A functional requirement for sex-determination M/m locus region lncRNA genes in Aedes aegypti female larvae

Latest Research

A screen identifies siRNAs that induce female-specific larval mortality

Recent high-throughput screens in which first instar (L1) larvae were briefly soaked in siRNAs led to the discovery of hundreds of protein-coding larval lethal genes and a new class of RNAi-based mosquito insecticides21,22,23. Given that the M/m locus is believed to be tightly linked to developmental genes that confer sex-specific lethal effects in A. aegypti8,9, lncRNA loci located both within, as well as flanking the M/m locus, were evaluated in a female larval lethal soaking screen that employed a similar strategy. These studies permitted functional assessment of the hypothesis that silencing A. aegypti M/m locus region lncRNA genes during larval development would induce female-specific lethality. A total of 50 siRNAs corresponding to lncRNA genes in and flanking the M/m locus, which is referred to herein as the M/m locus region (Supplementary Fig. S1), were screened (Supplementary Tables S1, S2, S3, S4).

The soaking screen uncovered a total of 19 siRNAs (Supplementary Tables S1 and S2) corresponding to M/m locus region lncRNAs that induced significant female-specific mortality and had no significant impact on male survival (Fig. 1). These siRNAs corresponded to target sites in 25 M/m locus region lncRNA genes, the identification numbers and chromosomal locations of which are provided in Supplementary Fig. S1 and Supplementary Table S3. Some of the siRNAs corresponded to target sites in singular M/m locus region lncRNA genes (Supplementary Table S1). Due to the highly repetitive nature of DNA located in this pericentric region10, several of the female-specific larvicidal siRNAs identified in the screen corresponded to target sites identically conserved in multiple lncRNA genes, at least one of which resides in the M/m locus region (Supplementary Table S2).

Figure 1
figure1

siRNAs that induce significant female-specific lethality. Significant female-specific larval mortality resulted from soaking treatments with the indicated siRNAs (* = p < 0.05, ** = p < 0.01, and *** = p < 0.001, Chi-square). No significant differences (p > 0.05, Chi-square) were observed in male survival following treatments with these siRNAs, and a control siRNA had no significant impact (p > 0.05, Chi-square) on survival of males or females. Data are represented as mean survival based on adult emergence, (n = 40 larvae/treatment), and error bars denote SEM.

The lncRNA genes identified in the female-specific lethal screen were located throughout the M/m locus region (Supplementary Table S3). A majority (17 of 25) of the lncRNA genes are intergenic, though several (8 of 25) are intragenic (Supplementary Table S3). None of the lncRNA genes identified in the screen have known orthologs reported in Vectorbase10, potentially because few lncRNA genes have been annotated in other mosquito species. Alternatively, a recent comparative analysis of the genomes of several Anopheles species revealed that female-biased protein-coding genes evolve more rapidly in sequence, expression, and genic turnover than male-biased protein-coding genes; this is an atypical pattern that is proposed to have resulted from sex-specific life history challenges, such as blood feeding, that are encountered by female mosquitoes24, and which could also apply to lncRNA genes. Several of the genes were located within the M locus in a region that was not thought to be found in female (genotype m/m) mosquitoes (Supplementary Fig. S1). For example, a perfect match for the siRNA 470 target sequence is only known to reside in AAEL026346, which lies between the two male-specific M locus genes myo-sex (AaegL5_1: 151,955,864–152,241,832) and Nix (AaegL5_1:152,616,641–152,718,167). Although siRNA 470 and the other siRNAs identified in the screen are not known to have identical target sites in mature transcripts that correspond to genes other than those noted in Supplementary Tables S1, S2, and S3, it is possible that the female-specific phenotypes observed could result, at least in part, from off-site targeting. Moreover, it is also possible that these siRNAs specifically target genes located in the known gap in the sequence at the sex determination locus, a region which has not yet been successfully sequenced but is believed to be highly repetitive7.

Finally, 31 of the siRNAs screened had identically conserved target sites in M/m locus region lncRNA genes, but had no significant impact on female or male larval survival (Supplementary Table S4). These genes (Supplementary Table S4) may lack sex-specific functions or could be active during different stages of the life cycle. It is also possible that siRNAs targeting different sites in these same genes could produce more effective silencing and yield female-specific killing. However, given the overall success of the screen, which had already identified multiple female-specific larvicides (Fig. 1), evaluation of additional siRNAs and further characterization of these particular lncRNA genes (Supplementary Table S4) were not pursued at this time.

Generation of a female-specific yeast larvicide that targets M/m locus region lncRNA genes

In recent years, S. cerevisiae has been developed as a system for inexpensive and scalable manufacture of larvicidal interfering RNAs23. The yeast can also be used as a delivery system for RNAi larvicides to mosquitoes, in which effective gene silencing is observed in larvae that consume the larvicides in the form of inactivated yeast tablets21,22,23. Yeast RNAi larvicide technology, which could also potentially facilitate scaled culturing and sex separation of male mosquitoes, was therefore used for further characterization of several lncRNAs identified in the screen. siRNA 478, which induced significant levels of female mortality, but which did not have a significant impact on male survival (Fig. 1), was down-selected for these studies. A yeast strain designed to express a short hairpin RNA (shRNA) corresponding to the siRNA 478 target site was generated. shRNA expression was confirmed in this strain, as well as a control interfering RNA strain, through PCR amplification of cDNA corresponding to the 3’ end of each hairpin and the terminator sequence that had been prepared from total RNA extractions from each strain (Supplementary Fig. S2). Dried inactivated yeast prepared from each of these strains was fed to larvae throughout larval development. Treatment with the yeast larvicide, but not control interfering RNA yeast, resulted in significantly higher male:female ratios among the surviving mosquitoes (Fig. 2a; p < 0.001). Yeast larvicide #478 was therefore characterized in further detail.

Figure 2
figure2

A female-specific lethal yeast interfering RNA larvicide targeting lncRNA genes. Significant female larval mortality resulted from oral consumption of yeast interfering RNA larvicide strain #478 [(a), *** = p < 0.001, Chi-square]. No significant death was observed in males following treatments with larvicide [(a); p > 0.05, Chi-square], and a control interfering RNA strain did not significantly impact survival of male or female larvae [(a) p > 0.05, Chi-square; data are represented as mean survival based on adult emergence following treatment of 180 total larvae, and error bars denote SEM]. (bd) display quantification of the identical AAEL020379, AAEL020813, and AAEL022952 transcripts, with mRNA levels normalized to levels of the rpS17 housekeeping gene; error bars denote standard deviations. The #478 larvicide target transcripts are expressed throughout larval development in untreated first (L1), second (L2), third (L3), and fourth (L4) instar larvae (b), with no significant differences in expression levels noted between the various larval stages (p > 0.05, ANOVA; the expression levels are shown relative to L4). No significant differences in transcript levels were noted between third instar male and female larvae [(c), t-test, p > 0.05]. Silencing of these lncRNA targets of larvicide #478 (d) was confirmed through qRT-PCR (*** = p < 0.001 vs. target gene levels in control interfering RNA-fed larvae, Student’s t-test; error bars denote standard deviation).

When larvae were reared on yeast larvicide #478, although no significant impact on male survival was observed, only 10 ± 2% of expected adult females emerged (p < 0.001), with the bulk of these animals dying as fourth instar larvae (Supplementary Fig. S2b). Yeast larvicide #478 targets three M/m locus region loci: AAEL020379, AAEL020813, and AAEL022952 (Supplementary Table S2). The sequences of these genes (both exons and an intron) are identical7,10 and correspond to a single transcript that is detected throughout larval development (Fig. 2b), is expressed at comparable levels in male and female larvae just prior to the time of death (Fig. 2c, p > 0.05), and which is silenced though treatment with yeast larvicide #478 (Fig. 2d; 71.1 ± 7.9% reduction in transcript levels with respect to larvae reared on control interfering RNA yeast, p < 0.001).

Scaled production of adult male mosquitoes

Male mosquitoes released en masse for control strategies such as the incompatible insect technique (IIT) and sterile insect technique (SIT) must successfully compete with wild-type males in areas in which they are mass-released2,25,26,27. It is therefore critical that yeast larvicides used for sex separation are specific to females and do not have undesired impacts on adult males. To examine if the impact of yeast larvicide #478 is specific to female larvae, life history traits were assessed in adult male mosquitoes that had been reared on the larvicide during larval development. Treatment with yeast larvicide #478 did not significantly impact the capacity of males to mate (Fig. 3a). The number of eggs laid (fertility) by wild-type females that mated with males treated with the larvicide, as well as the percentage of larvae that hatched from these eggs (fecundity) did not significantly differ from control male matings (Fig. 3b).

Figure 3
figure3

Yeast interfering RNA larvicide technology can be used for scaled production of males. Yeast larvicide #478 does not significantly impact male mating capacity [(a), p > 0.05, Student’s t-test], the number of eggs laid by females that mated with these males [(b), p > 0.05, Student’s t-test], or the percentage of larvae that hatched from these eggs [(b), p > 0.05, Student’s t-test]; results were compiled from 41 matings with #478-treated males and 72 matings with males treated with control interfering RNA yeast). Incorporation of the yeast larvicide into a larval diet used for mass-rearing (MR; n = 1200 total larvae per treatment) resulted in significant female mortality [(c), *** = p < 0.001, Chi-square] with no significant impact on male survival [(c), p > 0.05, Chi-square] or fitness [(d), p > 0.05, Student’s t-test; n = 83 control diet male wings, n = 40 #478-treated male wings]. Error bars denote SEM in all panels.

Mass rearing facilities utilize special larval diets that are optimized to produce fit male mosquitoes28. It is therefore helpful if yeast larvicides are compatible with these diets. Dried inactivated nutritional yeast is often a component of such diets28, suggesting that the nutritional yeast component could be replaced with female-specific yeast larvicides. To assess whether use of the larvicides would facilitate scaled production of males, a larval diet employed at mass-rearing mosquito facilities28 was modified by replacing the nutritional yeast component of the diet with dried inactivated yeast larvicide #478. The modified diet was tested on mosquitoes grown in mass-rearing trays containing 200 larvae/L of water. With respect to the control interfering RNA diet, larvicide #478 induced significant female mortality, resulting in 5 male:1 female ratios in emerging adults (Fig. 3c). The fitness of male survivors, which was ascertained through measurements of wing lengths, a proxy for body size and fitness, was not significantly different than males raised on the standard mass-rearing diet (Fig. 3d), providing further evidence that the larvicide is lethal to females, but does not impact male mosquitoes.

Although the yeast larvicides characterized here do not eliminate all females and could not be used in a stand-alone capacity, replacing nutritional yeast with the larvicidal yeast could further improve the efficacy of existing sex separation technologies2 or immensely reduce labor associated with hand separation strategies. Yeast interfering RNA technology, which could be implemented in remote and resource-limited locations, would likely benefit mass-rearing facilities worldwide. Moreover, the use of yeast interfering RNA larvicides would circumvent a need to further genetically manipulate existing mosquito strains that have already been developed for population control strategies, for which regulatory permits may have already been attained or might need to be acquired.

Conclusions and potential implications for understanding the evolution of sex chromosomes in A. aegypti

In summary, these studies functionally verified a female larval requirement for multiple lncRNA genes located at the M/m locus region (Figs. 1, 2, Supplementary Fig. S1). In multiple instances, silencing lncRNA genes resulted in significantly increased male:female ratios that resulted from female lethality, without any significant impact on male survival or fitness (Figs. 1, 2, 3). The complete phased structure of the male M locus and the female m locus have not yet been determined, and a ~ 163 kb gap in the sequence remains7. Completion of the entire phased sequence will undoubtedly facilitate further interpretation and a more sophisticated understanding of these lncRNA screen data. Nevertheless, as predicted by Matthews et al.7, the availability of the existing M locus assembly has provided the opportunity to study the evolution of A. aegypti homomorphic sex chromosomes. These initial lncRNA studies have elucidated key findings that may help shape our understanding of sex chromosome evolution.

The evolution of sex chromosomes is believed to occur in several stages29,30,31. Initially, a homologous pair of autosomes acquires sex-determining loci, forming a proto-Y chromosome bearing a male fertility locus (M) and a dominant female suppressor (SuF), as well as a proto-X chromosome carrying a female fertility locus (F) and a male sterility locus (m). Suppressed recombination in the sex-determining region evolves and eventually spreads over a larger portion of the proto-sex chromosomes. The A. aegypti homomorphic sex chromosomes appear to have evolved into proto-sex chromosomes bearing a sex determining M/m region3,4 which contains a male-determining factor, Nix5, that is present on the proto-Y chromosome. Nix regulates male-specific splicing of another chromosome 1 gene, doublesex (dsx), permitting expression of the male-specific splice form of dsx32 rather than the female splice form which is important for ovary development and fertility33. A sex-differentiated region of suppressed recombination has also evolved and is believed to have extended ~ 100 Mb beyond the M/m locus7,34,35.

The suppression of recombination on sex chromosomes permits accumulation of transposable elements and other non-coding sequences, as well as chromosomal rearrangements and the acquisition of sexually antagonistic genes with different alleles that differentially benefit either males or females29,30,31. Further loss of recombination between these genes and the sex-determination locus is expected to follow, eventually resulting in evolution of heteromorphic X and Y chromosomes31. Highly repetitive DNA, which comprises > 70% of the M locus and includes long terminal repeat retrotransposons7, has accumulated in A. aegypti. This investigation has revealed that functional lncRNA genes that are required in female larvae are located in this region. Given that retrotransposons can contribute to both the origin and diversification of lncRNAs36, one could speculate that accumulation of retrotransposons in A. aegypti has also contributed to the origin and diversification of M/m locus region lncRNA genes that evolved female-specific functions. It is predicted that these genes may eventually contribute to the formation of heteromorphic A. aegypti sex chromosomes and lead to genetic degeneration and reduced size of the Y chromosome29,30,31.

lncRNAs regulate a wide array of cellular activities that could contribute to sex-specific gene expression during sexually dimorphic development or differentiation, including the regulation of chromatin modifiers12. Although A. aegypti is not yet believed to possess dosage compensation, recent studies suggest that the region of non-recombination between M and m chromosomes is more extensive than previously believed7,34,35, suggesting that the evolution of such dosage compensation mechanisms could eventually initiate in A. aegypti. Interestingly, centromeric repeats in Saccharmocyes pombe produce dsRNA that targets formation and maintenance of heterochromatin through RNA interference (RNAi)37, which occurs through sequence-specific targeting of histone modifications regulated by small RNA silencing38. Woolcock et al.39 demonstrated that RNAi proteins interact with ncRNAs and retrotransposon long terminal repeats. The authors39 speculate that similar mechanisms could operate in other eukaryotes. Future studies will consider if lncRNAs regulate heterochromatin at the A. aegypti sex determination locus and elucidate the sex-specific molecular functions of lncRNAs in A. aegypti and other species of mosquitoes. Yeast interfering RNA technology, which may benefit efforts to mass produce male mosquitoes for emerging mosquito control programs, will likewise enhance future laboratory studies aimed at dissecting the molecular functions of mosquito lncRNAs during sex-specific development and differentiation.

Products You May Like

Leave a Reply

Your email address will not be published.