FAQ - Building a High Throughput Screening Facility in an Academic Setting

• Introduction
• Planning and Design of the Facility Workspace
• Staff
• Instrumentation
  • Liquid Handling
  • Compound Transfer
  • Robotic Integration
  • Assay Detection
    • Uniform Well Readout Assays--Plate Readers
    • Cell-based Imaging Assays--Automated Microscopes
  • Data Capture
• Data Analysis and Informatics
  • Data Collection and Analysis
  • Software Tools
• Compound Purchase
• Compound Storage and Handling
• Consumables
• Useful Resources

Introduction

The Institute of Chemistry and Cell Biology/Initiative for Chemical Genetics (ICCB/ICG) at Harvard Medical School established a Screening Facility in 1998 to facilitate the pursuit of Chemical Genetics as an academic discipline.  At the time, techniques for high throughput screening of small molecule libraries in biological assays were being developed in the biotechnology and pharmaceutical industries as means to speed the identification of lead compounds for drug discovery.

The ICCB/ICG Screening Facility was one of the first high throughput screening facilities to be opened in an academic setting.  Although the investigators who use our facility share some of the same goals as their counterparts in industry, their needs also differ in key ways.  For example, most industry screening programs focus their efforts on a relatively small number of disease-relevant target pathways and proteins. In contrast, in the university, investigators use chemical genetic screens to find small molecule research tools that perturb a wide variety of biological pathways in a diversity of organisms.  Much of it is very basic research, not necessarily immediately relevant to human disease.  Thus, our Screening Facility must have the flexibility to accommodate many different types of assays.

At the ICCB/ICG, most screens are carried out in the 384-well format and our screening instruments are set up in modular work stations.  Most have the capacity for integration with each other, in order to automate sequential steps of assay protocols when desirable.  Individual researchers carry out the bulk of the work for their own screening projects.  The Screening Facility staff assists by providing access to the ICCB/ICG compound collection, maintaining and operating the screening robots, and training screeners for independent operation of some machines.

The ICCB/ICG Screening Facility runs on average 12-15 screening sessions per week. In a typical screening session, 14,080 compound wells (20 plates in duplicate) are screened. Current comfortable capacity for the facility is ~350,000 compound wells (500 plates in duplicate) screened per week. Approximately 36-48 new screens enter our facility each year. As of November 2003, there were 140 screens ongoing at the ICCB/ICG. This includes screens in the piloting, HTS, and follow-up stages. A screen is considered complete once the results are published or when the investigator notifies us that it is no longer being pursued. A typical investigator-initiated screening project will screen 50,000-100,000 compound wells in duplicate.

We have written this document to answer questions frequently asked of the ICCB/ICG Screening Group by our colleagues in other departments and institutions who wish to set up their own high throughput screening facilities.

top of page

Planning and Design of the Facility Workspace

Space planning is the first concern in the design of the screening facility.  When it first opened, the ICCB Screening Facility was housed in an area of approximately 400 square feet that accommodated a small office for facility staff as well as three plate readers, several small liquid handling devices, and an automated pin-transfer robot.  The current ICCB/ICG facility is approximately 1000 square feet and accommodates all of the above as well as three additional plate readers, a large free-standing liquid handler, several robotic arms for integration of the screening instruments, one large freezer for compound storage, and a tissue culture area.  Additional equipment rooms house our automated screening microscopes and the other freezers in which our compounds are stored.

The design of facility workspace is wholly dependent upon the types of assays being performed and the equipment required to process the assays.  For example, mammalian cell-based assays require access to a tissue culture area with water-jacketed CO₂ incubators, whereas yeast or bacterial assays require only a standard 37ºC incubator.  All assays require an adequate amount of bench workspace for the individual researcher to prepare and carry out their screens.  In addition, consideration should be given to ensuring that adequate vacuum and gas (e.g. CO₂ and air) services are available for the facility as some instruments depend on them for their operation.  Computer network connections (data jacks) are essential for the assay detection and data capture phases of screens.  Large amounts of data are generated during high throughput screening and their final destination (e.g. data server and/or database) should be planned before screening begins.

Flexibility can be introduced into a facility from the outset by purchasing carts designed for laboratory equipment.  Although we have not implemented them at the ICCB, these carts can be used separately or connected together to provide modular, integrated workstations.  In lieu of this, however, standard lab benches and a reasonable amount of open floor space for a small number of freestanding machines is sufficient.

For tissue culture, a six-foot, laminar-flow tissue culture hood is recommended because it has room for an automated plate filler, which is required to dispense cells into assay plates.  The ICCB/ICG Screening Facility has three tissue culture incubators.  To accommodate a wide variety of assay conditions, one incubator is equipped with a circulating cooler to allow control of the temperature between 20ºC and 42ºC.  The clinical centrifuge used for tissue culture is equipped with microplate carriers and is used, as necessary, for compound stock plates as well as assay plates.  It is helpful to have a refrigerator in the Screening Facility, or access to a cold room nearby, for temporary storage of cell media and assay reagents.

With regard to facility infrastructure, the ICCB/ICG Screening Facility staff has found that house air and vacuum services can be unreliable and has thus purchased individual vacuum pumps and compressors.  High-end liquid handling instruments often require one or both of these services and it is worth the relatively minor cost to purchase these small items as needed.  Finally, it has been useful to maintain close working relationships with the local computing support group and the data management group to ensure a smooth transfer of data from the screening facility to the servers and databases.

top of page

Staff

A wide variety of skills are necessary to run a high throughput screening facility.  An aptitude for troubleshooting the computer and mechanical problems that invariably arise with complex instrumentation is essential.  At least one staff member should be proficient in the computer language Visual Basic, which is used to integrate the operation of individual screening robots with each other.  In addition to operating and maintaining laboratory equipment, staff members will likely also be asked to organize compound collections and assist investigators in performing screens.  Thus, some formal training in the biological sciences or chemistry is desirable.  For the most part, individuals with all of these qualifications have been employed in industry.  The ICCB/ICG has attracted experienced laboratory automation specialists from industry to the academic setting, but we have also successfully trained recent college graduates for Screening Facility staff positions.

top of page

Instrumentation

Prior to the purchase of any equipment, it is important to consider the particular requirements of the users.  For example, fairly inexpensive plate fillers are sufficient to perform all of the liquid handling steps of some screens, while more accurate, but very expensive, low-volume automated pipettors are required for other screens.  For assay detection, some instruments provide multi-mode assay detection and are therefore useful for multiple assay types, whereas others are specialized for individual assays. An additional consideration is whether radioactive assays will be performed in the screening facility.  Due to the diversity of user-requirements, the ICCB/ICG Screening Facility does not perform assays involving radioisotopes.

Instrument calibration and maintenance are important.  Liquid handling machines usually arrive from the factory with calibration certification, but the conditions under which this verification was performed do not necessarily correspond with the conditions that exist in the laboratory.  It is good practice, after delivery, to verify the accuracy and precision of a machine through a wide range of parameters and to continually monitor the findings at regular intervals (e.g. monthly or quarterly).  The initial in–facility calibration may involve leveling the machine itself, as bench tops and floors may not be level.  Because of the small size of dispense needles or pin arrays in some instruments, a surface that is only slightly off-level can produce significant inaccuracies in the volumes of liquid transferred.  Finally, a common option when purchasing a machine is an extended service contract.  These contracts can be very expensive, generally costing 10% of the purchase price per year.  The policy of the ICCB/ICG is to track maintenance costs for machines throughout the first year and to purchase a contract only if the yearly costs exceed the price of the contract.

The following sections discuss considerations to take into account when choosing instrumentation for high throughput screening.  Specific machines used at the ICCB/ICG are discussed.  Detailed specifications for these instruments are available from the manufacturers, and are also on our website (http://iccb.med.harvard.edu/screening/index.htm).

top of page

Liquid Handling

There are a wide variety of options available for high throughput liquid handling.  As noted above, subtle improvements in accuracy can often correspond to a large increase in price.  Also, machines capable of deep-well pipetting are often considerably more expensive than ones capable of only shallow-well pipetting.  Most high accuracy (“high-end”) liquid handling instruments combine accurate low-volume liquid handling with other functions such as library reformatting, cherry picking, or pin-transfer.  Significant training is required to program, run, and maintain high-end liquid handling instruments.  Thus, these instruments are generally operated exclusively by the Screening Facility staff.

The ICCB/ICG Screening Facility uses the Assay TekBench from TekCel for our high-end liquid handling needs.  In addition to being the facility’s primary choice for liquid handling requirements below 5ul, its 96-channel deep-well pipettor is capable of reformatting libraries from 96-deep well to 384-shallow well plates.  The Assay TekBench can pick up single pipet tips in succession to “cherry pick” compounds from multiple different compound source plates.  This task, one that until recently was tediously performed by hand in our facility, allows us to format a single plate to contain only the compounds that have scored as positives in a primary screen.  The automation of this task saves an enormous amount of time, freeing up the staff for other projects.

While the high-end liquid handlers are extremely useful, much smaller and less expensive plate fillers carry out the bulk of day-to-day liquid handling at the ICCB/ICG.  Plate fillers are machines that use a manifold (8- or 16-channel, for example) to rapidly dispense cells or reagents into assay plates with an acceptable level of accuracy.  As noted above, reliable performance from a liquid-handling machine is dependent upon regular maintenance and calibration.  Plate fillers tend to be straightforward in their operation and can be used independently by screeners after only a short training session led by Screening Facility staff.

The ICCB/ICG Screening Facility currently owns four Biotek Precision 2000 liquid handlers and four Biotek µFills (MicroFills).  These relatively inexpensive machines reliably rapid-dispense µL volumes of cells or reagents into 96-well or 384-well assay plates. The Precision 2000 accurately dispenses 20 µL to 300 µL volumes using an 8-channel manifold and it comes equipped with an 8-channel pipettor for smaller volumes and serial dilutions.  The Precision 2000 deck holds up to six plates at one time.  Its pipettor, while useful for some applications, is significantly slower than manifold dispensing. The MicroFill accurately dispenses from 5µL to 1500µL into 96-well and 384-well shallow or deep well assay plates, and uses a 16-channel manifold. The MicroFill dispense manifold and pump assembly is completely autoclavable, allowing for sterile dispensing if needed. While the MicroFill holds only one plate at a time, the accurate low volume transfer, sterile pathway, and increased speed are useful features.

Another key liquid handling function is plate washing.  Washing 384-well plates can be accomplished in a semi-automated fashion using an automated plate filler to add wash reagents and a hand held 24-channel adaptor (the Wand, available from V & P Scientific as Catalog # VP 186L) attached to a vacuum line for the aspiration step.  Fully automated plate washers are also available and are relatively inexpensive.  The ICCB/ICG Screening Facility has two of these machines, one from BioTek (configured with a 96-channel head) and one from Tecan (configured with a 384-channel head).  These time-saving devices can carry out a wash step (aspiration followed by buffer addition) on a 384-well assay plate in 30-60 seconds.  The downside of these machines is that the needles that perform these tasks become clogged easily, despite regular cleaning.  Therefore, plate washers tend to require more care and maintenance than other liquid handling machines.

top of page

Compound Transfer

Every small-molecule screen carried out at the ICCB/ICG Screening Facility requires transfer of compounds from library stock plates to assay plates.  While all other liquid handling is performed in the microliter range, the transfer of library compounds into assay plates (already containing cells, etc.) is performed in the nanoliter range and is accomplished using carefully-machined steel pin arrays.

Pin-transfer of library compounds conserves reagents, is cost effective, and is compatible with the types of assays carried out in our facility.  Compound libraries are stored in DMSO. Typically, it is desirable that the amount of DMSO transferred to an assay well is less than 5% of the final well volume (in most cases, <1% is preferred).  Since assay volumes usually range from 25-50 ul per well (in 384 well plates), approximately 100 nl of compound stock solution should be transferred to maintain the DMSO concentration in the desirable range.  Most commercial liquid handling systems cannot accurately pipette less than 1 ul of compound.  Although pin arrays require a substantial amount of time for calibration, we have found that stainless steel pins (in arrays purchased from V&P Scientific) can reliably transfer 100 nl of small molecules in DMSO into assay plates.  The pin array can be rapidly washed, dried with compressed air or blotted with inexpensive paper, and reused with undetectable levels of carryover between stock plates.  Another key advantage of the pin transfer system is cost effectiveness.  Steel pin arrays are much less expensive than a corresponding array of pipet heads and do not require the purchase and disposal of pipet tips. Therefore, the only consumable costs for the system are the methanol used to wash the pin arrays between transfers and blotting paper.  These pin arrays do wear over time, however, and must be sent back to the manufacturer approximately once a year for refurbishment.  The initial cost of a pin array can range from $5000 to $9000 and refurbishments are approximately $500.

The ICCB/ICG Screening Facility has three different machines that are capable of pin-transfer operations.  We custom built our first pin-transfer device, which is based on a Seiko cartesian robot.  The screening facility also uses a Cybio Cybi-Well with integrated stackers.  The Cybi-Well, originally purchased for its accurate low-volume pipetting capabilities, has been modified to perform pin-transfer using off-the shelf components sold by CyBio.  It is not possible, however, to use both the liquid handling capabilities and pin-transfer capabilities at the same time with this instrument.  Finally, the Assay TekBench mentioned above can also be used as a pin-transfer device.  Specifically, the integrated robotic arm on the Assay TekBench can pick up either a 96-channel pipettor or a 384-pin array.  Thus, it can perform pipetting steps on assay plates and then, without intervention from the user, transfer compound from library stock plates to those assay plates.

top of page

Robotic Integration

While screening instruments tend to be bought for their stand-alone capabilities, they can often be integrated with each other for automation of sequential steps in screening protocols.  This generally requires detailed discussion with knowledgeable salespeople or technicians so that appropriate software and hardware components can be purchased.  Integration of screening instruments can be accomplished either by contracting the vendor for the task or by employing an on-staff robotics programmer.  The ICCB/ICG Screening Facility staff creates custom robotic integrations using such tools as Visual Basic 6.0 and vendor-provided activeX controls.  This has been an effective strategy because the ongoing support for the effort remains in-house.  For example, the ICCB/ICG purchased a Twister2 robotic arm from Zymark and Screening Facility staff integrated it with three liquid handling devices and a plate reader.  The Twister2 shipped with scheduling software called CLARA (Computerized Logic for Automated Robotic Applications), but it was necessary for the staff to write software drivers to enable communication between CLARA and each instrument.  The result of this integration is an extremely flexible environment in which the Twister2 can serve microplates to any or all instruments or the instruments can be used in stand-alone mode.  If this integration had been purchased, it would have likely required future expenditures for maintenance and upgrades.

top of page

Assay Detection

The results of high throughput assays are typically detected using uniform well readout methods with a plate reader, or by imaging at the level of individual cells with an automated microscope.  

Uniform Well Readout Assays -- Plate Readers: Types of assays suitable for detection by uniform well readout methods include: luminescence or fluorescence intensity (FI), fluorescence polarization (FP), time-resolved fluorescence (TRF), fluorescence resonance energy transfer (FRET), and absorbance.  Examples of applications using these assay detection techniques are shown in Table I.  The ICCB/ICG Screening Facility has five multi-mode plate readers that read 96-well or 384-well assay plates: two Wallac Victor² readers (Perkin Elmer), two LJL Analyst systems (Molecular Devices), and a Biotek Synergy HT.  All five readers support multi-plate operations, either through stackers or robotic integration.  The Wallacs and Synergy HT are good basic plate readers, whereas the Analyst systems offer increased sensitivity but with a higher purchase cost.

Table I.  Assays using uniform well readout detection methods

Cell-based Imaging Assays--Automated Microscopes: Automated screening microscopes are used to monitor changes at the level of individual cells within an assay well.  These instruments perform iterative auto-focusing to acquire images of each well of a 384-well plate.  When performing cell-based imaging screens (also referred to as High Content Screening (HCS)), it is important to consider the level of throughput required.  More information is extracted than for uniform well readout methods and thus reading times are longer (e.g. 45-90 minutes/384-well plate by automated microscopy, versus 3 minutes by a plate reader).  Additional points to consider include whether multiple excitation/emission wavelengths will be used and whether incubation at temperatures higher than ambient will be required.

The ICCB/ICG Screening Facility has two automated microscopes for cell-based imaging; the AutoScope and the Discovery 1, both made by Universal Imaging Corp (see our website for lists of the lenses and filters we have fitted with each microscope, http://iccb.med.harvard.edu/screening/technology_screen_by_imag/index.htm ).  The Autoscope is a conventional microscope that was customized for high throughput use, whereas the Discovery 1 is an integrated system designed specifically for high throughput screening.  The Discovery 1 system is equipped with a Plate Crane that can hold stacks of plates, allowing the system to run unattended for continuous 24-hour screening.  While most imaging screens carried out at the ICCB/ICG are currently scored by eye, one image at a time, we are developing automated methods for analyzing images and quantitating the results of some imaging screens.  Other options for automated microscopy systems are available from Cellomics, Q3DM, Axon Instruments, and Amersham.

top of page

Data Capture

Capture and storage of raw data generated by high throughput assays is not a trivial point.  Plate readers generate data in the form of text files.  These are generally small in size (~ 3KB for a single plate) but fill up computer hard drives quickly.  In contrast, image files are large in size (~650 KB for a single image); thus an imaging screen of 20,000 compounds in which two separate wavelengths are imaged would generate ~ 51 GB of data!  At this point it becomes advisable to use a server for data storage.  Currently the ICCB/ICG Screening Facility is using an 8 TB server to facilitate storage of data from imaging screens.

top of page

Data Analysis and Informatics

Data Collection and Analysis

The methods used for analysis of data from high throughput screens are as important as the screening protocols.  There is no one correct method for data analysis and different possibilities should be evaluated for individual screens as the screens are being developed.  Some general considerations are highlighted below.

Most assays designed for high throughput screening have a high amount of inherent variability and error associated with them.  For this reason it is strongly recommended that all assays be run in duplicate when this is feasible. The best method for running duplicates is simply to duplicate the entire assay in a new set of assay plates.  This is far more reliable than re-analyzing or re-reading the same assay plates twice.  Dual data points from an assay allow the researcher to concentrate only on positive results detected in both assays and can result in reduction of false positive rates by up to one half.

Control readings are an essential part of a well-designed assay and every assay should make use of as many controls as possible.  In general there are two types of controls: plate-based controls and assay-wide controls.  Plate-based controls are controls that are placed on each individual assay plate.  These are essential in identifying plate by plate variability, and detecting assay background levels. Assays that are prone to plate-wise variability such as luciferase readouts (which decay over time) should make use primarily of plate-based controls and normalization (see also below).  Usually stock compound plates will be formatted with empty wells for the purposes of controls and it is good practice to use all available wells, with the researcher deciding upon the appropriate division of positive or negative controls needed. Assay-wide controls are separate plates containing only control wells and no screening compounds.  These are particularly useful for determining the background levels of an assay and should be used to help determine whether an assay has sufficient signal to be reliably detected.

Many assays involve a readout that is time dependent and therefore have background and intensity levels that will vary over time and by plate. Any screen that has an appreciable change in signal intensity and background from plate to plate should first be scaled using fold induction by dividing the observed value in each well by the plate median or the plate control well medians, depending upon experimental design.  In general, plate median is more reliable to use for re-scaling or normalization than plate mean as it is less affected by outlier values.  Screens without appreciable time-based or plate-based signal intensity variance should forego the fold-induction calculation and simply be normalized on a plate by plate basis by calculating the z-score, or number of standard deviations from the mean for each readout value.  These z-scores can then be used as an indication of the probability that a screening positive is not due to background noise. 

top of page

Software tools necessary to support high throughput screening fall into two categories: software to facilitate data collection and analysis, and software to organize and search compound structures. Some higher-end software packages combine these two functions into an integrated package.  The ICCB/ICG currently utilizes such an integrated package called ActivityBase, supplied by IDBS.  Other available integrated packages include HTS from Accelrys, and AssayExplorer from MDL.  These integrated packages are very expensive and should be considered only if there is adequate IT staff to support them. As an example, the package used by ICCB/ICG requires Oracle and therefore the related Oracle expertise to maintain.  In addition, these packages require much thought and work to configure correctly so that their features are utilized appropriately. The advantages to such packages are that they are capable of storing large amounts of data from many assays, and allow comparison of data across multiple assays and integrated access to chemistry information such as substructure searching.

In many cases high-end integrated packages are not feasible due to cost, personnel, or simply because the number of assays to be run is small.  It is possible to work with assay data using much simpler and cheaper tools.  For data analysis, most raw numeric data from plate readers can be easily handled within Excel or in any other good data analysis/statistics software package.  It is quite common for researchers to analyze the data from their own assays within Excel and then provide the results for entry in a centralized database for comparison with other screens or access to compound structures.  Very simple custom databases can be constructed to maintain analyzed assay data, but care should be taken to design a common format for data entry that is adhered to by all users. 

Some very simple and cost effective tools exist for dealing with chemistry data, the most common being ChemOffice from CambridgeSoft and ISIS from MDL.  These tools provide basic databases for cataloging and maintaining chemical compound collections.  Most compound collections are provided from the supplier with an associated SD format file, which contains the compound structure and re-order information. These files can easily be imported into either ISIS or ChemOffice ChemFinder so that structures can be browsed and searched.  Plate and well information can also be added for the formatted compounds if this information does not exist already.

top of page

Compound Purchase

There are many compound libraries available for purchase from commercial suppliers.  These fall into two broad classes: libraries assembled from compounds that were collected as discretes from chemistry labs worldwide (especially from Russia and Ukraine) and combinatorial chemistry libraries that are synthesized by labs affiliated directly with the vendors.  Generally, the collected compounds are cheaper than the vendor-synthesized compounds, but vendors often provide more follow-up options (e.g. guaranteed re-supply/re-synthesis) for their own compounds.

In choosing a compound supplier, one needs to consider not only the cost per library and the quality/purity of the compounds sold, but also the re-supply cost and availability of individual compounds for follow-up experiments.  As time passes and vendors add new compounds to their collections, availability of older compounds generally falls.  It is important to ask a potential supplier how long re-supply of their compounds is guaranteed (one year from the purchase date is typical) and what options are available once compounds are out-of-stock.

The ICCB/ICG has purchased compound libraries from ChemBridge, ChemDiv, Bionet, Maybridge, Peakdale, and CEREP.  We contacted these suppliers initially because of recommendations from colleagues in industry. Most were willing to provide discounts to academic institutions and, importantly, none place intellectual property restrictions on the compounds they sell.

Once a supplier is selected, the next step is to choose which compounds to purchase.  The degree of choice offered and the level of assistance provided by the vendors during this process varies greatly.  Some suppliers sell pre-assembled libraries for a set cost; others allow the buyer to choose individual compounds from a larger collection and charge per compound.  If possible, it’s very helpful to get advice from colleagues who are medicinal chemists or who have experience with small molecule screening.

When the ICCB/ICG last purchased compound libraries, we selected collections that are enriched for complex heterocyclic compounds and compounds of higher molecular weight (we favored an average mw of ~350-400) because we felt that these were more likely to provide interesting hits in our screens.  We sought to minimize the number of potentially “bad” compounds, those with groups that might make them unstable or toxic.  In particular, we tried to eliminate unstable imines, compounds with free carboxyl groups, and compounds with building block elements that might chelate metals.

When placing an order, one must specify how the compounds should be shipped, including the type of plate and the number of rows left empty per plate.  ICCB/ICG generally purchases 0.5 –1.0 mg of each compound, enough for several copies of each screening plate.  If possible, request that the compounds arrive already dissolved in DMSO--it’s quite time consuming to get >10,000 dry compounds into solution.  The ICCB/ICG commercial compound stocks are stored at 5 mg/ml (see below for more details about compound storage and handling).  Note that deep-well liquid handling capabilities may be required for compound re-formatting into 384-well screening stock plates as compounds are almost always shipped from the supplier in 96-deep well plates or in 96-tube racks.  We always ask for at least 1 column to be left empty in each 96-well plate.  This results in 2 empty columns per plate for controls once the compounds have been formatted into 384-well plates.

top of page

Compound Storage and Handling

Because the characteristics of individual compounds within screening collections vary greatly, there is no single ideal storage solution for compound libraries.  Typically, in industry, compound stocks are dissolved in DMSO and stored frozen, either at 4^oC or –20^oC.  Because DMSO is hygroscopic and because many compounds used in screening are not soluble at high concentration in water, compound stock plates are stored in a dessicated environment.  Finally, to promote compound stability, it is recommended that the number of freeze/thaw cycles experienced by compounds is limited (fewer than 15-25 cycles is preferred).

Expensive plate storage devices and servers are available to organize and store compound stock plates under controlled atmosphere at controlled temperatures.  These generally require bar-coding of stock plates and integration with other screening instruments via robotic arms.  Such storage systems are wonderful if money and space are unlimited, but are probably not practical purchases for starting screening facilities in academe.

At the ICCB/ICG, we store our compounds in DMSO in polypropylene 384-well plates made by Marsh (#AB-0781) , ABgene (#AB-1055), and by Genetix (#X5005).  The plates are sealed by hand using Corning/Costar aluminum seals (Costar #6570) that have a DMSO-resistant adhesive.  Compound stock plates are stored at –20°C in plate racks, in custom-made dessicators that fit into Revco freezers.  Each rack holds 66 plates, each dessicator holds 2 racks, and 8 dessicators fit into one freezer, for a total of 1056 plates per freezer.  Plates are thawed at room temperature in dessicators before they are used for screening.

We make five copies of each compound stock plate—four for screening (pin transfer into assay plates) and one for cherry picking.  Only one copy of each screening plate is active at a time; the others are held in reserve and put into use as the active copy ages or is depleted.  For our commercially purchased libraries, each copy plate starts with 40 ul (in a standard Marsh plate #AB-0781) or 20ul (in an Abgene low volume plate #AB-1055) per well of 5 mg/ml compound (this corresponds to ~10 mM for a compound of mw 500).  In a typical screen, 100 nl of compound stock will be transferred into a 30 ul assay volume (a 300-fold dilution of compound; we recommend screening in the range of 10-50 uM compound).  Because we have multiple copies of each library plate, we retire screening plates to deep storage after they have been through approximately 25 freeze/thaw cycles, and activate a new screening copy. However, this is still not ideal and we are currently exploring better options for storage and handling of our compound collection.

top of page

Consumables

For help in making estimates of the consumables costs for carrying out high throughput screens, please contact members of the ICCB/ICG screening group or consult the screening supplies section of our website: http://iccb.med.harvard.edu/screening/supplies.htm.

top of page

Useful Resources

Janzen, W.P. 2002. High Throughput Screening: Methods and Protocols. Humana Press, Totowa.

Seethala, R., Prabhavathi, B.F. 2001. Handbook of Drug Screening. Marcel Dekker, Inc., New York.

top of page