Reactive oxygen species are regulated by immune deficiency and Toll pathways in determining the host specificity of honeybee gut bacteria
Edited by Michael Strand, University of Georgia, Athens, GA; received November 16, 2022; accepted June 26, 2023
Significance
Animals have evolved immune systems to combat pathogens early in their evolutionary history. Here, we demonstrate that the honeybee has co-opted its innate immune system to maintain strict host–gut bacterium specificity, allowing proliferation of only its native strain of the core bacterium Gilliamella. The non-native Gilliamella strain, although sharing a common bacterial ancestor with the native strain, triggered a higher level of prostaglandin, inducing host IMD (immune deficiency) and Toll immune responses. These pathways subsequently regulated the Duox (Dual Oxidase) expression, which generated ROS (reactive oxygen species) as the immune effector, creating a hostile gut environment toward the non-native bacterium. Our study thus provides insights to host mechanisms in maintaining symbiotic specificity.
Abstract
Host specificity is observed in gut symbionts of diverse animal lineages. But how hosts maintain symbionts while rejecting their close relatives remains elusive. We use eusocial bees and their codiversified gut bacteria to understand host regulation driving symbiotic specificity. The cross-inoculation of bumblebee Gilliamella induced higher prostaglandin in the honeybee gut, promoting a pronounced host response through immune deficiency (IMD) and Toll pathways. Gene silencing and vitamin C treatments indicate that reactive oxygen species (ROS), not antimicrobial peptides, acts as the effector in inhibiting the non-native strain. Quantitative PCR and RNAi further reveal a regulatory function of the IMD and Toll pathways, in which Relish and dorsal-1 may regulate Dual Oxidase (Duox) for ROS production. Therefore, the honeybee maintains symbiotic specificity by creating a hostile gut environment to exotic bacteria, through differential regulation of its immune system, reflecting a co-opting of existing machinery evolved to combat pathogens.
Sign up for PNAS alerts.
Get alerts for new articles, or get an alert when an article is cited.
In animals, the gastrointestinal tract often harbors diverse and abundant symbionts (1), which benefit host animals in various ways. Host specificity, the restriction of microorganisms to particular host species, is observed in gut microbes in a wide range of animals, including mammals, birds, and social bees (2–4). Understanding mechanisms in the formation and maintenance of host specificity is important for developing general principles of host–microbe interaction. The host’s immune system plays a key role in defending against pathogens and in regulating the gut microbial composition (5–7). Yet, it remains largely unanswered whether the host immune regulation also functions at a finer scale in differentiating non-native bacteria from its native symbionts, including those derived from closely related host species. Or perhaps native bacteria avoid host immune stimulation or have developed a higher resistance to host immune effectors than non-native bacterial relatives.
Host specificity can be formed through codiversification of host and symbionts, which has been observed in primates (8, 9) and social bees (3). Gut microbes in social bees provide an excellent model to study mechanisms of host specificity. The eusocial corbiculate bees (including honeybees, bumblebees, and stingless bees) share five types of core gut bacteria, including Gilliamella, Snodgrassella, Bifidobacterium, Lactobacillus Firm-4, and Lactobacillus Firm-5 (10, 11). The different corbiculate bee species exhibit host-specific gut microbial compositions (11), and the core gut bacteria also show diverged lineages in different hosts (12–14). The ancestors of these symbionts likely developed symbiotic relationships with the common ancestors of extant social bees (11). Furthermore, host–gut bacteria specificity has been demonstrated between honeybees and bumblebees, where native strains of both core gut bacteria Snodgrassella alvi and Lactobacillus Firm-5 showed a higher inoculation success in the original hosts (12, 15). It is also known that the colonization of gut microbiota activates immune responses in honeybees (16–18), and different Lactobacillus strains stimulate varied gene expressions of the Toll immune pathway (19). However, whether the honeybee immune system is involved in differentiating non-native gut microbes from natives remains to be elucidated, as well as the specific molecular mechanism involved in subsequent regulation of host specificity.
In this study, we investigated the honeybee immune regulatory mechanisms in limiting exotic bacteria. Specifically, we explore whether the colonization of native and non-native Gilliamella is mediated by the host immune response, or by varied bacterial resistance to the same host immune effectors, and identify the key immune pathways and metabolites involved in restraining an exotic Gilliamella strain. We focus on the immune pathways involved in antimicrobial peptides (AMPs) and reactive oxygen species (ROS) generation, which are known to regulate gut microbiota and maintain gut homeostasis in insects (6, 20). Our results showed that the honeybee immune system inhibited the colonization of Gilliamella derived from bumblebee, but not its native strain. The non-native strain stimulated a higher level of prostaglandin (PG) in the honeybee gut and increased expression of genes in the immune deficiency (IMD) and Toll immune pathways. Dual Oxidase (Duox) expression was subsequently up-regulated by IMD and Toll pathways, by which ROS was generated to inhibit the non-native strain. The native strain did not show elevated resistance to host AMPs or ROS but avoided immune stimulation. Our study elucidates the regulatory mechanism of honeybee immune genes in filtering exotic gut bacteria that are closely related to the native strain and provides insights for understanding how hosts maintain their symbiotic specificity.
Results
Host-Specific Gilliamella Strains Trigger Varied Host Transcriptive Responses.
Through cross-host inoculation of bacterial strains derived from honeybees and bumblebees, we revealed apparent host specificity in the honeybee core bacterium Gilliamella. The Gilliamella strains W8127 (from Apis mellifera) and B19101 (from Bombus pyrosoma) were fed to both germ-free (GF) A. mellifera and Bombus terrestris (the phylogenetic relationship of the two strains is provided in SI Appendix, Fig. S1). In honeybee guts, quantitative PCR (qPCR) measurements indicated that the bacterial numbers reached 108 and 106, respectively, for native and non-native strains (Fig. 1A, P = 0.0002, Mann–Whitney test) 5 d upon mono-strain inoculation. We also measured the bacterial numbers using a plating approach, and the colony-forming units results confirmed the colonization advantage of the native bacteria (SI Appendix, Fig. S2A). Whereas in bumblebee guts, bacterial number of bumblebee- and honeybee-origin strains reached 107 and 104, respectively (Fig. 1A, P = 0.0002, Mann–Whitney test). The inoculation advantage for native strains remained, even when they were cocolonized with the non-native strains at only 1/10 abundance of the latter (Fig. 1B). We also measured bacterial numbers in honeybee hindguts at a series of time points (6 h, 12 h, 24 h, 48 h, and 5 d after inoculation). The native W8127 first showed a similar abundance as B19101 at 6 h but began to demonstrate significantly higher abundances after 12 h (SI Appendix, Fig. S2B).
Fig. 1.
This distinct colonization affinity indicated varied host responses to native and non-native bacteria. We compared gene expression of honeybee hindguts for GF, W8127, and B19101 colonization groups at the 12-h time point after inoculation (Fig. 1C), which was the earliest diverging time in bacterial colonization performance, and at the 48-h time point (SI Appendix, Fig. S3), when the honeybee gut community exhibited logarithmic growth (21). The gene expression profiles of the three groups were distinctly separated at both time points (Fig. 1C and SI Appendix, Fig. S3A). Compared to GF honeybees, the colonization of B19101 (bumblebee origin) caused more differentially expressed genes (DEGs) than W8127 (honeybee origin), with 487 shared DEGs (Fig. 1D). The colonization of B19101 resulted in 503 up-regulated genes and 657 down-regulated genes (Dataset S1), while W8127 colonization caused upregulation of 449 genes and downregulation of 372 genes relative to the GF group (Dataset S2). Compared to the W8127 colonized group, the B19101 group showed 37 up-regulated genes, and 276 down-regulated genes enriched in several metabolic pathways, phagosome and lysosome (Fig. 1E and Dataset S3). At 48 h, the B19101 group still showed more DEGs (SI Appendix, Fig. S3B) than the W8127 group when each was compared to the GF group, and the DEGs between W8172 and B19101 groups were enriched in the Hippo pathway and fatty acid metabolism pathways (SI Appendix, Fig. S3C and Datasets S4–S6). The results imply more pronounced host responses to non-native gut symbiont colonization.
The Non-Native Gilliamella Strain Induces More Pronounced Immune Responses in the Host Gut.
A series of differently expressed immune genes were identified in the aforementioned DEGs 12 h (Fig. 1F) and 48 h (SI Appendix, Fig. S3D) after inoculation. Among the typical insect immune pathways (IMD, Toll, and JAK-STAT) (22), the honeybee showed distinct expression patterns of B-gluc1 in the Toll pathway toward the two bacterial strains at both time points (Fig. 1F and SI Appendix, Fig. S3D). On the other hand, IMD pathway genes PGRP-S3 and ankyrin displayed a similar pattern between native and non-native strains when compared with the GF group 12 h after inoculation (Fig. 1F).
We measured the expression of AMPs using qPCR 12 h after inoculation. Our results showed that B19101 colonization triggered significant upregulations of Defensin-1, Abaecin, and Apidaecin than both W8127 and control groups. Conversely, native W8127 colonization produced no significant variations in AMP expression (Fig. 2A). Overall, our results imply that the non-native strain induced more pronounced early host immune responses.
Fig. 2.
The insect AMPs were mainly produced from IMD and Toll pathways, which are conserved in most insect lineages (23, 24), and the honeybee possess orthologs for the core members of both pathways (25). We then examined differential expression for genes belonging to IMD and Toll pathways using qPCR for the 12-h group, including recognition genes (PGRP-LC and B-gluc1), signal transduction genes (imd and spätzle), transcription factors (Relish and dorsal), and regulators (pirk and cactus). For the IMD pathway, the significant upregulations of PGRP-LC, imd, and Relish were observed in the B19101 group compared to the W8127 and control groups (Fig. 2B). The expression of pirk was up-regulated in the B19101 group (Fig. 2B), which is a negative regulator of the IMD pathway in Drosophila (26) and Aedes aegypti (27). For the Toll pathway, B-gluc1, dorsal-1, and cactus-2 were up-regulated in the B19101 group compared to both W8127 and control groups (Fig. 2C). Our results imply that the expression of both IMD and Toll immune signaling genes was more affected by the non-native strain, which might be related to the stronger immune response toward the non-native bacteria.
IMD and Toll Pathways Show Opposite Effects on Native and Non-Native Gilliamella Strains.
We then investigated whether honeybee immune responses could affect colonization success of Gilliamella strains. Newly emerged GF honeybees were fed with a combination of B19101 or W8127 bacterial strain and dsRNAs of Relish or dorsal-1 (Fig. 3A), which encode transcription factors involved in the IMD and Toll pathways, respectively. We fed mixed dsRNAs containing those designed for dorsal-1A and dorsal-1B to the dorsal-1 RNAi group and two different dsRNAs to the Relish RNAi group to avoid off-target effects. The expressions of Relish and two dorsal-1 transcripts (dorsal-1A and dorsal-1B) were significantly reduced in RNAi groups 12 h after dsRNA feeding, when compared with the GFP-RNAi control (Fig. 3 B and C and SI Appendix, Fig. S4). Interestingly, the abundance of B19101 was significantly increased in both RNAi groups (Fig. 3B and SI Appendix, Fig. S4), while that of W8127 did not change in the Relish RNAi group, but significantly reduced in the dorsal-1 RNAi group (Fig. 3C and SI Appendix, Fig. S4). The results showed that the IMD pathway did not affect the colonization of the native strain but inhibited the non-native strain. The Toll pathway facilitated the colonization of the native strain, while inhibiting the non-native strain.
Fig. 3.
Unexpectedly, in the B19101 group (12 h), none of the four AMP genes were reduced in expression after Relish and dorsal knock-down. Most AMP genes remained unaffected, while Defensin-1 and Apidaecin showed a significant increase after Relish and dorsal-1 RNAi, respectively (Fig. 3D). A similar pattern was observed at the 5-d time point, except that Apidaecin also showed significant increases after Relish RNAi (SI Appendix, Fig. S5). The Relish RNAi results for Hymenoptaecin, Abaecin, and Apidaecin were also confirmed by two additional dsRNAs designed to exclude off-target effects, while the response of Defensin-1 appeared to be unstable (SI Appendix, Fig. S4C). These results indicate that IMD and Toll regulation of the non-native bacterial strain did not involve AMPs.
The native W8127 group showed similar results. Relish RNAi did not change W8127 abundance or AMP expression after 12-h and 5-d colonization time points (Fig. 3C and SI Appendix, Fig. S5 B and C). Dorsal RNAi resulted in decreased W8127 abundances (Fig. 3C), but no variation in AMP expressions at the 12-h point (SI Appendix, Fig. S5B) and a slight increase in Defensin-1 (P = 0.043) at the 5-d time point (SI Appendix, Fig. S5C). Thus, variation in AMP expression patterns could not explain changes in W8127 abundance in the dorsal RNAi groups.
In addition, we tested the effects of honeybee AMPs on both Gilliamella strains in vitro. The results from inhibition zone assay and minimum inhibitory concentration showed that Apidaecin-1a and Apidaecin-1b (mature peptides from Apidaecin transcripts) could inhibit growth of both B19101 and W8127 (SI Appendix, Fig. S6). Notably, the non-native B19101 was even more resistant to honeybee Apidaecin than the native W8127, echoing findings in Kwong et al. (17). The other three AMPs did not demonstrate any inhibition to either strain. Apparently, the native W8127 has not developed greater resistance for AMP than the non-native B19101. Taken together, IMD and Toll pathways in the honeybee showed opposite effects on colonization abundance of native and non-native strains, but AMPs were not involved in this regulation process.
IMD and Toll Pathways Inhibit Non-Native Gilliamella Strain via Regulation of Duox and ROS.
Besides AMPs, ROS is also involved in the maintenance of insect microbial homeostasis (6, 20). We measured the ROS level in GF, W8127, and B19101 monocolonized honeybee hindguts. The ROS level significantly increased in the B19101 group since 12 h after inoculation but remained unchanged in the W8127 group (Fig. 4A).
Fig. 4.
Then, we ask whether ROS was produced in the hindgut or diffused from the midgut. In honeybee hindgut, the colonization of B19101 induced significant upregulation of candidate genes potentially involved in ROS generation, Duox and Nox (28, 29), and genes related to ROS degradation, Catalase, and GstD-1 (Fig. 4B), 12 h after inoculation. In contrast, no significant expression change was observed in the W8127 group (Fig. 4B). The results indicated that ROS can be generated in the hindgut.
We also used ROS-sensitive fluorescent probes to characterize the distribution and density of ROS in the guts for the GF, W8127, and B19101 groups at 12 h after inoculation (Fig. 4C). Both the midgut and hindgut were detected with the presence of ROS in the gut epithelium. In the midgut, the ROS level showed no significant elevation under bacterial colonization with W8127 or B19101 (Fig. 4 C, Bottom). In both the ileum and the rectum, ROS levels of the W8127 group (native) showed no significant variation from the GF control, while that of the B19101 group (non-native) exhibited significant elevation (Fig. 4 C, Bottom). In summary, the colonization of non-native bacteria has triggered differentiated production of ROS in the hindgut. However, we cannot rule out the potential involvement of the midgut, as there is a consistent basic ROS level (SI Appendix, Fig. S7), which could be related to generic bacterial responses of the honeybee.
We continued to investigate whether ROS had inhibited B19101. We fed GF honeybees with B19101 and vitamin C (an antioxidant that reduces ROS) and then measured the abundance of B19101. Interestingly, the abundance of B19101 increased in 5 d when GF bees were fed with 1 to 10 mM vitamin C over those fed with only bacteria (Fig. 4D). Congruently, the hindgut ROS level was significantly reduced 5 d after vitamin C treatment and colonization (Fig. 4E).
We then knocked down Duox and Nox, each with at least two sets of dsRNAs, to determine the gene underlying ROS production. Duox RNAi resulted in significant expression reduction 24 h after dsRNA feeding, which was slower than what was observed in Relish and dorsal RNAi (12 h) (Fig. 4F and SI Appendix, Figs. S8 and S9A). Duox silencing led to significant ROS reduction (Fig. 4G and SI Appendix, Fig. S9B) and subsequent increase of B19101 abundance in 5 d (Fig. 4H and SI Appendix, Fig. S9C). Although Nox was lowly expressed in the honeybee hindgut (SI Appendix, Fig. S10A), it still showed increased expression after B19101 colonization (Fig. 4B). However, Nox RNAi conducted at 24 h after dsRNA feeding (SI Appendix, Fig. S10B) did not change the ROS level (SI Appendix, Fig. S10C) or B19101 abundance in 5 d (SI Appendix, Fig. S10D). These results indicate that Duox, but not Nox, generated ROS in the honeybee gut, and the Duox-ROS pathway inhibited the proliferation of the non-native Gilliamella strain B19101. Despite the significant inhibitory effect of ROS on the in vivo colonization of the non-native strain, the hydrogen peroxide resistance assay in vitro showed that W8127 was notably more sensitive to ROS than B19101, when incubated with H2O2 at both 10 mM and 40 mM (SI Appendix, Fig. S11A). The results indicate that the proliferation of the native strain in the honeybee gut is not related to ROS resistance. Instead, the native strain avoids the activation of the Duox-ROS pathway via mechanisms yet unknown.
We then examined how Relish, dorsal, and Duox-ROS might have interacted in B19101 inhibition. Previous studies showed that ROS activated the Toll pathway to control dengue virus in mosquitos (30). Unexpectedly, Duox RNAi on honeybees did not change the expression levels of Relish and dorsal-1(SI Appendix, Fig. S9D), but Relish and dorsal-1 RNAi reduced Duox expression (Fig. 4I and SI Appendix, Fig. S9E) and ROS abundance (Fig. 4J and SI Appendix, Fig. S9F). Furthermore, we explored the regulation mechanism of Relish and dorsal-1 on Duox. Specifically, we examined whether honeybee Duox harbors binding sites for Relish and dorsal, which are NF-κB transcription factors. The prediction results suggested three potential NF-κB transcription factor binding sites on the promotor region of Duox (SI Appendix, Fig. S11B), suggesting a possible Duox regulation mechanism mediated by Relish and dorsal. In summary, ROS generated by Duox was the effector for colonization inhibition of the non-native Gilliamella strain and was regulated by IMD and Toll pathways.
PG Activates Honeybee Immunity to Inhibit Non-Native Gilliamella.
We further showed that different host immune responses toward the two Gilliamella strains might be triggered by key metabolites. To compare metabolite abundance in gut compartments of honeybees that were monocolonized by B19101 and W8127, we conducted untargeted metabolomics by LC-MS analysis after Gilliamella colonization at 12-h and 48-h time points after inoculation. The metabolic features of gut samples were distinctly differentiated between the two groups at both time points (Fig. 5A and SI Appendix, Fig. S12A). At 12 h, we identified 772 metabolites from whole guts of the B19101 and W8127 groups (Dataset S7). Compared with the W8127 group, B19101 colonization resulted in a significant increase in 3 metabolites and a decrease in 3 metabolites (SI Appendix, Fig. S12B). At 48 h, we identified 494 metabolites in the B19101 and W8127 colonized groups (Dataset S8). Compared with the W8127 group, B19101 colonization resulted in significant upregulation of 25 metabolites and downregulation of 22 metabolites (Fig. 5B). Notably, we found that four of the up-regulated metabolites in the B19101 group belonged to eicosanoids (Fig. 5C), of which PGJ2 belonged to PGs. PGB2 also showed 1.4-fold upregulation (P = 0.047; Variable Importance in Projection value = 1.97) 12 h after colonization (SI Appendix, Fig. S12B). Eicosanoids, especially PGs, regulate the initiation of cellular and humoral immune responses in some insects (31, 32).
Fig. 5.
Next, we examined the effect of PGs on host immune response toward B19101 colonization. The B19101 colonized honeybees were fed with naproxen or aspirin, both of which are PG synthesis inhibitors (33, 34). The B19101 abundance in the honeybee hindgut increased significantly in both naproxen-fed and aspirin-fed honeybees (Fig. 5D). And the naproxen results were consistent among groups applied with varied concentrations (SI Appendix, Fig. S13A). For the four genes involved in the immune signaling pathway, PGRP-LC, Relish, B-gluc1, and dorsal-1 that were up-regulated in B19101 colonization, PGRP-LC, Relish, and dorsal-1 were reduced 12 h after naproxen or aspirin feeding (Fig. 5 E–H). B-gluc1 expression was also reduced in the naproxen-fed group, while the aspirin-fed group showed the same trend, albeit without significance (P = 0.06). At the same time, Duox expression (Fig. 5I) and ROS level (Fig. 5J) were also significantly reduced in both PG-inhibited groups. Interestingly, the expression of all four AMP genes remained similar in both PG inhibited groups, when compared with controls (Fig. 5K), supporting our previous finding that AMPs were not involved in the inhibition of non-native bacterium.
Furthermore, we supplemented PG D2 (PGD2) in the naproxen feeding group in B19101 colonized honeybees. PGD2 metabolite 15-Deoxy-d-12,14-PGJ2 showed different abundances in gut metabolomes of the B19101 and W8127 groups. PGD2 addition to the naproxen-fed group decreased B19101 abundance, reaching a similar level as the control group (SI Appendix, Fig. S13B), indicating that PGD2 is functional in non-native strain inhibition. The supplementation of PGD2 did not change the reduction of gene expressions for PGRP-LC, B-gluc1, and Relish caused by naproxen feeding (SI Appendix, Fig. S13 C–E), but a significant recovery for dorsal-1 to the same level found in the control group (SI Appendix, Fig. S13F). For the Duox-ROS system, the expressions of Duox were similar between PGD2-supplemented and control groups, while that of the naproxen-fed group was significantly lower (SI Appendix, Fig. S13G). On the other hand, neither naproxen nor PGD2 feeding changed the expressions of the four AMPs (SI Appendix, Fig. S13H). These results showed that PGD2 was involved in honeybee immunity response toward non-native strain and was related to the Duox-ROS system. Overall, our results demonstrate that PG metabolites regulated immune-related genes of the host, preventing non-native Gilliamella strain from proliferating in the honeybee gut.
Arachidonic acid (AA) is the main precursor of PGs, which is mainly found in the sn-2 position of most membrane phospholipids. AA cannot be synthesized by either the honeybee or Gilliamella strain and must be obtained from food (35). Phospholipase A2 (PLA2) can hydrolyze phospholipids to release AA (36). There is evidence that bacterial infection affects host immunity concomitant with changes in host PLA2 expression (37). As expected, the transcriptome and qPCR results revealed that B19101 colonization also increased the expression of PLA2 genes in the honeybee guts (SI Appendix, Fig. S14 A and B). Furthermore, the expression of PLA2 was significantly and positively correlated with the immune DEGs B-gluc1 and spätzle (Pearson’s correlation, R2 > 0.7 for all the correlation analysis, SI Appendix, Fig. S14C). These results suggested that the production of PGs in the honeybee gut may be attributed to the release of AA from the cell membrane via PLA2 during B19101 colonization.
Taken all together, we proposed a schematic model for the role of honeybee immunity on the host specificity of the gut symbiont Gilliamella (Fig. 6). Compared to the native Gilliamella strain, the non-native strain triggers elevated PG production, which increased expression of immune-related genes in the IMD and Toll pathways. The Duox expression was then induced by Relish and dorsal, producing ROS to inhibit the non-native strain. In this model, PGs might have originated from honeybee PLA2 upregulation and AA generation from phospholipids on the membrane. Both Relish and dorsal might act as transcription factors, which directly bind to Duox. Compared to non-native bacterium, the native Gilliamella strain induces lower PG, which does not trigger subsequent ROS production during inoculation, thereby demonstrating colonization advantage in the host gut. In conclusion, the honeybee immune system reacts differently to native and non-native symbiont strains, favoring the colonization of the native bacterium, while rejecting the non-native lineage. Such an intriguing mechanism eventually leads to host specificity in Gilliamella.
Fig. 6.
Discussion
The interactions between animal hosts and symbiotic microbes have been recognized as a universal phenomenon across the animal kingdom. Recent efforts focused on varied animal models have significantly improved our understanding on the diversity of host–symbiont relationships and evolutionary history of their interactions (3, 8, 38). From a grand view of the tree of life, host–symbiont correlation ranges from seemingly loose and opportunistic co-occurrence to more intimate associations between host and microbes (39, 40). For the host–gut microbiota system, host specificity (2, 3) and phylosymbiosis (41, 42) denote a close interactive relationship developed along the evolutionary history of the host and its bacterial partners. Therefore, elucidating mechanisms underpinning host specificity for these evolutionary events will not only help to understand how cross-kingdom interactions function but also to understand whether similar mechanisms have been driving such reciprocal interferences across the history of life.
Regarding the roles played by the host and symbionts in maintaining this intimate interaction, research has focused on revealing bacterial features that may have caused the observed distinctions in affinity to a particular host (2, 4, 12, 15, 43, 44). Less is understood about the role of the host on bacterial affinity in animals. A few pilot studies suggested that genotypic variations were associated with Gilliamella and Bifidobacterium abundances at the strain level in the honeybee (45) and with the colonization success of Snodgrassella in bumblebees (46). Here, we further demonstrate that the honeybee can differentiate bacterial Gilliamella congeners, where the host maintains a higher affinity to its native bacterial strain through differential regulation of the immune system.
Although previous studies have demonstrated that host immunity plays an important role in resisting pathogenic microbes (47), our study indicates that the honeybees have also used the immune system to facilitate host specificity for its core gut bacteria. Under a codiversification model (11), closely related core bacteria (e.g., all direct descendants of the ancestral Gilliamella that first inhabited host ancestor) in corbiculate bees are unlikely to display strong deleterious effects on other corbiculate bee hosts (e.g., bumblebee Gilliamella toward honeybee). Therefore, we speculate that active bacterial filtering by the host may reflect potential advantages brought in by native bacteria. From an evolutionary perspective, the honeybee may have taken advantage of systems already involved in pathogen resistance to favor native core gut symbionts. Such a mechanism assures a more intimate host–bacterium relationship, without having to reinvent the wheel. Given the conservativeness of animal immunity, it is plausible that such a system is also involved in parallel host lineages, highlighting a working hypothesis for future research.
At the molecular level, our study illustrated the immune regulation model that rejects the non-native bacterial strain in the honeybee gut. By switching on the IMD and Toll pathways, the honeybee host creates a hostile gut environment, via elevated ROS, to prevent exotic bacterium (albeit of close phylogenetic relationship to its own) from proliferating. Bacterial immune resistance is not involved in host specificity.
Our results showed that ROS acts as a key effector in host-specific lineage selection and is regulated by IMD and Toll pathways. ROS is known to be produced mainly by Duox in insect gut immunity (48). Duox expression is regulated by the MEKK1-MKK3-p38-ATF2 pathway in fruit fly (Drosophila melanogaster) (49, 50) and by Mesh through MAPK JNK/ERK phosphorylation cascade in both D. melanogaster and the yellow fever mosquito (A. aegypti) (51). Additionally, Duox is modulated by serotonin in both oriental fruit fly (Bactrocera dorsalis) and A. aegypti (52). Our results thus revealed a mechanism, in which Duox expression is regulated by the IMD and Toll pathways. These pathways are best known for their roles in AMP production (23), and the present study adds yet another regulatory role in the Duox-ROS system.
Our results indicate that AMPs were not related to distinguishing between native and non-native Gilliamella, although the IMD and Toll pathways were involved. As a tissue-specific regulation pattern of AMP generation has been reported in Drosophila (53, 54), we expect that the AMP regulation pathways in the honeybee hindgut might also differ from those described in previous studies, e.g., by Relish or dorsal based on different honeybee tissues (55–57).
The immune regulatory mechanisms of insect to maintain gut microbial homeostasis is likely complex (48, 58). The final fate of the inoculated bacteria shall be determined by the combined effects of all relevant mechanisms. Much of the complexity is still unknown. In the present study, we only investigated the roles of potential immune factors AMPs and H2O2, while additional immune and nonimmune factors might also be involved in regulating host-specific gut bacteria. Regarding bacterial resistance to host immune factors, we showed that the native W8127 was more sensitive to both AMP and H2O2 than B19101. But these tests were conducted in vitro, and how W8127 and B19101 respond to both tested and untested immune factors in vivo remains unclear.
In addition to host regulation, the successful colonization of honeybee core bacteria is clearly influenced by multifaceted factors. For example, Gilliamella also relies on its interaction with Snodgrassella in vivo (15, 59), as well as variations in pollen diet (60) and other nutrients obtained by the host (14). How these parameters interact remains unanswered. Additionally, it is yet unknown how the host immunity regulation mechanism would respond to exotic bacteria that are more closely related to the native strain. For example, do honeybees show different affinity to Gilliamella strains derived from other honeybee species (e.g., the dwarf and giant honeybees) by also activating the same immune system? Moreover, our understanding of the molecular mechanism from the bacterial side that triggers different host immune responses remains to be examined.
Another aspect that deserves in-depth investigation lies in our finding that the elevation of dorsal from Toll pathway promoted the establishment of the native Gilliamella strain, while excluding the exotic bacterium. Bahuguna et al. (61) showed that the Toll pathway promotes gut symbiont colonization in Drosophila and is related to fat metabolism (61), providing a working hypothesis to be tested for the honeybees.
In conclusion, using transcriptomics, metabolomics, RNAi, in vitro, and in vivo experiments, our study reveals a host immune regulation mechanism that maintains host specificity for the native Gilliamella bacterium. We showed that the honeybee host may have co-opted the immune system, which has evolved to battle against pathogens, to strengthen host specificity for the gut symbionts. While the involvement of the immune system per se may suggest conservativeness across animals demonstrated with host–bacterium specificity, the specific molecular mechanism underlying each system deserves further scrutiny.
Materials and Methods
Detailed protocols are available in SI Appendix, SI Materials and Methods. The Gilliamella strain B19101 was isolated from the gut homogenate of bumblebee B. pyrosoma collected in Beijing, China, in 2019. The strain W8127 (62) isolated from the western honeybee was shared by Professor Hao Zheng of China Agricultural University. The genome of B19101 was sequenced with the Illumina NovaSeq platform. Genome assemblies and annotation of B19101 followed pipelines in Su et al. (60). The phylogenetic tree including B19101 and W8127 was built with RAxML (63). Monocolonized bees were obtained by feeding GF bees with pure cultures of W8127 or B19101. Co-colonized bees were obtained by feeding equal volumes of suspension from the two strains (OD600 = 0.1 for the native strain and OD600 =1 for the non-native strain). Gilliamella bacterial loads in gut samples were calculated with bacterial cultivation on agar plates and absolute qPCR using specific single-copy genes YagU (W8127) and YjhH (B19101) as reference. The hindguts of GF honeybees and those monocolonized with B19101 or W8127 were dissected, and RNA-seq was conducted on an Illumina NovaSeq platform. Expression levels of immune genes were determined by qPCR from RNA extracted from the hindgut. The gene silencing in each honeybee was conducted by feeding 3 μL of dsRNA solution mixture, in which the dsRNA and nanocarriers were mixed at a 1:1 ratio. Antimicrobial peptide resistance for both strains was determined by bacterial growth inhibition assay in liquid and inhibition zone assay. The ROS level was determined by staining with DCFDA and the hydrogen peroxide assay kit. Metabolomic profiles were determined from extracted whole guts using LC-MS. Each bee was fed 5 μL of naproxen, aspirin, or PGD2 solution.
Data, Materials, and Software Availability
Sequencing data have been deposited in the bioproject PRJNA874340 in the NCBI (64). The data that support the findings of this study are included in supporting information.
Acknowledgments
The work was supported by the National Natural Science Foundation of China (No. 32000343) to S.L., the National Special Support Program for High-level Talents (Ten-Thousand Talents Program), the 2115 Talent Development Program of China Agricultural University, and the program from the Sanya Institute of China Agricultural University (No. SYND-2021-30) to X.Z. We thank Prof. Hao Zheng from China Agricultural University for providing the W8127 strain. We thank Prof. Jie Shen and Associate Prof. Shuo Yan from China Agricultural University for lending materials and expertise on nanocarriers used in RNAi experiments. We thank Prof. Ken Tan and Associate Prof. Zhengwei Wang from Xishuangbanna Tropical Botanical Garden, Chinese Academy of Sciences, and Yifan Gu from China Agricultural University for their assistance in honeybee sampling. We thank Jiaming Liu and Yashuai Wu from China Agricultural University for their assistance in bacterial cultivation. We thank the Sanya Institute of China Agricultural University for providing essential infrastructure and instruments.
Author contributions
L.G., S.L., and X.Z. designed research; L.G. and J.T. performed research; M.T. contributed new reagents/analytic tools; L.G. analyzed data; and L.G., S.L., and X.Z. wrote the paper.
Competing interests
The authors declare no competing interest.
Supporting Information
Appendix 01 (PDF)
- Download
- 16.74 MB
Dataset S01 (XLSX)
- Download
- 248.21 KB
Dataset S02 (XLSX)
- Download
- 170.46 KB
Dataset S03 (XLSX)
- Download
- 67.20 KB
Dataset S04 (XLSX)
- Download
- 464.83 KB
Dataset S05 (XLSX)
- Download
- 373.72 KB
Dataset S06 (XLSX)
- Download
- 95.32 KB
Dataset S07 (XLSX)
- Download
- 264.53 KB
Dataset S08 (XLSX)
- Download
- 172.17 KB
References
1
R. D. Berg, The indigenous gastrointestinal microflora. Trends Microbiol. 4, 430–435 (1996).
2
P. L. Oh et al., Diversification of the gut symbiont Lactobacillus reuteri as a result of host-driven evolution. ISME J. 4, 377–387 (2010).
3
W. K. Kwong, N. A. Moran, Evolution of host specialization in gut microbes: The bee gut as a model. Gut Microbes 6, 214–220 (2015).
4
S. K. Sheppard et al., Genome-wide association study identifies vitamin B5 biosynthesis as a host specificity factor in Campylobacter. Proc. Natl. Acad. Sci. U.S.A. 110, 11923–11927 (2013).
5
N. H. Salzman et al., Enteric defensins are essential regulators of intestinal microbial ecology. Nat. Immunol. 11, 76–83 (2010).
6
Z. Yao et al., The dual oxidase gene BdDuox regulates the intestinal bacterial community homeostasis of Bactrocera dorsalis. ISME J. 10, 1037–1050 (2016).
7
A. Marra, M. A. Hanson, S. Kondo, B. Erkosar, B. Lemaitre, Drosophila antimicrobial peptides and lysozymes regulate gut microbiota composition and abundance. mBio 12, e0082421 (2021).
8
A. H. Nishida, H. Ochman, A great-ape view of the gut microbiome. Nat. Rev. Genet. 20, 195–206 (2019).
9
A. H. Moeller et al., Social behavior shapes the chimpanzee pan-microbiome. Sci. Adv. 2, e1500997 (2016).
10
W. K. Kwong, N. A. Moran, Gut microbial communities of social bees. Nat. Rev. Microbiol. 14, 374–384 (2016).
11
W. K. Kwong et al., Dynamic microbiome evolution in social bees. Sci. Adv. 3, e1600513 (2017).
12
K. M. Ellegaard et al., Genomic changes underlying host specialization in the bee gut symbiont Lactobacillus Firm5. Mol. Ecol. 28, 2224–2237 (2019).
13
L. Cornet et al., Phylogenomic analyses of Snodgrassella isolates from honeybees and bumblebees reveal taxonomic and functional diversity. mSystems 7, e0150021 (2022).
14
Y. Li, S. P. Leonard, J. E. Powell, N. A. Moran, Species divergence in gut-restricted bacteria of social bees. Proc. Natl. Acad. Sci. U.S.A. 119, e2115013119 (2022).
15
W. K. Kwong, P. Engel, H. Koch, N. A. Moran, Genomics and host specialization of honey bee and bumble bee gut symbionts. Proc. Natl. Acad. Sci. U.S.A. 111, 11509–11514 (2014).
16
O. Emery, K. Schmidt, P. Engel, Immune system stimulation by the gut symbiont Frischella perrara in the honey bee (Apis mellifera). Mol. Ecol. 26, 2576–2590 (2017).
17
W. K. Kwong, A. L. Mancenido, N. A. Moran, Immune system stimulation by the native gut microbiota of honey bees. R. Soc. Open Sci. 4, 170003 (2017).
18
R. D. Horak, S. P. Leonard, N. A. Moran, Symbionts shape host innate immunity in honeybees. Proc. Biol. Sci. 287, 20201184 (2020).
19
H. Lang et al., Specific strains of honeybee gut Lactobacillus stimulate host immune system to protect against pathogenic Hafnia alvei. Microbiol. Spectr. 10, e0189621 (2022).
20
N. Buchon, N. A. Broderick, B. Lemaitre, Gut homeostasis in a microbial world: Insights from Drosophila melanogaster. Nat. Rev. Microbiol. 11, 615–626 (2013).
21
J. E. Powell, V. G. Martinson, K. Urban-Mead, N. A. Moran, Routes of acquisition of the gut microbiota of the honey bee Apis mellifera. Appl. Environ. Microbiol. 80, 7378–7387 (2014).
22
G. Sheehan, A. Garvey, M. Croke, K. Kavanagh, Innate humoral immune defences in mammals and insects: The same, with differences? Virulence 9, 1625–1639 (2018).
23
N. Buchon, N. Silverman, S. Cherry, Immunity in Drosophila melanogaster–from microbial recognition to whole-organism physiology. Nat. Rev. Immunol. 14, 796–810 (2014).
24
W. J. Palmer, F. M. Jiggins, Comparative genomics reveals the origins and diversity of arthropod immune systems. Mol. Biol. Evol. 32, 2111–2129 (2015).
25
J. D. Evans et al., Immune pathways and defence mechanisms in honey bees Apis mellifera. Insect Mol. Biol. 15, 645–656 (2006).
26
A. Kleino et al., Pirk is a negative regulator of the Drosophila Imd pathway. J. Immunol. 180, 5413–5422 (2008).
27
M. Wang et al., Ecdysone signaling mediates the trade-off between immunity and reproduction via suppression of amyloids in the mosquito Aedes aegypti. PLoS Pathog. 18, e1010837 (2022).
28
T. Kawahara, M. T. Quinn, J. D. Lambeth, Molecular evolution of the reactive oxygen-generating NADPH oxidase (Nox/Duox) family of enzymes. BMC Evol. Biol. 7, 109 (2007).
29
S. Kumar, A. Molina-Cruz, L. Gupta, J. Rodrigues, C. Barillas-Mury, A peroxidase/dual oxidase system modulates midgut epithelial immunity in Anopheles gambiae. Science 327, 1644–1648 (2010).
30
X. Pan 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. 109, E23–E31 (2012).
31
Y. Kim, S. Ahmed, D. Stanley, C. An, Eicosanoid-mediated immunity in insects. Dev. Comp. Immunol. 83, 130–143 (2018).
32
A. B. F. Barletta et al., Prostaglandins regulate humoral immune responses in Aedes aegypti. PLoS Negl. Trop. Dis. 14, e0008706 (2020).
33
K. C. Duggan et al., Molecular basis for cyclooxygenase inhibition by the non-steroidal anti-inflammatory drug naproxen. J. Biol. Chem. 285, 34950–34959 (2010).
34
J. R. Vane, Inhibition of prostaglandin synthesis as a mechanism of action for aspirin-like drugs. Nat. New Biol. 231, 232–235 (1971).
35
J. Yu et al., The dietary arachidonic acid improved growth and immunity of honey bee (Apis mellifera ligustica). Bull. Entomol. Res. 112, 1–10 (2021).
36
M. Murakami, Y. Nakatani, G. I. Atsumi, K. Inoue, I. Kudo, Regulatory functions of phospholipase A2. Crit. Rev. Immunol. 37, 127–195 (2017).
37
S. M. Sajjadian, M. Vatanparast, Y. Kim, Toll/IMD signal pathways mediate cellular immune responses via induction of intracellular PLA2 expression. Arch. Insect Biochem. Physiol. 101, e21559 (2019).
38
N. A. Moran, H. Ochman, T. J. Hammer, Evolutionary and ecological consequences of gut microbial communities. Annu. Rev. Ecol. Evol. Syst. 50, 451–475 (2019).
39
M. McFall-Ngai et al., Animals in a bacterial world, a new imperative for the life sciences. Proc. Natl. Acad. Sci. U.S.A. 110, 3229–3236 (2013).
40
A. D. Kostic, M. R. Howitt, W. S. Garrett, Exploring host-microbiota interactions in animal models and humans. Genes Dev. 27, 701–718 (2013).
41
R. M. Brucker, S. R. Bordenstein, The hologenomic basis of speciation: Gut bacteria cause hybrid lethality in the genus Nasonia. Science 341, 667–669 (2013).
42
E. K. Mallott, K. R. Amato, Host specificity of the gut microbiome. Nat. Rev. Microbiol. 19, 639–653 (2021).
43
S. L. Foley, T. J. Johnson, S. C. Ricke, R. Nayak, J. Danzeisen, Salmonella pathogenicity and host adaptation in chicken-associated serovars. Microbiol. Mol. Biol. Rev. 77, 582–607 (2013).
44
S. A. Frese et al., Molecular characterization of host-specific biofilm formation in a vertebrate gut symbiont. PLoS Genet. 9, e1004057 (2013).
45
J. Wu et al., Honey bee genetics shape the strain-level structure of gut microbiota in social transmission. Microbiome 9, 225 (2021).
46
L. A. Sauers, B. M. Sadd, An interaction between host and microbe genotypes determines colonization success of a key bumble bee gut microbiota member. Evolution 73, 2333–2342 (2019).
47
A. P. Lourenco, K. R. Guidugli-Lazzarini, F. C. Freitas, M. M. Bitondi, Z. L. Simoes, Bacterial infection activates the immune system response and dysregulates microRNA expression in honey bees. Insect Biochem. Mol. Biol. 43, 474–482 (2013).
48
S. Bai, Z. Yao, M. F. Raza, Z. Cai, H. Zhang, Regulatory mechanisms of microbial homeostasis in insect gut. Insect Sci. 28, 286–301 (2021).
49
E. M. Ha et al., Regulation of DUOX by the Gαq-phospholipase Cβ-Ca2+ pathway in Drosophila gut immunity. Dev. Cell 16, 386–397 (2009).
50
S. Chakrabarti, M. Poidevin, B. Lemaitre, The Drosophila MAPK p38c regulates oxidative stress and lipid homeostasis in the intestine. PLoS Genet. 10, e1004659 (2014).
51
X. Xiao et al., A Mesh-Duox pathway regulates homeostasis in the insect gut. Nat. Microbiol. 2, 17020 (2017).
52
T. Zeng et al., Serotonin modulates insect gut bacterial community homeostasis. BMC Biol. 20, 105 (2022).
53
P. Tzou et al., Tissue-specific inducible expression of antimicrobial peptide genes in Drosophila surface epithelia. Immunity 13, 737–748 (2000).
54
N. Buchon, N. A. Broderick, M. Poidevin, S. Pradervand, B. Lemaitre, Drosophila intestinal response to bacterial infection: Activation of host defense and stem cell proliferation. Cell Host Microbe 5, 200–211 (2009).
55
H. Schluns, R. H. Crozier, Relish regulates expression of antimicrobial peptide genes in the honeybee, Apis mellifera, shown by RNA interference. Insect Mol. Biol. 16, 753–759 (2007).
56
A. P. Lourenco, M. M. Florecki, Z. L. P. Simoes, J. D. Evans, Silencing of Apis mellifera dorsal genes reveals their role in expression of the antimicrobial peptide defensin-1. Insect Mol. Biol. 27, 577–589 (2018).
57
G. Di Prisco et al., A mutualistic symbiosis between a parasitic mite and a pathogenic virus undermines honey bee immunity and health. Proc. Natl. Acad. Sci. U.S.A. 113, 3203–3208 (2016).
58
X. Pang et al., Mosquito C-type lectins maintain gut microbiome homeostasis. Nat. Microbiol. 1, 16023 (2016).
59
V. G. Martinson, J. Moy, N. A. Moran, Establishment of characteristic gut bacteria during development of the honeybee worker. Appl. Environ. Microbiol. 78, 2830–2840 (2012).
60
Q. Su et al., Significant compositional and functional variation reveals the patterns of gut microbiota evolution among the widespread Asian honeybee populations. Front. Microbiol. 13, 934459 (2022).
61
S. Bahuguna et al., Bacterial recognition by PGRP-SA and downstream signalling by Toll/DIF sustain commensal gut bacteria in Drosophila. PLoS Genet. 18, e1009992 (2022).
62
H. Zheng et al., Division of labor in honey bee gut microbiota for plant polysaccharide digestion. Proc. Natl. Acad. Sci. U.S.A. 116, 25909–25916 (2019).
63
A. Stamatakis, RAxML version 8: A tool for phylogenetic analysis and post-analysis of large phylogenies. Bioinformatics 30, 1312–1313 (2014).
64
L. Guo et al., Data from "Reactive oxygen species are regulated by immune deficiency and Toll pathways in determining the host specificity of honeybee gut bacteria". NCBI. https://www.ncbi.nlm.nih.gov/search/all/?term=PRJNA874340. Deposited 28 Aug 2022.
Information & Authors
Information
Published in
Classifications
Copyright
Copyright © 2023 the Author(s). Published by PNAS. This article is distributed under Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND).
Data, Materials, and Software Availability
Sequencing data have been deposited in the bioproject PRJNA874340 in the NCBI (64). The data that support the findings of this study are included in supporting information.
Submission history
Received: November 16, 2022
Accepted: June 26, 2023
Published online: August 9, 2023
Published in issue: August 15, 2023
Keywords
Acknowledgments
The work was supported by the National Natural Science Foundation of China (No. 32000343) to S.L., the National Special Support Program for High-level Talents (Ten-Thousand Talents Program), the 2115 Talent Development Program of China Agricultural University, and the program from the Sanya Institute of China Agricultural University (No. SYND-2021-30) to X.Z. We thank Prof. Hao Zheng from China Agricultural University for providing the W8127 strain. We thank Prof. Jie Shen and Associate Prof. Shuo Yan from China Agricultural University for lending materials and expertise on nanocarriers used in RNAi experiments. We thank Prof. Ken Tan and Associate Prof. Zhengwei Wang from Xishuangbanna Tropical Botanical Garden, Chinese Academy of Sciences, and Yifan Gu from China Agricultural University for their assistance in honeybee sampling. We thank Jiaming Liu and Yashuai Wu from China Agricultural University for their assistance in bacterial cultivation. We thank the Sanya Institute of China Agricultural University for providing essential infrastructure and instruments.
Author contributions
L.G., S.L., and X.Z. designed research; L.G. and J.T. performed research; M.T. contributed new reagents/analytic tools; L.G. analyzed data; and L.G., S.L., and X.Z. wrote the paper.
Competing interests
The authors declare no competing interest.
Notes
This article is a PNAS Direct Submission.
Authors
Metrics & Citations
Metrics
Citation statements
Altmetrics
Citations
Cite this article
Reactive oxygen species are regulated by immune deficiency and Toll pathways in determining the host specificity of honeybee gut bacteria, Proc. Natl. Acad. Sci. U.S.A.
120 (33) e2219634120,
https://doi.org/10.1073/pnas.2219634120
(2023).
Copied!
Copying failed.
Export the article citation data by selecting a format from the list below and clicking Export.
Cited by
Loading...
View Options
View options
PDF format
Download this article as a PDF file
DOWNLOAD PDFLogin options
Check if you have access through your login credentials or your institution to get full access on this article.
Personal login Institutional LoginRecommend to a librarian
Recommend PNAS to a LibrarianPurchase options
Purchase this article to access the full text.