A high throughput screen of synthetic small molecules was conducted on the model plant species, Arabidopsis thaliana. This protocol, developed for a liquid handling robot, increases the speed of forward chemical genetics screens, accelerating the discovery of novel small molecules affecting plant physiology.
Chemical genetics is increasingly being employed to decode traits in plants that may be recalcitrant to traditional genetics due to gene redundancy or lethality. However, the probability of a synthetic small molecule being bioactive is low; therefore, thousands of molecules must be tested in order to find those of interest. Liquid handling robotics systems are designed to handle large numbers of samples, increasing the speed with which a chemical library can be screened in addition to minimizing/standardizing error. To achieve a high-throughput forward chemical genetics screen of a library of 50,000 small molecules on Arabidopsis thaliana (Arabidopsis), protocols using a bench-top multichannel liquid handling robot were developed that require minimal technician involvement. With these protocols, 3,271 small molecules were discovered that caused visible phenotypic alterations. 1,563 compounds induced short roots, 1,148 compounds altered coloration, 383 compounds caused root hair and other, non-categorized, alterations, and 177 compounds inhibited germination.
In the past 20 years researchers in the field of plant biology have made great strides using chemical genetics approaches, both forward and reverse, improving our understanding of cell wall biosynthesis, the cytoskeleton, hormone biosynthesis and signaling, gravitropism, pathogenesis, purine biosynthesis, and endomembrane trafficking1,2,3,4,5. Employing forward chemical genetics techniques enables the identification of phenotypes of interest and allows researchers to understand the genotypic underpinnings of particular processes. Conversely, reverse chemical genetics seeks out chemicals that interact with a pre-determined protein target6. Arabidopsis has been at the forefront of these discoveries in plant biology because its genome is small, mapped, and annotated. It has a short generation time, and there are multiple mutant/reporter lines available to facilitate the identification of aberrant subcellular machinery7.
There are two major bottlenecks that slow the progress of forward chemical genetic screens, the initial screening process and determining the target of the compound of interest8. A major aid in increasing the speed of small molecule selection is the use of automation and automated equipment9. Liquid handling robots are an excellent tool for handling large libraries of small molecules and have been instrumental in driving progress in the biological sciences10. The protocol presented here is designed to alleviate the bottleneck associated with the screening process, enabling the identification of bioactive small molecules at a rapid rate. This technique decreases the burden of labor and time on behalf of the operator while also decreasing the economic cost to the principle investigator.
Thus far, most chemical libraries analyzed have held between 10,000 and 20,000 compounds, some with as many as 150,000 and some with as few as 709,11,12,13,14,15,16. The protocol introduced herein was implemented on a small molecule library of 50,000 compounds (see Table of Materials), one of the larger forward chemical genetics screens conducted on Arabidopsis to date. This protocol fits with the current trend towards increased efficiency and speed regarding forward chemical genetics, especially as it pertains to herbicide discovery, insecticide discovery, fungicide discover, drug discovery, and cancer biology17,18,19,20,21. Though implemented here with Arabidopsis, this protocol, could easily be adapted to cell cultures, spores, and potentially even insects in liquid medium within 96-, 384-, or 1536-well plates. Due to its small size, Arabidopsis is amenable to screening in 96 well plates. However, distributing seeds evenly among wells is a challenge. Hand seeding is accurate but labor intensive, and though there are devices designed to dispense seeds into 96-well plates, they are expensive to purchase. Here, we show how this step can be circumvented with just a small loss in accuracy.
The overall goal of this method was to make screening a large chemical library against Arabidopsis more manageable, without compromising accuracy, via the use of a liquid handling robot. The use of this method improves the efficiency of the researcher by decreasing the time taken to complete initial dilution series management and subsequent phenotypic screens, allowing quick visualization of samples under a dissecting microscope, and rapid identification of novel bioactive small molecules. Figure 1 depicts this protocol's key outcomes in 4 steps.
Figure 1: Overall workflow of the forward chemical genetics screen. An overview of the protocol to be described with some detail for each of the 4 key steps. 1: Receiving the Chemical Library, 2: Making the Dilution Library, 3: Making the Screening Plates, and 4: Incubating and visualizing the Screening Plates. Please click here to view a larger version of this figure.
1. Creating a Dilution Library
2. Adding Media-seed Mixture to Screening Plates
3. Adding Small Molecules to Screening Plates
4. Incubation and Visualization of Screening Plates
The ability to accurately and efficiently characterize phenotypes based on the addition of small molecules at screening concentrations under a dissecting microscope is the end goal of this method of forward chemical genetics on Arabidopsis. The phenotypes observed when all 50,000 compounds had been screened was diverse and could be broken into several distinct classes (Figure 2). Figure 3A-F depicts examples of phenotypes that were observed at low magnification under a dissecting microscope. Some phenotypes provided inconclusive results (Figure 3G, H). These had to be retested at lower concentrations to ensure that the chemical didn't provide a different phenotype at a lower dose.
Poor results can arise for many reasons. One is poor germination rates of the seeds. This can cause a screening plate to predominantly exhibit no germination or incomplete germination phenotypes (Figure 3G, H), which can be misleading. To overcome this, pre-test germination rates for all seeds used. Once germination rates have been established, and are greater than 95% for Arabidopsis, vernalization of seed prior to chemical addition is a key step ensuring simultaneous germination. Lack of simultaneous germination can lead to false positives in phenotyping. In addition to this, poor results can arise if media is allowed to evaporate during incubation. This lack of hydration prevents seeds from germinating and can be avoided through the use of desiccation-proof containers. Additionally, DMSO and media solutions are in the two external columns of every plate, ensuring proper micro climates and germination rates are obtained.
Satisfactory experimental results are achieved when germination rates are >95%, seeds are vernalized prior to addition into 96-well flat-bottom plates, ensuring simultaneous germination, and media does not evaporate during incubation. Ideally, chemicals would be tested at a concentration that allows all seeds to germinate and phenotypes to be assessed accurately (Figure 3A-F). The majority of the treatments produced seedlings with phenotypes that were visually indistinguishable from mock controls with no morphological phenotypes (Figure 3A), with the vast majority of aberrant phenotypes consisting of bleached and severely stunted roots (Figure 2).
Figure 2: The most common phenotypes observed and the proportion of each phenotype observed. A) A total of 3,271 small molecules were found to be bioactive at 100 µM after four days of incubation. The color indicates the severity of the phenotype (black = more severe, white = less severe). The most commonly observed phenotype pertained to root morphology, with more than 1,500 compounds inducing stunted roots of varying severity. Coloration was also commonly affected by the compounds in this library, with 1,148 seedlings recorded as entirely bleached or partially discolored. Just under 400 compounds produced distinctive root hair phenotypes – either stunted or both stunted and brightly colored. Finally, germination was affected by just under 200 compounds. In these cases, seeds either did not complete germination or did not even begin to germinate. B) Seedlings exhibiting abnormalities in root morphology, either being stunted or severely stunted, constituted almost half of all phenotypically aberrant seedlings. The next largest group were those that resulted in bleached or discolored seedlings. Phenotypes that pertained to root hair abnormalities also made up a sizeable portion while the inhibition of germination, either no germination or incomplete germination, only occurred in a small percentage of all bioactive compounds. Finally, there were a number of phenotypes that occurred at such a low frequency, they were grouped into the category 'other', an example of which was the production of green mucilage, seen in Figure 3. Please click here to view a larger version of this figure.
Figure 3: Images of representative phenotypes observed during the forward chemical genetics screen. No visible morphological abnormalities (A), brown root hairs (B), stunted root (C), severely stunted root (D), bleached (E), green mucilage (F), incomplete germination (G), and no germination (H). Please click here to view a larger version of this figure.
Figure 4: Overview of the Stacker 10 and deck set up before initiation of protocol. The Stacker Carousel is comprised of four Stacker 10's (Hotels A – D) which each accommodate ten Rooms. The Deck holds a variety of ALPs: the Tip Loader, the Stacker Shuttle, the Multichannel Tip Wash, and 13 static ALPs. Please click here to view a larger version of this figure.
Figure 5: Stacker 10 and Deck set up required for Creating a Dilution Library. The four Stacker 10's are loaded with boxes of AP96 P20 Pipette Tips in Rooms 1 and 6 of Hotels A – D and stacks of four 96-Well V-Bottom Plates in Rooms 2 – 5 and rooms 7 – 9 of Hotels A – D. The deck layout consists of two 300 mL Reservoirs on static ALPs P3 and P7. Please click here to view a larger version of this figure.
Figure 6: Stacker 10 and Deck set up for Adding Media-Seed Mixture to Screening Plates. The Stacker 10 is loaded with four 96-Well Flat-Bottom Plates in Rooms 1 and 2 of Hotel A. The deck layout consists of two 300 mL Reservoirs on Static ALPs P3 and P7 and a box of AP96 P250 Pipette Tips on TL1. Please click here to view a larger version of this figure.
Figure 7: Stacker 10 and Deck set up for Adding Small Molecules to Screening Plates. The Stacker 10 is loaded with a box of AP96 P250 Pipette Tips in Room 1, a stack of two 96-Well V-Bottom Dilution Plates in Rooms 2, 4, 6, and 8, and a stack of two 96-Well Flat-Bottom Plates filled with the media-seed mixture in Rooms 3, 5, 7, and 9. The deck layout consists of two 300 mL reservoirs on Static ALPs P3 and P7. Please click here to view a larger version of this figure.
This protocol is designed to aid researchers in accomplishing a forward chemical genetics screen on Arabidopsis. We provide representative results from a screen of 50,000 compounds (Figure 2 and Figure 3), one of the largest forward chemical genetics screens performed on Arabidopsis to date9,13,23. The use of a liquid handling robot enabled more efficient dilution library and screening library generation, improving the speed and efficiency of identification of novel compounds. Increasing the capacity to screen in a high throughput nature was accompanied by decreasing labor on behalf of the researcher. This technique was designed to be used with 96-well plates, which can accommodate small seeds or plants visible under a dissecting microscope. Utilizing plates with larger wells to fit larger seeds would require modifications to the throughput and design.
Additional limitations of this technique include the difficulty of using this piece of equipment with aseptic techniques; however, we did not encounter high percentages of contamination due to ½ MS media lacking sucrose. One could circumvent any contamination issue by placing the robot in a sterile room, allowing for sterile conditions and cell culture, or by using a liquid handling robot with a sterile chamber24,25. Another limitation is tip size and seed aspiration. A small pipette tip such as the AP96 P20 would clog with seeds; therefore, a larger pipette tip must be used for seed dispensing and solution mixing.
Critical steps within this protocol include the careful labeling of all plates in the dilution library and screening library, ensuring they are in the proper orientation along with correct order when feeding the robot. Clear labeling and systematic processing is straightforward and can overcome this issue. Another critical step is ensuring that the right equipment is in the correct place before starting the experiment, both within the Stacker 10's and on the deck. If the equipment is not properly placed on the deck, the 96-Channel 200 µL Head could crash, damaging the instrument and requiring maintenance. Another critical step is ensuring that the correct amount of liquid is placed within the 300 mL reservoirs and that this amount is entered correctly into the software. If the numbers don't match, the tips will not reach the liquid and aspiration will not occur.
It is also necessary to take steps to ensure that results obtained are accurate. One error that we noticed while developing the protocol was linked to tip life. After successive loading and unloading, the tips lose their ability to aspirate and dispense accurately. It is therefore imperative that each set of 96 tips is used a maximum of four times. It is also important to change the wash water regularly to avoid the potential for chemicals to be inadvertently added to screening plates. Finally, some chemicals have a tendency to precipitate out of solution26. To ensure each chemical is added at the correct concentration, mixing steps are incorporated into the dilution and screening protocol. Failure to mix could result in low quantities of chemicals being added from library to library, challenging interpretation of potential chemical induced phenotypes.
Using the correct plates for each part of the protocol is also very important. V-Bottom plates are designed to ensure that small volumes of liquid can be aspirated and are recommended for use in the creation of the Dilution Library. However, these plates are not suitable for the screening portion of the protocol, since their reflection of light leads to poor visualization of phenotypes. In order to observe the phenotypes of 3 – 4 day old seedlings, the screen must be carried out in flat-bottom plates.
Once the screening plates have been created, visualization is required. 96-well flat-bottom plates allow easy visualization under dissecting microscopes. It is imperative that the plates are stored in desiccation-proof containers to reduce the evaporation of the media. An alternative to microscopic visualization is using a high-resolution scanner. Images produced at high resolution reveal the majority of the phenotypes observed in this screen and provide an archive of the results that can be revisited in the future. Once visualization is complete, and the library of your choice screened, this method could then be performed on different organisms or with a different chemical library. Modifications to the equipment could allow for sterile culture, allowing ventures into the realms of stem cells, fungus, insects, and small plants2,18,25,27.
The authors have nothing to disclose.
We thank Jozsef Stork, Mitchel Richmond, Jarrad Gollihue, and Andrea Sanchez for constructive and critical discussion. Dr. Sharyn Perry for the phenotypic photographs. This material is based upon work supported by the National Science Foundation under Cooperative Agreement No. 1355438.
Keyboard | Local Provider | N/A | Used for protocol design and operating the Biomek FX |
Mouse | Local Provider | N/A | Used for protocol design and operating the Biomek FX |
Computer Screen | Local Provider | N/A | Used for protocol design and operating the Biomek FX |
Computer | Local Provider | N/A | Used for protocol design and operating the Biomek FX |
DIVERSet Diverse Screening Library | ChemBridge | N/A | Chemical library |
Biomek Software | Beckman Coulter | N/A | Runs and designs the Biomek FX |
Device Controller | Beckman Coulter | 719366 | Operates the water pump/tip washing station |
Stacker Carousel Pendent | Beckman Coulter | 148240 | Manual operation of Biomek Stacker Carousel |
Biomek Stacker Carousel | Beckman Coulter | 148520 | Rotary unit that houses all FX Stacker 10's |
FX Stacker 10 | Beckman Coulter | 148522 | Elevator unit that houses components for screen |
FX Stacker 10 | Beckman Coulter | 148522 | Elevator unit that houses components for screen |
FX Stacker 10 | Beckman Coulter | 148522 | Elevator unit that houses components for screen |
FX Stacker 10 | Beckman Coulter | 148522 | Elevator unit that houses components for screen |
Biomek FX | Beckman Coulter | https://www.beckman.com/liquid-handlers | Robot that performs the desired operations |
Accuframe | Artisan Technology Group | 76853-4 | Frames arm to place components corretly |
Framing Fixture | Beckman Coulter | 719415 | Centers arm in the Accuframe |
Multichannel Tip Wash ALP | Beckman Coulter | 719662 | Washes the tips after the ethanol bath |
Tip Loader ALP | Beckman Coulter | 719356 | Pneumatically loads tips onto the arm |
Air Compressor | Local Provider | N/A | Provides air for pneumatic tip loading |
MasterFlex Console Drive | Cole-Parmer | 77200-65 | Pump used to circulate water through the Multichannel Tip Washer |
Air Hose | Local Provider | N/A | Provides air from air compressor to Tip Loader |
Water Hose | Local Provider | N/A | Provides water from 5 Gallon Reserviour to Tip Washer |
Static ALP's | Beckman Coulter | Comes with Biomek FX | Supports equipment for the Screen |
5 Gallon Reserviour | Local Provider | N/A | Recirculates the dirty water from cleaning the tips |
Grippers | Beckman Coulter | Comes with Biomek FX | Grabs and moves the equipment to the correct places |
96-Channel 200 µL Head | Beckman Coulter | Comes with Biomek FX | Holds the 96 tips used within the screen |
AP96 P200 Pipette Tips | Beckman Coulter | 717251 | Used to make the screening library |
96 Well Flat Bottom Plate | Costar | 9018 | Aids in visulization of screen |
96 Well V-Bottom Plate | Costar | 3897 | Aids in storing of dilution library |
AlumaSeal 96 Sealing Film | MedSci | F-96-100 | Seals for storage both the chemicle library and dilution library |
Plastic ziplock sandwich bags | Local Provider | N/A | Used to ensure a humid environment for screen |
AP96 P20 Pipette Tips | Beckman Coulter | 717254 | Used in the dilution library creation |
Growth Chamber | Percival | AR36L3 | Germinates seeds for phenotypic visulization |
Spatula | Local Provider | N/A | Holds seeds to add into wells where liquid seeding failed seed adequatly |
Toothpick | Local Provider | N/A | Pushes seeds from spatula to wells |
Murashige and Skoog Basal Salt Mixture | PhytoTechnology Laboratories | M524 | Add to MS media mixture |
MES Free Acid Monohydrate | Fisher Scientific | ICN19483580 | Added to MS media to decrease pH |
Agar Powder | Alfa Aesar | 9002-18-0 | Increases thickness of media to support seed suspension |
5M KOH | Sigma-Aldrich | 484016 | Increases pH to adequate levels |
1L Media Storage Bottle | Corning | 1395-1L | Holds enough media for a screen |
Polypropylene Centrifuge Tubes | Corning | 431470 | Sterilizes seeds prior to vernilization |
pH Probe | Davis Instruments | YX-58825-26 | Used for making media |
ALPs (Automated Labware Positioners) Users Manual | Beckman Coulter | PN 987836 | Aids in setting up the accompaning equipment for the Biomek FX |
Biomek 2000 Stacker Carousel Users Guide | Beckman Coulter | 609862-AA | Aids in setting up the Stacker Carousel |
Biomek FX and FXP Laboratory Automation Workstations Users Manual | Beckman Coulter | PN 987834 | Used to frame the Multichannel Pod |
Biomek FXP Laboratory Automation Workstation Customer Startup Guide | Beckman Coulter | PN B32335AB | Used to aid in setting up the Biomek FX |
Biomek Software User's Manual | Beckman Coulter | PN 987835 | Used to set up and understand the Software |