Advertisement

  • Loading metrics

Open Access

Peer-reviewed

Research Article

Gut bacteria induce heterologous immune priming in Rhodnius prolixus encompassing both humoral and cellular immune responses

  • Carissa A. Gilliland,

    Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Visualization, Writing – original draft, Writing – review & editing

    Current address: Department of Entomology, The University of California, Riverside, California, United States of America

    Affiliation Entomology Department, The University of Georgia, Athens, Georgia, United States of America

    ⨯
  • Loretta Mugo-Kamiri,

    Roles Investigation, Methodology

    Affiliation Entomology Department, The University of Georgia, Athens, Georgia, United States of America

    ⨯
  • Sara Martin,

    Roles Investigation, Methodology

    Affiliation Department of Genetics, The University of Georgia, Athens, Georgia, United States of America

    ⨯
  • Kevin J. Vogel

    Roles Conceptualization, Data curation, Formal analysis, Funding acquisition, Project administration, Resources, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing

    kjvogel@uga.edu

    Affiliation Entomology Department, The University of Georgia, Athens, Georgia, United States of America

    ⨯

Abstract

Insects lack the adaptive, antibody mediated responses of vertebrates, yet they possess a robust innate immune system capable of defending themselves against pathogens. Immune priming has been observed in multiple insect species, wherein exposure to a pathogen provides protection against subsequent infections by the pathogen. Heterologous immune priming has also been described, where presence of one bacterial species provides protection against another. We determined that Rhodococcus rhodnii, a gut symbiont of the kissing bug Rhodnius prolixus, induces strong heterologous immune priming, while axenic bugs lacking gut bacteria are highly susceptible to pathogens. Commensal Escherichia coli provides less robust protection. R. rhodnii must be alive within the insect as dead bacteria do not stimulate immune priming and pathogen resistance. Removal of R. rhodnii from the gut reduces resistance to pathogens while restoring it to axenic bugs improves pathogen resistance, though not completely. Unlike most other examples of symbiont-mediated immune priming, we find no evidence that R. rhodnii ever leaves the gut, despite activating a potent immune response in the hemocoel and fat body. R. rhodnii and E. coli activate both the IMD and Toll pathways indicating cross-activation of the pathways, while silencing of either pathway leads to a loss of the protective effect. Several antimicrobial peptides are induced in the fat body by presence of gut bacteria. When E. coli is in the gut, expression of antimicrobial peptides is often higher than when R. rhodnii is present, while R. rhodnii induces proliferation of hemocytes and induces a stronger melanization response than E. coli. Hemolymph from R. rhodnii bugs has a greater ability to convert the melanin precursor DOPA to melanization products than axenic or E. coli-harboring bugs. These results demonstrate that R. rhodnii’s benefits to its host extend beyond nutritional provisioning, playing an important role in the host immune system.

Author summary

Insects often form beneficial relationships with bacteria allowing them to eat nutritionally deficient diets. In insects that only consume blood, symbionts are necessary to provide B vitamins absent in the host diet. There is a growing appreciation that in some of these symbiotic associations, the bacteria provide services beyond nutrition. We show that in kissing bugs, which feed exclusively on vertebrate blood and require bacterial symbionts for development, these symbiotic bacteria are important in activating the insect immune system. Insects with no gut bacteria are highly susceptible to infection and cannot mount an effective immune response. The bacteria reside exclusively in the insect gut yet protect against infections in the rest of the insect’s body. The bacteria must be alive to prime the immune system, and the response is dependent on the species of bacteria in the gut, with symbiotic bacteria providing stronger protection against infection and inducing a broader array of immune responses than commensal bacteria. This study expands our understanding of the role of beneficial bacteria in insect immunity and demonstrates that immune systems differ between major groups of insects.

Citation: Gilliland CA, Mugo-Kamiri L, Martin S, Vogel KJ (2025) Gut bacteria induce heterologous immune priming in Rhodnius prolixus encompassing both humoral and cellular immune responses. PLoS Pathog 21(12): e1012947. https://doi.org/10.1371/journal.ppat.1012947

Editor: Pedro F. Vale, The University of Edinburgh, UNITED KINGDOM OF GREAT BRITAIN AND NORTHERN IRELAND

Received: January 29, 2025; Accepted: November 19, 2025; Published: December 1, 2025

Copyright: © 2025 Gilliland et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: Data files are included in the supplemental information.

Funding: This work was funded by National Science Foundation (https://www.nsf.gov/) award number 2239595 to KJV. The funders did not play any role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Insect immunity is often presented as a set of discrete innate immune pathways targeting broad classes of pathogens. The Toll pathway acts against Gram-positive bacteria and fungi, the IMD pathway against Gram-negative bacteria, and the JAK/STAT pathway against viral infections [1]. These pathways are activated by microbe associated molecular pattern (MAMPs) and trigger humoral and cellular immune defenses to limit pathogen damage and proliferation. These insights have been gained primarily through studies undertaken in a small number of model systems, nearly all of which are holometabolous. Among the paurometabolous orders (insects that transition from a nymph to an adult without undergoing true metamorphosis), Hemiptera is the most diverse and arguably best studied, yet our understanding of the immune function in this group is far behind Drosophila and other holometabolous insects. Morphological, genomic, and experimental studies suggest that the immune systems of hemipterans are surprisingly different from holometabolous insects [2–4]. Morphologically, the hemipteran gut lacks the chitinous peritrophic matrix which forms a barrier between gut microbes and the midgut epithelium in most insect groups. In its place is the perimicrovillar membrane, a lipoprotein-based structure that has some similarities to the peritrophic matrix but is not entirely analogous [5]. The presence of the perimicrovillar membrane may promote different interactions between gut microbes and the host immune response relative to insects with a peritrophic matrix.

The first genome sequence of a hemipteran, the pea aphid Acyrthosiphon pisum, revealed that it lacks several genes in the IMD pathway thought to be essential for its function [6]. Subsequent hemipteran genomes have shown similar losses of core IMD genes, indicating these losses occurred early in the evolution of Hemiptera [4,7–9]. Functionally these two immune pathways are not as compartmentalized as in Drosophila, suggesting that these gene losses have significant consequences for the activation of the hemipteran immune system [4,10]. One explanation for the loss of these immune genes is that they are related to the frequent association of hemipterans with bacterial symbionts. Many hemipterans feed on nutritionally imbalanced diets and rely on microbial partners to supply the missing nutrients [11] and the need to support these symbionts may have led to a loss of immune genes that might activate host responses against these necessary microbes [12]. Despite the absence of canonical pathway components, many hemipterans can mount robust and effective immune responses [4,7,10,13–15].

While symbionts may have contributed to the loss of immune genes in hemipterans, they are increasingly appreciated for their role in host immune function in hemipterans and other insects. For example in the bean bug Riptortus pedestris, environmentally-acquired Burkholderia escape the midgut and subsequently prime the host immune response by stimulating AMP expression, resulting in increased survival following heterologous bacterial challenges [16,17]. This phenomenon is also seen in holometabolous insects. In honeybees (Apis mellifera), the presence of Snodgrassella alvi elevates production of the antimicrobial peptide (AMP) Apidaecin in the bee gut and provides protection against hemocoelic infection with E. coli [18]. In Glossina sp., when larvae are reared without their symbiont Wigglesworthia glossinidia, the resulting adults have impaired immune function, including reduced melanization, lower expression of AMPs, and reduced hemocyte titer [19]. Subsequent studies have demonstrated that Wigglesworthia are essential for hematopoiesis during Glossina development [20].

Rhodnius prolixus, a triatomine kissing bug (Hemiptera: Reduviidae), is a vector of Trypanosoma cruzi, a stercorarian parasite that causes Chagas disease in humans. R. prolixus harbors symbiotic Gram-positive Actinobacteria, Rhodococcus rhodnii, within the midgut which are essential for successful insect development [21–23]. A previous study demonstrated that gut bacteria can influence the host immune response to enteric pathogens, but surprisingly found that the immune system appeared to be minimally influenced by the presence of R. rhodnii [24]. That study employed antibiotic clearing of established communities of microbes from R. prolixus followed by recolonization with target microbes including R. rhodnii. This approach may have failed to clear non-target microbes from the gut, and systemic effects of antibiotics may have also influenced microbiota reestablishment and observed changes in immune function [25].

We previously developed a system for producing axenic (bacteria-free) nymphs without the use of antibiotics, which we leveraged to manipulate the host gut microbiome by re-introducing bacteria via a blood meal using artificial membrane feeders [22]. This experimental platform allows us to interrogate the effects of the symbiont on host physiology. We used this approach to ask if R. rhodnii had a significant impact on the R. prolixus immune system, whether other bacteria have similar effects to R. rhodnii, and what elements of the host immune system may be influenced by presence or absence of the symbiont. We show that R. rhodnii specifically is necessary for proper host immune function and involves both the Toll and IMD pathways. The symbiont activates both humoral and cellular immune responses in distal host tissues without leaving the insect gut, ultimately providing robust protection against hemocoel infection. These results expand our understanding of how symbionts contribute to host immune function and provide potential avenues to exploit towards limiting transmission of T. cruzi.

Results

The presence of bacteria protects R. prolixus against bacterial infection

To investigate the effects of gut bacteria on host immune function, we removed all bacteria through surface sterilization of eggs using our previously described protocol [22]. Bugs were either reared in axenic conditions (RproAxn), or experimentally infected through a blood meal with E. coli MG1655 (RproEc), or their symbiont R. rhodnii ATCC 35071 (RproRr). Twenty four unfed 4th instar nymphs from each gnotobiotic condition were injected with 106 CFUs of E. coli, R. rhodnii, the Gram-positive Micrococcus luteus, or sterile saline into their thorax and survival after infection was monitored. Almost all RproAxn individuals died within 5 days after injection with either E. coli (9% survival, Fig 1A) or M. luteus (17% survival, Fig 1A). In contrast, RproRr bugs showed significantly higher survival after E. coli or M. luteus infection (79% and 68% survival, respectively) than RproAxn (p < 0.0001 for both pathogens, log-rank test). RproEc bugs had significantly higher survival rates compared to RproAxn individuals for both challenges (52% survival, p < 0.0001, and 50% survival, p = 0.003, respectively, log-rank test, Fig 1A) and lower survival relative to RproRr challenged with E. coli (p = 0.018, log-rank test) but not when challenged with M. luteus (p = 0.07, log-rank test). To test if the mortality we saw was due to bacterial infection or due to wounding alone, we also injected RproAxn, RproEc, and RproRr insects with sterile saline. All insects subject to sterile saline injection had high survival and were not significantly different from each other (p > 0.05, log-rank test, Fig 1A). We also tested whether lower infectious doses elicited the same pattern. We challenged RproRr and RproAxn with 102 and 104 CFU of E. coli via the same injection protocol. We found no significant difference in survival between RproRr and RproAxn at these lower doses which may reflect inoculations more frequently experienced by R. prolixus in nature (S1 Fig).

thumbnail
Download:
Fig 1. Gut bacteria promote immune priming against bacterial infection in R. prolixus.

(A) Survival curves of RproAxn, RproEc, and RproRr after injection with 106 CFU of E. coli, M. luteus, or sterile saline into their hemocoel. RproRr and RproEc had significantly higher survival than RproAxn regardless of the bacteria used for challenge (p < 0.0001, log-rank test), while there was no difference in survival of bugs injected with sterile saline (p ≥ 0.15, log-rank test) indicating that the presence of gut microbes plays a critical role in insect defense against pathogens. When challenged with E. coli there was a significant difference in survival (26.8%) between RproRr and RproEc (p = 0.018, log-rank test). The difference in survival between RproRr and RproEc was smaller (18%) and approached but did not reach significance when challenged with M. luteus (p = 0.072, log-rank test). Lines connected by different letters are significantly different (p < 0.05, log-rank test). (B) Gnotobiotic R. prolixus limit the growth of E. coli in the hemocoel. Boxplots of E. coli CFUs in R. prolixus hemolymph collected at 1 and 5 days post-infection (DPI). Points represent individual bugs. RproRr bugs had fewer E. coli CFU at both 1 and 5 DPI than RproEc or RproAxn. RproEc had fewer E. coli CFU at both 1 and 5 DPI than RproAxn (** p < 0.002, * p < 0.05, Wilcoxon test). (C) Hemolymph from RproRr bugs suppresses growth of both E. coli and M. luteus in vitro more than RproEc or RproAxn bugs. ***p < 0.001, Tukey’s HSD.

https://doi.org/10.1371/journal.ppat.1012947.g001

Our survival curves indicate that bugs reared with bacteria in their gut can survive infection with E. coli or M. luteus while nearly all RproAxn die after infection. Insects can overcome infection through either resistance – killing or clearing of a pathogen – or through tolerance – minimizing fitness costs of infection without reducing the pathogen load. To determine if the observed difference in survival was due to tolerance or resistance, we measured E. coli titer in the hemolymph after infection over time. Because both the E. coli and M. luteus treatment groups had similar mortality rates we chose to further investigate only E. coli for this experiment. We injected 6–8 bugs with 106 CFU of live, kanamycin-resistant E. coli, then collected hemolymph at 1- and 5- days post injection. Hemolymph was spread onto plates containing kanamycin to calculate CFUs of the kanamycin-resistant E. coli present in the hemolymph. The presence of bacteria in the gut had a significant impact on the number of KanR E. coli hemolymph (F2,32 = 123.2, p < 0.0001, aligned-rank transformed ANOVA). One day post injection with E. coli, RproRr insects had on average 1.6 x 103 CFUs of E. coli, RproEc had 1.82 x 104, and RproAxn individuals had 4.6 x 105 CFUs, higher than either gnotobiotic treatment (p = 0.002, Wilcoxon signed-rank test, Fig 1B). Five days after injection with E. coli most RproAxn had died; those who remained alive had on average 4.3 x 106 CFUs of E. coli. RproEc bugs had significantly fewer E. coli than RproAxn bugs (1.8 x 104 CFUs, p = 0.014, Wilcoxon signed-rank test, Fig 1B), while RproRr bugs had the least E. coli in their hemolymph of any treatment (2 x 102 CFUs, p = 0.014, Wilcoxon signed-rank test). The identity of the gut bacteria influences the extent to which the gnotobiotic bugs can clear the bacteria from their hemolymph, with RproRr bugs having significantly fewer bacteria at day 5 than RproEc bugs (p = 0.014, Wilcoxon signed-rank test). Though neither gnotobiotic treatment completely cleared E. coli from their hemolymph during the observation window, they appear to suppress the pathogen to a level that is survivable.

We next sought to determine if the decrease in bacteria in the hemocoel of challenged RproAxn, RproEc, and RproRr bugs was due to bacterial killing via antimicrobial factors in the hemolymph. To explore if presence of gut bacteria results in differences in antimicrobial activity in the bugs’ hemolymph, we conducted a zone of inhibition assay [26,27]. We injected nine RproAxn, RproEc, and RproRr 4th instar bugs with 106 cells of either heat-killed E. coli or M. luteus. We used heat-killed bacteria for this assay to avoid bacterial growth on the plates that could potentially confound our zone of inhibition measurements. RproRr and RproEc bugs had significantly larger zones of inhibition when infected with E. coli than RproAxn bugs (p < 0.0001, Tukey’s HSD, Fig 1C). Interestingly, only RproRr bugs had significantly larger zones of inhibition after being infected with M. luteus (p < 0.0001, Tukey’s HSD, Fig 1C). These results indicate that RproRr and RproEc bugs have higher antimicrobial activity in their hemolymph when challenged with Gram-negative E. coli while only RproRr bugs had higher antimicrobial activity after infection with the Gram-positive M. luteus, possibly due to activation of the Toll pathway in RproRr. Naïve individuals did not show a significant difference in antimicrobial activity from each other, suggesting this is an induced response (S2 Fig). These data demonstrate the presence of gut bacteria promotes resistance to pathogens in R. prolixus through a reduction in the number of bacteria in the hemolymph of gnotobiotic bugs but not RproAxn bugs.

The squash bug Riptortus pedestris harbors environmentally acquired Burkholderia bacteria that provide an immunoprotective effect to the host, which is induced upon migration from the M3 region of the midgut to the hemolymph [16,17]. We repeatedly isolated hemolymph from RproRr bugs and plated it on LB media that readily grows R. rhodnii. We screened dozens of RproRr yet did not observe any growth of R. rhodnii on plates, suggesting that R. rhodnii does not escape the gut, and that the immune priming effect of R. rhodnii occurs without direct contact with the insect hemocoel.

The protective effect of R. rhodnii requires live bacteria and is lost upon removal

Insects sense bacteria through detection of MAMPs such as peptidoglycan, lipopolysaccharides, and lipoteichoic acid, which activate signaling cascades via serine proteases. We therefore asked whether the immune priming effect of R. rhodnii was primarily due to the presence of MAMPs, and so we fed RproAxn bugs blood meals that were spiked with an amount of heat-killed R. rhodnii that was equivalent to 108 CFUs/bug (RproAxn+HK Rr) from the first blood meal until they developed into 4th instars, then challenged them as before with injection of E. coli. All RproAxn+HK Rr bugs died within the 10-day window and their survival was not statistically different from the survival of RproAxn bugs (p = 1.00, Cox proportional hazards model, Fig 2A). From these results we conclude that live bacteria are likely necessary for immune priming in this system, though we cannot rule out a role for heat-labile small effector molecules produced by R. rhodnii in stimulating the immune response.

thumbnail
Download:
Fig 2. R. prolixus requires live bacteria to mount an effective immune response.

(A) Survival after bacterial injection of 4th instar RproAxn nymphs fed heat-killed R. rhodnii throughout development (RproAxn + HK Rr) was similar to 4th instar RproAxn nymphs not fed bacteria (p = 0.69, log-rank test). Survival of RproRr was significantly higher (p < 0.0001, log-rank test). (B) Survival after immune challenge of RproAxn bugs after restoration of R. rhodnii (RproAxn + Rr). Restoring R. rhodnii via a blood meal significantly increases survival relative to axenic nymphs (p < 0.0001, log-rank test) but there was still significantly lower survival in the RproAxn + Rr than in RproRr (p < 0.0001, log-rank test). (C) Clearance of R. rhodnii from RproRr significantly reduced survival following immune challenge with E. coli. RproRr-KanR bugs treated with kanamycin retained their immune priming effect. Within a panel, lines not connected by the same letter are significantly different (p < 0.001, log-rank test). Each experiment consisted of 15–25 bugs per treatment.

https://doi.org/10.1371/journal.ppat.1012947.g002

The main function of R. rhodnii is thought to be supplementation of B vitamins to R. prolixus [22,28]. These nutrients are important in numerous essential processes in the host and their absence throughout RproAxn development may underlie the higher survival of RproRr following immune challenge. Alternatively, the protective effect of R. rhodnii may be independent of its nutrient provisioning services. If the effect were nutritional, we would expect that removal of R. rhodnii from RproRr bugs shortly before an immune challenge would not have a dramatic effect on survival, while if the effect were primarily immune priming, addition of R. rhodnii to RproAxn bugs shortly before challenge would lead to increased survival relative to RproAxn individuals.

To investigate this, we fed 3rd instar RproAxn nymphs a blood meal containing 106 CFU/mL of R. rhodnii. Nymphs developed into 4th instars and a subset were sacrificed and qPCR was used to confirm that sacrificed nymphs harbored R. rhodnii (S3 Fig). Two weeks after molting, insects were injected with live E. coli as previously outlined. RproAxn+Rr bugs had higher survival than RproAxn bugs (Fig 2B, p = 0.0047, Cox proportional hazards) but still suffered increased mortality compared to RproRr individuals. Interestingly, when we removed R. rhodnii from RproRr bugs by feeding kanamycin to 3rd instar insects via the blood meal (RproRr + Kan), they also suffered from increased mortality compared to bugs that were fed kanamycin but had been inoculated with kanamycin resistant R. rhodnii (RproRr-KanR + Kan, Fig 2C). Thus, we conclude that the nutritional role of R. rhodnii in R. prolixus is not sufficient to explain the observed immune priming effects, but that nutrient provisioning may contribute to the protective effects of R. rhodnii. We did not test this in RproEc bugs, as their B vitamin synthesis capabilities are similar to R. rhodnii, and we do not anticipate that nutritional provisioning by E. coli would be able to rescue immune priming when R. rhodnii could not.

Immune priming by gut bacteria acts through both the Toll and IMD pathways

Insect immune responses to bacteria are thought to be mediated by two pathways, Toll and IMD [29]. Our zone of inhibition assays suggested that some factor(s) in the hemolymph are involved in suppressing bacterial proliferation, which may be regulated by Toll or IMD. To investigate whether the Toll or IMD pathways are essential to R. rhodnii-mediated immune priming, we first measured expression of two transcription factors, dorsal and relish, which drive expression of Toll and IMD response genes including antimicrobial peptides (AMPs). For both genes, we assessed expression via qPCR in insects injected or blood-fed with live E. coli or M. luteus, as described above. We tested the fat bodies of naïve or injected insects and the guts of blood-fed insects. We tested RproRr, RproEc, and RproAxn bugs, and found that naïve, RproAxn bugs had low expression of both dorsal and relish (Fig 3A and 3B).

thumbnail
Download:
Fig 3. Immune priming by gut bacteria is dependent on Toll and IMD pathways.

(A, C) Expression of the Toll pathway transcription factor dorsal in RproRr, RproEc, and RproAxn in naïve bugs or bugs challenged with E. coli or M. luteus (A) injection or (C) via an inoculated blood meal. (B, D) Expression of the IMD pathway transcription factor relish in RproRr, RproEc, and RproAxn in (B) naïve bugs or bugs challenged with E. coli or M. luteus injection or (D) orally via an inoculated blood meal. Bars connected by different letters are significantly different (p < 0.05, Tukey’s HSD). (E) Survival curves of RproRr bugs treated with dsRNA against dorsal, relish, or a control egfp sequence demonstrate that silencing of relish or dorsal via RNAi significantly reduced the survival of RproRr bugs, indicating that these pathways are necessary for R. rhodnii-mediated immune priming. Different letters indicate treatments with significantly different survival (p < 0.0001, log-rank test). (F) Silencing dorsal or relish reduces the bacteriostatic factors in hemolymph of RproRr. Different letters indicate treatments with significantly different survival (p < 0.0001, Tukey’s HSD).

https://doi.org/10.1371/journal.ppat.1012947.g003

For each gene, there was a highly significant effect of the gnotobiotic state of the bugs, the immune challenge, and the interaction between state and challenge (see S1 Table for statistical details). Naïve RproEc bugs had higher expression of dorsal and relish than RproAxn, indicating that E. coli in the gut stimulates expression of both transcription factors independent of immune challenge via injection. Naïve RproRr bugs had low expression of dorsal, equivalent to RproAxn bugs but had higher expression of relish than RproAxn. In bugs challenged with injection of E. coli, both dorsal and relish expression were elevated in their respective gnotobiotic bugs relative to RproAxn. M. luteus also stimulated robust expression of relish and dorsal, with RproRr bugs having the strongest expression of dorsal while relish expression was high in both RproRr and RproEc. Feeding of RproRr and RproAxn with E. coli or M. luteus induced a smaller change in expression of relish or dorsal, though for most treatments, RproRr had significantly higher gut expression of these genes than RproAxn (p < 0.05, Tukeys HSD, Fig 3C and 3D).

From our expression analysis of dorsal and relish, we observe significant cross activation of the Toll and IMD pathways, with both Gram-negative and Gram-positive bacteria activating both pathways. The Toll pathway, and by extension dorsal, is thought to primarily respond to Gram-positive bacteria while IMD and relish are thought to be activated by Gram-negative bacteria, but significant crosstalk between these pathways has been observed in other hemipterans [4]. Absence of gut bacteria does not eliminate the ability of bugs to mount an immune response, but gnotobiotic bugs almost always have higher expression of these genes. In all bugs tested, both dorsal and relish can be induced by the presence of bacteria in the hemolymph, but we were surprised to see that RproEc often induces higher expression of dorsal and relish than RproRr, given the latter’s stronger protective effect.

We next asked whether the immune priming effect of gut bacteria is dependent on the Toll or IMD pathway. We suppressed each pathway via RNAi knockdown of the transcription factors relish (IMD) or dorsal (Toll). RNAi of relish or dorsal was confirmed via qPCR revealing over a 90% reduction in dorsal or relish transcripts after injection with dsRNA (S4 Fig). After injection with dsRNA, 106 CFUs of E. coli were injected into the bugs’ hemocoel, and survival rates were measured (Fig 3C). RproRr treated with either dorsal or relish dsRNA succumbed to E. coli infection at a higher rate than RproRr individuals treated with dsegfp (25% and 34% survival respectively, p = 0.002, p = 0.0004, respectively, log-rank test).

We then examined the change in antimicrobial activity in relish or dorsal-silenced R. prolixus by comparing the inhibitory effects of hemolymph on microbial growth using the zone of inhibition assay described above, though we only tested the effect of hemolymph from dsRNA-treated bugs against E. coli. There was a significant effect of silencing either relish or dorsal as silenced bugs had smaller zones of inhibitions compared to the control dsegfp-injected group (Fig 3D, p < 0.0007, p < 0.0001, Tukey’s HSD). The decrease in survival after challenge with E. coli in either relish or dorsal knockdown bugs suggests immune pathway cross activation in R. prolixus, as silencing of the Toll pathway transcription factor dorsal led to significantly higher mortality following challenge with Gram-negative bacteria. The reduction in antimicrobial activity in hemolymph following knockdown of dorsal or relish suggests that the activation of the Toll and IMD pathways by R. rhodnii has functional consequences for host immune responses.

R. rhodnii influences the cellular immune system

Cellular immune responses are often the first line of defense against pathogens, clearing many bacteria from the hemolymph before upregulation of antimicrobial peptide gene expression [26]. Given the robust immunoprotective effect of R. rhodnii, we wondered if presence of R. rhodnii induced cellular immune responses. We first measured the number of circulating hemocytes in the hemolymph of five RproRr, RproEc, and RproAxn bugs. Fourth instar nymphs were immune challenged with E. coli and hemolymph was collected at 0 and 4 hours post-injection and a hemocytometer was used to count the number of hemocytes present. Immediately after injection RproRr bugs had a significantly higher number of circulating hemocytes compared to RproAxn bugs (p = 0.0001, Fig 4A) and RproEc bugs (p = 0.02, Tukey’s HSD) and there was no significant difference in hemocyte counts between RproAxn and RproEc (p = 0.11, Tukey’s HSD). These results demonstrate that RproRr bugs have a higher initial number of hemocytes than RproEc or RproAxn. Four hours after injection, all hemocyte titers had increased (Fig 4B, p < 0.001 for all comparisons, Tukey’s HSD), but RproRr insects still had significantly higher hemocyte counts than both RproAxn and RproEc bugs (p = 0.0008 and p = 0.001, respectively, Tukey’s HSD).

thumbnail
Download:
Fig 4. RproRr bugs have higher baseline and induced numbers of hemocytes relative to RproEc and RproAxn bugs.

(A) Hemocyte counts from hemolymph extracted at 0 h post-challenge with E. coli. There was no significant difference in the number of hemocytes between RproAxn and RproEc but RproRr had significantly more than either (p = 0.0001, p = 0.02, Tukey’s HSD). (B) At 4 h post challenge, all bugs had more hemocytes but RproRr still had significantly more than RproEc or RproAxn, which were not significantly different from one another (p = 0.0008, p = 0.001, p = 0.23 respectively, Tukey’s HSD). (C) Fluorescent latex beads are consumed by a hemocyte (arrow) in RproRr. Top panel: light microscopy, middle panel: 395 nm, bottom panel: merge. (D) Hemocytes are essential for R. rhodnii-mediated immune effects. Inactivation of hemocytes by injection of latex beads dramatically reduced survival of RproRr bugs after challenge with E. coli (**** p < 0.0001, log-rank test).

https://doi.org/10.1371/journal.ppat.1012947.g004

To investigate whether the differences in hemocyte counts among the treatments were a major factor in infection outcome, we suppressed hemocyte phagocytosis by pre-injecting fluorescent latex beads into the hemocoel of RproRr bugs [30]. The beads were visually confirmed to be phagocytosed by the hemocytes in vitro via microscopy (Fig 4C). Twenty-four hours after bead injection, E. coli was injected into the hemolymph and survival was monitored as previously described. Insects with suppressed cellular immunity (RproRr/bead) succumbed to bacterial infection faster than RproRr/control insects (p < 0.05, log-rank test, Fig 4D), indicating that R. rhodnii-mediated cellular immunity represents an important component of R. rhodnii-based immune priming, but that E. coli in the gut does not induce as strong of a cellular response.

Expression of immune-related genes is influenced by the gut microbiome

Our experiments demonstrated that dorsal and relish were important mediators of the immune priming effect seen in RproRr. We next wanted to see if antimicrobial peptides were upregulated in gnotobiotic insects. To test this, we evaluated the fat body expression profiles of AMPs in 4th instar RproAxn, RproEc, and RproRr bugs that were either uninfected or 1-day post-inoculation with either E. coli or M. luteus. We measured the expression of the AMPs prolixicin, two defensins (RPRC012182 and RPRC012184, subsequently referred to as defensin82, and defensin84), and a lysozyme (RPRC015442 subsequently referred to as lysozyme42, Fig 5A).

thumbnail
Download:
Fig 5. Antimicrobial peptide (AMP) genes are differentially expressed among naïve, RproEc, and RproRr bugs.

Induction of AMP expression was generally higher in insects (A) injected with bacteria than in insects that were (B) fed blood meals inoculated with bacteria. Within each AMP panel, bars connected by different letters are significantly different (p < 0.05, Tukey’s HSD).

https://doi.org/10.1371/journal.ppat.1012947.g005

The interaction of gnotobiotic state by injection was highly significant for all genes tested (p < 0.0001, aligned ranks transformation ANOVA for non-parametric interactions). Our expression analysis of immune genes revealed several interesting patterns. First, RproAxn insects often had lower expression of AMP genes regardless of the immune challenge (Fig 5A), which may be a consequence of lower relish and dorsal expression in RproAxn bugs (Fig 3A and 3B). The overall reduced expression of immune genes may partially explain the heightened susceptibility of RproAxn to pathogens. A second, surprising pattern is that RproEc bugs often have higher expression of AMPs than RproRr bugs, despite RproRr having a significantly higher survival rate when challenged with pathogens. In the RproEc bugs challenged with E. coli, expression of all immune genes except defensin84 were significantly higher than when RproRr bugs were challenged with E. coli, indicative of a strong immune priming effect of E. coli in the gut against E. coli in the hemolymph (Fig 5A, p < 0.05, Tukey’s HSD). RproRr bugs exhibited strong induction of immune gene expression in response to bacterial challenge relative to naïve RproRr or RproAxn, apart from defensin84, suggesting that humoral immune responses are important in the heterologous immune priming seen in RproRr following pathogen challenge. Oral challenge with E. coli induced expression of AMPs defensin82 and defensin84, but the induction was independent of gnotobiotic state. RproRr had higher expression of prolixicin than RproAxn bugs, but this was independent of pathogen challenge (Fig 5B).

The third pattern we observed was further evidence of cross activation of the Toll and IMD immune pathways, consistent with the earlier experiments examining dorsal and relish (Fig 3). Defensins, AMPs which act primarily against Gram-positive bacteria [31], were highly expressed in RproEc in response to challenge with Gram-negative bacteria (Fig 5, defensin 84, 82). Likewise, lysozyme42 expression was also induced by Gram-negative bacteria (Fig 5A) despite canonically being considered an AMP against Gram-positive bacteria [32]. A similar pattern was seen with expression of prolixicin, an ortholog of Drosophila diptericin, as it was induced in RproRr bugs after challenge with M. luteus. This is despite Diptericin being active against Gram-negative bacteria [33]. Our data supports and expands on earlier work in the stink bug Plautia stali which demonstrated that expression of relish and several immune effectors can be induced by both E. coli and M. luteus in Hemiptera [4].

Melanization and phenol oxidase activity are dependent on the presence of R. rhodnii

We observed throughout our experiments that RproAxn individuals exhibited reduced wound healing. RproAxn bugs did not develop a robust, dark melanization scar at injection sites but rather a light, thin scar, while RproRr developed a characteristic thick, dark, scar (Fig 6A). Interestingly we also saw a lack of dark, thick scar tissue in RproEc bugs. The dark scars following wounding are due to deposition of melanin produced by the action of phenol oxidases and other enzymes in the hemolymph which convert tyrosine to melanin [34]. The cascade leading to melanization can be triggered by both wounding and MAMPs [35]. Based on these observations and the role of melanization in insect immunity, we decided to further investigate differences in melanization in RproRr, RproEc, and RproAxn insects.

thumbnail
Download:
Fig 6. R. rhodnii enables successful melanization in R. prolixus.

(A) Melanization of wounds is impaired in RproAxn and RproEc relative to RproRr. Wound area is outlined. (B) Hemolymph DOPA conversion assay of RproAxn, RproEc, and RproRr measured at 1 and 24 h post-hemolymph collection with E. coli (circles) or M. luteus (triangles). At 1 h, there was no significant difference in conversion of DOPA to melanin between the bugs or bacterial challenges (F2,44 = 0.11 p = 0.74, ANOVA). By 24 h post-collection, there was a significant increase in DOPA conversion in all treatments relative to 0 h (p < 0.05, Tukey’s HSD). RproRr bugs had higher DOPA conversion than RproAxn and RproEc bugs (p < 0.05, Tukey’s HSD), which were not significantly different from each other (p > 0.05, Tukey’s HSD). As with the 1 h time point, there was no significant effect of the bacterial species used in the challenge (F1,44 = 0.21, p = 0.648, ANOVA). (C) Wounding induces activation of the melanization response 24h post-hemolymph collection. RproRr had higher DOPA conversion in response to wounding than RproAxn. Bars connected by different letters are significantly different (p < 0.05, Tukey’s HSD n = 5 insects per treatment).

https://doi.org/10.1371/journal.ppat.1012947.g006

We measured phenol oxidase activity via a dihydroxyphenylalanine (DOPA) conversion assay [36] to assess how the presence of gut microbes influences melanization. DOPA is a precursor metabolite that is converted into melanin via the action of phenol oxidases. Hemolymph DOPA conversion was measured at 1 and 24 hours following hemolymph collection. There was a highly significant effect of the gnotobiotic state (Fig 6B, F2,54 = 23.8, p = 3.88 x 10-8, ANOVA) and time (F1,54 = 681, p < 2 x 10-16, ANOVA) on the amount of DOPA conversion. Surprisingly, there was not an effect of the microbe injected, as both E. coli and M. luteus injection triggered similar melanization among the different gnotobiotic states (Fig 6B, F3,54 = 0.09, p = 0.96). There was no significant difference in DOPA conversion at 1 h post-collection, but at 24 h post-collection there was a significant increase after injection of either E. coli or M. luteus. Compared to RproEc and RproAxn, RproRr had greater DOPA conversion regardless of challenge. We also investigated whether a stab wound alone was sufficient to trigger an increase in melanization. We stabbed RproAxn and RproRr bugs with a sterile insulin syringe, then collected hemolymph and measured DOPA conversion as before at 1 and 24 h post-collection. We found that similar to injections with E. coli or M. luteus, stabbing with a sterile needle induced the ability to convert DOPA in both RproAxn and RproRr hemolymph by 24 h (Fig 6C, p < 0.001, Tukey’s HSD) but not 1 h (p = 0.99, Tukey’s HSD), with a larger increase in conversion seen in the RproRr bugs (p = 0.014, Tukey’s HSD). Our DOPA conversion assays suggest that R. rhodnii is important for successful melanization in R. prolixus.

Discussion

We demonstrate that the presence of gut microbes is integral to a functional immune system in R. prolixus and key to their ability to overcome infection with facultatively pathogenic microbes. We provide multiple lines of evidence that gut microbes in R. prolixus induce a strong immune response against hemocoelic pathogens, “priming” the immune system by influencing humoral and cellular immunity, and that symbiotic R. rhodnii induce a stronger priming and protective effect than commensal E. coli. This enhanced immune response encompasses both cellular and humoral immunity and is active against diverse bacterial pathogens, suggesting a hyperactivation of immune responses by the symbiont R. rhodnii. Bugs that harbored a microbe in their gut exhibited a strong humoral immune response by expressing significantly higher levels of antimicrobial peptides than microbe-free RproAxn individuals, that corresponded to higher antimicrobial activity in their hemolymph, and subsequently an increased ability to reduce the number of infecting microbes compared to RproAxn bugs. Our results suggest that different gut microbes activate the immune system in different ways and to different extents. R. rhodnii produces a stronger effect and modulates both cellular and humoral immunity while E. coli mainly stimulates humoral immune factors and does not have as strong of a priming effect.

Immune priming has been revealed to be a broadly important facet in insect immunity. Despite their lack of antibody-mediated adaptive immunity, many insect species exhibit elevated immune responses after a second exposure to a nonlethal dose of a pathogen or pathogen derived material. This priming manifests as increased production of AMPs and mobilization or proliferation of circulating hemocytes [37], ultimately leading to the insect being resistant to subsequent infections [16,38–41]. Immune priming was initially characterized through experiments where an insect is repeatedly exposed to a single pathogen, known as homologous immune priming [42]. More recently, heterologous immune priming, in which one pathogen induces a protective immune response towards a different secondary pathogen has been observed [38]. The growing appreciation for symbiotic associations between insects and microbes has also illuminated the role of symbionts in host immune function [13,16–20,43,44]. Our results suggest that both the Gram-positive symbiotic R. rhodnii and Gram-negative commensal E. coli stimulate heterologous immune priming against a different bacterial species, though to varying degrees.

In other systems where symbiont-mediated immune priming has been observed, the bacteria responsible have at least some contact with the host hemolymph or extracellular tissues beyond the gut. Tsetse flies (Glossina sp.), which are also obligately exclusively hematophagous, harbor an intracellular symbiont, Wigglesworthia glossinidia. The symbiont is essential for proper immune function in Glossina morsitans, as aposymbiotic larvae develop into adults with impaired immune responses, reduced hemocyte proliferation, and inability to melanize properly [19,20,43]. To achieve an immune priming effect, Wigglesworthia must be present during larval development [19,20,43]. During this time Wigglesworthia can be found in the lumen of the mother’s milk glands and extracellularly in larval fat body [45]. In the bean bug, Riptortus pedestris, their gut symbiont Caballeronia, which resides in specialized crypts within the gut, strongly contributes to the activation of host immunity [17]. However, stronger immune priming is seen when a soil-derived Burkholderia colonizes R. pedestris and escapes the gut, increasing AMP expression and hemocyte number [16]. The priming observed in Riptortus by Burkholderia sp. involves extracellular localization of the bacterium allowing for direct contact with the host hemolymph and presumably hemocytes [16].

Mosquito immunity has also been shown to be influenced by symbiotic bacteria. Wolbachia upregulates immune-related genes of its natural Drosophila host as well as when artificially introduced to mosquitoes, thereby conferring protection against a variety of pathogens [46–50]. Though Wolbachia is nearly always intracellular, it can infect a variety of insect tissues including the immunologically important fat body [46]. The mosquito symbiont Asaia activates AMP production in Anopheles stephensi mosquitos, and though primarily associated with the gut, it is found in several other tissues [51].

In contrast to the Glossina-Wigglesworthia, Riptortis-Burkholderia, and Aedes-Wolbachia or Asaia systems, we never detected R. rhodnii outside of the gut of R. prolixus. R. rhodnii is environmentally acquired each generation and does not escape the insect’s alimentary canal. Despite this, it potentiates a strong and multifaceted immune response beyond the insect gut. We observe mobilization of hemocytes, increased expression of AMPs in the fat body, and elevated melanization potential in the hemolymph. We initially hypothesized that R. rhodnii MAMPs may escape the gut and activate the host immune system, and thus expected dead R. rhodnii would be able to recapitulate RproRr immune priming. In other systems, dead bacteria can elicit immune priming [30], but dead bacteria did not stimulate a protective response. We are left to conclude that R. rhodnii produces some extracellular signal or activates a host signaling mechanism in the gut that reaches the fat body and hemocytes to induce immune protection. Several candidate molecules including nitric oxide and prostaglandins have been previously shown to be important in R. prolixus immunity [24,52,53], and may be necessary for symbiont mediated immune priming.

Differences in the gut anatomy and physiology of kissing bugs may contribute to the ability of R. rhodnii to activate the host immune system without leaving the gut. Most insects possess an acellular chitinous and proteinaceous peritrophic matrix that lines the midgut epithelium and is responsible for protecting the midgut cells from direct contact with gut microbes. This protective barrier modulates immune activation by the gut microbiome [54]. Hemipterans, including kissing bugs, do not have a peritrophic matrix (PM), but rather a lipid-based structure called the perimicrovillar membrane (PMM). While the PMM forms a barrier between the gut lumen and its resident microbiota, it is possible that the PMM is not as significant a barrier to microbes or MAMPs and may permit direct contact of gut microbes with the gut epithelia. Such direct contact could potentially activate host immune responses to a greater extent than if bacteria did not contact the epithelial cells directly [5]. Additional studies are necessary to understand the extent to which gut bacteria or MAMPs in R. prolixus encounter the epithelium, and how this influences immune responses.

The immune priming response we observed in RproRr was very strong in comparison to similar experiments in Drosophila, which used lower doses of bacteria for immune challenges. Immune priming by prior exposure to Enterococcus faecalis lead to ~30% of challenged flies surviving a low dose infection with ~3,000 CFU of E. faecalis and only ~25% of flies surviving a high dose infection of 30,000 CFU [55]. Drosophila primed against Streptococcus pneumoniae had higher survival against 250 or 35,000 CFU of live S. pneumoniae [30]. Another study using heat-killed Providencia rettgeri followed by challenge with > 100 CFU of live bacteria found variable protection from subsequent infections even at this low dose [56]. In our experiments RproRr exhibited high survival rates (65% to 75%) when challenged with 106 CFU of E. coli or M. luteus, demonstrating that the immune priming effect of R. rhodnii in R. prolixus is highly effective at protecting the host from these infections. The ability to survive challenge with higher bacterial loads may be related to the specific pathogens tested. The priming observed in the E. faecalis experiments was attributed primarily to tolerance, as primed flies had higher bacterial loads at death than un-primed flies [55]. In the experiments using P. rettgeri, both primed and un-primed insects displayed clearance of the pathogen, suggesting that the clearance was independent of priming [56]. Again, our results highlight differences between R. prolixus and Drosophila, as RproRr were able to suppress and reduce the number of bacteria following challenge, indicating that RproRr survival is related to resistance to pathogens rather than tolerance.

Though many bacterial pathogens likely invade through ingestion and subsequently escape the midgut, we chose to focus our experiments on injection directly into the hemocoel. Recent work has demonstrated that up to 30% of wild-caught insects have scars that cannot be explained by parasitism by wasps or mites, demonstrating that injury is common in insects [57]. Much like vertebrates, the risk of infection following injury is likely significant, as insects have evolved a robust and rapid immune response to wounding that involves activation of insect immune pathways, mobilization of hemocytes, and production of AMPs [58–61]. Injuries can lead to resulting infections, as seen in parasitic mite bites which lead to secondary bacterial infections in A. mellifera [62], and in Drosophila where hemolymph-feeding mites transfer pathogens between flies [63]. Kissing bugs may be particularly susceptible to injury and subsequent infection as they exhibit kleptohematophagy, where bugs will attack conspecifics and steal blood from fed individuals [64]. Furthermore, they are long-lived relative to many other model insects, potentially increasing the likelihood of injury during their lifetime [65,66]. A second rationale for introduction of pathogens via injection over feeding is to isolate the effects of the immune system from confounding factors that would have influenced a feeding-based infection. These factors include direct and indirect bacterial competition between gut bacteria and those being used for immune challenge, niche exclusion of bacteria used for challenges, and vertebrate immune proteins in the blood meal.

While wounding is likely an underappreciated route for infection in insects, it is unlikely that insects such as R. prolixus would experience an inoculation of the same magnitude as we employed in our experiments. We chose a titer of bacteria that induced significant mortality in axenic but not gnotobiotic insects which clarified the fitness effects of gut symbionts in relation to infection. While we did not observe significant mortality in experiments with lower inoculations of bacteria, we did not follow up with subsequent studies of the outcomes of these infections regarding long-term survival, successful development, or reproduction. More field-relevant inoculations may still have significant but subtler fitness effects that our experiments were not designed to capture. However, the protective effects of R. rhodnii against high titers of pathogens suggests that the symbiont likely would reduce the fitness costs of field-realistic infections.

Studies of immune priming in insects examined several insect lineages including Drosophila melanogaster [67,68], Aedes aegypti [69], Anopheles sp. [70–72], Tenbrio [73], Gallaria mellonella [74], Tribolium castaneum [75] among others. These studies have primarily explored holometabolous insects and relatively few have been performed in paurometabolous orders. Our results contribute to a growing understanding that hemipteran immunity is substantially different from many holometabolous model organisms such as Drosophila. Genomic data has revealed that many hemipteran genomes lack elements of the canonical insect immune pathways, including aphids [2], bedbugs [76], scale insects [77], and plant hoppers [78]. The genome sequence of R. prolixus was initially thought to be missing key components of the IMD pathway, though subsequent analysis has revealed that despite lack of the genes imd and kenny, R. prolixus does indeed possess a functioning IMD pathway [3,79]. Our results support these earlier findings, as both R. rhodnii and E. coli stimulate expression of relish and IMD-associated AMPs, while inactivation of IMD signaling through RNAi silencing of relish leads to increased mortality of R. prolixus. The absence of imd and kenny suggests that IMD immune signaling functions differently in R. prolixus highlighting the need for further investigation into immune signaling in R. prolixus and hemipterans in general.

Immune signaling in insects was initially described in Drosophila as a linear response where different classes of pathogens trigger discrete immune pathways: Gram-negative bacteria activate the IMD pathway, while Gram-positive bacteria and fungi activate the Toll pathway [29]. Yet as insect immune studies have moved beyond Drosophila and into other insects, it appears that crosstalk between or cross activation of immune pathways may be more common than previously thought. Work in the hemipteran stink bug, Plautia stali, revealed that not only are both IMD and Toll pathways present but there is a blurred functional differentiation, as immune challenge with Gram-negative or Gram-positive bacteria elicited expression of immune effector genes of both pathways [4]. We see similar patterns in our expression data as immune challenge with Gram-positive and Gram-negative bacteria trigger expression of both dorsal and relish immune transcription factors belonging to the Toll and IMD pathway respectively. Knockdown of either dorsal or relish also results in increased mortality after challenge with Gram-negative E. coli, and AMPs thought to be active against either Gram-positive or Gram-negative bacteria are expressed in response to both Gram-negative and positive bacteria. Crosstalk has been seen in other insects, including the beetle Tenebrio molitor [73], and other species of kissing bugs. In Triatoma pallidipennis, silencing of Toll pathway genes led to increased mortality when bugs were challenged with Gram-negative bacteria [10]. Interestingly, silencing of relish in T. pallidipennis did not result in increased mortality. Even in Drosophila, gut bacteria can elicit immune priming against heterologous pathogens including fungi [30] and viruses [80]. Together, these studies and others broaden our understanding of insect immunity, regarding both the Toll and IMD signaling pathways and immune priming by heterologous bacteria.

While R. rhodnii appears to be a ubiquitous member of the gut community in members of the genus Rhodnius [23], surveys of the microbiome of wild-caught Rhodnius pallescens describe a moderate diversity community that varies with age, geographic location, and T. cruzi status [81]. Our experiments were performed in a long-term lab colony of R. prolixus maintained at the Centers for Disease Control and Prevention since at least 1989 (Ellen Dotson, personal communication). At both the CDC and in our lab, the colony is maintained on sterile defibrinated rabbit blood supplemented with R. rhodnii using artificial membrane feeders [82]. This has likely altered the microbiome of the bugs, as other lab-reared triatomines have distinct microbiomes relative to wild-caught individuals [83]. While our previous results demonstrate that R. rhodnii in isolation is a necessary and sufficient symbiont of R. prolixus [22] and the current study demonstrates that R. rhodnii provides robust immune priming to R. prolixus, it is possible that other microbes in the R. prolixus gut influence the insect’s immune response or alter the impact of R. rhodnii on host immune function. Our findings that RproEc individuals had higher expression of several AMP genes tested when compared to RproRr bugs demonstrate that different gut microbes have different modes of immune priming of R. prolixus. The natural, more complex microbiomes of wild triatomines may induce even stronger protection against pathogens than R. rhodnii alone does for R. prolixus.

Immune function has been linked to nutrition in many different insects. Starving or rearing insects on nutritionally poor diets alters the humoral and cellular immunity leading to increased mortality after infection [84]. Diet interacts with the immune response in R. prolixus as well. Thirty days of starvation post-ecdysis resulted in increased mortality after infection due to changes in the cellular immune system, and similar results were found when bugs were fed an incomplete diet of plasma alone [85]. All bugs used in our experiments were two weeks post-molt to control for any effects starvation has on survival after infection. We further attempted to disentangle the impacts of nutrition on survival by clearing R. rhodnii from RproRr with a blood meal containing antibiotics. These cleared bugs were confirmed to have no R. rhodnii present and were challenged with E. coli. Interestingly, these bugs suffered high mortality rates, though not as high as RproAxn bugs, suggesting that R. rhodnii’s influence on the immune system is not primarily mediated by nutritional factors. Conversely, axenic individuals that were fed a blood meal containing R. rhodnii at their 3rd instar then challenged with E. coli shortly after had higher survival than RproAxn but significantly lower survival than our RproRr group. These results taken together suggest that nutritional supplementation via the microbiome may play some role in immunity, but other factors are likely more important than this interaction. Direct experiments with B vitamin supplementation and B vitamin auxotrophic R. rhodnii will be necessary to fully resolve the role of symbiont-provisioned nutrients in R. prolixus immunity.

Kissing bugs are the vectors of Trypanosoma cruzi, the causative agent of Chagas disease. T. cruzi is a stercorarian parasite, residing exclusively in the gut of its triatomine host during the insect phase of its development. As a result, it may be directly or indirectly influenced by gut bacteria. The activation of the immune system by gut bacteria may have consequences for T. cruzi persistence and transmission, as has been seen in Anopheles mosquitoes where presence of gut bacteria induces a strong immune priming effect via hemocyte differentiation following ookinete escape from the midgut, and subsequently reduces the survival of Plasmodium in the mosquito [70]. However, our results suggest that within the gut, there is lower activation of the host immune response relative to immune factors in the hemolymph.

The triatomine microbiome has been implicated in diverse interactions with T. cruzi, recently reviewed in [86]. Surveys of microbes in kissing bug guts infected with T. cruzi describe variable impacts on microbial diversity and abundance [81,87–90]. Tripartite interactions of T. cruzi, the host immune system, and microbiome have been documented [15,52,53,91–95], including activation of various host immune effectors by T. cruzi which suppress or alter microbial communities in the gut. In R. prolixus, infection with T. cruzi induces expression of antimicrobial genes which reduces the microbial community [91,93]. In T. infestans, infection with T. cruzi leads to expression of TiAP, an antimicrobial peptide active against Gram-negative bacteria [92]. Taken together these results indicate that T. cruzi manipulates the host immune system and, either directly or indirectly, the microbiome. How these interactions alter immune priming by R. rhodnii or how immune priming influences T. cruzi infection in R. prolixus are important outstanding questions in the field.

Our study provides evidence that the gut microbiome in kissing bugs plays an essential role in activating the host immune system against pathogens in the hemocoel. The nature of the immune priming appears to vary based on the identity of the gut microbe in question, as symbiotic microbes provide a stronger protective effect than non-symbiotic commensals. Both Gram-positive and Gram-negative bacteria were able to activate both the Toll and IMD pathways, which were both essential for immune activation regardless of the bacterial challenge, revealing that our understanding of insect immunity in kissing bugs and possibly other hemipterans, largely based on studies in Drosophila and other holometabolous insects, is not complete.

Materials and methods

Insect maintenance

Rhodnius prolixus were obtained from the lab of Dr. Ellen Dotson at the Centers for Disease Control and Prevention through BEI Resources. Insects were reared at 28 °C with a photoperiod of 12 h of light and 12 h of dark and 80% relative humidity. General colony insects were kept in 1 L Nalgene containers and regularly fed defibrinated rabbit blood (Hemostat Laboratories, Dixon, CA) inoculated with R. rhodnii bacteria in the exponential phase of growth through an artificial membrane feeder.

Generation of axenic and gnotobiotic nymphs

R. prolixus eggs were collected 7 days after being laid then placed in a sterile cell collection basket and washed with 70% ethanol for 5 minutes followed by 3 minutes in 10% povidone-iodine solution, then another 5-minute wash in 70% ethanol, followed by three rinses in autoclaved deionized water. Sterilized eggs were then transferred to autoclaved glass containers enclosed in sterile Nalgene containers with gas-exchange tape covering an air hole. Sterility was validated by screening total genomic DNA from insects with PCR to amplify a 16S rDNA gene with the universal primers 27F and 1492R (S1 Table). No bands were observed in axenic nymphs. Gnotobiotic nymphs were generated by feeding axenic first instar nymphs a blood meal inoculated with 106 CFU/mL of R. rhodnii or E. coli. Nymphs were fed at every instar approximately 2 weeks after molting. Gnotobiotic states were confirmed through qPCR on DNA extracted from whole bodies of nymphs using primers specific to the gyrB sequence of each bacteria (S1 Table).

Bacterial strains

Rhodococcus rhodnii (NRRL B-16535) was obtained from ATCC and grown at 28 °C in liquid Luria-Broth (LB). Escherichia coli MG1655 was a gift of Eric Stabb and was grown in liquid LB at 37 °C. Micrococcus luteus NCTC 2665 bacteria was a gift from Michael Strand and was grown in liquid LB at 37 °C. Bacterial titers were determined by measuring the OD600 of cultures on a Beckman Coulter DU640 spectrophotometer and then plating out serial dilutions of culture on LB agar plates to correlate OD600 with Colony Forming Units (CFU) counts.

Bacterial immune challenge

Bacterial immune challenge in kissing bugs was performed on either RproAxn, RproEc, or RproRr 4th instar nymphs that were two weeks post molt. Bacteria were injected intrathoracically with 2 μl of 108, 106, or 104 CFU/ml of an overnight culture. For per os infection, bugs were given a blood meal inoculated with cells pelleted from 0.5 ml of 108 CFU/ml of an overnight culture. Nymphs were challenged with either R. rhodnii, M. luteus, E. coli, or sterile Aedes saline. Bacteria were collected by centrifugation at 8,000 x g for 5 minutes and resuspended in sterile Aedes saline. A group of nymphs received a stab wound without injection and a group of nymphs were unaltered and left as a naïve treatment group. Twenty-four hours after injection, whole guts and fat body were dissected out in sterile PBS and stored at -80 °C. For oral infection, 4th instar RproAxn or RproRr nymphs that were two weeks post molt were fed a blood meal containing 106 CFU/ml of either E. coli or M. luteus. Guts and fat body were collected via dissection 24h post-blood meal and tissues were frozen at -80 °C until RNA extraction was performed.

Real-time quantitative PCR (qPCR)

For analysis of gene expression, axenic and gnotobiotic individuals’ total RNA was isolated from homogenized tissues (gut or fat body) using the Direct-zol 96 RNA MagBead Kit (Zymo Research) and KingFisher Apex extraction system. Total RNA was subject to DNAse treatment with the Turbo DNA-free kit (Thermo Fisher Scientific) according to manufacturer’s instructions. Purified, DNased RNA quantification and purity was validated using a NanoDrop spectrophotometer to measure absorbance ratios. One hundred nanograms of RNA was reverse transcribed using iScript Reverse Transcription Supermix (Bio-Rad). qPCR was performed on synthesized cDNA in quadruplicate using the QuantiNova SYBR Green master mix (Qiagen) in a total volume of 20 µl with 0.5 mM of each primer on a Roche LightCycler96 system or a Qiagen Rotor Gene system. For each gene tested, 4 technical replicates and 5 biological replicates were performed. Absolute quantification of genes was performed as previously described [22] using standard curves of pSCA plasmids containing qPCR products. All qPCR primer pairs had an efficiency of > 0.85.

Survival analysis

All bacteria were grown in LB media overnight to an OD600 = 1. E. coli, and M. luteus were grown at 37 °C while R. rhodnii, was grown at 28 °C. Cells were centrifuged at 6,000 rpm and resuspended in sterile Aedes saline to a concentration of 5 x 107 CFU/ml. Fourth instar RproAxn, RproRr, or RproEc nymphs 2 weeks post-molt were injected with 2 µl (106 CFU) of either E. coli, R. rhodnii, M. luteus, or sterile saline with a sterile Hamilton syringe using a Micro4 syringe pump controller (World Precision Instruments). Lower infective doses were tested in RproAxn and RproRr by diluting cultues of E. coli to 102 CFU and 104 CFU. Nymphs were placed individually into wells of sterile 24-well polystyrene cell culture plates for observation and mortality was observed daily for 21 days post-injection.

Bacterial clearance in hemolymph

Bacterial abundance in hemolymph was measured by collecting hemolymph from 4th instar nymphs after challenge with kanamycin resistant E. coli as described above. To test for the presence of R. rhodnii in the hemolymph, hemolymph was collected from RproRr bugs not injected with E. coli. All legs were removed with forceps, then an individual was placed inside a sterilized, filtered p1000 pipette tip inserted into a 2 mL microcentrifuge tube, which was then centrifuged at 2000 RCF for 10 min at 15 °C, resulting in the collection of 2–3 µl of hemolymph from an individual. Hemolymph of 4 individuals per treatment was pooled, diluted in 20 µl of sterile PBS, and spread on LB agar plates with 50 µg/ml of kanamycin sulfate, then incubated for 24 h at 37 °C and the number of CFUs were counted. For each treatment, three biological replicates were performed.

Quantification of antimicrobial hemolymph activity

Antimicrobial activity of kissing bug hemolymph was measured using a zone of inhibition assay as described by [26,27]. A culture of M. luteus was grown overnight in LB at 37 °C, then 1 mL of culture was added to 10 mL of sterile, cooled liquid LB agar. The M. luteus- LB agar solution was mixed and poured into Petri plates. After solidifying, 1 mm diameter holes were created in the agar using a sterile glass Pasteur pipette. Hemolymph was collected from individual nymphs as described above, and 1 µl of hemolymph was placed in each hole and the plates were incubated overnight at 37 °C. The diameters of the individual zones of bacterial growth inhibited were measured using an ocular micrometer on a stereo dissecting microscope. Two independent trials were conducted, and each trial consisted of 10 bugs per treatment group.

Antibiotic clearing/recolonization of R. rhodnii

To investigate the impacts on survival of microbiome recolonization, we inoculated 3rd instar RproAxn bugs with R. rhodnii (RproAxn + Rr) through a blood meal as previously described. Bugs were allowed to develop to the 4th instar when the presence of R. rhodnii was confirmed in a subset of bugs via qPCR. Two weeks after molting bugs were immune challenged with E. coli and survival was monitored as previously described. To determine the effects of removal of R. rhodnii on host immune function, 3rd instar RproRr were fed a bloodmeal containing either 150 µg/ml of kanamycin or a bloodmeal containing 150 µg/ml of kanamycin along with kanamycin resistant R. rhodnii (RproRr + Kan and RproRr-KanR + Kan). Bugs were confirmed to be removed of R. rhodnii or confirmed to still harbor R. rhodnii via qPCR as described above. All bugs were then allowed to molt to the 4th instar then immune challenged as previously described and survival was monitored.

RNAi-mediated immune suppression

PCR Primers containing the minimal T7 promoter sequence were designed to amplify 400–500 bp of relish, dorsal, or egfp (S1 Table). Total RNA was extracted from the fat bodies of 4th instar nymphs, DNased, and reverse transcribed as described above. The PCR product was subsequently cloned to the pSCA vector using the Strataclone PCR cloning kit (Agilent). Target DNA was amplified by PCR from isolated plasmid DNA. dsRNA was synthesized using the MEGAscript RNAi kit (ThermoFisher Scientific) according to the manufacturer’s instructions. Synthesized dsRNA was precipitated with sodium acetate and ethanol, then resuspended to 2 μg/μl in Aedes saline. Fourth instar nymphs were injected with 1 μl of dsRNA, then allowed to recover for 1 week before immune challenge as described previously. Knockdown of relish or dorsal was confirmed by qPCR on cDNA extracted from treated nymphs as described previously.

Hemocyte quantification and inactivation

Hemolymph was collected from 4th instar nymphs that were 2 weeks post-molt. Hemolymph was collected via perfusion of 100 µL of cold Aedes saline injected through the abdomen via insulin syringe. The samples were stored on ice until hemocyte numbers were counted using a Neubauer hemocytometer and an inverted stereo microscope. Two counts per insect were conducted and the average of the two counts was used for each of 5 individuals per treatment.

To reduce hemocyte activity, fluorescent latex microbeads (Polyscience, Fluoresbrite Microspheres 2.00 µm) were diluted to 108 beads/μl in Aedes saline and injected into the hemolymph of 22 4th instar nymphs that were two weeks post-molt. Beads were confirmed to be engulfed by host hemocytes by observation with an epifluorescence microscope (Leica). Four hours after injection with beads, nymphs were challenged with injection of either E. coli or sterile Aedes saline then monitored for survival as previously described.

DOPA conversion assay

Hemolymph was collected as described above from RproAxn, RproEc, or RproRr injected with either E. coli or M. luteus. DOPA conversion was measured as described in [35]. Briefly, 100 µl of perfused hemolymph was suspended into 100 µl of PBS containing 4 mg/ml DOPA, and added to the wells of a sterile 96 well plate. The plate was then incubated at 28 °C for 1 h in a humidified chamber and absorbance was read at 470 nm and on a µQuant plate reader (BioTek), then returned to the chamber and measured again at 24h. Prior to analysis, background absorbance of blank wells was subtracted from the values of test wells. Four to six bugs per treatment were tested.

Statistical analyses

Statistical analysis of insect survival was determined using a Cox-proportional hazards model followed by a log-rank test for pairwise comparisons via R package survminer and survival. Bacterial clearance was analyzed using a Wilcoxon rank-sum test with a Benjamini-Hochberg correction for multiple comparisons. Statistical analysis of gene expression and hemocyte counts was performed using an aligned-rank transformed ANOVA test followed by a Tukey post-hoc test using the R package ARTool [96]. DOPA conversion assay data was analyzed via ANOVA and Tukey post-hoc tests. Data files and R scripts used in the study are found in the supplemental online data.

Supporting information

S1 Fig. Survival of axenic and gnotobiotic insects challenged with different pathogen doses.

https://doi.org/10.1371/journal.ppat.1012947.s001

(EPS)

S2 Fig. Zone of inhibition assays of naïve insects.

https://doi.org/10.1371/journal.ppat.1012947.s002

(EPS)

S3 Fig. Titer of bacteria in antibiotic-treated insects.

https://doi.org/10.1371/journal.ppat.1012947.s003

(EPS)

S4 Fig. qPCR confirmation of dorsal and relish RNAi knockdown.

https://doi.org/10.1371/journal.ppat.1012947.s004

(EPS)

S1 Table. Primers used in this study.

https://doi.org/10.1371/journal.ppat.1012947.s005

(XLSX)

S2 Table. Statistical details of qPCR expression experiments.

https://doi.org/10.1371/journal.ppat.1012947.s006

(XLSX)

S1 Data. Zipped folder of R scripts and datafiles used in the manuscript.

https://doi.org/10.1371/journal.ppat.1012947.s007

(ZIP)

Acknowledgments

We would like to thank Logan Harrell, Bradley Mackett, Betsy Jackson, and Cassandra Armstrong for their assistance in maintaining our colony of R. prolixus. Dr. Michael Strand provided guidance on hemocyte quantification and inactivation experiments. Sandra Mendiola provided helpful feedback on the manuscript.

References

  1. 1. Zhang W, Tettamanti G, Bassal T, Heryanto C, Eleftherianos I, Mohamed A. Regulators and signalling in insect antimicrobial innate immunity: Functional molecules and cellular pathways. Cell Signal. 2021;83:110003. pmid:33836260
  2. 2. Laughton AM, Garcia JR, Altincicek B, Strand MR, Gerardo NM. Characterisation of immune responses in the pea aphid, Acyrthosiphon pisum. J Insect Physiol. 2011;57(6):830–9. pmid:21439291
  3. 3. Salcedo-Porras N, Guarneri A, Oliveira PL, Lowenberger C. Rhodnius prolixus: Identification of missing components of the IMD immune signaling pathway and functional characterization of its role in eliminating bacteria. PLoS One. 2019;14(4):e0214794. pmid:30943246
  4. 4. Nishide Y, Kageyama D, Yokoi K, Jouraku A, Tanaka H, Futahashi R, et al. Functional crosstalk across IMD and Toll pathways: insight into the evolution of incomplete immune cascades. Proc Biol Sci. 2019;286(1897):20182207. pmid:30963836
  5. 5. Silva CP, Silva JR, Vasconcelos FF, Petretski MDA, Damatta RA, Ribeiro AF, et al. Occurrence of midgut perimicrovillar membranes in paraneopteran insect orders with comments on their function and evolutionary significance. Arthropod Struct Dev. 2004;33(2):139–48. pmid:18089029
  6. 6. International Aphid Genomics Consortium. Genome sequence of the pea aphid Acyrthosiphon pisum. PLoS Biol. 2010;8(2):e1000313. pmid:20186266
  7. 7. Ma L, Liu L, Zhao Y, Yang L, Chen C, Li Z, et al. JNK pathway plays a key role in the immune system of the pea aphid and is regulated by microRNA-184. PLoS Pathog. 2020;16(6):e1008627. pmid:32584915
  8. 8. Salcedo-Porras N, Lowenberger C. The innate immune system of kissing bugs, vectors of chagas disease. Dev Comp Immunol. 2019;98:119–28. pmid:31014953
  9. 9. Zumaya-Estrada FA, Martínez-Barnetche J, Lavore A, Rivera-Pomar R, Rodríguez MH. Comparative genomics analysis of triatomines reveals common first line and inducible immunity-related genes and the absence of Imd canonical components among hemimetabolous arthropods. Parasit Vectors. 2018;11(1):48. pmid:29357911
  10. 10. Alejandro A-D, Lilia J-P, Jesús M-B, Henry RM. The IMD and Toll canonical immune pathways of Triatoma pallidipennis are preferentially activated by Gram-negative and Gram-positive bacteria, respectively, but cross-activation also occurs. Parasit Vectors. 2022;15(1):256. pmid:35821152
  11. 11. Cornwallis CK, van ’t Padje A, Ellers J, Klein M, Jackson R, Kiers ET, et al. Symbioses shape feeding niches and diversification across insects. Nat Ecol Evol. 2023;7(7):1022–44. pmid:37202501
  12. 12. Gerardo NM, Hoang KL, Stoy KS. Evolution of animal immunity in the light of beneficial symbioses. Philos Trans R Soc Lond B Biol Sci. 2020;375(1808):20190601. pmid:32772666
  13. 13. Kim JK, Lee JB, Jang HA, Han YS, Fukatsu T, Lee BL. Understanding regulation of the host-mediated gut symbiont population and the symbiont-mediated host immunity in the Riptortus-Burkholderia symbiosis system. Dev Comp Immunol. 2016;64:75–81. pmid:26774501
  14. 14. Vieira CS, Moreira OC, Batista KKS, Ratcliffe NA, Castro DP, Azambuja P. The NF-κB inhibitor, IMD-0354, affects immune gene expression, bacterial microbiota and Trypanosoma cruzi infection in Rhodnius prolixus midgut. Front Physiol. 2018;9:1189. pmid:30233391
  15. 15. Batista KK da S, Vieira CS, Florentino EB, Caruso KFB, Teixeira PTP, Moraes C da S, et al. Nitric oxide effects on Rhodnius prolixus’s immune responses, gut microbiota and Trypanosoma cruzi development. J Insect Physiol. 2020;126:104100. pmid:32822690
  16. 16. Jang S, Ishigami K, Mergaert P, Kikuchi Y. Ingested soil bacteria breach gut epithelia and prime systemic immunity in an insect. Proc Natl Acad Sci U S A. 2024;121(11):e2315540121. pmid:38437561
  17. 17. Kim JK, Lee JB, Huh YR, Jang HA, Kim C-H, Yoo JW, et al. Burkholderia gut symbionts enhance the innate immunity of host Riptortus pedestris. Dev Comp Immunol. 2015;53(1):265–9. pmid:26164198
  18. 18. Kwong WK, Mancenido AL, Moran NA. Immune system stimulation by the native gut microbiota of honey bees. R Soc Open Sci. 2017;4(2):170003. pmid:28386455
  19. 19. Weiss BL, Wang J, Aksoy S. Tsetse immune system maturation requires the presence of obligate symbionts in larvae. PLoS Biol. 2011;9(5):e1000619. pmid:21655301
  20. 20. Benoit JB, Vigneron A, Broderick NA, Wu Y, Sun JS, Carlson JR, et al. Symbiont-induced odorant binding proteins mediate insect host hematopoiesis. Elife. 2017;6:e19535. pmid:28079523
  21. 21. Brecher G, Wigglesworth VB. The transmission of Actinomyces rhodnii Erikson in Rhodnius prolixus stål (hemiptera) and its influence on the growth of the host. Parasitology. 1944;35(4):220–4.
  22. 22. Gilliland CA, Patel V, McCormick AC, Mackett BM, Vogel KJ. Using axenic and gnotobiotic insects to examine the role of different microbes on the development and reproduction of the kissing bug Rhodnius prolixus (Hemiptera: Reduviidae). Mol Ecol. 2023;32(4):920–35. pmid:36464913
  23. 23. Filée J, Agésilas-Lequeux K, Lacquehay L, Bérenger JM, Dupont L, Mendonça V, et al. Wolbachia genomics reveals a potential for a nutrition-based symbiosis in blood-sucking Triatomine bugs. bioRxiv. 2023:2022.09.06.506778.
  24. 24. Batista KKS, Vieira CS, Figueiredo MB, Costa-Latgé SG, Azambuja P, Genta FA, et al. Influence of Serratia marcescens and Rhodococcus rhodnii on the humoral immunity of Rhodnius prolixus. Int J Mol Sci. 2021;22(20):10901. pmid:34681561
  25. 25. Wu J, Wang Q, Wang D, Wong ACN, Wang G-H. Axenic and gnotobiotic insect technologies in research on host-microbiota interactions. Trends Microbiol. 2023;31(8):858–71. pmid:36906503
  26. 26. Haine ER, Pollitt LC, Moret Y, Siva-Jothy MT, Rolff J. Temporal patterns in immune responses to a range of microbial insults (Tenebrio molitor). J Insect Physiol. 2008;54(6):1090–7. pmid:18513740
  27. 27. League GP, Estévez-Lao TY, Yan Y, Garcia-Lopez VA, Hillyer JF. Anopheles gambiae larvae mount stronger immune responses against bacterial infection than adults: evidence of adaptive decoupling in mosquitoes. Parasit Vectors. 2017;10(1):367. pmid:28764812
  28. 28. Baines S. The role of the symbiotic bacteria in the nutrition of Rhodnius prolixus (Hemiptera). J Exp Biol. 1956. Available from: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.548.9092&rep=rep1&type=pdf
  29. 29. Lemaitre B, Hoffmann J. The host defense of Drosophila melanogaster. Annu Rev Immunol. 2007;25:697–743. pmid:17201680
  30. 30. Pham LN, Dionne MS, Shirasu-Hiza M, Schneider DS. A specific primed immune response in Drosophila is dependent on phagocytes. PLoS Pathog. 2007;3(3):e26. pmid:17352533
  31. 31. Li Z, Mao R, Teng D, Hao Y, Chen H, Wang X, et al. Antibacterial and immunomodulatory activities of insect defensins-DLP2 and DLP4 against multidrug-resistant Staphylococcus aureus. Sci Rep. 2017;7(1):12124. pmid:28935900
  32. 32. Zdybicka-Barabas A, Mak P, Klys A, Skrzypiec K, Mendyk E, Fiołka MJ, et al. Synergistic action of Galleria mellonella anionic peptide 2 and lysozyme against Gram-negative bacteria. Biochim Biophys Acta. 2012;1818(11):2623–35. pmid:22705262
  33. 33. Hanson MA, Dostálová A, Ceroni C, Poidevin M, Kondo S, Lemaitre B. Synergy and remarkable specificity of antimicrobial peptides in vivo using a systematic knockout approach. Elife. 2019;8:e44341. pmid:30803481
  34. 34. Lavine MD, Strand MR. Insect hemocytes and their role in immunity. Insect Biochem Mol Biol. 2002;32(10):1295–309. pmid:12225920
  35. 35. Clark KD, Strand MR. Hemolymph melanization in the silkmoth Bombyx mori involves formation of a high molecular mass complex that metabolizes tyrosine. J Biol Chem. 2013;288(20):14476–87. pmid:23553628
  36. 36. Clark KD, Lu Z, Strand MR. Regulation of melanization by glutathione in the moth Pseudoplusia includens. Insect Biochem Mol Biol. 2010;40(6):460–7. pmid:20417279
  37. 37. Fallon JP, Troy N, Kavanagh K. Pre-exposure of Galleria mellonella larvae to different doses of Aspergillus fumigatus conidia causes differential activation of cellular and humoral immune responses. Virulence. 2011;2(5):413–21. pmid:21921688
  38. 38. Sheehan G, Farrell G, Kavanagh K. Immune priming: the secret weapon of the insect world. Virulence. 2020;11(1):238–46. pmid:32079502
  39. 39. Prakash A, Agashe D, Khan I. The costs and benefits of basal infection resistance vs immune priming responses in an insect. Dev Comp Immunol. 2022;126:104261. pmid:34536466
  40. 40. Korša A, Lo LK, Gandhi S, Bang C, Kurtz J. Oral immune priming treatment alters microbiome composition in the red flour beetle Tribolium castaneum. Front Microbiol. 2022;13:793143. pmid:35495655
  41. 41. Cooper D, Eleftherianos I. Memory and specificity in the insect immune system: current perspectives and future challenges. Front Immunol. 2017;8:539. pmid:28536580
  42. 42. Hoang KL, King KC. Symbiont-mediated immune priming in animals through an evolutionary lens. Microbiology (Reading). 2022;168(4). pmid:35442184
  43. 43. Weiss BL, Maltz M, Aksoy S. Obligate symbionts activate immune system development in the tsetse fly. J Immunol. 2012;188(7):3395–403. pmid:22368278
  44. 44. Horak RD, Leonard SP, Moran NA. Symbionts shape host innate immunity in honeybees. Proc Biol Sci. 2020;287(1933):20201184. pmid:32842927
  45. 45. Balmand S, Lohs C, Aksoy S, Heddi A. Tissue distribution and transmission routes for the tsetse fly endosymbionts. J Invertebr Pathol. 2013;112 Suppl(0):S116-22. pmid:22537833
  46. 46. Moreira LA, Iturbe-Ormaetxe I, Jeffery JA, Lu G, Pyke AT, Hedges LM, et al. A Wolbachia symbiont in Aedes aegypti limits infection with dengue, Chikungunya, and Plasmodium. Cell. 2009;139(7):1268–78. pmid:20064373
  47. 47. Kambris Z, Cook PE, Phuc HK, Sinkins SP. Immune activation by life-shortening Wolbachia and reduced filarial competence in mosquitoes. Science. 2009;326(5949):134–6. pmid:19797660
  48. 48. Pan X, Zhou G, Wu J, Bian G, Lu P, Raikhel AS, et al. Wolbachia induces reactive oxygen species (ROS)-dependent activation of the Toll pathway to control dengue virus in the mosquito Aedes aegypti. Proc Natl Acad Sci U S A. 2012;109(1):E23-31. pmid:22123956
  49. 49. Pinto SB, Mariconti M, Bazzocchi C, Bandi C, Sinkins SP. Wolbachia surface protein induces innate immune responses in mosquito cells. BMC Microbiol. 2012;12 Suppl 1(Suppl 1):S11. pmid:22375833
  50. 50. Rancès E, Ye YH, Woolfit M, McGraw EA, O’Neill SL. The relative importance of innate immune priming in Wolbachia-mediated dengue interference. PLoS Pathog. 2012;8(2):e1002548. pmid:22383881
  51. 51. Favia G, Ricci I, Damiani C, Raddadi N, Crotti E, Marzorati M, et al. Bacteria of the genus Asaia stably associate with Anopheles stephensi, an Asian malarial mosquito vector. Proc Natl Acad Sci U S A. 2007;104(21):9047–51. pmid:17502606
  52. 52. Castro DP, Moraes CS, Gonzalez MS, Ribeiro IM, Tomassini TCB, Azambuja P, et al. Physalin B inhibits Trypanosoma cruzi infection in the gut of Rhodnius prolixus by affecting the immune system and microbiota. J Insect Physiol. 2012;58(12):1620–5. pmid:23085484
  53. 53. Whitten M, Sun F, Tew I, Schaub G, Soukou C, Nappi A, et al. Differential modulation of Rhodnius prolixus nitric oxide activities following challenge with Trypanosoma rangeli, T. cruzi and bacterial cell wall components. Insect Biochem Mol Biol. 2007;37(5):440–52. pmid:17456439
  54. 54. Erlandson MA, Toprak U, Hegedus DD. Role of the peritrophic matrix in insect-pathogen interactions. J Insect Physiol. 2019;117:103894. pmid:31175854
  55. 55. Cabrera K, Hoard DS, Gibson O, Martinez DI, Wunderlich Z. Drosophila immune priming to Enterococcus faecalis relies on immune tolerance rather than resistance. PLoS Pathog. 2023;19(8):e1011567. pmid:37566589
  56. 56. Prakash A, Fenner F, Shit B, Salminen TS, Monteith KM, Khan I, et al. IMD-mediated innate immune priming increases Drosophila survival and reduces pathogen transmission. PLoS Pathog. 2024;20(6):e1012308. pmid:38857285
  57. 57. Subasi BS, Grabe V, Kaltenpoth M, Rolff J, Armitage SAO. How frequently are insects wounded in the wild? A case study using Drosophila melanogaster. R Soc Open Sci. 2024;11(6):240256. pmid:39100166
  58. 58. Chambers MC, Jacobson E, Khalil S, Lazzaro BP. Thorax injury lowers resistance to infection in Drosophila melanogaster. Infect Immun. 2014;82(10):4380–9. pmid:25092914
  59. 59. Altincicek B, Vilcinskas A. Septic injury-inducible genes in medicinal maggots of the green blow fly Lucilia sericata. Insect Mol Biol. 2009;18(1):119–25. pmid:19076250
  60. 60. Morejon B, Michel K. The expanded immunoregulatory protease network in mosquitoes is governed by gene coexpression. Proc Natl Acad Sci U S A. 2025;122(18):e2425863122. pmid:40305045
  61. 61. Dekmak AS, Yang X, Zu Dohna H, Buchon N, Osta MA. The route of infection influences the contribution of key immunity genes to antibacterial defense in Anopheles gambiae. J Innate Immun. 2021;13(2):107–26. pmid:33207342
  62. 62. Kanbar G, Engels W. Ultrastructure and bacterial infection of wounds in honey bee (Apis mellifera) pupae punctured by Varroa mites. Parasitol Res. 2003;90(5):349–54. pmid:12684884
  63. 63. Jaenike J, Polak M, Fiskin A, Helou M, Minhas M. Interspecific transmission of endosymbiotic Spiroplasma by mites. Biol Lett. 2007;3(1):23–5. pmid:17443956
  64. 64. Verly T, Costa S, Lima N, Mallet J, Odêncio F, Pereira M, et al. Vector competence and feeding-excretion behavior of Triatoma rubrovaria (Blanchard, 1843) (Hemiptera: Reduviidae) infected with Trypanosoma cruzi TcVI. PLoS Negl Trop Dis. 2020;14:e0008712.
  65. 65. Promislow DEL, Flatt T, Bonduriansky R. The biology of aging in insects: from Drosophila to other insects and back. Annu Rev Entomol. 2022;67:83–103. pmid:34590891
  66. 66. Chaves LF, Hernandez M-J, Revilla TA, Rodríguez DJ, Rabinovich JE. Mortality profiles of Rhodnius prolixus (Heteroptera: Reduviidae), vector of Chagas disease. Acta Trop. 2004;92(2):119–25. pmid:15350863
  67. 67. Ryu J-H, Kim S-H, Lee H-Y, Bai JY, Nam Y-D, Bae J-W, et al. Innate immune homeostasis by the homeobox gene caudal and commensal-gut mutualism in Drosophila. Science. 2008;319(5864):777–82. pmid:18218863
  68. 68. Iatsenko I, Boquete JP, Lemaitre B. Microbiota-derived lactate activates production of reactive oxygen species by the intestinal NADPH oxidase Nox and shortens Drosophila lifespan. Immunity. 2018;49:929-942.e5. https://doi.org/10.1016/j.immuni.2018.09.017 pmid:30446385
  69. 69. Hixson B, Chen R, Buchon N. Innate immunity in Aedes mosquitoes: from pathogen resistance to shaping the microbiota. Philos Trans R Soc Lond B Biol Sci. 2024;379(1901):20230063. pmid:38497256
  70. 70. Dong Y, Manfredini F, Dimopoulos G. Implication of the mosquito midgut microbiota in the defense against malaria parasites. PLoS Pathog. 2009;5(5):e1000423. pmid:19424427
  71. 71. Contreras-Garduño J, Rodríguez MC, Hernández-Martínez S, Martínez-Barnetche J, Alvarado-Delgado A, Izquierdo J, et al. Plasmodium berghei induced priming in Anopheles albimanus independently of bacterial co-infection. Dev Comp Immunol. 2015;52(2):172–81. pmid:26004500
  72. 72. Brown LD, Shapiro LLM, Thompson GA, Estévez-Lao TY, Hillyer JF. Transstadial immune activation in a mosquito: adults that emerge from infected larvae have stronger antibacterial activity in their hemocoel yet increased susceptibility to malaria infection. Ecol Evol. 2019;9(10):6082–95. pmid:31161020
  73. 73. Keshavarz M, Jo YH, Patnaik BB, Park KB, Ko HJ, Kim CE, et al. TmRelish is required for regulating the antimicrobial responses to Escherichia coli and Staphylococcus aureus in Tenebrio molitor. Sci Rep. 2020;10(1):4258. pmid:32144366
  74. 74. Bergin D, Murphy L, Keenan J, Clynes M, Kavanagh K. Pre-exposure to yeast protects larvae of Galleria mellonella from a subsequent lethal infection by Candida albicans and is mediated by the increased expression of antimicrobial peptides. Microbes Infect. 2006;8:2105–12.
  75. 75. Greenwood JM, Milutinović B, Peuß R, Behrens S, Esser D, Rosenstiel P, et al. Oral immune priming with Bacillus thuringiensis induces a shift in the gene expression of Tribolium castaneum larvae. BMC Genomics. 2017;18(1):329. pmid:28446171
  76. 76. Benoit JB, Adelman ZN, Reinhardt K, Dolan A, Poelchau M, Jennings EC, et al. Unique features of a global human ectoparasite identified through sequencing of the bed bug genome. Nat Commun. 2016;7:10165. pmid:26836814
  77. 77. Becchimanzi A, Nicoletti R, Di Lelio I, Russo E. Immune gene repertoire of soft scale insects (Hemiptera: Coccidae). Int J Mol Sci. 2024;25(9):4922. pmid:38732132
  78. 78. Bao Y-Y, Qu L-Y, Zhao D, Chen L-B, Jin H-Y, Xu L-M, et al. The genome- and transcriptome-wide analysis of innate immunity in the brown planthopper, Nilaparvata lugens. BMC Genomics. 2013;14:160. pmid:23497397
  79. 79. Salcedo-Porras N, Noor S, Cai C, Oliveira PL, Lowenberger C. Rhodnius prolixus uses the peptidoglycan recognition receptor rpPGRP-LC/LA to detect Gram-negative bacteria and activate the IMD pathway. Curr Res Insect Sci. 2020;1:100006. pmid:36003603
  80. 80. Sansone CL, Cohen J, Yasunaga A, Xu J, Osborn G, Subramanian H, et al. Microbiota-dependent priming of antiviral intestinal immunity in Drosophila. Cell Host Microbe. 2015;18(5):571–81. pmid:26567510
  81. 81. Kieran TJ, Arnold KMH, Thomas JC 4th, Varian CP, Saldaña A, Calzada JE, et al. Regional biogeography of microbiota composition in the Chagas disease vector Rhodnius pallescens. Parasit Vectors. 2019;12(1):504. pmid:31665056
  82. 82. Sutcliffe AC, Dotson EM. Laboratory maintenance and care of Rhodnius prolixus (Hemiptera: Reduviidae) and other Triatominae. J Med Entomol. 2024;61(6):1345–67. pmid:39298679
  83. 83. Waltmann A, Willcox AC, Balasubramanian S, Borrini Mayori K, Mendoza Guerrero S, Salazar Sanchez RS, et al. Hindgut microbiota in laboratory-reared and wild Triatoma infestans. PLoS Negl Trop Dis. 2019;13(5):e0007383. pmid:31059501
  84. 84. Ponton F, Tan YX, Forster CC, Austin AJ, English S, Cotter SC, et al. The complex interactions between nutrition, immunity and infection in insects. J Exp Biol. 2023;226(24):jeb245714. pmid:38095228
  85. 85. Azambuja P, Garcia ES, Mello CB, Feder D. Immune responses in Rhodnius prolixus: influence of nutrition and ecdysone. J Insect Physiol. 1997;43(6):513–9. pmid:12770413
  86. 86. Omondi ZN, Caner A, Arserim SK. Trypanosomes and gut microbiota interactions in triatomine bugs and tsetse flies: a vectorial perspective. Med Vet Entomol. 2024;38(3):253–68. pmid:38651684
  87. 87. Díaz S, Villavicencio B, Correia N, Costa J, Haag KL. Triatomine bugs, their microbiota and Trypanosoma cruzi: asymmetric responses of bacteria to an infected blood meal. Parasit Vectors. 2016;9(1):636. pmid:27938415
  88. 88. Eberhard FE, Klimpel S, Guarneri AA, Tobias NJ. Exposure to Trypanosoma parasites induces changes in the microbiome of the Chagas disease vector Rhodnius prolixus. Microbiome. 2022;10(1):45. pmid:35272716
  89. 89. Dumonteil E, Pronovost H, Bierman EF, Sanford A, Majeau A, Moore R, et al. Interactions among Triatoma sanguisuga blood feeding sources, gut microbiota and Trypanosoma cruzi diversity in southern Louisiana. Mol Ecol. 2020;29(19):3747–61. pmid:32749727
  90. 90. Gumiel M, da Mota FF, Rizzo V de S, Sarquis O, de Castro DP, Lima MM, et al. Characterization of the microbiota in the guts of Triatoma brasiliensis and Triatoma pseudomaculata infected by Trypanosoma cruzi in natural conditions using culture independent methods. Parasit Vectors. 2015;8:245. pmid:25903360
  91. 91. Soares TS, Buarque DS, Queiroz BR, Gomes CM, Braz GRC, Araújo RN, et al. A Kazal-type inhibitor is modulated by Trypanosoma cruzi to control microbiota inside the anterior midgut of Rhodnius prolixus. Biochimie. 2015;112:41–8. pmid:25731714
  92. 92. Buarque DS, Gomes CM, Araújo RN, Pereira MH, Ferreira RC, Guarneri AA, et al. A new antimicrobial protein from the anterior midgut of Triatoma infestans mediates Trypanosoma cruzi establishment by controlling the microbiota. Biochimie. 2016;123:138–43. pmid:26905205
  93. 93. Vieira CS, Waniek PJ, Castro DP, Mattos DP, Moreira OC, Azambuja P. Impact of Trypanosoma cruzi on antimicrobial peptide gene expression and activity in the fat body and midgut of Rhodnius prolixus. Parasit Vectors. 2016;9:119. pmid:26931761
  94. 94. Castro DP, Moraes CS, Gonzalez MS, Ratcliffe NA, Azambuja P, Garcia ES. Trypanosoma cruzi immune response modulation decreases microbiota in Rhodnius prolixus gut and is crucial for parasite survival and development. PLoS One. 2012;7(5):e36591. pmid:22574189
  95. 95. Araújo CAC, Pacheco JPF, Waniek PJ, Geraldo RB, Sibajev A, Dos Santos AL, et al. A rhamnose-binding lectin from Rhodnius prolixus and the impact of its silencing on gut bacterial microbiota and Trypanosoma cruzi. Dev Comp Immunol. 2021;114:103823. pmid:32800901
  96. 96. Wobbrock JO, Findlater L, Gergle D, Higgins JJ. The aligned rank transform for nonparametric factorial analyses using only anova procedures. Proceedings of the SIGCHI Conference on Human Factors in Computing Systems. New York (NY): ACM; 2011. https://doi.org/10.1145/1978942.1978963