12–14) to create a series of deletions in the 86E-F region. These deletions, along with Df T7, were tested against the mutant allele mus309D2 , and four P element–insertion mutations in the area, to generate a more detailed deletion map of 86E-F (Fig. 1A). Themus309 phenotypes tested were mutagen sensitivity and female sterility. The results ruled out the reported identity betweenDmKu70 and mus309, but were consistent with identity between Dmblm and mus309.

Figure 1

(A) Deletion mapping and complementation tests. Each of four deletions was tested against each of four insertion and mus309D2 mutations. Complementation is indicated by “+” and failure to complement by “−”. Flies heterozygous for the mutation and a balancer chromosome were crossed to balancer heterozygotes for each deletion (11, 13). Viability was tested forprosS067102 , l(3)j1D8, orl(3)S11046/deletion. The expected number of insertion/deletion offspring was computed as the average number of all other viable genotypes in the cross. The observed and expected numbers are in (29). Male fertility was tested fortho1 /deletion. Wild-type females were crossed to tho1 /deletion, and the fertility was indicated by a low ratio of pupae to eggs laid. Formus309D2 /deletion, two phenotypes were tested. First, female fecundity was measured as pupae per eggs laid (29) after wild-type males were crossed tomus309D2 /deletion. Second, MMS sensitivity was measured using a cross similar to those used for the viability tests except that a 0.1% MMS solution was added to the culture after egg laying as described (7). These counts ofmus309D2 /deletion offspring are available (29). The end points of deletion Df T7 are bracketed by our complementation tests but are not known exactly. The map was drawn to scale as described (29). (B) Sequence of mus309 mutations.Dmblm DNA was cloned by PCR frommus309D3 /Df T7 andmus309D2 /Df T7 flies and sequenced by standard methods. The intron structure comes from analysis of Drosophila cDNA, and positions of the helicase motifs (black bars) are as determined previously (6). (C) Complementation of MMS sensitivity in mus309mutants by Dmblm transgenes. Two independent insertions on chromosome 2 of P{hspDmblm,w +}, designated 18-8 and 12-2, were tested for their ability to rescue MMS-sensitivity in mus309mutant flies. Females heterozygous for mus309 and the balancer chromosome TM3 (11) were crossed to males of the genotype (transgene or +)/CyO; Df T7/TM3. The progeny were treated with 0.1% MMS as described (7). In crosses with heat shock treatments, larvae were placed at 37°C for 1 hour on days 5, 6, and 7 after the eggs were laid. The numbers reported are the ratios of observed to expected mus309/Df T7 adult progeny, where the expected numbers were computed as the average of the two heterozygous classes, mus309/TM3 andTM3/Df T7, which are approximately equally frequent. We scored an average of 242 progeny per cross.

We next sequenced the Dmblm genes from the two existing mutant alleles of mus309, and found that both carry mutational changes that could potentially impair or abolish the function of Dmblm (Fig. 1B). The less severe allele,mus309D3 , encodes lysine instead of glutamic acid within a conserved region of helicase motif II, plus another amino acid substitution near the COOH-terminus. The more severe allele,mus309D2 , has a stop codon between helicase motifs III and IV, and is expected to yield a severely truncated polypeptide. Both mus309 alleles were induced by ethylmethane sulfonate (7), which generates predominantly G → A transitions (15), and notably, the two suspect changes in the helicase region are both G → A transitions (Fig. 1B).

We constructed a P element–borne transgene carrying theDmblm cDNA driven by an hsp70 heat shock promoter (16), and we tested its ability to rescuemus309 phenotypes. The methylmethane sulfonate (MMS) sensitivity was rescued, at least partially, by each of two transgene insertions tested (Fig. 1C). Position effects normally preclude complete rescue by transgenes of this kind. Both mutantmus309 alleles were rescued by thehsp-Dmblm transgene, and the degree of complementation was enhanced by heat shock for both transgene insertions (Fig. 1C). The sterility phenotype of mus309males and females was also complemented by the Dmblmtransgene (Fig. 2) (17).

Figure 2

Complementation of mus309 male sterility by Dmblm and Ku70 transgenes. Each male of the indicated genotype was crossed to two wild-type females (14). After 2 days of egg laying, the number of embryos produced in each cross was scored, and 5 to 8 days later, the number of pupae generated was scored. Two measures of sterility were used: the percent of inviable offspring, calculated as (inviable eggs)/(total eggs), and the percent of males that were completely sterile. The numbers of eggs counted (left to right) were 544, 520, 1197, 3452, 1553, and 1542. The numbers of males tested were 20, 20, 35, 75, 39, and 26. The Dmblm transgene was insertion 18-8.

To examine the potential interaction of Dmblm andKu70, we obtained two stocks carrying Ku70transgenes. One had a genomic Drosophila sequence,DmKu70, and the other carried a hsp70-driven human cDNA, hsp-HsKu70. Both theDrosophila and human Ku70 transgenes partially rescued the MMS sensitivity and the sterility ofmus309D2 /Df T7 mutants (Figs. 2 and3) (17). This finding is consistent with a previous report that the Drosophilagenomic Ku70 transgene partially rescues MMS sensitivity inmus309D2 /D3 heterozygotes (8). We can rule out the possibility that the twomus309 alleles, as well as Df T7, carry a second lesion at the DmKu70 locus because all three were induced independently on unrelated chromosomes (7, 10). Our interpretation, therefore, is that a third copy of Ku70partially compensates for a mutation at Dmblm.

Figure 3

Complementation of mus309 MMS sensitivity by Dmblm and Ku70 transgenes. Progeny from the indicated cross were grown with MMS (or water) as described (7). The balancer chromosomes were CyO,Roi and TM3, Sb e (11). The numbers plotted are the ratios of observed to expectedmus309D2 /Df T7 progeny, where the expected numbers are as in Fig. 1C with other surviving genotypes being in approximately equal frequencies. Heat shocks were applied only to the lower three groups as described in Fig. 1C. The Dmblmtransgene used was 18-8 in the upper three groups and 12-2 for the lower three. Each of the 18 crosses was set up with 8 males and 20 females, and yielded an average of 687 progeny.

Chromosome loss and nondisjunction are potential phenotypes of mutations affecting DNA repair. We measured the frequency of these events in mus309 mutant males by scoring for XY and nullo-X sperm. The results revealed that both nondisjunction and chromosome loss increased more than tenfold inmus309D2 /Df T7 males relative to the wild-type and heterozygous controls (Table 1). Partial rescue was again seen with thehsp-Dmblm transgene. These events are apparently independent of meiotic recombination, which does not occur inDrosophila males. The absolute frequency of XY sperm was considerably greater than that of nullo-X sperm in themus309 mutant males. This disparity is expected if nondisjunction occurs in the premeiotic germ cells where the nullo-X cells would be inviable.

Table 1

Nondisjunction and chromosome loss inmus309 mutant males, and Dmblm complementation. We used a marked Y chromosome, designated y + YBS, with dominant markers translocated to each end (11) to detect nondisjunction and chromosome loss events. Males carrying this Y chromosome and the indicated genotypes on chromosomes 2 and 3 were crossed to y cv f homozygous females (11). XY sperm were detected as y + cv + BSf + daughters. Nullo-X sperm were indicated byy cv f sons. Cases of marker loss were identified by progeny, usually sons, that inherited one of the two Y-borne markers,y + andBS , without the other. All crosses were set up with individual males so that clustered events could be detected. Out of 10 XY nondisjunctional offspring females frommus309D 2/Df T7 fathers without the transgene, six were from one father.

View this table:

We also observed an elevated frequency of sperm in which one of the two dominant markers on the Y chromosome had been lost in the germ line of the mus309 mutant males. This effect was also partially rescued by the hsp-Dmblm transgene (Table 1). These two markers [y +and BS (11)] were present on opposite ends of the Y chromosome. We postulate that these events represent chromosome breakage and partial loss of the Y chromosome. To investigate this possibility, we recovered 23 such sons frommus309D2 /D3 fathers and tested them for fertility. If part of the Y chromosome had been lost, it is likely that one or more of the male fertility factors located on the Y chromosome would be lost as well. All but one of the 23 males proved to be sterile, consistent with the chromosome breakage hypothesis.

The identification of mus309 as the DrosophilaBloom syndrome gene opens up new avenues for studying its function. First is the unexpected finding that Dmblm andKu70 interact. One possible basis for this interaction is that an extra copy of Ku70 facilitates use of alternative pathways of double-strand break repair. However, more direct interactions cannot be ruled out. Indeed, biochemical interactions between Ku70 and another RECQ family member, WRN, have been observed (18, 19). In addition, both Ku70 and BLM have DNA helicase activity (20, 21), which suggests possible redundancy.

The elevated frequency of chromosomal nondisjunction and chromosome loss in the germ line is likely to promote frequent germ cell loss during development. This process could explain the sterility seen in mus309 mutants, as well as in humans with Bloom syndrome. In addition, previous studies have shown that mutant alleles of mus309, as well as those of mei41, aDrosophila homolog of the ataxia telangiectasiagene, increase sensitivity to chromosome breakage (8,22, 23). However, mei41, but notmus309, mutants display defects in a DNA damage checkpoint assay (24, 25). In contrast to a recent proposal that BLM functions to prevent chromosome breakage during replication (26, 27), our data, combined with previous studies of mus309 phenotypes, suggest that theDrosophila BLM homolog functions in a repair role after double-strand break formation. BLM could perform this role during replication, as well as in response to incidental DNA damage.

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


Stay Connected to Science

Navigate This Article