Our study provides the first ever comprehensive in silco evidence across a 1,779 genome-wide database for the highly species-specific conservation of mosquito retroposons. In the absence of a molecular biomarker for mosquitoes, entomological studies of mosquito bionomics have so far involved physical taxonomic classification. We therefore felt it necessary to identify a molecular target that may serve as a biomarker. Such a biomarker, it is envisioned, may enable mosquito speciation to be established by molecular entomology. Our work offers the first support for the hypothesis that mosquito retroposons may be exploited for that purpose. Overall, while several authors have documented the vertical inheritance of mosquitoes 1011, most of these studies have involved too few taxa to support the concept that mosquito retroposons are highly conserved.

First, we have shown that among all three retroposons investigated, a sequence tag for one mosquito species could only be used to identify the derivative retroposon from that species. For instance, searching the entire 1,779 genome-wide nucleotide sequence database using retroposon sequence tags of the Culex pipiens retroposon 5 Cx pip, clone 1 and Culex pipiens retroposon 7 Cx pip, clone 3, yielded hits with contigs of Culex pipiens quinquefasciatus strain JHB, the sub-species of origin (see Tables 1 and [additional files 1 and 2]). This view is further supported by the finding that, since the Anopheles sinensis genome is not included in the genome-wide 201-eukaryote database, no hits irrespective of score were obtained with the sequence tag of Anopheles sinensis retroposon 1 An sin, clone 5 as the query. Moreover, despite the presence of other related species such as Anopheles gambiae str. PEST and Aedes aegypti, none of their retroposons were identical to those of Culex pipien and A.sinensis. It may be argued that the power of our findings is limited by the absence from the 201-eukaryote database of more mosquito species genomes for which horizontal transfer has previously been reported 10, such as those involving (i) Ae.aegypti and Ae.Albopictus, where three cloned PCR products from Ae.albopictus are nearly identical to sequences from Ae.aegypti, (ii) C.quinquefasciatus, for which the PCR sequence groups have homology with C.nigripalpus, O.atropalpus and O.epactius sequences. However, Biedler and Tu 10 have recently shown that such previously-reported relationships are to be expected, since these species belong to the same species complex in which there may be introgression. Moreover, related and more comprehensive experiments based on cloning and PCR analysis of the inheritance of the Mosquito Jockey (JM1-Juan A and Juan C, JM2 and JM3) plus the CR1 clade elements have shown divergence in all groups compared, even among Culicine vs. Culicine (Cul/Cul), with increasing evolutionary distance: Culicine vs. Anopheles (Cul/An), mosquito vs. non-mosquito dipterans (Msq/dip), mosquito vs non-dipteran neoptarans (Msq/Neo) and mosquito vs. vertebrates (Msq/vert) 11.

Second, we note that among lower taxa, species-specific conservation of mosquito retroposons is common. Specifically, although several eukaryotes that contain integral mobile elements or transposable elements were part of the 201-eukaryote genome database searched, including (i) Bombyx mori (R2Bm element), (ii) yeast (al1 and aI2), (iii) dipterans and others 7891213, none of their respective retroposons were found to be identical to mosquito retroposons. In addition, no retroposon of bacterial origin was similar to mosquito retroposons (see Figure 1 for taxonomic tree of organisms searched bacteria and archea).

In polarity, several low score hits were found by querying the human genome with the 5 Cx pip, clone 1 retroposon tag of C. pipiens and the 1 An sin, clone 5 retroposon tag from An. sinensis. Note that whereas about 25–35% of the human genome 141516 comprises Long Interspersed Nuclear Elements (LINE) or non-Long Terminal Repeats (LTR), spliceosomal elements that are considered to be ancestors of all group II introns 717, the low score blast hits found by aligning the three query mosquito retroposon tags against the human genome do not support consideration as homologs for which a minimum (>95%) identity score was set. Moreover, most have functions diverging – as shown by such examples of hits as integrin, alpha 1 precursor and spermatogenic leucine zipper 1 – from that of the six moieties of introns including reverse transcriptase or maturase activity [additional file 3] 717. However, it is noticeable that most of these hits occur at extremities of the human genes or contigs, regions often flanked (interspersed) by spliceosomes. While several strategies have been used to differentiate orthologs from paralogs including the use of a protein clock and genome cross-referencing or XREFdb 18192021, we found it appropriate and easier to determine the possible relationships between mosquito retroposons and human splicesomal elements by functional and positional inference. Specifically, despite an outright absence of homology, the localization of all hits at regions occupied by splicesomes within the human genome supports prior work that identifies human genomic spliceosomal elements as possible ancestors of all group II introns 717. While several bioinformatics algorithms and software with greater capacity to predict identity are available, such as space-efficient spliced alignment 22, our choice of the BLAST-N tool 23 in this study was based on its ease of access and link to the organismal genomes of interest. It is therefore likely that insignificant differences may be found when other tools are used 22. This work, however, also serves to uniquely emphasize how simple yet reliable bioinformatics tools like BLAST may still be useful in resolving hypothetic-driven biomedical research questions and hence advancing novel drug, vaccine and diagnostic discovery. Specifically, two potential" highly innovative" applications are likely to accrue for mosquito retroposons given our findings.

First, because the foregoing evidence shows that mosquito retroposons are highly conserved within species, they may be ideal targets for research and development of mosquito-specific molecular biomarkers to employ in molecular entomology. Specifically, DNA probes or retroposon-specific monoclonal antibodies (MAbs) may be mounted on to existing platforms for the molecular characterization of pathogens, such as Polymerase Chain Reaction (PCR), DNA chips or immunohistochemistry.

Second, and more speculative, is the possibility that such mosquito RST-specific MAbs may be conjugated to DDT to enhance targeted delivery of DDT to a mosquito of interest. Specifically, DDT may be conjugated to MAbs through a two step emulsion process, first incorporating DDT into the polyester PLGA, and subsequently into MAbs to form nanoparticles. The choice of design specifying the dissolution of the DDT-PLGA emulsion into MAbs is aimed at manufacturing nanoparticles coated with mosquito RST-specific MAbs. The resultant emulsion may then be allowed to nanoparticulate (precipitate) through magnetic steering as described elsewhere 2425. Hence, these model nanoparticles (see Figure 2) would combine DDT with MAb(s) generated from mosquito RSTs (MAb_RST). Overall, the DDT immunoconjugate strategy is predicted to enhance the accumulation of DDT in the target rather than other organisms. We presume that the proposed DDT immunoconjugates will have the potential to eliminate the ethical controversies surrounding the cumulative toxic effects of conventional DDT. Using DDT immunoconjugates has additional advantages including the fact that, since they may ensure mosquito strain-specific toxicity, one may choose to target DDT to only those mosquito strains or species that are known vectors for pathogen(s) of public health control interest (thereby ensuring that other mosquito species not associated with disease are preserved), which is not possible with conventional unconjugated DDT. DDT immunoconjugates have another advantage in that DDT may be used at lower concentrations than are normally sprayed (dosages subtoxic to other organisms), but still attaining the levels required to kill the target species. Moreover, since DDT is bound to accumulate within the target host, resistance to DDT immunoconjugates is likely to be minimal.

While the proposed use of mosquito RST as a precursor for bioengineering mosquito-specific molecular markers is highly feasible, several concerns are apparent in the equally "highly innovative" DDT immunoconjugate model presented. First, unless novel strategies are devised that enable the MAbs to be stabilized to prolong their t1/2 on exposure to the environment, their faster biodegradation relative to DDT would render the proposed nano-constructs ineffective after a short period in the environment. However, one may still argue that, once sprayed directly into the breeding areas of mosquito larvae, namely stagnant water for the anopheles, these nanoparticles may achieve their purpose if they come quickly into contact with the larvae. It therefore becomes necessary to determine the functional t1/2 of MAbs within the DDT immunoconjugates to establish exactly how long the said nanoparticles could remain viable. Although difficult, ELISA assays may be designed to achieve such measurements, say by taking timed samples of nanoparticles exposed to harsh environments and analyzing them for binding affinity to the specified antigen (mosquito tissue sample). Also, the minimal identity among mosquito retroposons and human splicesomal elements implies that more DDT may accumulate in humans than with conventional DDT. Lethal doses of DDT among humans are however high, although comparative carcinogenic and tumogenic levels for DDT immunoconjugates would have to separately be established 6.

Second, the issue of cost is significant, unless the proposed DDT immunoconjugates are used sparingly, say by spraying directly on to the larvae within stagnant water. Specifically, since MAbs are expensive to synthesize, the cost of the proposed DDT nanoparticles would be high relative to conventional DDT. Therefore, the overall cost effectiveness of DDT immunoconjugates is debatable. In our opinion, in view of the devastating impact of diseases such as malaria on individuals and nations within malaria endemic areas, if such nanoparticles are shown by trials to have promise for eradicating malaria, it may be justifiable to invest funds in them. For instance, economists estimate that malaria accounts for approximately 40% of public health expenditure in Africa and causes an annual loss of $12 billion, or 1.3%, of the continent's gross domestic product 2627. This figure could be re-channeled to DDT immunoconjugates. Of course, results from actual feasibility and efficacy studies will be necessary to convince donors to decide in favor of such opinions.

We offer evidence to support the species-specific conservation of mosquito retroposons. Retroposons may therefore constitute a unique biomarker for mosquito species that may be exploited in molecular entomology. The model proposing the use of mosquito RST-specific MAbs to synthesize mosquito species-tailored insecticides (DDT), however, remains speculative and highly contentious, and calls for further feasibility and effectiveness studies.

There are no potential sources of financial conflicts of interest to declare. BW, KH and WM are all affiliated to Restrizymes Biotherapeutics (U) LTD.

BW and WM conceived of the study, conducted the in silco analyses and contributed to drafting the final manuscript. BW, WM and KH participated in data analysis and writing the final manuscript. All authors read and approved the final manuscript.