Mike Benway, Larry Simonson and John Ung
Caliper Life Sciences Inc, Applications Laboratory, Hopkinton, Massachusetts, USA
"Assay-Ready" plates prepared from compound or library stock plates in real time greatly accelerates the High Throughput Screening (HTS) process. The ability to transfer and dilute 0.5μL volumes is a prerequisite for this function. Here we present preliminary data on 0.5μL dry dispensing using the Caliper Life Sciences Sciclone Advanced Liquid Handler (ALH) Workstation, configured with a Low Volume Head and 384/1536 Fixed-Cannula Array.
Materials and Methods
All testing was performed using a Sciclone ALH with a Low Volume Dispensing Head, and a Bulk Reagent Dispenser (Caliper Life Sciences). Absorbance was measured at 450nm using a Wallac Victor2 1420 Multilabel Counter (PerkinElmer).
Tartrazine was obtained from Sigma Chemical Company, Catalogue number T-0388. A stock dispensing solution was prepared by dissolving 2.8g of tartrazine into 1L of DMSO. The stock solution was prepared for dispensing volumes of 0.5μL. Deionized water (DI) was used as a diluent after dispensing into the dry plate.
Dispense accuracy is determined before measuring dry dispense precision. A calibration is performed to determine the volume dispensed. For the calibration, 50μL of deionized water is pipetted in to a 384 Well Flat Bottom, Untreated Polystyrene Assay Plates (COSTAR, Catalogue # 3702). Next, 0.5μL of the tartrazine solution is manually injected into each of 20 wells, successively, from a filled 10μL Hamilton micro syringe. The wells are manually mixed by multiple aspirate- dispense cycles using a 50μL hand pipettor. Any air bubbles are removed, and the plate is read for absorbance. The average absorbance reading for these manual standards is used in a single point calibration routine.
Here a single aspiration step is performed, followed by priming dispenses, then incremental dispenses into dry plates as follows:
Data Acquisition and Processing
Absorbance data is acquired on a Wallac Victor2 reader using the same read protocol as for the calibration plate. The raw results are exported to ExcelTM and %CV is calculated. The volume dispensed is calculated by comparison of signal to a manually prepared calibration standard.
Testing has been done for three and five incremental plates using the method described above. For each run, the volume dispensed and %CV for each plate is calculated and summarized (Table 1). The results for the two runs are represented graphically (Figure 3).
In this preliminary work we have found that the Sciclone ALH equipped with a Low Volume Head is capable of delivering 0.5μL of liquid into dry microplates with a precision of less than 6%CV. This is performed with no direct contact between the cannulas and the plate. In the above results, a "first plate" effect is observed, in which the first plate in a series has a higher volume. We believe this is due to residual tartrazine solution of the external surfaces of the cannulas that is removed as the first plate is dispensed.
Initial dry plate work was carried out using a 384 Cannula Array. The benefits of this array are that it can access both deep well plates and standard microplates. However, the larger cannulas do not permit formation of well formed droplets. Without such droplets, physical contact of the cannulas with the plates is required when transferring sub-microliter volumes to dry plates, potentially resulting in variable %CV's (unpublished data). For the above results, we used a shorter, finer 384/1536 Cannula Array designed for accessing 384 well plates or 1536 plates (by dispensing into four quadrants). This type of a cannula array can fully access standard microplates, or half height deep well plates. The main benefit of the smaller cannulas is that uniform droplets can form on the ends of the finer cannulas, allowing the dispense of the droplet onto the plate without contacting the plate.
The success of dry dispensing depends on assuring the Cannula Array and destination plates are parallel. Therefore, it is necessary to level any position in which sub-microliter dry dispensing is attempted. The deck of the Sciclone ALH is designed to be adjustable at each corner. The individual plate locators must also be level. It is also necessary to determine a precise dispense height for any non-contact dispense operation.
Physical factors are an important determinant of overall dispensing performance. Such factors include microplate composition, microplate surface treatments, solvent concentrations and compositions, solute (compound) characteristics and solute concentration. The current study represents 2.8g/L tartrazine dissolved with DMSO and dispensed into untreated, flat-bottomed, polystyrene plates only. Under these conditions we demonstrated that it is feasible to dispense sub microliter volumes to dry plates with very good precision.