Molecular evolution of neuropeptides in the genus Drosophila

Abstract

Background: Neuropeptides comprise the most diverse group of neuronal signaling molecules. They often occur as multiple sequence-related copies within single precursors (the prepropeptides). These multiple sequence-related copies have not arisen by gene duplication, and it is debated whether they are mutually redundant or serve specific functions. The fully sequenced genomes of 12 Drosophila species provide a unique opportunity to study the molecular evolution of neuropeptides.

Results: We data-mined the 12 Drosophila genomes for homologs of neuropeptide genes identified in Drosophila melanogaster. We then predicted peptide precursors and the neuropeptidome, and biochemically identified about half of the predicted peptides by direct mass spectrometric profiling of neuroendocrine tissue in four species covering main phylogenetic lines of Drosophila. We found that all species have an identical neuropeptidome and peptide hormone complement. Calculation of amino acid distances showed that ortholog peptide copies are highly sequence-conserved between species, whereas the observed sequence variability between peptide copies within single precursors must have occurred prior to the divergence of the Drosophila species.

Conclusion: We provide a first genomic and chemical characterization of fruit fly neuropeptides outside D. melanogaster. Our results suggest that neuropeptides including multiple peptide copies are under stabilizing selection, which suggests that multiple peptide copies are functionally important and not dispensable. The last common ancestor of Drosophila obviously had a set of neuropeptides and peptide hormones identical to that of modern fruit flies. This is remarkable, since drosophilid flies have adapted to very different environments.

Figure 1

Terminology and amino acid distances. (ai) Peptide copy terminology exemplified by three aligned ASTa prepropeptides from species a1-3. (aii) Processing at dibasic processing sites (indicated in red in (ai)) yields the four neuropeptides ASTa1-4. The C-terminal glycine is further processed to yield the C-terminal amidation. Peptide copies aligning at the same position in the precursor (for example, ASTa1 of species a1-3) will be referred to as orthocopies, which do not have to be sequence-identical. The different copies in a precursor of a single species are paracopies (for example, ASTa1-4 of species a1) = not at the same location. Paracopies may or may not be sequence-identical. (b) Different types of amino acid distances obtained by pairwise comparisons. (bi) The average distance D_o between orthocopies is the arithmetic mean of all individual pairwise distances. It does not contain distances between different paracopies. (bii) The average distance between all peptides within a family D_f is the arithmetic mean of all individual pairwise distances. It contains all pairwise distances between orthocopies and all paracopies. (biii) The net distance D_np between paracopies is similar to D_f after subtraction of D_o. It does not contain the pairwise distances between each set of orthocopies.

Figure 2

Direct peptide profiling of the ring gland of different Drosophila species. (ai-di) Mass range 900-1,600 Da. The protonated mass of AKH is not visible, but the Na⁺ and K⁺ adducts are prominent. (aii-dii) Mass range 2,050-2,250 Da. Only one mass peak corresponding to CPPB is visible.

Figure 3

Direct peptide profiling of the dorsal neural sheath of different Drosophila species. (ai-di) Thoracic portion containing FMRFa-like peptides. Note that peak intensity corresponds with isocopy number in the FMRFa prepropeptide. Small peaks corresponding to CAPA peptides from overlapping Va neurites are visible in D. virilis and D. mojavensis. (aii-dii) Abdominal portion containing CAPA peptides. Peaks corresponding to drosokinin and IPNa in D. pseudoobscura and D. sechellia represent contaminations with ganglionic neurites or the segmental nerve.

Figure 4

Direct peptide profiling of tracheal preparations containing the peritracheal cells of different Drosophila species. Peaks corresponding to the [M+H]⁺ or [M+Na]⁺ adducts of the two ETHs are visible besides the typical and possibly non-peptidergic tracheal peaks [22].

Figure 5

Plot of the average distance between orthocopies and ortholog spacers. Each data point represents the average amino acid distance D_so between orthocopies or ortholog spacer regions. With the exception of FMRFa-7, the peptide orthocopy distances have values below 0.3 and do not follow a Poisson distribution as is seen for the spacers.

Figure 6

Plot of the average distance between all paracopies in a family. Each data point represents the average amino acid distance D_af between all paracopies of a peptide family for each amino acid position throughout the species as outlined in (a). (b-d) The mean ± standard deviation of the data (b) for single copy peptides (c) and multiple copy peptide families (d). Paracopy number is color-coded in (d): black, 2; blue, 3; red, 4; green, 5; purple, 6; and brown, 10-17. The asterisks indicate a significant difference between multiple and single copy peptides.

Figure 7

Plot of the average distance between orthocopies for each amino acid position. Each data point represents the average amino acid distance D_ao between the orthocopies for each amino acid position throughout the species as outlined in (a). (b) The D_ao for multiple copy peptide families. (c) The mean D_ao ± standard deviation for single (black) and multiple (red) copy peptide families (see Figure 6c). The different shapes code for paracopy numbers: filled square, 1; filled triangle, 2; inverted filled triangle, 3; filled diamond, 4; filled circle, 5; open square, 6; open triangle, 7; open triangle, 8; open diamond, 9; open circle, 10; cross, 11; plus sign, 12; asterisk, 13.

Figure 8

Plot of the net distance between paracopies for each amino acid position. (a) Each data point represents the average net amino acid distance D_anp ± standard deviation between the paracopies for each amino acid position throughout the species. (b) The mean ± standard deviation of the data for multiple copy peptides compared to the mean D_ao ± standard deviation of single copy peptides (see Figure 6c). The asterisks indicate a significant difference between multiple and single copy peptides.

Figure 9

Plot of the net distance between paracopies for each amino acid position. Each data point represents the average net amino acid distance D_anp ± standard deviation between the paracopies for each amino acid position throughout the species for peptide families with one (open black squares) or two (closed red triangles) known receptors.