Virtual Wells: Combinatorial Biology Demands
Ultra-High-Throughput Screening.
Kathlene A. Thompson
ChromaXome Corporation
11111 Flintcote Avenue, STE A
San Diego, CA 92121
http://www.netsci.org/Science/Screening/feature04.html
Introduction
Where would we be without the 96-well plate? This simple-looking little plastic tray has displaced untold tons of glass test tubes, and corresponding generations of glassware washing students over the past 40 years. An entire support industry exists to provide ancillary supplies and compatible devices for multi-well plates. Regardless of your specialty, there are probably new 96-well (or other multi-well) plate protocols available that would make your work easier or more productive.
However, in some fields, such as the pharmaceutical industry, those familiar trays have the effect of a ball and chain. In this article, I will review the historical role of multi-well plates, and discuss the dire need --and potential-- for alternative sample handling formats, using combinatorial biology as an example.
Development of the Multi-Well Format
First established in the 1950's in Europe with 100 wells, the 96-well plate first appeared in the US in the early 1960's (1). Until the 1980's, these plates lived up to their trademarked name of "Microtiter" plates, since they were best known for their use by immunologists to measure antibody titers on a micro scale.
However, a burst of new and sensitive research tools, and a general trend toward dealing with increased numbers of samples has led to an explosion of other applications for multi-well plates. Different fields have established a spectrum of protocols. The hundreds of applications fall into three general categories: production, storage and analysis. Some of the most massive users are in clinical testing (for ELISA assays), basic research (for screening of monoclonal antibody clones, for tissue culture of mammalian cells, or for production or screening of ordered array genome libraries of yeast or bacterial cells), and the pharmaceutical industry (for drug discovery screens, and for production, receipt and /or storage of combinatorial chemistry libraries).
An entire support industry has grown up to provide devices, reagents and equipment for multi-well plate applications. These include multi-channel pipettors, plate readers, massive robotics, thermal cyclers, pH meters, and incubators, all tailored to fit the dimensions, or "footprint", of a standardized multi-well plate. Within the confines of the standard footprint, the plates themselves are generated in a dizzying range of configurations, including material, number and shape of wells, chemical or biological coating, level of sterility, transparency to selected wavelengths of light, plate depth, and built-in filter panels. The multi-well plates are disposable, without exception, so they are cheap enough to use for archival storage and everyday use.
This fruitful symbiosis between plate users and the support companies is a remarkable example of how entire fields of endeavor can shape themselves around a simple decision. In this case the decision, conscious or not, was to accept the multi-well plate as a standard format for sample handling. Each additional multi-well reagent or compatible device that has been developed strengthens this position, resulting in a rich and growing resource of high-efficiency tools for scientists.
However, conscious or not, this standard also introduces an arbitrary scale of numbers: units of 96. A scientist with a set of tools designed to manipulate 96 samples at a time will think of experiments in terms of multiples of 96. But what if you need to be thinking of multiples of 1,000,000?
The Word "Combinatorial" Means "Astronomical Numbers"
Whether the sample is a crude extract of tree bark, an HPLC fraction, or a complex mix of combinatorial chemicals, the entry point into the drug discovery process is inescapably a multi-well plate. Information management systems can track the identity, source, history, ownership, and activities of the contents of each individual well. Over the past decades, this has proven to be a godsend in the pharmaceutical industry, allowing drug discovery groups to examine escalating numbers of samples for the select few that will progress to the next round of screening, and eventually to market.
However, in the last 8 or so years, the scale of sample numbers has shot off the charts. First combinatorial chemistry introduced the concept of screening millions of chemicals. Now combinatorial biology (more below) has arrived as the latest manifestation of a chemistry source capable of generating astronomical numbers of samples, each needing screening.
The number of assays for each of those individual samples has also been escalating. These include an incredibly creative and rapidly evolving profile of biological assays (using purified proteins, as well as whole organisms or cells) and chemical assays (tracking chemical characteristics). The read-out of the assays may be fluorescence, luminescence, radioactivity, visible coloration, peaks on a graph or straight numbers, but by far, the most common assay format is still 96-well plates. This standardized format has allowed automation to be installed for at least some component of all major screening programs.
Logarithmic Consequences of Escalation
Thus you need to think in bigger multiples than 96 to succeed in the business of drug discovery. Decreasing returns on increasing investments of capital and manpower, in addition to rising costs of clinical studies and FDA regulations have intensified the hunt for new samples and new assays, on the costly route to finding new drugs. New assays spring up --newly-cloned receptors or enzymes-- and are replaced within months. New sources of chemical samples, such as combinatorial chemistry and biology, can generate astronomical numbers of samples. The resulting matrix of assay:sample combinations is infinitely large.
The size of this sample:assay matrix has resulted in a new problem in the pharmaceutical industry. Despite the extensive use of robotics, there are just too many samples and too many assays -- 96-wells are not enough. Multi-well plate assays are simply too inefficient, and thus too expensive for high-volume and high-risk samples. Unfortunately, the high-risk samples are a very promising frontier as discovery groups hunt for new, untapped sources of chemical diversity.
Furthermore, even for discovery groups that are not overwhelmed by the numbers, the expense alone is a compelling reason to look beyond 96 wells. It is true that the wells in a multi-well plate are so small that they are measured in microliters, but they are immense in the eyes of a purified protein (so to speak). For better or worse, some of the most exciting and creative new assays use purified, cloned proteins that sell for more than $1,000,000 per gram, so the cost per assay can easily pass $100. As a result, there is an irresistible movement toward miniaturization and nanoassays. Fortunately, many of the new lines of assays and assay reporting mechanisms are highly amenable to miniaturization.
Numbers and Success in the Pharmaceutical Industry
The winners in this race may be the ones who make the smart choices in that matrix, but that will not always be possible. After all, we don't really know that much about what makes a good drug. Many of the top drugs that are currently available have structures that no chemist would have predicted. In that case, the winners in this race will be the ones who look at the most points in the ever-expanding matrix of sample:assay combinations (and some intelligent choices may not hurt either). Clearly, the contenders will have to think in multiples closer to 1,000,000 than to 96.
However, the question remains: if not 96-well plates, then what will be the format?
Thinking Outside of the Lines
If you were to start from scratch in 1996 to design alternative sample handling and assay formats, you would consider these points:
- all of the sample and assay ingredients are small
- some of the ingredients are very expensive
- powerful technology is available for visualization of assay results
- many samples (millions) should be assayed during each use
- each sample should have minuscule requirements for space, time, and cost of reagents and equipment
- most samples will be worth nothing
- each sample must have an identity allowing replication and scale-up
- the format should be compatible with multi-well plates
- the format should be simple to perform and analyze to allow rapid set-up of new assays
- the format should be amenable to scale-up and automation
Let's look at a brief example.
The Combinatorial Biology Example
First, a quick primer on combinatorial biology:
Many people are familiar with combinatorial synthesis. In that process, a series of steps in chemical synthesis pathways are interwoven to result in a greatly expanded set of products. The 96-well plate grid positions provide a working identity for each resulting chemical sample, allowing the researcher to return to an identification key upon selection of a sample of interest. Thus for this application, one of the primary roles of the multi-well plate is the assignment of identity to samples.
For combinatorial biology, the name indicates the use of combinatorial techniques, but in a process that involves living organisms. In this case, it is not the chemist's recipes that are rearranged, but instead it is the biological material (the DNA, or genes) that is rearranged. The result can be the same; a fortuitous rearrangement of genes can generate a rearranged biochemical synthetic pathway. Certainly, one result is the same: massive numbers of samples can theoretically be generated, starting at millions and going beyond human comprehension. The limits are imposed only by the efficiency of the production tools, and the ability to process the resulting samples.
An immediate advantage of combinatorial biology is that the individual samples are automatically encoded. In other words, each microbe that we engineer to act as a tiny natural product chemist has a built-in identity code: it has a unique DNA sequence, replicated automatically whenever the microbe is grown for use.
This encoding automatically gives the combinatorial biologists the freedom to move away from multi-well plates.
At ChromaXome, we have developed an alternative format, called Virtu-WellTM, that is very effective with combinatorial biology libraries. In this format, each individual sample (in this case living organisms, each with a unique DNA sequence) is encased in a resilient 1-3 mm matrix that is permeable to liquids and chemicals, but not to the organism. The organisms can grow, produce chemicals, and be screened, all while individually isolated within these virtual wells, as if in 96-well plates.
Set-up time for this format is less than in 96-well plates since virtual wells are self-forming, shifting upon cue from a liquid to a stable solid phase. All of the wells are seeded simultaneously while the wells are in a liquid phase. Once the virtual wells are allowed to form, the seeded cells are then locked in place. The virtual wells are then free to move in three dimensions, and fill less space as the result of more efficient packing. The overall costs is less since plates are used only for positive samples, and more significantly, the labor otherwise required for handling samples individually in the 96-well plate format is minimized.
A wide range of extant assays are compatible with this format, and we have generated a series of analytical and interfacing devices for routinely processing batches of 1,000,000+ virtual wells. Set-up, assay incubation, screening, and transferring of positive samples into 96-well plates can be done by one operator in 24-48 hours, depending on the assay.
The low cost and high efficiency of the Virtu-Well format give us the capability to explore the frontier of combinatorial biology. We can fearlessly plan to screen 10,000,000 samples to track down only enough positives to fill one 96-well plate, and consider it time well spent. The number of positives may be low, but the probability that the selected samples are truly unique is high.
Conclusion: We Need Multi-Wells and Virtual Wells
Where could we go without the 96-well plate? One destination is obvious: out of the plate. Multi-well plates have proven to be extraordinarily versatile, and will always be with us. However, a variety of new tools are available for encoding samples, as exemplified by combinatorial biology, that allow freedom from two-dimensional spatial arrays. These virtual wells should allow the drug discovery industry to take advantage of the combinatorial phenomena with greater efficiency. The flexibility of the virtual well format leads to various applications in both sample delivery and sample assays, with facile interfacing into the multi-well format.
References
1. "History of the Microplate", 1st ed., Roy L. Manns, Polyfiltronics Inc., 1995.
NetSci, ISSN 1092-7360, is published by Network Science Corporation. Except where expressly stated, content at this site is copyright (© 1995 - 2010) by Network Science Corporation and is for your personal use only. No redistribution is allowed without written permission from Network Science Corporation. This web site is managed by:
- Network Science Corporation
- 4411 Connecticut Avenue NW, STE 514
- Washington, DC 20008
- Tel: (828) 817-9811
- E-mail: TheEditors@netsci.org
- Website Hosted by Total Choice