Ultra-high-throughput screening (uHTS) is an automation-based methodology for conducting hundreds of thousands of biological or chemical screening tests per day. The cutoff between high-throughput screening (HTS) and ultra-high-throughput is somewhat arbitrary. “There is no fixed boundary,” says Simon Sheard, Ph.D., business development manager at RTS Life Science (Manchester, UK), which supplies automated sample management equipment used in uHTS. The generally accepted crossover point today is 100,000 tests per day.

uHTS is conducted in microtiter plates. To provide numerical perspective, 100,000 tests per day require 1,450 96-well plates (by far the most commonly used type), 261 384-well plates, or 65 1536-well plates. uHTS programs that exceed 1 million screens per day use ten times as many plates.

Equipment for conducting uHTS is indistinguishable from a standard microplate handling system, consisting of a robotic microplate handler, a liquid dispenser, and a plate reader. Additional components for washing, agitation, bar code reading and incubation are also possible.

uHTS achieves its speed through a combination of higher-density microtiter plates and multichannel (384 and higher) liquid dispensing. Equally important in achieving high throughput, however, is assay simplicity. Most ultrafast screens involve simple binding and rapid reading of results. For this reason, uHTS lends itself most readily to drug screening where, classically, tens of thousands or hundreds of thousands of wells are plated with entries from a large compound library, and the assay reagents (protein, enzyme, cell, or receptor, plus reporting reagent) remain constant in every well. Depending on the nature of the detection event, the interaction between compound and target is read as fluorescence or luminescence.

It is possible to “cheat” in HTS/uHTS by utilizing unpurified compounds, mixtures of compounds, or even multiple targets, a technique known as high-content screening because a multiple of the information normally available is collected. Wells that “light up” are examined more closely, for example by purifying mixtures or plating components individually.

The pharmaceutical connection

Parallel screening methods have been used for decades in the pharmaceutical industry. The advent of automated plate-handling and reading instrumentation, and the replacement of radiolabeling assays with luminescence- and fluorescence-based screens, created the opportunity for the several-hundredfold improvement in throughput represented by uHTS. Original equipment was expensive, but over the past decade instrumentation prices have fallen in terms of cost per assay per day, to the point where uHTS is now accessible to small drug discovery firms and academic groups. Numerous service providers also conduct uHTS services for organizations that lack this capability or whose own systems are overcommitted.

Wei Zheng, Ph.D., a group leader at the NIH Chemical Genomics Center (Rockville, Md.) learned the HTS and uHTS trades while screening drug candidates at Merck and Amgen. One of the instruments in use at the NIH Center is a plate-handling robotic system, codeveloped by Zheng at Merck, that processes hundreds of thousands of wells per day and has 1,536-well capability. “It runs between half a million and a million screens per day, depending on the assay,” Zheng told Lab Manager Magazine. The system uses plate readers from PerkinElmer and GE, and core robotics from Kalypsys Systems.

Zheng’s group uses 1,536-well plates almost exclusively, as do most pharmaceutical labs. “Miniaturization saves time and enables higher throughput at reduced cost,” he notes. However, minuscule assay volumes sometimes create difficulties for cell-based assays. “It’s often difficult to deliver the number of cells you need for an assay at such low volumes. In these circumstances the screens cannot be run at 1,536-well density.”

Recently, researchers from the Chemical Genomics Center, in collaboration with scientists at Trinity College (Dublin, Ireland) reported on a screen of 17,143 FDA-approved and experimental drugs. The biological target in this case was a panel of human liver enzymes that metabolize drugs, and hence are critical to a medicine’s effectiveness.

uHTS received a bad reputation around the beginning of the decade, based on a perceived low success rate in identifying new drugs. The fault, says Zheng, was not with uHTS methods but with the drug companies’ choice of screening targets.

Simon Sheard agrees. “We hear comments about the failure of the ‘law of big numbers’ regularly. That’s a generalization, and the approach of cranking the handle faster has not completely fallen out of use. Nevertheless, what we have seen during the last few years is a shift away from uHTS to automated screening of smaller compound sets through assays that provide more information per well, or higher-quality data.”

HTS and uHTS systems don’t differ much in terms of instrumentation. What changes is the trend towards modularity. “Both systems employ a collection of instruments linked by software and robotics,” Sheard observes. As assay strategies become more sophisticated and screens more numerous, the number of components increases. uHTS is greatly facilitated, for example, by dedicated compound management systems that store compounds directly in readyto- test plates. At some point, Sheard notes, “It may not be sensible to have a single robot feeding plates to numerous instruments.” And all this added functionality necessitates software products that tie everything together seamlessly.

Angelo DePalma holds a Ph.D. in organic chemistry and has worked in the pharmaceutical industry. You can reach him at angelo@adepalma.com.