Here, we present a framework to relate broad-range dietary restriction to gene expression and lifespan. We describe protocols for broad-range dietary restriction and for quantitative imaging of gene expression under this paradigm. We further outline computational analyses to reveal underlying information processing features of the genetic circuits involved in food-sensing.
Sensory systems allow animals to detect, process, and respond to their environment. Food abundance is an environmental cue that has profound effects on animal physiology and behavior. Recently, we showed that modulation of longevity in the nematode Caenorhabditis elegans by food abundance is more complex than previously recognized. The responsiveness of the lifespan to changes in food level is determined by specific genes that act by controlling information processing within a neural circuit. Our framework combines genetic analysis, high-throughput quantitative imaging and information theory. Here, we describe how these techniques can be used to characterize any gene that has a physiological relevance to broad-range dietary restriction. Specifically, this workflow is designed to reveal how a gene of interest regulates lifespan under broad-range dietary restriction; then to establish how the expression of the gene varies with food level; and finally, to provide an unbiased quantification of the amount of information conveyed by gene expression about food abundance in the environment. When several genes are examined simultaneously under the context of a neural circuit, this workflow can uncover the coding strategy employed by the circuit.
All organisms need to be able to sense and respond to changes to the environment to ensure their survival. In animals, the nervous system is the primary detector and transducer of information about the environment and coordinates the physiological response to any change that might affect the organism's survival1. Food abundance is an environmental cue that is well studied in multiple contexts that not only regulates food-related behaviors, such as foraging2, but also impacts the longevity of an animal. The modulation of lifespan by changes in food abundance is a phenomenon known as dietary restriction (DR), and has broad evolutionary conservation3.
The nematode Caenorhabditis elegans is a powerful model system for addressing fundamental biological questions. A plethora of techniques have been developed that allow for manipulation of the worm genome, such as RNAi and in vivo gene editing techniques. The small physical size of the worm and its optical transparency also lend themselves to in vivo imaging of both transcriptional and translational fluorescent reporters and utility of high-throughput technologies such as microfluidics4. Together, these tools can be harnessed to examine how neural circuits direct animal behavior.
C. elegans is a bacterivore and several methods have been published that allow for the precise control of food abundance by manipulating bacterial concentration5,6,7,8. Within the C. elegans research community, DR has been studied in two different contexts. The first can be termed 'classical DR', as it mirrors the changes seen in response to decreasing food levels in other organisms. In this context, decreasing food abundance from ad libitum levels results in an increasing lifespan up until an optimum is reached, after this point longevity decreases with further reduction of food6,7,9. The second context under which DR has been studied in C. elegans is dietary deprivation in which the longevity of the worms is increased by the complete removal of any bacterial food source10,11. In Entchev et al. (2015)12, we showed that the complexity in DR resulting from these two different paradigms can be examined simultaneously under a context we term 'broad-range DR'. By using the protocol outlined below, we identified a new class of genes involved in DR that bidirectionally modulate the lifespan response to food abundance and are involved in neural circuits that sense food12 (Figure 1).
The response of an animal to changes in the environment integrates a sequence of biological processes that link the sensory system to complex regulatory interactions conveying environmental information to physiology. Although the mechanistic details of such "information flow" are often unknown, genetic tools can be used to acquire an insight into how this complex computation is organized among different biological components. In our recent work, we showed that daf-7 and tph-1 are involved in the transmission of environmental information about food abundance through a food-sensing neural circuit that modulates lifespan in C. elegans12,13. By applying the mathematical framework of information theory14, we were able to quantify the amount of environmental information, in terms of bits, that is represented by the gene expression changes in daf-7 and tph-1 in specific neurons across different food levels. From this, we were then able to uncover the encoding strategy employed by this neural circuit and how it is genetically controlled (Figure 2).
In the following protocol, we outline the steps required to understand what the effects of genes of interest expressed in specific neurons are and how they participate to the flow of food information from environment to lifespan. Broadly, our framework is split in two experimental protocols and a computational workflow. For the experimental aspects, it is critical to have mutants of the genes of interest that can be examined under broad-range DR. Faithful transcriptional reporters are also necessary to quantify the expression level of the genes at different food levels. To be able to carry out the computational analysis discussed in our method, the dataset needs to be of sufficient size to provide meaningful estimates of expression distributions. Even though we provide template source codes for the analyses, the user needs to be familiar with the language of information theory that is extensively used throughout our computational framework. The source codes are written in R and C++. Therefore, a certain level of programming proficiency is also required to apply them in a meaningful way.
1. Preparation of Bacterial Cultures and Plates for General Worm Culture
2. Preparation of Bacterial Cultures and Dilutions for Experiments
3. Setting Up Lifespan Experiments
NOTE: Lifespan assays are performed on 6 cm treated tissue culture filled with 12 mL of NGM agar supplemented with streptomycin and carbenicillin (NSC), both at a final concentration of 50 μg/mL. These tissue culture plates are well suited for lifespan assays at no or very low bacterial concentrations, as the worms are significantly less prone to sticking to the wall of the plate and desiccating. The use of two different antibiotics prevents the OP50 strain from developing drug-resistance, which is critical in controlling the bacterial concentration on the plate. Also, by inactivating bacteria with antibiotics, we minimize the effects on worm lifespan due to pathogenic infection16. This allows us to consider only the food-related components of bacterial concentration in these experiments. Many C. elegans ageing studies supplement NGM plates with the chemical fluoro-2′-deoxyuridine (FUdR) to render the adults sterile. However, the use of FUdR can be problematic as its use can cause several confounding issues with longevity assays17,18,19,20,21. Entchev et al. (2015)12 circumvented this issue by exposure of L4 larvae to RNAi of egg-522,23, which inhibits the oocyte-to-embryo transition by blocking the formation of the eggshell of fertilized C. elegans oocytes resulting in their death.
4. Setting Up Imaging Experiments
NOTE: The steps outlined in this section are sufficient to generate enough worms per strain for imaging one experimental food condition at a given temperature. This protocol can be scaled to fit the number of conditions that will be imaged on any given day. However, care should be taken in the experimental design to ensure that the timeframe is reasonable and that animals across different strains do not have an age difference greater than 12 hours on the day of imaging. It is highly recommended that the transcriptional reporters being imaged are single-copy transgenics, as this will more closely resemble the native regulation of the gene. The protocol outlined below, and summarized in Table 3, also provides a streamlined method for obtaining large synchronous age matched populations of strains, which can be used for other experimental workflows.
5. Initial Culture of Reporter Strains
6. Egg-collection and Synchronization of Reporter Strains
NOTE: Some genes of interest in the C. elegans aging research community result in growth and egg-laying defects when mutated, which makes it more difficult to generate large synchronous populations of animals of different genotypes. For example, The daf-7(ok3125) mutation causes a severe egg-laying defect in comparison to the wild type N2 strain. Therefore, to get sufficient numbers of synchronous L4 larvae of different strains for quantitative imaging experiments requires a more robust methodology than manually picking worms. For this reason, transcriptional reporter strains were subjected to a sodium hypochlorite (NaClO)/ sodium hydroxide (NaOH) solution treatment of gravid adults to break open the animals and liberate their eggs, a process commonly referred to as 'bleaching' by the C. elegans research community15.
7. Treatment of Reporter Strains with egg-5 RNAi
8. Initiation of Broad-range DR
NOTE: After the 24 h egg-5 RNAi treatment, the L4 larvae that were originally deposited on the plates will have become 1-day old adults.
9. Microfluidic Imaging of Reporter Strains
10. Data Assembly
NOTE: The fluorescence intensities of all neurons analyzed by the image processing software are combined into the filtered expression data file (FED) which is used to estimate the distribution profiles of gene expression (template R and C++ scripts are available at https://github.com/giovannidiana/templates).
11. Estimation of Information Encoded
NOTE: The following procedure describes how to quantify the information about specific environmental conditions encoded by the set of gene expressions. In Diana et al. (2017)13, the information encoded about food abundance in the environment was examined, however, the method itself is applicable to any discrete number of environmental states. The essential ingredient to quantify information theoretic variables such as information entropies or redundancy is the joint probability distribution of the neural responses under the set environmental stimuli considered. To perform such estimation, it is crucial to have a sufficient sampling of the response across populations of worms. Gaussian distributions can be estimated from relatively small samples; however, it is important to have an idea of the expected shape of the expression distribution to quantify the appropriate sample size for a reliable density estimation. Due to the unavoidable variability across different trials, it is essential to check that the central values of the distributions obtained from different repeats of the same experiment are not systematically shifted or that any of the statistical features of the expression distribution are not significantly altered across trials. In case the trial-to-trial variability is compatible with the variability within each trial, it is crucial to balance the number of trials versus the number of worms within trials to average out those environmental/biological factors that affect trial-to-trial variability. Undersampling those factors could heavily bias the information-theoretic analysis.
12. Calculation of Redundancy, Noise and Signal Correlation
By conducting lifespan experiments on the mutants of the genes of interest alongside the wild type N2 strain, one can establish whether these genes have a role in the food response to broad-range DR. The wild type response should be comparable to the one depicted in Figure 1A. Any modulation of this response by the mutants, reflected by a non-uniform effect across food conditions, indicates that these genes affect the ability of the worm to correctly respond to changes in food abundance, at which point further investigation of the expression responses of these genes to broad-range DR is warranted. If, however, the longevity response of the mutants is not significantly different from the wild type then the genes have no role in transducing the effects of broad-range DR, at least at the level of mean lifespan. If the mutations cause a uniform shift of the whole lifespan response then the genes have a food-independent effect on longevity. This does not rule out the possibility that the expression of the genes of interest is food-responsive, in which case the information carried by these genes is not transmitted to lifespan.
The next stage of the protocol is to determine how expression levels change under broad-range DR for the genes of interest. In Figure 1B, we illustrate this through the expression levels of a transcriptional reporter of daf-7, which shows a response to changes in food level in the ASI sensory neurons. In a daf-7(-) mutant, the expression response of the transcriptional reporter is altered. If the genes of interest are truly food-responsive at the level of lifespan then one can expect that their expression will also change with food. Correspondingly, a transcriptional reporter in the mutant background should have an altered expression profile in response to broad-range DR. However, it is also possible that the transcriptional reporter of the gene of interest in a wild type background does not show any food-responsive changes in expression. In this situation, this may indicate a post-transcriptional regulatory effect that falls outside the scope of this protocol.
In Diana et al. (2017)13, we extracted expression values for daf-7 in ASI and tph-1 in ADF and NSM. In Figure 2A, we illustrate the estimation of the expression distribution in ASI and ADF for a given food level. Having multiple readouts from single worm images allows us to analyze not only the amount of information encoded independently by each neuron but also the combinatorial information of the whole neural circuit (Figure 2B-2C). Combining these two information-theoretic measures allows us to characterize the system in terms of the encoding strategy employed by the neurons to convey information about food. The amount of redundancy in the circuit can be obtained by taking the sum of the mutual information for each neuron and subtracting the joint mutual information (channel capacity) obtained by considering the combinatorial readouts of the circuit. A positive value of such difference denotes a redundant character of the encoding strategy because the cumulative information among the parts is larger than the actual information encoded by the whole circuit. Conversely a negative value reflects a synergistic strategy because the true information encoded is larger than the sum of its components (Figure 2B). Information and redundancy can be compared across different genotypes to explore possible higher order roles of gene regulation, for instance in Diana et al. (2017)13 the effect of daf-7 mutation switches the encoding strategy from redundant to synergistic (Figure 2C-2D).
Figure 1: Response of lifespan and gene expression under broad-range DR. (A) The mean lifespan of the wild type N2 strain (black line) displays a complex response to broad-range DR. The magnitude of this response is attenuated in a null mutant of the daf-7 gene (red line). Error bars represent standard error of the mean, pooled data from Entchev et al. (2015)12. (B) The mean expression levels of a transcriptional reporter for the daf-7 gene in wild type background (black line) also display a complex non-monotonic response to broad-range DR. In daf-7(-) genetic background the expression of this transcriptional reporter is highly attenuated and shows little response to changes in food level. Error bars represent standard error of the mean, data from a single trial in Entchev et al. (2015)12. Please click here to view a larger version of this figure.
Figure 2: Computational methodology. (A) Illustration of two-dimensional density estimation of tph-1 expression in ADF and daf-7 expression in ASI as obtained from the 'ks' R package using a grid dimension 30 x 30. (B) Visualization of the information encoded by the joint expression of tph-1 and daf-7 (whole) and individually (sum of parts) for ADF, ASI and NSM neurons. Redundant and synergistic characters of the encoding are represented by the difference between the height of the stacked bars on the right and the information encoded by the full circuit. (C) Comparison between food information encoded by wild type animals and daf-7(-) mutants. (D) The reduction of mutual information observed in the mutants is accompanied by a switch towards synergistic encoding. Panels B-D are adapted from Diana et al. (2017)13. Please click here to view a larger version of this figure.
Bacterial Concentration (cells/ml) | Optical Density (600nm) | Dilution Factor (From Previous) |
1.12E+10 | 56.000 | 0.00 |
2.00E+09 | 10.000 | 5.60 |
6.32E+08 | 3.160 | 3.16 |
6.32E+07 | 0.316 | 10.00 |
2.00E+07 | 0.100 | 3.16 |
0 (S basal) | 0.000 | NA |
Table 1: Food levels and dilution factors used in broad-range DR. Bacterial concentrations (cells/mL) used in the broad-range DR protocol, along with their respective OD600 measurements and the dilution factor required to achieve each concentration from the previous one.
Experimental Temperature of Lifespan | |||||||
Day | 12.5°C | 15°C | 17.5°C | 20°C | 22.5°C | 25°C | 27.5°C |
-12 | Chunk all strains to fresh NGM plates and maintain at 20°C | ||||||
-11 | |||||||
-10 | Set up P0 generation of daf-7(-) strains and maintain at 20°C (1 L4 per plate, 5 plates) | ||||||
-9 | Set up P0 generation of wild type strains and maintain at 20°C (1 L4 per plate, 5 plates) | ||||||
-8 | |||||||
-7 | |||||||
-6 | |||||||
-5 | Set up F1 generation of daf-7(-) strains and maintain at 20°C (1 L4 per plate, 90 plates) | ||||||
-4 | Set up F1 generation of wild type strains and maintain at 20°C (1 L4 per plate, 30 plates) | ||||||
-3 | |||||||
-2 | |||||||
-1 | |||||||
0 | Pick F2 L4 larvae to >egg-5 RNAi plates and maintain at 20°C (15 L4 per plate, 24 plates) | ||||||
1 | Move 1-day old adults to NSC plates seeded with 2.0E+9 cells/ml food level and maintain at 20°C | ||||||
2 | Move 2-day old adults to NSC plates seeded with experimental food conditions and to experimental temperature | ||||||
3 | Transfer | Transfer | Transfer | Transfer | Transfer | Transfer | Transfer |
4 | |||||||
5 | Transfer | Transfer | Transfer | Transfer | Transfer | Transfer | Transfer |
6 | |||||||
7 | Transfer | Transfer | Transfer | Transfer | Transfer | Transfer | Transfer |
8 | |||||||
9 | Transfer | Transfer | Transfer | Transfer | Transfer | Transfer | Transfer |
10 | |||||||
11 | Transfer | Transfer | Transfer | Transfer | Transfer | Transfer | |
12 | |||||||
13 | |||||||
14 | Transfer | Transfer | Transfer | Transfer | |||
15 | |||||||
16 | |||||||
17 | |||||||
18 | Transfer | Transfer | |||||
19 | |||||||
20 | |||||||
21 | |||||||
22 | Transfer |
Table 2: Schedule for lifespans conducted at different temperatures. Schematic outline of the steps needed to set up broad-range DR lifespan experiments at different temperatures using daf-7(-) and wild type strains as examples. The number of transfers to fresh plates of each experimental food level decreases with increasing temperature. This is to account for the fact that animals at higher temperatures are aging more rapidly and so a more prone to physical damage per transfer.
Days | Imaging Pipeline |
-14 | Chunk reporter strains in daf(-) background. Maintain at 20 °C. |
-13 | Chunk reporter strains in wild type background. Maintain at 20 °C. |
-12 | Set up P0 generation of daf-7(-) reporter strains. Use 3 L4 larvae per 10cm NGM plate. Use 2 plates and maintain at 20 °C. |
-11 | |
-10 | Set up P0 generation of wild type reporter strains. Use 3 L4 larvae per 10cm NGM plate. Use 2 plates and maintain at 20 °C. |
-9 | |
-8 | Set up F1 generation of daf-7(-) reporter strains. Use 3 L4 larvae per 10cm NGM plate. Use 12 plates and maintain at 20 °C. |
-7 | |
-6 | Set up F1 generation of wild type reporter strains. Use 3 L4 larvae per 10cm NGM plate. Use 4 plates and maintain at 20 °C. |
-5 | |
-4 | |
-3 | Bleach daf-7(-) reporter strains in afternoon (~5pm) and deposit eggs on 3 10cm NGM plates and maintain at 20 °C. |
-2 | Bleach wild type reporter strains in morning (~10am) and deposit eggs on 3 10cm NGM plates and maintain at 20 °C. |
-1 | |
0 | Wash L4 to 10 cm egg-5 RNAi plates. Use 3 plates per strain and maintain at 20 °C. |
1 | Wash 1-day adults to NSC plates seeded with 2.0E+9 cells/ml. Use 3 plates per strain and maintain at 20 °C. |
2 | Wash 2 day adults to NSC plates seeded with experimental food levels. Use 3 plates per strain and shift to experimental temperature. |
3 | Transfer to fresh NSC plates. Use 3 plates per strain and maintain at experimental temperature. |
4 | |
5 | Transfer to fresh NSC plates. Use 3 plates per strain and maintain at experimental temperature. |
6 | Pick animals off plates and prepare for imaging. |
Table 3: Schedule for imaging pipeline. Schematic outline of the steps needed to set up broad-range DR imaging experiments using fluorescent transcriptional reporter strains in daf-7(-) and wild type backgrounds at different temperatures as examples.
Here, we present a new method for dietary restriction that encapsulates a much broader range of food concentrations than previously published protocols. This method links two previously separate phenomena seen in C. elegans DR literature, bacterial deprivation and classical dietary restriction, allowing both dietary effects to be studied under one protocol. Using the new broad-range DR paradigm, we present a general framework for examining single cell gene expression in response to a specific environmental cue and determining how this cell encodes information. Our framework consists of two experimental protocols that illustrate how to perform lifespans and quantitative imaging, respectively, under broad-range DR. Data from these experimental protocols can then be examined with the computational analyses provided in this framework to quantify the information encoded by changes in the gene expression levels or lifespans across different food conditions.
Lifespan experiments using broad-range DR paradigm involve six distinct food levels (Table 1). This necessitates a more labor-intensive approach than examining longevity under fewer food levels, such as dietary deprivation10,11 or using the eat-2 genetic background35. However, examining at lifespan under a single condition can limit the interpretations of a gene's role in DR. For example, we recently showed that daf-7 mutants have a bidirectional attenuation of the response to food concentration compared to wild type animals12 (Figure 1A). In the absence of food, daf-7 mutants display a shortening of their lifespan compared to wild type animals. If we had only considered dietary deprivation, we would have interpreted that the daf-7 gene as being necessary for only lifespan extension, when in fact daf-7 role is more complex. Therefore, the critical outcome of this part of the protocol is to establish whether a gene of interest is involved in modulating the overall response of lifespan to changes in food abundance.
One major advantage of this protocol compared to other methods is that it uses a novel method to eliminate progeny production in the animals undergoing lifespan analysis. Most studies use the drug FuDR to inhibit proliferation of the germline in adults rendering them sterile. However, recent studies have shown FuDR treatment can have condition- and gene-specific effects on lifespan17,18,19,20,21, calling into question its general applicability. In this protocol, elimination of progeny production is achieved through a 24 h treatment of animals with RNAi targeting the egg-5 gene, which inhibits the formation of the chitin eggshell of fertilized C. elegans oocytes resulting in their death22,23. The advantage of this method is that it is very late-acting and so does not interfere with the germline, which is a major regulator of longevity in C. elegans.
One potential caveat of the broad-range DR protocol is its reliance on the use of the antibiotics to control bacterial proliferation to ensure tight control of bacterial concentration. Bacterial proliferation within the gut of the worm is known to be a major cause of death in C. elegans16. Thus, the use of bacteriostatic antibiotics, such as carbenicillin, in NGM agar prevents bacterial proliferation and increases lifespan of worms compared to non-antibiotic controls16. Certain types of antibiotics, such as rifampicin36 and members of the tetracycline family37,38, have been shown to extend lifespan in C. elegans independently of their effect on bacterial proliferation. However, there is no evidence in the literature that either carbenicillin or streptomycin can increase longevity independently of their effect on bacterial proliferation.
Lifespan can be viewed as the output of a complex computation where environmental information, routed by gene-expression in neuronal networks, is transmitted to physiology. Our protocol provides a methodology to understand how specific genes affect this flow of environmental information. To address this question, we need reliable image processing to determine the distribution of gene expression responses at the single-cell level. Being able to estimate not only the average response of gene expression to changes in food abundance but also the full statistical distribution from large populations represents an important requirement for the applicability of our method. This accurate description of gene expression responses to food abundance allows the application of information theory to quantify the information encoded by the specific neurons as well as the coding strategy employed by the neural circuit.
The imaging and computational aspects of the methods outlined in this protocol are applicable to a greater set of biological contexts. In our work, we focused on a small neural network involved in food sensing, however, the analyses of information-processing features are not limited to a specific cell type or specific environmental cues. In the future, these methodologies can potentially be extended to a larger variety of input variables, affecting any physiological output. These approaches will contribute to a greater understanding of how gene regulatory networks encode, process and transmit information.
The authors have nothing to disclose.
We thank the Bargmann and Horvitz labs for reagents. Some strains were provided by the CGC, which is funded by NIH Office of Research Infrastructure Programs (P40 OD010440). We also thank M. Lipovsek for comments on the manuscript. This research was supported by the Wellcome Trust (Project Grant 087146 to Q.C.), BBSRC (BB/H020500/1 and BB/M00757X/1 to Q.C.), European Research Council (NeuroAge 242666 to Q.C.), US National Institutes of Health (R01AG035317 and R01GM088333 to H.L.), and US National Science Foundation (0954578 to H.L., 0946809 GRFP to M.Z.).
Carbenicillin di-Sodium salt | Sigma-Aldrich | C1389-5G | Antibiotic |
Streptomycin Sulphate salt | Sigma-Aldrich | S6501-50G | Antibiotic |
Isopropyl β-D-1-thiogalactopyranoside (IPTG) | Sigma-Aldrich | I6758-10G | Inducer for RNAi plates |
Sodium Chloride (NaCl) | Sigma-Aldrich | 71380-1KG-M | Used in S basal, and NGM agar |
di-Potassium Hydrogen Phosphate(K2HPO4) | Sigma-Aldrich | 1.05104.1000 | Used in S basal, and NGM agar |
Potassium di-Hydrogen Phosphate (KH2PO4) | Sigma-Aldrich | P9791-1KG | Used in S basal, and NGM agar |
Magnesium Sulphate (MgSO4) | Sigma-Aldrich | M2643-1KG | Used in NGM agar |
Calcium Chloride (CaCl2) | Sigma-Aldrich | C5670-500G | Used in NGM agar |
Sodium Hydroxide (NaOH) | Sigma-Aldrich | 71687-500G | Used for bleaching |
Pluronic-F127 | Sigma-Aldrich | P2443-1KG | Used in imaging |
Sodium Hypochlorite (NaClO) | Sigma-Aldrich | 1.05614.2500 | Used for bleaching |
LB Broth | Invitrogen | 12780052 | Used to grow bacteria |
Adavanced TC 6 cm Tissue Culture plates | Greiner Bio-One | 628960 | Plates for lifespan |
CellStar 10cm Tissue Culture plates | Greiner Bio-One | 664160 | Plates for imaging |
Low Retention P200 tips | Brandt | 732832 | Tips for handling worms in liquid |
Agar | BD | 214510 | Agar for NGM, RNAi and NSC plates |
Bacto-peptone | BD | 211820 | Peptone for NGM, RNAi and NSC plates |