The Hmr and Lhr hybrid incompatibility genes suppress a broad range of heterochromatic repeats

Abstract

Hybrid incompatibilities (HIs) cause reproductive isolation between species and thus contribute to speciation. Several HI genes encode adaptively evolving proteins that localize to or interact with heterochromatin, suggesting that HIs may result from co-evolution with rapidly evolving heterochromatic DNA. Little is known, however, about the intraspecific function of these HI genes, the specific sequences they interact with, or the evolutionary forces that drive their divergence. The genes Hmr and Lhr genetically interact to cause hybrid lethality between Drosophila melanogaster and D. simulans, yet mutations in both genes are viable. Here, we report that Hmr and Lhr encode proteins that form a heterochromatic complex with Heterochromatin Protein 1 (HP1a). Using RNA-Seq analyses we discovered that Hmr and Lhr are required to repress transcripts from satellite DNAs and many families of transposable elements (TEs). By comparing Hmr and Lhr function between D. melanogaster and D. simulans we identify several satellite DNAs and TEs that are differentially regulated between the species. Hmr and Lhr mutations also cause massive overexpression of telomeric TEs and significant telomere lengthening. Hmr and Lhr therefore regulate three types of heterochromatic sequences that are responsible for the significant differences in genome size and structure between D. melanogaster and D. simulans and have high potential to cause genetic conflicts with host fitness. We further find that many TEs are overexpressed in hybrids but that those specifically mis-expressed in lethal hybrids do not closely correlate with Hmr function. Our results therefore argue that adaptive divergence of heterochromatin proteins in response to repetitive DNAs is an important underlying force driving the evolution of hybrid incompatibility genes, but that hybrid lethality likely results from novel epistatic genetic interactions that are distinct to the hybrid background.

Figure 1. Hmr forms a complex with Lhr and HP1a and is required to stabilize Lhr.

(A) mel-Hmr-HA (green) colocalizes with HP1a (top) and H3k9me2 (middle; both red) in nuclear cycle 14 embryos. The HP1a costain is in a mel-Hmr-HA background, while the H3k9me2 costain is in a Hmr³; mel-Hmr-HA background. A negative control shows no HA signal in w¹¹¹⁸ embryos lacking the mel-Hmr-HA transgene (bottom). Scale bars represent 10 µm. (B) mel-Hmr-HA (green) colocalizes with 2L3L, dodeca and GA-rich satellites but not with the 359 bp repeat satellite in mel-Hmr-HA (all FISH probes red). Scale bars represent 5 µm. (C) mel-Hmr-HA (red) colocalizes with the nucleolar marker Fibrillarin (green) in mel-Hmr-HA early embryos. Scale bars represent 10 µm. (D) mel-Lhr-HA and mel-Hmr-FLAG co-immunoprecipitate from D. melanogaster embryo extracts derived from flies expressing both transgenes (left 4 lanes) but not from flies expressing only Lhr-HA (right 4 lanes). Extracts were IP'd with the indicated antibodies, and then probed by Western Blots (WB) with the same or different antibodies. (E) Lhr-HA, Hmr-FLAG and HP1a co-immunoprecipitation from embryo extracts. Specificity is indicated by lack of immunoprecipitation of histone H3. Asterisk indicates the antibody light chain. (F) Lhr and Hmr interact in a yeast-two hybrid assay. Interactions were detected by growth on complete media (CM) lacking histidine (his); growth controls were performed on CM lacking tryptophan (trp) and leucine (leu). The top 4 panels test for interactions between orthologs from the same species; the bottom 4 between heterospecific orthologs. AD, activation domain; BD, DNA binding domain. (G) Lhr-HA is detectable in Hmr³ and localizes to heterochromatin, as indicated by co-localization with HP1a. Note that a higher gain was used in the Hmr³ panels compared to the Hmr⁺ panels in order to detect Lhr-HA, and is reflected in the higher background. Western blots confirm that Lhr-HA levels are reduced in Hmr³. HP1a is used as a loading control. (H) Hmr-HA maintains its localization to DAPI-dense heterochromatin in Lhr^KO; Hmr-HA embryos. Scale bars represent 10 µm.

Figure 2. Lhr mutant females have reduced fertility.

Total adult progeny from single Lhr^KO/Lhr^KO (A) or Lhr^KO/Df(2R)BSC44, Lhr⁻ (B) females were compared at 27° to heterozygous female siblings (Lhr^KO/+ for (A); Lhr^KO/SM6a for (B)). The difference between the fertility of genotypes was tested by a two-tailed t-test. n.s = not significant, **p<0.01,***p<0.001. The number of individuals tested for each experiment is shown at the bottom of the bars. The error bars represent S.E.M. Crosses were performed at 27°.

Figure 3. TE misregulation in Lhr and Hmr mutants.

(A and B) Analysis of Lhr^KO (A) and Hmr⁻ (B) ovaries. Reads with zero mismatches were mapped separately to the individual-insertion or consensus-sequence TE databases. A subset of TEs that are significantly different between genotypes are shown and include those with the 25 lowest p-values obtained from individual-insertion mapping analysis, but excluding all centroid repeats . Additionally shown are TAHRE, which is only found in the consensus-sequence database, as well as TARTB1 for Lhr^KO, which is significant but not among the 25 top hits in the Lhr^KO individual-insertion analysis. (C) 49 TEs are upregulated at least 2 fold in both Lhr^KO and Hmr⁻. TE families include those resulting from mapping reads to the insertion database, as well as families found only when reads were mapped to the consensus database. (D) Reads from Hmr mutant or wildtype male larvae with up to three mismatches were mapped to the individual-insertion or consensus-sequence TE databases. All TE families, excluding centroids, that were significantly upregulated in the insertion sequence based analysis are shown here. Note the different Y-axis scales in A, B and D. Classification of DNA, LTR and non-LTR elements is from reference .

Figure 4. Reduced expression of heterochromatic genes in Lhr and Hmr mutants.

Loss of Lhr (A) and Hmr (B) leads to a statistically significant reduction in the expression of heterochromatic genes. Significance of difference was calculated using the Wilcoxon rank sum test with continuity correction (for (A) p = 3.549e-05, for (B) p = 1.461e-09). Box plots show log₂ fold change of 7838 euchromatic and 370 heterochromatic genes for (A) and 7451 euchromatic and 344 heterochromatic genes for (B). The definition of the euchromatin-heterochromatin boundary for all chromosomes comes from experiments done in S2 tissue culture cells, except for 3R, which comes from the cytogenomic border .

Figure 5. Lhr and Hmr are telomere cap proteins required for regulating telomere length.

Lhr-HA (A) and Hmr-HA (B) localize to telomeres. Co-immunostaining with anti-HA and anti-HP1a shows that both proteins colocalize at the cap (arrowheads). The merged images include DAPI to stain DNA, shown in blue. Lhr^KO (C) and Hmr³ (D) have increased HeT-A copy number. qPCR was used to estimate the abundance of HeT-A and rp49 from Lhr^KO, Lhr⁺, y w Hmr³, a matched y w Hmr⁺ control, and the wild-type Canton S strain. Genomic DNA was isolated from carcasses of females whose ovaries were removed in order to minimize the amount of polytenized DNA present. Relative Het-A copy number is the ratio of Het-A/rp49. The error bars represent S.E.M for three replicates. The significance of the differences between the genotypes was calculated using two tailed t-test; * = p<0.05; ** = p<0.01. Scale bars = 5 µm.

Figure 6. Lhr and Hmr repress satellite DNA transcription.

(A) Fold increase in satellite transcripts of Lhr^KO versus Lhr⁺. Numbers within the bars show normalized reads mapping to each satellite, the numerator from Lhr^KO and the denominator from Lhr⁺. All differences have p<0.01 by F.E.T. test. (B) Lhr-HA (green) colocalizes with GA-rich and AACAC satellites (red) in ovarian nurse cell nuclei (arrowheads). DAPI is shown in the merged images in blue. Scale bar = 10 µm. (C) Fold increase in satellite transcripts in Hmr⁻ versus Hmr^+/−. Numbers within the bars show normalized reads mapping to each satellite, the numerator from Hmr⁻ and the denominator from Hmr^+/−. All differences have p<0.001 by F.E.T. test.

Figure 7. Small RNA patterns are largely unaffected in Lhr^KO.

(A) VASA (green) marks the peri-nuclear nuage (white arrowheads) and shows no difference in localization between Lhr⁺ and Lhr^KO ovaries. (B) siRNA (17–22 nt) without mismatches and piRNA (23–30 nt) with up to one mismatch were mapped to a reference sequence set containing the D. melanogaster r5.68 genome, D. melanogaster sequences from Repbase and the repeat-sequence database. The number of mapped Lhr^KO reads was normalized to the total number of mapped Lhr⁺ reads. (C) Filtered piRNA reads were mapped uniquely to the Repbase TE consensus sequences with one allowed mismatch. 121 TE families producing > = 1000 reads summed over both genotypes are shown. Black circles represent TE families whose fold change between Lhr^KO and Lhr⁺ is greater than 2 fold (p<0.001). (D) Ping-pong scores of TE families in Lhr^KO and Lhr⁺. Black circles represent TE families whose fold change in ping-pong score between Lhr^KO and Lhr⁺ is greater than 2 fold (Table S10). (E) Plot shows the number of unique piRNAs mapped to piRNA clusters, with one allowed mismatch and normalized between genotypes. piRNA clusters with > = 500 reads summed over both genotype are shown. Black arrowheads point to sub-telomeric piRNA clusters. Black circles indicate clusters whose fold change between Lhr^KO and Lhr⁺ is greater than 2-fold (p<0.001). (F) Unique siRNA (17–22 nt) were mapped as in (C), except no mismatches were allowed. 96 TE families are plotted that have > = 1000 reads summed over both genotypes. Black circles represent TEs whose siRNA levels changed by >2 fold. siRNA mapping to the TAS repeat HETRP are almost completely lost (arrow). For (C, D, F) significance values were calculated using F.E.T., implemented in DEG-seq.

Figure 8. Analysis of Lhr function in D. simulans.

(A) Immuno-FISH experiment shows that the brightest mel-Lhr foci colocalize with dodeca (red, arrow) and GA satellites (white, arrowhead) in D. melanogaster (upper panel). The brightest sim-Lhr foci either colocalize or are juxtaposed with dodeca (arrow) but are not associated with GA-rich satellites (arrowhead). All panels contain DAPI shown in blue. Scale bar = 10 µm. (B) Fold changes in TE expression between w⁵⁰¹; Lhr¹ and w⁵⁰¹; Lhr⁺ were calculated for uniquely mapping reads with zero mismatches to the individual-insertion database and with three mismatches to the consensus-sequence database. Three mismatches are required to account for the divergence of TE insertions in D. simulans from the consensus sequences, which are largely defined from D. melanogaster TEs. The 25 most significantly derepressed TE families in the individual-insertion sequence based analysis are shown here (excluding centroids), as well as TAHRE, which is found only in the consensus-sequence database. Classification of DNA, LTR and non-LTR elements is from reference . (C) Comparison of TE misregulation between D. melanogaster and D. simulans Lhr mutations. The diagram includes all TE families that were upregulated at least two fold, including those in individual-insertion database analysis as well as those that are only represented in the consensus-sequence database analysis. (D) Comparison of euchromatic and heterochromatic gene expression in D. simulans w⁵⁰¹; Lhr¹, as described in Figure 4. The euchromatin-heterochromatin border has not been experimentally determined in D. simulans and was defined from D. melanogaster, Analysis includes 7479 euchromatic and 350 heterochromatic genes (p = 0.12, Wilcoxon rank sum test with continuity correction).

Figure 9. Hmr orthologs have diverged in their effects on a small subset of TEs.

(A) sim-Hmr-HA colocalizes with HP1a (red) in nuclear cycle 14 D. melanogaster Hmr³; sim-Hmr-HA embryos. The sim-Hmr-HA transgene was transformed into D. melanogaster at the identical attP2 site used for mel-Hmr-HA above (Figure 1). DAPI is shown in blue. (B) mel-Hmr-FLAG does not fully complement TE derepression in Hmr⁻. 9 TE families are 2–9× more highly expressed in Hmr⁻; ø{mel-Hmr-FLAG}/+ compared to Hmr^+/−. (C) Comparison of TE expression in Hmr⁻; ø{mel-Hmr-FLAG}/+ and Hmr⁻; ø{sim-Hmr-FLAG}/+. For B and C, reads were mapped to the individual-insertion database. TEs are considered differentially expressed in the pairwise comparisons if there was at least a 2× fold change and p<0.001.

Figure 10. TE misregulation in hybrid males.

(A) Fold change of TEs up- or downregulated ≥2-fold in Hmr⁺ hybrid male larvae relative to both D. melanogaster and D. simulans male larvae. Uncharacterized centroids are not shown. (B) Fold change of TEs with significantly higher expression in lethal Hmr⁺ versus viable Hmr⁻ hybrid male larvae. “H” indicates TEs that are significantly upregulated in Hmr⁻ D. melanogaster male larvae compared to Hmr⁺ D. melanogaster male larvae from Figure 3D. Note the different Y axis scales between panels A and B. TE families include those resulting from mapping reads to the individual-insertion database, as well as families found only when reads were mapped to the consensus-sequence database. Reads unique to each TE class were mapped allowing for up to 3 mismatches.