A High-throughput Assay for the Prediction of Chemical Toxicity by Automated Phenotypic Profiling of Caenorhabditis elegans

* These authors contributed equally

Your institution must subscribe to JoVE's Environment section to access this content.

Fill out the form below to receive a free trial or learn more about access:


Enter your email below to get your free 10 minute trial to JoVE!

We use/store this info to ensure you have proper access and that your account is secure. We may use this info to send you notifications about your account, your institutional access, and/or other related products. To learn more about our GDPR policies click here.

If you want more info regarding data storage, please contact gdpr@jove.com.



A quantitative method has been developed to identify and predict the acute toxicity of chemicals by automatically analyzing the phenotypic profiling of Caenorhabditis elegans. This protocol describes how to treat worms with chemicals in a 384-well plate, capture videos, and quantify toxicological related phenotypes.

Cite this Article

Copy Citation | Download Citations

Gao, S., Chen, W., Zhang, N., Xu, C., Jing, H., Zhang, W., Han, G., Flavel, M., Jois, M., Zeng, Y., Han, J. D., Xian, B., Li, G. A High-throughput Assay for the Prediction of Chemical Toxicity by Automated Phenotypic Profiling of Caenorhabditis elegans. J. Vis. Exp. (145), e59082, doi:10.3791/59082 (2019).


Applying toxicity testing of chemicals in higher order organisms, such as mice or rats, is time-consuming and expensive, due to their long lifespan and maintenance issues. On the contrary, the nematode Caenorhabditis elegans (C. elegans) has advantages to make it an ideal choice for toxicity testing: a short lifespan, easy cultivation, and efficient reproduction. Here, we describe a protocol for the automatic phenotypic profiling of C. elegans in a 384-well plate. The nematode worms are cultured in a 384-well plate with liquid medium and chemical treatment, and videos are taken of each well to quantify the chemical influence on 33 worm features. Experimental results demonstrate that the quantified phenotype features can classify and predict the acute toxicity for different chemical compounds and establish a priority list for further traditional chemical toxicity assessment tests in a rodent model.


Along with the rapid development of chemical compounds applied to industrial production and people's daily life, it is important to study the toxicity testing models for the chemicals. In many cases, the rodent animal model is employed to evaluate the potential toxicity of different chemicals on health. In general, the determination of lethal concentrations (i.e., the assayed 50% lethal dose [LD50] of different chemicals) is used as the traditional parameter in a rodent (rat/mouse) model in vivo, which is time-consuming and very expensive. In addition, due to the reduce, refine, or replace (3R) principle that is central to animal welfare and ethics, new methods that allow for the replacement of higher animals are valuable to scientific research1,2,3. C. elegans is a free-living nematode that has been isolated from soil. It has been widely used as a research organism in the laboratory because of its beneficial characteristics, such as a short lifespan, easy cultivation, and efficient reproduction. In addition, many fundamental biological pathways, including basic physiological processes and stress responses in C. elegans, are conserved in higher mammals4,5,6,7,8. In a couple of comparisons we and others have made, there is a good concordance between C. elegans toxicity and toxicity observed in rodents9. All of this makes C. elegans a good model to test the effects of chemical toxicities in vivo.

Recently, some studies quantified the phenotypic features of C. elegans. The features can be used to analyze the toxicities of chemicals2,3,10 and the aging of worms11. We also developed a method that combines a liquid worm culturing system and an image analysis system, in which the worms are cultured in a 384-well plate under different chemical treatments12. This quantitative technique has been developed to automatically analyze the 33 parameters of C. elegans after 12-24h of chemical treatment in a 384-well plate with liquid medium. An automated microscope stage is used for experimental video acquisition. The videos are processed by a custom-designed program, and 33 features related to the worms' moving behavior are quantified. The method is used to quantify the worm phenotypes under the treatment of 10 compounds. The results show that different toxicities can alter the phenotypes of C. elegans. These quantified phenotypes can be used to identify and predict the acute toxicity of different chemical compounds. The overall goal of this method is to facilitate the observation and phenotypic quantification of experiments with C. elegans in a liquid culture. This method is useful for the application of C. elegans in chemical toxicity evaluations and phenotype quantifications, which help predict the acute toxicity of different chemical compounds and establish a priority list for further traditional chemical toxicity assessment tests in a rodent model. In addition, this method can be applied to the toxicity screening and testing of new chemicals or the compound as the food additive agent pollution, pharmacautical compounds, environmental exogenous compound, and so on.

Subscription Required. Please recommend JoVE to your librarian.


The protocol follows the animal care guidelines of the Animal Ethics Committee of the Beijing Center for Disease Prevention and Control in China.

1. Chemical preparation

  1. Obtain chemicals (Table 1 and Table of Materials).
  2. Determine the highest and lowest dosage of the individual chemicals for a minimum concentration of 100% lethality (LC100, 24 h) and a maximum concentration of 100% nonlethality (LC0, 24 h) to worms. Use at least six dilutions of the highest concentration (Table 1).
    NOTE: Conduct a preliminary worm lethality test9 to explore LC100 and LC0 for a new chemical, to determine the dosage.
  3. Dilute each chemical with K-medium (Table of Materials) to 2x the required concentration. Use K-medium as a control to compare the phenotype alterations caused by the chemicals.
    1. For example, prepare 7 gradient concentrations of cadmium chloride (CdCl2) (Table 1). To prepare 2x the highest concentrated aqueous solution (4.64 mg/mL), dissolve 92.8 mg of CdCl2 solid powder in 8 mL of K-medium and fill up to 10 mL after the powder has fully dissolved. Prepare the other concentration levels by dilution with K-medium.
  4. Prepare eight parallel wells for every concentration in the chemical gradient. Each well contains 50 µL of the 2x chemical solution. Prepare at least three groups of eight parallel wells of K-medium as controls (Table 2).
    NOTE: In brief, a volume of 500 µL of 2x working solution is necessary for a single dose of each chemical.

2. Worm preparation

  1. Obtain wild-type N2 worms and Escherichia coli OP50 strains from the Caenorhabditis Genetics Center (CGC).
  2. Obtain synchronized L4 worms.
    1. Pick a single colony of E. coli OP50 from the streak plate. Aseptically inoculate the colony in 100 mL of LB broth and grow it overnight at 37 °C.
      NOTE: The E. coli OP50 solution is now ready for seeding to nematode growth medium (NGM, Table of Materials) plates.
    2. Pour NGM into a 90 mm plastic Petri plate. Seed each plate with 300 µL of E. coli OP50 solution the day after pouring. Incubate N2 worms on the NGM plates with OP50 at 20 °C for about 2-3 days until most of the worms have reached the adult stage.
    3. Harvest gravid worms into a 15 mL sterile conical centrifuge tube with sterile H2O. Let the worms settle down for at least 2 min, aspirate the H2O, and add 5 mL of bleach buffer (Table of Materials).
    4. Vortex the tube for 5 min, spin the tube for 30 s (at 1,300 x g) to pellet the eggs, and discard the supernatant.
    5. Wash the eggs with 5 mL of sterile H2O and vortex the tube for 5 s. Centrifuge the tube for 30 s (at 1,300 x g), remove the supernatant, and wash again.
    6. Pipette the eggs onto a new NGM plate with OP50. Incubate them at 20 °C. Monitor the hatched L1 worms the next morning; the worms will reach the L4 stage in approximately 40 h.
  3. Wash the L4 worms off the 90 mm Petri plates with K-medium into a 50 mL sterile conical tube. Adjust the concentration of worms to ~40 animals per 100 µL of K-medium under a stereomicroscope. Add 50 µL (~20 worms) into each well of the 384-well plate. These synchronized worms (L4 stage) are ready for the following treatment by chemicals.

3. Chemical treatment and video capture

NOTE: In a 384-well plate, worms (50 µL in each well) are treated to six to seven dosages of an individual chemical (Table 1). Prepare eight parallel wells, each containing 50 µL of the 2x chemical solution for every dosage (eight wells are filled with the same chemical and the same concentration, Table 2). All videos are collected using a digital camera attached to an inverted microscope (Table of Materials). The chemical treatment experiment lasts for 24 h. Do not add bacterial food to each well during the 24 h chemical treatment experiment.

  1. Before adding the chemicals, set the 384-well plate with the synchronized worms on the automatic stage and take videos of each well with the programmed acquisition procedure (7 frames per second for 2 s; it takes ~25 min to scan each plate).
  2. Add 50 µL of the 2x chemical stock prepared according to section 1 for each well (Table 2). Set the time as the 0 h point.
  3. Incubate the 384-well plate at 20 °C and shake it at 80 rpm in an incubator shaker.
  4. Remove the plate from the incubator and transfer it to an automatic stage. Take videos of each well of the whole plate, at 12 h and at 24 h, to check the phenotypes of the worms for each specific chemical treatment in K-medium. Approximately 25 min are required for one plate screen.

4. Experiment video processing

NOTE: A program for experimental video and images processing was written and packaged. It can be freely downloaded (see Table of Materials). The experimental video is stored in the form of an image frame sequence, and the frame sequence of each video is stored in a specific directory. The program can recognize worms and quantify phenotypes automatically.

  1. In the graphical user interface (GUI, Figure 1), add the parameters, such as the frame sequence directory, the output directory, the worm size parameter, and the movement threshold parameter. Click the Analyze button to process the experimental images.
    1. Click the Select button to choose the source images directory.
    2. Add the middle result directory in the interface.
      NOTE: The middle results include the segmented images. These middle results are useful for the visual observation of the processed images.
    3. Add the final result directory in the interface.
    4. Add the average worm size parameter in the Worm Size textbox in the interface.
      NOTE: The size parameter used in the experiments is 2,000.
    5. Add the Threshold of moved ratio in the interface.
      NOTE: The ratio used in the experiments is 0.93.
    6. Click the Analyze button to start the image processing. Click the Reset button to clear the added parameters.
      NOTE: There are 33 features defined and quantified for worms. All the defined phenotypes are sorted by categories (listed in Table 3). These features can be quantified from experimental images. A quantitative comparison among different chemicals, which have different toxicities, can be done by comparing these features.

Subscription Required. Please recommend JoVE to your librarian.

Representative Results

We have tested the phenotypes of worms exposed to different concentrations of more than 10 chemicals12. In the test, 33 distinct features were quantified for each chemical compound at three time points (0 h, 12 h, and 24 h). Previously, a comparison between a manual and an automatic analysis of a lifespan assay was done11,12. In this assay, we found that chemicals and concentrations can influence the worm phenotypes. An overview of this method is shown in Figure 2.

The results (Figure 3 and Figure 4c,d) showed that the worms died quickly as the chemical concentration increased. At higher concentrations, the worms became straighter and less curved than at lower concentrations or in control groups (Figure 3 and Figure 4b). In the beginning (at 0 h), there was no significant difference between the control (K-medium) and chemical treatments for all phenotypes. After 12 h of treatment with a given chemical dosage, the phenotypes of worms showed different degrees of differences among control and different concentration groups. For example, the major axis length increased as time increased. There is also a gradient trend from lower to higher chemical concentrations. The gradient trend of different chemical concentrations was also significant in the minor axis length (Figure 4a,b).

In this assay, the worm's motility was calculated in two ways, based on the area the worm moved in and the motility ratio (Figure 4c,d). Motility results of both ways showed similar patterns. There were no significant differences of the worm motility among different concentrations and control groups at the beginning (at the 0 h time point). As time passed, the worms in the control groups showed a stable decrease in motility. At 12 h, the worms that underwent chemical treatments at different concentrations showed significant differences in motility compared with control groups. In addition, the worms under higher concentration treatments showed weak motility compared to the worms under lower concentration treatments. This indicates that worms under higher concentration treatments became less motile and died quicker (Figure 4c,d). These results suggest that the designed method is useful for chemical toxicity assessments, and the quantified phenotypes of C. elegans are useful markers for chemical toxicity identification.

Figure 1
Figure 1: The interface of the software. Please click here to view a larger version of this figure.

Figure 2
Figure 2: The pipeline of a high-throughput assay for the prediction of chemical toxicity by automated phenotypic profiling of Caenorhabditis elegans. Please click here to view a larger version of this figure.

Figure 3
Figure 3: Experimental images of worms under 4.64 mg/mL CdCl2 (upper panel), 0.464 mg/mL CdCl2 (middle panel), and K-medium (bottom panel), at different time points. The images show the status changes of worms under chemical treatment or in a control group in one representative well of the 384-well plate throughout time. Please click here to view a larger version of this figure.

Figure 4
Figure 4: The quantified features of worms under different concentrations of CdCl2. (a) The quantified major axis length. (b) The quantified minor axis length. (c) The quantified motility by the moved area. (d) The quantified motility by the moved area/worm size. The bar plots show the average quantification for each feature on single worms. The error bars denote ± standard deviation (SD). The concentration unit = mg/mL. Please click here to view a larger version of this figure.

Table 1
Table 1: Exposure concentration of 10 chemicals for the 384-well-plate C. elegans acute toxicity test.

Table 2
Table 2: A schematic of the 384-well plate layout.

Table 3
Table 3: Defined phenotypes of worms.

Subscription Required. Please recommend JoVE to your librarian.


The advantages of C. elegans have led to its increasing usage in toxicology9, both for mechanistic studies and high-throughput screening approaches. An increased role for C. elegans in complementing other model systems in toxicological research has been remarkable in recent years, especially for the rapid toxicity assessment of new chemicals. This article provides a new assay of high-throughput, quantitative screening of worm phenotypes in a 384-well plate for the automatic identification and assessment of chemical toxicity. This assay is ideal for acute toxicity testing of chemicals within 24 h, and it could be applied to subacute toxicity testing as well when more time points of data are collected and food source (OP50) is supplied for the worms.

The medium used for diluting the chemicals can vary; we chose K-medium in the assay by referring to Sofieet al.13. Worms were cultured in K-medium in both the control and chemical treatment groups. An artificial freshwater solution or a soil solution with low ionic strength could be alternatives to K-medium.

Chemicals with different toxicities can alter the phenotypes of C. elegans in different patterns. Chemicals used in this test were chosen from the third to sixth categories of the Globally Harmonized System of Classification and Labeling of Chemicals (GHS). C. elegans were exposed to chemicals at six or more dosage levels, which covered the 0%-100% mortality dosage range. For those chemicals with low water solubility, DMSO is recommended to promote the chemical dissolution in water. As a high concentration of DMSO may affect worm development and lifespan14, no more than 0.2% DMSO should be used for aquatic tests.

The automatically quantified features show significant difference among different toxicities, which demonstrates that these quantified phenotypes of worms are very useful in identifying the toxicity of chemicals. It indicated that phenotypic profiling revealed conserved functions to classify and predict the toxicity of different chemicals using nematode C. elegans as an in vivo model organism.

The US National Toxicology Program (NTP) established the Tox21 community through a memorandum of understanding with the U.S. Environmental Protection Agency (EPA) and the National Institutes of Health (NIH) Chemical Genomics Center, now the National Center for Advancing Translational Sciences (NCATS). Tox21 uses high-throughput in vitro screening and in vivo alternative animal model testing to identify mechanisms of toxicity, to prioritize chemicals for additional in vivo toxicity testing, and to develop predictive models of human toxicological responses. As part of that effort, C. elegans was used to screen the EPA's ToxCast Phase I and Phase II libraries, which contain 292 and 676 chemicals, respectively, for chemicals leading to decreased larval development and growth15. The COPAS (Complex Object Parametric Analyzer and Sorter) platform has also been used for the worm toxicological screening studies2. However, the COPAS platform only quantifies few features, such as worm width, worm length, and the fluorescence intensity. This method is an improvement to current methods using worms to rapidly prescreen the toxicity of new chemicals.

There are several critical steps within the protocol: the worm culture in a 384-well plate, the chemical treatment, the experimental image capture, and the phenotype quantification. Compared to traditional toxicity evaluation methods, this protocol can quantify some phenotypes of worms that are difficult to calculate manually and useful to reflect the toxicities of every chemical, such as the worm motility, worm width, worm size, and gray intensity. Clearly, this high-throughput assay for the prediction of chemical toxicity will be a valuable toxicity model approach and could be used for the prescreening of chemicals before rodent animal experiments.

In summary, this technique paves a way on rapid toxicity assessment in multiple areas. Researchers could apply the method to the emergency analysis of toxicity in foodborne toxicosis, the safety evaluation of pharmaceutical compounds, as well as the acute toxicity screening and detection of new chemicals and environmental exogenous compounds.

Subscription Required. Please recommend JoVE to your librarian.


The authors have nothing to disclose.


The authors thank CGC for kindly sending the C. elegans. This work was supported by National Key Research and Development Program of China (#2018YFC1603102, #2018YFC1602705); National Natural Science Foundation of China Grant (#31401025, #81273108, #81641184), The Capital Health Research and Development of Special Project in Beijing (#2011-1013-03), the Opening Fund of the Beijing Key Laboratory of Environmental Toxicology (#2015HJDL03), and the Natural Science Foundation of Shandong Province, China (ZR2017BF041).


Name Company Catalog Number Comments
2-Propanol Sigma-Aldrich 59300
384-well plates Throme 142761
Agar Bacto 214010
Atropine sulfate Sigma-Aldrich PHL80892
Bleach buffer 0.5 mL of 10 M NaOH, 0.5 mL of5% NaClO, 9 mL ofultrapure water
Cadmium chloride Sigma-Aldrich 202908
Calcium chloride Sigma-Aldrich 21074
CCD camera Zeiss AxioCam HRm Zeiss microscopy GmbH
Cholesterol Sigma-Aldrich C8667
Copper(II) sulfate Sigma-Aldrich 451657
Ethanol Sigma-Aldrich 24105
Ethylene glycol Sigma-Aldrich 324558
Glycerol Sigma-Aldrich G5516
K-Medium 3.04 g of NaCl and 2.39 g of KCl in 1 L ultrapure water
LB Broth  10 g/L Tryptone, 5 g/L Yeast Extract, 5 g/L NaCl 
Magnesium sulfate heptahydrate Sigma-Aldrich 63140
NGM Plate 3 g ofNaCl, 17 g ofagar, 2.5 g ofpeptone in 1 L of ultrapure water, after autoclave add 1 mL of cholesterol (5 mg/mL in ethanol), 1 mL of MgSO4 (1 M), 1 mL of CaCl2 (1 M), 25 mL of PPB buffer
Peptone Bacto 211677
Potassium chloride Sigma-Aldrich 60130
Potassium phosphate dibasic Sigma-Aldrich 795496
Potassium phosphate monobasic Sigma-Aldrich 795488
PPB buffer 35.6 g of K2HPO4, 108.3 g of KH2PO4 in 1 L ultrapure water
shaker ZHICHENG ZWY-200D
Sodium chloride Sigma-Aldrich 71382
Sodium fluoride Sigma-Aldrich s7920
Sodium hydroxide Sigma-Aldrich 71690
Sodium hypochlorite solution Sigma-Aldrich 239305
The link of program https://github.com/weiyangc/ImageProcessForWellPlate
Tryptone Sigma-Aldrich T7293
Yeast extract Sigma-Aldrich Y1625
Zeiss automatic microscope  Zeiss AXIO Observer.Z1 Zeiss automatic microsco with peproprietary software Zen2012 and charge coupled device(CCD) camera



  1. Anderson, G. L., et al. Assessing behavioral toxicity with Caenorhabditis elegans. Environmental Toxicology and Chemistry. 23, (5), 1235-1240 (2004).
  2. Boyd, W. A., et al. A high-throughput method for assessing chemical toxicity using a Caenorhabditis elegans reproduction assay. Toxicology and Applied Pharmacology. 245, (2), 153-159 (2010).
  3. Boyd, W. A., Williams, P. L. Comparison of the sensitivity of three nematode species to copper and their utility in aquatic and soil toxicity tests. Environmental Toxicology and Chemistry. 22, (11), 2768-2774 (2003).
  4. Dengg, M., van Meel, J. C. Caenorhabditis elegans as model system for rapid toxicity assessment of pharmaceutical compounds. Journal of Pharmacological and Toxicological Methods. 50, (3), 209-214 (2004).
  5. Schouest, K., et al. Toxicological assessment of chemicals using Caenorhabditis elegans and optical oxygen respirometry. Environmental Toxicology and Chemistry. 28, (4), 791-799 (2009).
  6. Sprando, R. L., et al. A method to rank order water soluble compounds according to their toxicity using Caenorhabditis elegans, a Complex Object Parametric Analyzer and Sorter, and axenic liquid media. Food and Chemical Toxicology. 47, (4), 722-728 (2009).
  7. Wang, D., Xing, X. Assessment of locomotion behavioral defects induced by acute toxicity from heavy metal exposure in nematode Caenorhabditis elegans. Journal of Environmental Sciences (China). 20, (9), 1132-1137 (2008).
  8. Leung, M. C., et al. Caenorhabditis elegans: an emerging model in biomedical and environmental toxicology. Toxicological Sciences. 106, (1), 5-28 (2008).
  9. Li, Y., et al. Correlation of chemical acute toxicity between the nematode and the rodent. Toxicology Research. 2, (6), 403-412 (2013).
  10. Boyd, W. A., et al. Effects of genetic mutations and chemical exposures on Caenorhabditis elegans feeding: evaluation of a novel, high-throughput screening assay. PLoS One. 2, (12), 1259 (2007).
  11. Xian, B., et al. WormFarm: a quantitative control and measurement device toward automated Caenorhabditis elegans aging analysis. Aging Cell. 12, (3), 398-409 (2013).
  12. Gao, S., et al. Classification and prediction of toxicity of chemicals using an automated phenotypic profiling of Caenorhabditis elegans. BMC Pharmacology and Toxicology. 19, (1), 18 (2018).
  13. Moyson, S., et al. Mixture effects of copper, cadmium, and zinc on mortality and behavior of Caenorhabditis elegans. Environmental Toxicology and Chemistry. 37, (1), 145-159 (2018).
  14. Wang, X., et al. Lifespan extension in Caenorhabditis elegans by DMSO is dependent on sir-2.1 and daf-16. Biochemical and Biophysical Research Communications. 400, (4), 613-618 (2010).
  15. Boyd, W. A., et al. Developmental Effect of the ToxCast Phase I and Phase II Chemicals in Caenorhabditis elegans and Corresponding Responses in Zebrafish, Rats, and Rabbits. Environmental Health Perspectives. 124, (5), 586-593 (2016).



    Post a Question / Comment / Request

    You must be signed in to post a comment. Please or create an account.

    Usage Statistics