The Large-Scale Cultivation of Nematodes to Study Their Collective Behaviors

Summary

Here, a system is reported for studying the collective behaviors of nematodes by culturing them in bulk using dog food agar medium. This system allows researchers to propagate large numbers of dauer worms and can be applied to Caenorhabditis elegans and other related species.

Abstract

Animals exhibit dynamic collective behaviors, as observed in flocks of birds, schools of fish, and crowds of humans. The collective behaviors of animals have been investigated in the fields of both biology and physics. In the laboratory, researchers have used various model animals such as the fruit fly and zebrafish for approximately a century, but it has remained a major challenge to study large-scale complex collective behavior orchestrated by these genetically tractable model animals. This paper presents a protocol to create an experimental system of collective behaviors in Caenorhabditis elegans. The propagated worms climb on the lid of the Petri plate and show collective swarming behavior. The system also controls worm interactions and behaviors by changing the humidity and light stimulation. This system allows us to examine the mechanisms underlying collective behaviors by changing environmental conditions and examining the effects of individual-level locomotion on collective behaviors using mutants. Thus, the system is useful for future research in both physics and biology.

Introduction

Both nonscientists and scientists are fascinated with animals' collective behaviors, as in flocks of birds and schools of fish. Collective behaviors have been analyzed in a broad range of fields, including physics, biology, mathematics, and robotics. In particular, active matter physics is a growing research field that focuses on systems composed of self-propelled elements, that is, dissipative systems, such as flocks of birds, schools of fish, biofilms of motile bacteria, cytoskeletons composed of active molecules, and groups of self-propelled colloids. The theory of active matter physics maintains that however complex the behaviors of individuals are, the collective motions of enormous numbers of living things are governed by a small number of simple rules. For example, the Vicsek model, a candidate for a unified description of the collective motion of self-propelled particles, predicts that short-range alignment interaction of moving objects is required to form a long-range ordered phase with eccentric fluctuation in 2D, as in herds of animals¹. Top-down experimental approaches pertaining to the physics of active matter are developing rapidly. Previous experiments confirmed the formation of a long-range ordered phase in Escherichia coli². Other recent works employed cells^3,4, bacteria⁵, motile colloids⁶, or moving proteins^7,8. Simple minimal models such as the Vicsek model successfully described these real phenomena. In contrast with these unicellular experimental systems, collective behaviors by animals are usually observed in the wild, as no one could hope to perform controlled experiments with 10,000 real birds or fish.

Biologists share the same interest as physicists: how individuals interact with each other and functionally behave as a group. One of the traditional research fields for analyzing individual behavior is neuroscience, in which the mechanisms underlying behavior have been examined at the neuronal and molecular levels. Many neuroscientific bottom-up approaches have been developed thus far. Top-down approaches in physics and bottom-up approaches in biology can be facilitated using model animals such as the fruit fly, the worm Caenorhabditis elegans, and the mouse⁹. However, there have been few findings on the large-scale collective behavior of these model animals in the laboratory¹⁰; it is still difficult to prepare genetically tractable model animals on a large scale in the laboratory. Therefore, in current research on collective behaviors in biology and physics, it has been difficult for scientists who usually do research in the laboratory to study animals' collective behaviors.

In this study, we established a method for the large-scale cultivation of nematodes to study their collective behaviors. This system allows us to change environmental conditions and examine the effect of individual-level locomotion on collective behaviors using mutants¹⁰. In active matter physics, the parameters of the mathematical model can be controlled in both experiments and simulations, which enables verification of that model for identifying unified descriptions. Genetics is used to understand the neural circuit mechanism underlying collective behavior¹¹.

Protocol

1. Preparation of worms

NOTE: Prepare the wild-type N2 Bristol strain¹² and ZX899 strain (lite-1(ce314); ljIs123[mec-4p::ChR2, unc-122p::RFP])¹³ for the observation of collective behaviors and optogenetic experiments, respectively. Maintain the ZX899 strain under dark conditions.

1. Deposit four well-fed adult worms on a 60 mm plate containing 14 mL of nematode growth medium (NGM) with agar and seeded with E. coli OP50¹².
2. Grow F1 worms to starvation in the NGM plate at 23 °C for 7 days. The yield of F1 worms is approximately 100 worms/plate at this point. The stages of worms comprise a mixed population of dauer and starved L1 larvae.

2. Preparation of dog food agar (DFA) medium plates

1. Autoclave a glass bottle containing 2 g of powdered dog food and 5 mL of 1% agar medium and cool it to room temperature (Figure 1A).
   NOTE: Other dog foods from different manufacturers can be used in this experiment.

3. Inoculation of worms to DFA medium plates

1. Transfer small amounts (approximately 0.5 g) of DFA medium onto the center of an NGM plate seeded with E. coli OP50 (Figure 1B). For optogenetic experiments, pour 40 µL of 50 µM all-trans-retinal, the cofactor of channel rhodopsin 2, onto the DFA before inoculation of worms.
2. Collect the starved worms from four NGM plates using autoclaved water.
3. Place a small fragment of dog food (approximately 0.5 g) on the DFA medium, approximately 2 mm away from the lid of the plate.
4. Illuminate the NGM plate with ultraviolet light for 15 min inside a clean bench to prevent contamination.
5. Inoculate the collected worms (approximately 400 worms) onto the DFA medium on NGM plates. Do not seal the plate with parafilm to avoid increasing the humidity inside the Petri plate and generating water droplets that trap worms on a lid.
6. Propagate the worms at 23 °C and allow them to climb up to the lid of the plate for approximately 10-14 days.
   ​NOTE: Because the number of worms on the lids hardly increased after 10-14 days, it was presumed that the worms had likely run out of food.

4. Observation of collective behavior

1. On the day of the experiment, place a new NGM plate that did not include E. coli and dog food agar medium onto an aluminum plate on the stage of a macro zoom microscope (Figure 2A). Keep the bottom of this new NGM plate at 23 °C using a Peltier temperature controller unit for at least 5 min (Figure 2B). Then, replace the lid of this new NGM plate with the lid of the plate that the worms climbed up onto. Use the objective lens (x2, NA = 0.5) as a low-magnification objective (Figure 2A).
2. Increase the temperature of the bottom of the Petri plate from 23 °C to 26 °C to alter the humidity inside that plate (Figure 2).
3. Acquire images of the inner surface of the plate lid with the camera at 20 frame s⁻¹ (Supplementary Video S1).
4. Save the acquired images in the Tagged Image File format.

5. Optogenetic experiment

1. Use a 100 W mercury lamp to deliver blue light, filtered with a filter set. Control the illumination time precisely using an electromagnetic shutter system (Figure 2B).
2. Maintain ZX899 on DFA under these conditions for 5 min before blue light illumination.
3. Illuminate ZX899 worms attached to the lid of a Petri plate on the microscope stage maintained at 23 °C.
4. Acquire images of the inner surface of the plate lid with a camera at 20 frame s⁻¹ (Supplementary Video S2).

Representative Results

Here, wild-type dauer worms were used for collective behavior observations. Worms were cultivated at 23 °C for approximately 10-14 days and climbed up to the inner surface of the lid of a DFA medium plate. On the experimental day, only the lid was transferred to a new NGM plate without E. coli and DFA medium. The bottom of this Petri plate was initially kept at 23 °C using the Peltier system, and then its temperature was increased to 26 °C. A movie was taken under the microscope. Figure 3 shows snapshots of the movie. Worms dynamically remodeled their network patterns during humidity change. As humidity increases, the compartment sizes of the network also become larger. Finally, the networks collapsed, and dormant worm clusters remained on the inner surface of the lid.

Figure 1: Photos of DFA medium for cultivating large numbers of worms. (A) Photo of DFA medium prepared in a glass bottle. (B) Photo of NGM plate with DFA medium just after collected worms were inoculated. Abbreviations: DFA = dog food agar; NGM = nematode growth medium. Please click here to view a larger version of this figure.

Figure 2: Experimental system for the observation of collective behavior. (A) Microscopy for observation of collective behavior. (B) Mechanical shutter controller and temperature control system using the Peltier system. Please click here to view a larger version of this figure.

Figure 3: Representative data of the dependence of the collective network pattern on humidity. Dependence of the C. elegans network on ambient humidity. The camera frame rate is 1 fps. Scale bar = 4 mm. Please click here to view a larger version of this figure.

Supplementary Video S1: Collective network formations. Wild-type dauer worms were propagated using DFA on NGM in a Petri plate. The worms self-organized inside the lid. The humidity was changed using a Peltier device. Images were taken from above the lid. The movie plays 80 times faster than the real-time recording rate. Abbreviations: DFA = dog food agar; NGM = nematode growth medium. Please click here to download this File.

Supplementary Video S2: Optogenetic manipulation of worms' collectives. Optogenetics was performed with 1, 2, 4, 8, 32, and 128 s blue light illuminations. This activation initially caused the arborization and collapse of bundles. Finally, a network different from the initial structure was formed. The movie is played 20 times faster than the real-time recording rate. Please click here to download this File.

Discussion

In this study, we show a protocol for preparing a system for the large-scale collective behavior of C. elegans in the laboratory. The DFA-based method was originally established with Caenorhabditis japonica¹⁴ and Neoaplectana carpocapsae Weiser¹⁵, both of which are non-model animals. However, this method was not applied to investigate collective behaviors. The C. elegans is a genetically tractable model animal^11,12. Behavioral genetic studies using C. elegans have contributed to the investigation of individual-level behavioral research. However, in the long history of C. elegans research, although a simple clumping pattern was observed^16,17,18,19, no reports have demonstrated dynamic pattern formation via the group-level behavior of C. elegans. A key idea of this study is using DFA medium to facilitate the maintenance of a huge number of worms for a long time in a Petri plate. Using DFA medium, we present the observation of dynamic collective behavior by C. elegans, thus introducing a new behavioral paradigm.

Previously, several mass worm production methods have been reported. In comparison with these methods, the advantage of this method is to enable the investigation of collective behavior on a lid without a process for the isolation of dauer worms. Recently, we published a paper reporting the transfer of nictating dauer worms across a gap between a lid and DFA medium using electrostatic interactions with the lid²⁰. This worm transfer occurs when worms form a nictation column composed of approximately 100 worms. This study shows that only dauer worms can transfer when dauer formations are induced by too much crowding in DFA. The number of worms produced by this method is likely fewer than other methods such as egg yolk-based methods. However, to perform a behavioral assay on a lid, we can use the population of the dauer worms, which hardly include other stage worms such as starved L1 larvae, whereas previous methods require a process for the isolation of dauer worms. Thus, this method allows for a more precise collective behavioral examination using dauer worms. In addition, the experimenter can also control the density of worms in the following procedure. First, autoclaved water was used to collect and wash the worms that moved to the lid. Then, the concentration of worms in water was determined by counting the worms in an aliquot of the worm suspension, and the worm suspension was dropped onto a substrate. Taken together, our system is more controllable in terms of the worm stage and density for behavioral experiments.

Collective behavior has been analyzed from the perspective of active matter physics, which seeks to identify unified descriptions of collective motions of living and nonliving self-propelled particles. Toward this goal, many experimental systems have been developed for nonliving self-propelled particles and cells, but fewer systems have been developed for multi-cellular organisms, which exhibit much more complex behaviors based on the neural circuit. Therefore, our system extends the possibility that the unified description of collective motions exists. Regarding humidity manipulation, our previous numerical simulation based on a model suggested that attraction forces between worms, likely induced by humidity in the experiment, induce pattern changes, which were qualitatively consistent with humidity-induced pattern changes¹⁰. However, we think that there is no deterministic experimental evidence showing that pattern changes were induced by humidity rather than temperature. Therefore, the experimenter should exercise caution in whether the collective behavior changes can be attributed solely to the humidity change rather than the temperature change or not.

Understanding the neural mechanism underlying collective behaviors in animals is a new challenge in the field of biology. Collective behaviors lead to the emergence of a new function that does not appear at the individual level. As animals have a nervous system, they have memory and learning abilities, and it is intriguing to examine the differences in these neural functions at the individual and population levels. It has been noted that collective behaviors improve the detection sensitivity for foreign organisms and prey and enhance the ability for correct decision-making^21,22,23. C. elegans also has a nervous system comprised of 302 neurons and thereby memorizes the past cultivation temperature²⁴ and migrates to a location with preferable humidity²⁵. Thus, it would be interesting to investigate the relationship between neural functions and collective behaviors in C. elegans. Furthermore, one can expect to extract mechanical parameters through observation of the behavior of a worm population. For example, observation of the viscoelastic properties in C. elegans crowds would make it possible to estimate the elasticity of a single worm and the surface tension between worms. The size distribution of worm clumps should relate to the surface tension between them. The propulsive force of the C. elegans individual can also be estimated from the frequency that the worm moves out in response to the surface tension. Thus, we can expect to estimate mechanical parameters at the level of individual worms based only on macroscopic information of worm population.

In conclusion, active matter physics aims to identify unified descriptions of collective behaviors, and this field requires more experimental verification of the proposed mathematical models by controlling parameters. Additionally, the functional significance of each animal's collective pattern formation and its mechanical relevance to neural functions are important open questions. Furthermore, given that one of the objectives of 'soft robotics' is the precise control of collectives of robots, we hope that an algorithm can be established through the experiments of worms' collective behaviors for application to controlling the collective motions of soft robots.

Materials

Name	Company	Catalog Number	Comments
Escherichia coli and C. elegans strains
E. coli OP50	Caenorhabditis Genetics Center	OP50	Food for C. elegans. Uracil auxotroph. E. coli B.
lite-1(ce314); ljIs123[mec-4p::ChR2, unc-122p::RFP]	author	ZX899	lite-1(ce314) mutant carrying the genes expressing ChR2 and RFP under the control of the mec-4 and unc-122 promoter, respectively
N2 Bristrol	Caenorhabditis Genetics Center		Wild-type C. elegans strain
For worm cultivation
Agar purified, powder	Nakarai tesque	01162-15	For preparation of NGM plates
All-trans retinal	Sigma-Aldrich	R2500	For optogenetics
Bacto pepton	Becton Dickinson	211677	For preparation of NGM plates
Calcium chloride	Wako	036-00485	For preparation of NGM plates
Cholesterol	Wako	034-03002	For preparation of NGM plates
di-Photassium hydrogenphosphate	Nakarai tesque	28727-95	For preparation of NGM plates
Dog food	Nihon Pet Food	VITA-ONE	For preparation of dog food agar medium
LB broth, Lennox	Nakarai tesque	20066-95	For culture of E. coli OP50
Magnesium sulfate anhydrous	TGI	M1890	For preparation of NGM plates
Petri dishes (60 mm)	Nunc	150270	For preparation of NGM plates
Potassium Dihydrogenphosphate	Nakarai tesque	28720-65	For preparation of NGM plates
Sodium Chloride	Nakarai tesque	31320-05	For preparation of NGM plates
Observation
Computer	CT solution	CS6229	Windows10 Pro with Intel Xeon Gold 6238R CPU and 768 GB of RAM
CMOS Camera	Hamamatsu photonics	ORCA-Lightning C14120-20P	For data acquisition
CMOS Camera	Olympus	DP74	For data acquisition
Microscope with SZX-MGFP set	Olympus	MVX10	For data acquisition
x2 Objective lens	Olympus	MV PLAPO 2XC	Working distance 20 mm and numerical aperture 0.5
Shutter control
Shutter	OptoSigma	BSH2-RIX	For controlling temporal pattern of  light illumination
Shutter controller	OptoSigma	SSH-C2B-A	For controlling temporal pattern of  light illumination
Temperature control
Peltier temperature controller unit	VICS	WLVPU-30	For controlling humidity inside a Petri plate
UNI-THEMO CONTROLLER	Ampere	UTC-100	For controlling humidity inside a Petri plate
Data acquisition software
HCImage	Hamamatsu photonics		For data acquisition