Application of RNA interference in triatomine (Hemiptera: Reduviidae) studies

Abstract

Abstract  Triatomines (Hemiptera: Reduviidae) are obligate hematophagous insects. They are of medical importance because they are vectors of Trypanosoma cruzi, the causative agent of Chagas disease in the Americas. In recent years, the RNA interference (RNAi) technology has emerged as a practical and useful alternative means of studying gene function in insects, including triatomine bugs. RNAi research in triatomines is still in its early stages, several issues still need to be elucidated, including the description of the molecules involved in the RNAi machinery and aspects related to phenotype evaluation and persistence of the knockdown in different tissues and organs. This review considers recent applications of RNAi to triatomine research, describing the major methods that have been applied during the knockdown process such as the double-stranded RNA delivery mechanism (injection, microinjection, or ingestion) and the phenotype characterization (mRNA and target protein levels) in studies conducted with the intent to provide greater insights into the biology of these insects. In addition to the characterization of insect biomolecules, some with biopharmacological potential, RNAi may provide a new view of the interaction between triatomine and trypanosomatids, enabling the development of new measures for vector control and transmission of the parasite.

Introduction

In recent years, the increase in the number of genomic and transcriptomic sequences from vectors of diseases available in Databanks has provided a vast amount information to the scientific community for further studies. Most of these sequences have unknown function and only a small portion have been subjected to functional studies. There are entire families of proteins awaiting functional characterization. The challenge now is to insert this large amount of information in the biological context of the arthropods. In this context, the RNA interference (RNAi) has emerged as an attractive and useful tool to study gene function. The technology has already being widely used in gene studies in several areas of biology. Among arthropods, it has been applied to Drosophila melanogaster (Belles, 2010), which is one of the major model organisms for genetic studies, to the coleoptera Tribolium sp. (Tomoyasu et al., 2008), to lepdopiterans (Terenius et al., 2010), to several tick species (de la Fuente et al., 2007), and to triatomine bugs (Araujo et al., 2006). Triatomines (Hemiptera: Reduviidae) are obligate hematophagous insects. They are of medical importance because they are vectors of Trypanosoma cruzi, the causative agent of Chagas disease in the Americas. Currently, 140 triatomine species are formally recognized and they are all capable of transmitting T. cruzi. Despite this, only a small number of species have epidemiological significance as vectors of the pathogen to humans (Schofield & Galvao, 2009), including among them are species such as Triatoma infestans, Triatoma brasiliensis, and Rhodnius prolixus. In addition to being the major vector of Chagas disease parasites in the northern part of South America and in Central America, R. prolixus is also a classical model in insect physiology studies (Fresquet & Lazzari, 2011).

Mechanisms of RNAi

The term “RNA interference” was first used by Fire et al. (1998) and refers to the posttranscriptional gene knockdown in response to the introduction of a double-stranded RNA (dsRNA) in a cell. The success of the RNAi technique is because of its simplicity, ease, and speed with which it can be applied in vivo, using whole organisms, or in vitro, by knockdown in cell or tissue culture (Fjose et al., 2001).

The RNAi mechanism was partially elucidated mainly because of the studies with model organisms such as the nematode Caenorhabditis elegans and the fly D. melanogaster (Roignant et al., 2003; Bischoff et al., 2006; Miller et al., 2008). Among the insects, studies performed in the red flour beetle Tribolium castaneum also contributed to the understanding of the process (Tomoyasu & Denell, 2004; Arakane et al., 2008; Konopova & Jindra, 2008; Minakuchi et al., 2009).

The knockdown of the target gene basically occurs in response to the introduction of a homologous dsRNA into the cells, where the dsRNA is cleaved into smaller fragments (21–25 nucleotides) by a nuclease known as Dicer (Elbashir et al., 2001). This enzyme is similar to type III RNases from Escherichia coli and has several conserved motifs, including a dsRNA binding domain and a helicase domain (Bernstein et al., 2001). The small dsRNA fragments, known as small interfering RNAs (siRNAs), correspond to the “sense” and “antisense” strands of the RNA target. After their production, they associate with cellular proteins to form a multimeric complex called RNA-induced silencing complex (RISC). The RISC contains several proteins, including a helicase that opens the double-stranded siRNA; then, one of the siRNA strands is discarded and the complex containing the antisense strand will associate with the target mRNA and mediate the knockdown (Nykanen et al., 2001). An endoribonuclease also present in the complex degrades the complex by cleaving the target mRNA. The RISC complex is also composed of the proteins of the Argonaute family. This class of proteins forms essential components of the complex. Once siRNA molecules bind to argonaute proteins (Ago), they act as a guide to the target mRNA (Meister & Tuschl, 2004). Ago contain two different domains: a PAZ and a PIWI domain, which are involved in siRNA binding and RNase activity, respectively (Lingel & Sattler, 2005; Tomoyasu et al., 2008).

An additional event in the RNAi pathway is the systemic spread of RNAi from cell to cell throughout the organism and its potential systemic transfer to subsequent generations (Timmons et al., 2003). In plants and C. elegans, two proteins were identified as key molecules for the systemic RNAi pathway: a transmembrane protein known as SID-1 and an enzyme called RNA-dependent RNA polymerase (RdRP). SID-1 is required for receiving the systemic RNAi signal and to promote the cells ability to uptake dsRNA (Winston et al., 2002; Tomoyasu et al., 2008), whereas RdRP amplifies the RNAi effect by synthesizing endogenous dsRNA. In plants, RdRP uses the cleaved transcript fragments as templates to synthesize long dsRNA, which were then diced into secondary siRNAs, whereas in worms (as C. elegans), RdRP synthesizes the secondary siRNAs (exclusively antisense to their mRNA targets), which were bound to secondary Argonautes and can cleave its mRNA targets (Ghildiyal & Zamore, 2009).

The absence of RdRP genes in the fly D. melanogaster (despite the detection of RdRP-like activity) and in mammalian genomes indicates that the RNAi persistent effect might not be possible in all organisms. However, in D. melanogaster, this effect was recently related to a subunit of the RNA polymerase II core elongator complex, known as D-elp1, which proved to have RdRP activity and is able to produce endogenous secondary siRNAs, responsible for extending the knockdown effects (Lipardi & Paterson, 2009). This fact suggests that the RNAi amplification step in insects may possibly be based on different mechanisms, which are still unclear.

With the accomplishment and the availability of data generated by the genome sequencing of R. prolixus, several aspects of the mechanisms of RNAi in triatomine bugs will be elucidated. Among the several contigs obtained from the R. prolixus genome project (genome. wustl.edu/genomes/view/rhodnius_prolixus/), gene sequences similar to the PIWI and PAZ domains (GenBank: ACPB02032227.1) from argonaute family proteins were found. A contig (GenBank: ACPB02024930.1) with high similarity to a Dicer from C. elegans was also found in the R. prolixus genome. However, as in almost all studied insect species, homologs of sid protein (systemic RNAi defective), RdRP, and D-Elp1 were not found. Meanwhile, the persistence of the RNAi effects in triatomines is still unclear, and more studies are necessary to unveil the duration of knockdown effects and the molecules possibly responsible for these persistence.

Methods for achieving RNAi in triatomines

For an efficient and reliable application of the RNAi technique, several factors must be considered including the characteristics of the inducing agent to promote the knockdown, the choice of an unspecific dsRNA to use as a control, the method chosen to deliver the dsRNA into the insect, and the evaluation of the knockdown phenotype. Below we discuss the techniques used for triatomines and the factors to be borne in mind when choosing the best method to be applied.

Delivery of dsRNA

The means chosen to introduce dsRNA into the insect has a direct bearing on the success of the experiment. The efficiency, the ease and practicality of each method, and the advantages and disadvantages of each one must be evaluated according to the experimental design and the aim of the study (de la Fuente et al., 2007). Until now, just a few methods have been used to deliver dsRNA in triatomines: (i) injection (Fig. 1A) or microinjection (Fig. 1B) and (ii) feeding.

Figure 1. Injection of dsRNA into third-instar nymphs of the triatomine R. prolixus. (A) Injection of dsRNA through a thin needle (13 × 3, 30 G, 1/2″) coupled to a Hamilton syringe. (B) Microinjection of dsRNA through microdispensers (about 0.05 mm of diameter) coupled to a microinjector (Nanoinjector; Drummond). Tip thickness of the needle used in conventional dsRNA injection (C) and of the beveled capillary used in dsRNA microinjection (D).

• (i) 
  Injection of dsRNA. The most used delivery methodology in the majority of insect studies has been injection/microinjection of small amounts of in vitro synthesized dsRNA into the insect hemolymph (Price & Gatehouse, 2008). The most used method was the injection of dsRNA into the triatomines hemolymph with a Hamilton syringe connected to a fine needle (Araujo et al., 2007; Mury et al., 2009; Lavore et al., 2012). One of the main problems caused by the injection is the damage caused by the needle, which impairs the use of some insects in further experiments (such as feeding behavior experiments) and dramatically increases the mortality rates. It also complicates interpretation of experiments because of the physiological processes set in train by the damage. This is a particular problem, often ignored, for studying immune responses. Although the mortality rate of the injected insects was not reported, our experience in RNAi technique over the last years showed that the damage caused by the injection can cause very high mortality (up to 100%) and significantly affect the insect's physical integrity, impairing some phenotype studies using silenced insects. In addition to the needle thickness, other parameters such as the amount of dsRNA and the volume injected may contribute to increase the mortality of the triatomine bugs after the injection. The use of a microinjector (Nanoinjector; Drummond, Broomall, Pennsylvania) was a significant improvement for the application of the RNAi technique (Araujo et al., 2009; Paim et al., 2011), especially in small nymphs (first, second, and third instar) of triatomine bugs. In addition to permit a better control of the volume injected, the use of a microinjector with beveled capillary tips favors a smoother injection, causing just a small puncture in the thorax of the bug. The thickness of the capillary used attached to the microinjector (about 0.05 mm of diameter) is much smaller than the thinnest needle (13 × 3, 30 G, 1/2″) used in the conventional injection with a Hamilton syringe (Figs. 1C and 1D). Microinjected T. brasiliensis nymphs show little mortality (<5%) and were as active as noninjected ones (Paim et al., 2011).
• (ii) 
  dsRNA ingestion. The delivery of in vitro synthesized dsRNA to triatomines by feeding was used because of the difficulty in introducing the dsRNA into first- and second-instar nymphs by injection. Second-instar nymphs were fed with an artificial solution containing 1 mg/mL of the target dsRNA, in an artificial feeder (Araujo et al., 2006). Initially, Price and Gatehouse (2008) thought that the use of dsRNA feeding for gene knockdown would be unfeasible in insects, but the method was successfully used in both triatomines (Araujo et al., 2006) and other arthropods (Soares et al., 2005; Turner et al., 2006; Walshe et al., 2009). The RNAi effects in insects due to the dsRNA ingestion could be explained by two views: the dsRNAs are absorbed by the midgut epithelium and transferred to the hemolymph and from there to the the target tissue (Price & Gatehouse, 2008). Alternatively, the dsRNA molecules could pass directly through the intercellular spaces of the midgut epithelium to reach the hemolymph (Araujo et al., 2006) because molecules bigger than dsRNA, including host immunoglobulin G, have been shown to cross the undamaged gut epithelium in a variety of insect species (Schlein et al., 1976; Hatfield, 1988; Ramasamy et al., 1988; Lackie & Gavin, 1989; Allingham et al., 1992). The disadvantages of the feeding method are the difficulty in controlling the volume ingested by nymphs and consequently regulating the quantity of dsRNA received by each insect in the experimental group. In addition, the knockdown by dsRNA ingestion (about 42%) proved to be less effective than injection (about 75%) (Araujo et al., 2006). Despite these disadvantages, the dsRNA ingestion method could be useful for screening a large number of genes by RNAi in small nymphs (first instar). On the contrary, the knockdown by dsRNA feeding causes less damage to the nymphs and keeps them healthier, with mortality considerably reduced (Araujo et al., 2006).

Novel methods for dsRNA delivery have been developed to be used in the RNAi, such as the production of transgenic insects expressing gene-specific dsRNA or by the introduction of a plasmid system, which would generate dsRNAs under the control of an inducible promoter (Kang & Hong, 2008). These methods are already used in Drosophila studies (Giordano et al., 2002; Dietzl et al., 2007) but so far have not been published in works with triatomine bugs. A significant knockdown verified on mRNA levels (more than 99%) and phenotype (46% of reduction in the eggs laid) was achieved in R. prolixus females by ingestion of the bacteria strain E. coli HT115(DE3) expressing dsRNA of the Rhodnius heme binding protein (Taracena et al., 2011, unpublished data), which suggests that this method is another alternative for dsRNA introduction in triatomines.

Knockdown inducers

Knockdown is typically induced through the introduction of molecules of dsRNA or siRNA. In worms, plants, and insects, the introduction of long molecules of dsRNA is the main approach used to trigger RNAi knockdown. However, in mammals, the introduction of long dsRNAs promotes a nonspecific type I interferon response leading to cell death. Thus, only siRNA molecules, which can bypass the interferon response (Elbashir et al., 2001), are efficiently used for RNAi studies in mammals (Waters & Swedlow, 2007). In a few cases, the direct delivery of siRNA was adopted to promote knockdown in insects (Levin et al., 2005; Mutti et al., 2006; Zhou et al., 2006), but dsRNA molecules, which are as effective as the siRNAs, are easier and less expensive to produce and are currently the most frequent molecules used for insects.

An important aspect to be considered in RNAi assays is the feature of the introduced dsRNA molecules, as the length and the nucleotide sequence. The specificity of the knockdown might be compromised by promoting undesirable effects due to the repression of nontarget mRNA molecules (Semizarov et al., 2003). The off-target effects related to the nucleotide sequence of the dsRNA should be considered if it shows, even at low level, some similarity to short sequences of other mRNAs (Tschuch et al., 2008). In triatomines, off-target knockdown was reported by Araujo et al. (2006), when in addition to the targeted gene [nitrophorin (NP)2], others NP transcripts with high homology were also knocked down (69% of reduction in NP1 and 83% of reduction in NP3 transcripts). One molecule of siRNA can promote the knockdown of the target gene and also of genes with regions with greater than 80% similarity to the specific siRNA sequence. siRNA sequences are often composed of 21–25 nucleotides; consequently, it is not difficult to find knockdown in genes with similarities of this extent (Ramakrishnan et al., 2005). Among the measures used to minimize off-target effects are the requirement to avoid targeting gene regions that encode highly conserved domains during dsRNA design and the checking of nucleotide sequences from a careful search in the database or in available algorithms in an attempt to find homologies to others expressed sequences (Kulkarni et al., 2006). Besides the databank comparison by BLAST similarity searches (http://blast.ncbi.nlm.nih.gov/Blast.cgi), some algorithms were specifically developed to provide accurate off-target searches for long dsRNA sequences. Among them are the software dsCheck (Naito et al., 2005), the Web tool from http://rnai.cs.unm.edu/rnai/off-target (Qiu et al., 2005), and the Web-service e-RNAi (Horn & Boutros, 2010). At present, these freely available softwares do not check for the possibility of off-target with triatomine sequences because complete triatomine databases, wherein the dsRNA sequences can be compared, are still not available. However, many species of insects have databases available in these sites such as some mosquito species (Anopheles gambiae, Aedes aegypti), D. melanogaster, and T. castaneum.

The problem of nonspecific knockdown can also be related to the use of long dsRNA molecules (Qiu et al., 2005; Kulkarni et al., 2006; Moffat et al., 2007). Despite the high success rates, the large dsRNA molecules are more likely to generate a nonspecific response (Hannon, 2002). The quantity of dsRNA to be introduced also needs to be planned to promote the maximum knockdown with the minimum dsRNA dose. The careful titration of dsRNA dose used for RNAi experiments in triatomines, such as that carried out by Paim et al. (2011) for the brasiliensin gene in T. brasiliensis, can avoid the use of excessive and unnecessary dsRNA and minimize the chance of side effects (Waters & Swedlow, 2007). The ideal dose of dsRNA to be introduced must be evaluated for each species, life stage, and target gene. Although the gene that encodes the intestinal anticoagulant brasiliensin was effectively knocked down in T. brasiliensis third-instar nymphs by two injections (with 48 h of interval) of 3 μg of dsRNA (Paim et al., 2011), the same dose of infestin dsRNA (an ortholog of brasiliensin gene) was unable to induce knockdown in third-instar nymphs of T. infestans (unpublished data). A possible reason for this is that T. infestans has more than twofold anticoagulant activity in its intestine compared with T. brasiliensis, possibly requiring a higher dose of dsRNA to achieve knockdown.

RNAi can even act on the innate immune response in mammals by interferon activation, which is triggered by siRNA in cell cultures (Sledz et al., 2003). Insects, in contrast to mammals, do not have interferon response, but other molecules that comprise the immunity system can be affected by the dsRNA introduction into organism. The dsRNA machinery in insects is part of their immune system and that one of its major functions is dealing with viruses that produce dsRNA at some stage in their life cycle. De Marco et al. (2010) reported an immune response induction in the triatomine T. infestans by dsRNA. Both the hemolymph injection of the E. coli bacteria and unrelated dsRNAs were able to increase the expression of the first pacifastin domain (TiPP1) in the bug fat body. Pacifastin belongs to a protease inhibitor family, which seems to participate in several processes related to the innate immune system in arthropods, although its general function is still not known. The authors also observed upregulation of TiPP1 expression in T. infestans after feeding on mice infected with T. cruzi. So, like pathogens such as bacteria and trypanosomatids, dsRNA molecules (even unrelated dsRNAs) also demonstrated competence to affect the insect immune system that can compromise or mask the knockdown effects. The act of piercing the insects can itself stimulate immune responses (Randolt et al., 2008). In this case, a noninjected group can be used as an additional control (Paim et al., 2011).

Therefore, caution must be exerted in the interpretation of data from experiments using RNAi technology to avoid reaching confounding knockdown effects. To minimize the occurrence of mistakes in the knockdown evaluation, it is essential to use a nonspecific dsRNA introduced in the same conditions of the specific dsRNA as another control group. The β-lactamase gene (Araujo et al., 2006, 2007; Lavore et al., 2012), the β-galactosidase gene (Mury et al., 2009), and the mice keratin gene (Paim et al., 2011) have been successfully used as control dsRNA in triatomine studies. For other insects, one of the most widely used nonspecific gene is the green fluorescent protein (Mutti et al., 2006), which despite having been proven to be an effective and relatively inert dsRNA, has not been used for triatomine studies until now.

Phenotype evaluation

The application of RNAi in triatomines is affected by the lack of information about the physiological factors that modulate the transcription/translation of the target genes. This fact is crucial to determine the ideal time and route to introduce the dsRNA into the bug and to evaluate the knockdown phenotype. This last aspect is a common complication in the use of the RNAi technique, especially if the target is a gene with completely unknown function. When the gene to be knocked down has a suggested function, it is easier to test the loss-of-function phenotype, as was performed by all triatomine studies listed in Table 1. Usually, in triatomine studies, the reduction in mRNA levels can be seen quickly within 48 h after dsRNA injection (Araujo et al., 2006, 2007, 2009; Mury et al., 2009; Paim et al., 2011), the mRNA levels were already significantly reduced in the target tissues. However, the effective knockdown requires the reduction of the relevant protein levels, which normally requires periods as long as the half-life of the protein coded. So, although new molecules of mRNA are eliminated, the existing target protein is not affected by RNAi and still needs to be spent or degraded before the knockdown phenotype becomes apparent. Consequently, the start and the persistence of RNAi effects may be correlated with the half-life and the turnover rate of the target protein (Huvenne & Smagghe, 2010). Thus, there is not a direct relation between the mRNA levels and protein expression, which depends on the control mechanisms of gene expression, including transcriptional control mechanisms and control of protein half-life in eukaryote cells, suggesting that it is difficult to predict correlation between the mRNA levels and corresponding protein levels (Gygi et al., 1999).

Table 1.  Overview on use of RNAi in triatomines with dsRNA being applied through different methods.

Organism	Target genes	Location	Stage	Application method	Amount dsRNA	dsRNA length (bp)	mRNA knockdown	Knockdown evaluation methods	Reference
The insect species and stage, the target gene and its location are given. The application method, the amount and length of dsRNA is mentioned if the data were present, the knockdown level, the recovery of the gene expression, and evaluation method are summarized.
Rhodnius prolixus	Nitrophorin 2	Salivary glands	Fourth-instar nymphs	Hemolymph injection	2 × 15 μg	548	38% ± 7% (One injection)	RT-PCR	Araujo et al. (2006)
							75% ± 14% (Two injections)	Anticlotting activity
R. prolixus	Nitrophorin 2	Salivary glands	Second-instar nymphs	Ingestion (artificial feeder)	13 μg	548	42% ± 10%	RT-PCR	Araujo et al. (2006)
								Anticlotting activity
Triatoma brasiliensis	Brasiliensin	Anterior midgut	Fourth-instar nymphs	Hemolymph injection	2 × 15 μg	575	42% ± 14% (One injection)	RT-PCR	Araujo et al. (2007)
							71% ± 11% (Two injections)	Thrombin inhibitory activity
								Anticlotting activity
								Blood intake
T. brasiliensis	Brasiliensin	Anterior midgut	Third-instar nymphs	Hemolymph microinjection	2 × 3 μg	575	99%	qPCR	Paim et al. (2007)
								Anticlotting activity
								Cibarial pump electromyogram
R. prolixus	Nitrophorin1	Salivary glands	Fourth-instar nymphs	Hemolymph microinjection	2.6 μg of each NP (10.5 μg on total)	380	99% for all NPs	qPCR	Araujo et al. (2009)
	Nitrophorin 2					548		Anticlotting activity
	Nitrophorin 3					566		Hemeprotein content
	Nitrophorin 4					547		Salivary gland aspect (color)
								SDS PAGE
								Cibarial pump electromyogram
R. prolixus	Gap gene giant	Embryos	Adult female	Hemolymph injection	2 μg	–	–	Defect in the formation of the pregnathal, mandibulary, maxillary, and anterior abdominal segments.	Lavore et al. (2012)
R. prolixus	α-Glucosidase	Midgut	Adult female	Hemolymph injection	2 or 10 μg	–	Drastic reduction	qPCR	Mury et al. (2009)
								α-Glucosidase activity
								Hemozoin formation
								Mortality
								Oviposition

Even when the function of the target genes tested is well known or suggested, the evaluation of the phenotype may present some peculiarities. An interesting aspect of phenotype evaluation can be exemplified by the knockdown process of genes expressed in the salivary glands of triatomines, when the phenotype is apparent only after the molt to the next instar, despite the fact that mRNA levels reduction could be observed 48 h after dsRNA introduction. This fact occurs because the saliva is stored in a reservoir, and the phenotype of the silenced salivary gland can only be observed after the saliva is resynthesized, without or with low levels of the target protein, to refill the saliva reservoir. In addition, the salivary glands volume increases considerably after the molt with a doubling in size (Guarneri et al., 2003). The knockdown phenotype of nitrophorins RNAi (NPs1–4) could be easily verified in fifth-instar R. prolixus nymphs (dsRNA injected in fourth-instar nymphs) simply by observing the salivary glands (Fig. 2). The salivary glands from the knockdown group lost their typical reddish color (because of the presence of the hemeproteins). When the saliva of the two lobes was spent after feeding, both lobes appeared as a light red color. Although the two lobes of the gland have been knocked down, the salivary gland lobe that has not spent saliva remained its normal reddish color.

Figure 2. Rhodnius prolixus salivary glands dissected 7 days after molt to fifth instar of (A) control group, and (B) NPs knockdown group of the two lobes and (C) NPs knockdown group of only one lobe. Modified from Araujo et al. (2006).

In contrast to what occurs in the salivary gland, the knockdown phenotype in the midgut could be observed just a few days after injection (by the fourth day after the first injection or 2 days after the second injection, for brasiliensin knockdown). The possibility to promptly evaluate the phenotype of gene knockdown in the triatomine midgut was reported by Araujo et al. (2007), Mury et al. (2009), and Paim et al. (2011).

It is important to emphasize that the RNAi technique may have its effectiveness compromised, once some dominant phenotypes may cover redundant or secondary functions of a gene, which would make the evaluation of the silenced phenotype unfeasible (Kuttenkeuler & Boutros, 2004). Several members of the same gene family or completely different genes may code different proteins, which can perform similar functions in the organism. This redundancy could result in compensation of the loss of function of the targeted genes, leading to a masking of the silenced phenotype (Kan & Kessler, 2005). The influence of the presence of redundant proteins was observed in the knockdown of salivary apyrase of the mosquito A. gambiae (Boisson et al., 2006). Despite the high reduction in the mRNA levels of apyrase gene (85%), the apyrase activity in salivary gland extracts from dsRNA-injected mosquitoes was reduced only by 35% compared with controls. Anopheles mosquitoes carry in their saliva two different members of the 5′ nucleotidase family, and both can act as apyrases (Ribeiro et al., 2010), suggesting that the mosquito saliva contains redundant molecules with apyrase activity, but only one was reduced by RNAi.

The complex and redundant system of saliva of bloodsucking animals is well known and involves molecules with antihemostatic activities, such as anticoagulants, platelet aggregation inhibitors, and vasodilators, which facilitate the intake of blood from the vertebrate host (Ribeiro & Francischetti, 2003). In triatomines, the concomitant knockdown of the four major NPs (NPs1–4) in the salivary gland of R. prolixus nymphs was necessary for the effective evaluation of the loss-of-function phenotype (Araujo et al., 2009).

RNAi studies in triatomines

Among the hemipteran bugs, the first RNAi-based study was aimed to investigate gene function in the development of the milkweed bug Oncopeltus fasciatus (Hughes & Kaufman, 2000), which is commonly studied using the RNAi technique (Hrycaj et al., 2008; Panfilio, 2009).

The RNAi was applied for the first time in triatomines by Araujo et al. (2006), with the knockdown of a salivary gland gene of R. prolixus. Since then, five other studies were published, all of them aimed to investigate aspects related to the biology and physiology of these insects by knockdown genes expressed in the salivary glands, the intestine, the hemolymph, the embryos, and the eggs from the species T. brasiliensis and R. prolixus. Despite the small number of publications, abstracts, and works presented at scientific meetings show that the RNAi technique has been widely employed by several research groups in several biological situations, especially after the release of the first results of the annotation of the genome of R. prolixus. It is estimated that in the next few years the number of publications will be significantly increased. The features of works already carried out of RNAi in triatomine bugs are summarized in Table 1.

In the work by Araujo et al. (2006), NP2 was knocked down from the R. prolixus salivary glands and the authors showed that the knockdown could be achieved by both dsRNA injection and ingestion. Two injections of 15 μg of NP2 dsRNA proved to be more efficient in fourth-instar nymphs (about 75% knockdown) than that of the ingestion of approximately 13 μg of dsRNA by second-instar nymphs (about 42% knockdown). The NP2 knockdown reduced the plasma coagulation time by approximately fourfold and the salivary gland content lost their dark red color. The biological implication of the NPs knockdown was shown, later on, by Araujo et al. (2009). In that study, the four redundant and major NPs (NPs1–4) were simultaneously silenced from the salivary gland of R. prolixus nymphs. The authors achieved high levels of knockdown of all target NPs (99% of reduction of mRNA levels and 82% of reduction in hemeproteins content) and showed that insects in which NPs1–4 had been knocked down by RNAi had lower ingestion rates and longer contact time with the host when they were fed on mice dorsal surface (small diameter vessels) but not when fed in the lateral tail vein of mice, indicating that for the feeding in large vessels, the salivary molecules (and therefore the NPs) are not important.

Salivary glands are also the objective of RNAi studies in other hematophagous arthropods such as ticks (Ramakrishnan et al., 2005) and mosquitoes (Boisson et al., 2006). These studies often emphasize the role of salivary gland components on the insect blood feeding and on transmission of pathogens. For ticks, it is well known that their saliva is extremely important to ensure a successful feeding to counteract the host hemostasis throughout the long-lasting feeding process in which the tick remains in contact with its host (Ramakrishnan et al., 2005). Unlike the triatomine bugs, ticks do not store saliva inside their salivary glands. The start of the feeding process stimulates a substantial increase in saliva secretion, which enables the tick to concentrate the blood meal by returning excess water and ions to the tick host. The salivary gland undergoes a rapid structural organization after adherence to the host permitting it to execute its functions effectively (Sauer et al., 1995). Thus, amounts of dsRNA as low as 1 μg, are sufficient to promote an effective knockdown (Ramakrishnan et al., 2005). On the contrary, for the mosquito A. gambiae, an efficient knockdown of salivary gland genes required the injection of high amounts of dsRNA in the insect hemolymph (between fivefold and 30-fold more dsRNA than that necessary for knockdown in hemocytes or midgut).

In addition to the salivary glands, the intestinal tract of triatomines, especially the midgut, is another tissue commonly investigated by RNAi studies. The molecules of the midgut environment were reported to influence the feeding process according to Araujo et al. (2007) and Paim et al. (2011). In both studies, the focus was on brasiliensin, a molecule with an anticoagulant activity, described in the anterior midgut of T. brasiliensis. In the paper of Araujo et al. (2007), the brasiliensin knockdown (about 71% reduction in mRNA levels) was achieved by two injections (at 48 h interval) of 15 μg of the specific dsRNA. In this work, the knocked down nymphs ingested approximately 39% less blood than controls. Later, Paim et al. (2011) performed a more detailed analysis of the role of brasiliensin in T. brasiliensis nymphs feeding. The authors verified that smaller amounts of dsRNA (two injections of only 3 μg) were sufficient to significantly reduce the brasiliensin expression in the insect midgut (a reduction of >99% in transcript levels). The lower feeding performance of silenced nymphs observed by the authors was because of the difficulty the nymphs had in maintaining the cibarial pump frequency at high levels throughout the feeding process resulting in lower ingestion rates, even under favorable conditions. The difficulty in feeding was reversed when the silenced bugs were fed on heparinized mice, confirming the role of the midgut anticoagulant activity in the feeding efficiency.

The midgut was also the target organ in the evaluation of heme degradation and hemozoin formation activity by α-glucosidase in R. prolixus by using the RNAi technique (Mury et al., 2009). The knockdown effect was more pronounced by injection of 10 μg of α-glucosidase dsRNA. The authors observed a reduction in α-glucosidase activity in the silenced group, which also had hemozoin formation compromised in the midgut. A high mortality rate (44%) occurred in the knockdown group at 4 days after feeding, which may indicate the vital role of α-glucosidase in the insects’ physiology (Mury et al., 2009). Lavore et al. (2012) studied the embryogenesis process of R. prolixus by RNAi. In this work, the target gene was the gap giant gene, which is expressed in ovaries and maternally supplied in the early embryo and is important to divide the embryo in broad fields. According to the authors, the parental RNAi (dsRNA injected in females) succeeded in compromising embryo head and abdomen formation. The parental RNAi phenotype could be evaluated in embryos of dsRNA-injected bugs, but there is no information about nymphal stages derived from these females.

Prospects of RNAi in triatomines

In recent years, molecular studies on triatomines have provided a wealth of transcriptomic, EST, and cDNA libraries, including the salivary gland transcriptomes of R. prolixus (Ribeiro et al., 2004), T. brasiliensis (Santos et al., 2007), T. infestans (Assumpcao et al., 2008), T. dimidiata (Kato et al., 2010), Dipetalogaster maxima (Assumpcao et al., 2011), and Panstrongylus megistus (Bussacos et al., 2011). The transcriptome of the ovarian follicle tissue was also accomplished (Medeiros et al., 2011) and efforts are being made to accomplish the sequencing of organ-specific cDNA libraries such as the digestive tract, fat body, testes, Malpighian tubules, and entire body of nymphs and adults of R. prolixus. In most studies, parts of the obtained transcript sequences were classified as unknown. Even among transcripts classified as codifying secreted proteins, some of them have only a presumed function, but their biological role has yet to be confirmed. The number of genes with unknown function increases when the approximately five million contigs from the R. prolixus genome project are taken into account (genome.wustl.edu/genomes/view/rhodnius_prolixus/).

Considering the wide range of genetic information currently available for triatomine bugs, RNAi becomes a powerful tool in the study of gene function. This technique is especially important in triatomines (as other nonmodels organisms) once other means of gene silencing, as knockout for example, are usually not possible.

The technically simple methodology and the possibility of evaluating the phenotypes in vivo favor the use of this technique. RNAi research in triatomines is still in the early stages. Just a few studies have been published and most of them used RNAi to silence genes whose function was previously known or presumed. From now onward, the main challenge is to use the technique to functionally characterize genes that encode proteins with unknown function through evaluation of loss-of-function phenotypes. Currently, several research groups are using the RNAi technology in such context and novel publication concerning novel genes might come up soon. Among the main areas where we expect novel discoveries using RNAi is the characterization of molecules with biopharmacological potential and with a role in the interface between vector and parasite, which might allows the development of further ideas and tools for the design of vector and parasite transmission control measures. The possibility of the use of RNAi to control vector-borne diseases was previously showed for malaria. The knockdown of genes with a critical role for parasite development resulted in an impaired development of Plasmodium in Anopheles mosquitoes, decreasing the vector competence (Osta et al., 2004).

Acknowledgments

This work was supported by the Conselho Nacional de Desenvolvimento Científico e Tecnológico, the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior, the Fundação de Amparo à Pesquisa do Estado de Minas Gerais, and the INCT-Entomologia Molecular.

Disclosure

We declare that no competing interests exist.