Applications for Open Source Microplate-Compatible Illumination Panels

Summary

Microplate Assistive Pipetting Light Emitter (M.A.P.L.E.) is a computer-driven device that systematically illuminates microtiter wells to provide guidance for the manual preparation of microplates.  M.A.P.L.E. improves the accuracy of microplate preparation while automating data recordkeeping.  In addition, it can assist with examining microplate quality or aid in the detection of errors.   

Abstract

Microplates are commonly used in the modern laboratory environment for a wide variety of tasks both in small-scale laboratory benchtop operations as well as large-scale high-throughput screening (HTS) campaigns. Though laboratory automation has greatly increased the utility of microplates there remain instances where automation-based instrumentation is not feasible, cost-effective or compatible with microplate formatting needs. In these cases, microplates must be manually prepared. Problematic to manual microplate manipulations is that a number of difficulties can arise pertaining to the accurate tracking of sample operations, data record keeping and quality control (QC) inspection for well artifacts or formatting errors. As microplate well densities increase (i.e., 96-well, 384-well, 1536-well) the potential for introducing errors also drastically increases.  Moreover, for small bench-top laboratory operations there exists a need to improve the ease and accuracy of sample handling in a cost-effective fashion. Herein, we describe a system that acts as a semi-automated pipetting guide referred to as the Microplate Assistive Pipetting Light Emitter (M.A.P.L.E.).  M.A.P.L.E. has multiple uses for supporting compound hit-picking and microplate preparation for assay development in high-throughput screening or laboratory benchtop operations, as well as QC/quality assurance (QA) diagnostic evaluation of microplate quality or visualizing well formatting errors.

Introduction

As recently published¹, the Lead Identification laboratory at Scripps Research² has developed and released an open-source illumination panel for microplate preparation referred to as the Microplate Assistive Pipetting Light Emitter (M.A.P.L.E.). Manual preparation of microplates, whether they are made for compound management or bio-assay needs, can be prone to human errors that drastically increase as well density of the microplate increases. In addition, proper recordkeeping and data-logging of microplate content/format is also prone to manual entry errors. In high throughput screening (HTS) automation facilities these issues are mitigated through the use of computer-driven robotic workstations that are integrated with automated database recordkeeping; minimizing manual manipulations and reducing the potential of formatting and data recording errors. However, there remain many instances where automation-based instrumentation is simply not feasible or compatible with microplate formatting needs, requiring manual intervention. Moreover, there is also a need to support small-scale laboratory operations that require compact and cost-effective semi-automated devices to improve their throughput, accuracy and automate data-recordkeeping of microplate preparation.    

While other microplate illumination systems exist, they are proprietary commercial solutions^3,4,5,6,7 limited to select microplate formats and their proprietary closed-source nature prevents user-driven modifications that would allow the adaptation these devices for specialized operations.  M.A.P.L.E. was designed to be an inexpensive open-source device, with source code and all design files available for free online⁸. Users with knowledge of surface mount soldering techniques can assemble their own M.A.P.L.E. devices with the code and design files available on GitHub, or they can modify the provided printed circuit boards (PCBs) designs, 3D print enclosure computer-aided design (CAD) models and code to meet their specific needs. A full list of parts needed to fabricate the light guide PCBs can be found in Supplementary Tables 1 and 2 and further details regarding the design and implementation of the light panels can be found in recently published documentation¹. Users who wish to purchase pre-assembled light guide PCBs based off the open-source files can find them listed online⁹.

M.A.P.L.E. provides the user with an easily controllable illumination panel which has a microplate-based footprint and LED-to-LED spacing matched to Society for Biomolecular Screening (SBS) specifications for microplates¹⁰. M.A.P.L.E. was developed to support 96- and 384-well density microplates and allow users to illuminate wells in any desired configuration, color and intensity. These light panels can be used to illuminate microplates for pipetting operations¹¹, to simulate laboratory formatting operations or instruments such as a microplate reader^12,13 for educational and demonstration purposes. The open source nature of the project allows users to easily modify the panels, firmware or graphical user interface (GUI) software to support any new desired functionality. Guidance and data recordkeeping are computer-driven and can be integrated with spreadsheets or ported to a database system. Because M.A.P.L.E. is designed to work with plaintext comma delimited files, any spreadsheet or database software that is able to import or export CSV formatted files can be easily extended to work with M.A.P.L.E. Further, the project enclosure which has been designed for this system tilts the microplate towards the user during pipetting operations, increasing ergonomics by providing a more natural posture for the user while at the lab bench. Specific operational features to the M.A.P.L.E. system include: (i) Facilitating compound management efforts in preparing customized plates by illuminating single source well and destination well across microplates for manual pipetting guidance; assisted through a computer script that can be saved as an electronic record post completion. (ii) M.A.P.L.E. can illuminate any number of wells across microplate rows or columns; which is ideally suited for rapid serial dilution guidance or placement of select replicate controls. (iii) M.A.P.L.E. can be used in a demonstration mode to facilitate laboratory training needs or highlight formatting requirements with respect to sample and control placements or dedicated well usage (e.g., edge-effect barrier gap). (iv) M.A.P.L.E. can backlight transparent/translucent wells to allow visualization of artifacts such as precipitation/crystallization, bubbles, well heterogeneity, empty wells; which also allows end-user to easily photograph plate images for documentation needs

Protocol

1. Semiautomated "plate to plate" sample transfer preparation

1. Generate a CSV file as shown in Figure 1 containing source and destination plates using a spreadsheet editing application. The CSV file that is generated must have the following header columns in the order listed: Source_barcode; Destination_barcode; Source_well; Dest_well; Transfer_volume.   
2. Under the header columns, make sure to include one row in the CSV file for each desired pipetting operation (i.e., sample transfer) with the following information:
   1. Source_barcode: Alphanumeric barcode of the source microplate, e.g., S1007372; leave blank if no barcode is associated.
   2. Destination_barcode: Alphanumeric barcode of the destination microplate, e.g., D0573282; leave blank if no barcode is associated.
   3. Source_well: Alphanumeric row and column identifier for well to be pipetted out of from source plate, e.g., H10 for row H (8^th row), column 10 (ANSI/SLAS standard well designations, e.g., A1, C10…).
   4. Dest_well: Alphanumeric row and column identifier for well to be pipetted out of from destination plate, e.g., A3 for row A (1^st row), column 3 (ANSI/SLAS standard well designations, e.g., A1, C10…).
   5. Transfer_volume: Volume to be transferred from source_well in source_barcode to dest_well in destination_barcode (numeric and unitless: typically in µL).
3. Open the Microplate Assistive Pipetting Light Emitter Plate to plate GUI application, shown in Figure 2, by opening the Light Guide program (Maple-LightGuide.exe).
4. Click the Select cherrypick file button in the upper left corner of the GUI.
5. Use the file browser window, shown in Figure 3, to navigate to the CSV file generated in steps 1.1 and 1.2 above and click the Open button. The application will parse the first row of the CSV file and illuminate the corresponding wells in the source and destination plates.
6. Use the Previous well and Next well buttons, in the upper right corner of the GUI, shown in Figure 4, to traverse the CSV file as desired. The GUI will highlight in grey any rows which have been previously illuminated and highlight in brown the currently active row.
7. Perform pipetting operations as needed to transfer samples between source well of source plate to destination well of destination plate. An example of a M.A.P.L.E. unassisted hand pipetting operation can be seen in Figure 5, with a comparison of the current user pipetting view seen in Figure 4 and Figure 6. In addition to the user GUI, plate barcodes can be verified via the LCD displays attached to the illumination panels.
8. Continue until the end of the CSV file is reached via the Next well button. To load a new CSV file, the Select cherrypick file may be clicked at any time. To exit the program the red X at the top right corner of the GUI can be clicked.

2. Multi-well illuminations for parallel transfers and serial dilutions

1. Open the Microplate Assistive Pipetting Light Emitter 'Serial dilution' application by opening the serial dilution program (Maple-SerialDilution.exe).
2. Use the GUI, shown in Figure 7 and Figure 8, to specify the desired titration mode (column or row), plate density and start row(s) or column(s). The GUI also allows users to specify a column or row mask to control which LEDs in a given row or column are illuminated. This allows a subset of LEDs in a row or column to be illuminated instead of illuminating the entire row or column.
3. Use the Next and Previous buttons to step through the rows or columns in sequence from the initial start row or column to the last row or column in the plate. Each time the Next or Previous button is clicked, the light panel will illuminate the corresponding LEDs of the microplate.
4. Continue until the end of the titration sequence is reached. To exit the program, click the red X at the top right corner of the GUI.

3. Laboratory training: assay development and screening format techniques

1. Place a 96- or 384-well microplate into the portable light guide. The portable light guide contains a battery and all electronics necessary to be used independently of a computer. This allows the portable lightguide to be used in a handheld mode which can be controlled with built-in pushbuttons to toggle between demonstration modes.
2. Use the power toggle switch on the portable light guide enclosure to power the system on.
3. Determine the mode for the portable light guide to be in. By default, the portable light guide will load into the default HTS demo mode which provides users with a visual representation of a typical assay plate as seen in Figure 9. In this mode, one can use the right pushbutton switch at the top of the portable light guide to toggle through the following sample illumination patterns.
   1. All wells illuminated with red color to simulate the reagent dispense of an assay, e.g., (suspended cells in media).
   2. All wells illuminated with a yellow color to simulate dye reagent addition.
   3. First column and last column of wells illuminated green, remaining middle 'sample field' columns illuminated blue to indicate plate being read on microplate reader. Random wells in the sample field will also have green color of varying intensity to represent hits.
4. To toggle the light guide between HTS demo mode and Titration demo mode, push the left pushbutton switch. Doing so will switch the portable light guide to the Titration demo mode which provides a visual guide to users to understand how titrations can be performed in compound plates. When the light guide enters the Titration demo mode, the following will occur.
   1. All wells in column 3 and 13 are illuminated with yellow color.
   2. Subsequent presses of the rightmost pushbutton switch illuminates columns in sequence, e.g., (4 and 14, 5 and 15, etc.).
   3. When the pushbutton is pressed after columns 12 and 22 are reached, wells in columns 4-12 and 13-22 are illuminated in decreasing intensity of yellow to represent the titration.
5. To modify the default behavior of the light guide, plug the portable light guide to a computer via a USB cable and follow the detailed instructions for updating the default firmware via the Arduino IDE which can be found on the project GitHub page⁸. By updating the firmware, you can modify these modes to display other sequences or sets of LEDs.

4. Illumination of artifacts in microplates

1. Place a 96- or 384-well microplate into the portable light guide.
2. Switch the light guide to Illumination mode by pressing the leftmost pushbutton switch two times.
   NOTE: An example of the practical use of this mode can be seen in Figure 10 and Figure 11, where compounds have precipitated out of solution and can be observed on the bottom of the microplates. Without backlit illumination, most of the precipitate is invisible to the naked eye, but M.A.P.L.E. backlighting reveals precipitate for user inspection and photographic documentation.
3. Use the right pushbutton to toggle between a set of predefined colors as needed for the application. The light panel will turn all LEDs on to the following colors in sequence with each push of the right button: red, blue, green, orange, white, violet, yellow and indigo.
4. As an optional step, use a camera or smartphone to photograph the illuminated plate for recordkeeping or documenting the work.

Representative Results

The M.A.P.L.E. platform is capable of illuminating wells in 96- and 384-well microplates in a variety of user configurable ways, allowing straightforward and independent control of color and light intensity in each well. By helping to reduce opportunities for error in manual pipetting operations, M.A.P.L.E. helps users prepare microplates with increased confidence that each well contains the desired contents. The transfer of samples between plates and the preparation of serial dilution plates, such as the examples seen in Figure 12 and Figure 13, can be accomplished without concern that the user will be distracted during their work and lose track of what pipetting operations remain. When pipetting work is completed, the M.A.P.L.E. platform can then be used to help illuminate the microplate to help the user identify potential artifacts such as precipitate, empty wells, partially filled wells or air bubbles. By detecting these artifacts at the time the microplate is created, users can take measures to ameliorate samples before providing them to downstream laboratory processes.

To demonstrate the functionality of M.A.P.L.E., a head-to-head test was performed to measure the speed and accuracy of pipetting operations using a printed worklist versus the steps described in protocol section 1. For this test, seven users in the Lead Identification laboratory  performed the test using the same worklist was used for both offline versus M.A.P.L.E-guided. These seven users represented a variety of pipetting experience, ranging from many years in the laboratory to novice pipetting users. The only difference being the user hand-annotating a printed sheet for manual mode and using the computer GUI in the M.A.P.L.E.-guided mode. This worklist consisted of 49 pipetting operations from two 384-well source microplates containing a random assortment of colored dyes in DMSO (Figure 14A,B) that spell 'jove' in a single 384-well destination microplate (Figure 14C). In this configuration, the layout of the wells in the destination plate confirm the user has pipetted into the correct wells of the destination plate and the color pattern of the wells in the destination plate can be used to identify errors where the user did not pipette from the correct well of the source plates as seen in Figure 14D which shows an example of pipetting errors in wells K2, F22, F23 which occurred while a user was following a printed worklist. Table 1 contains the results of this head-to-head test which shows an average time saving of 50% when users performed this test using M.A.P.L.E. versus an offline printed worklist. Not only was the process speed increased when M.A.P.L.E. was used, but the error rate of plates created using M.A.P.L.E. was 0% for all users, while a 6% error rate was observed for one novice user when using a worklist for the sample preparation task (Figure 14D).

Figure 1: Example CSV file used for sample preparation application. Example CSV file used for sample preparation application including the five columns required to annotate the transfer volume, microplate barcodes and well locations for both source and destination plates. Please click here to view a larger version of this figure.

Figure 2: Sample preparation application GUI. The sample preparation application GUI is displayed to the user upon start of the application. From this interface the user can select a CSV file for use in the sample preparation process. Please click here to view a larger version of this figure.

Figure 3: File open dialog box. The file open dialog box allows the user to navigate to the CSV files of interest to be used in the sample preparation process. Please click here to view a larger version of this figure.

Figure 4: The GUI interface that is visible to the user after a CSV file has been selected and loaded into the application. The contents of the CSV file are displayed in a spreadsheet-style format and the active row is highlighted. Users can step forward or backward through the file by using the 'Previous well' or 'Next well' buttons which update the active row and send the appropriate illumination commands to M.A.P.L.E. Please click here to view a larger version of this figure.

Figure 5: Example of typical manual sample preparation process prior to M.A.P.L.E. showing user referencing printed list of plate barcodes and well locations to be pipetted. Please click here to view a larger version of this figure.

Figure 6: Current manual sample preparation process with M.A.P.L.E. illuminating wells of interest and displaying barcodes of microplates needed for current pipetting operation. Illuminated wells and barcode metadata are automatically updated based on user input from GUI seen in Figure 4. Please click here to view a larger version of this figure.

Figure 7: GUI interface for controlling M.A.P.L.E. in titration mode, allowing user control of illumination by specifying columns of interest. In addition to titration mode (row or column), users can specify plate density and step forward or backward through the columns by clicking the 'Next column' or 'Previous column' buttons. Please click here to view a larger version of this figure.

Figure 8: GUI interface for controlling M.A.P.L.E. in titration mode, allowing user control of  illuminating by specifying rows of interest. In addition to titration mode (row or column), users can specify plate density and step forward or backward through the rows by clicking the 'Next row' or 'Previous row' buttons. Please click here to view a larger version of this figure.

Figure 9: Wells illuminated with green and blue lights to represent wells fluorescing in microplate reader for HTS demo mode. Please click here to view a larger version of this figure.

Figure 10: Example of 96-well vendor-provided microplate containing newly purchased compounds with solubility issues back illuminated with a 96-well M.A.P.L.E. light panel with blue light. Illuminating the bottom of the microplate makes it much easier to identify compounds which have precipitated out of solution and need remediation prior to further liquid handling. Please click here to view a larger version of this figure.

Figure 11: Microplate examples. (A) Example of 384-well microplate containing compounds without any back illumination. (B) 384-well microplate backlight with green light, revealing many wells containing precipitate. (C) Closeup of 384-well microplate with green backlighting. Please click here to view a larger version of this figure.

Figure 12: Example of destination 384-well microplate containing 320 individual samples which have been transferred from many different source microplates. This example represents a typical sample preparation known as a hitpicked or cherrypicked microplate which is seen at the confirmation screen stage of an assay. Please click here to view a larger version of this figure.

Figure 13: Example of typical 384-well microplate with 10-point serial dilutions starting in columns 3 & 13 with a mask filter for rows 1–16 (all rows included). Please click here to view a larger version of this figure.

Figure 14: Sample preparation pipetting test. (A, B) 384-well microplates containing various colored dyes solvated in dimethyl sulfoxide (DMSO) to be used for sample preparation pipetting test. (C) 384 well microplate resulting from sample preparation pipetting test containing correct colors of samples in correct locations. (D) 384-well microplate resulting from sample preparation pipetting test containing errors (K2, F22, F23) when the user followed the manual worklist method. Please click here to view a larger version of this figure.

Figure 15: LED output spectra as measured by a spectrometer. Please click here to view a larger version of this figure.

User	Worklist Time	Worklist error rate	M.A.P.L.E. Time	M.A.P.L.E. error rate	% speed increase
Experienced user #1	15 min 8 s	0%	9 min 39 s	0%	36%
Experienced user #2	17 min 54 s	0%	7 min 59 s	0%	55%
Experienced user #3	18 min 34 s	0%	10 min 25 s	0%	44%
Experienced user #4	20 min 50 s	0%	10 min 13 s	0%	51%
Novice user #1	26 min 52 s	0%	11 min 03 s	0%	59%
Novice user #2	35 min 49 s	6%	15 min 29 s	0%	57%
Novice user #3	22 min 44 s	0%	11 min 30 s	0%	49%

Table 1: Results of manual sample preparation versus M.A.P.L.E.-guided sample preparation, including time spent by each user to process full worklist in each mode.

96w Microplate M.A.P.L.E. parts list	Vendor	Vendor part #	Cost per item	Quantity needed for assembly	Parts cost per prototype
96 well RGB prototype PCB	OSH Park		$28.38	1	$28.38
RGB 3535 SK6812 RGB SMD LED	Aliexpress (BTF-Lighting)	SK6812mini 3535	$0.10	96	$9.63
0.1 µF capacitor SMD (0805)	Digi-Key	478-3351-1-ND	$0.16	96	$15.36
Adafruit Metro Mini 328 – 5V	Digi-Key	1528-1374-ND	$12.50	1	$12.50
Total					$65.87

Supplementary Table 1: List of components needed to fabricate a 96-well RGB light guide.

384w Microplate M.A.P.L.E. parts list	Vendor	Vendor part #	Cost per item	Quantity needed for assembly	Parts cost per prototype
384 well RGB prototype PCB	OSH Park		$28.38	1	$28.38
RGB 2427 SK6805 RGB SMD LED	MOKUNGIT	SK6805 2427	$0.09	384	$34.20
0.1uF capacitor SMD (0603)	Digi-Key	478-10679-6-ND	$0.05	384	$18.05
Adafruit Metro Mini 328 – 5V	Digi-Key	1528-1374-ND	$12.50	1	$12.50
Total					$93.13

Supplementary Table 2: List of components needed to fabricate a 384-well RGB light guide.

Discussion

By releasing M.A.P.L.E. as an open-source platform, we have introduced a laboratory tool that provides utility, but can also be easily extended to meet the evolving needs of the end user. Benchtop microplate sample preparation is a common task that is performed in a wide variety of laboratory environments and this task can be demonstrably improved with a technology such as M.A.P.L.E.

The M.A.P.L.E. platform has been specifically engineered with adaptability to future applications in mind. Each component (electronic illumination panel, firmware, GUI, enclosure) can be extracted for use individually, used as part of the larger system or any intermediate combination thereof. For example, the 3D printed project enclosure can be used without the illumination panel simply to improve the ergonomics of benchtop pipetting. The illumination panel has a straightforward three-wire interface that can be attached to any system capable of generating a +5 V control signal, +5 V source and ground (GND). The behavior and utility of the software GUIs can be modified using Python code, the illumination panel circuitry can be modified in KiCad and the microcontroller firmware used to control the panels can be edited in the Arduino IDE. With this flexibility, the M.A.P.L.E. platform is extensible to meet future needs.

Of similar devices previously developed for use for illuminating microplates^3,4,5,6,7, M.A.P.L.E. is the only device that is fully open source. This provides a great amount of flexibility to the end user to extend the existing functionality to meet their specific need. This extended functionality may take the form of additional user input control devices (foot pedals, buttons, etc.) or other metadata display devices. The open source nature of the device also helps to prevent device obsolescence due to reliance on any one specific vendor for device production, development or support. Users can choose to keep M.A.P.L.E. as a compact single microplate formfactor device, or to extend it to illuminate multiple microplates simultaneously, both applications of which have been demonstrated in this manuscript. Lastly, the components necessary to assemble a M.A.P.L.E. system have a cost which is lower than any previously available commercial solutions.

Potential limitations of the system include illumination and visualization interference caused by dark colored compounds. The sample preparation functionality also currently requires that M.A.P.L.E. be tethered to a computer via USB. We also suggest that laboratory processes utilizing light sensitive compounds or reagents are tested prior to extended use with M.A.P.L.E. Light sensitive compound transfers are a problem in any lab situation, but M.A.P.L.E. can select for wavelengths that are less prone such as red light. M.A.P.L.E. also allows users to adjust LED color and intensity via firmware updates to the microcontroller to provide the specific illumination desired. The spectral output of the LEDs has been provided as per Figure 15 so that the user can avoid wavelengths at which the compound is known to absorb light.

The components of M.A.P.L.E. could also be reused to investigate alternative uses such as photochemistry, phase separation in compound libraries or modified with different wavelength LEDs (e.g., UV) to extend its functionality for other applications. Likewise, colorimetry or absorbance spectroscopy can be performed inexpensively with M.A.P.L.E.'s select emissions as shown in Figure 15 and camera or smartphone app to capture RGB output values. In conclusion, the M.A.P.L.E. has been designed for immediate use to support sample microplate preparation, but as an open-source platform it can be adapted for use in many other applications.

Materials

Name	Company	Catalog Number	Comments
96 or 384 well microplate			https://en.wikipedia.org/wiki/Microplate
Microplate Assistive Pipetting Light Emitter	Open source		https://github.com/pierrebaillargeon/Microplate-Assistive-Pipetting-Light-Emitter
Pipettor			https://www.jove.com/science-education/5033/an-introduction-to-the-micropipettor
Spectrometer	Ocean Optics	USB-650 Red Tide