Constitutive activation of cellular immunity underlies the evolution of resistance to infection in Drosophila

Abstract

Organisms rely on inducible and constitutive immune defences to combat infection. Constitutive immunity enables a rapid response to infection but may carry a cost for uninfected individuals, leading to the prediction that it will be favoured when infection rates are high. When we exposed populations of Drosophila melanogaster to intense parasitism by the parasitoid wasp Leptopilina boulardi, they evolved resistance by developing a more reactive cellular immune response. Using single-cell RNA sequencing, we found that immune-inducible genes had become constitutively upregulated. This was the result of resistant larvae differentiating precursors of specialized immune cells called lamellocytes that were previously only produced after infection. Therefore, populations evolved resistance by genetically hard-wiring the first steps of an induced immune response to become constitutive.

###This manuscript is in revision at eLife

The decision letter after peer review, sent to the authors on June 23, 2020, follows.

Summary

It has been previously shown that resistance to parasitoid wasps can emerge upon selection in wild-type Drosophila populations, and that this increased resistance correlates with a higher number of hemocytes. This paper combined experimental evolution and single cell transcriptomics to show that increased resistance to parasitoids upon several rounds of selection is caused by the presence of a differentiated subset of hemocytes (pre-lamellocyte) in the unchallenged state, which is usually found only upon wasp infestation. This led the authors to conclude that intense pathogen pressures can shift the immune system from inducible to constitutive, consistent with a theoretical framework indicating that elevated and constant pathogen pressure should lead to the emergence of constitutive defense. The approach is interesting, the paper well-written and the notion tested interesting. An important concern is the degree of advance over previous studies. Initial papers investigating how selection increases resistance to wasps have already shown that this was linked to an increase in hemocyte number. In a certain sense, this could be considered as a demonstration of a change from inducible to constitutive defense, although the emphasis of these papers was not on this point. In addition, the current work provides so far only limited information on this specific population of pre-lamellocytes.

Essential Revisions

Analysis of single cell RNA-seq (Figure 3). Several RNA-seq papers have been published and it is important that the authors better relate their hemocyte clusters to other scRNA-seq datasets using the nomenclature of some of these papers. Would their data be deposited in a database? It would also be great to better describe the transcriptional profile, and not only focus on two genes, Attila and PPO3.
Discrepancy between the transcriptional and morphological changes in the hemocytes.

2a) Earlier studies, both on hemocyte flow cytometry and in other scRNA-seq experiments (as cited in the manuscript) revealed that the transdifferentiation into lamellocytes is a dynamic / continuous process, which may derive from several hemocyte lineages and from different hematopoietic organs. The authors here showed a discrepancy in the transcriptional and morphological changes in the hemocytes, and revealed that the plasmatocyte lineage was already starting the resemble the lamellocytes (in gene expression), without needing the induction by infection. Yet, they were not yet fully differentiated hemocytes based on morphology, and still needed infection to reach that stage. Therefore, the conclusion that the selected lines had "hard-wired" the inducible response into a constitutive response is not fully warranted (they do not fully differentiate, but proceed partially towards that state). Also, the differentiation of lamellocytes is fully attributed to originate from lymph glands and as originating from the plasmatocytes, while different organs and hemocyte lineages appear to contribute to the population of lamellocytes. The reviewer feel that all these aspects should be further explored and would deserve some mentioning in the discussion.

2b) Along this line, the authors could do a better job in characterizing the hemocyte populations of the evolved lines using available antibody, cooking and other melanization assays, phalloidin treatment...

2c) Third instar hemocytes are found in the sessile state, in circulation or in the lymph gland. It could not be excluded that some of the changes they observed relate more to changes in hemocyte localization rather than differentiation. According to the material and methods, the authors has collected only the circulating hemocytes in the unchallenged state as they did not vortex larvae. It is very important to better compare the lymph gland, sessile and circulating compartments of the evolved and the non-evolved lines. This can be done by using various staining methods. The paper is written in such a way that selection acted only on circulating hemocytes but it could also act on hemocytes localization (decrease sessility), lymph gland maturation....

Gene expression was measured in circulating hemocytes at 48h after infection.

The authors measured gene expression in circulating hemocytes, 48h after infection, at which stage hemocyte proliferation, lamellocyte differentiation and parasitoid encapsulation is already well underway. The induction of the critical two processes, hemocyte proliferation and lamellocyte differentiation, may not be fully detectable from gene expression of only the circulating hemocytes themselves at this late stage of the immune response. Clearly, the authors do show that differentiation from circulating plasmatocytes can be detected, using pseudotime, and also revealed changes in gene expression in uninfected selected larvae. Yet, how induction in the lymph glands or sessile clusters has changed by experimental evolution, and whether the inducible response had indeed proceeded towards a constitutive response, requires further investigation along a wider time course (e.g. during early larval development) and perhaps in different tissues (e.g. lymph glands). If the author cannot address this, this aspect would need some discussion.

The changes in gene expression after selection can be presented clearer.

The description of these results (from L111 onwards), and Figure 2, difficult to read and understand, while it is key to the claim that the inducible response has become hardwired into a constitutive response. In the text it starts out with saying that "data was pooled to investigate global changes" (L116-117), but then it refers in Figure 2 to the x-axis which only provides the data for the control lines. This figure 2 is difficult to grasp, as the strong positive correlation in a) means something different (i.e. stronger constitutive response) than the very similar positive correlation in b) (weaker induced response), while c shows that control and selected larvae respond the same to infection. Is there a better way to tease apart these patterns in a figure, and to explain them in the text? Also, the data is all expressed in log2 fold changes (relative to non-infected control line individuals?). Also, for a subset of approximately 170 genes, the authors showed that the increase in expression had already started without the infection in the selection lines. Do the functional annotations of these genes reveal anything of interest for hemocyte proliferation and the differentiation towards lamellocytes?

Other studies came to partially contrasting, partially similar conclusions.

Transcriptomics on whole larvae after experimental evolution for high parasitism was done for Drosophila, using a different parasitoid species. In this study, they also found the typical increased density of hemocytes in Drosophila selected for increased parasitoid resistance, without being infected. However, contrary to the authors, this study concluded this increase in hemocytes could not be attribute to a pre-activation of the immune response. Additionally, the genes for hematopoiesis and for several effector genes showed opposite patterns to those that would explain the increased density of hemocytes in selected lines or for an pre-activation of the inducible response (Wertheim et al, 2011, Molecular Ecology). However, in line with the findings for the current study, whole-larvae RNAseq after parasitoid infection did not result in substantial gene expression differences between selected lines and control lines (Salazar et al, 2017, BMC Genomics), while substantial differences were reported in uninfected larvae of selection and control line larvae (Wertheim et al, 2011, Molecular Ecology). These whole-body transcriptomics experiments lacked the resolution to measure specifically what changed in hemocytes, but both studies indicate that much of the increased resistance after selection is likely caused by changes in constitutive immunity, not by increasing the acute/inducible immune response.

Another concern is related to the parasitoid species. Leptopilina boulardi is a parasitoid that relies partly on VLPs to overcome the host defense. This is not discussed, not even mentioned. Some older work (Fellowes et al 1999, Evolution), shows that, while resistance evolves readily against L. boulardi, populations resistant against L. boulardi are also cross-resistant to another Leptopilina species. The immune effectors studied in this manuscript are obviously playing a significant role, but how do the evolved flies cope with the VLPs? The paper would benefit from at least discussing this issue.
The selection of larvae for the single cell work warrants some clarification. According to figure 1b just under 50% of parasitoid resistant larvae showed an increased encapsulation response. This is presumably also related to the increase of expression of immune effectors. How is this accounted for in the single cell work? And if not, do you have any way to get an estimate of the variance in the response variables?

Read the original source

Constitutive activation of cellular immunity underlies the evolution of resistance to infection in Drosophila

This article has been Reviewed by the following groups

Listed in

Abstract

Article activity feed