an indel polymorphism in the hybrid incompatibility gene ... · a population survey revealed that...

15
INVESTIGATION An Indel Polymorphism in the Hybrid Incompatibility Gene Lethal Hybrid Rescue of Drosophila Is Functionally Relevant Shamoni Maheshwari and Daniel A. Barbash 1 Department of Molecular Biology and Genetics, Cornell University, Ithaca, New York 14853 ABSTRACT Hybrid incompatibility (HI) genes are frequently observed to be rapidly evolving under selection. This observation has led to the attractive conjecture that selection-derived protein-sequence divergence is culpable for incompatibilities in hybrids. The Drosophila simulans HI gene Lethal hybrid rescue (Lhr) is an intriguing case, because despite having experienced rapid sequence evolution, its HI properties are a shared function inherited from the ancestral state. Using an unusual D. simulans Lhr hybrid rescue allele, Lhr 2 , we here identify a conserved stretch of 10 amino acids in the C terminus of LHR that is critical for causing hybrid incompatibility. Altering these 10 amino acids weakens or abolishes the ability of Lhr to suppress the hybrid rescue alleles Lhr 1 or Hmr 1 , respectively. Besides single- amino-acid substitutions, Lhr orthologs differ by a 16-aa indel polymorphism, with the ancestral deletion state xed in D. melanogaster and the derived insertion state at very high frequency in D. simulans. Lhr 2 is a rare D. simulans allele that has the ancestral deletion state of the 16-aa polymorphism. Through a series of transgenic constructs we demonstrate that the ancestral deletion state contributes to the rescue activity of Lhr 2 . This indel is thus a polymorphism that can affect the HI function of Lhr. W HAT evolutionary forces drive speciation? A signi- cant step toward answering this question has been the identication of hybrid incompatibility (HI) genes, that is, genes with incompatible substitutionsthat cause break- down in interspecic hybrids. The next challenge is describ- ing the evolutionary basis for the origin of such incompatible substitutions. The classic DobzhanskyMuller (DM) model elegantly explains how substitutions incompatible only in an interspecic context can evolve; however, it is agnostic on the nature of the intraspecic evolutionary forces that cause them (Presgraves 2010; Maheshwari and Barbash 2011). The model is equally consistent with incompatible substitu- tions evolving as functionally neutral mutations drifting to xation or as functionally advantageous mutations being driven to xation by natural selection. It is therefore particularly intriguing that so many HI genes show high rates of sequence divergence driven by positive selection. If this divergence corresponds to the incompatible substitutions then there is a direct link between the pheno- type under selection and HI. This is very likely for the hybrid sterility gene OdsH, where the signature of selection is con- centrated within the DNA-binding homeodomain, because functional analysis of OdsH orthologs has implicated diver- gent DNA-binding activity in hybrid incompatibility (Ting et al. 1998; Bayes and Malik 2009). However, such a direct link between sequence divergence and function remains to be established for other rapidly evolving HI genes. The HI gene Lethal hybrid rescue (Lhr) poses an interest- ing paradox. Lhr causes F 1 hybrid male lethality in crosses between Drosophila melanogaster and D. simulans (Watanabe 1979; Brideau et al. 2006). The classic DM model describes HI as the negative ectopic interaction between two derived loci, thus setting up the expectation that selection-driven divergence of Lhr led to incompatible substitutions in one of the hybridizing lineages. Surprisingly, however in transgenic assays, Lhr orthologs from both hybridizing species cause hybrid dysfunction (Brideau and Barbash 2011; Maheshwari and Barbash 2012). This argues against the expectation that the hybrid lethal activity of Lhr is solely the outcome of selection-driven substitutions in its protein coding sequence (CDS) specic to D. simulans. Moreover, our recent results Copyright © 2012 by the Genetics Society of America doi: 10.1534/genetics.112.141952 Manuscript received May 10, 2012; accepted for publication July 17, 2012 Supporting information is available online at http://www.genetics.org/lookup/suppl/ doi:10.1534/genetics.112.141952/-/DC1. 1 Corresponding author: Department of Molecular Biology and Genetics, Cornell University, Ithaca, NY 14853. E-mail: [email protected] Genetics, Vol. 192, 683691 October 2012 683

Upload: others

Post on 01-Nov-2020

0 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: An Indel Polymorphism in the Hybrid Incompatibility Gene ... · A population survey revealed that the ancestral non-insertion form is segregating at a very low frequency in some D

INVESTIGATION

An Indel Polymorphism in the HybridIncompatibility Gene Lethal Hybrid Rescue

of Drosophila Is Functionally RelevantShamoni Maheshwari and Daniel A. Barbash1

Department of Molecular Biology and Genetics, Cornell University, Ithaca, New York 14853

ABSTRACT Hybrid incompatibility (HI) genes are frequently observed to be rapidly evolving under selection. This observation has led tothe attractive conjecture that selection-derived protein-sequence divergence is culpable for incompatibilities in hybrids. The Drosophilasimulans HI gene Lethal hybrid rescue (Lhr) is an intriguing case, because despite having experienced rapid sequence evolution, its HIproperties are a shared function inherited from the ancestral state. Using an unusual D. simulans Lhr hybrid rescue allele, Lhr2, we hereidentify a conserved stretch of 10 amino acids in the C terminus of LHR that is critical for causing hybrid incompatibility. Altering these10 amino acids weakens or abolishes the ability of Lhr to suppress the hybrid rescue alleles Lhr1 or Hmr1, respectively. Besides single-amino-acid substitutions, Lhr orthologs differ by a 16-aa indel polymorphism, with the ancestral deletion state fixed in D. melanogasterand the derived insertion state at very high frequency in D. simulans. Lhr2 is a rare D. simulans allele that has the ancestral deletion stateof the 16-aa polymorphism. Through a series of transgenic constructs we demonstrate that the ancestral deletion state contributesto the rescue activity of Lhr2. This indel is thus a polymorphism that can affect the HI function of Lhr.

WHAT evolutionary forces drive speciation? A signifi-cant step toward answering this question has been

the identification of hybrid incompatibility (HI) genes, thatis, genes with “incompatible substitutions” that cause break-down in interspecific hybrids. The next challenge is describ-ing the evolutionary basis for the origin of such incompatiblesubstitutions. The classic Dobzhansky–Muller (D–M) modelelegantly explains how substitutions incompatible only in aninterspecific context can evolve; however, it is agnostic onthe nature of the intraspecific evolutionary forces that causethem (Presgraves 2010; Maheshwari and Barbash 2011).The model is equally consistent with incompatible substitu-tions evolving as functionally neutral mutations drifting tofixation or as functionally advantageous mutations beingdriven to fixation by natural selection.

It is therefore particularly intriguing that so many HI genesshow high rates of sequence divergence driven by positive

selection. If this divergence corresponds to the incompatiblesubstitutions then there is a direct link between the pheno-type under selection and HI. This is very likely for the hybridsterility gene OdsH, where the signature of selection is con-centrated within the DNA-binding homeodomain, becausefunctional analysis of OdsH orthologs has implicated diver-gent DNA-binding activity in hybrid incompatibility (Tinget al. 1998; Bayes and Malik 2009). However, such a directlink between sequence divergence and function remains tobe established for other rapidly evolving HI genes.

The HI gene Lethal hybrid rescue (Lhr) poses an interest-ing paradox. Lhr causes F1 hybrid male lethality in crossesbetween Drosophila melanogaster and D. simulans (Watanabe1979; Brideau et al. 2006). The classic D–M model describesHI as the negative ectopic interaction between two derivedloci, thus setting up the expectation that selection-drivendivergence of Lhr led to incompatible substitutions in one ofthe hybridizing lineages. Surprisingly, however in transgenicassays, Lhr orthologs from both hybridizing species causehybrid dysfunction (Brideau and Barbash 2011; Maheshwariand Barbash 2012). This argues against the expectation thatthe hybrid lethal activity of Lhr is solely the outcome ofselection-driven substitutions in its protein coding sequence(CDS) specific to D. simulans. Moreover, our recent results

Copyright © 2012 by the Genetics Society of Americadoi: 10.1534/genetics.112.141952Manuscript received May 10, 2012; accepted for publication July 17, 2012Supporting information is available online at http://www.genetics.org/lookup/suppl/doi:10.1534/genetics.112.141952/-/DC1.1Corresponding author: Department of Molecular Biology and Genetics, CornellUniversity, Ithaca, NY 14853. E-mail: [email protected]

Genetics, Vol. 192, 683–691 October 2012 683

Page 2: An Indel Polymorphism in the Hybrid Incompatibility Gene ... · A population survey revealed that the ancestral non-insertion form is segregating at a very low frequency in some D

argue that the divergent hybrid lethal activities of Lhr ortho-logs can be largely attributed to their asymmetric expressionin the hybrid background (Maheshwari and Barbash 2012).The D. simulans Lhr allele is expressed twofold higher thanthe D. melanogaster ortholog in the F1 hybrid. But it is stillan open question whether divergence of the CDS might alsobe contributing to the differential hybrid lethal effects of Lhr.

Lhr orthologs have �50 fixed differences between D. mel-anogaster and D. simulans scattered throughout a proteinsequence of only �330 residues. Additionally, Lhr from eachof the sibling species D. simulans, D. mauritiana, andD. sechellia has a 16-amino acid (aa) insertion relative to theD. melanogaster ortholog. The insertion is absent in outgroupspecies and is therefore identified as a derived state, specificto the common ancestor of the sibling species. This 16-aainsertion is also interesting because it may affect the struc-ture of a predicted leucine zipper in the LHR protein andhad been proposed as a candidate for mediating functionaldifferences between the D. melanogaster and D. simulans Lhrorthologs (Brideau et al, 2006).

The discovery of D. simulans Lhr2 motivated us to furtherexplore the effect, if any, of this 16-aa region on the hybridlethal activity of Lhr orthologs. Lhr2 partially suppresses hybridmale lethality, strongly suggesting that it is a loss-of-functionallele. Interestingly, Lhr2 lacks the 16-aa insertion foundin most other D. simulans Lhr alleles. However, the Lhr2

allele also has a complex deletion in its C terminus withina sequence of high conservation (Supporting Information,Figure S1). Furthermore, it was not tested whether Lhr2 iswild type in expression level, which is critical because Lhr1

is strongly reduced in expression (Brideau et al. 2006). Thus,even if the hybrid rescue property of the D. simulans Lhr2

strain is a function of the unusual CDS of the Lhr2 allele, itis unclear whether one or both of the aforementioned twomajor mutations are responsible for its hybrid rescue activity.

A population survey revealed that the ancestral non-insertion form is segregating at a very low frequency insome D. simulans populations (Nolte et al. 2008). Nolte et al.(2008) tested two D. simulans strains in hybrid crosses thatcarried Lhr alleles lacking the 16-aa insertion but wild typeat the C terminus. Neither of these strains produced viablehybrid sons, leading them to conclude that the hybrid res-cue property of the D. simulans Lhr2 strain is not caused bythe ancestral noninsertion form of the 16-aa region, leavingthe complex C-terminal mutation as the most likely candi-date. However, whether the presence or absence of this16-aa region makes any contribution either to functionaldifferences between mel-Lhr and sim-Lhr, or to the hybridrescue properties of Lhr2, remains untested. Here we describea series of transgenic assays to address these questions.

Materials and Methods

Drosophila stocks and culturing

All crosses were done at room temperature or at 18� whereexplicitly stated. At least two replicates were done for each

cross. Each interspecific cross was initiated with �15–20 one-day-old D. melanogaster virgin females and �30–40 three- to4-day-old sibling-species males. Genetic markers, deficien-cies, and balancer chromosomes are described on FlyBase(McQuilton et al. 2012).

Nomenclature

The abbreviations mel-Lhr and sim-Lhr refer to the Lhr ortho-logs from D. melanogaster and D. simulans, respectively. Werefer generically to the 16-aa region that is present in sim-Lhrand absent in mel-Lhr as the “16-aa indel.” Because it is a de-rived insertion in the D. simulans lineage but absent in thesim-Lhr2 allele, we refer to it as the “16-aa deletion” in sim-Lhr2 and in mel-Lhr and as the “16-aa insertion” in sim-Lhr.

DNA constructs

PCR primers are listed in Table S1. To generate constructsfor transgenic experiments (Figure 1), first the wild-type LhrCDS in p{sim-Lhr} from (Maheshwari and Barbash 2012)was replaced by the Lhr2 CDS using a three-piece fusionPCR strategy. The first and last PCR products, containingupstream and downstream genomic regions, were amplifiedusing p{sim-Lhr} as the template, with primer pairs 691/938 and 941/664, respectively. The central PCR product con-taining the Lhr2 CDS was amplified from D. simulans Lhr2

genomic DNA, with primer pair 939/940. The three over-lapping PCR products were then used as templates for thefusion PCR using primers 691/664, cloned into the pCR-BluntII vector to create the plasmid p{sim-Lhr2}, and se-quenced completely.

A triple-HA tag in frame with the C terminus of Lhr2 CDSwas synthesized using a two-piece fusion PCR strategy. Twooverlapping PCR products were amplified using p{sim-Lhr2}as the template, with primer pairs 882/728 and 729/664.Fusion PCR was then performed using these products asthe templates with primers 882/664, and the resulting prod-uct was TOPO cloned into the pCR-BluntII vector. This

Figure 1 Schematic of the Lhr2 constructs. The mel-Lhr-HA and sim-Lhr-HA constructs are described in (Maheshwari and Barbash 2012). All otherconstructs contain the full sim-Lhr2 coding sequences fused to the HAepitope tag (green) with the UTRs and genomic DNA from the D. simulansw501 strain. White boxes represent the 16-aa deletion and C-terminal muta-tions. Triangles represent replacement of Lhr2 CDS with sequence fromwild-type D. simulans Lhr.

684 S. Maheshwari and D. A. Barbash

Page 3: An Indel Polymorphism in the Hybrid Incompatibility Gene ... · A population survey revealed that the ancestral non-insertion form is segregating at a very low frequency in some D

intermediate construct was digested with SacII and ApaIand the fragment released was subcloned into p{simLhr2},generating p{sim-Lhr2-HA}. The full insert was sequencedcompletely and subcloned into the multiple cloning site ofpCasper4\attB using NotI and KpnI restriction enzymes.

To synthesize the construct p{sim-Lhr2-HA + 16aa},the 16-aa insertion was inserted into the Lhr2 CDS usinga two-piece fusion PCR strategy. The two overlapping PCRproducts were amplified using p{sim-Lhr2-HA} as the tem-plate, with primer pairs 691/945 and 946/664. These frag-ments were used as templates for the fusion PCR withprimers 691/664, and the gel-purified product was TOPOcloned into the pCR-BluntII vector and sequenced com-pletely. The insert was then subcloned into pCasper4\attBexactly as in p{sim-Lhr2-HA}. The construction of p{sim-Lhr2-HA +Cter}, where the complex mutation in the C ter-minal mutation in Lhr2 CDS was replaced by 10 residuesof wild-type D. simulans Lhr sequence, was done as aboveusing primer pairs 691/942 and 943/664.

For yeast two-hybrid experiments, the Lhr2 CDS was am-plified from genomic DNA using primer pair 404/405 andcloned into pENTR-DTOPO (Invitrogen) according to themanufacturer’s instructions, and verified by sequencing. Theentry vector was recombined with the destination vectors ina standard LR Clonase (Invitrogen)-mediated reaction. Thedestination vectors used were pGADT7-AD and pGBKT7-DNA-BD (K. Ravi Ram, A. Garfinkel, and M. F. Wolfner,personal communication).

Transgenic fly lines

FC31-mediated transformants of D. melanogaster wereperformed by Genetic Services. The integration site usedwas M{3xP3-RFP.attP}ZH-86Fb at cytological position 86Fb(Bischof et al. 2007). Site specificity of integrations weretested using the PCR assays described in Maheshwari andBarbash (2012).

Recombination mapping of the D. simulans Lhr2

rescue activity

The D. simulans Lhr2 rescue strain was outcrossed to thenonrescuing D. simulans v strain. From this, seven indepen-dent recombination lines were established by backcrossing8–10 F1 daughters to 8–10 males from the D. simulans vstrain. Sons from this cross were used to set up three hybridcrosses. Each hybrid cross was set up with �30 recombinantsons, aged for 3 days, and twenty 0- to 1-day-old virginD. melanogaster w1118 females. Individual viable F1 hybridsons, which by definition inherit the mutation responsiblefor rescue, were PCR genotyped for their Lhr alleles. To de-termine whether hybrid sons inherited the wild-type Lhr orthe Lhr2 allele from the D. simulans father, we used primerpairs 409/410 to PCR across the 16-aa indel. If sons in-herit wild-type D. simulans Lhr we expect to see two bands,the smaller band corresponding to the ancestral state inD. melanogaster Lhr and the larger size corresponding to theinsertion in wild-type D. simulans Lhr; however, if they in-

herit the Lhr2 alelle, we expect to see only one band corre-sponding to the ancestral state.

RT–PCR, immunofluorescence, and yeast two hybrid

RT–PCR and immunofluorescence were performed as previ-ously described (Maheshwari and Barbash 2012). Yeast two-hybrid assays were performed as in Brideau and Barbash(2011).

Sequence and phylogenetic analyses

We examined Lhr sequences from a recent large-scale rese-quencing of D. melanogaster populations and found all 158strains contain the 16-aa deletion (Mackay et al. 2012). Wealso searched the short-read archive from this project, usingas the query a 100-bp sequence frommel-Lhr flanking the siteof the 16-aa indel. All 26 traces from 454 sequencing fullymatched the query. In combination with our previous poly-morphism sampling of mel-Lhr (Brideau et al. 2006), weconclude that D. melanogaster is fixed for the deletion formof the 16-aa indel. The phylogenetic tree was built by MEGA5.05 using the maximum parsimony method (Tamura et al.2011). The Lhr alleles used for the analysis are published inBrideau et al. (2006). For phylogenetic analysis the regioncorresponding to the C-terminal mutation in Lhr2was excludedfrom the alignment.

Results

D. simulans Lhr2 is mutant in its coding sequence

A cross between wild-type D. melanogaster females andD. simulans males produces only sterile daughters and nosons. The genetic basis of male lethality appears to be fixedbetween the two species, as crosses between many differentwild-type strains fail to produce hybrid sons (Sturtevant1920; Lachaise et al. 1986). The only two exceptions arestrains with mutations in D. melanogaster Hmr or D. simulansLhr (Watanabe 1979; Hutter and Ashburner 1987).

Although we and others implicitly assumed in previousanalyses that rescue in the D. simulans Lhr2 strain is due toits unusual Lhr allele, this point has not been established(Brideau et al. 2006; Nolte et al. 2008). We therefore firstdid a crude mapping experiment to test whether the hybridrescue function is associated with the Lhr2 locus. We out-crossed D. simulans Lhr2 to wild-type D. simulans and testedfor linkage between the Lhr2 locus and hybrid rescue. Wegenotyped by PCR 48 viable hybrid sons, which by definitionhave inherited the rescue locus, and found that all of themalso inherited the Lhr2 allele from the D. simulans parent.This pattern of cosegregation supports the hypothesis thatthe Lhr2 allele is responsible for suppressing hybrid malelethality instead of an unrelated mutation segregating in thesame genetic background.

We next sequenced 4 kb of genomic DNA spanning theLhr2 locus and found only several SNPs but no insertions,deletions, or rearrangements in its noncoding regions,suggesting that the Lhr2 allele is unlikely to be mutant

Functional Effect of an Indel Polymorphism 685

Page 4: An Indel Polymorphism in the Hybrid Incompatibility Gene ... · A population survey revealed that the ancestral non-insertion form is segregating at a very low frequency in some D

in its expression. Using quantitative RT–PCR we deter-mined that Lhr expression in D. simulans Lhr2 is not sig-nificantly different from wild type (t test, P = 0.2) (Figure2A), demonstrating that the hybrid rescue property of D.simulans Lhr2 is different from the original rescue alleleLhr1, which is an expression mutant having nearly undetect-able levels of Lhr. The Lhr2 CDS is unusual in two respects(Figure 1, Figure S1). First, Lhr2 lacks the 16-aa insertion thatis present in frequencies near fixation in other sim-Lhralleles. Second, Lhr2 has a complex mutation in a conservedsequence near its C terminus, which includes a 12-bp in-frame deletion and nonsynonymous mutations causingunique substitutions in 6 adjacent amino acids.

Considering that the D. simulans Lhr2 allele containsthe melanogaster-like ancestral state at the 16-aa indel, itraised the possibility that Lhr2 is a recent introgression ofD. melanogaster Lhr into D. simulans. This was rejected,however, by phylogenetic analysis that firmly groups Lhr2

with alleles from the sibling species (Figure 2B). Interest-ingly, Lhr2 appears to be a relatively old allele that clustersseparately from other sim-Lhr alleles.

To test conclusively whether the coding sequence of theLhr2 allele is defective for hybrid lethal activity, we used atransgenic assay to compare it with wild-type sim-Lhr. Weused the FC31 site-specific integration system to generatea D. melanogaster strain carrying a D. simulans Lhr2 trans-

gene at the attP86Fb site on the third chromosome (Figure 1).The Lhr2 CDS was C-terminally tagged with HA and placedunder the control of wild-type D. simulans regulatory se-quences (from strain w501), to generate the F{sim-Lhr2-HA}construct.

Hybrid lethal activity was assayed using the D. simulansLhr1 complementation test (Maheshwari and Barbash 2012).D. melanogaster mothers heterozygous for an experimentalor control transgene were crossed to D. simulans Lhr1 fathers,Lhr1 being a loss-of-function mutation that acts as a domi-nant suppressor of HI. If the transgene has hybrid lethalactivity, it is expected to suppress rescue by the Lhr1 muta-tion. In the control cross with F{sim-Lhr-HA} no hybrid sonsinheriting the transgene were recovered (Table 1, cross 1).This full suppression of rescue is consistent with our pre-vious results (Maheshwari and Barbash 2012). In contrast,F{sim-Lhr2-HA} only partially suppressed rescue, with via-bility in the range of 35–40% relative to the control class(Table 1, cross 2). This assay demonstrates that the Lhr2

CDS has significantly reduced ability to cause HI but it isnot a null allele. This conclusion is consistent with the ob-servation that D. simulans Lhr1 rescues more strongly thanD. simulans Lhr2. When crossed to D. melanogaster w1118

at room temperature, the viability of hybrid males withD. simulans Lhr1 is �73% relative to hybrid females (51 F1males and 70 F1 females), while with D. simulans Lhr2, it

Figure 2 The Lhr2 allele is not an expression mutant or a D. melanogaster introgression. (A) Quantitative RT–PCR analysis comparing Lhr expression inD. simulans Lhr2 with D. melanogaster w1118 and D. simulans v strains, both of which are Lhr+. RNA was isolated from 6- to 10-hr-old embryos.Lhr abundance was measured relative to rpl32. Expression levels were normalized by setting D. melanogaster w1118 strain to 1. Error bars representstandard error among biological replicates, n $ 3. (B) Evolutionary history of Lhr2 in the melanogaster subgroup was inferred using the maximumparsimony method. Arrowhead indicates the branch on which the 16-aa insertion originated. Percentage of replicate trees in which the associated taxaclustered together in the bootstrap test (500 replicates) is shown next to the branches. Bootstrap values are not shown for the terminal nodes within theD. melanogaster and D. simulans clades.

686 S. Maheshwari and D. A. Barbash

Page 5: An Indel Polymorphism in the Hybrid Incompatibility Gene ... · A population survey revealed that the ancestral non-insertion form is segregating at a very low frequency in some D

is �49% (107 F1 males and 219 F1 females). Lower levelsof rescue with Lhr2 compared to Lhr1 were also observed ina previous study (Barbash 2010).

Assaying the function of the two major structuralmutations in Lhr2

To individually test the contribution of the complex C-terminalmutation and the 16-aa deletion to hybrid lethal activity,each was individually replaced in sim-Lhr2-HA with wild-type sequence to generate F{sim-Lhr2-HA,+Cter} andF{sim-Lhr2-HA,+16aa}, respectively (Figure 1). Initial ex-periments suggested that the presence or absence of theC-terminal mutation has a much more significant impacton Lhr function compared to the 16-aa indel. We thereforecompared them to different references in our genetic assays.For F{sim-Lhr2-HA,+Cter}, where we reverted the C-terminalmutation to the wild-type sequence, we compared its activ-ity to the wild-type F{sim-Lhr-HA} control and found thatit also fully suppresses rescue (Table 1, cross 3). This resultdemonstrates that the conserved C-terminal region is essen-tial for wild-type Lhr function. Because F{sim-Lhr2-HA,+Cter}contains the ancestral deletion state of the 16-aa indel, thisresult also demonstrates that the presence or absence of the16-aa indel region in an otherwise wild-type sim-Lhr alleledoes not affect the ability of sim-Lhr to suppress hybrid rescueby Lhr1.

For F{sim-Lhr2-HA,+16aa}, where we added the 16-aainsertion to the Lhr2 allele, a comparison to the constructF{sim-Lhr2-HA} tests whether the 16-aa deletion has anyfunctional effect in the background of an allele that is par-tially impaired because it carries the C-terminal deletion. Inour Lhr1 complementation assay, we detected a significantdifference in viability between the two genotypes of hybridmales [Table 1, cross 2 vs. 4, two-tailed Fisher’s exact test(FET), P = 0.000]. The relative viability of hybrid sons inher-iting F{sim-Lhr2-HA,+16aa} was reduced to �16% com-pared to �35–41% for F{sim-Lhr2-HA}. This demonstratesthat having the ancestral deletion state significantly contrib-utes to the hybrid rescue activity of Lhr2. This result thus showsthat the polymorphic 16-aa indel does affect Lhr function, atleast in the presence of the second C-terminal mutation.

To further explore the functional effects of the 16-aaindel, we turned to a more sensitive genetic assay for Lhrfunction involving its interacting partner Hmr. We have pre-viously shown that in the background of the hypomorphicHmr1 mutation, Lhr orthologs exhibit significantly differentdegrees of hybrid lethality (Maheshwari and Barbash 2012).We therefore introduced each of our Lhr2 transgenes intoan Hmr1 mutant background and tested the effect of thetransgenes on hybrid male viability in crosses to D. mauritiana(Table 2). D. mauritiana was chosen as the male parent be-cause in interspecific crosses between D. melanogaster femalesand sibling species males, D. mauritiana hybrids show thehighest viability (Hutter and Ashburner 1987). The crosseswere also done at both room temperature and 18� becausehybrid viability is temperature dependent, with viability in-creasing at lower temperatures (Hutter and Ashburner 1987).Crosses with the wild-type Lhr transgenes recapitulated ourprevious experiments (Maheshwari and Barbash 2012): Hmr1

hybrid sons carrying sim-Lhr-HA were essentially inviable atroom temperature, while hybrid sons inheriting mel-Lhr-HAhad 32–44% viability (Table 2, crosses 1 and 2).

Surprisingly, hybrid sons carrying the F{sim-Lhr2-HA}transgene were fully viable at 18� relative to their controlbrothers and had substantially higher viability relative tocontrol brothers at room temperature (138.8–144.3%, Table2, cross 3). Lhr2 is therefore acting as a null allele or evenan antimorph in this Hmr1 interaction assay. Reverting theC-terminal mutation to the wild-type sequence fully restoredhybrid lethal effects to wild-type levels, with hybrid viabilitynot significantly different than the wild-type F{sim-Lhr-HA}transgene (Table 2, cross 2 vs. 4, two-tailed FET, P = 1.0 atboth room temperature and 18�). These results demonstratethat the C-terminal region is critical for the strong loss-of-function/antimorphic activity of Lhr2 in this assay.

We then compared the F{sim-Lhr2-HA} allele to F{sim-Lhr2-HA,+16aa}, which differ only by the presence or ab-sence of the 16-aa indel. We found that although hybridsons inheriting F{sim-Lhr2-HA,+16aa} have viabilities com-parable to the control class, the relative viabilities of hybridsons inheriting this transgene are less than that for theF{sim-Lhr2-HA} transgene. This reduction in viability is

Table 1 Testing the two major mutations in Lhr2 for suppression of hybrid rescue by D. simulans Lhr1

No. of hybrid males

Relative viability ofF{ } males (%)Cross Transgenic construct No. of hybrid females

Genotype 1+/Lhr1; +/+

Genotype 2+/Lhr1; F{ }/+

1 F{sim-Lhr-HA} 135 74 0 02 F{sim-Lhr2-HA} 494 226 80 35.4

308 185 75 40.543 F{sim-Lhr2-HA+Cter} 269 175 0 0

187 104 0 04 F{sim-Lhr2-HA+16aa} 337 178 28 15.73

224 164 26 15.85

Crosses were between D. melanogaster females heterozygous for the different transgenes (genotype w; F{ }/+) and D. simulans Lhr1 males. The transgenes carried a copy of thew+ gene so the hybrid sons inheriting the transgene, +/Lhr1; F{ }/+ (genotype 2) were distinguished from their +/Lhr1; +/+ siblings (genotype 1) by their eye color. All crosses werecarried out at room temperature. Relative viability is the ratio of the number of hybrid sons inheriting the transgene (genotype 2) compared to the control class (genotype 1).

Functional Effect of an Indel Polymorphism 687

Page 6: An Indel Polymorphism in the Hybrid Incompatibility Gene ... · A population survey revealed that the ancestral non-insertion form is segregating at a very low frequency in some D

significant at 18� (Table 2, cross 3 vs. 5, two-tailed FET, P =0.022). These results again show that the 16-aa indel doeshave a detectable effect on Lhr function in the backgroundof the C-terminal mutation.

The molecular properties of the LHR2 protein

We next asked whether the LHR2 mutant protein is alteredfor molecular functions of LHR. LHR localizes to specificregions of heterochromatin through interaction with hetero-chromatin protein 1 (HP1) (Brideau et al. 2006; Greil et al.2007; Brideau and Barbash 2011). We therefore askedwhether the reduced hybrid lethal activity of Lhr2 was reflect-ing a defect in heterochromatin association. We performedyeast two-hybrid assays and found that the interaction be-tween LHR2 and HP1 was indistinguishable from the wild-typecontrol (Figure 3A). Consistent with this result, LHR2-HA lo-calized to heterochromatin in vivo and immuno-FISH experi-ments showed colocalization with the dodeca satellite ina manner indistinguishable from wild-type LHR (Maheshwariand Barbash 2012), providing further support for wild-typeassociation with heterochromatin (Figure 3B). We concludethat the reduced hybrid lethal activity of Lhr2 is not becauselocalization to heterochromatin is defective.

Discussion

The C-terminal mutation in Lhr2 identifies a regioncritical for Lhr function

In this study, we demonstrate conclusively that Lhr2 is a mu-tant allele of the Lhr hybrid lethality gene and further show

that its mutant properties are due to changes in its CDS. Lhr2

is a weaker mutant allele than Lhr1 in its hybrid rescueability, and in transgenic assays sim-Lhr2 complements Lhr1

more weakly than does a wild-type sim-Lhr allele (Table 1).By these criteria, Lhr2 would appear to be hypomorphic. Incontrast, results from the Hmr1 interaction assay suggest thatLhr2 has no wild-type activity or is even antimorphic (Table 2).

We therefore devised modified Lhr2 alleles to individuallyassay specific regions for effects on hybrid lethal activity(Figure 1). We find that a highly conserved stretch of10 residues in the C terminus of Lhr is critical for wild-typelevels of hybrid lethal activity in both genetic assays. Thisconclusion is consistent with the observations of Nolte et al.(2008) who found wild-type hybrid lethal activity for twoD. simulans Lhr alleles that have the deletion state for the16-aa indel but are wild type for the C-terminal mutation.Because this region is highly similar between mel-Lhr andsim-Lhr, this result also supports published results that Lhrorthologs from both species can cause incompatibility (Brideauand Barbash 2011; Maheshwari and Barbash 2012). Our datahere suggest that the C-terminal region is especially criticalfor interactions with Hmr because F{sim-Lhr2-HA} has nowild-type activity for complementing Hmr1 (Table 2, cross 3),but whether this reflects a direct physical interaction remainsunknown. The Lhr1 complementation assay is perhaps morestraightforward to interpret since one is asking whether dif-ferent Lhr alleles complement a loss-of-function allele of Lhr.Since only half of the hybrid sons inheriting the F{sim-Lhr2-HA} transgene are viable (Table 1 cross 2), it is clear thatthe C-terminal deletion does not fully account for the hybrid

Table 2 Testing the two major mutations in Lhr2 for suppression of hybrid rescue by D. melanogaster Hmr1

No. of hybrid males

Transgenicconstruct Temp. No. of hybrid females

Genotype 1Hmr1/Y; +/+

Genotype 2Hmr1/Y; f{ }/+

Relative viability off{ } males (%)

1 F{mel-Lhr-HA} RT 446 78 25 32.1RT 324 67 30 44.818� 462 184 127 69.018� 689 265 140 52.8

2 F{sim-Lhr-HA} RT 305 55 1 1.8RT 180 35 0 0.018� 692 283 90 31.818� 504 198 111 56.1

3 F{sim-Lhr2-HA} RT 354 79 114 144.3RT 361 80 111 138.818� 782 264 283 107.218� 742 253 250 98.8

4 F{sim-Lhr2-HA+Cter} RT 226 47 1 2.1RT 382 59 0 0.018� 561 236 106 44.918� 393 197 74 37.6

5 F{sim-Lhr2-HA+16aa } RT 86 32 34 106.3RT 182 52 57 109.618� 538 253 214 84.618� 373 191 155 81.2

The different Lhr transgenes were tested for interaction with an Hmr hypomorphic allele, Hmr1. D. melanogaster w Hmr1 v; F{transgene, w+}/+ femaleswere mated to D. mauritiana Iso105 males. Hybrid male progeny that inherit the transgene are orange eyed (genotype 2), while the sibling brothersare white eyed (genotype 1). Relative viability is the ratio of the number of hybrid sons inheriting the transgene compared to the control class.

688 S. Maheshwari and D. A. Barbash

Page 7: An Indel Polymorphism in the Hybrid Incompatibility Gene ... · A population survey revealed that the ancestral non-insertion form is segregating at a very low frequency in some D

lethal activity of wild-type Lhr. Therefore additional regionsof the LHR protein must also contribute to its incompatibilityproperties.

An effect of the 16-aa indel polymorphism on hybridlethal activity was excluded by Nolte et al. (2008) using apopulation survey. They tested two D. simulans lines thatretain the ancestral state of lacking the 16-aa insertion,but neither of them rescued hybrid sons. However, in thetransgenic assay we find a significant difference in hybridlethal activity of the Lhr2 allele with and without the inser-tion (Table 1 cross 2 vs. cross 4). We also detected a signif-icant difference in the Hmr1 interaction assay (Table 2, cross3 vs. cross 5). The lack of any phenotypic effects observed byNolte et al. (2008) is most likely because the effect of the16-aa indel is revealed only in a sensitized background. Inthis transgenic assay the C-terminal mutation in Lhr2 lowersthe lethal activity of Lhr, providing us with the sensitivity toassess the contribution of the 16-aa deletion.

Differential hybrid lethal activity of Lhr orthologs:coding or regulatory?

Lhr has strongly asymmetric effects on hybrid viability, asmutations in sim-Lhr but not mel-Lhr produce viable hybrids(Brideau et al. 2006). This finding led to the hypothesis that

the hybrid lethal activity of Lhr is due to coding sequencedivergence that is specific to the D. simulans lineage. Sur-prisingly, we subsequently found that hybrid lethal activityis an ancestral property shared by the coding sequences ofboth Lhr orthologs (Brideau and Barbash 2011; Maheshwariand Barbash 2012). The different hybrid rescue effects ofLhr orthologs instead appear to be largely the consequenceof divergent gene regulation that causes sim-Lhr to be ex-pressed more highly in hybrids thanmel-Lhr (Maheshwari andBarbash 2012). Our results here are consistent with thesefindings. First, we have identified the site of the C-terminalmutation in Lhr2 as critical for HI. Since this region is nearlyidentical between D. melanogaster and the sibling species, itwas likely present in the ancestral Lhr allele. Second, our pre-vious transgenic comparisons ofmel-Lhr and sim-Lhr alleles didnot exclude the possibility that coding sequence divergencemay make some contribution to functional divergence. Ourfinding here that the 16-aa indel has a functional effect, butis only detectable on the background of the C-terminal de-letion, is indicative of coding sequence divergence making asmall contribution to differences in the hybrid lethal activity ofLhr. Interestingly though, since this difference betweenmel-Lhrand sim-Lhr is an indel it does not contribute to the signatureof adaptive evolution discovered for Lhr (Brideau et al. 2006).

Figure 3 The sim-LHR2 protein interacts with HP1 andlocalizes to heterochromatin. (A) Interaction with HP1.Wild-type D. simulans LHR was used as a positive control.Yeast two-hybrid interactions were detected by activationof HIS3 and growth on media lacking histidine; loadingcontrols [complete media (CM) 2Leu 2Trp] contain histi-dine. (B) Localization of sim-LHR2-HA to heterochromatinin D. melanogaster cycle 12–14 embryos. Top, sim-LHR2-HA was detected with anti-HA (green) and localizes toapical heterochromatin, detected by TOPRO3 staining(red) of DNA at the embryo surface. Bottom, immuno-FISHexperiment with anti-HA (green) detecting sim-LHR2-HAin interphase nuclei. LHR2-HA shows no overlap with the359-bp (red) satellite but partially colocalizes with the dodecasatellite (blue).

Functional Effect of an Indel Polymorphism 689

Page 8: An Indel Polymorphism in the Hybrid Incompatibility Gene ... · A population survey revealed that the ancestral non-insertion form is segregating at a very low frequency in some D

Rigorous identification of incompatible substitutions hasonly been attempted for yeast interstrain and interspecificHI genes. Single-amino-acid changes have been identified ineach of two interacting genes that cause a defect in mismatchrepair (Heck et al. 2006). In the case of AEP2, a translationfactor that causes mitonuclear incompatibility betweenS. cerevisiae and S. bayanus, it was narrowed down to mul-tiple mutations within a region of 148 aa. In the case ofMRS1, a splicing factor that also causes mitonuclear incom-patibility between the same two yeast species, it was pareddown to only three nonsynonymous substitutions (Lee et al.2008; Chou et al. 2010). There is no evidence of selectionacting on either of these latter two HI genes and both haveexperienced relatively limited sequence divergence. Thereare at least six HI genes known that are rapidly divergingunder selection (Presgraves 2010; Maheshwari and Barbash2011). Although it is implicitly assumed that this divergenceis the basis of HI, this hypothesis remains largely unexamined.

Functional effects of indels and polymorphisms

While indels are a common type of sequence variation, theyare rarely considered in evolutionary studies. The reason forthis is that their origins and functional consequences arepoorly understood. Analysis of indels within protein sequen-ces supports the view that they affect protein folding, andcomputational analysis of high-throughput protein interac-tion datasets suggests that indels modify protein interactioninterfaces, thereby significantly rewiring the interaction net-works (Hormozdiari et al. 2009; Zhang et al. 2011). More-over, studies comparing patterns of evolution of Catsper1, asperm-specific calcium channel, found evidence of positiveselection for elevated rates of indel substitutions within itsintracellular domain across multiple primate and rodent spe-cies (Podlaha and Zhang 2003; Podlaha et al. 2005). Theauthors suggest that the selection for indels might be a con-sequence of their effect on the regulation of the Catsper1channel, which can affect sperm motility, an importantdeterminant in sperm competition.

Large structural polymorphisms are not unique to Lhr;other HI genes such as Hmr and Prdm9 have multiple in-frameindels, as does the segregation distorter RanGAP (Presgraves2007; Maheshwari et al. 2008; Oliver et al. 2009). So farthe primary focus of evolutionary analysis has been single-amino-acid substitutions, and indel variation has been largelyignored in the assessment of functional divergence. Recenthigh throughput analyses on human tissues has catalogedthe occurrence of coding indels in hundreds of conservedand essential genes as well as in protein isoforms via alter-native splicing, thus highlighting indels as an abundantsource of structural variation (Wang et al. 2008; Mills et al.2011). Our characterization of an indel polymorphism in Lhrpresents one functional argument supporting the predictionthat coding indels play an important evolutionary role. Thelow frequency of the deletion state of the 16-aa indel inD. simulans and its monomorphic state in D. melanogasterdo not suggest an obvious role for selection in maintaining

it. Our experiments here nevertheless demonstrate that thisindel does affect Lhr function.

Acknowledgments

We thank Greg Smaldone and Shuqing Ji for help scoringflies and Tawny Cuykendall, Heather Flores, P. Satyaki,Michael Nachman, and the anonymous reviewers for helpfulcomments on the manuscript. This work was supported byNational Institutes of Health grant 2R01GM074737.

Literature Cited

Barbash, D. A., 2010 Genetic testing of the hypothesis that hybridmale lethality results from a failure in dosage compensation.Genetics 184: 313–316.

Bayes, J. J., and H. S. Malik, 2009 Altered heterochromatin bind-ing by a hybrid sterility protein in Drosophila sibling species.Science 326: 1538–1541.

Bischof, J., R. K. Maeda, M. Hediger, F. Karch, and K. Basler,2007 An optimized transgenesis system for Drosophila usinggerm-line-specific phiC31 integrases. Proc. Natl. Acad. Sci. USA104: 3312–3317.

Brideau, N. J., and D. A. Barbash, 2011 Functional conservationof the Drosophila hybrid incompatibility gene Lhr. BMC Evol.Biol. 11: 57.

Brideau, N. J., H. A. Flores, J. Wang, S. Maheshwari, X. Wang et al.,2006 Two Dobzhansky-Muller genes interact to cause hybridlethality in Drosophila. Science 314: 1292–1295.

Chou, J.-Y., Y.-S. Hung, K.-H. Lin, H.-Y. Lee, and J.-Y. Leu,2010 Multiple molecular mechanisms cause reproductive iso-lation between three yeast species. PLoS Biol. 8: e1000432.

Greil, F., E. de Wit, H. J. Bussemaker, and B. van Steensel,2007 HP1 controls genomic targeting of four novel hetero-chromatin proteins in Drosophila. EMBO J. 26: 741–751.

Heck, J. A., J. L. Argueso, Z. Gemici, R. G. Reeves, A. Bernard et al.,2006 Negative epistasis between natural variants of the Sac-charomyces cerevisiae MLH1 and PMS1 genes results in a defectin mismatch repair. Proc. Natl. Acad. Sci. USA 103: 3256–3261.

Hormozdiari, F., R. Salari, M. Hsing, A. Schönhuth, S. K. Chanet al., 2009 The effect of insertions and deletions on wiringsin protein-protein interaction networks: a large-scale study.J. Comput. Biol. 16: 159–167.

Hutter, P., and M. Ashburner, 1987 Genetic rescue of inviablehybrids between Drosophila melanogaster and its sibling species.Nature 327: 331–333.

Lachaise, D., J. R. David, F. Lemeunier, L. Tsacas, and M. Ashburner,1986 The reproductive relationships of Drosophila sechellia withD. mauritiana, D. simulans, and D. melanogaster from the Afro-tropical region. Evolution 40: 262–271.

Lee, H.-Y., J.-Y. Chou, L. Cheong, N.-H. Chang, S.-Y. Yang et al.,2008 Incompatibility of nuclear and mitochondrial genomescauses hybrid sterility between two yeast species. Cell 135:1065–1073.

Mackay, T. F. C., S. Richards, E. A. Stone, A. Barbadilla, J. F. Ayroleset al., 2012 The Drosophila melanogaster Genetic ReferencePanel. Nature 482: 173–178.

Maheshwari, S., and D. A. Barbash, 2011 The genetics of hybridincompatibilities. Annu. Rev. Genet. 45: 331–355.

Maheshwari, S., and D. A. Barbash, 2012 Cis-by-Trans regulatorydivergence causes the asymmetric lethal effects of an ancestralhybrid incompatibility gene. PLoS Genet. 8: e1002597.

Maheshwari, S., J. Wang, and D. A. Barbash, 2008 Recurrentpositive selection of the Drosophila hybrid incompatibility geneHmr. Mol. Biol. Evol. 25: 2421–2430.

690 S. Maheshwari and D. A. Barbash

Page 9: An Indel Polymorphism in the Hybrid Incompatibility Gene ... · A population survey revealed that the ancestral non-insertion form is segregating at a very low frequency in some D

McQuilton, P., S. E. St Pierre, and J. Thurmond; and FlyBase Con-sortium, 2012 FlyBase 101–the basics of navigating FlyBase.Nucleic Acids Res. 40: D706–D714.

Mills, R. E., W. S. Pittard, J. M. Mullaney, U. Farooq, T. H. Creasyet al., 2011 Natural genetic variation caused by small inser-tions and deletions in the human genome. Genome Res. 21:830–839.

Nolte, V., D. Weigel, and C. Schlötterer, 2008 The impact ofshared ancestral variation on hybrid male lethality–a 16 co-don indel in the Drosophila simulans Lhr gene. J. Evol. Biol.21: 551–555.

Oliver, P. L., L. Goodstadt, J. J. Bayes, Z. Birtle, K. C. Roach et al.,2009 Accelerated evolution of the Prdm9 speciation geneacross diverse metazoan taxa. PLoS Genet. 5: e1000753.

Podlaha, O., and J. Zhang, 2003 Positive selection on protein-length in the evolution of a primate sperm ion channel. Proc.Natl. Acad. Sci. USA 100: 12241–12246.

Podlaha, O., D. M. Webb, P. K. Tucker, and J. Zhang, 2005 Positiveselection for indel substitutions in the rodent sperm proteincatsper1. Mol. Biol. Evol. 22: 1845–1852.

Presgraves, D. C., 2007 Does genetic conflict drive rapid molecularevolution of nuclear transport genes in Drosophila? Bioessays29: 386–391.

Presgraves, D. C., 2010 The molecular evolutionary basis of spe-cies formation. Nat. Rev. Genet. 11: 175–180.

Sturtevant, A. H., 1920 Genetic studies on Drosophila simulans.I. Introduction. Hybrids with Drosophila melanogaster. Genetics5: 488–500.

Tamura, K., D. Peterson, N. Peterson, G. Stecher, M. Nei et al.,2011 MEGA5: molecular evolutionary genetics analysis usingmaximum likelihood, evolutionary distance, and maximum par-simony methods. Mol. Biol. Evol. 28: 2731–2739.

Ting, C. T., S. C. Tsaur, M. L. Wu, and C. I. Wu, 1998 A rapidlyevolving homeobox at the site of a hybrid sterility gene. Science282: 1501–1504.

Wang, E. T., R. Sandberg, S. Luo, I. Khrebtukova, L. Zhang et al.,2008 Alternative isoform regulation in human tissue transcrip-tomes. Nature 456: 470–476.

Watanabe, T. K., 1979 A gene that rescues the lethal hybridsbetween Drosophila melanogaster and D. simulans. Jpn. J.Genet. 54: 325–331.

Zhang, Z., J. Huang, Z. Wang, L. Wang, and P. Gao, 2011 Impactof indels on the flanking regions in structural domains. Mol.Biol. Evol. 28: 291–301.

Communicating editor: M. W. Nachman

Functional Effect of an Indel Polymorphism 691

Page 10: An Indel Polymorphism in the Hybrid Incompatibility Gene ... · A population survey revealed that the ancestral non-insertion form is segregating at a very low frequency in some D

GENETICSSupporting Information

http://www.genetics.org/lookup/suppl/doi:10.1534/genetics.112.141952/-/DC1

An Indel Polymorphism in the HybridIncompatibility Gene Lethal Hybrid Rescue

of Drosophila Is Functionally RelevantShamoni Maheshwari and Daniel A. Barbash

Copyright © 2012 by the Genetics Society of AmericaDOI: 10.1534/genetics.112.141952

Page 11: An Indel Polymorphism in the Hybrid Incompatibility Gene ... · A population survey revealed that the ancestral non-insertion form is segregating at a very low frequency in some D

S.  Maheshwari  and  D.  A.  Barbash  

 

2  SI  

 

 

 

                                                                                         Figure  S1      Alignment  of  D.  simulans  Lhr2  protein  sequence  with  wild  type  orthologs.  The  16  aa  indel  polymorphism  and  C-­‐terminal  mutations  are  underlined.      

Page 12: An Indel Polymorphism in the Hybrid Incompatibility Gene ... · A population survey revealed that the ancestral non-insertion form is segregating at a very low frequency in some D

S.  Maheshwari  and  D.  A.  Barbash  

 

3  SI  

Table  S1      Primers  used  in  the  Materials  and  Methods.    

No.   Sequence   Capitalized  region  

405     caccatgagtaccgacagcgccgaggaa    

405   tcatgttctcagcgtaggccg    

409   gtagctttctcttggcgctctt    

410   gtaagtgaactgaagctgcgttgg    

664   tcgcatAAGCTTctggcaggtggtaaccgatacgg   HindIII  

691   tactatAAGCTTtggttgttccacacgactttatcg   HindIII  

728  TGCATAGTCCGGGACGTCATAGGGATAGCCCGCATAGTCAGGAACATCGTATGGGTA

CATtgttctcagcgtaggccg  3xHA  tag  

729  CCCTATGACGTCCCGGACTATGCAGGATCCTATCCATATGACGTTCCAGATTACGCTtg

actttctttcgtataaaatgc  3xHA  tag  

882   tgtcgcccgcggaacgtcgcc    

938   cgtttcctcggcgctgtcggtactcat    

939   atgagtaccgacagcgccgaggaaacg    

940   tcatgttctcagcgtaggccgcctgg    

941   ccaggcggcctacgctgagaacatga    

942   ccaTTATAGCTTATTCTTTTATTGGCACTTGctacgttgggtcttatgttgcg   Cter  fill-­‐in  

943   CAAGTGCCAATAAAAGAATAAGCTATAAtggtgttagcaatgaatcaaatgatgtc   Cter  fill-­‐in  

945  GATTTGCAATTTGTGTACATCGTTCATCTCCCGCCACAGAGGTTCAGTgatttgccctttgg

cagccgc  16aa  fill-­‐in  

946  ACTGAACCTCTGTGGCGGGAGATGAACGATGTACACAAATTGCAAATCcctgaacctctg

tttcgggtg  16aa  fill-­‐in  

 

 

Page 13: An Indel Polymorphism in the Hybrid Incompatibility Gene ... · A population survey revealed that the ancestral non-insertion form is segregating at a very low frequency in some D

GENETICSSupporting Information

http://www.genetics.org/lookup/suppl/doi:10.1534/genetics.112.141952/-/DC1

An Indel Polymorphism in the HybridIncompatibility Gene Lethal Hybrid Rescue

of Drosophila Is Functionally RelevantShamoni Maheshwari and Daniel A. Barbash

Copyright © 2012 by the Genetics Society of AmericaDOI: 10.1534/genetics.112.141952

Page 14: An Indel Polymorphism in the Hybrid Incompatibility Gene ... · A population survey revealed that the ancestral non-insertion form is segregating at a very low frequency in some D

S.  Maheshwari  and  D.  A.  Barbash  

 

2  SI  

 

 

 

                                                                                         Figure  S1      Alignment  of  D.  simulans  Lhr2  protein  sequence  with  wild  type  orthologs.  The  16  aa  indel  polymorphism  and  C-­‐terminal  mutations  are  underlined.      

Page 15: An Indel Polymorphism in the Hybrid Incompatibility Gene ... · A population survey revealed that the ancestral non-insertion form is segregating at a very low frequency in some D

S.  Maheshwari  and  D.  A.  Barbash  

 

3  SI  

Table  S1      Primers  used  in  the  Materials  and  Methods.    

No.   Sequence   Capitalized  region  

405     caccatgagtaccgacagcgccgaggaa    

405   tcatgttctcagcgtaggccg    

409   gtagctttctcttggcgctctt    

410   gtaagtgaactgaagctgcgttgg    

664   tcgcatAAGCTTctggcaggtggtaaccgatacgg   HindIII  

691   tactatAAGCTTtggttgttccacacgactttatcg   HindIII  

728  TGCATAGTCCGGGACGTCATAGGGATAGCCCGCATAGTCAGGAACATCGTATGGGTA

CATtgttctcagcgtaggccg  3xHA  tag  

729  CCCTATGACGTCCCGGACTATGCAGGATCCTATCCATATGACGTTCCAGATTACGCTtg

actttctttcgtataaaatgc  3xHA  tag  

882   tgtcgcccgcggaacgtcgcc    

938   cgtttcctcggcgctgtcggtactcat    

939   atgagtaccgacagcgccgaggaaacg    

940   tcatgttctcagcgtaggccgcctgg    

941   ccaggcggcctacgctgagaacatga    

942   ccaTTATAGCTTATTCTTTTATTGGCACTTGctacgttgggtcttatgttgcg   Cter  fill-­‐in  

943   CAAGTGCCAATAAAAGAATAAGCTATAAtggtgttagcaatgaatcaaatgatgtc   Cter  fill-­‐in  

945  GATTTGCAATTTGTGTACATCGTTCATCTCCCGCCACAGAGGTTCAGTgatttgccctttgg

cagccgc  16aa  fill-­‐in  

946  ACTGAACCTCTGTGGCGGGAGATGAACGATGTACACAAATTGCAAATCcctgaacctctg

tttcgggtg  16aa  fill-­‐in