page 52 Lab Times 6-2017 Products P ipetting is the most essential task in a life science lab that nobody can es- cape. Though it is basically an easy job that may literally be run by robots, it is also very demanding and may ruin eve- ry experiment, if not done properly. What makes pipetting especially hard, is its re- petitive nature, which not only stresses the thumb but is also mentally exhausting – keeping the concentration high during a tedious pipetting marathon can be a very tough experience. No wonder then that many technicians, PhD students or postdocs, especially in wet labs, dream of pipetting robots aka auto- mated liquid handlers and dispensers that take over their daily pipetting routine. Sup- posing their group leader’s pockets are not too empty, they may choose from a variety of commercial liquid handling robots, rang- ing from small benchtop budget models to full-blown workstations with “slightly” big- ger footprints. Liquid handlers are rather simple con- structed instruments, built of three main parts: a deck, holding standardised micro- plates or other reaction vessels and contain- ers; a liquid handling arm, installed above the deck, that slides back and forth along the x- and y-axes; and a pipetting head, or- thogonally joined to the arm, that moves up and down the z-axis. This very basic set- up is usually found in affordable, stripped down liquid handlers, intended for simple standard liquid handling tasks, such as se- rial dilutions. Modular platforms If you wish to do more, modular plat- forms expandable with additional robot arms, adapters, carriers for standard lab- ware, barcode-reading modules and ex- changeable pipetting heads with single or multichannel tips may be a good choice. But the higher flexibility comes at a higher price and, pretty soon, you end up at very complex instruments that usually need ex- pert knowledge for programming. Similar to manual or electronic pipettes, pipetting heads of liquid handlers function as air displacement or direct displacement pipettes. In the former case, tip content and piston are separated by a moveable air cush- ion that aspires the liquid into the tip or pushes it out. Direct displacement heads usually apply disposable syringes that get in direct contact with the pipetted liquid. Both types of heads can handle small microlitre volumes very efficiently and reliably. But they will get in big trouble at lower nanoli- tre volumes that disobey classical physical laws of liquids and follow the sometimes obscure rules of microfluidics instead. Transferring tiny nanolitre volumes falls into the realm of automated dispens- ers, based on contact or non-contact dis- pensing techniques. In simple contact dis- pensers, a liquid drop is transferred via a thin pen. The pen is loaded with a defined volume by dipping it into a liquid reser- voir. The liquid drop then sticks to the tip by adhesion forces and is deposited on the surface of the reaction vessel when the tip smoothly touches it. It is pretty obvious that the touchdown of the pen on the surface is very critical to the liquid transfer: slight- ly too hard and the surface or the pen are gone, slightly too soft and parts of the liquid drop will remain stuck to the pen. To circumvent these problems, engi- neers designed non-contact dispensers based on piezoelectric, solenoid or acoustic dispensing. Especially acoustic dispensers have gained considerable interest in the last years. The physical principles of acoustic dispensing have been known for decades: liquid drops are ejected from the surface of a liquid, if high-energy acoustic waves are directed against the liquid. But engineers struggled for years, to find a way of adjusting and focussing the energy of the acoustic waves, to obtain drops, with exactly defined size and vol- ume. Once they had fixed this problem, the doors were open for today’s acoustic dis- pensers, working “inverse” to traditional liquid handlers: the drops are ejected from a reservoir into the wells of a receiving mi- croplate, placed upside-down above the res- ervoir. Acoustic dispensing stirred up quite a fuss in the drug-screening communi- ty, when researchers from British pharma giant AstraZeneca reported in 2010 about discrepancies in drug potency data (IC 50 values) obtained after adjusting drug con- centrations, either by direct dilution, ap- plying an acoustic dispenser, or serial di- lution with a conventional liquid handler. The researchers found considerably lower IC 50 values (higher affinity values, respec- tively), for all tested drugs after direct di- lution with the acoustic dispenser. Disturbing results Inspired by the AstraZeneca report, a group led by Sean Ekins, then working for Collaborations in Chemistry, comput- ed pharmacophores, i.e., computer-gener- ated interactions of macromolecules with ligands, based on IC 50 values, and analysed whether different dilution techniques had an impact on the modelling ( PLoS ONE 8(5): e62325). The results were even more disturbing. According to Ekins et al ., pharmacophores calculated after serial dilution with a liquid handler were not viable, in contrast to phar- macophores generated after direct dilution with an acoustic dispenser. To get a deeper understanding of this di- lution conundrum, Ekins joined forces with drug designer John Chodera from the Me- morial Sloan Kettering Institute, USA. To cut a long story short, Ekins and Chode- ra are pretty sure that the serial dilution of small volumes with fixed washable tips and a liquid handler leads to a significant accumulation of bias, even after a short (8-point) dilution series ( Journal of Com- puter-Aided Molecular Design , 29, 12, 1073- 86). Though Ekins and Chodera speculate that the adsorption of liquids to the plas- tic surface of the tips may be responsible for the error during serial dilution, the ex- act reason is still not clear. But whatever it is, AstraZeneca has taken action and com- pletely switched to acoustic dispensing in its drug discovery facilities. Harald Zähringer Product survey: Automated liquid handlers and dispensers Dropwise Traditional liquid handling robots are very versatile instruments, covering a wide volume range. But sometimes, it is wiser to switch beforehand to automated dispensers, capable of dispensing tiny nanolitre droplets.