TO DOWNLOAD A COPY OF THIS POSTER, VISIT WWW.WATERS.COM/POSTERS ©2017 Waters Corporation

INVESTIGATION INTO MALDI SAMPLE PREPARATION COMPARED TO DESI FOR MULTIMODAL MSI IN PRE-CLINICAL BREAST CANCER

Lee Gethings1, Jonathan Sleeman2; Kirill A. Veselkov3; Mark Towers1 , Philippa Hart1; Emmanuelle Claude1 1Waters Corporation, Wilmslow, UK; 2Centre for Biomedicine and Medical Technology Mannheim, Mannheim, Germany & Institute for Toxicology and Genetics, Karlsruhe Institute of Technology (KIT), Karlsruhe, Germany; 3Department of Surgery and Cancer, Imperial College, London, UK

INTRODUCTION Breast cancer is a complex and heterogeneous disease that has distinct biological features and clinical phenotypes. The biochemical understanding of breast cancer, which is the most frequently diagnosed form of cancer and the second leading cause of death in Western women, has greatly profited from research using genetically modified mouse models1.

Mass spectrometry imaging (MSI) is an established analytical tool for biomolecular research which can accurately determine the spatial location of molecules in a tissue section. MALDI has been a widely for MSI, but in the last few years, other techniques such as DESI have become increasingly popular.

In this study, we present data comparing several sample preparation protocols for MALDI MSI and contrasting with DESI MSI using mouse breast normal and tumour samples.

METHODS Tissue sample preparation

Experiments were conducted on normal (control) and tumour samples from the polyoma middle T oncoprotein (PyMT) mouse model of breast cancer, which were sectioned using a cryotome and deposited onto standard microscope slides preserved at -80C degrees until analysis by mass spectrometry. For MALDI experiments, a SunCollect nebulizing spray device was used to evenly apply the different matrices in several coats: 1) α-Cyanop-4-hydroxycinnamic acid (CHCA) 5 mg/mL solution in acetonitrile/water (70/30 v/v), 2) CHCA 5mg/mL solution in methanol/water, (70/30 v/v) and 3) 2,5-Dihydroxybenzoic (DHB) acid in 5mg/mL solution in methanol/water (70/30 v/v) .

DESI imaging experiments require no sample preparation as desorption and ionization are initiated by charged droplets (95% MeOH, 5% water) impacting directly on the surface.

Mass spectrometry

All experiments were carried out on a MALDI SYNAPT HDMS G2-Si mass spectrometer in positive ionisation mode, with a mass range of m/z 100—2,000. When the intermediate vacuum MALDI source was in operation, a solid-state diode-pumped ND:YAG laser with a repetition rate of 1KHz was used. When the 2D DESI stage was mounted, the MALDI source was uncoupled and the electrospray inlet block was installed along with an inlet capillary.

When the Prosolia 2D DESI stage was mounted, the MALDI source was uncoupled and the electrospray inlet block was installed along with an inlet capillary.

MALDI-MS Laser: Nd:YAG laser (355 nm)

Pulse rate: 1000 Hz

Spatial resolution: 45 µm (lateral)

DESI-MS Flow rate: 1.5 µl/min

Capillary voltage: 4.5 kV

Nebulising gas: 5 bar

Spatial resolution: 50 µm (lateral)

Stage speed: 100 µm/sec

Data management

Slides were scanned using a flatbed scanner and regions to be imaged were defined in High Definition Imaging Software (HDI) v1.4 (Waters).

DESI and MALDI imaging datasets were mined using MassLynx and DriftScope as well as processed and visualized using HDI 1.4 where the ion mobility dimension is fully utilized and integrated.

Extensive statistical analyses were carried out using EZ Info (Umetrics). Regions of Interest (ROIs) defined in HDI 1.4 and associated intensities were averaged and TIC normalised in the form of a .csv file which was loaded directly into EZInfo.

References 1. Elaine Y.Lin et all, Progression to Malignancy in the Polyoma Middle T Oncoprotein Mouse

Breast Cancer Model Provides a reliable Model for Human Diseases, American Journal of Pathology, Vol. 163, No. 5, 2003

RESULTS Control breast tissue analysis

Initial experiments were carried out using the control tissue consecutive tissue sections, prepared with the three sample preparation methods for the MALDI experiment and also analyzed using DESI. Furthermore an additional tissue was H&E stained. The control breast tissue contained mostly adipose tissue, surrounding the remaining lymph node tissue for two of the tissue sections (DESI and MALDI CHCA MeCN/Water).

CONCLUSION Control and tumour mouse breast tissue sections were analyzed by DESI and MALDI .

Three sample preparation methods for the MALDI imaging experiments were tested: CHCA in MeCN, CHCA in MeOH and DHB in MeOH.

From the control tissues which contain a high level of adipose tissue, phosphatidylcholine and triglyceride class of lipids were detected.

Depending on the type of TGs, the protonated ones were detected outside of the tissue, whereas potassiated species were detected within the tissue, mainly by DESI.

From the tumor tissues, phosphatidylcholines were predominantly detected with the difference depending on the ionisation techniques or MALDI sample preparation method.

For Research use only, not for use in diagnostic procedures.

Figure 1 displays MS spectra from specific regions, highlighting the difference in relative intensities of the different lipids ionized with either the various ionization techniques or/and MALDI sample preparation methods. A large number of phospholipids and triglycerides were detected directly from the tissue under each condition. However triglycerides (TG) were preferentially ionized using DESI. After mining of the data using PCA, it was observed that a series of TG were detected within the tissue whereas a second series TG were detected mainly outside of the tissue, especially with all MALDI sample preparation experiments (figure 2).

Figure 1. MS spectra from different tissue types of control and tumour mouse breast tissue sections, analyzed by DESI-MS im-aging.

Figure 2. Principal Component analysis score plot (A) for the com-parison of the breast control tissue analyses. (B) is the loading plot from the PCA.

The TG which are delocalized are the peaks highlighted in grey on the MS spectra in figure 1. The TG localized within the tissue are highlighted in yellow. To understand the difference between the two series, MS/MS experiments were carried out (see figure 3) and identification was possible using the fragment information from Lipid Maps. From table in figure 3, it can be seen that the delocalized TG are protonated with (16:0) and (20:4) fatty acid chains whereas the localized TG are potassiated with a (16:0) fatty acid chain.

Figure 3: MS/MS spectra of the TGs, (A) localised TG within the breast tissue section TG (16:0/18:3/18:1)K+, (B) delocalised TG outside the breast tissue section TG (16:0/18:3/20:4)H+.Table of potential identification of the localised and delocalised TGs de-tected.

Tumor breast tissue analysis

The tumor breast sections contained 95 % of tumor tissue according to pathologist annotation from the consecutive H&E stained image. From the statistical PCA results (figure 4) and database search by mass accuracy from Lipid Maps, it can be seen that most of the lipids present in the data were phosphatidylcholine and phosphatidic acid: PC and PA. From the Total Ion Current (TIC) normalized ion images, PA either sodiated and posstasiated were detected using MALDI, with slightly more intense ion images generated with CHCA in MeCN. Protonated PC’s were better ionized using MALDI prepared using the DHB matrix. Potassiated PC’s were ionized preferentially using DESI whereas sodiated PC’s were less abundant from the MALDI CHCA in MeOH dataset. TGs were also detected on the edge of the tumor tissue, especially from the DESI and CHCA in MeOH datasets.

Delocalized TG Localised TG Triglyceride Identification

851.7197 TG (16:0/16:2/20:4)H+

853.7308 TG (16:0/16:1/20:4)H+

855.7396 TG (16:0/16:0/20:4)H+

867.6866 TG (16:0/17:2/17:1)K+

869.7008 TG (16:0/17:1/17:1)K+

871.7135 TG (16:0/17:0/17:1)K+

877.7267 TG (16:0/18:3/20:4)H+

879.7429 TG (16:0/18:3/20:4)H+

881.7567 TG (16:0/18:2/20:4)H+

893.7017 TG (16:0/18:3/18:1)K+

895.7165 TG (16:0/18:2/18:1)K+

897.7307 TG (16:0/18:1/18:1)K+

Figure 4. A&B) Principal Component analysis score plot (A) for the comparison of the breast control tissue analyses. (B) is the loading plot from the PCA, C) MS images of the breast control tissues from the four experiments.