Introduction

RNA-binding proteins (RBPs) are the proteins that bind RNAs and regulate their life cycle starting from synthesis to decay. RNAs after transcription are associated with RBPs. RBPs and other cis-regulatory elements regulate RNA processing, transportation, localization, stability, modifications, translation, and degradation (Glisovic et al. 2008). Due to their role in several aspects of RNA biology, Diaz-Munoz and Turner in 2018 have aptly described RBPs as “readers, writers, editors, and erasers of the transcriptome” where “RNA writers” include the proteins involved in splicing, capping, and polyadenylation, “readers” are the RBPs involved in subcellular localization and translation, “editors” constitute methyltransferases and deaminases, and “erasers” include the destabilizing factors and nucleases that cause RNA instability and degradation (Díaz-Muñoz and Turner 2018). The regulation of all these aspects after transcription is known as post-transcriptional gene regulation (PTGR), and RBPs play a major role in PTGR.

RBPs make up a significant part of the proteome in eukaryotes. They regulate cell differentiation and homeostasis during developmental stages, environmental challenges, and diseases/infections (Wurth 2012; Brinegar and Cooper 2016; Li et al. 2020). Several RBPs have been associated with different types of cancers and genetic disorders (Wurth 2012; Li et al. 2020). RBPs also play a role in the growth and development of plants (Muleya and Marondedze 2020), animals (Kerner et al. 2011), bacteria (Holmqvist and Vogel 2018), and insects (Norvell et al. 1999; Yoon et al. 2018). Interestingly, many RBPs have been found to regulate early embryo development in Drosophila. For example, mutations in the Squid gene disrupted Gurken-dependent dorsoventral patterning during oogenesis in Drosophila resulting in female sterility (Norvell et al. 1999; Gamberi et al. 2006). Smaug, another RBP, is involved in maternal to zygotic transition (Benoit et al. 2009). Localization and translational regulation of maternal transcripts during oogenesis in Drosophila rely on RBPs (Bansal et al. 2020). Staufen, a double-stranded RBP, has been identified as a key player in RNAi in certain Coleopteran insects (Yoon et al. 2018). However, RBPs in mosquitoes are not much explored except in reports where mosquito RBPs were found to control the replication of flaviviruses (Diosa-Toro et al. 2020; Yeh et al. 2022). RNA viruses use host RBPs for their replication and survival, and such RBPs can act as antivirals. For example, AeStaufen reduced genomic and sub-genomic flaviviral RNA copies in several mosquito tissues, including the salivary glands indicating its role in dengue transmission (Yeh et al. 2022). Taschuk et al. (2020) described the antiviral role of DDX56, Dead box helicase in Drosophila and human cell lines (Taschuk et al. 2020). Thus, the identification and characterization of mosquito RBPome would be helpful for such studies.

RBPs are generally identified by the presence of different RNA-binding structural motifs such as RNA recognition motif (RRM), the zinc fingers, K homology (KH), and others. These motifs locate and interact with their specific RNA targets. Among these motifs, RRM is one of the most abundant and well-characterized RNA-binding motifs in eukaryotes (Venter et al. 2001). RRM is found in all forms of life, including bacteria and viruses. RRM-containing proteins are involved in several post-transcriptional aspects like alternative splicing, translation, degradation, and processing of precursor ribosomal RNA (pre-rRNA). RRM-containing proteins form a prominent class of RBPs performing important biological functions.

In this study, we screened Ae. aegypti genome and identified the RRM-containing proteins using various in silico approaches. We also carried out expression analysis of the genes encoding the RBPs to elucidate their association with mosquito growth and developmental stages using stage-specific transcriptome data (Akbari et al. 2013).

Material and methods

Identification and characterization of RRM-containing proteins in Ae. aegypti

To search for RRM-containing proteins in Ae. aegypti genome, we used multiple in silico approaches. First, “RRM and Ae. aegypti” was used as the keyword in GenBank to retrieve proteins containing RRM. Second, 134 RRM-containing proteins of D. melanogaster (Gamberi et al. 2006) were used as a protein query against Ae. aegypti in BLASTp search. Third, the UniProt entries for Ae. aegypti for RRM-containing proteins were retrieved. Fourth, RRM profiles from PROSITE for IDs PDOC00030, PDOC51472, and PDOC51939 were used as a query for BLASTp against Ae. aegypti database with default parameters. The proteins obtained from each strategy were carefully examined to remove duplicates, and a list of unique proteins was obtained. This list was further used in the BLASTp search against Ae. aegypti protein database to retrieve proteins that might have been missed from earlier searches.

The list of proteins obtained was analyzed using SMART and Conserved Domain (CD) databases to confirm the presence of RRM. The domain organization for each protein was retrieved from the SMART database. The chromosomal locations, protein length, number of transcripts, and protein IDs were obtained from the VectorBase (https://vectorbase.org/). Functional annotation was done using Blast2GO (Conesa et al. 2005), and the descriptions for each protein were based on available annotations in Genbank/VectorBase or sequence similarity with D. melanogaster (> 30% identity).

Phylogenetic analysis, chromosomal mapping, and gene structure

Phylogenetic analysis of full-length amino acid sequences of Ae. aegypti RBPs was aligned using CLUSTAL W, and neighbour-joining tree with 500 bootstraps was constructed using MEGA7.0 (Kumar et al. 2016). The most similar proteins in D. melanogaster and Bombyx mori were retrieved from VectorBase (https://vectorbase.org/vectorbase/app) and Ensembl (https://metazoa.ensembl.org/index.html), respectively. Annotations were added to the tree using the iTOL server (Letunic and Bork 2021). Chromosomal mapping of the genes on Ae. aegypti chromosomes was done using Mapchart 2.32 (Voorrips 2002). Intron-exon gene structure schematics were generated using GSDS an online server (Hu et al. 2015).

Expression profiling of the genes encoding RRM-containing proteins

RNA sequencing data from 41 different developmental time points throughout the Ae. aegypti life cycle from eggs up to the adult stage were obtained from Sequence Read Archive (SRA) a public repository database, with accession number SRP026319 (Akbari et al. 2013). Transcriptomic data were mapped to the Ae. aegypti AaegL5.0 reference genome using HISAT2 (Kim et al. 2019) to identify their genomic positions. The reads mapped to each gene were assembled and quantified using Cufflinks (Trapnell et al. 2012). The read count was quantified as the number of fragments per kilobase of exons per million mapped fragments (FPKM). The FPKM for Ae. aegypti RRM-containing RBPs were used for clustering and heatmap generation using Morpheus, (https://software.broadinstitute.org/morpheus).

Results and discussion

Overview of RBPs in Ae. aegypti

Using multiple in silico strategies, Ae. aegypti genome was screened, and 135 RRM-containing RBPs were identified (Supplementary Table 1) which is comparable to the number of RBPs in D. melanogaster (n = 134) (Gamberi et al. 2006; Sysoev et al. 2016) and B. mori (n = 123) (Wang and Zhou 2009). Orthologs of 133 RBPs were identified in D. melanogaster (Supplementary Table 1). Annotations were taken as available in Genbank or VectorBase, and the remaining unannotated genes were putatively characterized based on sequence similarity with D. melanogaster. One hundred twenty-six proteins could be described (Supplementary Table 1), while nine RBPs could not be assigned any description. Six of these nine RBPs contained multiple RRMs: AAEL022693 and AAEL012243 had six RRM copies; AAEL019864 and AAEL023907 contained five copies; AAEL022113 and AAEL004699 had four copies. Three of the nine RBPs contained a single RRM (AAEL022876, AAEL025927, AAEL019879).

RRM is often found as a single copy or in multiple copies and is also commonly found in combination with other motifs (Maris et al. 2005; SenGupta 2013; Loerch and Kielkopf 2015). Out of 135 RBPs, 41 (30%) had a single RRM, and 52 proteins had multiple RRMs (37%). The remaining 42 RBPs contained RRM along with other motifs (33%). Zinc finger (Znf) was the most common as earlier seen in many other organisms (Maris et al. 2005; Mahalingam and Walling 2020). Znf is a small protein motif with multiple finger-like protrusions that make tandem contacts with their target molecule. Many superfamilies of Znf have binding affinities to DNA/RNA/proteins based on their sequence and structure. For example, CCHHs are known to bind double-stranded DNA/RNA, and CCCH and CCHC types bind single-stranded RNA (Summers 1991; Iuchi 2001; Michel et al. 2003; Wang et al. 2021). A total of 13 Znf motifs including five C3H1, three C2HC, two C2H2, one ring finger, and three RBZ were observed in 11 proteins. The ring finger is involved in protein-protein interaction, and the Znf-RBZ is related to the ubiquitin-binding function (Lorick et al. 1999).

Other motifs included PWI, KH, and G-patch, which are also known for their RNA-binding affinities (Szymczyna et al. 2003; Dong et al. 2004; Valverde et al. 2008; Aksaas et al. 2011; Zhang et al. 2016). Apart from this, motifs like SURP, SAP, LA, SPOC, RPR, HAT, Lsm_interact, methyltransferase, muHD, and CID are involved in transcriptional and translational processes. For example, SURP is found in splicing regulatory proteins (Kuwasako et al. 2006). SAP motif has been identified in proteins involved in DNA repair, RNA processing, and apoptotic chromatin degradation (Aravind and Koonin 2000). Three proteins contained the LA motif. LA motif is involved in the maturation of RNA polymerase III transcripts (Alfano et al. 2004; Dong et al. 2004). LA motif also recognizes mRNAs with a 5′-terminal oligopyrimidine motif (5′TOP) important for protein synthesis (Pellizzoni et al. 1997). One protein had RRM in combination with the cleavage stimulation factor (CSTF) which is known for 3′-end cleavage and polyadenylation of pre-mRNAs (Mandel et al. 2008). HAT, a helical repeat motif, is involved in RNA-binding and RNA remodeling (Hammani et al. 2012). Nuclear transport-like factor (NTF2) was identified in one Ae. aegypti RBP. NTF2 functions as a cytosolic factor for nuclear import and also interacts with nuclear pore complex protein (Paschal and Gerace 1995). The ubiquitin-binding–associated (UBA) motif was found in one of the proteins. UBA mediates ubiquitination (Hurley et al. 2006). The motifs having specific functions may modulate the binding affinity, specificity, and versatility of RBPs. They may also be responsible for locating and interacting with multiple RBPs.

RRMs are also commonly seen in multiple copies within a protein. Fifty-two proteins contained multiple RRMs ranging from two to seven. Among proteins with multiple RRM copies, AAEL021950 contained the maximum number of seven RRMs, and four RBPs, namely, AAEL004075, AAEL012243, AAEL020196, and AAEL022693, contained six RRMs. Some of the RBPs are known to have a specific number of copies. For example, a single RRM is found in spliceosomal U1 70 kDa protein and two in U2 auxiliary factor 65 kDa subunit, three RRM copies are generally seen in ELAV protein, and the polyadenylate binding proteins (PABPs) are known to have four RRM copies (SenGupta 2013). Although the function of each RRM among multiple copies is unknown, they may increase specificity by binding to the long stretch of RNA sequence and functioning cooperatively (Maris et al. 2005). They may also have different specificities and thus diversify the biological functions of the protein. Moreover, it is also possible that all of the RRMs may not be functional or may not be able to bind RNAs (SenGupta 2013).

Chromosomal mapping and functional enrichment

RBPs were mapped onto the Ae. aegypti chromosomes, and 135 RBP encoding genes were found evenly distributed across the length of all the three chromosomes (Fig. 1). The largest chromosome (chr_2) contained a maximum of 53 genes; chromosome 3 (chr_3) had 45, and the smallest chromosome (chr_1) contained 37, the lowest number of genes. The length of proteins varied from 91 to 5680 amino acids including 43 small proteins (< 320 amino acids) and 32 large proteins (> 700 amino acids). The majority (75%) of the proteins that had only one exon were small proteins. The number of predicted transcripts for each protein varied from one to 29, and 56% (76/135) of the total genes had only one transcript.

Fig. 1
figure 1

Chromosomal distribution of the genes encoding RNA-binding proteins containing RNA recognition motif (RRM) across three chromosomes (chr_1, chr_2, and chr_3) of Aedes aegypti. The numbers on the left of each chromosome are the locations of the genes (in Mb), and the protein IDs are shown on the right side

Full size image

Gene ontology (GO) analysis with Blast2GO categorized RBPs based on their putative functions. As expected, RNA-binding was the most enriched term among molecular functions, and in biological processes, Ae. aegypti RBPs were found to be associated with gene expression regulation, nucleic acid and protein metabolism, macromolecule biosynthesis, cellular component assembly, protein-containing complex, macromolecule modification, and ribonucleoprotein complex. This indicates the involvement of RBPs in several biological processes as seen in several other organisms (Fig. 2).

Fig. 2
figure 2

Gene ontology classifications of the biological process, molecular function, and cellular localization of the genes encoding RNA-binding proteins containing RNA recognition motif (RRM) in Aedes aegypti. GO distribution for the top 20 levels has been shown

Full size image

Expression analysis of RBP genes using RNA sequencing data

Since developmental stage-specific transcriptomes of Ae. aegypti were available in a public database for several time points throughout its life cycle, starting from early embryo up to adult mosquitoes (Akbari et al. 2013), it was possible to explore the stage-specific expression of RBPs in Ae. aegypti. From the analysis of RNA seq data, we retrieved FPKM values for 135 RBP encoding genes at 41 developmental time points. All the genes had > 1.0 FPKM in more than one stage and thus were considered expressed. Hierarchical clustering of the gene expression data indicated that many genes exhibited preferentially high expression in the embryonic stages and in blood-fed ovaries (Fig. 3). A subset of RBPs were found highly expressed in early embryonic stages, and interestingly, those RBPs were also highly expressed in blood-fed ovaries (Fig. 3). This indicates that these RBPs might be maternally inherited and play essential roles in embryonic development. Some of these genes included SR proteins, cytoplasmic polyadenylation-binding protein, La protein, gawky protein, boule, and among others. Further characterization of these genes might explain molecular basis of embryogenesis in mosquitoes. Two RBPs (AAEL010665 and AAEL013869) were also identified in the Ae. aegypti egg shell proteome analysis (Marinotti et al. 2014) suggesting their possible role in egg development. The detailed characterization of these genes is warranted to identify the candidates for developing new mosquito growth regulators or insecticides.

Fig. 3
figure 3

Expression profiles of the genes encoding RNA-binding proteins containing RNA recognition motif (RRM) in Aedes aegypti across 41 developmental time points starting from eggs up to adult mosquitoes. These developmental stages include specific time intervals of embryonic development, larval stages, sex-biased expression of male versus female pupae and carcass, blood-fed versus non-blood-fed ovary and female carcass, and testes-specific expression. Colors from white to black represent increasing gene expression. The genes showing high expression in the early embryonic stages and the blood-fed ovaries are shown within brown box

Full size image

Phylogenetic relationships, gene structure, and domain organization

RRM-containing proteins belonged to a wide range of functional categories including serine arginine (SR)-rich proteins, heterogeneous nuclear RNP (hnRNPs), small nuclear ribonucleoproteins (snRNPs), nucleolysins, polyadenylate-binding proteins (PABP), ELAV-like proteins, and eukaryotic translation initiation factors (Table 1). We analyzed the gene structure, domain organization, and phylogenetics of some of the major Ae. aegypti RBP classes and compared with D. melanogaster and B. mori.

Table 1 Summary of the functional classes of proteins and the number of genes in each class
Full size table

Spliceosome-associated proteins

Pre-mRNA splicing needs core spliceosomes, cis-sequence elements, and several RBPs (Fredericks et al. 2015; Vuong et al. 2016). Among RBPs, snRNPs form the core components of the spliceosomes, and SR and hnRNPs proteins are the major splicing factors. Thirty-six proteins with a putative role in splicing were identified, including 12 SR proteins, eight hnRNPs, three pre-mRNA splicing factors, eight snRNPs, and five other splicing factors (Supplementary Table 1). Brooks and co-workers (Brooks et al. 2015) reported 56 Drosophila RBPs involved in splicing. A BLAST search using 56 proteins against Ae. aegypti genome identified 37 proteins containing RRM. These 37 proteins belong to different categories including six SR proteins, two pre-mRNA splicing factors, one each from hnRNP, PABP, and ELAV along with two nucleolysins, and two translation initiation factors. This is possible, as many RBPs are known to perform multiple functions. However, the functional characterization of Ae. aegypti may provide accurate information.

Serine arginine-rich (SR) proteins

A total of 12 SR proteins were identified in Ae. aegypti. SR proteins contain SR repeat regions at the C terminal and RRM at the N terminal. Nine proteins had a single RRM and five had two RRMs. One of the proteins contained Znf along with RRM, and another protein had an RPR motif which helps in protein-protein interaction. SR proteins are known to be a part of spliceosome assembly and regulate the alternate splicing (Jeong 2017), while RRM in the protein binds to the splicing enhancers to regulate processing events and the SR-rich region binds to the other proteins or RNAs.

We analyzed 12 Ae. aegypti putative SR proteins along with ten SR proteins from D. melanogaster and nine from B. mori. Phylogenetic analysis revealed that SR proteins were highly conserved among the three insect species. Ae. aegypti proteins clustered closely with their orthologs from both D. melanogaster and B. mori (Fig. 4). Protein length and domain organization were conserved (Matthews et al. 2018), while gene length and the intron-exon structure were highly variable across the species. The number of introns and their organization also varied among the species. These findings are in corroboration with the findings of genomic comparisons between Ae. aegypti, Anopheles gambiae, and D. melanogaster (Nene et al. 2007) which observed that Ae. aegypti genes are generally longer and contain longer introns. This suggested that the genes have evolved across the species but they are highly conserved at the protein level indicating the biological importance of these proteins and the conserved functional roles played by them.

Fig. 4
figure 4

Phylogenetic tree of Aedes aegypti, Drosophila melanogaster, and Bombyx mori serine/arginine-rich proteins containing RNA recognition motif (RRM). The tree was constructed using the neighbour joining (NJ) method with 500 bootstraps. The domain organization of each RBP and the intron-exon architecture of the genes encoding them are also shown. Bootstrap values are represented by the size of the red circle on each branch

Full size image

Small nuclear ribonucleoproteins (snRNPs)

Removal of introns from pre-mRNA (splicing) is a crucial step in eukaryotes’ gene expression. This is carried out by a ribonucleoprotein complex called spliceosome which is consisted of snRNPs and various other proteins. Seven snRNPs containing the RRM were identified in Ae. aegypti. snRNP is a complex of proteins and snRNA. snRNPs associated with specific U-rich snRNAs are the core components of the spliceosomes involved in splicing (Solymosy and Pollák 1993). The most abundant and well-characterized snRNPs include U1, U2, U4, U5, and U6 which form the major components of spliceosomes (Solymosy and Pollák 1993), and the minor spliceosomes contain U11, U12, U4atc, U5, and U6atac (Will and Lührmann 2011; Matera and Wang 2014). snRNP in Ae. aegypti included three U1 snRNPs, one U11/12, two proteins with U2 auxiliary factor, and one with U2-associated SURP motif-containing protein. Phylogenetic analysis of these snRNPs along with D. melanogaster and B. mori revealed a similar pattern as observed in SR proteins. The proteins and the domain organization were highly conserved, but variations were observed in gene length and intron-exon organization (Fig. 5).

Fig. 5
figure 5

Phylogenetic tree of Aedes aegypti, Drosophila melanogaster, and Bombyx mori small nuclear RNA nucleoproteins (snRNPs) containing RNA recognition motif (RRM). The tree was constructed using the neighbour-joining (NJ) method with 500 bootstraps. The domain organization of each RBP and the intron-exon architecture of the genes encoding them are also shown. Bootstrap values are represented by the size of the red circle on each branch

Full size image

Heterogeneous nuclear ribonucleoproteins (hnRNPs)

RRM is also found in several hnRNPs. Many hnRNPs are known to regulate alternative splicing and protein components of snRNPs. hnRNPs associate with nascent RNAs and lead to their export, translation, localization, and stability (Krecic and Swanson 1999; Geuens et al. 2016). A total of eight hnRNPs were identified, five of which encoded two RRMs and three proteins had three RRMs. All the proteins are clustered in a phylogenetic tree as per their domain organization (Fig. 6). Two of the Ae. aegypti hnRNPs (AAEL005515 and AAEL005049) were clustered with Drosophila squid protein. Squid is the most abundant hnRNP protein which is expected to bind most cellular RNAs. Squid plays an important role in dorsoventral axis formation during oogenesis by localizing GurkenRNA (Norvell et al. 1999; Steinhauer and Kalderon 2005; Cáceres and Nilson 2009). Mutants of Squid have shown its importance in oogenesis and embryo development. For example, homozygous mutants laid severely dorsal eggs that could not mature into adulthood (Kelley 1993; Matunis et al. 1994); complete deletion of squid genes lead to lethality (Matunis et al. 1994), while deletion in the first exon of the squid resulted in larval death (Matunis et al. 1994). Ae. aegypti proteins AAEL005515 and AAEL005049 have not been characterized. The amino acid sequences of both proteins are > 75% identical to the Drosophila squid protein (Supplementary Figure 1). However, the gene structure and the overall gene length were different (Fig. 6). Protein-protein interaction analysis with STRING database (Szklarczyk et al. 2021) using default parameters revealed interactions between AAEL005515 and AAEL005049 as well as with the other nine proteins (Fig. 7). Seven of the nine proteins coded for RNA-binding proteins. In particular, AAEL005515 and AAEL005049 were found interacting with all other nine proteins indicating their importance in the network. The sequence conservation across the species indicates functional relevance. However, an in-depth analysis of these genes could provide greater insights and envision.

Fig. 6
figure 6

Phylogenetic tree of Aedes aegypti, Drosophila melanogaster, and Bombyx mori heterogeneous nuclear ribonucleoproteins (hnRNPs) containing RNA recognition motif (RRM). The tree was constructed using the neighbour-joining (NJ) method with 500 bootstraps. The domain organization of each RBP and the intron-exon architecture of the genes encoding them are also shown. Bootstrap values are represented by the size of the red circle on each branch

Full size image
Fig. 7
figure 7

The STRING protein-protein interaction network for Aedes aegypti protein AAEL005515. Colored lines between the proteins indicate interactions

Full size image

Polyadenylate-binding proteins (PABPs)

PABPs, another RRM containing RBPs, are involved in protein synthesis, mRNA stability, and mRNA biogenesis (Bernstein et al. 1989; Kühn and Wahle 2004). As per published records, only one PABP has been predicted in D. melanogaster (Smith et al. 2014), three in humans, while 12 in Arabidopsis thaliana (Lorković and Barta 2002), five in rice (Belostotsky 2003) and 14 in barley (Mahalingam and Walling 2020). In Ae. aegypti, we identified five PABPs and two cytoplasmic polyadenylation element-binding (CPEB) proteins. Similar to other eukaryotes, four PABPs had multiple RRMs ranging from two to six. However, only one of them contained a polyA domain with four RRMs which is the well-characterized domain pair identified in PABPs. One protein had two RRMs, and three had a single RRM copy. Similar to plants, where PABPs with one or two RRMs have been identified (Lorković and Barta 2002; Mahalingam and Walling 2020), however, the functional characterization of the proteins could shed light on their exact role and functional capabilities.

Conclusion

This study has generated the first reference database for Ae. aegypti RBPs containing RRMs and their expression analysis specific to different developmental time points of the mosquito life cycle. This database would be useful for future studies to select and characterize candidate proteins and elucidate their role in mosquito growth and development. The proteins having a significant impact on the mosquito life cycle can be targeted to develop growth regulators or insecticides. This will also be a useful reference base for future studies to characterize the entire RBP repertoire of Ae. aegypti as well as other mosquito species. It can further be extended to identify and validate the genes experimentally through interactome capture from the different developmental stages of mosquitoes to identify candidate genes.