A Test of Double Interspecific Introgression of Nucleoporin Genes in Drosophila

In interspecific hybrids between Drosophila melanogaster and Drosophila simulans, the D. simulans nucleoporin-encoding Nup96sim and Nup160sim can cause recessive lethality if the hybrid does not also inherit the D. simulans X chromosome. In addition, Nup160sim leads to recessive female sterility in the D. melanogaster genetic background. Here, we conducted carefully controlled crosses to better understand the relationship between Nup96sim and Nup160sim. Nup96sim did not lead to female sterility in the D. melanogaster genetic background, and double introgression of Nup96sim and Nup160sim did not generally lead to lethality when one was heterozygous and the other homozygous (hemizygous). It appears that introgression of additional autosomal D. simulans genes is necessary to cause lethality and that the effect of the introgression is dominant to D. melanogaster alleles. Interestingly, the genetic background affected dominance of Nup96sim, and double introgression carrying homozygous Nup96sim and hemizygous Nup160sim resulted in lethality. Thus, Nup96sim and Nup160sim seem to be two components of the same incompatibility.

aspects of the same incompatibility. In this context, it is notable that protein2protein interactions between Nup96 and Nup160 are species-specific, as revealed in yeast sibling species and their hybrids (Leducq et al. 2012).
We conducted interspecific crosses of Drosophila to address the following three questions. (1) Does Nup96 sim lead to female sterility in the D. melanogaster genetic background as seen with Nup160 sim introgression? (2) Does the Nup96 sim and Nup160 sim double introgression lead to lethality when one is heterozygous and the other homozygous (or hemizygous) in the D. melanogaster background (Figure 1, C and D)? (3) Does the Nup96 sim and Nup160 sim double introgression lead to lethality when both are homozygous (or hemizygous) in the D. melanogaster background ( Figure 1E)? Based on these three tests, we ask whether the double introgression of Nup96 sim and Nup160 sim is necessary and sufficient condition for the incompatibility to the gene(s) on the D. melanogastere X chromosome. Dominance of the genes and the possible involvement of different genes to the hybrid lethality will also be discussed.
A P{w + Nup96 sim } e Nup98-96 339 chromosome was made by recombination between P{w + Nup96 sim } and e Nup98-96 339 chromosomes in the w genetic background ( Figure 2). Here w + (68A4; red eye color) and e (93C7-D1; ebony/dark body color) were used as visible markers, and Nup98-96 is at 95B1-5. To confirm that the recombinant chromosome carried the Nup98-96 339 mutation and that it was not lost by rare double recombination between e and Nup98-96 339 , P{w + Nup96 sim } was removed from the established chromosome by further recombination with a wild-type chromosome using the w + and e markers. The resultant chromosome again exhibited recessive lethality that was not complemented by the Nup98-96 deficiencies (Df(3R)Exel9014 and Df(3R)BSC489), thus confirming that the chromosome examined carried Nup98-96 339 . A balancer chromosome, TM3, was used to isolate the recombinant chromosome in a heterozygous state, and CyO and SM1 were used as a chromosome 2 balancer. Int(2L)D+S is a chromosome 2 D. simulans introgression covering two cytological regions that include Nup160 sim (Sawamura et al. 2000). Of note, the Int(2L)D+S introgression also carries other Nup loci (Nup107 and Nup154), but we do not believe that this could affect our overall conclusion of this study. When carrying one X chromosome from D. melanogaster, but females carrying two D. melanogaster X chromosomes can also be obtained using the attached-X system (Presgraves et al., 2003;Tang & Presgraves 2009). The genotypes in (C), (D), and (E) are females carrying two D. melanogaster X chromosomes or males carrying one D. melanogaster X chromosome. necessary, only Nup160 sim was made hemizygous by a deficiency of the Nup160 locus, Df(2L)Nup160M190 (Maehara et al. 2012).

RESULTS
First, we established a D. melanogaster line carrying an extra segment of D. simulans chromosome 3 (including CG10208, Nup98-96, and mbc) inserted at cytological position 68A4 of the same chromosome. Note that Nup98-96 is a dicistronic gene that produces the proteins Nup98 and Nup96 by autoproteolysis (Presgraves et al. 2003). Then, the endogenous Nup98-96 at 95B1-5 of the line was replaced by the recessive lethal Nup98-96 339 mutant allele (Figure 2), which has a stop codon at amino acid position 1726 (therefore, only Nup96 was affected; Presgraves et al. 2003). Thus, we obtained a D. melanogaster chromosome 3 carrying Nup96 sim instead of the D. melanogaster wildtype allele of Nup96. The resultant chromosome (P{w + Nup96 sim } e Nup98-96 339 ) is referred to as the Nup96 sim introgression. Both male and female Nup96 sim introgression homozygotes (and hemizygotes) were viable and fertile, and the strain homozygous for Nup96 sim could be maintained indefinitely. Although females that were homozygous for Nup96 sim and hemizygous over Df(3R)BSC489 exhibited lower fertility than heterozygous controls (x 2 = 94.5, P , 0.001 and x 2 = 6.6576, P , 0.05, respectively), fertility was not decreased in Nup96 sim hemizygotes over Df(3R)Exel9014 (x 2 = 1.5958, P . 0.2) (Table 1). Therefore, Nup96 sim does not lead to female sterility in the D. melanogaster genetic background. We note the possibility that the chromosome harboring Nup96 sim might have a second-site recessive gene or genes responsible for lower female fertility.
Next, to examine possible synergistic and/or additive effects of Nup160 sim and Nup96 sim introgression, we produced w; Int(2L)D+S, Nup160 sim /CyO; Nup96 sim e/+ males by conventional crosses. Then, these males were crossed to females heterozygous for a balancer and a mutation (or a deficiency) of Nup160 or Nup98-96. If the introgressions were behaving similar to the F 1 hybrid, then Nup160 sim / (Nup160 sim or Df-Nup160); Nup96 sim /+ is expected to be lethal; however, that is not what is observed. Instead, the Nup160 sim homozygotes (or hemizygotes) were viable in the Nup96 sim heterozygous background ( Figure 1C and Table 2). If the introgressions were behaving similar to the F 1 hybrid, then Nup160 sim /+; Nup96 sim /(Nup96 sim or Df-Nup96) is expected to be lethal; however, that is not what is observed. Instead, the Nup96 sim homozygotes (or hemizygotes) were viable in the Nup160 sim heterozygous background ( Figure 1D and Table 3). Thus, the Nup96 sim and Nup160 sim double introgression did not lead to lethality when one was heterozygous and the other homozygous (or hemizygous).

DISCUSSION
We found that D. melanogaster females homozygous (or hemizygous) for the Nup96 sim introgression were fertile (Table 1), in contrast to what has been observed for the Nup160 sim introgression, for which eggs produced by homozygotes (or hemizygotes) display karyogamy n  n failure and female pronuclei never fuse to wild-type male pronuclei (Sawamura et al. 2004). Although Nup96 and Nup160 are functionally and structurally in close proximity in the Y-shaped Nup1072160 complex, the effects of interspecific substitution of these two components differed. The structural position of Nup96 and Nup160 might reflect the functional difference; Nup160 is on the surface of the pore ring (Bilokapic and Schwartz 2012;Szymborska et al. 2013) and might have more interactions with other proteins important for NPC function.
We found that flies with genotypes indicated in Figure 1, C and D were viable (Table 2 and Table 3), in contrast to the lethality observed for those with genotypes indicated in Figure 1, A and B (Presgraves et al. 2003;Tang and Presgraves 2009;Sawamura et al. 2010). The primary difference between these flies is the genetic background, with the remaining autosomal genes being from D. melanogaster in our flies and from D. melanogaster and D. simulans (heterozygous) in the previous studies. Apparently the presence of additional autosomal D. simulans genes is necessary to cause lethality, and these genes are dominant to the D. melanogaster alleles. Thus, more genes (maybe encoding other Nups) are involved in this hybrid incompatibility. Nup107 and Nup154 are excluded from the candidates because Int (2L)D+S also carries these genes from D. simulans but did not exhibit the dominant effect. One candidate for the interactant is Nup75, presumably the Drosophila homolog of Nup85. Further investigation of this system is necessary to better understand the genetic mechanisms of reproductive isolation.
Interestingly, dominance of Nup96 sim was changed by the presence of a balancer TM3 (Table 4). Reproductive isolation might be easily affected by the genetic background, as has been suggested in the other hybrid incompatibility (Lhr vs. Hmr) in the same species cross (Matute et al. 2014;Shirata et al. 2014). Finally, double introgression carrying homozygous Nup96 sim and hemizygous Nup160 sim resulted in lethality in the hybrids (Table 4 and Figure 1E). This is the first evidence suggesting that Nup96 sim and Nup160 sim are two components of the same incompatibility.