The results of this study showed that the virulence determined by the in vitro PCR technique for E. coli virulence factors did not correspond to the virulence of E. coli expressed in C. elegans. It could due to the complex pathogenesis of E. coli, whose ability to exert virulence on the host depends on the balance between the state of the host and the presence and expression of E. coli virulence factors. Therefore, the in vitro results of virulence factors cannot reflect the in vivo conditions and the complex immune reactions after E. coli infection [35,36,37]. In addition, although O serotyping is important to determine the virulence of E. coli [38, 39], the method has its own limitations: antisera produced against specific O groups may cross-react, and another being that some E. coli may lack O antigens and cannot be isolated or identified [40]. The above discussion shows that in vitro detection of E. coli virulence factors and identification of O serotypes cannot reflect the real pathogenicity of E. coli in cattles, the in vitro detection data to determine the virulence of clinically isolated E. coli also lacks accuracy, which must be confirmed by in vivo tests using animal models to verify their virulence. Since it could be expensive and not be practical to use cattle as preliminary testing animals, animal models could offer the possibility of establishing causality links, allowing robust interpretations of the real influence of one system on the other [41, 42]. So we choosed C. elegans and mice to study the virulence of bovine pathogenic bacteria, both of which were very typical model organisms. Our results demonstrated that all 13 strains of bovine-derived E. coli in this study showed the same lethality in C. elegans and Kunming mice, indicating that the virulence of E. coli was the same in nematodes and mice, and the bovine-derived E. coli infection model was successfully established in C.elegans. Moreover, a criterion for determining the strong, medium and weak pathogenicity of E. coli to nematodes was also established by comparing the half lethality time of 13 E. coli strains and the nematodes’ survival rates, which provided a reference standard for the subsequent classification of the pathogenicity of clinical isolated E. coli strains. Since both nematodes and mice could get the same test results, it reflected the importance of predicting the success of a trial at an early stage of research, as it involves time and spending. An SPF mouse needs ¥12 or more and subsequent experimental animal handling fee is also required. The mouse test could get result in one week but cannot guarantee that the pathogenic E. coli can be accurately screened. C. elegans perfectly fits such requirements:it is a valuable tool in pathogenicity and can predict pathogenicity outcomes in mammals. It is inexhaustible and can get result in 2 ~ 3 days. The most important is that mice are mammals, and there are ethical issues involved in the use of them, while C. elegans completely ignores this issue and guarantees that results can be obtained rapidly [43, 44]. So we concluded that C. elegans is more suitable than mice to study the virulence of pathogenic bacteria.
The immune genes and antimicrobial peptides of C. elegans were detected by fluorescence quantitative PCR. The effects of different pathogenic E. coli strains were also analyzed by observing multiple critical genes from nematode immune signaling pathways, which provided a strong theoretical foundations for clinical prevention and the potential anti-infection treatment. The fluorescence quantitative PCR results of this study showed that E. coli with different pathogenicity differed significantly in the regulation of nematode immune gene and antimicrobial peptide expression. In the TGF-β signaling pathway, E. coli with high pathogenicity significantly upregulated nematode Dbl-1 gene expression, while E. coli with weak pathogenicity had almost no effect on Dbl-1 expression, indicating that Dbl-1 expression was directly influenced by the pathogenicity of E. coli [45]. Dbl-1 is an important ligand of the TGF-β signaling pathway in Caenorhabditis elegans and is expressed in the pharynx, subcutaneous tissues and intestine Dbl-1 is a homolog of nematode bone morphogenetic protein 2/4 (BMP2/4), which is involved in regulating nematode growth and development and intrinsic immunity, and is essential for sensory-motor responses [46, 47]. Since Caenorhabditis elegans feeds on bacteria, it is able to learn and recognize the odor of different bacteria, and is able to regulate its interaction with pathogenic bacteria through olfactory learning for those bacteria that are pathogenic or survival threatening [48, 49], and Dbl-1 mutants are usually more sensitive to pathogens, thus it can be inferred that this gene plays an extremely important role in defending against harmful bacteria [50] . Zhang’s findings suggest that nematodes do not induce olfactory aversion when exposed to non-pathogenic bacteria such as Pseudomonas fluorescens, while short-term exposure to the pathogenic bacterium Serratia marcescens induces aversion learning in adults [51]. Moreover, Dbl-1-deficient mutants are defective in avoiding learning of the pathogenic bacterium Serratia marcescens, but the introduction of DNA from the Dbl-1 genome again rescues the mutant’s defect in learning [51] . In addition, Dbl-1 is important for maintaining the natural abundance of Enterobacteriaceae members in the natural microbiota of Caenorhabditis elegans, and deletion of the Dbl-1 gene changes the role of these bacteria from symbiotic to pathogenic [52]. The above studies suggest that this gene is essential for the learned behavior of nematodes in the face of potentially harmful bacterial food for aversive olfaction. This is consistent with the results of the present experiment: when a highly pathogenic E. coli infected the nematode, the organism recognized the bacterium as a harmful bacterium and triggered a significant upregulation of the Dbl-1 gene to enhance the ability of the organism to escape from the harmful bacterium, whereas a weakly pathogenic E. coli was minimally lethal to the nematode, invading the organism, the pathogen was not recognized as a harmful bacterium, and did not trigger olfactory aversion to the pathogen. Thus, no upregulation of Dbl-1 was activated to escape the pathogen.
The p38 mitogen-activated protein kinase (p38 MAPK) signaling pathway of Caenorhabditis elegans is highly conserved with mammals, involved in responses to various physiological stimuli and environmental stresses, and plays an important role in the intrinsic immunity of nematodes. The results of this experiment showed that in the TIR-1 → NSY-1 → SEK-1 → PMK-1 → SKN-1 cascade of the p38 MAPK signaling pathway, the expression of each immune gene in the nematode p38 MAPK signaling pathway was upregulated to different degrees by pathogenic E. coli, among which the upstream Tir-1, Nsy-1, Sek-1 and Pmk-1 were upregulated The up-regulation of Tir-1, Nsy-1, Sek-1 and Pmk-1 was obvious, and the up-regulation of terminal Skn-1 was relatively small, but it was still higher than the expression of Skn-1 in E. coli group with weak pathogenicity. The weakly pathogenic E. coli only activated the upregulation of midstream Nsy-1 and Sek-1, and did not activate the expression of Pmk-1 and Skn-1, indicating that Nsy-1 and Sek-1 are more sensitive in response to E. coli infection, which is consistent with the result of Dennis’ screening of Nsy-1 and Sek-1, two pathogen resistance genes, from a variety of nematode mutants with increased susceptibility to P. aeruginosa lethality, suggesting that both are essential genes for resistance to pathogenic bacteria and play an important role in resistance to pathogen infection [53]. Nsy-1 and Sek-1 are homologs of the mammalian apoptosis signal-regulated kinases ASK1 and MKK3/6, respectively. Nsy-1 is a direct activator of Sek-1 and is expressed in many tissue types, including the intestine [53]. p38 MAPK PMK-1 is activated by TIR-1, MAPKKK NSY-1 and MAPKK SEK-1 and is a key gene in resistance to pathogenic bacterial infection. Bolz et al. were able to enhance Yersinia pestis susceptibility by mutation or RNAi inhibition of Pmk-1 / p38, indicating an important role of Pmk-1 / p38-regulated immune-related effectors in resistance to Yersinia pestis [54]. In mediating responses to pathogens, Salmonella-induced programmed cell death in Caenorhabditis elegans hosts appears to be associated with protective responses, and inactivation of RNAi Pmk-1 blocks Salmonella-induced programmed cell death. The transcription factor Skn-1 is a homolog of the mammalian Nrf protein, and the DNA binding mechanism causes Skn-1 to be less active than other proteins, exhibiting nuclear translocation only when activated by Pmk-1 to function as an activator of the oxidative stress response. The inactivity of Skn-1 clarifies the relatively small upregulation of Skn-1 in this study and the fact that E. coli with high pathogenicity are more sensitive to oxidative stress than E. coli with low pathogenicity. The inactivity of Skn-1 clarified that the up-regulation of Skn-1 in this study was more pronounced in E. coli with relatively low pathogenicity than in E. coli with low pathogenicity.
The insulin-like signaling is an evolutionarily conserved pathway with significant functions in phosphorylation. Insulin-like signaling is well known in controlling metabolism and lifespan growth, which also regulates the immune responses of C. elegans. The Daf-16 is a homolog of the FOXO family in C. elegans, known as the primary transcription factor of nematodes’ insulin-like cascades and the main downstream target of the insulin-like receptor DAF-2 [55]. The Daf-16 is also essential for maintaining nematode longevity in both wild-type and germline-deficient contexts. In C. elegans, either overexpressing Daf-16 or simply increasing Daf-16 protein activities could increase their resistance to various pathogens.
Age-1 is a homologue of the phosphatidylinositol-3-hydroxyl kinase PI3K, which plays a role to some extent in the formation of the Dauer phase, resistance and lifespan direction in nematodes. Age-1 may reduce fertility in hermaphrodites or other unknown metabolic/ physiological changes that has a lifespan extension effect. Mutations of Age-1 causes nematode growth to stagnate at the Dauer stage, shifting metabolism to fat accumulation and extending its lifespan, which similar to the function of mammalian insulin in metabolic regulation. In this experiment, EC3 with high pathogenicity stimulated the expression of nematode Age-1, while EC10 increased the expression of Daf-16 and Age-1. Whereas weak pathogenicity strains EC5 and EC13 induced lower expressions of nematode Daf-16 and Age-1 genes when comparing with high pathogenicity E. coli infected groups. It indicates that the pathogenicity of E. coli directly affected the expression of life span genes in nematodes.
Antimicrobial peptides involve in antimicrobial activity and have important roles in innate immunity of C. elegans [56]. As a respond to pathogens, C. elegans produces specific proteins such as C-type lectins, hydrolases, Lysozyme and Spp-1. The caenopore-1 protein is encoded by Spp-1 gene and specifically expressed in the intestine of C. elegans. The Salmonella infection could strongly enhance the expression Spp-1 and inhibit Salmonella reproduction [57]. When infected with Enterotoxigenic Escherichia coli (ETEC), C. elegans with Spp-1 mutants shown a significantly shorter lifespan compared to wild types, indicating that Spp-1 has an important role in defense ETEC. Antimicrobial factor (Abf) is an antimicrobial peptide identified in Caenorhabditis elegans, and six ABFs (Abf-1 to 6) have been identified from Caenorhabditis elegans [58], which play a direct role in the intrinsic immunity of nematodes. Abf-3 is usually expressed in the intestine, and the results of this experiment indicate that the expression of Spp-1 is more sensitive to the pathogenicity of E. coli, with different pathogenicity The expression of Spp-1 was significantly up-regulated after the infection of nematodes by E. coli with different pathogenicity, which may play an early immune response at the early stage of the organism’s response to foreign pathogens, and is a rapid response and more sensitive defense mechanism. In contrast, the expression of Abf-3 and Clec-60 was positively correlated with the pathogenicity of E. coli, and the expression was significantly up-regulated under the aggression of E. coli with strong pathogenicity, but did not show significant changes for E. coli with weak pathogenicity, indicating that these two immune genes are susceptible to activation only when external adverse stimuli reach a certain level. It was shown that the expression level of the C-type lectin Clec-60 was significantly reduced after Pmk-1 or Sek-1 deletion in response to Acinetobacter candida infection, indicating that the induction range of C-type lectin is highly correlated with the p38 MAPK pathway [31]. The expression of Abf-2, Clec-85, and Lys-7 did not correlate significantly with the pathogenicity strength of E. coli. This study is the first time to isolate strains of bovine E. coli from East and West China, identifying their unique O-serotype types and evaluating their carriAge of virulence factors. Using these bovine E. coli strains, this study also established the first C. elegans pathogenic infection model to establish an in vivo system for efficient, convenient and rapid detection of virulence E. coli infections and to explore the immune responses of these strains among C. elegans. Limited by the time and resources, in this study only three general categories of E. coli were analyzed: the strong, moderate, and weak pathogenicity based on the survival rate and the half-lethal time of the nematodes. In future studies, we will generate a more detailed analyzing system which could including more features in testing the pathogenicity of E. coli strains and to analyzing their immune responses in C. elegans. In this study, only the most and least virulent E. coli strains were used, because we tend to explore the spectrum of immune reactivities for C. elegans and their regulatory effects by pushing the testing condition close to the boarder. For the next stage, we plan to extend our study to those bovine E. coli strains with moderate pathogenicity features, and plan to use both sets of data to create a relevant database of C. elegans. Additionally, there are numerous immune genes and pathways which functional in E. coli infections among nematodes, but with limited resources, only TGF-β, p38 MAPK and the insulin-like signaling pathways were studied together with some classical immune genes could be selected in this study based on previously reported results. We consider these pathways as hubs that controlling nematodes immune reactivities, particularly in immune defense functions. These genes and antimicrobial peptides can be better compared and discussed with previous studies.
Overall, this study demonstrated that C. elegans is an effective infection model in testing the immune responses of diverse kinds of bovine E. coli strains, although in-depth studies still needed to correlate their pathogenic toxicities and underlying immune processes. This set of results also illustrates the virulence of E. coli of bovine origin, and diverse molecular mechanisms that regulate the three immune signaling pathways. We expect this C. elegans model could not only be valuable in bovine sourced bacteria identification and immunology mechanism research, but also contribute for the future control of bovine diseases.