Increasing evidence reveals that different classes of environmental toxicants can have substantial effects on the function and homeostasis of the immune system of multiple animal species [see reviews (1–7)]. The complexity of the immune system provides a multitude of possible mechanisms by which toxicants may impact immune-related responses. Immunotoxicity may occur through direct interactions with molecules implicated in canonic immunological pathways, or by disrupting the crosstalk between the immune system and other physiological systems (2).

Recent research has shown that energy metabolism and immune responses are intimately connected: Metabolic shifts may directly modulate immune function, and molecules and cells of the immune system regulate metabolic pathways (8). These connections are studied by the recently established research field of immunometabolism (9). Studies in immunometabolism have revealed mitochondria as a key modulator of immune responses (10), which can be also defined as mitoimmunity (11). Some of the main findings regarding the mitochondrial control of immune responses reveal that (i) mitochondrial components, such as mitochondrial DNA and cardiolipin, act as a danger-associated molecular pattern (DAMP) activating cell-autonomous and cell non-autonomous immune responses; (ii) molecules present at the mitochondrial outer membrane form a platform for anti-viral signaling and inflammasome activation; (iii) mitochondrial dynamics (biogenesis, fusion, and fission) have roles in immune-cell activation; and (iv) Krebs cycle intermediates are necessary for immune-related responses and inflammation, in both innate and adaptive immune cells (12). Therefore, it is reasonable to consider that toxicant-induced mitochondrial damage or disruption may have profound effects on the homeostasis of the immune system. Indeed, increasing evidence suggests that mitochondrial dysfunction caused by environmental toxicants is implicated in the onset of several diseases and immune-related disorders (13). Moreover, mitochondria can be particularly vulnerable targets to several classes of environmental toxicants, including metals, polycyclic aromatic hydrocarbons (PAHs), and pesticides (14, 15). Despite this awareness, little is known about how mitotoxicants affect immunometabolic processes.

Caenorhabditis elegans, a free-living nematode, has traditionally been utilized in the areas of neuroscience, aging, and developmental biology. Over the past few decades, this model has also been recognized as a valuable tool in revealing fundamental principles in innate immunity and host-pathogen interactions (16). Most recently, C. elegans was also established as an important model for unraveling key concepts in immunometabolism (17). Most of the success in the use of C. elegans in research relies on the fact that this species is a valuable representative of metazoans, including humans. For example, 83% of the worm proteome contains human homologs (18, 19). Importantly, various molecular pathways involved in immunometabolic responses in mammals are conserved in C. elegans ( Figure 1 ) (17, 20, 21). Moreover, the key role of mitochondria as a signaling hub for innate immune responses also seems to be conserved between vertebrates and C. elegans (22, 23).

Over the past years, C. elegans has also emerged as a prevalent toxicological test organism, being successfully used in drug discovery screens (24) and shown to be predictive of chemical toxicity in mammals, with several reported cases of a conserved mode of toxic action (24). C. elegans has also been proposed as a prominent model in environmental toxicological studies (25, 26). Thus, research using this model may reveal valuable insights regarding immunometabolism toxicity mechanisms and environmental health.

In this study, we exposed nematodes to the pesticide rotenone, a classic complex I inhibitor, to investigate the role of specific mitochondrial disruption on modulating conserved immuno-metabolic molecular pathways and disease susceptibility. We found that rotenone disrupted nematode growth and mitochondrial bioenergetics, as expected, and modulated the expression of genes and pathways mostly involved in detoxification, energy metabolism, and pathogen defense, including genes related to conserved immunometabolic pathways such as UPR^mt and HIF. Among the down-regulated DEG, most of the known genes were involved in immune defense, revealing a major role of rotenone in suppressing immune-related genes, which (according to the WormExp gene overlap analysis tool) included genes upregulated by Pseudomonas aeruginosa resistance. In line with these results, rotenone-exposed animals were more susceptible to P. aeruginosa (PA14) challenges. In contrast, rotenone rendered animals more resistant when challenged with another pathogen, Salmonella enterica (SL1344). Finally, by evaluating the expression of the UPR^mt reporter gene hsp-6, we propose that the UPR^mt pathway might mediate rotenone-induced resistance to S. enterica.

Previous studies from our research group and data from the literature show that different classes of toxicants can promote physiological stress in organisms by changing energy metabolism (14, 38–41) and innate immune responses (1–7, 42–48). However, although the fields of immunometabolism and mitoimmunity have recently been the focus of extensive research, most studies use vertebrate models, and little attention is directed to mitoimmune dysfunction in the context of pollutant exposures.

To start tackling the hypothesis that mitochondrial toxicants can also act as immunometabolic disruptors and alter disease susceptibility, we exposed nematodes to the pesticide rotenone, which is a classic and specific mitochondrial complex I inhibitor. In this study, we opted to perform a developmental rotenone exposure because, at this stage, animals are usually more sensitive to mitochondrial toxicants. This greater sensitivity is because while early developmental stages and growth involve biological processes which require high levels of mitochondrially-derived energy, at these stages, several mitochondrial parameters are limited or greatly reduced (49–53). Moreover, mitochondrial parameters and function are highly variable with developmental stage; thus, it is likely that mitochondrial toxicants can have more profound effects on some developmental windows of susceptibility (14, 41). Furthermore, it is expected and frequently observed that defects should occur during exposures at these early life stages with mitochondrial toxicants (40, 54). Indeed, our results demonstrated that rotenone caused significant impairment in growth ( Figures 2B–D ) and bioenergetic efficiency ( Figures 3A–E ). Associations between growth and decreases in mitochondrial efficiency (respiration coupled to ATP production) have already been reported (55, 56). It is important to highlight, however, that despite these rotenone-induced effects in growth and mitochondrial function, no mortality was observed, and animals were able to fully develop to adulthood and reproductive maturation without affecting total brood size (pilot study data in Supplementary Figure 2 ). Sublethal rotenone exposures as well as mild inhibition of mitochondrial complex I genes have been reported to increase C. elegans lifespan (57–59). Therefore, it is reasonable to believe that, with our rotenone exposure protocol, we did not cause a severe and generalized organismal toxicity but rather a relatively mild mitochondrial stress.

Next, we investigated if this mild and specific mitochondrial stress in rotenone-exposed nematodes could lead to the modulation of immunometabolic pathways through transcriptomic analysis. Gene ontology analysis of the DEG revealed that the great majority of the known genes were implicated in energy metabolism and stress response ( Figure 4A ). Most of the DEG associated with energy metabolism were up-regulated and annotated as involved in fatty acid β-oxidation ( Figure 4C ). This result is consistent with a previous study from our group which also detected, through metabolomics analysis, increased fatty acid β-oxidation in adult nematodes exposed to rotenone (39). This increase might be related to a need to replenish acetyl-CoA levels or to boost ATP production upon Complex I inhibition, although the latter hypothesis was not confirmed in our previous work (39). Fatty acid β-oxidation plays a significant role in immunometabolism. For example, increased fatty acid β-oxidation is important for innate immune responses in higher vertebrates such as macrophage polarization towards an anti-inflammatory state (60). In C. elegans, although, to our knowledge, direct associations between fatty acid β-oxidation and innate immunity have not been yet explored, several studies describe the importance of certain fatty acids for pathogen resistance [see review (17)]. For example, the fatty acid oleate, which is synthesized by fat-6 and fat-7 desaturases, is required in C. elegans for the pathogen-mediated induction of immune defense genes, and disruption of oleate synthesis increase nematode susceptibility to P. aeruginosa (61). The fat-7 gene was highly down-regulated by our rotenone treatment, which could have contributed to our observations of increased susceptibility of the rotenone-exposed nematodes to this same pathogen ( Figure 5A ). Additionally, fat-7 down-regulation was reported to cause growth delay, reduced size, and increased expression of genes in β-oxidation (62, 63). Interestingly, all of these phenotypes were also observed in our rotenone-treated nematodes. These literature reports, together with our findings, indicate that fat-7 down-regulation –thus potentially oleate synthesis disruption– might be a central player governing the rotenone-induced effects on immunometabolic and pathogen susceptibility responses in C. elegans.

The stress-response DEGs were further separated into three subcategories: detoxification, C-type lectins, and pathogens ( Figure 4B ). Among the xenobiotic metabolism (“detoxification”)-related DEG are genes belonging to the Cytochrome P450 (CYP) and UDP-glucuronosyltransferase (UGT) families, which were identified as up-regulated by our rotenone treatment ( Figure 4D ). The enzymatic modification (biotransformation) of rotenone in rats and fish has already been shown to involve oxidation by CYPs (64). Evidence of rotenone detoxification by CYPs was also reported in invertebrates such as insects and helminths (65, 66). While we could not find studies describing the direct involvement of UGTs in rotenone detoxification, similarly to our findings, rotenone also induced several UGT genes in the fruit fly Drosophila melanogaster (67). These results suggest that the mechanisms of rotenone biotransformation might be conserved amongst vertebrates and invertebrates, although in general the C. elegans xenobiotic biotransformation processes, and in particular the substrate specificities of the relevant enzymes, are relatively poorly understood (68). Together with our growth and mitochondrial dysfunction observations, these findings strengthen the suitability of the C. elegans model to address conserved modes of action of mitotoxicants.

It is not surprising that exposure to a xenobiotic that inhibits mitochondrial function elicited a transcriptomic response fitting with metabolic restructuring and xenobiotic metabolism/detoxification. Perhaps more surprisingly, we also saw gene expression changes consistent with altered immune responses. For example, among the other prominent stress-response-related subcategories were “C-type lectins”. These pattern recognition receptors (PRRs) are well-recognized for their role in the innate immune response. They can detect pathogen-derived carbohydrate ligands and activate immune signaling pathways to protect against infection (69, 70). In C. elegans, to our knowledge, the specific function of proteins belonging to this family has not been characterized. Nonetheless, it has been suggested based on sequence analogy, that several C. elegans C-type lectins may have a role in recognizing non-self carbohydrates (71). Most of the C-type lectin DEG in our database were up-regulated ( Figure 4D ), but a few of them were also found to be down-regulated ( Figure 4E ). It would be interesting to investigate if some of these differentially-expressed C-type lectins could be specific to one or the other pathogen used in this study and potentially involved in the rotenone-induced differential pathogen resistance phenotype. Although, to our knowledge, the role of C-type lectins in immunometabolism has not been much explored, there is evidence of metabolic changes of vertebrate immune cells induced through C-type lectin-signaling pathways (72). Thus, future studies should further explore the role of this PRR family in the metabolic regulation of C. elegans.

The final of the three most represented stress-response-related subcategories was “Pathogens”. This subcategory comprised only a few genes that were up-regulated ( Figure 4D ) with a majority that were down-regulated ( Figure 4E ), suggesting a role of rotenone in shifting nematodes to a more immunosuppressive state. To compare how this set of genes is usually modulated during infections, we took advantage of another C. elegans bioinformatics tool, WormExp (35), and performed a gene enrichment analysis using a list comprising only our DEG related to innate immunity and/or pathogen defense (19 genes which are highlighted in Supplementary Tables 1, 2 and from now on described as “immune-related DEG”). We have identified significant overlap with different gene expression data sets related to P. aeruginosa infection and, interestingly, all of these data sets were described as up-regulated upon P. aeruginosa infection ( Supplementary Table 2 ). By compiling all our immune-related DEG that overlapped with these data sets describing genes up-regulated upon P. aeruginosa infection, we obtained a list of fifteen genes in total and found that eleven of these were downregulated upon rotenone treatment. In fact, two of them were among the list of genes confirmed for P. aeruginosa resistance: dod-24 and irg-5 (73). This major role of rotenone in suppressing immune-related pathways and specific genes up-regulated by P. aeruginosa might contribute to the observed increased susceptibility to P. aeruginosa infection ( Figure 5A ). In another study, however, Chikka et al. (74) found that rotenone exposures induced a distinct immunomodulatory response in C. elegans. Rotenone activated the p38-MAPK pathway, which is a conserved immune-related pathway required for resistance to a diverse range of pathogens (75). Interestingly, none of the p38-MAPK-dependent genes that were up-regulated upon their rotenone exposures were differentially-expressed within our dataset. However, it is unknown whether this rotenone-induced p38-MAPK pathway activation altered the ability of nematodes to survive P. aeruginosa infections. Similarly to the findings from Chikka et al. (74), Campos et al. (59) reported that a point mutation in the complex I subunit gene nuo-6 which resulted in a mild impairment of mitochondrial function also activated the p38-MAPK pathway. Moreover, these mutants were more resistant to P. aeruginosa infections (59). Although the results from these two studies may seem to contradict our findings, it is important to note that the exposure conditions from Chikka et al. (74) were very distinct from our study including differences in rotenone concentration, developmental exposure window, and the nematode age at which these gene expressions were assessed. Similarly, in the study from Campos et al. (59), nuo-6 mutants were subjected to a continuous mitochondrial impairment throughout their lifetime and for several generations, thus probably eliciting a series of adaptive molecular mechanisms, which may not be present during a time-limited chemically-induced mitochondrial impairment. Most importantly, however, all these studies are in line with a key role of mitochondrial dysfunction in the modulation of immune pathways in the C. elegans model.

Despite the down-regulation of several immune-related genes induced by rotenone, exposed nematodes did not display increased susceptibility to S. enterica. In fact, in two out of our three experiments, nematodes displayed a clear and significant increased resistance ( Figure 5B and Supplementary Table 4 ). Unfortunately, we were unable to gain any mechanistic insights using the WormExp tool, as we did not identify any significant overlap with gene data sets related to S. enterica infection. This is probably because, to our knowledge, to this date, the WormExp database lacks gene sets with this pathogen species alone (there are some data sets available, but they represent combined treatments, e.g., “UP by S. enterica SL2048 under fer-1 mutant”).

In C. elegans, S. enterica and P. aeruginosa trigger distinct pathogen-defense mechanisms. For example, the ZIP-2 transcription factor is required for resistance against PA14 and regulates the expression of irg-1. irg-1 is up-regulated by PA14, but not S. enterica and other less pathogenic strains of P. aeruginosa (76). Moreover, in C. elegans, the pathogenicity of S. enterica relies on disseminated oxidative stress during infection (77), whereas such prominent oxidative stress levels, to our knowledge, have not been described during P. aeruginosa infections. With such distinct mechanisms of pathogenicity, it is reasonable to expect that host survival will also be dependent on different defense-related pathways.

One mechanism evolved in C. elegans to survive pathogens is an evasion strategy, also known as avoidance behavior, which is chiefly mediated by chemosensory neurons (78). Pathogen avoidance plays a significant role in the C. elegans defense response to P. aeruginosa but not to S. enterica (79). Therefore, one could argue that the differential pathogen susceptibility response caused by rotenone in this study could be related to effects on avoidance behavior. Indeed, chemicals may impact chemosensory behavior controlling the C. elegans ability to be attracted or avoid certain natural products. For example, we have recently shown that a sublethal dose of silver nanoparticles can completely block the C. elegans ability to be attracted to a food-like odor (80). Contributing to this idea is the fact that rotenone is also well known for causing neurotoxic effects (81). Having this in mind, all our pathogen challenges were conducted in Petri plates seeded with each bacterial culture spread over the entire plate (full lawn). With this approach, potential effects on pathogen avoidance are not likely to have affected our survival results.

Colonization and proliferation of bacterial pathogens in the C. elegans intestine causes luminal distension and correlates with C. elegans death (82). For example, S. enterica typically causes a remarkable luminal distention in nematodes, and S. enterica mutants that are less lethal to nematodes exhibit reduced colonization capabilities (82). Thus, it would be interesting to address in future studies if the increased resistance of rotenone-exposed animals to S. enterica could be related to an enhanced ability of the rotenone-exposed worms to destroy and eliminate bacterial cells and prevent gut colonization. Similarly, it would be interesting to test whether the increased susceptibility to P. aeruginosa could be related to a decreased capability of C. elegans exposed to rotenone in controlling intestinal bacterial colonization and proliferation.

Our results suggest that the UPR^mt pathway could be involved in our rotenone-induced differential pathogen susceptibility. This pathway is probably the most studied in C. elegans linking mitochondria and innate immunity. Upon mitochondrial stress, the ATFS-1 transcription factor is translocated to the nucleus and triggers the expression of mitochondrial protective and immune-related genes (22, 59, 83). In our study, this pathway was shown to be up-regulated by rotenone which was evidenced in one of our transcriptomic dataset analyses (“module-weighted annotations” analysis; Supplementary Figure 1 ). This induction was sustained even after 24 h of S. enterica exposure. However, the UPR^mt was suppressed by P. aeruginosa, and rotenone was only able to restore it to the same levels of solvent control animals (not exposed to rotenone) ( Figure 6 ). Indeed, the role of P. aeruginosa in suppressing the C. elegans UPR^mt was already reported (84). Therefore, it is possible that the rotenone-induced activation of UPR^mt was involved in the higher resistance of nematodes to S. enterica but was not sufficiently up-regulated during the P. aeruginosa infection to promote protection. Moreover, P. aeruginosa is known to produce mitotoxic compounds (85, 86) and cause mitochondrial dysfunction (84, 87, 88), which suggests that rotenone-exposed nematodes further infected with P. aeruginosa might be suffering from a double-hit in mitochondria, possibly contributing to high energy deficits and death. Corroborating to this hypothesis, Revtovich et al. (87) showed that C. elegans fed with an E. coli strain that disrupts mitochondrial homeostasis are more susceptible to P. aeruginosa. Nonetheless, future studies are needed to address the double-hit hypothesis between mitotoxicant chemical exposures and pathogens that target mitochondrial function and its role in promoting premature death.

Pesticides are one of the most common and important classes of pollutants in the environment, mostly due to their intensive use in agriculture worldwide (89). Although these substances have societal benefits, their use and limitations in describing their negative effects pose serious risks to human, animal, and environmental health. Importantly, besides pesticides, several other classes of environmental pollutants are also known to cause mitochondrial toxicity (14, 15). Here, we provide evidence that mitotoxicant compounds have the potential to act as immunometabolic disruptors and affect host-pathogen interactions altering disease susceptibility, highlighting the importance of careful regulation of the production and use of mitotoxic chemicals. This work also opens new roads for future studies to further characterize C. elegans immunometabolic pathways, which are currently understudied. It also provides potential pathways implicated in pathogen resistance and susceptibility, which can be further investigated, shedding light on potential therapeutic targets.

The original contributions presented in the study are publicly available. This data can be found here: GEO database; accessions GSE195584; GSM5841177-GSM5841184.

DM designed the experiment, performed the experiments and data analysis, prepared the graphical art, and wrote the main manuscript. CB contributed with experimental design and with performing experiments and data analysis. KF, RC, SB, YW, and AC contributed with experiments and data analysis. LB and JM contributed with experimental design, reagents, materials, and analysis tools. All authors contributed to the article and approved the submitted version.

This work was funded by the National Institute of Health (R01ES028218 and P42ES010356). Some strains were provided by the Caenorhabditis Genetics Center, which is funded by NIH Office of Research Infrastructure Programs (P40 OD010440).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.