Serotonin modulates insect hemocyte phagocytosis via two different serotonin receptors

Essential revisions:

The work done in the caterpillar is well done and nicely controlled. There are similar results already published for other insect species (Kim GS, Nalini M, Kim Y, Lee DW, Arch Insect Biochem Physiol. 2009; Kim GS and Kim Y, J Insect Physiol. 2010) and these need to be acknowledged up front in the introductory section with a sentence justifying why this study is still relevant. The reviewers found that the butterfly experiments are more detailed than in the older literature and that the extension to Drosophila opens up new possibilities for extending this research. Unfortunately, it is the fruit fly results that are not substantial enough to warrant an eLife publication. If some more characterization is presented in terms of phenotype and mechanisms, then this will achieve that goal. The authors are the best judges of the experiments that will further consolidate the fruit fly story. The manuscript will benefit from such addition. Of course, if necessary changes will take longer than 2 months, the authors could resubmit an expanded manuscript. The following major questions are provided as examples of questions that need to be addressed. 1) Is the receptor 5-HT_1B required autonomously in hemocytes? Does hemocyte-autonomous RNAi of 5-HT_1B affect the phagocytosis response, CFU after infection, and overall survival after infection? The authors could use Hml∆-GAL4 and ideally two independent UAS-RNAi lines. For suitable readouts of CFU and phagocytosis see below.

We thank the reviewers for raising this point. As the reviewers suggested, we have used Hml∆-GAL4 and two independent UAS-5-HT_1B RNAi lines to address this question. As show in Figure 7, hemocyte-specific knockdown of 5-HT_1B decreased hemocyte phagocytic capacity, increased the bacterial load and susceptibility to S. aureus.

2) Does 5-HT_1B indeed mediate phagocytic behavior in hemocytes and bacterial clearance as measured by CFU (or other hemocyte response/s)? Regarding the quantification of CFU after bacterial infection, it is unclear if bacterial strains with a resistance gene were used? This is not mentioned in the methods and needs to be clarified. Use of bacteria with a resistance marker is needed so injected bacteria can be distinguished from the naturally occurring bacterial load.

To carefully evaluate phagocytosis the authors should, at a minimum, use the trypan blue phagocytosis assay as described by Elrod-Erickson et al. Curr. Biol. 2000.

We appreciate the reviewers’ constructive comments and suggestions. We are sorry for the unclear information regarding the bacterial strains we used. The bacterial strain S. aureus is resistant to tetracycline and the LB plates contain tetracycline. We have added the information in the Methods.

The bacteria we used to evaluate phagocytosis are E. coli and S. aureus labeled with pHrodo, a dye that fluoresces in the acidic environment of a mature phagosome upon fusion with lysosomes. This method reduces signal variability and improves assay reproducibility since wash steps and quencher dyes are not needed. It is now widely used in mammals (Fiala M et al., Proc Natl Acad Sci USA. 104:12849-12854, 2007; Berger SB et al., Nat Immunol.11:920-927,2010; Neaga A et al., J Immunol Methods. 2011) and insects including Drosophila (Cuttell L et al., Cell 135, 524-534, 2008; Ulvila J et al., J Leukoc Biol. 89(5): 649-59, 2011; Stone EFet al., PLoS Pathog. 8(1): e1002445, 2012; Lombardo Fet al., PLoS Pathog. 9(1): e1003145, 2013; Regan JC et al., PLoS Pathog. 9(10): e1003720, 2013; Garg A and Wu LP. Cell Microbiol. 16(2): 296-310, 2014). Besides, we also tested the classical trypan blue phagocytosis assay described by Elrod-Erickson et al. (Curr. Biol. 2000) to check the hemocyte phagocytic difference between 5-HT_1B null flies and control flies. The results are in Author response image 1 and similar to that by the pHrodo dye assays (Figure 6D). Thus, we still used the pHrodo dye in phagocytosis assays of this paper.

Quantification of in vivo phagocytosis of E. coli.

3) Does infection or LPS challenge indeed trigger an increased phagocytic hemocyte response through 5-HT_1B?

Does hemocyte-specific kd of 5-HT_1B dampen phagocytosis in general, or is there indeed a lack of inducibility, e.g. after LPS challenge? Comparison of injecting fluorescent beads -/+LPS would allow to address this. A wt control should be used in parallel to confirm inducibility by the immune challenge here. The experiment could also be performed by injecting fluorescent beads with -/+ 5-HT.

We thank the reviewers for this valuable comment. Following the reviewers’ suggestion, we used latex beads with -/+ 5-HT to address this question. Flies injected with PBS all show similar phagocytic capacity. However, when flies are injected with 5-HT, the flies expressing 5-HT_1B RNAi in blood cells take up significantly less latex beads than the control (Figure 7E). These results indicate that hemocyte-specific knockdown of 5-HT_1B flies do not have a more limited phagocytic capacity, but instead become unable to enhance phagocytosis upon exposure to 5-HT.

4) Do 5-HT levels increase upon bacterial infection in Drosophila? It would be easy to perform a hemolymph ELISA (as in Figure 1D) to see if 5-HT levels increase upon infection in flies.

We thank the reviewers for this suggestion. However, we have tried many times with different protocols but still failed to culture hemocytes in the medium without contamination for a longer time (>30 min). The small body size makes the Drosophila larvae extremely difficult to isolate clean hemolymph like we did on the caterpillar.

5) Both 5-HT_1B and 5-HT_2B seem to be expressed in hemocytes of uninfected adult Drosophila (Figure 6A). But the entire focus of the infection study phenotype is limited to the 1B mutant. What effects, if any, does inhibition/reduction of 2B have on phagocytosis and the lifespan of immune challenged animals? Would it be the same as in the caterpillar? If there is no mutant for 2B, the authors could minimally check levels of 2B expression in the hemocytes of infected animals, and perhaps one would see a down regulation similar to that in the caterpillar. There are MiMIC and TRiP lines available at BDSC, and minimally, 2B RNAi lines could be used to determine if this leads to increased phagocytosis and increased survival when challenged with bacterial infection.

Following these comments, we have used Hml∆-GAL4 and two independent UAS-5-HT_2B RNAi lines to investigate the role of 5-HT_2B in phagocytosis and the lifespan of immune challenged animals. As Figure 8 shows, the reduced levels of 5-HT_2B in hemocyte led to the defect in phagocytosis of E. coli and S. aureus (Figure 8A-D) and that was associated with higher bacterial load (Figure 8I) and increased susceptibility to S. aureus (Figure 8E-G) in the flies. The results suggest that activation of 5-HT_2B may promote phagocytosis in Drosophila, different from the results in the caterpillar.