Figure 1.
PCR amplification of full length hymenoptaecin gene and cDNA.
The PCR-products from gDNA (lane 1) and cDNA (lane 2) were separated on a 1.2% agarose gel alongside molecular size markers (lane M, GeneRuler 1 kb DNA Ladder, Fermentas) and analyzed with EtBr staining. The major bands correspond to the full length hymenoptaecin gene- (3356 bp lane 1) and cDNA-product (2536 bp, lane 2). The minor bands are technical artefacts with variable repeat numbers caused by the tandem repeats.
Figure 2.
Alignment of HLD (hymenoptaecin-like domain) and all HDs (hymenoptaecin domains) from the same C. floridanus hymenoptaecin multipeptide precursor protein.
Grey boxes indicate conserved residues. The insertion in the hymenoptaecin-like domain (top) is clearly visible.
Figure 3.
Structure of the C. floridanus hymenoptaecin gene locus.
(A) Schematic structure of the hymenoptaecin gene containing a single intron within the region coding for the hymenoptaecin propeptide. The deduced multipeptide precursor peptide consists of a signal-sequence (Pre, grey hatched box) and a pro-sequence (Pro, white hatched box), followed by a hymenoptaecin-like domain (HLD, light grey box) and six repeated hymenoptaecin domains (HD 1–6, dark grey boxes). The hymenoptaecin domains are flanked by the two putative processing sites EAEP (white boxes) and RR (black boxes). (B) The nucleotide and deduced amino acid sequence of a hymenoptaecin repeat unit are shown and the putative processing sites are boxed.
Figure 4.
Southern blot (A) with C. floridanus genomic DNA using a 32P-labelled hymenoptaecin fragment corresponding to one of the repeats as a probe.
Genomic DNA (35 µg per lane) was digested with EcoRI (lane 1) and with Bsu15I (lane 2), separated by gel electrophoresis and hybridized with the above mentioned DNA fragment. Northern blot (B) with total RNA of C. floridanus using a 32P-labelled cDNA fragment corresponding to the 5′end of the hymenoptaecin gene as a probe. Total RNA (25 µg per lane) was isolated from midgut and fat body of major workers injected with heat-killed Serratia marcescens (2×105 bacteria/ant) in the haemocoel (lane 1) or untreated animals (lane 2). The major band corresponds to the spliced mature transcript, while the minor band very likely is the unspliced precursor. The position of molecular size markers is indicated on the left side of each figure. All hybridizing bands have the expected molecular size.
Figure 5.
Antibacterial activity of 4.5 nmol recombinant Cfl-hym peptide (A) against E. coli D31.
Dialysis buffer alone (B) was applied as a negative control and 4 µg kanamycin (C) as a positive control.
Figure 6.
Schematic structure of the hymenoptaecin precursors from different hymenopteran species:
A) Apis mellifera (GenBank Acc. No.: NP_001011615) or Bombus ignitus (GenBank Acc. No.: ACA04900); B) Nasonia vitripennis: (GenBank Acc. No.: NP_001165829 XP_001607881); C) Harpegnathos saltator 1 (GenBank Acc. No.: EFN79831); D) Harpegnathos saltator 2 (GenBank Acc. No.: EFN79832); E) Camponotus floridanus (GenBank Acc. No.: HQ315784); F) Acromyrmex echinatior (hymenoptaecin multipeptide precursor deduced from genome draft). The various domains are marked as follows: signal-sequence (grey hatched box), pro-sequence (white hatched box), hymenoptaecin-like domain (HLD, light grey box), hymenoptaecin domains (HD 1–6, dark grey boxes), proline-rich AMP-like peptide (AMP 1–2, white dotted boxes). The hymenoptaecin domains are flanked by the putative processing sites EAEP (EANP for Harpegnathos) (white box) and RR (or RxxR) (black box).
Figure 7.
Phylogenetic analysis of hymenoptaecin domains from different ant species.
Shown is the unrooted tree of the single domains of the hymenoptaecins of the ant species, N. vitripennis, and A. mellifera. The proteins were cleaved at the sites predicted by ProP, followed by the alignment by translatorX. The tree was reconstructred by PhyML with a GTR+I+G+F model with 100 bootstrap replicates. The domains of the species with a complete hymenoptaecin protein form clades and are named as groups according there genus name and are indicated by their grey background. The domains which are outside these groups result from missing data from the predicted genes. The gene models are incomplete due to long N-stretches in the genomic sequences based on the scaffolding process. Nevertheless, the distinct groups formed by the complete proteins suggest an intra-species mechanism for the accumulation of the single domains.
Figure 8.
Phylogenetic analysis of defensins from different ant species.
All defensin sequences were aligned by MUSCLE and a BioNJ-tree with 100 bootstrap replicates was calculated. Branches with a bootstrap support below 40 were removed. Other bootstrap values are indicated. The genes for the gene tree were generated using the gene prediction pipeline maker and hand curated. The gene tree was rooted at the defensin from Ixodes scapularis (GenBank Acc. No.: XP_002436104.1). The A. mellifera defensin-1 forms a clade with proteins formerly described as defensin-2, which gives a first indication that they could be renamed accordingly. However, some of the proteins form a clade with the A. mellifera defensin-2. Moreover, two species S. invicta and C. floridanus, own both defensins. Therefore, we suggest a duplication event at the LCA of A. mellifera and the ant species.
Figure 9.
Schematic structure of the defensin genes from C. floridanus.
(A) The gene encoding defensin-1 (GenBank Acc. No.: JN989495) is composed of three exons and two introns. The first intron is located within the region coding for the propeptide and the second is located within the region coding for the mature defensin peptide. (B) The gene encoding defensin-2 (GenBank Acc. No.: JQ693412) contains only one intron, which is also located within the propeptide coding region. Both deduced precursor peptides consist of a signal-sequence (Pre, grey hatched box) and a pro-sequence (Pro, white hatched box), followed by the mature defensin peptide (Def, grey box).
Figure 10.
Reconciled species tree of ant defensins.
This tree was generated by Notung. The ant species L. niger (GenBank Acc. No.: ACB46517.1), Myrmica scabrinodis (GenBank Acc. No.: ACB46524.1), and F. aquilonia (GenBank Acc. No.: Q5BU36.1) are examples for the sequences generated by [12], [27]. For these species no whole genome is available, which is indicated by the question mark behind the species. The gain and loss events are indicated by green and red boxes. The gain event at the LCA of A. mellifera and the ant species generated the defensin-1/2 peptide. The LCA of L. niger, C. floridanus and F. aquilonia had an additional duplication event of its defensin-1 peptide, but the species lost either their defensin-1a or their defensin-1b.