1
Towards 3-Dimensional MALDI MS Molecular Imaging of the Optic Chiasm David M. G. Anderson 1 , Raf Van de Plas 1 , Kristie L. Rose 1 , Kevin L. Schey 1 , Anne Solga 2 , David H. Gutmann 2 Richard M. Caprioli 1 . Mass Spectrometry Research Center, Vanderbilt University School of Medicine, Nashville, TN 1 . Department of Neurology, Washington University School of Medicine, St. Louis, MO 2 . 3-D Ion Volumes WP152 3-D Workflow method [email protected] 1. Arrange tissue in desired orientation; 2. Pre-embedded tissue in carboxymethyl cellulose (CMC); 3. Place pre-embedded block onto hypodermic needle in custom- made mold with carbohydrate rods perpendicular to the tissue; 4. Section and thaw mount tissue onto poly-lysine coated ITO coated glass slide. Plates scanned using differential interference contrast microscopy Matrix applied and data acquired from 35 sections 10 μm thick at 40 um spatial resolution, mass range 4000-26000 Da. Myelin basic protein isoform 8 (m/z 14126) Myelin basic protein isoform 8 (m/z 14126) Truncated myelin basic protein (1-43 ) (m/z 4808) Myelin basic protein isoform 8 (m/z 14126) Acyl CoA Binding Protein (m/z 9914) Myelin basic protein isoform 8 (m/z 14126) Truncated myelin basic protein (1-43) (m/z 4808) Acyl CoA Binding Protein (m/z 9914) Calcyclin, S100-A6 (m/z 9962, intensity threshold 75.7) (Identified via LC-MS MudPIT experiments, theoretical 9962 Da) Calpactin 1 light chain, S100-A10 (m/z 11090, intensity threshold 55.4) (Identified via LC-MS MudPIT experiments, theoretical 11097 Da) Truncated myelin basic protein (1-43) (m/z 4808, intensity threshold 275.6) (Identified via LC-MS top down experiments, theoretical 4809 Da) Dynein light chain 1 (m/z 10275, intensity threshold 99.4) (Identified via LC-MS MudPIT experiments, theoretical 10277 Da) Fit IMS data to registered and transformed DIC images Signals Biologically Relevant to Nf1 Registered data reconstructed in 3-D volume 100% 0% 1 5 2 3 4 6 7 8 *2-D images represented from section 14 which contains glioma confirmed by a pathologist . *2-D images of section 14 m/z 4808 *2-D images of section 14 m/z 9962 *2-D images of section 14 m/z 10275 *2-D images of section 14 m/z 11090 2-D images of section 3 m/z 14126 2-D images of section 3 m/z 14126 m/z 4808 2-D images of section 3 m/z 9914 m/z 14126 2-D images of section 3 m/z 9914 m/z 14126 m/z 4808 Results and Discussion Images displayed in the central panel (1-4) show signals which describe the native structure of the optic nerve chiasm. Figure 1 displays the distribution of myelin basic protein isoform 8 (identified with LC-MS ETD experiment) at m/z 14126 which is present throughout the central body of both the tissues with similar abundance. Images in figure 2 show a truncated myelin basic protein (1-43) overlaid with myelin basic protein isoform 8, the truncated version appears to be present in the meninges, which surround the optic nerve and is more abundant in the Nf1 model (figure 5). Figure 3 displays the distribution of an Acyl CoA binding protein at m/z 9914 overlaid with myelin basic protein isoform 8. Acyl CoA binding protein is present in the dura mater region of the meninges. These distributions become more apparent when all three signals are overlaid in figure 4. Figures 5-8 display signals which vary between the Nf1 optic glioma and wild-type nerve chiasms. The truncated myelin basic protein observed in the meninges can be seen in figure 5 to be more abundant throughout the majority of the Nf1 optic glioma and can be seen to proliferate into central regions of the tissue. The 2-D image also shows high abundance in the region known to contain a glioma. Figure 6 displays the distribution of calcyclin S100-A6 which can be seen in the 3-D reconstruction to be prevalent throughout the outer regions of the Nf1 optic glioma. The 2-D image also shows high abundance in the glioma region and the optic nerve tract leading towards the chiasm body. Figure 7 displays the distribution of dynein light chain 1 which is also observed with high abundance in the Nf1 optic glioma for both the 3-D reconstruction and the 2-D image of section 14. Figure 8 displays the distribution of calpactin 1 light chain S100-A10 at m/z 11090 which also present in the glioma region although present this signal displays a lower abundance then the previous signals. Signals observed in fig 6-8 have been observed previously in glioma using MALDI-MS from human tissue and cell lines 4 . Conclusion Embedding tissue along with external fiducial markers improved transformation and registration of 2-D IMS into a 3-D volume by making the process more accurate and efficient 5 . Reconstructing the data from 2-D to 3-D provided a view into the ion distributions in the chiasm closer to the 3- dimensional reality. Due to the very small nature of the starting material, 3-D reconstruction of multiple 2-D mass spectrometry images resulted in a number of proteins with differing expression between the wild type and Nf1 optic glioma being observed with a higher global expression. Acknowledgements This project was supported by the National Institutes of Health (NIH/NIGMS 5RO1GM058008-14) and NCI MMHCC grant (1U01CA141549-01). I would like to thank Salisha Hill (Vanderbilt) for her assistance with the MudPIT analysis. References 1. Deutskens F, Junhai Y, Caprioli RM. High spatial resolution imaging mass spectrometry and classical histology on a single tissue section. J Mass Spectrom 2011, 46,568-571. 2. Schey K. L, Anderson D. M, Rose K. L. Spatially-Directed Protein Identification from Tissue Sections by Top-Down LC-MS/MS with Electron Transfer Dissociation. Anal Chem. 2013. Article in press. 3. Wang Z, Hill S, Luther J. M, Hachey D. L, Schey K. L. Proteomic analysis of urine exosomes by multidimensional protein identification technology (MudPIT). Proteomics. 2012 Jan;12(2):329-38. 4. Schwartz S. A., Weil RJ, Thompson R. C., Shyr Y., Moore J. H., Toms S. A., Johnson M.D., Caprioli R.M. Proteomic-based prognosis of brain tumor patients using direct-tissue matrix-assisted laser desorption ionization mass spectrometry. Cancer Res. 2005 Sep 1;65(17):7674-81. 5. Amstalde van Hove, Erika R.; Blackwell, Tiffany R.; Klinkert, Ivo; Eijkel, Gert B.; Heeren, Ron M. A.; Glunde, Kristine. Multimodal Mass Spectrometric Imaging of Small Molecules Reveals Distinct Spatio-Molecular Signatures in Differentially Metastatic Breast Tumor Models. Cancer Research (2010), 70 (22), 9012-9021. Nf1 Nf1 Wild type Wild type Nf1 Wild type Nf1 Wild type Nf1 Wild type Overview • Optic nerve chiasm from a genetically-engineered mouse model of neurofibromatosis type 1-associated optic glioma compared to wild-type mice using MALDI-Imaging MS. • MALDI-IMS technology was applied on multiple sections utilizing fiducials incorporated in the embedding medium. • Fiducials provided a high degree of accuracy for transforming and registering images for 3-D reconstruction. • The resulting 3-D images provided a view into the ion distributions in the chiasm closer to the 3-dimensional reality. Introduction • The optic chiasm is the region where the optic nerve tracts coming from the eyes, meet and cross over before entering into the brain. The optic chiasm plays a key part in delivering information from the retina to the visual cortex. • Individuals with the neurofibromatosis type 1 (NF1) inherited cancer syndrome are prone to develop optic nerve glioma, leading to visual loss. • Nf1 genetically-engineered mice have been shown to exhibit similarities to humans with NF1-associated optic glioma. Using this Nf1 mouse model, we applied MALDI-IMS technology in 2-D measurements towards a 3-D reconstruction of the chiasm in order to achieve a greater understanding of the mechanisms involved in glioma development. Methods • Wild-type optic nerves along with optic nerves from Nf1 mice with optic nerve glioma were excised, and tissues were aligned in the correct orientation for sagittal sectioning before embedding in 2.7% carboxymethyl cellulose (CMC) solution within a custom-made mold. • Carbohydrate rods (angel hair pasta) were inserted into designated regions of the mold, ensuring a marker triangle for registration of the 2-D experiments to each other. • 47 sections 10 μm thick were placed onto poly-lysine coated ITO coated glass slides (45x45 mm), washed with 70% , 100% ethanol for 30 seconds each then 2 minutes in a Carnoy’s wash (60:30:10 ethanol: chloroform: acetic acid) followed by 100% ethanol, 100%, deionized water and 100% ethanol for 30 seconds each 1 . • 6 sections 10 μm thick were thaw mounted onto separate ITO coated glass slides, subject to the same wash protocol and utilized for instrument optimization and intact protein identification by LC-MS ETD analysis. • Sinapinic acid matrix was applied by spray coating with a 5 mg/mL solution 90:9.7:0.3 ACN:H 2 O:TFA using a HTX Technologies TM sprayer (8 passes with a flow rate of 0.1 mL/min). • IMS experiments were performed on a Bruker Speed TOF/TOF using a SmartBeam 1 kHz laser with a mass range of 4000- 26000 Da and 40 μm spatial resolution. • Post data acquisition the matrix was removed using 100% methanol and sections were H&E-stained so that regions containing glioma could be interpreted by a pathologist. • Re-assembly of 35 2-D IMS experiments into a virtual 3-D volume was performed with in-house software developed in MATLAB. Identification work • Whole washed and desiccated tissue sections were solubilized in 55% acetonitrile 0.15% TFA and LC-MS/MS analysis were performed as described in Schey et al 2 for intact protein identification. • 2D LC-MS/MS analysis was performed on digested proteins from a separate Nf1 mouse optic chiasm whole tissue. Analysis of tryptic peptides was completed as described in Wang et al 3 . • All LC-MS analysis were completed on an LTQ Orbitrap Velos mass spectrometer. 1 mm Nf1 Wild type 1 mm 100% 10% 60% 0% Nf1 Wild type 1 mm Nf1 Wild type 1 mm 15% 10% 0% Nf1 Wild type 1 mm Nf1 Wild type 1 mm 60% 10% Nf1 Wild type 1 mm 80% 15% Nf1 Wild type 1 mm 100% 20% Nf1 Wild type Nf1 Wild type Nf1 Wild type Nf1 Wild type 60% 100% 60% 60% 15% 100% 10% 1. 2. 3. 4. y z x y z x y z x y z x y z x y z x y z x y z x 60% 5% Wild type Nf1
