Abstract
To survive in challenging environments, animals must develop a system to assess food quality and adjust their feeding behavior accordingly. However, the mechanisms that regulate this chronic physiological food evaluation system, which monitors specific nutrients from ingested food and influences food-response behavior, are still not fully understood. Here, we established a low-quality food evaluation assay system and found that heat-killed E. coli (HK-E. coli), a low sugar food, triggers cellular UPRER and immune response. This encourages animals to avoid low-quality food. The physiological system for evaluating low-quality food depends on the UPRER (IRE-1/XBP-1) – Innate immunity (PMK-1/p38 MAPK) axis, particularly its neuronal function, which subsequently regulates feeding behaviors. Moreover, animals can adapt to a low-quality food environment through sugar supplementation, which inhibits the UPRER –PMK-1 regulated stress response by increasing vitamin C biosynthesis. This study reveals the role of the cellular stress response pathway as physiological food evaluation system for assessing nutritional deficiencies in food, thereby enhancing survival in natural environments.
Introduction
Food is essential for the survival, growth and fitness of all animals. To adapt to fluctuating environments with a wide range of food sources, animals have developed a food evaluation system. This system enables them to identify nutrient-rich food and avoids low-quality or toxic food, thereby maximizing their survival prospects 1, 2, 3. Various sensory neuron evaluation systems in animals has evolved to evaluate food quality through vision 4, 5, olfactory 3, 6, 7, 8, 9, 10, 11,, 12 and gustatory senses 13, 14, 15. Besides these sensory systems that facilitate quick feeding decisions, animals may also initiate cellular stress response programs to detect nutrition/toxin and trigger food response behaviors 16, 17. This could be one of physiological food quality evaluation systems that monitor the nutritional status of consumed food. However, the signaling events in cellular stress responses involved in evaluating of specific nutrients and the mechanisms that connect these signaling activities to food behaviors are largely unexplored. More specifically, while cellular stress response through UPRER 18 and PMK-1/p38 MAPK 19 dependent immunity in response to pathogens have been extensively studied, the functions of these cellular stress response in sensing and evaluating specific nutrients from food remain unclear.
Vitamin C is an essential micronutrient that cannot be synthesized by humans due to the loss of a key enzyme in the biosynthetic pathway 20. Animals obtain vitamin C from their diet and possibly also from gut microbes 21. Vitamin C is an important physiological antioxidant and a cofactor for a family of biosynthetic and gene regulatory monooxygenase and dioxygenase enzymes. It is also required for the biosynthesis of collagen, L-carnitine, and certain neurotransmitters 22, 23. Vitamin C has been associated with various human diseases including scurvy, immune defect and cardiovascular disease 20. Therefore, in animals, having robust food evaluation systems to detect vitamin C levels could significantly impact their survival in the wild. However, the potential involvement of the cellular stress response pathway in this food evaluation system for sensing and assessing vitamin C remains largely unexplored.
In this study, using the low-quality food evaluation assay system we established 24, we elucidated the mechanism by which the cellular stress response pathway operates as a physiological food evaluation system. This pathway assesses the deficiency of D-glucose in food and the subsequent vitamin C content in animals through the unfolded protein response (UPRER) – innate immunity (PMK-1/p38 MAPK) axis. This mechanism promotes animals to leave low-quality food and is critical for their survival in nature environments.
Results
Low-quality food induces stress response in animals
Our previous studies have shown that Heat-killed E. coli (HK-E. coli), which lacks certain molecules, is considered a low-quality food that is unable to support animal growth 24, 25. Moreover, through metabolic-seq analysis, we identified significant changes in a large numbers of derivatives (Figure 1 – figure supplement 1A, Table S1), including lipids and their derivatives (Figure 1-figure supplement 1B, Table S1), amino acid and their metabolites (Figure 1-figure supplement 1C, Table S1), as well as coenzymes and vitamins (Figure 1-figure supplement 1D, Table S1). Interestingly, we observed a significant decrease in carboxylic acids and their derivatives (Figure 1-figure supplement 1E, Table S1) in E. coli after being heat-killed (Figure 1-figure supplement 1F, Table S1). This suggests that HK-E. coli is nutritionally deficient for C. elegans when compared to normal E. coli food.
Next, we conducted two behavior assays to facilitate the analysis of the food evaluation process in animals by seeding L1 animals in assay plates (Figure 1A and 1B). In the avoidance assay, wild-type animals avoided the HK-E. coli OP50 (HK-OP50) food (Figure 1A). Interestingly, in the food choice assay, animals initially showed no preference between the two types of food (1-2h), but eventually exhibited a preference for high-quality food (Live E. coli) up until the 17h mark (Figure 1B, Figure 1-figure supplement 1G). This suggests that worms depart from the HK-E. coli after recognizing it as low-quality food source through ingestion.
In order to investigate the pathways in animals that respond to HK-E. coli, we performed transcriptomics analysis on worms that were cultured with both HK-E. coli and Live E. coli. Gene-expression profiling revealed that stress response genes, including those related to biotic stimulus, immune response and response to stress, are up-regulated in animals fed with HK-E. coli OP50 (HK-OP50) (Figure 1C, Table S2). Among these up-regulated genes, we identified 11 out of 62 of UPRER target genes (Figure 1D, Figure 1-figure supplement 1H and Table S2). Additionally, about 50%-80% of up-regulated genes overlap with genes responding pathogenic bacteria 26, 27 (Figure 1E, Table S2). Consistent with the results of the RNA sequencing (RNA-seq) analysis, the UPRER reporter (Phsp-4::GFP)28 and immunity reporter (Pirg-5::GFP)29 were strongly induced in intestine (Figure 1F-G) and neurons (Figure 1-figure supplement 2A) by feeding unfavorable food (HK-E. coli OP50), suggesting that UPRER and immune pathways may respond to low-quality food (HK-E. coli OP50). As intestinal fluorescence (Phsp-4::GFP or Pirg-5::GFP) is easy observation and scoring, the further analyses were done in the intestine.
Moreover, UPRMt reporter (Phsp-6::GFP) 30 was weakly induced under HK-E. coli feeding condition (Figure 1-figure supplement 2B), and starved worm did not induce UPRER and immunity (Figure 1-figure supplement 2C-D).
Together, these findings suggest that low-quality food (HK-E. coli OP50) triggers a stress response pathway in animals, including UPRER and innate immune pathway. This implies that animals may assess the quality of food through UPRER and innate immune pathway.
Animals evaluate food quality through UPRER-immune-dependent physiological food quality evaluation system
To determine whether the UPRER and innate immune pathways play a role in evaluating low-quality food, we first examined whether the activation of the UPRER by HK-E. coli was dependent on the known signaling components of the UPRER branches, including IRE/XBP-1, PERK/ATF-4 and ATF-6 31, 32. We observed no difference in Phsp-4::GFP induction with atf-4 (Figure 2-figure supplement 1A) or atf-6 (Figure 2-figure supplement 1B) RNAi-mediated knockdown in animals fed with HK-E. coli. However, knockdown of ire-1/xbp-1 or mutation of xbp-1 reduced GFP fluorescence (Figure 2A, 2B). Among the 11 differentially expressed UPRER target genes in animals fed with HK-E. coli from RNA-seq (Figure 1D), 64% of the genes are IRE-mediated genes (Figure 1 – figure supplement 1H, Table S2). The mRNA level of IRE-1-mediated splicing of xbp-128 is also induced in animals fed with HK-E. coli OP50 (Figure 2-figure supplement 1C). However, UPRER is not affect in animals feeding live-E. coli by RNAi of ire-1, xpb-1, atf-4 and atf-6 (Figure 2-figure supplement 1D). These data suggest that activation of the UPRER by low-quality food (HK-E. coli) depends on the IRE-1/XBP-1.
To further analyze whether XBP-1-dependent UPRER activation is critical for animals to leave low-quality food, we tested food avoidance behavior using xbp-1 mutant. The results show that xbp-1 mutants had a significantly decreased likelihood of leaving of HK-E. coli, which was rescued by expressing xbp-1 in neuron rather than intestine (Figure 2C). This indicates that XBP-1-dependent UPRER activation in neuron is critical for animals to specific evaluate low-quality food (HK-E. coli).
We then investigated which innate immune pathway is involved in evaluating low-quality food. First, we analyzed HK-E. coli induced genes from RNA-seq. Among these up-regulated genes, 82 out of 409 of PMK-1-dependent genes 33 were identified (Figure 2-figure supplement 1E, Table S2). Second, we confirmed the induction of several well-known PMK-1 target genes in RNA-seq data34 (Figure 2-figure supplement 1F) and reporter analysis (Figure 2-figure supplement 1G-H). Moreover, the induction of Pirg-5::GFP was abolished in pmk-1 knockdown animals fed with HK-E. coli (Figure 2D). Third, we found that the phosphorylated PMK-1 (p-PMK-1) level was prominently increased in wild-type N2 animals fed HK-E. coli compared to feeding E. coli OP50 (Figure 2E). Finally, pmk-1 mutants had a decreased likelihood of leaving of HK-E. coli, which was rescued by expressing pmk-1 in neurons rather than intestine (Figure 2F). Moreover, Pirg-5::GFP is not affect in animals feeding live-E. coli by RNAi of pmk-1 (Figure 2-figure supplement 1I).These data suggest that PMK-1 regulated immune pathway evaluates low-quality food, especially the neuronal PMK-1 has a critical function for food quality response.
XBP-1 and PMK-1 are in the same pathway for evaluating food quality
Next, we explored the connection between UPRER (IRE-1/XBP-1) and innate immunity (PMK-1 p38 MAPK) in food quality evaluation. We found that Pirg-5::GFP induction (Figure 2-figure supplement 1I, 2A) and PMK-1 activation (Figure 2E) were decreased in animals with xbp-1 mutation or knockdown when fed with HK-E. coli, suggesting that XBP-1 could regulate PMK-1 under this condition. Additionally, Phsp-4::GFP induction under HK-E. coli was not affected in animals with pmk-1 RNAi (Figure 2-figure supplement 1D, 2B), indicating that XBP-1-dependent UPRER activation is not regulated by PMK-1. Finally, we constructed a double mutant of xbp-1 and pmk-1 and found that the food avoidance phenotype of the double mutant was similar to the pmk-1 mutant (Figure 2G), indicating that PMK-1 is downstream of XBP-1 in responding to low-quality food.
We then asked whether UPRER (IRE-1/XBP-1) – Innate immunity (PMK-1/p38 MAPK) axis is specific to evaluate low-quality food (HK-E. coli). We performed behavior assay in N2, pmk-1 and xbp-1 mutant animals by feeding normal E. coli food, inedible food (Saprophytic staphylococci) 35 and pathogenic food (Pseudomonas aeruginosa-PA14) 18. We found that N2, pmk-1, and xbp-1 mutant worms did not exhibit avoidance behavior when presented with normal food (OP50). However, both N2 and xbp-1 mutant worms were able to escape from inedible food (N2 was predominantly found on the border areas of the bacterial lawn and xbp-1 mutant worms on border and in), Saprophytic staphylococci, whereas pmk-1 mutant worms did not exhibit this avoidance behavior. Notably, N2 and xbp-1 mutant worms exhibited even more pronounced avoidance behavior when exposed to Pseudomonas aeruginosa, whereas pmk-1 mutant worms were more susceptible to infection by this pathogen (Figure 2-figure supplement 2C). These findings suggest that the UPR-Immunity pathway plays a crucial role in helping animals avoid low-quality food (HK-E. coli) by triggering an avoidance response. In contrast, the Innate immunity pathway, which is mediated by PMK-1/p38 MAPK, appears to play a key role in evaluating unfavorable food sources, such as HK-E. coli, Saprophytic staphylococci, and Pseudomonas aeruginosa, and helping animals avoid these environments.
Sugar deficiency in HK-E. coli food induces stress response and avoidance behavior in animals
We then investigated which nutrients/metabolites are sensed by animals through the XBP-1-PMK-1 axis for food quality evaluation. First, we hypothesized that the nutrition status is improved in E. coli mutant (HK-treatment), which could inhibit UPRER and immune response in animals. We established a system for screening the E. coli mutant Keio library (Figure 3 – figure supplement 1A), and identified 20 E. coli mutants that did not induce Phsp-4::GFP through the UPRER reporter (Phsp-4::GFP) after three rounds of screening (Table S3). From these 20 E. coli mutants, we identified 9 E. coli mutants that did not induce Pirg-5::GFP through the immunity reporter (Pirg-5::GFP) screening (Figure 3-figure supplement 1B-C, Table S3). Animals fed HK-yfbR, which catalyzes carbohydrate derivative metabolic process36, had a decreased ability to leave food (Figure 3A, Table S3), indicating that HK-yfbR may be a higher quality food for animals compared to HK-K12.
Secondly, we performed a metabolomics analysis of different quality food (HK-K12, HK-yfbR and Live-K12). We found that the level of 13 metabolites were similar between HK-yfbR and Live-K12, but significantly changed in HK-K12 (Figure 3B, Figure 3-figure supplement 2A, and Table S1). We also found that genes involved in glycolysis/gluconeogenesis were up-regulated in animals fed with HK-E. coli (Figure 3-figure supplement 2B), suggesting that glycolysis/gluconeogenesis metabolism is disordered in animals fed with HK-E. coli, which may result from changes in sugar/carbohydrate intake. The carbohydrates (D-trehalose, lactose, and D-(+)-sucrose) were also decreased in HK-E. coli (Figure 3B), suggesting that carbohydrate deficiency may induce stress response and avoidance behavior in animals feeding HK-E. coli.
Thirdly, to determine which carbohydrate inhibits stress response in animals, we supplemented each metabolite to HK-E. coli and found that only Lactose, and D-(+)-sucrose inhibited HK-E. coli induced UPRER (Figure 3C and 3D, Figure 3-figure supplement 2C). Moreover, we found from our metabolomic data that the sugar level, including lactose, and D-(+)-sucrose, and D-(+)-glucose, was also decreased in HK-E. coli (Figure 3B, Table S1). Since lactose and D-(+)-sucrose are hydrolyzed to produce glucose37, 38, we wondered whether glucose also inhibits the stress response in animals. We found that D-(+)-glucose supplementation also inhibited HK-E. coli induced UPRER (Figure 3E), immune response (Figure 3F, 3G and Figure 3-figure supplement 2D) and avoidance (Figure 3H). Moreover, sugar supplementation did not affect UPRER and immunity in normal food (OP50) or starved condition (NGM) (Figure 3 – figure supplement 2E-F). While sugar effectively inhibits the HK-E. coli-induced UPRER and immune response, it does not fully suppress it to the extent observed with live-E. coli (Figure 3C-F). This implies that additional nutrients present in live-E. coli might also contribute to the inhibition of UPRER and immune response.
Previous studies have shown that heat-killed E. coli (HK-E. coli) is a low-quality food source that cannot support the growth of C. elegans larvae24, 25, whereas supplementation with vitamin B2 (VB2) can restore animal growth24. Here, we found that sugar deficiency in HK-E. coli induces the UPRER-immune response and avoidance behavior in C. elegans. Given this, we investigated whether sugar supplementation could promote animal growth when fed HK-E. coli. To our surprise, supplementing HK-E. coli with carbohydrates (D-Glc, D-GlcA) did not support animal development (Figure 3-figure supplement 2G), suggesting that carbohydrates are not essential for supporting animal growth on this food source. However, we did find that carbohydrates are critical for inhibiting the UPRER-immune response induced by sugar deficiency in HK-E. coli.
Together, these findings suggest that HK-E. coli induces a stress response and avoidance behavior in animals, which can be inhibited by D-(+)-glucose supplementation. This implies that animals may evaluate the sugar deficiency from HK-E. coli through the activation of UPRER and immune responses.
Animals could overcome a low-quality food environment by sugar supplementation through vitamin C biosynthesis
We discovered that D-(+)-glucose supplementation inhibited HK-E. coli induced UPRER (Figure 3E), immune response (Figure 3F, 3G and Figure 3-figure supplement 2D) and avoidance (Figure 3H). Simultaneously, vitamin C (VC), which is synthesized by glucuronate pathway using D-glucose 39, 40 (Figure 4A), was found to contribute to neuroprotective41, 42, immune defense43, 44, and inhibits inflammatory and ER stress45, 46. This led us to question whether the vitamin C biosynthesis pathway is involved in evaluating low-quality food by using D-glucose.
Firstly, we observed an increase in the vitamin C level in C. elegans when fed with HK-yfbR (Figure 4B), a high carbohydrate food compared to HK-E. coli (Figure 3B and Figure 1-figure supplement 1E). However, the VC level in bacteria is the same (Figure 4-figure supplement 1A). The VC level also increased when D-glucose (Figure 4-figure supplement 1B) or D-glucuronate (D-GlcA) was added to HK-E. coli (Figure 4C), which was abolished by knocking down VC biosynthesis genes (Figure 4C and Figure 4-figure supplement 1B). This suggests that addition of sugar supplementation promotes VC synthesis in animals fed with HK-E. coli.
Secondly, we hypothesized that animals could overcome a low-quality food (HK-E. coli) environment by inhibiting the stress response through increasing vitamin C biosynthesis. We found that VC or D-glucuronate (D-GlcA) supplementation inhibits HK-E. coli induced UPRER (Figure 4D), immune response including irg-5/sysm-1 reporter expression (Figure 4E and Figure 4 – figure supplement 1C) and p-PMK-1 (Figure 4F and 4G), as well as food avoidance (Figure 4H).
Finally, we asked whether inhibition of stress response and avoidance by sugar supplementation depends on the vitamin C biosynthesis pathway. We found that suppression of HK-E. coli induced UPRER (Figure 4I), immune response (Figure 4J and Figure 4-figure supplement 1D) and food avoidance (Figure 4K) by D-GlcA/sugar supplementation was abolished in animals with RNAi of VC biosynthesis genes. Food selection behavior assays showed that D-GlcA/sugar supplementation increased the preference for heat-killed bacteria, which was also suppressed by knocking down VC biosynthesis genes (Figure 4-figure supplement 1E). However, VC supplementation still suppressed the UPRER (Figure 4I), immune response (Figure 4J and Figure 4-figure supplement 1D) and food avoidance (Figure 4K), and increased the food preference (Figure 4 – figure supplement 1E) in animals with or without RNAi of VC biosynthesis genes. This suggests that VC, as the final metabolite synthesized from D-glucose, is critical for evaluating low-quality food response in animals.
Together, these data indicate that the vitamin C biosynthesis pathway is critical for evaluating whether food is of higher quality and can be eaten by animals. It also suggests that animals could improve their VC levels to adapt to bad food environment.
Animals evaluate sugar and vitamin C through neuronal XBP-1 and PMK-1
As D-GlcA/sugar and VC supplementation suppressed HK-E. coli induced UPRER, immune response and food avoidance behavior, we investigated whether animals evaluate sugar and VC through XBP and PMK-1 dependent pathways. We performed a food selection behavior assay by adding D-Glc, D-GlcA or VC to the NGM, E. coli and HK-E. coli (Figure 5A). The food selection behavior assays revealed that supplementation with D-Glc, D-GlcA, or VC inhibits the animals’ choice of sugar or VC on E. coli-OP50 feeding conditions (Figure 5-figure supplement 1A). This suggests that supplementation with D-Glc, D-GlcA, or VC may alter the metabolites of live bacteria, leading to avoidance by the animals. There was no preference observed on NGM plate (no food condition) supplementation with D-Glc and VC (Figure 5-figure supplement 1B), indicating that the intake of sugar or VC alone does not influence animal preference. However, alone D-GlcA could influence worm physiology which induces preference change (Figure 5-figure supplement 1B). Interestingly, D-Glc and D-GlcA (Figure 5B and 5C) or VC (Figure 5D) supplementation increased the preference for heat-killed bacteria, which was suppressed in xbp-1 or pmk-1 mutant animals. However, this preference was also rescued in xbp-1 or pmk-1 mutant animals by expressing XBP-1 or PMK-1 in neurons rather than intestine (Figure 5B-D), indicating that neuronal XBP-1 and PMK-1 are critical for physiological food elevation system for monitoring the level of sugar and VC under low-quality food condition.
Discussion
To better survive, animals must evolve a system to recognize and evaluate the quality of their food. This includes the sensory neuron evaluation system for immediate response and feeding decision 2, as well as physiological food evaluation system for chronic response to ingested food. In our previous study, we discovered that the TORC1-ELT-2 pathway, acting as master regulators in intestine, evaluates vitamin B2 deficiency in low-quality food (HK-E. coli) and regulates gut digestive activity to impact animal’s food behavior 24. To further identified the mechanism by which animals evaluate low-quality food (HK-E. coli), we performed metabolomics and transcriptomics analyses to identify specific nutrition deficiencies in low-quality food and the cellular response pathways that are involved in food evaluation pathway. This study identified a physiological food evaluation mechanism by which animals recognize food quality through UPRER (IRE-1/XBP-1) – Innate immunity (PMK-1/p38 MAPK) regulated cellular stress response program in neurons that dictates food avoidance and selection behaviors (Figure 6).
One of the cellular stress response pathways, the Unfolded Protein Response (UPRER), is activated by various stresses, including infection and nutrition deficiency, which disrupt the homeostasis of the endoplasmic reticulum (ER)31. The activation of UPRER, specifically in the nervous system, has been shown to promote changes in feeding and foraging behavior47. The p38 PMK-1 pathway is also crucial for regulating the expression of secreted innate immune effectors and is essential for survival during infection48. Therefore, these two pathways play a critical role in ensuring animals’ survival in changing environments. However, it is still unclear whether UPRER and innate immunity evaluate food quality under physiological conditions. Our study provides evidence that low-quality food (HK-E. coli) activates both UPRER and p-PMK-1, leading to animals leaving the low-quality food. Previously finding have shown that inhibition of ER function promotes C. elegans avoid to the toxic food, which employs the MLK-1/MEK-1/KGB-1 pathway49. Notably, HK-E. coli induced avoidance behavior is independent of the KGB-1 pathway (Figure 5-figure supplement 1C). Additionally, our study reveals that neuronal UPRER and PMK-1 are essential for evaluating low-quality food, suggesting that the nervous system plays a critical role in assessing food quality.
Previous studies have shown that XBP1 deficiency in intestinal epithelial cells leads to IRE1a hyperactivation and increased JNK phosphorylation in the epithelial compartment in vivo50. The IRE1-XBP1 axis has been identified as a critical protective branch of the Unfolded Protein Response (UPR) induced secondary to an innate immune response in the presence of P. aeruginosa 18, 51. The p38 MAPK has also been shown to directly act on the phosphorylation of IRE-1 to promote the stress response52, 53. Interestingly, IRE-1 has been found to confer cold resistance independently of XBP-1 by activating JNK-1 MAPK49. In contrast, our study reveals a new mechanism where the UPRER (IRE-1/XBP-1) positively regulates Innate Immunity (PMK-1/p38 MAPK) under HK-E. coli food conditions, establishing a novel physiological food evaluation system that activates the cellular stress response program.
A previous study has shown that activating innate immunity (PMK-1 MAPK) leads to a reduction in translation54. Our own previous research has also demonstrated that PMK-1 activation causes a shutdown of food digestion in animals35, likely to reduce protein translation and cellular metabolism. To investigate this further, we measured the translation level of animals fed with HK-E. coli and found that total translation ability is significantly reduced in these animals (Figure 5-figure supplement 1D). This finding suggests that activating innate immunity (PMK-1 MAPK) may serve as a mechanism to slow down translation progress, thereby alleviating the pressure on the unfolded protein response (UPR) and preventing excessive UPRER activation.
Vitamin C (VC) is an important physiological antioxidant and a cofactor for a family of biosynthetic and gene regulatory monooxygenase and dioxygenase enzymes. It is also required for the biosynthesis of collagen, L-carnitine, and certain neurotransmitters 22, 23. Meanwhile, VC helps animals to protect neuron41, 42, defend excessive immune 43, 44, and inhibit inflammatory and ER stress45, 46 in order to better survive. The synthesis of vitamin C (VC) occurs through the glucuronate pathway, utilizing D-glucose as a precursor 39, 40 (Figure 4A). This led us to investigate whether the vitamin C biosynthesis pathway is involved in evaluating low-quality food by using D-glucose. In this study, we found that animals feeding live E. coli, which should produce more VC, exhibit higher glucose levels. However, our results show that animals maintain similar VC levels when fed ideal food (live E. coli) compared to low-quality food (HK-E. coli) (Figure 4B), suggesting that animals do not stimulate VC biosynthesis under favorable food conditions. In contrast, when animals are fed low-quality food (HK-OP50), we found that supplementing D-GlcA in HK-E. coli or E. coli-yfbR mutation can improve VC levels and inhibit UPRER-immunity (Figure 4C). These data indicate that glucose boosts the animal’s ability to adapt to unfavorable food environments by increasing VC levels, but not in favorable food conditions.
Unlike the sensory neuron evaluation system, which permits rapid feeding decisions through smell and taste, the cellular stress response as physiological food evaluation system describe here requires a slow and multi-step signal transduction process after the ingestion of food. The disruption of cellular homeostasis by ingested of low-quality or toxic food can activate stress response mechanisms that both increase the cellular ability to withstand and adapt to this disruption of homeostasis and promote behavioral strategies to avoid these conditions and lessen their impact on the organism. These cellular stress response mechanisms include heat shock response, unfolded protein response, oxidative stress response 55. Therefore, this slow physiological food evaluation system is an evolutionary adaptation mechanism for detecting nutrition deficiencies in food that was not detected by quick sensory nervous system.
One limitation of our study is the lack of explanation for why HK-E. coli activates UPRER and immunity. We hypothesized that when heat-killed, HK-E. coli may lack or contain altered levels of certain metabolites that either activate or inhibit UPRER and immunity, respectively. Additionally, we speculated that E. coli mutants killed by heat may lack metabolites that activate UPRER and immunity, or conversely, have increased levels of metabolites that inhibit these pathways. Fortunately, our investigation led to the discovery of the E. coli mutant yfbR, which inhibits UPRER and immunity by increasing carbohydrates that aid in resisting these stress pathways. Moving forward, we intend to further explore the intricate relationship between HK-E. coli and UPRER-immunity. This will be a key focus of our future research efforts.
Collectively, this study uncovers the unexpected function of UPRER (IRE-1/XBP-1) – Innate immunity (PMK-1/p38 MAPK) as a physiological food evaluation system for evaluating and sensing food quality in animals. It also highlights the utility of the HK-E. coli (low-quality food) – C. elegans interaction as a means to dissect the mechanism of food evaluation system in assessing food. Most importantly, it reveals that animals are capable of altering their nutrient (Vitamin C) levels through in vivo synthesis or food intake to adapt to a poor food environment when better food choices are not available.
Author Contributions
P. L designed, performed experiments, analyzed data. X. L constructed all transgenic animals. B.Q. designed research, supervised this study, and wrote the paper with inputs from P. L.
Acknowledgements
We thank the Caenorhabditis Genetics Center (CGC) (funded by NIH P40OD010440) for strains; Dr. Zhao Shan for suggestions. This work was supported by the Ministry of Science and Technology of the People’s Republic of China (2019YFA0802100, 2019YFA0803100), the National Natural Science Foundation of China (32170794), Yunnan Applied Basic Research Projects (202302AP370005, 202201AT070196, K264202230211, 202001AV070011), the Yunnan University Startup Program.
Declaration of interests
The authors declare no competing interests.
Supplemental Figure legends
Supplemental Tables
Table S1. Metabolices analysis.
Table S2. RNA-seq analysis.
Table S3. Screening data for E. coli mutant keio library.
Table S4. Metabolomics-seq data of HK-K12, HK-yfbR and K12.
Table S5. RNA-seq data of animals fed with HK-E. coli OP50 and E. coli OP50.
Raw-data. Raw data for experiments.
Star Methods
Resource Availability
Lead contact
Further information and requests for reagents may be directed to the lead contact Bin Qi (qb@yun.edu.cn).
Materials availability
All reagents and strains generated by this study are available through request to the lead contact with a completed Material Transfer Agreement.
Data and code availability
Metabolomics-seq data are accessible in Table S4
RNA-seq data are accessible in Table S5.
This paper does not report original code.
Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.
Experimental model and subject details
C. elegans strains and maintenance
Nematode stocks were maintained on nematode growth medium (NGM) plates seeded with bacteria (E. coli OP50) at 20℃.
1) The following strains/alleles were obtained from the Caenorhabditis Genetics Center (CGC) or as indicated:
N2 Bristol (wild type control strain);
AU78: agIs219 [T24B8.5p::GFP::unc-54 3’ UTR + Pttx-3::GFP::unc-54 3’ UTR];
SJ4005: zcIs4 [Phsp-4::GFP];
AY101: acIs101 [F35E12.5p::GFP + rol-6(su1006)];
SJ17: xbp-1 (zc12);
KU25: pmk-1(km25);
AY102: pmk-1(km25) IV; acEx102 [Pvha-6::pmk-1::GFP + rol-6(su1006)];
YNU108: Ex[Prgef-1::pmk-1::GFP;Podr-1::RFP]35;
xbp-1(tm2482) 56;
KU21: kgb-1(km21);
AU133: agIs17 [Pmyo-2::mCherry + Pirg-1::GFP] IV;
SJ4100: zcIs13 [Phsp-6::GFP + lin-15(+)].
2) The following strains were constructed by this study:
YNU242: xbp-1(tm2482); pmk-1(km25) double mutant was constructed by crossing: xbp-1(tm2482) with KU25[pmk-1(km25)].
YNU240: ylfEx149 [xbp-1(tm2482); Prgef-1::xbp-1::GFP; Podr-1::RFP] transgene strain was constructed by injecting plasmid Prgef-1::xbp-1::GFP with Podr-1::RFP in xbp-1(tm2482) background
YNU241: ylfEx150 [xbp-1(tm2482); Pges-1::xbp-1::GFP; Podr-1::RFP] transgene strain was constructed by injecting plasmid Pges-1:xbp-1::GFP with
Podr-1::RFP in xbp-1(tm2482) background
Bacterial strains
E. coli-OP50, Saprophytic staphylococci, Pseudomonas aeruginosa-PA14, E. coli-K12 (BW25113), and E. coli-K12 mutant were cultured at 37°C in LB medium. A standard overnight cultured bacteria was then spread onto each Nematode growth media (NGM) plate.
Culture Medium
MGN Medium: Sigma agar: 20g/L; Bacto Peptone: 2.5g/L; NaCl: 3g/L; MgSO4: 0.12g/L; CaCl: 0.111g/L; PPB: (KH2PO4 0.8M; K2HPO4·3H2O 0.2M) 25ml/L; Cholesterol: 0.005g/L.
LB broth: TPYPTONE: 10g/L; Yeast Extract: 5g/L; NaCl 5g/L.
Method Details
Generation of transgenes
1) To construct the C. elegans plasmid for expression of xbp-1 in neuron, 3057bp promoter of rgef-1 and genomic DNA of xbp-1 was inserted into the PPD95.77 vector. DNA plasmid mixture containing Prgef-1::xbp-1::GFP (25ng/ul) and Podr-1::RFP (25ng/ul) was injected into the gonads of adult xbp-1(tm2482).C. elegans plasmid for expression of xbp-1 in neuron, 3057bp promoter of rgef-1 and genomic DNA of xbp-1 was inserted into the PPD95.77 vector. DNA plasmid mixture containing Prgef-1::xbp-1::GFP (25ng/ul) and Podr-1::RFP (25ng/ul) was injected into the gonads of adult xbp-1(tm2482).
2) To construct the C. elegans plasmid for expression of xbp-1 in intestine, 2549 bp promoter of ges-1 and genomic DNA of xbp-1 was inserted into the PPD95.77 vector. DNA plasmid mixture containing Pges-1::xbp-1::GFP (25ng/ul) and Podr-1::RFP (25ng/ul) was injected into the gonads of adult xbp-1(tm2482).C. elegans plasmid for expression of xbp-1 in intestine, 2549 bp promoter of ges-1 and genomic DNA of xbp-1 was inserted into the PPD95.77 vector. DNA plasmid mixture containing Pges-1::xbp-1::GFP (25ng/ul) and Podr-1::RFP (25ng/ul) was injected into the gonads of adult xbp-1(tm2482).
Preparation and feeding of worm food
We followed an established protocol 24, 25 to prepare heat-killed (HK) E. coli. Briefly, a standard OD600=0.5-0.6 of E. coli OP50 and E. coli K12 grown in LB broth was concentrated to 1/20 vol and was then heat-killed at 80℃ for 180 min. About 150ul of the heat-killed bacteria was spread onto each 35mm NGM plate.
Preparation of HK-E. coli + carbohydrate or vitamin C food
1) 100ul of water, 100ul of L-ascorbic acid (dissolved in water at a concentration of 100mg/ml, Sangon Biotech, 100143-0100) or 100 ul of D-glucuronic acid (dissolved in water at a concentration of 100mg/ml, Adamas, 1102520) was mixed with 500 ul of HK-E. coli, then 150ul of the mixture was spread onto 35mm NGM plates.E. coli, then 150ul of the mixture was spread onto 35mm NGM plates.
2) 12.5ul of water or 12.5ul of D-(+)-glucose (dissolved in water at a concentration of 100mg/ml, Sangon Biotech, A501991-0500) was mixed with 500 ul of HK-E. coli, then 150ul of the mixture was spread onto 35mm NGM plates.E. coli, then 150ul of the mixture was spread onto 35mm NGM plates.
Behavioral assay
C. elegans selection assays
For C. elegans to have enough space to evaluate food, we add 18ul of the sample onto a 35mm NGM plate. This creates a round lawn with a radius of 5mm, which occupies about 8% of the total plate area.
1) 18ul of heat-killed OP50, live OP50, and LB broth (as the buffer for bacteria) was added into 35mm NGM plate in an equilateral triangle pattern. Then, synchronized L1 worms were seeded in the center of NGM plate for 16-17h at 20℃ (as indicated in Figure1B).
2) 18ul of heat-killed OP50 and heat-killed OP50 with D-GlcA or vitamin C was added into 35mm NGM plate in an equilateral triangle pattern, then synchronized L1 worms were seeded on equilateral triangle of NGM plate for 16-17h at 20℃ (as indicated in Figure 5A).Figure 5A).
Here is Selection rate formula:
C. elegans aversion assays
18ul food was spread out the center of NGM plate, then synchronized L1 by bleach solution (NaOH: 1M, NaClO:4-6%) worms were seeded on center of food for 16-17h at 20 ℃.
Three to ten replicates for each condition were performed for each assay, and the experiments were duplicated on different days.
Analysis of the fluorescence intensity in worms
The synchronized L1 worms carrying either UPRER reporter (Phsp-4::GFP) or innate immunity reporter (Pirg-5::GFP; Psysm-1p::GFP; Pirg-1p::GFP) were seeded on NGM with indicated food and incubated for 24h at 20℃. For fluorescence imaging, worms were anesthetized with 25 mM levamisole and photographed using either an Olympus BX53 microscope or Olympus MVX10 dissecting microscope equipped with a DP80 camera.
The fluorescence intensity in entire intestinal region was quantified using ImageJ software and normalized to the body area.
E. coli Keio collection screen
The whole Keio E. coli single mutant collection (Baba et al., 2006) was screened. Mutant bacteria strains, as well as the wild-type control strain BW25113, were cultured in LB medium with 50µg/ml kanamycin at 37℃ until an OD600 of 0.5-0.6 was reached. The bacteria were then heat-killed following our established protocol 24, and 150 ul of the heat-killed mutant E. coli was spread onto 35 mm NGM plates. Synchronized L1 worms carrying UPRER reporter (Phsp-4::GFP) were seeded and cultured for 24h at 20℃. The fluorescence was then examined by using an Olympus MVX10 dissecting microscope, progressive screening three times. Next, a 4th screen was performed using immune reporter (Pirg-5::GFP) animals fed with HK-E. coli mutants that reduced the Phsp-4::GFP fluorescence. Finally, an aversion behavior assay was performed using HK-E. coli mutants that both reduced Phsp-4::GFP and Pirg-5::GFP. HK-E. coli mutants that reduced UPRER, immune and avoidance behavior were identified through this screening.
RNAi treatment
RNAi plasmid is delivered in a E. coli strain, HT115, from either the MRC RNAi library 57 or the ORF-RNAi Library 58. RNAi plates were prepared by adding IPTG to NGM agar to a final concentration of 1 mM. Overnight E. coli cultures (LB broth containing 100 ug/ml ampicillin and 100uM IPTG) of specific RNAi strains and the control HT115 strain were seeded onto RNAi feeding plates and cultured at room temperature until dry. Synchronized L1 worms were treated RNAi by feeding (Ahringer, Reverse genetics, WormBook 2006) for the first generation and allowed to grow to maturity. The worms were then bleached and hatched in M9 buffer for 18hr. The synchronized L1 worms were then seeded on the indicated feeding plate.
Western blot
To measure the level of p-PMK-1, worms (feeding different food for 4h) were analyzed by standard western blot methods and probed with anti-p38 (dilution = 1:5,000; Cell Signaling, 9212S), anti-p-p38 (dilution = 1:5,000; Cell Signaling, 4511S) and anti-α-tubulin (dilution = 1:10,000; Sigma T5168) as a loading control.
To measure the level of protein translation, worms (feeding different food for 24h) were analyzed by standard western blot methods and probed with anti-Puromycin (dilution = 1:10,000; Sigma-Aldrich, MABE343) and anti-α-tubulin (dilution = 1:10,000; Sigma T5168) as a loading control.
Total content of ascorbic acid (TAA) assay
The total content of ascorbic acid was measured using the kits (Beijing Biotech-Pack-analytical Scientific Co., Ltd., Beijing, China, BKWB132 http://biotech-pack-analytical.foodmate.net/) according to the manufacturer’s protocol. Briefly, L1 worms were seeded on the different feeding assay plate and cultured for 4 hours. The worms were then lysed in ice-cold conditions using lysis buffer. Equal amounts of protein were used for the normalization. Here is formula for getting TAA concentration
V1 – the volume of supernatant of for experiment
Cpr – the concentration of supernatant protein
D – Dilution ratio of supernatant
a – the intercept of standard curve
b – the slope of standard curve
the standard curve y = 0.0611x + 0.0003 for Figure 4C; y = 0.0258x + 0.0066 for Figure 4B, Figure 4-figure supplement 1A and Figure 4-figure supplement 1B.
Preparation of samples for RNA sequencing
RNA-seq was done with three biological replicates that were independently generated, collected, and processed. Adult wild type (N2) worms were bleached and then the eggs were incubated in M9 for 18 hours to obtain synchronized L1 worms. L1 worms were cultured in the NGM plate with HK-E. coli or E. coli for 4hrs at 20℃. L1 worms were then collected for sequencing.
RNA sequencing and data processing
For the RNA sequencing assay, cDNA libraries were constructed, and single-end libraries were sequenced using the Illumina platform (Novogene, Beijing, China). HISAT2 59 was used to map the clean reads to the reference gene sequence (Species: Caenorhabditis_elegans; Source: NCBI; Reference Genome Version: GCF_000002985.6_WBcel235), and then “featureCounts” tool in subread software 60 was used to calculate the gene expression level of each sample. Read counts were inputted into DESeq2 61to calculate differential gene expression and statistical significance. Differentially expressed genes (DEGs) were screened using following criteria: |log2(FoldChange)| > 1 & padj<= 0.05.
Preparation of samples for metabolome sequencing
Metabolome-seq of bacterial was done with four biological replicates that were independently generated, collected, and processed. Total of three group E. coli sample including: E. coli K12 (Con), HK-E. coli K12 (HK), and HK-E. coli yfbR mutant (HY). All bacteria were overnight cultured to the same OD (OD600=1). E. coli K12 and E. coli yfbR mutant are heat-killed (80, 180min), E. coli K12, HK-E. coli K12 and HK-E. coli yfbR mutant were then spread out NGM plate for 72hs at room temperature. Finally, sample was collected into 1.5ml tube by using sterile cell scraping.
Metabolome sequencing and data processing
Metabolome were sequenced using the Ultra Performance Liquid Chromatography (UPLC) (ExionLC AD, https://sciex.com.cn/) and Quadrupole-Time of Flight (TripleTOF 6600, AB SCIEX) for Non-targeted; Ultra Performance Liquid Chromatography (UPLC) (ExionLC AD, https://sciex.com.cn/) and Tandem mass spectrometry (MS/MS) (QTRAP®, https://sciex.com/) for Broad targeting (Metware, Wuhan, China). Multiple reaction monitoring (MRM) was used to calculate the expression level of each metabolite. Differential metabolites were screened through Fold change ≥ 2 or Fold change ≤ 0.5 and VIP ≥ 1 (Variable Importance in Projection of OPLS-DA model).
Microscopy
Analysis of fluorescence was performed with an Olympus BX53 microscope, CLSM (Zeiss LSM900), or Olympus MVX10 dissecting with a DP80 camera.
Quantification and statistical analysis
Quantification
ImageJ software was used for quantifying fluorescence intensity of UPRER and Innate immunity reporter. ImageJ software was used for counting the number of worms about selection and aversion behavior.
Statistical analysis
All statistical analyses were performed in Graphpad prism 8.0. Two-tailed unpaired t test was used for statistical analysis of two groups of samples, one-way or two-way ANOVA was used for statistical analysis of more than two groups of samples. Data are presented as Mean ± SD, and p<0.05 was considered a significant difference, “*” represents p<0.05, “**” represents p<0.01, “***” is represents < 0.001, “****” represents p<0.0001, “ns” represents no significant difference. For all figures, ‘‘n’’ represents the number of worms scored from at least three independent experiments.
References
- 1.Feeding State Modulates Behavioral Choice and Processing of Prey Stimuli in the Zebrafish TectumNeuron 90:596–608
- 2.Food allergy as a biological food quality control systemCell 184:1440–1454
- 3.Diverse states and stimuli tune olfactory receptor expression levels to modulate food-seeking behaviorElife 11
- 4.Viewing images of foods evokes taste quality-specific activity in gustatory insular cortexProc Natl Acad Sci U S A 118
- 5.Fruit scent and observer colour vision shape food-selection strategies in wild capuchin monkeysNat Commun 10
- 6.Chemosensation in C. elegansWormBook :1–29
- 7.Bargmann CI. odr-10 encodes a seven transmembrane domain olfactory receptor required for responses to the odorant diacetylCell 84:899–909
- 8.Reprogramming chemotaxis responses: sensory neurons define olfactory preferences in C. elegansCell 91:161–169
- 9.Olfaction and olfactory learning in Drosophila: recent progressCurr Opin Neurobiol 17:720–726
- 10.Dissecting a circuit for olfactory behaviour in Caenorhabditis elegansNature 450:63–70
- 11.Olfactory perception of food abundance regulates dietary restriction-mediated longevity via a brain-to-gut signalNat Aging 1:255–268
- 12.Functional organization of a neural network for aversive olfactory learning in Caenorhabditis elegansNeuron 68:1173–1186
- 13.Gustatory Processing in Drosophila melanogasterAnnu Rev Entomol 63:15–30
- 14.Antagonistic sensory cues generate gustatory plasticity in Caenorhabditis elegansEmbo J 25:312–322
- 15.A gustatory receptor paralogue controls rapid warmth avoidance in DrosophilaNature 500:580–584
- 16.Feeding is inhibited by sublethal concentrations of toxicants and by heat stress in the nematode Caenorhabditis elegans: Relationship to the cellular stress responseJ Exp Zool 284:147–157
- 17.The gut-to-brain axis for toxin-induced defensive responsesCell 185:4298–4316
- 18.An essential role for XBP-1 in host protection against immune activation in C. elegansNature 463:1092–1095
- 19.A conserved p38 MAP kinase pathway in Caenorhabditis elegans innate immunityScience 297:623–626
- 20.Vitamin C and Immune FunctionNutrients 9
- 21.Vitamins for the Gut MicrobiomeTrends Mol Med 26:137–140
- 22.New developments and novel therapeutic perspectives for vitamin CJ Nutr 137:2171–2184
- 23.Toward a new recommended dietary allowance for vitamin C based on antioxidant and health effects in humansAm J Clin Nutr 69:1086–1107
- 24.A vitamin-B2-sensing mechanism that regulates gut protease activity to impact animal’s food behavior and growtheLife 6
- 25.Microbial Siderophore Enterobactin Promotes Mitochondrial Iron Uptake and Development of the Host via Interaction with ATP SynthaseCell 175:571–582
- 26.Contrasting invertebrate immune defense behaviors caused by a single gene, the Caenorhabditis elegans neuropeptide receptor gene npr-1BMC Genomics 17
- 27.System wide analysis of the evolution of innate immunity in the nematode model species Caenorhabditis elegans and Pristionchus pacificusPLoS One 7
- 28.IRE1 couples endoplasmic reticulum load to secretory capacity by processing the XBP-1 mRNANature 415:92–96
- 29.A Conserved PMK-1/p38 MAPK Is Required in Caenorhabditis elegans Tissue-specific Immune Response to Yersinia pestis Infection*The Journal of Biological Chemistry 285:10832–10840
- 30.Compartment-specific perturbation of protein handling activates genes encoding mitochondrial chaperonesJournal of Cell Science 117:4055–4066
- 31.Mechanisms, regulation and functions of the unfolded protein responseNat Rev Mol Cell Biol 21:421–438
- 32.Signal integration in the endoplasmic reticulum unfolded protein responseNat Rev Mol Cell Biol 8:519–529
- 33.Global transcriptional regulation of innate immunity by ATF-7 in C. elegansPlos Genet 15
- 34.Measurements of Innate Immune Function in C. elegansMethods Mol Biol 2144:145–160
- 35.Gut commensal E. coli outer membrane proteins activate the host food digestive system through neural-immune communicationCell Host Microbe 30:1401–1416
- 36.The Deoxycytidine Pathway for Thymidylate Synthesis in Escherichia coliJournal of Bacteriology 189:7922–7926
- 37.Sucrose Phosphorylase and Related Enzymes in Glycoside Hydrolase Family 13: Discovery, Application and EngineeringInternational Journal of Molecular Sciences 21
- 38.Lactose induced redox-dependent senescence and activated Nrf2 pathwayInt J Clin Exp Pathol 12:2034–2045
- 39.The invertebrate Caenorhabditis elegans biosynthesizes ascorbateArch Biochem Biophys 569:32–44
- 40.L-Ascorbate Biosynthesis Involves Carbon Skeleton Rearrangement in the Nematode Caenorhabditis elegansMetabolites 10
- 41.The role of vitamin C in stress-related disordersThe Journal of Nutritional Biochemistry 85
- 42.Ascorbate regulation and its neuroprotective role in the brainTrends in Neurosciences 23:209–216
- 43.Update: Effects of Antioxidant and Non-Antioxidant Vitamin Supplementation on Immune FunctionNutrition Reviews 65:181–217
- 44.Selected vitamins and trace elements support immune function by strengthening epithelial barriers and cellular and humoral immune responsesBritish Journal of Nutrition 98:S29–S35
- 45.Vitamin C protects against hypoxia, inflammation, and ER stress in primary human preadipocytes and adipocytesMolecular and Cellular Endocrinology 556
- 46.Hepatoprotective benefits of vitamin C against perfluorooctane sulfonate-induced liver damage in mice through suppressing inflammatory reaction and ER stressEnvironmental Toxicology and Pharmacology 65:60–65
- 47.Tyramine Acts Downstream of Neuronal XBP-1s to Coordinate Inter-tissue UPR(ER) Activation and Behavior in C. elegansDev Cell 55:754–770
- 48.p38 MAPK regulates expression of immune response genes and contributes to longevity in C. elegansPlos Genet 2
- 49.Inactivation of conserved C. elegans genes engages pathogen– and xenobiotic-associated defensesCell 149:452–466
- 50.XBP1 links ER stress to intestinal inflammation and confers genetic risk for human inflammatory bowel diseaseCell 134:743–756
- 51.Survive an innate immune response through XBP1Cell Research 20:506–507
- 52.Alleviating chronic ER stress by p38-Ire1-Xbp1 pathway and insulin-associated autophagy in C. elegans neuronsPLOS Genetics 16
- 53.Activation of p38 MAPK Signaling-Mediated Endoplasmic Reticulum Unfolded Protein Response by Nanopolystyrene ParticlesAdvanced Biosystems 3
- 54.Non-Canonical Caspase Activity Antagonizes p38 MAPK Stress-Priming Function to Support DevelopmentDev Cell 53:358–369
- 55.The regulation of animal behavior by cellular stress responsesExp Cell Res 405
- 56.Physiological IRE-1-XBP-1 and PEK-1 signaling in Caenorhabditis elegans larval development and immunityPLoS Genet 7
- 57.Systematic functional analysis of the Caenorhabditis elegans genome using RNAiNature 421:231–237
- 58.Toward improving Caenorhabditis elegans phenome mapping with an ORFeome-based RNAi libraryGenome Res 14:2162–2168
- 59.Mapping and quantifying mammalian transcriptomes by RNA-SeqNat Methods 5:621–628
- 60.featureCounts: an efficient general purpose program for assigning sequence reads to genomic featuresBioinformatics 30:923–930
- 61.Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2Genome Biol 15
Article and author information
Author information
Version history
- Sent for peer review:
- Preprint posted:
- Reviewed Preprint version 1:
- Reviewed Preprint version 2:
- Version of Record published:
Copyright
© 2024, Liu et al.
This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.
Metrics
- views
- 715
- downloads
- 65
- citations
- 2
Views, downloads and citations are aggregated across all versions of this paper published by eLife.