Introduction

As one of the first steps in the drug discovery process, high-throughput screening (HTS) has become a key tool in the discovery of pharmacological leads. Robotics and automation are now absolutely essential to perform the tremendous number of tests that a company conducts each year to discover new drugs in every step in the drug discovery process from lead optimization to clinical trials. Before robotics and automation were widely used in screening laboratories, it was customary for a researcher or laboratory technician to take one full year to test only a few hundred compounds [1]. Now, with robotics and other laboratory automation, it is possible to screen 100,000 or more compounds in a 24-hour period.

In addition to primary and secondary screens, robotic systems are now used in additional areas in the drug discovery process, such as early in vitro ADME (absorption, distribution, metabolism, and excretion) and toxicity studies. Because robotic systems can perform multiple assays simultaneously, early ADME and toxicity testing can now be performed at the same time as other HTS assays, allowing compounds to be profiled earlier in the drug discovery process. With the results of in-vitro ADME and toxicity tests, lead candidates can be selected that have the greatest chance of passing the later in vivo ADME and toxicity studies, ultimately leading to more successful drug candidates. Early drug profiling is useful because it allows companies to focus their efforts on lead compounds that have the greatest chance of becoming successful drugs.

Later in the drug discovery process, robotic systems again make a significant contribution by testing the data arising from the clinical trials. When a drug candidate goes to clinical trials, the high uptime and productivity of robotic systems provides fast sample analysis, quick turnaround time (TAT), accurate testing, and easy scale-up, without dramatic changes in the level of the human workforce. This provides more flexibility in the clinical trial and allows a higher trial population, because, within the system limits, sudden increases in the clinical sample load do not require a sudden increase in the number of laboratory technicians to analyze the samples. With more data reported and analyzed faster, the time for regulatory approval can sometimes be shortened, bringing the drug to market faster. One company recently estimated that by saving a year in drug development, sales increased by $580 million in the first year alone [2]. A typical robotic system is shown in Figure 1.

To put the term in perspective, automation is the technique of making a process automatic [3]. Fortunately, for those who work in today's laboratories, modern-day instruments are highly automated with sample changers, injectors, ''sipping'' cannulae, and so forth, which free the technician to perform other tasks. Indeed, the laboratory of today contains instruments that accept the sample,

perform the analytical test, reduce the data, and send the results onto the laboratory data network or laboratory information management systems (LIMS).

Although in recent years many laboratory analytical instruments have gradually become more automated through improvements in hardware and software, laboratory robots have been a relatively new introduction to pharmaceutical laboratories. Robots were generally first applied to laboratory tasks in the early 1980s, with Zymark Corp. (Hopkinton, MA) being credited as the first company to have developed a multifunctional robot specifically for widespread applications in analytical laboratories.

The classical guidelines for a task to be a candidate for robotic automation are the ''four H's,'' i.e., hot, heavy, hazardous, and high-cost labor. Although laboratory tasks might be hazardous, depending on the chemical compounds being used, they usually are not hot or heavy, so on first glance it might have seemed unusual that robots would be used in the laboratory. However, although laboratory applications do not have much in common with industrial robotic applications, the laboratory has many elements that promote the use of robotics: high cost of labor, repetitious tasks, tasks that require accuracy, and the demand for higher productivity.

The high cost and limited supply of scientific and laboratory labor, coupled with the increasing workloads in drug discovery organizations, has required the use of many types of laboratory automation, including, of course, laboratory robotics. Technicians and scientists are too highly paid and too valuable to the organization simply to perform the simple, repetitious pipetting, incubation, and microplate transfer tasks in a typical screening assay.

Another reason for the widespread use of robotics is the need for precision. As microplate densities have increased and compound volumes decreased, it has become even more important to perform all steps in a laboratory method in the same manner. Liquid transfer and reagent addition, incubation times, washing or filter harvesting techniques are among a few of the steps that must be performed identically in order for the results to be comparable from run to run. Otherwise, erroneous conclusions may be drawn from the data, easily resulting in an incorrect conclusion about a compound, or a lost opportunity, both of which can be very costly in today's competitive drug discovery environment.

Unlike the automation found on typical analytical instruments, a robot is a reprogrammable mechanical device capable of gripping, for example, a test tube or microplate and moving it through the work envelop [4]. Of the other types of automation commonly used in drug screening, automated pipettors (or ''liquid handlers'' as they are sometimes called) are probably one of the most common automated instruments in the laboratory. These instruments usually do not have gripping capabilities but are often equipped with multiple fixed cannulae or, alternately, disposable pipetting tips. Automated pipettors began appearing in laboratories in the early 1980s and now have become standard equipment in most screening laboratories. There are many different types of automated pipettors from a range of suppliers, including Tecan, Inc. (Research Triangle Park, NC), Packard Instrument Company (Meriden, CT), Hamilton Company (Reno, NV), and Rosys, Inc. (Wilmington, DE) to name a few. A typical liquid handler is shown in Figure 2.

With a liquid handler, the operator will place the various test tubes, microplates, or other containers on the pipetting surface and the liquids will be transferred as programmed. After the liquid transfers are completed, the operator replaces the containers, possibly refills the reagent reservoirs and any other consumables, and the automated pipettor continues its tasks on the next set of containers. Thus, the liquid transfer steps are performed automatically, but the instrument relies on an operator to change the vessels on the deck and replenish reagents and other consumables.

On the other hand, laboratory robotic systems can run for hours or even days with little operator attention. The operator is only needed to replenish the

Figure 2 Robot arm placing a microplate onto a typical liquid handler.

Was this article helpful?

0 0

Post a comment