MSACL 2019 EU Abstract

High-Throughput Automated Tissue Imaging by DESI-MS, Three-Dimensional Tissue Imaging or Large Sample Cohort Mapping

Emrys Jones (1), Emmy Hoyes (1), Praveen Harapanahalli (1), Vincen Wu (2), Richard Chapman (1), Zoltan Takats (2)
(1) Waters Corporation, Wilmslow, UK. (2) Imperial College London, London, UK

Warning: Undefined variable $headshot in /var/www/html/view_abstract/view_abstract_in_program.php on line 704

Emrys Jones (Presenter) Waters Corporation

Relevant Financial Disclosures (within past 24 months) Salary Waters Corporation Discussing Products? YES, presenter will present educational content related to products/services of one or more ineligible company(ies) listed above. Conflict mitigation IN PROCESS.

Introduction
One barrier to the adoption of imaging MS as a tissue imaging tool for clinical applications has been the time that is required per sample, which includes the sample preparation, set up of the experiment and then the acquisition time. With ambient techniques such as desorption electrospray ionisation there is no sample preparation step, and with the time per sample now in the range of minutes, not hours, then it is a highly suitable system to investigate automation.
Objectives
To demonstrate that it is possible to obtain reproducible and clinically significant data from large batches of tissue samples in an automated manner by DESI-MS imaging.
Methods
A DESI (Prosolia, Indianaopolis, US) stage was modified such that it could be coupled to a 200-capacity slide loader (Prior Scientific, Cambridge, UK). In house software was written to communicate between the mass spectrometer acquisition software and the control software of the slide loader. The result is a system where a number of slides can be placed into the cassettes and added into a queue. Each slide is sequentially loaded onto the stage where object detection algorithms locate the tissue section(s) and direct the analysis to that area. Realtime data processing software can monitor these analyses and give immediate results to the user if required.
Results
One method to demonstrate the applicability of this approach is to take many sections from the same tissue at different depths. These were then automatically imaged using the system described here and the data reconstructed to form three dimensional chemical maps of the tissues. Cancer related samples such as spheroids and mouse xenograft tumour have been mapped by this approach with regions of hypoxia and proliferation distinguished.
With the utility of the approach demonstrated, it was then used as a platform for quickly mapping a full tissue classification study. The benefit of being able to automatically conduct such a study in one batch is that the experimental variability is minimised, allowing the biologically significant differences both intra-sample but more importantly inter-sample to be more easily determined. After the batch analysis was completed, the slides were stained and annotated by a pathologist. The tissue specific regions of interest were extracted from each MS dataset and analysed by a host of multivariate techniques.
Conclusion
With improved performance and robustness, the implementation of an automated DESI-MS approach had become feasible. Demonstrations of the technology include multiple section imaging from the same tissue, for 3D mapping, and multiple sections from various tissues for building up tissue-specific mass spectrometry databases.