(M1130-06-32) A Chilled Liquid Handler Deck Enables the Automation of Lipid LC-MS/MS Bioanalytical Assays

Purpose: Liquid handling robots significantly improve the efficiency and robustness of bioanalytical assays. However, many assays often require reagents and study samples to be cooled on ice. It is challenging to integrate ice baths on liquid handlers because the labware positions must be fixed on the liquid handler deck. This poster presents a modular, low-cost, and easy-to-assemble deck that keeps reagents and samples chilled on a robotic liquid handler. The deck allows for flexibility in the number of cooled units, labware types, and orientation. We demonstrate the performance of an automated lipid LC-MS/MS bioanalytical assay, which requires the preparation of QCs, standards, and blanks at 4°C, using a custom designed and assembled chiller on a Hamilton Star liquid handler.

Methods: The chilled deck consists of a cooler, 5 SBS thermal transfer decks, and vessel holders assembled on an acrylic base (Figure 1a). The thermal plates are connected to the cooler in a series configuration with coolant flowing through the plates. This design ensures even temperature distribution throughout the chilled deck. The thermal plates are positioned on a custom-machined base that ensures precise xyz-alignment and facilitates easy assembly on standard liquid handlers. Labware is positioned on tube and plate holders that are tailored to fit the specific geometries of vessels used in lipid LC-MS/MS assays, thereby preventing air gaps that could lead to nonuniform temperatures. Custom solutions were required to address the need for fixing specialized tube holders to the deck, including 3D-printing a clamp for integrating an autosampler vial rack and machining a commercially available tube rack for holding study samples (Figures 1b-c). The chilled deck was used to develop and qualify an automated lipid extraction method on a Hamilton STAR liquid handler that performs all aspiration, dispense, and mixing steps for the quantification of an ionizable lipid by LC-MS/MS (Figure 2). The manual version of the same lipid quantification method requires all QCs, standards, blanks, and study samples to be prepared and held on ice.

Results: The temperature at 5 locations across each thermal plate was 5.3 +/- 0.3°C (mean± SD), which demonstrated the feasibility of using the chilled deck to maintain a consistent temperature (Figure 1d). The lipid LC-MS/MS method consisted of 7 working calibration standards from 10 to 1,000 ng/mL prepared in blank Sprague-Dawley rat K2EDTA plasma and three QC levels that span the range of the calibration curve. The calibration standards had a CV between duplicate wells of ≤20% (≤25% at the LLOQ), an RE of ±20% (±20% at the LLOQ), and at least 75% of the calibration standards with a minimum of 6 standard levels passed acceptance criteria. The overall CV of all working standards ranged from 4.8 to 11.9%. Quality control samples were prepared at 30 ng/mL (LQC), 150 ng/mL (MQC), and 750 ng/mL (HQC) in blank rat K2EDTA plasma. The QCs met target acceptance criteria of a CV between duplicate wells of ≤20% and an RE of ±20% of at least 1 QC at each level and 6 out of 8 total QCs (LQC, MQC, and HQC) for all 3 batches. The overall CV was 11.9% (LQC), 6.5% (MQC), and 7.3% (HQC) and the overall RE was 3.9% (LQC), 1.1% (MQC), and -8.5% (HQC) (Table 1). This result demonstrated the feasibility of using chilled deck to replicate manual benchtop procedures that require handling and incubation steps on ice. The ability to perform the lipid extraction process on a liquid handler reduced the hands-on time requirement for this assay from 3.5 h to 1 h. The total cost of the chilled deck is approximately $9,740, which offers significant cost savings compared to standalone commercially available heating/cooling modules.

Conclusion: A chilled deck that maintains reagents near 4°C was assembled for the development of automated methods that require steps to take place on ice. The deck positions labware in a consistent xyz-orientation and maintains consistent temperature across 5 SBS positions. We developed a lipid LC-MS/MS method on a Hamilton STAR liquid handler using the chilled deck and demonstrated that the assay performance met non-GLP qualification criteria. Previously, the assay required the preparation of standards, QCs, and blanks on ice; performing it with robotics now significantly saves time and increases efficiency. The chilled deck can accommodate other types of assays because the tube and plate holders can be swapped to meet specific labware requirements. Additionally, we envision that this platform can be integrated with other types of automated systems beyond liquid handlers that use SBS plate formats such as tube labelers and automated synthesizers.

Acknowledgements: We thank the entire Beam Team for supporting this work. We want to particularly thank the Automation, Nonclinical Development, and Analytical Research & Development teams for their contributions.
This abstract was written with assistance from ChatGPT 4.


Figure 1. Photo of the chilled deck and a graph of the temperature measured across the deck.


Figure 2. Schematic of the lipid LC-MS/MS assay.


Table 1. QC performance of the lipid bioanalytical assay.