Transcriptome exploration of the sex pheromone gland of Lutzomyia longipalpis (Diptera: Psychodidae: Phlebotominae)

Abstract

Background: Molecules involved in pheromone biosynthesis may represent alternative targets for insect population control. This may be particularly useful in managing the reproduction of Lutzomyia longipalpis, the main vector of the protozoan parasite Leishmania infantum in Latin America. Besides the chemical identity of the major components of the L. longipalpis sex pheromone, there is no information regarding the molecular biology behind its production. To understand this process, obtaining information on which genes are expressed in the pheromone gland is essential.

Methods: In this study we used a transcriptomic approach to explore the pheromone gland and adjacent abdominal tergites in order to obtain substantial general sequence information. We used a laboratory-reared L. longipalpis (one spot, 9-Methyl GermacreneB) population, captured in Lapinha Cave, state of Minas Gerais, Brazil for this analysis.

Results: From a total of 3,547 cDNA clones, 2,502 high quality sequences from the pheromone gland and adjacent tissues were obtained and assembled into 1,387 contigs. Through blast searches of public databases, a group of transcripts encoding proteins potentially involved in the production of terpenoid precursors were identified in the 4th abdominal tergite, the segment containing the pheromone gland. Among them, protein-coding transcripts for four enzymes of the mevalonate pathway such as 3-hydroxyl-3-methyl glutaryl CoA reductase, phosphomevalonate kinase, diphosphomevalonate descarboxylase, and isopentenyl pyrophosphate isomerase were identified. Moreover, transcripts coding for farnesyl diphosphate synthase and NADP+ dependent farnesol dehydrogenase were also found in the same tergite. Additionally, genes potentially involved in pheromone transportation were identified from the three abdominal tergites analyzed.

Conclusion: This study constitutes the first transcriptomic analysis exploring the repertoire of genes expressed in the tissue containing the L. longipalpis pheromone gland as well as the flanking tissues. Using a comparative approach, a set of molecules potentially present in the mevalonate pathway emerge as interesting subjects for further study regarding their association to pheromone biosynthesis. The sequences presented here may be used as a reference set for future research on pheromone production or other characteristics of pheromone communication in this insect. Moreover, some matches for transcripts of unknown function may provide fertile ground of an in-depth study of pheromone-gland specific molecules.

Figure 1

Tissues of L. longipalpis male included in this study. The stereoscope image of a L. longipalpis male illustrates the three abdominal segments (3^rd, 4^th and 5^th) analyzed. Characteristics of these segments are shown by scanning electron microscopy. Pheromone disseminating structures, which are present in the 4^th abdominal segment are indicated by the arrowhead. Macrotrichias present in the 3^rd and 5^th segment are indicated by the asterisks, and microtrichias, which are present in all abdominal segments, are indicated by the star.

Figure 2

Distribution of the number of ESTs with homology to previously described proteins grouped into functional classes from the three tissues: LL4SEG (4^th segment containing the pheromone gland), LL3SEG (3^rd abdominal segment) and LL5SEG (5^th abdominal segment). Functional classification of the transcripts was based on their sequence similarity to the proteins in the reference databases. A Chi-square test was performed for each EST tissue group across the functional categories to examine the differences between tissues. The double-asterisks represents functional classes where both differences in the number of ESTs were statistically significant (4^th and 3^rd; 4^th and 5^th) and the single asterisk represents functional classes where only one difference in the number of ESTs was statistically significant.

Figure 3

Scheme of the classical isoprenoid pathway exhibiting enzymes for which transcripts have been identified. Enzymes for which transcripts have been identified in the L. longipalpis pheromone gland only are highlighted in red.

Figure 4

3-hydroxy-3-methyl-glutaryl-CoA reductase (HMG-R) sequence analysis. Sequence alignment of Culex quinquefasciatus (Cq), Aedes aegypti (Aa), and L. longipalpis (LLphg-contig_215) HMG-R. Identical residues are highlighted in black and similar residues highlighted in gray (>70%). The illustrated sequences represent a subset of this alignment and include residues that extend from positions four hundred to seven hundred. The predicted HMG-CoA binding motif (ENVIG) is shown in the blue box and the predicted NADP(H) binding motif (DAMGXN) is shown in the green box. The accession numbers of the sequences used are between parentheses.

Figure 5

Phosphomevalonate kinase (PMK) sequence analysis. (A) Neighbor-joining tree of putative PMK from the L. longipalpis pheromone gland (LLphg-contig_437) Aedes aegypti (Aa), Culex quinquefasciatus (Cq), Dendroctonus pondersae (Dp) and Bombix mori (Bm). The accession numbers of the sequences used are between parentheses and node support is indicated by the bootstrap values. (B) Sequence alignment of PMK sequences. Identical residues are highlighted in black and similar residues highlighted in gray (>70%). Portions of the N-terminal region of these putative proteins with the phosphate binding loop is shown in the blue box and the putative mevalonate 5-phosphate binding site in the green boxes.

Figure 6

Isopenteyl diphosphate isomerase (IDI) sequence analysis. (A) Sequence alignment of IDIs from L. longipalpis (contigs 53, 54, 55 and 117) to reference sequences. Identical residues are highlighted in black and similar residues highlighted in gray (>70%). (B) Neighbor-joining tree of IDI full transcripts of L. longipalpis pheromone gland (contigs 53, 54, 55 and 117), Dendroctonus jeffreyi (Dj), Dendroctonus ponderosae (Dp), Bombix mori (Bm), Culex Aedes aegypti (Aa) and quinquefasciatus (Cq). The accession numbers of the sequences used are between parentheses and the node support indicated by the bootstrap values.

Figure 7

Farnesyl diphosphate synthase (FPPS) sequence analysis. (A) Subset of the sequence alignment of FPPS including residues extending from positions one hundred to four hundred forty two. Sequence alignment of the largest predicted FPPS sequences of our library (contig 94 and 93) and Aedes aegypti (Aa), Drosophila melanogaster (Dm), Glossina morsitans morsitans (Gm), Dendroctonous ponderosae (Dp), and Dendroctonus jeffreyi (Dj) sequences. Identical residues are highlighted in black and similar residues highlighted in gray. The aspartate-rich motifs are shown in a blue square. (B) Neighbor-joining tree of FPPS sequences of L. longipalpis pheromone gland (contigs 93 and94), Aedes aegypti (Aa), Drosophila melanogaster (Dm), Glossina morsitans morsitans (Gm), Dendroctonus jeffreyi (Dj), Dendroctonus ponderosae (Dp). The accession numbers of the sequences used are between parentheses and the node support indicated by the bootstrap values.

Figure 8

Farnesol dehydrogenase FDH sequence analysis. (A) Neighbor-joining tree and (B) sequence alignment of two Lutzomyia longipalpis farnesol dehydrogenases sequences and previously described enzymes from Culex quinquefasciatus (Cq) and Aedes aegypti (Aa). Identical residues are highlighted in black and similar residues in gray.