In the present study we investigated and analyzed the expression profile of a defensin gene in L. longipalpis developmental stages, adult females infected orally with Gram-positive or negative bacteria and L. mexicana, or injected with E. coli or L. mexicana.
The L. longipalpis LlDef1 defensin gene contains two exons (134 and 172 nt respectively) interspersed with a 63 nt intron. The presence of six cysteines at positions 52, 57, 61, 71, 77 and 79 on the predicted amino acid sequence, with the potential to create three disulfide bonds, characterizes a defensin signature sequence [Pfam 01097]. We also sequenced 511 nt of the LlDef1 5'UTR and the analysis revealed that this gene is potentially under the control of at least two immune-related transcription factors: caudal and dorsal. Caudal encodes a DNA-binding nuclear transcription factor that plays a crucial role during development and innate immune response in Drosophila. In Drosophila, Dorsal has its nuclear localization enhanced upon microbial challenge, interacting with Pelle, Tube, and Cactus during Toll activation to translocate and bind to NFκB-related sequences of AMP genes inside the nucleus . The phylogenetic analysis showed that LlDef1 is similar to defensin sequences from other nematoceran diptera, being closely related to a P. duboscqi defensin .
High transcription levels were detected in non-feeding L. longipalpis L4 larvae and pupae. In Anopheles gambiae, defensin expression was detected in non-challenged third and fourth instar larvae and pupae, reaching high expression levels after E. coli injections . A Drosophila defensin was detected in third instar larvae only after bacterial challenge, although expression was detected in non-challenged pupae , similarly to what was observed in L. longipalpis and A. gambiae. No previous study explored the immune response in naturally feeding versus non-feeding larvae in the Diptera group. Since transstadial passage of bacteria from larvae to pupae and adult flies has been already reported for sandflies [12, 48, 49], L. longipalpis non-feeding L4 and pupae may trigger defensin expression to control and select gut microbiota during late L4 through pupation to emerged adult.
L. longipalpis were orally exposed to five different Gram-positive and Gram-negative bacteria. Defensin expression was found to increase with time upon infection with the Gram-negative E. coli, Ochrobactrum sp. and S. marcescens. Ochrobactrum sp. is acquired by P. duboscqi larvae from the environment  and it is plausible to consider that it is recognized by the insect immune system as a foreign antigen as much as E. coli. S. marcescens is entomopathogenic and was shown to trigger the L. longipalpis immune system through ROS increase . Interestingly, infection with the Gram-negative P. agglomerans showed an initial decrease of defensin expression and a very constant level of expression over time matching control levels. This may be due to the fact that this bacterium, commonly found in Anopheles stephensi gut, is not pathogenic  and may not be recognized as a hazard by L. longipalpis. Insect defensins are known to be active mainly against Gram-positive bacteria [23, 26, 52]. Accordingly, flies exposed to the Gram-positive M. luteus showed a sharp up-regulation of defensin mRNA during the early stages of infection (24 h post-feeding). Although defensin gene expression dropped considerably during the following 3 days, transcription was still significantly increased at 48 and 72 h post-feeding in comparison to controls.
These results suggest that sand flies are capable of mounting different innate immune responses against distinct bacterial species. A previous study that used the synergistic effects of lysozyme with antibacterial peptides revealed that L. longipalpis can successfully mount a humoral response against bacterial challenge and this response specifically discriminates between M. luteus and E. coli. Although an increase of expression of a 4 kDa peptide was detected in the hemolymph of both M. luteus and E. coli-injected L. longipalpis in comparison to mock-injected controls, an unknown 33 kDa peptide could be detected in the hemolymph of the sand fly only when insects were challenged with M. luteus but not with E. coli. These findings, and our present results, suggest that specific and discriminating immune responses are probably produced against the Gram-positive and Gram-negative bacteria in L. longipalpis.
At 48 h after artificial blood feeding and artificial infection with L. mexicana adult female sand flies showed a dramatic increase of defensin expression that slowly decreased over time. This initial increase in defensin expression may be a response to the proliferation of sand fly gut microbiota caused by the ingestion of a nutrient-rich blood meal as it was seen in P. duboscqi and Aedes aegypti[54, 55]. Interestingly, a defensin down regulation was observed starting at 72 h after Leishmania infection, reaching statistical significance at 144 h in comparison to blood-fed controls. Late infections were previously correlated with high numbers of Leishmania promastigotes within the sand fly gut . Our present results indicate that high parasite number is correlated to low defensin expression. One explanation of this may be due to low levels of defensin expression at later time points after bloodfeeding, allowing for parasite survival and multiplication. On the other hand, if the defensin expression response is primarily towards bacterial molecular factors then the significant fall in defensin expression may be due to suppression of the gut bacterial population, via a competitive exclusion effect, in the presence of Leishmania.
A different transcription profile was reported in P. duboscqi infected with Leishmania major, where low levels of defensin expression were observed in the first day of infection whereas expression was strongly induced at four days after the Leishmania infection . It is plausible that different phlebotomine sand flies and different Leishmania species may trigger diverse immune responses. This has been reported in mosquitoes, where different immune-related genes were modulated upon infection with various Plasmodium species [57, 58].
Expression of defensin in L. longipalpis after L. mexicana or E. coli intra-thoracic injection was also investigated. Pricked and LB medium-injected sand flies showed an increase in defensin expression in comparison to uninjected sugar-fed controls at 24 and 48 h post-injection. These results indicate that trauma by injection was sufficient to activate the innate immunity and induce defensin transcription in L. longipalpis. Cuticle pricking and mock-injection of dsRNA into the sand flies’ hemocoel was shown to reduce the number of L. mexicana promastigotes within the midgut of L. longipalpis, possibly by nonspecific activation of the IMD pathway . In A. aegypti, the injection of sterile saline induced the mosquito immune response and produced low but detectable levels of defensin mRNA . Previous work in L. longipalpis showed that antimicrobial activity increased in sham-injected insects when compared to non-injected controls . Similarly, our results demonstrated that control L. longipalpis microinjected with medium showed a significant increase in defensin expression at 24 h in comparison to controls, which was maintained until 48 h post-injection. In Drosophila Toll and IMD pathways can regulate different AMPs  and both can act synergistically . This much is not yet explored in L. longipalpis.
Nimmo et al.  observed a significant increase in L. longipalpis humoral response against E. coli or M. luteus estimated by inhibition zone assays using hemolymph from bacteria-challenged insects. In addition, P. duboscqi inoculated with Erwinia carotovora showed higher defensin expression in comparison to naive insects and bacteria-fed sand flies . Although in line with results obtained for P. duboscqi, our results show a much subtler defensin expression in L. longipalpis upon bacterial injection. Similar results were obtained in A. aegypti inoculated with E. coli and M. luteus which showed 3 times higher levels of defensin peptides in their hemolymph when compared to sterile saline-injected insects . These results confirm that mosquitoes and sand flies can mount an immune response through defensin expression upon bacterial challenge in their hemolymph.
L. longipalpis injected with L. mexicana showed a significant increase of defensin expression at 72 h post infection. Although the presence of Leishmania in the hemolymph does not occur in nature, it is possible that the ectopic presence of parasites within the hemolymph induced an immune response. It has been shown that Drosophila is capable of producing an immune response against injected Plasmodium gallinaceum oocytes . Defensin reduction at 24 and 48 h after Leishmania injection may be a counterbalance caused by activation of the IMD pathway triggering other AMP, but not DefLl1. Later, at 72 h, L. longipalpis is able to express high levels of defensin. To our knowledge, this is the first report of an immune response in sandflies after parasite injection. Investigation of other Toll or IMD related AMPs could address and clarify this hypothesis related to the sandflies immune response to Leishmania injection in hemolymph, but none has been described up to date.