Long-Term Culture and Monitoring of Isolated Caenorhabditis elegans on Solid Media in Multi-Well Devices

Summary

Presented here is an optimized protocol for culturing isolated individual nematodes on solid media in microfabricated multi-well devices. This approach allows individual animals to be monitored throughout their lives for a variety of phenotypes related to aging and health, including activity, body size and shape, movement geometry, and survival.

Abstract

The nematode Caenorhabditis elegans is among the most common model systems used in aging research owing to its simple and inexpensive culture techniques, rapid reproduction cycle (~3 days), short lifespan (~3 weeks), and numerous available tools for genetic manipulation and molecular analysis. The most common approach for conducting aging studies in C. elegans, including survival analysis, involves culturing populations of tens to hundreds of animals together on solid nematode growth media (NGM) in Petri plates. While this approach gathers data on a population of animals, most protocols do not track individual animals over time. Presented here is an optimized protocol for the long-term culturing of individual animals on microfabricated polydimethylsiloxane (PDMS) devices called WorMotels. Each device allows up to 240 animals to be cultured in small wells containing NGM, with each well isolated by a copper sulfate-containing moat that prevents the animals from fleeing. Building on the original WorMotel description, this paper provides a detailed protocol for molding, preparing, and populating each device, with descriptions of common technical complications and advice for troubleshooting. Within this protocol are techniques for the consistent loading of small-volume NGM, the consistent drying of both the NGM and bacterial food, options for delivering pharmacological interventions, instructions for and practical limitations to reusing PDMS devices, and tips for minimizing desiccation, even in low-humidity environments. This technique allows the longitudinal monitoring of various physiological parameters, including stimulated activity, unstimulated activity, body size, movement geometry, healthspan, and survival, in an environment similar to the standard technique for group culture on solid media in Petri plates. This method is compatible with high-throughput data collection when used in conjunction with automated microscopy and analysis software. Finally, the limitations of this technique are discussed, as well as a comparison of this approach to a recently developed method that uses microtrays to culture isolated nematodes on solid media.

Introduction

Caenorhabditis elegans are commonly used in aging studies because of their short generation time (approximately 3 days), short lifespan (approximately 3 weeks), ease of cultivation in the laboratory, high degree of evolutionary conservation of molecular processes and pathways with mammals, and wide availability of genetic manipulation techniques. In the context of aging studies, C. elegans allow for the rapid generation of longevity data and aged populations for the analysis of late-life phenotypes in live animals. The typical approach for conducting worm aging studies involves manually measuring the lifespan of a population of worms maintained in groups of 20 to 70 animals on solid agar nematode growth media (NGM) in 6 cm Petri plates¹. Using age-synchronized populations allows the measurement of lifespan or cross-sectional phenotypes in individual animals across the population, but this method precludes monitoring the characteristics of individual animals over time. This approach is also labor-intensive, thus restricting the size of the population that can be tested.

There are a limited number of culture methods that allow for the longitudinal monitoring of individual C. elegans throughout their lifespan, and each has a distinct set of advantages and disadvantages. Microfluidics devices, including WormFarm², NemaLife³, and the "behavior" chip⁴, among others^5,6,7, allow the monitoring of individual animals over time. Culturing worms in liquid culture using multi-well plates similarly allows the monitoring of either individual animals or small populations of C. elegans over time^8,9. The liquid environment represents a distinct environmental context from the common culture environment on solid media in Petri plates, which can alter aspects of animal physiology that are relevant to aging, including fat content and the expression of stress-response genes^10,11. The ability to directly compare these studies to the majority of data collected on aging C. elegans is limited by differences in potentially important environmental variables. The Worm Corral¹² is one approach developed to house individual animals in an environment that more closely replicates typical solid media culture. The Worm Corral contains a sealed chamber for each animal on a microscope slide using hydrogel, allowing the longitudinal monitoring of isolated animals. This method uses standard brightfield imaging to record morphological data, such as body size and activity. However, animals are placed in the hydrogel environment as embryos, where they remain undisturbed throughout their lifespan. This requires the use of conditionally sterile mutant or transgenic genetic backgrounds, which limits both the capacity for genetic screening, as each novel mutation or transgene needs to be crossed into a background with conditional sterility, and the capacity for drug screening, as treatments can only be applied once to the animals as embryos.

An alternative method developed by the Fang-Yen lab allows the cultivation of worms on solid media in individual wells of a microfabricated polydimethylsiloxane (PDMS) device called a WorMotel^13,14. Each device is placed into a single-well tray (i.e., with the same dimensions as a 96-well plate) and has 240 wells separated by a moat filled with an aversive solution to prevent the worms from traveling between wells. Each well can house a single worm for the duration of its lifespan. The device is surrounded by water-absorbing polyacrylamide gel pellets (referred to as "water crystals"), and the tray is sealed with Parafilm laboratory film to maintain the humidity and minimize the desiccation of the media. This system allows healthspan and lifespan data to be gathered for individual animals, while the use of solid media better recapitulates the environment experienced by animals in the vast majority of published C. elegans lifespan studies, thus allowing more direct comparisons. Recently, a similar technique has been developed using polystyrene microtrays that were originally used for microcytotoxicity assays¹⁵ in place of the PDMS device¹⁶. The microtray method allows for the collection of individualized data for worms cultured on solid media and has improved capacity for containing worms under conditions that would typically cause fleeing (e.g., stressors or dietary restriction), with the trade-off being that each microtray can only contain 96 animals¹⁶, whereas the multi-well device utilized here can contain up to 240 animals.

Presented here is a detailed protocol for preparing multi-well devices that is optimized for plate-to-plate consistency and the preparation of multiple devices in parallel. This protocol was adapted from the original protocol from the Fang-Yen laboratory¹³. Specifically, there are descriptions for techniques to minimize contamination, optimize the consistent drying of both the solid media and the bacterial food source, and deliver RNAi and drugs. This system can be used to track individual healthspan, lifespan, and other phenotypes, such as body size and shape. These multi-well devices are compatible with existing high-throughput systems to measure lifespan, which can remove much of the manual labor involved in traditional lifespan experiments and provide the opportunity for automated, direct longevity measurement and health tracking in individual C. elegans at scale.

Protocol

1. Preparation of stock solutions and media

NOTE: Before beginning the preparation of the multi-well devices, prepare the following stock solutions and media.

1. Stock solutions for nematode growth media (NGM) and low-melt NGM (lmNGM):
   1. Prepare 1 M K₂HPO₄: Add 174.18 g of K₂HPO4 to a 1 L bottle, and fill it up to 1 L with sterile deionized water. Autoclave (121 °C, 15 psig) for 30 min, and store at room temperature (RT).
   2. Prepare 1 M KP_i, pH 6.0: Add 136.09 g of KH₂HPO₄ to a 1 L bottle, and fill it up to 1 L with sterile deionized water. Titrate with 1 M K₂HPO₄ to pH 6.0. Autoclave (121 °C, 15 psig) for 30 min, and store at RT.
   3. Prepare 1 M CaCl₂: Add 73.5 g of CaCl₂ to a 500 mL bottle, and fill it up to 500 mL with sterile deionized water. Autoclave (121 °C, 15 psig) for 30 min, and store at RT.
   4. Prepare 1 M MgSO₄: Add 123.25 g of MgSO₄ to a 500 mL bottle, and fill it up to 500 mL with sterile deionized water. Autoclave (121 °C, 15 psig) for 30 min, and store at RT.
   5. Prepare 5 mg/mL cholesterol: Combine 2.5 g of cholesterol, 275 mL of 100% ethanol, and 25 mL of sterile deionized water in a 500 mL amber bottle. Store at RT.
   6. Prepare 50 mM floxuridine: Combine 0.1231 g of floxuridine and 10 mL of sterile deionized water in a 15 mL conical tube. Filter-sterilize it with a 0.22 µm filter using a 10 mL syringe. Aliquot 1 mL each into 10 microcentrifuge tubes, and store them at −20 °C.
   7. Prepare 50 mg/mL carbenicillin: In a 15 mL conical tube, combine 500 mg of carbenicillin with 10 mL of sterile deionized water. Filter-sterilize it with a 0.22 µm filter using a 10 mL syringe. Aliquot 1 mL each into 10 microcentrifuge tubes, and store them at −20 °C.
   8. Prepare 1 mM isopropyl ß-D-1-thiogalactopyranoside (IPTG): In a 15 mL conical tube, combine 2.38 g of IPTG with 10 mL of sterile deionized water. Filter-sterilize it with a 0.22 µm filter using a 10 mL syringe. Aliquot 1 mL each into 10 microcentrifuge tubes, and store them at −20 °C.
   9. Prepare 100 mg/mL ampicillin: In a 15 mL conical tube, combine 1 g of ampicillin with 10 mL of sterile deionized water. Filter-sterilize it with a 0.22 µm filter using a 10 mL syringe. Aliquot 1 mL each into 10 microcentrifuge tubes, and store them at −20 °C.
2. Preparing nematode growth media (NGM) for general C. elegans maintenance
   1. To make 50 plates, in a 1 L flask, combine 10 g of Bacto agar, 1.5 g of NaCl, 1.25 g of Bacto peptone, and 486 mL of ultrapure water. Autoclave on a liquid cycle (121 °C, 15 psig) for at least 30 min to sterilize.
   2. Post autoclave, after the media has cooled to 55 °C, add 12.5 mL of 1 M KP_i, 500 µL of 1 M MgSO₄, 500 µL of 1 M CaCl₂, and 500 µL of 5 mg/mL cholesterol.
   3. Using a sterile technique, pour 10 mL of media into each 60 mm plate (50 total). After the media has solidified (at least 30 min after pouring), pipette 300 µL of bacterial culture (prepared according to step 4 and step 10) that has been grown overnight onto the center of the plate. Leave the plate on the bench, allowing the bacterial culture to dry and grow thicker (1-2 days). Store the plates at 4 °C.
3. Prepare the low-melt nematode growth media (lmNGM) pre-mixture:
   1. In a 500 mL flask, combine 4 g of low-melt agarose, 0.5 g of Bacto Peptone, and 195 mL of ultrapure water. Autoclave on a liquid cycle (121 °C, 15 psig) for at least 30 min to sterilize.
   2. While the media is still molten, distribute 10 mL aliquotsinto sterile glass test tubes. Seal each test tube with Parafilm followed by a test tube cap to preserve the lmNGM pre-mixture and prevent desiccation. The media will solidify and can be stored for ~2 weeks at RT before use.
4. Prepare 142.8 mM NaCl: In a 250 mL bottle, combine 0.8345 g of NaCl and 100 mL of sterile deionized water. Autoclave (121 °C, 15 psig) for 30 min, and store at RT.
5. Prepare detergent solution: In a 15 mL conical tube, combine 3 mL of Tween 20 and 7 mL of sterile deionized water. Mix thoroughly, filter-sterilize with a 0.22 µm filter using a 10 mL syringe, and store at RT.
6. Prepare copper sulfate solution: In a sterile 50 mL bottle, combine 20 mL of sterile deionized water, 0.5 g of CuSO₄, and 100 µL of detergent solution (prepared in step 1.5). Mix until the CuSO₄ has dissolved. Store at RT.
7. Prepare water-absorbing polyacrylamide crystals: In an autoclave-safe squeeze bottle, combine 150 mL of deionized water and 1 g of Type S super absorbent polymer. Mix gently by inverting the bottle several times. Autoclave on a liquid cycle (121 °C, 15 psig) for at least 20 min, and store at RT.
8. Prepare lysogeny broth (LB): In a 1 L or larger beaker, combine 20 g of LB powder with 1 L of deionized water. Aliquot into two 500 mL beakers. Autoclave (121 °C, 15 psig) for 30 min, and store at RT
9. Prepare lysogeny broth (LB) agar plates:
   1. In a 1 L flask, combine 10 g of LB powder, 7.5 g of Bacto Agar, and 500 mL of deionized water. Autoclave on a liquid cycle (121 °C, 15 psig) for 30 min.
   2. If desired, add optional antibiotics (post-autoclave, after media has cooled to 55 °C): 500 µL of 100 mg/mL ampicillin. Using a sterile technique, pour 20 mL of media into each 100 mm plate (25 total), and store at 4 °C.

2. Printing the 3D multi-well device mold

NOTE: Each device is molded from PDMS using a custom 3D-printed mold. A single mold can produce as many devices as needed; however, if attempting to prepare multiple devices at the same time, one 3D-printed mold is required for each device to be made in parallel.

1. Download the STL file (see Supplementary File 1).
2. Print the mold with a high-resolution 3D printer.
   NOTE: The resolution of the printer must be sufficiently high to allow consistent well and moat volumes. The suggested printer resolution is a minimum XY resolution of ~40 µm and a layer resolution of 28 µm. The material used by the 3D printer is equally important, as many material types will prevent the PDMS from curing. From experience, molds produced by high-resolution PolyJet printers (resolution: X-axis = 600 dpi, Y-axis = 600 dpi, Z-axis = 1,600 dpi) using the Vero Black print material work consistently. The 3D printing of the molds used in this study was outsourced (printing details can be found in the Table of Materials).

3. Preparation of the multi-well device

NOTE: This section describes how the 3D-printed mold is used to create the PDMS multi-well device.

1. Set the curing oven to 55 °C to allow preheating.
2. Determine the number of devices to be prepared in one batch. For each device, mix 60 g of PDMS base and 6 g of curing agent in a large, disposable weigh boat (or similar disposable container) using a disposable spatula.
3. Place the PDMS mixture in a vacuum chamber for 30 min at −0.08 MPa to remove bubbles.
4. Pour the PDMS mixture into each 3D-printed mold. Fill the mold completely, with the PDMS mixture extending slightly above the top of the mold. Keep excess PDMS mixture to refill the mold in case of spills.
5. Put the filled mold and any extra PDMS mixture in the vacuum chamber for 25 min at −0.08 MPa to remove any bubbles created during the pouring process.
6. Remove the filled mold from the vacuum chamber, and pop any remaining bubbles with a disposable spatula. Use the spatula to brush any visible debris or remaining bubbles to the edge of the mold away from the wells. Any remaining debris or bubbles can potentially interfere with the imaging in the later steps.
7. Ensure that the mold is filled completely with the PDMS mixture; it is common for a small portion of the mixture to spill out while in the vacuum chamber or during transfer. Refill using the extra PDMS mixture that was degassed in step 3.5. This should not create bubbles, but if any do form, they can be gently brushed to the side of the mold with the spatula.
8. Place a flat acrylic sheet on top of the PDMS-filled mold by first placing one edge and slowly lowering the other until the acrylic sits flat on top of the mold, displacing any PDMS mixture extending above the edge. If bubbles form while placing the acrylic sheet, slowly remove the acrylic, refill the mold with PDMS mixture if needed, and restart the acrylic placement. The acrylic sheet will form a flat bottom for the molded device and ensure that all the wells are at the same level relative to the base.
9. Place a 2.5 lb weight on top of the acrylic. Multiple molds, each with a separate acrylic sheet, can be stacked. Place the weighted molds in the oven, and let them cure at 55 °C overnight.
10. The next day, remove the weights and the molds from the oven.
11. Carefully work a razor blade between the acrylic and cured PDMS to break the seal. Carefully remove the acrylic from the top of the mold. The PDMS will have set by this point, and removing the acrylic can take some force.
12. Using a razor or a metal spatula, carefully loosen the sides of the cured PDMS from the mold. It is easy to tear the PDMS at this stage; work slowly and gently around the edge to remove the PDMS device intact.
    NOTE: Freshly printed molds are particularly sticky for the first three uses, and the PDMS often tears while trying to remove the device. With more uses, it becomes easier to remove the cured PDMS from the mold.
13. Wrap the device in aluminum foil, and seal it with autoclave tape. Autoclave on a dry cycle for at least 15 min at 121 °C, 15 psig to sterilize. After autoclaving, store the wrapped devices on the bench, and use them as needed.

4. Streaking the bacteria

NOTE: Begin preparing the bacteria that will be used as the worms' food source while they are on the multi-well device. The most common bacteria is Escherichia coli strain OP50 (or strain HT115 for RNAi experiments). Complete this step at least 2 days prior to adding the worms to the device.

1. Streak the bacteria onto a fresh LB plate. Ensure that any strain-specific selective additives (e.g., ampicillin to select a plasmid conferring ampicillin resistance to the bacteria) are included in the LB plates.
2. Incubate the plate at 37 °C overnight (~18 h) to allow the colonies to grow.

5. Preparation of the multi-well device for media loading

NOTE: The surface of the silicone PDMS material that makes up the device is hydrophobic, which prevents the small-volume wells and aversive moats from being filled with NGM and copper sulfate, respectively. To circumvent this problem, an oxygen plasma is used to temporarily modify the surface properties of the device to be hydrophilic, allowing the wells and moat to be filled within a limited time window (up to ~2 h). This section lays out the steps for completing the plasma-cleaning process. Complete this step at least 1 day before spotting the device wells with bacteria, as lingering effects of the plasma clean can interfere with spotting. Given the timing of sections 5-7, the practical limit for these steps per technician is three devices in parallel.

1. Preheat a dry bead bath incubator to 90 °C in preparation for section 6.
2. As the total media volume needed to fill the wells of a multi-well device is small compared to that for Petri plates, prepare a large batch of NGM, and aliquot it into test tubes (see step 1.3).
   NOTE: This can be done ahead of time, as the tubes of media can be stored on the bench for ~2 weeks. If the media has been prepared and aliquoted into tubes, proceed to step 5.3.
3. While wearing gloves, unwrap an autoclaved multi-well device; avoid touching any of the wells. Cut a small notch in the top-right corner of the device to indicate how many times it has been used/reused (see section 15 below for instructions and recommendations for cleaning and reusing each device).
4. Place up to three unwrapped devices into the plasma cleaner with the wells facing up. Run the plasma cleaner.
   NOTE : The following detailed instructions are for the Plasma Etch PE-50 plasma cleaner. The specific steps and settings need to be adapted and possibly re-optimized for other plasma cleaners.
   1. Check that the plasma cleaner power level is set to 75%.
   2. Ensure that the vent and isolation valves are both in the OFF position.
   3. Open the main valve on the oxygen tank. Wait until the regulator pressure drops to between 15 psig and 20 psig, and then turn the isolation valve to the ON position.
   4. Turn on the vacuum pump and plasma cleaner.
   5. Enter the following settings on the plasma cleaner (first use), or verify that the previously programmed settings are correct:
      Plasma Time: 3:00 min
      Vacuum Set Point: 149.5 mTorr
      Atmospheric Vent: 45 s
      Purge Vent: 5 s
      Gas Stabilize: 15 s
      Vacuum Alarm: 3:00 min
      Auto Cycle-Off: ON
   6. Press Enter to go to the Commands Menu. Use the Right Arrow key to select Setup Menu, and then press Enter. Scroll through all the settings by pressing Enter to confirm each current setting and then the Right Arrow to move to the next setting.
   7. Return to the Commands Menu by pressing the Up Arrow and then the Left Arrow.
   8. On the Commands Menu screen, press Enter. Select Commands PLASMA by pressing Enter. The system will cycle through the following phases:
      Plasma Pumpdown
      Gas Stabilize: During this phase, adjust the Gas 1 knob on the plasma cleaner until the flow meter is at 10 cc/min.
      Plasma Time
      Purge Pumpdown
      Plasma Cycle Complete
      Chamber Vent
      NOTE: This process should take approximately 5-10 min to complete. When all the phases are complete, the Commands Menu is displayed on the screen again. If during the Plasma Pumpdown phase, the pressure does not get low enough, the plasma cleaner will not proceed to the next phase, and the screen will display an error message. If this happens, check the vacuum pump oil. If the oil levels are low or the oil is cloudy/dirty, changing the oil can allow the vacuum pressure to reach the necessary level.
   9. Turn off the main valve on the oxygen tank.
   10. Very slowly, turn the vent valve to the ON position, and allow the regulator pressure of the oxygen tank to return to zero.
   11. Turn the isolation valve to the OFF position.
   12. Turn off the vacuum pump and plasma cleaner.
       NOTE: The hydrophilic surface modification of the multi-well device by the plasma cleaner is temporary and becomes progressively less effective over time (up to about 2 h). Proceed through section 6 and section 7 as quickly as possible.

6. Filling the wells with lmNGM

NOTE: A dry bead bath incubator should be on and preheated from step 5.1. Ensure that the bath has reached 90 °C.

1. Sterilize one single-well polystyrene tray per device by spraying the inside of the tray with 70% ethanol and wiping it dry with a task wipe.
2. Remove each device from the plasma cleaner with a gloved hand, and place it in a cleaned tray.
3. Place a 25 mL disposable pipette basin in the bead bath incubator after heating it to 90 °C.
4. Gather one tube of solidified lmNGM pre-mixture per device to be filled (see step 1.3).
5. Place the lmNGM pre-mixture test tubes in a 200 mL glass beaker, and remove the caps and Parafilm. Microwave for ~20 s until the media melts sufficiently to pour (the presence of some solid media is fine at this stage).
6. Combine the molten lmNGM pre-mixture from multiple test tubes into another sterile 200 mL beaker. Continue microwaving for an additional 20 s to reach a total of 40 s. If the lmNGM pre-mixture begins to boil over, stop the microwave, and allow the lmNGM pre-mixture to settle before continuing.
7. Remove the molten lmNGM pre-mixture from the microwave, and allow it to cool to ~60 °C.
   NOTE: At this stage, the media will cool and resolidify after ~5 min. Proceed immediately to step 6.8 and step 6.9.
8. For each 10 mL of lmNGM pre-mixture, add the following (in order): 250 µL of 1 M KP_i, 10 µL of 1 M MgSO₄, 10 µL of 1 M CaCl₂, and 10 µL of 5 mg/mL cholesterol.
   1. To prevent egg hatching when monitoring adults on the device, add 10 µL of 50 mM floxuridine.
   2. To select for an RNAi plasmid and to induce the expression of the RNA, add 5 µL of 50 mg/mL carbenicillin and 12 µL of 1 mM IPTG.
      ​NOTE: The antibiotics or other additives can be altered depending on the design of the experiment. Test compounds/drugs can also be added to the media at this stage.
9. Pour the molten lmNGM into the 25 mL basin in the bead bath.
10. Fill the device wells with lmNGM using a 200 µL multichannel repeater pipette.
    1. Set the repeater to dispense aliquots of 14 µL (14 µL can be dispensed up to 14 times if using a 200 µL pipette tip).
    2. Mount five pipette tips. The device's wells have the same spacing as a 384-well plate. Standard multichannel pipettes are spaced such that the user can pipette into every other well in a row/column.
    3. Load the tips with molten lmNGM. Dispense the first 14 µL back into the lmNGM-containing basin.
    4. Moving quickly but carefully, dispense 14 µL into the inner wells (white in Figure 1B), starting with those marked with an "x" in Figure 1B and moving across the plate to the right and then down, dispensing a total of 12 times (60 wells).
    5. Dispense the remaining lmNGM back into the basin, as the final aliquot is typically less than 14 µL.
    6. Repeat steps 6.10.3-6.10.5 until all the inner wells have been filled.
    7. Next, set the repeater to dispense aliquots of 15 µL. Repeat steps 6.10.3-6.10.5, but instead of the inner wells, fill the outermost ring of wells(gray in Figure 1B), beginning with the wells marked with "+" in Figure 1B and moving around the outer edge of the plate.
       NOTE: Work quickly so that the media does not solidify in the pipette tips. Avoid spilling any lmNGM into the moat. The outer wells tend to dry more quickly. The extra 1 µL of lmNGM allows the final dried well to have a level surface in the same drying time as the inner wells.
11. As a quality control step, examine each well to identify those with flaws that may interfere with imaging, including wells that are underfilled (the lmNGM surface is sunken below the edge of the well), overfilled (the lmNGM surface has a domed top), and contain bubbles or debris.
12. Remove the solidified lmNGM from the unsatisfactory wells with a short platinum wire pick or vacuum aspirator. Refill the empty wells with fresh molten lmNGM. Also, remove any media that overflowed into the moat with a short platinum wire pick or vacuum aspirator.

7. Adding copper sulfate to the moat

NOTE: This device's wells are surrounded by a continuous moat. Here, the moat is filled with copper sulfate, which acts as a repellent and deters the worms from fleeing from their wells.

1. Using a 200 µL pipette, dispense 200 µL of copper sulfate solution (see step 1.6) into the device moat in each corner, 2x per corner. This should be a large enough volume for the copper sulfate to flow through the entire moat. Be careful not to overfill the moat at any point; the copper sulfate should not touch the top surface of the wells.
2. If the copper sulfate does not flow easily through the entire moat, use a short platinum wire pick to help break the tension and drag the copper sulfate through the moat.
3. After the copper sulfate has flowed through the entirety of the moat, remove as much copper sulfate as possible from the moat using a 200 µL pipette or aspiration with a vacuum. The residue left behind will be sufficient to deter the worms from leaving their wells. Leaving the moat full of copper sulfate solution risks copper sulfate spilling into the wells, which would cause worms to flee.

8. Adding autoclaved water crystals

NOTE: To maintain humidity within the plate and prevent desiccation of the lmNGM, each device is surrounded by saturated water-absorbing polyacrylamide crystals.

1. Prepare the water crystals (see step 1.7).
2. Using a squeeze bottle, add the water crystals in the spaces between the device and the tray walls. Close the tray lid, and wrap all four sides with a piece of Parafilm. Add two additional pieces of Parafilm to completely seal the plate.
   NOTE: Water crystals can be prepared in a beaker and scooped into the tray with a sterilized spatula. However, this adds time to the procedure and increases the time that each tray is open and exposed to potential contaminants.
3. Leave the sealed device on the bench until the next day when the bacteria are ready to be spotted. Ensure that the devices are stored with the wells facing up after the lmNGM has been added.

9. Preparation of an age-synchronized population of worms

NOTE: The following steps yield a synchronized population of worms that are ready to add to the multi-well device at the fourth larval stage (L4). However, worms at different stages of development can also be added. This step should be completed 2 days before adding the worms to the device if L4s are desired. Adjust the timing of synchronization for the desired life stage.

1. For consistent aging studies, maintain C. elegans on standard NGM plates (see step 1.2 at 20 °C under well-fed conditions).
2. Obtain a synchronized population of animals from a stock plate via standard methods, for example, bleaching¹⁷ or timed egg laying¹.
3. Add isolated eggs to an NGM plate spotted with bacteria. The eggs will hatch on this plate, and the worms will reach the L4 larval stage in 2 days for wild-type animals.

10. Inoculating the bacterial culture

NOTE: Bacteria are used as the primary food source for C. elegans, most commonly E. coli strains OP50 or HT115. The bacteria are concentrated 10-fold, which should be accounted for in the volume of the prepared culture. Prepare a bacterial culture the day before spotting the device.

1. From the LB plate prepared in step 4, pick a single colony, and inoculate 12 mL of sterile LB per device to be spotted. Include selective agents if necessary for the bacteria strain being used.
2. Grow the bacterial culture overnight (~18 h) in a 37 °C incubator with shaking at ~250 rpm.

11. Spotting the wells with concentrated bacteria

NOTE: A small volume of concentrated bacteria is added to each well, which is sufficient to feed the worms for their entire lifespan on the device. The bacterial culture needs to be dried before the worms can be added to the wells. As the media volume in each well is small (14-15 µL) relative to the bacteria volume added (5 µL), the chemical content of the bacterial media can impact the chemical environment of the well. To account for this, the bacteria are concentrated and resuspended in salt water to remove depleted LB while avoiding hypoosmotic stress. There is no salt added to the lmNGM recipe (see steps 1.3-1.4) as it is added at this stage.

1. After growing the bacteria overnight as described in section 10, concentrate the bacterial culture 10x. Pellet the bacteria by centrifuging at ~3,400 x g for 20 min, dispose of the supernatant, and resuspend the pellet in 1/10 of the original culture volume of 142.8 mM NaCl (see step 1.4). For example, to spot one 240-well device, spin down 12 mL of culture and resuspend in 1.2 mL of 142.8 mM NaCl.
   NOTE: Test compounds/drugs can be added to the resuspended bacterial culture before spotting.
2. Using a repeat pipette, spot each well with 5 µL of the concentrated bacteria. Avoid direct contact with the lmNGM surface, as the pipette tip can pierce the lmNGM and allow the worm to burrow below the surface of the well. Be careful not to let the bacteria spill into the moat because the worms will be attracted to it and flee into the moat.
3. Dry the spotted bacteria with the tray lids removed. This step is important for the long-term integrity of the well environment, and there are a variety of ways to dry the spotted devices. Whatever method is used, ensure that the device remains in a sterile environment, as it needs to sit uncovered while the bacteria dry. Increased airflow greatly reduces the drying time. Drying can be performed as described in the steps below:
   1. Leave the device uncovered in a clean, sealed container, such as a plastic bin that has been cleaned with 10% bleach, followed by 70% ethanol. This method of drying can take several hours.
   2. Place the devices uncovered in a sterile laminar flow hood (similar to the preferred method below).
   3. Dry the devices using a custom-built "drying box" (the optimized method followed in this study), which can be constructed at minimal cost using computer case fans and HEPA filters (see the Supplementary File 1).
      NOTE: Regardless of the method used, the drying step has to be optimized for the local environment. Frequently monitor how quickly the wells are drying until the typical drying time has been identified. In a low humidity environment, the use of a drying box results in a drying time of ~30-40 min. Do not let the plates get too dry, as the media in the wells will shrink and sink. It is better to remove the device from drying while a few wells are still wet than to let it dry longer and potentially over-dry a large number of wells.
4. Check the water crystals. If the majority of the water crystals in the tray have dried out and lost volume during the drying process, add more to the tray.
5. After the bacteria are dry, add the worms immediately (section 12) or seal the tray to use later. To seal, close the tray lid, and wrap all four sides with a single piece of Parafilm. Repeat twice for a total of three Parafilm layers. After wrapping the tray completely, the device can be left on the bench at room temperature for up to 4 days before adding the worms (section 12).

12. Adding worms to the multi-well device

1. Manuallyadd one worm per well to each well using a platinum pick to transfer the animals from the plates of worms prepared in section 9. Only pick worms that are at the desired life stage and age.
   ​NOTE: Taking more than 1 h to add the worms to the device can result in lmNGM desiccation, so the worms should be added as quickly as possible. Picking multiple worms (20+) at a time before adding them to the device wells can help increase the speed.

13. Finishing the preparation of the device for long-term use

NOTE: These steps ensure that the device wells remain hydrated for the duration of the experiment.

1. Examine the water crystals. Ensure that the water crystals are level with the top of the device but not overflowing onto it. Add additional water crystals to the tray if necessary.
2. Add a drop of prepared detergent solution (see step 1.5) to the inside of the tray lid, and rub in with a task wipe until the detergent solution has dried. This prevents fogging on the lid after the device is sealed inside the tray so that the worms are clearly visible.
3. Wrap the tray with three pieces of Parafilm using a specific technique that promotes the long-term integrity of the Parafilm seal.
   1. Slightly stretch a piece of Parafilm such that it covers only two sides of the tray. Repeat this with a second piece of Parafilm to cover the other two sides.
   2. Layered on top of the first layer of Parafilm, take a final piece of Parafilm, stretch it fully, and wrap all four sides. If properly sealed, the device wells should remain hydrated for ~2 months.
      ​NOTE: The Parafilm integrity should be monitored every 1-2 weeks, and the Parafilm should be replaced if broken.
4. Remove any fingerprints from the top of the tray using a task wipe wetted with 70% ethanol.

14. Collection of the data

NOTE: The purpose of this study is to describe the culture methodology. Once populated, multi-well devices are compatible with the longitudinal monitoring of a variety of phenotypes. Here, basic guidance for measuring several of the most common parameters is provided.

1. Lifespan: Monitor the lifespan by tapping the plate or exposing the worms to a bright blue light every 1-3 days. Score as dead if no movement is observed. These multi-well devices are also compatible with automated imaging and analysis pipelines for estimating lifespan^13,18.
2. Activity: Monitor the activity of individual animals throughout life by taking static images or videos of the animals in the device wells and assessing the distance traveled, speed, or other metrics of movement. The devices are also compatible with automated imaging and analysis pipelines for estimating activity^13,18.
3. Body size and shape: Monitor the changes in body size and shape by taking static images of the animals in the device wells and quantifying the visual parameters using standard imaging techniques. In principle, this culture method should be compatible with existing software for a more sophisticated assessment of worm body shape^19,20,21.

15. Reusing the devices

NOTE: After an experiment is complete, the multi-well devices can be cleaned and reused up to three times. Additional reuse begins to impact the worm phenotypes, possibly caused by chemicals from the media or bacteria building up in the walls of the PDMS material.

1. Discard the Parafilm, and remove the device from the tray. Check the notches on the top-right corner of the device, which indicate how many times it has been used. If it has been used three times, discard the device. If it has been used fewer than three times, clean and reuse it.
   NOTE: The trays are made from polystyrene and are not directly in contact with the lmNGM, bacteria, or any additives to the media. They can be reused many times as long as they remain visually clear.
2. Rinse out any remaining water crystals from the tray. Spray the inside of the tray with 10% bleach solution, and let it sit for 10 min. Rinse with deionized water and dry immediately to avoid water spots.
3. Hold the device under running water to begin cleaning the media out of the wells. Gently bend and twist the device under the water to help loosen the media from the wells, but do not bend so far that the device tears. Pick out any remaining media stuck in wells with a 200 µL pipette tip.
4. Fill a 2 L beaker with a stir bar ~3/4 full with deionized water, and bring to the boil on a hot plate.
5. Add one or two multi-well devices and a stir bar to the boiling water. Turn on the magnetic stirring to a low speed (~200 rpm) to gently agitate the devices in the water. Allow the devices to boil for 10 min.
6. Remove the devices from the water with metal tweezers, and place them well-side down on paper towels to drain. Allow the devices to dry at a minimum overnight.
7. When the devices are completely dry, wrap them in foil, and seal them with autoclave tape. Write on the autoclave tape the number of times the device has been used. Autoclave on a dry cycle for 15 min at 121 °C to sterilize. The devices are now ready for reuse.

Representative Results

The WorMotel culture system can be used to gather a variety of data, including regarding lifespan, healthspan, and activity. Published studies have utilized multi-well devices to study lifespan and healthspan^13,14, quiescence and sleep^22,23,24, and behavior²⁵. Lifespan can be scored manually or through a collection of images and downstream imaging analysis. In the former approach, the worms can be manually observed following a stimulus (e.g., tapping the plate or exposure to blue light) every 1-3 days and scored as dead if no movement is observed, similar to standard methods on Petri plates¹. The latter approach is similar, except that worm movement can be determined by comparing frame-to-frame differences between the images taken after the stimulus has been applied. This provides an added benefit in that movement provides information both on the activity level of individual animals at that time point and provides a metric by which lifespan (e.g., cessation of movement) and healthspan (multiple definitions have been proposed) can be determined. The images can further be used to extract additional physiological parameters such as body size, body shape, and body posture.

To demonstrate the capacity of the system, we examined the classical epistatic relationship between the insulin receptor, encoded by the gene daf-2, and the downstream FOXO family transcription factor encoded by daf-16 in the context of lifespan, healthspan, and daily activity for individual animals. Wild-type (strain N2) and daf-16(mu68) loss-of-function (strain CF1038) C. elegans fed E. coli (strain HT115) expressing either control (empty vector; EV) or daf-2 RNAi feeding constructs were cultured in multi-well devices, and each animal was monitored for lifespan (Figure 2A), healthspan (Figure 2B), and daily activity (Figure 2C). The activity was monitored daily by taking a series of still images every 5 s for 2 min, with the worms being exposed to bright blue light for 5 s at 1 min to stimulate activity (as per Churgin et al.¹³). The daily activity for each animal was estimated by normalizing the background across wells and images, identifying the worm area in each image, and calculating the change in area between adjacent images. Lifespan was defined as the age at which activity was last observed for each worm, and healthspan was defined as the age at which a worm could no longer move a full body length. As expected from numerous previous studies (e.g., Kenyon et al.²⁶, Murphy et al.²⁷), the daf-16(mu86) mutation resulted in short lifespans and prevented lifespan extension from the RNAi knockdown of daf-2 (Figure 2A). A similar pattern was observed for healthspan (Figure 2B). As an advantage of using multi-well device culture systems, the capacity to track individual animals throughout life allows a detailed analysis of the individual variation in each measured phenotype across the population. For instance, the variation in lifespan and healthspan across individual animals can be compared in either absolute (Figure 2D) terms or as a fraction of the total lifespan (Figure 2E). Early-life phenotypes can further be compared to late-life phenotypes, including lifespan, in individual animals across a population. For instance, the cumulative activity for each individual animal across the lifespan (i.e., the area under the curve [AUC] for individual activity) correlated better with lifespan (Figure 2F) than cumulative lifespan up to day 5 of life (Figure 2G) across all the conditions measured. We emphasize that the purpose of this work is to provide a detailed protocol for constructing the multi-well environment for tracking individual animals over time, not for measuring a specific phenotype using the device. The representative results presented in Figure 2 provide just one example of the phenotypes that can be measured in this system. Once constructed, the multi-well environment is compatible with a wide range of techniques for measuring the phenotypes of free-crawling worms on solid media.

Figure 1: Schematic of the microfabricated multi-well devices. (A) Individual C. elegans are cultured on solid low-melt nematode growth media (lmNGM) agarose pads seeded with bacterial food in individual wells. The space between the wells is coated with an aversive chemical (copper sulfate) to isolate each worm within its well. Each device is secured inside a single-well tray. The perimeter of the tray is filled with water crystals to maintain the humidity. The tray is sealed with Parafilm to allow oxygen exchange. Image created with BioRender.com. (B) Overview of the multi-well device indicating the suggested order for loading the wells. The inner wells (white) receive 14 µL of lmNGM. The outer wells (gray) receive 15 µL of lmNGM. Please click here to view a larger version of this figure.

Figure 2: Correlation of measured phenotypes across populations in individual animals using multi-well devices. All the panels provide data from the same experiment comparing four groups of animals: wild-type (N2) animals subject to empty vector EV(RNAi) (N = 138), wild-type animals subject to daf-2(RNAi) (N = 151), daf-16(mu86) animals subject to EV(RNAi) (N = 123), and daf-16(mu86) animals subject to daf-2(RNAi) (N = 135). (A) The lifespan extension from daf-2(RNAi) is blocked by the daf-16(mu86) null mutation. Pairwise significance between groups determined by the log-rank test (survdiff function in R). (B) The healthspan-defined here as the day at which an animal can no longer move a full body length-extension from daf-2(RNAi) is blocked by the daf-16(mu86) null mutation. Pairwise significance between groups determined by the log-rank test (survdiff function in R). (C) The 3 day rolling mean of activity across the lifespan is reduced by both daf-16(mu86) and daf-2(RNAi). Significance calculated by the Mann-Whitney U test to compare the area under the curve for activity across the lifespan for individual animals between groups. (D) Healthspan and lifespan for each population as absolute values (mean ± standard error of the mean). (E) Healthspan and lifespan for each population normalized to total lifespan within each group (mean ± standard error of the mean). (F) The cumulative activity across the lifespan (area under the curve [AUC] across the lifespan) for individual animals correlates better with lifespan than (G) the activity for individual animals at any specific day across the lifespan (the activity correlation on day 8, representing the point at which the mean activity is maximized, is shown), as calculated by linear regression (lm function in R). n.s. = not significant, * p < 0.05, ** p < 0.01, *** p < 0.001. All p-values were adjusted for multiple comparisons using the Bonferroni method for comparisons made within each panel. Please click here to view a larger version of this figure.

Supplementary File 1: STL file for printing the 3D multi-well device mold Please click here to download this File.

Discussion

The WorMotel system is a powerful tool for gathering individualized data for hundreds of isolated C. elegans over time. Following the earlier studies using multi-well devices for applications in developmental quiescence, locomotory behavior, and aging, the goal of this work was to optimize the preparation of multi-well devices for the long-term monitoring of activity, health, and lifespan in a higher-throughput manner. This work provides a detailed protocol for preparing multi-well devices that optimizes many of the steps from the original protocol¹³, highlights key points that may present technical difficulties, and provides a discussion about reusing the plates and other materials.

For the purposes of scaling-as a lab that currently prepares between 10 and 20 devices in a typical week-a top consideration was whether the devices could be reused and, if so, to what degree. There is a higher cost in terms of both time and money in preparing PDMS devices relative to conducting traditional culture on Petri plates, but these higher costs can be reduced by reusing the PDMS or other components of the system. With many reuses, the PDMS began to develop a yellow coloration, likely reflecting the accumulation of compounds from the lmNGM media or bacteria. The animals cultured on these plates also displayed a higher rate of fleeing and reduced lifespans. Based on dozens of experiments, three uses are optimal for reusing these PDMS devices, allowing age-related phenotypes to be assessed without a measurable impact from PDMS degradation while reducing the number of new devices that need to be molded (thus saving on costs). We further confirmed that experiment-matched animals grown on devices on their first, second, and third use produced survival curves that were nearly indistinguishable and not statistically different (data not shown). The trays used to contain the devices are made of polystyrene and can be cleaned and reused indefinitely if they remain free of scratches or other marks that could interfere with visualizing the worms.

A key challenge for preparing multi-well devices for applications that last more than ~2 weeks is the prevention of plate contamination by environmental bacteria and fungus. There are multiple steps at which sterilization is critical for preventing contamination. These include autoclaving all the devices prior to use, boiling the devices intended for reuse, autoclaving the absorbent water crystals used to maintain humidity, cleaning the trays that contain the devices with both bleach and ethanol before use, and filter-sterilizing the detergent solution that is applied to the lid of the tray to prevent fogging and added to the copper sulfate solution. Implementing each of these changes notably reduced contamination events, allowing the sealed devices to be consistently used for longitudinal monitoring across the lifespan, even under pro-longevity conditions (e.g., knockout of daf-2) that require monitoring for >45 days. The protocol described here includes two modifications to the original protocol for long-term applications designed to maintain consistent well drying and prevent desiccation. First, the overall depth of the device was increased by 2 mm to increase the capacity for water crystals. Over long experiments, particularly in a low-humidity environment, too few water crystals in the tray led to the desiccation of the lmNGM. Along with more water crystals, it was necessary to increase the volume of agarose that was added to the wells along the edge of the device. These wells tended to dry first following the well loading (section 6) and shrink. Using 14 µL of agarose on the inside wells was enough volume to fill the wells completely without creating the domed top that results from over-filling the wells. Adding 15 µL of agarose to the outer wells provided enough volume that, when the wells began drying out, they shrank to a level comparable to the 14 µL added in the inner wells.

One of the largest deviations from the original protocol was to reverse the order in which the user loads the lmNGM (section 6) and copper sulfate (section 7). Originally, the copper sulfate was added to the moat first, followed by filling the wells with lmNGM¹³. It was observed that filling the wells with lmNGM as soon after plasma cleaning as possible improved the adherence of the lmNGM to the well walls. Waiting too long after the plasma cleaning resulted in wells with bubbles and domed tops, which can interfere with visualizing the worms. Prioritizing filling the wells over adding the copper sulfate is particularly important when preparing multiple devices at the same time to ensure consistent, high-quality lmNGM surfaces. A downside of filling the wells first is that the hydrophilic surface modification produced by the plasma cleaner will have worn off noticeably by the time the user moves on to adding the copper sulfate. Copper sulfate does not flow easily through the moat when the surface becomes less hydrophilic, thus making it challenging to achieve complete coverage. Adding detergent to the copper sulfate solution to act as a surfactant improves the flow of the solution through the moat. A platinum wire pick can also be used to gently guide the copper sulfate through the moat by breaking the tension at any points where the copper sulfate is having difficulty flowing. Furthermore, if the copper sulfate solution were left in the moat, it would easily spill into the wells and contaminate the lmNGM surface when tilting the tray. The nature of the loading process makes it nearly impossible to keep the device sufficiently level throughout to prevent copper from contaminating a subset of the wells. To account for this, the copper sulfate solution is removed from the moat (step 7.3), and the residual copper sulfate left behind is sufficient to deter most worms from fleeing. As a final note on the use of copper sulfate as an aversive barrier, some repellents can affect age-associated phenotypes, including lifespan. The use of copper sulfate in these multi-well devices was examined by Churgin et al.¹³ and found to have no detectable impact on either lifespan or development.

Other minor updates to the protocol focus on improving the steps taken to prepare the PDMS. An extra degassing step after mixing the PDMS base and curing agent was added, as this is a process that generates many bubbles. Removing the majority of the bubbles before adding the mixture to the device molds minimizes the bubbles remaining after the second degassing step. To ensure that the bottom of the device is entirely flat and even, a feature that is important for whole-plate imaging, a piece of glass or acrylic was laid across the top of the filled mold. This step is not essential-though still helpful-for applications that only examine one well at a time, as the user can manually adjust the focus for each well. Finally, curing the PDMS at a higher temperature (55 °C) was necessary at the location where this version of the protocol was optimized (Tucson, Arizona, USA), in contrast to the 40 °C indicated by the original protocol (optimized in Philadelphia, Pennsylvania, USA). This suggests that differences between locations (such as climate or the precise reagents and equipment used) can affect the specific steps in the protocol, such as the curing temperature or drying techniques, and that these may need to be optimized at each site. For instance, environmental humidity plays a large role in determining the drying time for the plates after spotting, and this can vary greatly between locations or seasons.

In principle, this multi-well device system can be used to collect data on any phenotype that can be measured under standard brightfield microscopy on Petri plates-lifespan, healthspan, unstimulated/stimulated movement, body size and shape, movement geometry-with the added capability of tracking those metrics for individual worms over time. As noted in the introduction, there are other methods for acquiring individual whole-lifespan data. Microfluidics devices²⁸ or multi-well plates²⁹ offer the capability to follow the life history of individual animals, but only by culturing the worms in liquid media. A liquid environment can alter the worms' transcriptome^10,11,30 relative to solid media and induces distinct physiological and behavior changes, including episodic swimming³¹, increased energy expenditure^10,32, and elevated oxidative stress¹⁰. The degree to which lifespan and other metrics related to healthy aging can be directly compared between liquid and solid media is unclear. Solid culture multi-well devices allow single worm tracking in an environment that is comparable to standard group culture on Petri plates. The curved structure of the device wall allows for imaging of worms anywhere inside the well, making these devices compatible, in principle, with more sophisticated tools for single-worm analysis, such as Worm Tracker³³.

The ability to monitor individual animals throughout the course of an experiment on a multi-well device is accompanied by several limitations. First, even with an optimized protocol, these devices require more time to set up per animal than standard culture on Petri plates, thus providing single-animal data at the cost of overall sample size. However, the worms are also isolated to individualized wells, removing some of the complications associated with automated lifespan measurement, such as automatically detecting individual animals that stop moving while touching one another. Automation more than recaptures the time lost to the more complex setup and provides the opportunity for more high-throughput screening^13,18. Second, experimental conditions that the worms dislike more than the copper sulfate moat, such as strong stress-inducing agents (e.g., paraquat) or dietary restriction, will lead to the worms fleeing from the wells and into the moat. Third, while the PDMS devices allow brightfield and darkfield microscopy, the PDMS material tends to have both high and uneven background fluorescence, the latter likely resulting from dust or other microparticles becoming embedded in the PDMS during molding, which limits their application in fluorescent microscopy. Finally, the discoloration observed in the PDMS for devices reused over multiple experiments and the associated reduction in lifespan and increase in fleeing suggests that the media components and other added chemicals leech into the PDMS over time. The degree to which this may impact aging phenotypes or drug treatments is currently being explored. As mentioned previously, there is an alternative method for a single-worm culture that is similar to PDMS multi-well devices but instead uses commercially available microtrays to culture isolated animals¹⁶. These microtrays are made from polystyrene, avoiding the potential issue of leeching media components, and have consistent fluorescence background, allowing for direct in vivo fluorescence imaging directly on the plate. The microtray wells are also surrounded by palmitic acid in place of the copper sulfate used in the protocol described here, which is more aversive and reduces the fraction of worms that leave their wells, even in the case of stress-inducing agents and dietary restriction. These benefits are realized at the cost of lower packing efficiency of the wells, as the microtrays allow a maximum of 96 worms to be cultured in a single tray in contrast to the 240 worm capacity of the PDMS devices. The specifics of the desired outcome of an experiment will influence which of these systems should be utilized.

Materials

Name	Company	Catalog Number	Comments
2.5 lb weight	CAP Barbell	RP-002.5
Acrylic sheets (6 in x 4 in x 3/8 in)	Falken Design	ACRYLIC-CL-3-8/1224	Large sheet cut to smaller sizes
Ampicillin sodium salt	Sigma-Aldrich	A9518
Autoclavable squeeze bottle	Nalgene	2405-0500
Bacto agar	BD Difco	214030
Bacto peptone	Thermo Scientific	211677
Basin, 25 mL	VWR	89094-664	Disposable pipette basin
Cabinet style vacuum desiccator	SP Bel-Art	F42400-4001	Do not need to use dessicant, only using as a vacuum chamber.
CaCl2	Acros Organics	349615000
Caenorhabditis elegans N2	Caenorhabditis Genetics Center (CGC)	N2	Wildtype strain
Carbenicillin	GoldBio	C-103-25
Centrifuge	Beckman	360902
Cholesterol	ICN Biomedicals Inc	101380
Compressed oxygen tank	Airgas	UN1072
CuSO4	Fisher Chemical	C493-500
Dry bead bath incubator	Fisher Scientific	11-718-2
Escherichia coli OP50	Caenorhabditis Genetics Center (CGC)	OP50	Standard labratory food for C. elegans
Ethanol	Millipore	ex0276-4
Floxuridine	Research Products International	F10705-1.0
Hybridization oven	Techne	731-0177	Used to cure PDMS mixture, any similar oven will suffice
Incubators	Shel Lab	2020	20 °C incubator for maintaining worm strains and 37 °C incubator to grow bacteria
Isopropyl ß-D-1-thiogalactopyranoside (IPTG)	GoldBio	I2481C100
K2HPO4	Fisher Chemical	P288-500
KH2PO4	Fisher Chemical	P286-1
Kimwipes	KimTech	34155	Task wipes
LB Broth, Lennox	BD Difco	240230
Low melt agarose	Research Products International	A20070-250.0
MgSO4	Fisher Chemical	M-8900
Microwave	Sharp	R-530DK
Multichannel repeat pipette, 20–200 µL LTS EDP3	Rainin	17013800	The exact model used is no longer sold, a similar model's catalog number has been provided
NaCl	Fisher Bioreagents	BP358-1
Nunc OmniTray	Thermo Scientific	264728	Clear polystyrene trays
Parafilm M	Fisher Scientific	13-374-10	Double-wide (4 in)
Petri plate, 100 mM	VWR	25384-342
Petri plate, 60 mM	Fisher Scientific	FB0875713A
Plasma cleaner	Plasma Etch, Inc.	PE-50
PLATINUM vacuum pump	JB Industries	DV-142N
PolyJet 3D printer	Stratasys	Objet500 Connex3	PolyJet 3D printing services provided by ProtoCAM (Matrial: Vero Rigid; Finish: Matte; Color: Gloss; Resolution: X-axis: 600 dpi, Y-axis: 600 dpi, Z-axis: 1600 dpi)
Shaking incubator	Lab-Line	3526CC
smartSpatula	LevGo, Inc.	17211	Disposable spatula
Superabsorbent polymer (AgSAP Type S)	M2 Polymer Technologies	Type S	Referred to in main text as "water crystals"
SYLGARD 184 Silicone Elastomer base	The Dow Chemical Company	2065622
SYLGARD 184 Silicone Elastomer curing agent	The Dow Chemical Company	2085925
Syringe filter (0.22 µm)	Nest Scientific USA Inc.	380111
Syringe, 10 mL	Fisher Scientific	14955453
TWEEN 20	Thermo Scientific	J20605-AP	Detergent
Vacuum pump oil	VWR	54996-082
VeroBlackPlus	Stratasys	RGD875	Rigid 3D printing filament
Weigh boat	Thermo Scientific	WB30304	Large enough for PDMS mixture volume