Inflation-Fixation Method for Lipidomic Mapping of Lung Biopsies by

Mar 30, 2016 - University of Maryland, School of Pharmacy, Department of Pharmaceutical Sciences, 20 North Pine Street, Baltimore, Maryland 21201, Uni...
2 downloads 9 Views 1MB Size
Subscriber access provided by UOW Library

Article

Inflation-fixation method for lipidomic mapping of lung biopsies by matrix assisted laser desorption/ionization – mass spectrometry imaging Claire L Carter, Jace W. Jones, Ann M Farese, Thomas J. MacVittie, and Maureen A Kane Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.6b00165 • Publication Date (Web): 30 Mar 2016 Downloaded from http://pubs.acs.org on April 2, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 14

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Inflation-fixation method for lipidomic mapping of lung biopsies by matrix assisted laser desorption/ionization – mass spectrometry imaging

Claire L Carter1, Jace W. Jones1, Ann M. Farese2, Thomas J. MacVittie2 and Maureen A. Kane1

1

University of Maryland, School of Pharmacy, Department of Pharmaceutical Sciences, Baltimore, MD

2

University of Maryland, School of Medicine, Department of Radiation Oncology, Baltimore, MD

Correspondence: Maureen A. Kane University of Maryland, School of Pharmacy Department of Pharmaceutical Sciences 20 N. Pine Street, Room 723 Baltimore, MD 21201 Phone: (410) 706-5097 Fax: (410) 706-0886 Email: [email protected] Abstract Chronic respiratory diseases are among the leading causes of deaths worldwide and major contributors of morbidity and global disease burden. To appropriately investigate lung disease the respiratory airways must be fixed in their physiological orientation and should be inflated prior to investigations. We present an inflation-fixation method that enables lipidomic investigations of whole lung samples and resected biopsy specimens by matrix-assisted laser desorption/ionization-mass spectrometry imaging (MALDI-MSI). Formalin-inflation enables sample preparation to parallel standard clinical and surgical procedures, in addition to greatly reducing the complexity of analysis, by decreasing the number of analytes in the MALDI plume and reducing adduct formation in the resulting mass spectra. The reduced complexity increased sensitivity and enabled high resolution imaging acquisitions without any loss in analyte detection at 10 and 20 µm scans. We present a detailed study of over 100 lipid ions detected in positive and negative ion modes covering the conducting and respiratory airways, and parts of the peripheral nervous tissue running through the lungs. By defining the resolution required for clear definition of the alveolar space and thus the respiratory airways we have provided a guideline for MSI investigations of respiratory diseases involving the airways, including the interstitium. This study has provided a detailed map of lipid species and their localization within larger mammalian lung samples, for the first time, thus categorizing the lipidome for future MALDI-MSI studies of pulmonary diseases.

Introduction Pulmonary diseases are debilitating and the clinical outcome following diagnosis is often poor.1 Diseases characterized by poor gas exchange result from abnormalities of the lung parenchyma and exhibit complex pathophysiology, which is heterogeneous, and often sporadically dispersed

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

throughout the lung lobes.2 To date investigations of lung disease to ascertain pathophysiology, determine markers for faster diagnosis, and identify effective therapeutic targets for drug design and treatment, have mainly involved histology, targeted analysis of known markers using labeled techniques such as immunohistochemistry,3 or a variety of analytical and biological assays on bronchoalveolar lavage fluid (BALF).4 The latter of which loses spatial information and extracts only substances in the alveolar space and thus does not document tissue-specific changes, such as those that occur during interstitial thickening, and the parenchymal remodeling that occurs during the development of lung disease. Matrix-assisted laser desorption/ionization-mass spectrometry imaging MALDI-MSI is being applied to biomedical, preclinical and clinical studies to provide molecular information at the spatial level as a novel approach to unravelling the (patho)physiology of organs systems,5-8 in addition to drug distribution, pharmacokinetics/pharmacodynamics (PK/PD) and absorption, distribution, metabolism and excretion (ADME) investigations.9-11 Whist MALDI-MSI has been applied to several investigations mapping drug distribution in lung tissue for treatment of TB10,12 and cancers,9,13-15 minimal studies have been applied to the investigation of parenchymal lung disease,5 or preneoplastic lesions that will identify early markers of tumorigenesis. This is in part due to problems associated with the analysis of fresh un-inflated lung sections. The analysis of fresh tissue, as is often required for MSI, results in sections that do not resemble the lung in its natural air-filled inflated state. Frozen sections result in collapsed lung parenchyma, which makes interpretation of the sections incredibly difficult as they appear hypercellular, thus masking or imitating disease states. From a pathological standpoint this is undesirable during the investigation of interstitial lung disease as these sections mimic alveolar thickening and hyperplasia, due to high cell counts.16 Collapsed lung parenchyma also masks investigations of pre-neoplastic lesions as the hyper-cellular appearance of collapsed lung is difficult to differentiate from diseased regions.17 Additionally, as the level of collapse varies widely across and between frozen lung sections, interpretation of MALDI-MSI results from the investigation of endogenous changes can problematic, as differing amounts of material are being ablated with each laser spot. Investigators thus have to pay extra attention to the level of collapse to ensure that any increase in signal intensity is not due to more material being ablated as a relationship to the level of collapse, than actual increases in specific molecules within these regions as a result of disease processes. Many methods have been developed and applied to the inflation of lung tissue following lobe resection or partial resection during the acquisition of biopsies, in addition to frozen sections to aid in the rapid diagnosis of pulmonary neoplasms.16-18 These methods are currently not suitable for MSI investigations of the lung lipidome as the processing method used in the clinic removes lipids from the tissue. The method of inflation for frozen sections utilizes the cryomold to inflate the lung parenchyma, which contains polyethylene glycols (PEGs) that cause ion suppression during MS analysis and contaminates the spectra, resulting in poor sensitivity.19 To-date two methods have been published for the inflation of whole mouse lungs, one contains PEG 2000 which is less than desirable for analysis of molecules over m/z 1500,20 the other needs multiple synthesis steps which could not be routinely transferred to clinic.21 Work presented herein combines the inflation-fixation method developed by Andrew Chug for the analysis of interstitial lung disease from human biopsy samples,16 with a MALDI-MSI compatible embedding media,22 for a novel method of lung inflation that can be applied to whole lung samples and biopsy samples for lipidomic investigations by MALDI-MSI. We also investigate the resolution required for efficient analysis of the lung parenchyma and map, for the first time, the distribution of lipids in positive and negative ion mode from larger mammalian

ACS Paragon Plus Environment

Page 2 of 14

Page 3 of 14

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

airways. Results presented cover both the conducting and respiratory airways with an aim to provide a lipidomic map to aid as a reference for future investigations into pulmonary disease processes. Materials and Methods Materials. Analytical grade solvents were purchased from Fisher Scientific (Pittsburgh, PA, USA). DHB was purchased from Acros (New Jersey, USA) and 9-AA was purchased from Sigma-Aldrich (St. Louis, MO, USA). ITO-coated glass slides were obtained from Bruker Daltonics (Bruker Daltonics, Bremen, Germany). Hematoxylin and eosin (H&E) staining kit was purchased from Thermo Scientific (San Jose, CA, USA). Animals. Male rhesus macaques (Macaca mulatta, Chinese origin) were housed and cared for as previously described.23 All animals were of good health, free from signs of disease, and tested negative for simian immunodeficiency virus, simian T-cell leukemia virus type-1, malaria, herpes B virus and tuberculosis. Following euthanasia the lungs were removed and biopsy samples were taken from multiple lobes, samples were snap frozen and stored at -80C until analysis. All animal procedures were conducted in accordance with the NIH guidelines for the care and use of laboratory animals and experiments were performed with prior approval from the University of Maryland Institutional Animal Care and Use Committee (IACUC). Lung Inflation and Sample Preparation for MSI. Samples were removed from the -80C freezer and allowed to thaw at room temperature. Samples were inflated by slowly injecting formalin through the pleura into the lung using a 25 gauge needle as previously reported for surgical specimens.16 For larger lung biopsies that have subdivided regions separated by interlobular septa, multiple regions were inflated with formalin, this ensured the whole biopsy was inflated. Following expansion the lungs were placed in formalin and fixed for 8 h or overnight. After overnight fixation the samples were removed, gently patted dry to remove any excess formalin and rapidly embedded in gelatin (100 mg/mL) on dry ice. Samples were then stored at -80C until sectioning. Inflated-embedded samples were removed from the -80ºC freezer, transferred to a cryostat chamber, and allowed to equilibrate for 1 hour prior to sectioning. Frozen sections were taken at 10 µm using a Microm HM550 cryostat (Themo Fisher Scientific, MA, USA), thaw-mounted onto indium tin oxide-coated conductive glass slides (Bruker Daltonics, Bremen, Germany), and prepared immediately for MSI or again stored at 80C. Matrix deposition was carried out using the HTX sprayer (HTX Technologies, North Carolina, USA) applying 9-AA for negative ion mode analysis and DHB for positive ion mode analysis as previously described.5 Serial sections taken for histology were stained with hematoxylin and eosin (H&E) and scanned using the Aperio slide scanner (Leica Microsystems Inc., IL, USA) for image analysis. Mass Spectrometry. Imaging experiments were performed using an UltrafleXtreme TOF/TOF mass spectrometer (Bruker Daltonics, Bremen, Germany) equipped with a smartbeam II Nd:YAG laser, 355 nm. Voltages, delay time, detector gain, and laser fluence were optimized for superior ion transmission, spectral quality and resolution. For the image resolution investigation the laser diameter was operated at the medium setting (~50 µm) during the 100 and 50 µm acquisitions, and the minimum setting (~20-25 µm) for the 30, 20 and 10 µm acquisitions, the latter using oversampling. In all cases 300 laser shots were accumulated at 1kHz per raster position. High resolution imaging experiments were performed at 20 µm resolution using the minimum laser setting at 1 kHz, covering the mass range m/z 400-1000, for positive and negative ion mode. Instrument calibration was performed on known matrix and lipid ions as previously determined during fresh tissue investigation of lung samples.5 On-tissue tandem MS investigations were carried out using LID-LIFT on selected ions with a mass window of ±1-2 Da.

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Data Analysis. Imaging datasets were analyzed using FlexImaging version 4.1, (Bruker Daltonics, Bremen, Germany) with and without skyline projection enabled. No normalization was applied to any imaging datasets analyzed and presented. Single point spectra were extracted to FlexAnalysis version 3.4 for further analysis. Spectra were internally calibrated on known lipid ions for positive and negative ion modes prior to database searches. Batch searches were carried out using both LipidMaps (www.lipidmaps.org) and Metlin (www.metlin.scripps.edu), and tentative assignments of lipid species were based on mass and/or fragmentation data. If assignments could not be made the ions were not reported, or as in the case of the nervous tissue, the ions were reported but not assigned.

Results and Discussion Lung Inflation and Spatial Resolution. Cryosections of fresh lung tissue more often than not result in sections demonstrating collapse of the alveolar epithelium, with the interstitium appearing several layers thick, resembling a solid organ and not representative of the porous architecture that should be present. Figure 1A-C shows the type of histology sections and MSI data that are often obtained from frozen lung sections. The MS and histology images for the fresh lung section resemble a solid organ, the parenchyma is completely collapsed in places, which would make comparisons to disease states and characterization of the normal interstitium difficult. The only anatomical features that are clearly discernable are those of the larger airways, including the bronchi (data not shown), and the alveolar ducts, as shown by the larger spaces in the tissue sections in Figures 1B and C. For a direct comparison to inflated lung sections the biopsy lung presented in Figures 1A-C was thawed and inflated with formalin as per the methods section, sections from this methodology resulted in MS and histology images showing clear interstitium and alveolar space, as demonstrated in Figures 1D-F. The anatomical features of the lung parenchyma are clearly discernable, with defined interstitium and alveolar space, obvious upon comparison of the enlarged regions of the H&E sections presented in Figures 1C and F for fresh and inflated, respectively. Inflation with formalin enabled high resolution MSI datasets to be obtained, as shown in by the image presented in Figure 1D, which was obtained at 20 µm resolution. The images in Figure 1 show clear differences between inflated and non-inflated tissue, however, the MSI data were acquired at different resolutions, thus additional studies were carried out on inflated lung tissue to determine the resolution required for sufficient analysis and image quality of the lung parenchyma following inflation. Results from lung tissue acquired at resolutions of 100, 50, 30, 20 and 10 µm is presented in Figures 2A-E, respectively. The ion at m/z 835.7, identified as the sodium adduct of SM(d18:1/24:1), was chosen for comparison of the resolution datasets as this ion highlights the lung parenchyma. Upon comparison to histology it can be seen that the MS image from the 100 µm dataset enabled identification of the larger bronchi and alveolar ducts, with few alveolar air spaces obvious (Figure 2A). The 100 µm resolution dataset from the inflated lung preparation (Figure 2A) still demonstrated a more defined image, with superior anatomical characterization, than that obtained from the fresh section acquired at 75 µm (Figure 1A). The images acquired from the 50 and 30 µm datasets presented in Figures 2B and C appear similar, more alveolar air spaces are identifiable and the image quality is improved compared to that of the 100 µm image, as would be expected. The images at 20 and 10 µm show clearly defined alveolar spaces and interstitium, as can be seen by comparing the MS images with their respective H&E stained sections in Figures 2D and E. The latter two data sets are of far superior quality, showing crisp images demonstrating differences in signal intensity along the interstitium. As the tissue is fixed, the MALDI plume is believed to be much less complex due to the chemical crosslinking nature of the fixative,

ACS Paragon Plus Environment

Page 4 of 14

Page 5 of 14

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

meaning there will be inherently less analytes in the gas phase, thus increasing the sensitivity in the analysis of species that are desorbed and ionized. This is evident as no loss in sensitivity at the smaller laser diameter was observed, as demonstrated by the single pixel spectra taken from the lung parenchyma at each resolution, presented in Figure S1A-E. Additionally because there are less analytes in the gas phase, the spectral quality is greatly improved. A detailed description of the species identified in positive and negative ion mode, and their localization within lung tissue, is presented in the following sections. Fixation and Crosslinking of Aminophospholipids. The analysis of fixed lung tissue resulted in a loss of phosphatidylethanolamine (PE) detection and a reduction in phosphatidylserine (PS) detection, agreeing, partially, with the recent data presented by Gaudin et al.24 This group showed a loss of PE and PS species following total lipid extraction from fixed brain tissue. The authors based findings on formalin-induced alterations of human brain lipidome and stated that severe degradation of the phospholipidome occurred in formalin fixed (FF) samples, and concluded that PE and PS species were modified by formalin resulting in derivatized lipids. We disagree with the main findings of this paper because if modifications of aminophospholipids were to be the predominant factor following formalin fixation we would expect to detect these lipids as their derivatized or modified forms. As PE species are predominantly detected as the [M+Na]+ ion by MALDI-MS,25 the high sodium content, as is the environment of buffered formalin, should potentially favor their detection. We observed a complete loss in the detection of PE species from our imaging datasets and believe chemical crosslinking and thus ‘fixing’ of these lipids is the predominant reason for the loss of PE detection and the reduction of PS’. To support our hypothesis we conducted untargeted ultraperformance liquid chromatography-tandem mass spectrometry using data-independent MSE experiments on fixed and fresh whole mouse lung lobes (Figures S2 – S4). The negative ion mode analysis from fixed and fresh mouse lung tissue demonstrated a complete loss of PE species following overnight fixation, as shown by comparing the retention times between 9 and 14 minutes in the chromatograms presented in Figure S2A and S2B. The fragmentation data for the most abundant ion in the fresh chromatogram at m/z 766.5 is shown in Figure S3 and was identified as PE(38:4) based on the product ions at m/z 140, 283 and 303 representing the ethanolamine head group, and the fatty acids C18:0 and C20:4, respectively. As it is a complete loss of PE’s, and not a shift in the retention times, it is not possible to explain this phenomenon by derivatization alone. Results from this study also demonstrated a reduction in PS species, and not a total loss, as can be observed in the mass spectra presented in Figure S4. Data obtained from the MSE studies correlate with data obtained from the MSI investigations, as will be presented in the forthcoming sections. The loss of PE and the reduction in PS lipid classes is thought to be due to the symmetry of these lipids in the plasma membrane and the physical mechanisms by which formalin fixes tissue. Formalin action is by diffusion into tissue samples and as PS lipids are predominantly located on the inner leaflet of the plasma membrane they will be fixed at a slower rate compared to the PE lipids on the outer leaflet of the membrane.26 The lung is porous and thus fixes more rapidly than a solid organ, which is why overnight fixation was sufficient, if this was prolonged to 48 h we believed we would see a complete loss in PS species following fixation. We also believe the increase in fatty acids, diacylglycerides and lysoPLs presented by Gaudin et al24 is predominantly a result of degradation by enzymes and natural tissue processes before the tissue is completely fixed. Formalin is reported to penetrate tissue at 1 mm/h for solid organs, and fixation time depends on the size or thickness of the tissue sample, in addition to the temperature of the fixative solution and the ratio of formalin-totissue.27 We hypothesize the aminophospholipids, PE and PS are readily crosslinked with proteins and metabolites through their primary amine group in the same manner that is reported

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

for proteins27 and that an optimized antigen retrieval method will enable the detection of these lipids in FF and formalin fixed paraffin embedded (FFPE) tissue by MSI. Detection and Distribution of Lipids in the Positive Ion Mode. The detection and distribution of lipids species covering all anatomies of the larger mammalian lungs has not been presented before, the results presented here are those of multiple lung sections, with images representative of regions covering the conducting (bronchi, as evident by the presence of cartilage) and respiratory airways (alveolar space). The positive ion mode analysis of lipids species from fixed lung sections resulted in spectra that consisted predominantly of phosphatidylcholine (PC) and sphingomyelin (SM) species as the [M+H]+ and [M+Na]+ ions, with suppression of the [M+K]+ ions that are routinely observed during MSI investigations of fresh tissue sections. Data obtained thus agrees with those previously reported for the positive ion mode analysis of lipids from fixed brain tissue.28 Over 28 lipid species were identified based on MS/MS data or mass, following internal calibration of MS datasets. Ions that could not be tentatively assigned were excluded from the results. As PE species were not detected they were eliminated from the database search reducing the number of possible lipid assignments. The most abundant ion detected in positive ion mode from lung tissue was m/z 756.6, followed by m/z 734.6 representing the [M+Na]+ and [M+H]+ of PC(32:0), the predominant PC in lung surfactant, and previously reported as the species detected with highest intensity from fresh sections of mouse lung tissue.20 Figure 3 demonstrates the main anatomical-based molecular profiles obtainable from high resolution MALDI-MSI of lung sections. The ions chosen are representative of the base peak for each region of lung tissue as an example of lipid distribution within lung sections. The ion at m/z 616.2 identified as heme B highlights the blood that is still present within the blood vessel. The ion at m/z 725.6 was identified as the sodium adduct of SM(d18:1/16:0) based on MS/MS analysis (Table S1), and was detected with highest intensity in the connective tissue and smooth muscle regions of the blood vessels and bronchi, with lower intensity signal in the respiratory airways, agreeing with the distribution of this lipid in mouse lung.20 The MALDI-MS image of m/z 780.6 presented in Figure 3 displayed increased intensity in the bronchial epithelium followed by the respiratory airways with little-to-no signal being detected in the connective tissue and smooth muscle regions of the blood vessels and bronchi. Two isobaric species are detected at this mass, both PC(34:2) and PC(36:5) as the [M+Na]+ and [M+H]+ ions respectively, are believed to contribute to the signal detected at m/z 780.6. The last single ion image presented in Figure 3 is of m/z 782.6 identified as the [M+Na]+ ion of PC(34:1) based on MS/MS analysis (Table S1), and was detected with highest intensity in the bronchial epithelium, respiratory airways, the bronchiassociated lymphoid tissue (BALT), and showed little-to-no signal in the regions highlighting the pulmonary vessels. The bronchi lumen (BL) is labeled in the MS images to enable the reader ease of orientation upon comparison to the labeled H&E section presented at the bottom left of Figure 3. The single point spectra presented above each MS image were taken from anatomical regions highlighted by the selected ions and demonstrate the differences in the base peak and relative intensity of lipid species across the different lung anatomies. The spectrum for the image of m/z 616 was taken from the center of the vessel, m/z 725 was taken from the top right of the blood vessel wall as labeled in the histology image, m/z 780 was taken from the highest intensity region of the respiratory airways, as presented in the Figure, and m/z 782 was taken from the bronchial epithelium region. A merged ion image of m/z 616, 725, 780 and 782 is presented, clearly highlighting the different anatomies, as can be seen on comparison to the labeled H&E section presented at the bottom of Figure 3. A full list of lipid species identified during positive ion mode analysis, their diagnostic product ions following tandem MS, and their distribution within lung tissue is presented in Table S1.

ACS Paragon Plus Environment

Page 6 of 14

Page 7 of 14

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

A key aspect of analyzing lipids from FF tissue in positive ion mode is the reduction in spectral complexity due to suppression in the formation of potassium adducts. Of particular interest is the detection of the ion at m/z 742.6, as this ion would usually be masked under the higher abundant C13 isotope of the [M+K]+ ion of SM(d18:1/16:0), during the analysis of fresh tissue sections. The distribution of m/z 742.6 was mapped with highest intensity to the bronchiassociated lymphoid tissue (BALT), a cluster of immune cells that form follicles around the bronchi, followed by lower intensity signal in regions of the respiratory airways, as demonstrated by Figure S5. Tandem MS investigation of the ion at m/z 742.6 resulted in diagnostic fragment ions at m/z 683 and 147 corresponding to a neutral loss of trimethylamine, and the sodium adduct of the cyclic 1,2-phosphodiester ion, characteristic ions following the dissociation of a sodium adduct of PCs. A database search of the [M+Na]+ ion of m/z 742.6 identified this species as PC(O-32:0). Detection and identification of this lipid from lung tissue is of significant importance as ether lipids such as the, 1-alkyl, 2-acetylglycerophosdphocholine identified here, are remodeled via hydrolysis of the acyl chain during the formation of Platelet Activating Factor (PAF) and thus plays important roles in signaling and inflammation. It should be noted that the [M+H]+ ion of PC(O-32:0) was not detected. Detection and Distribution of Lipids in the Negative Ion Mode. Analysis of fixed-inflated lung tissue in the negative ion mode resulted in the detection of over 100 ions, half of which were detected with high abundance but low frequency, meaning they were highly localized. The MALDI-MS images presented in Figure 4 were again selected based on their differential distribution and thus highlight the different anatomical features of the lung. The ion at m/z 631.2 was detected with high abundance in the blood accumulated within the large vessel with low signal intensity in the parenchyma.Fragmentation of this ion did not provided any diagnostic lipid peaks, thus this ion is believed to be a result from metastable decay of a larger glycolipid anchored to erythrocytes, further investigation of this ion is out of the scope of this manuscript. The ion at m/z 764.5 was detected with highest abundance in the smooth muscle and connective tissue regions surrounding the blood vessels and bronchi, this ion was tentatively assigned as PS(34:0) based on a database search combined with the exclusion of PEs. The cartilage tissue that supports bronchi in larger mammalian airways is highlighted by the ion detected at m/z 794.5, which was highly localized to this region, as shown in Figure 4. This ion was identified as C16(OH) sulfatide, based on characteristic fragmentation ions representative of sulfate, the galactose 3-sulfate moiety, and the galactose 3-sulfate moiety with a loss of water, at m/z 97, 241 and 259, respectively. This data also agrees with those previously published by our group for the detection of this species in cartilage.5 The ion at m/z 869.6 shows higher signal intensity within the respiratory airways, with increased accumulation in one region, as can be seen by comparing the MS image to the labeled H&E section in the bottom left of Figure 4. Dissociation of this ion resulted in the detection of diagnostic fragment ions for PIs at m/z 241 and 223 representing the inositol phosphate ion and the inositol phosphate ion minus water, respectively. A number of product ions from the fatty acid side chains were also detected, at m/z 281 (C18:1), 283 (C18:0) and 303 (C20:4), resulting from dissociation of the isobaric phosphatidylinositol (PI) species at m/z 869.5, identified as PI(O38:5) and PI(P-38:4) from a database search. Again the decrease in complexity due to the absence of PEs, in addition to the increase in sensitivity due to their being less analytes in the gas phase, enabled the detection of a number of PIs, phosphatidylglycerol (PG) species, PSs and HexCers that were not previously detected in fresh lung sections.5 Table S2 lists lipid species detected and their tentative assignments, based on database searches and/or MS/MS analysis, along with their distribution within lung tissue. The final image in Figure 4 displays the distribution of the ion at m/z 890.6, which was detected with high abundance but low frequency, as can be seen by the highly localized nature of this ion within the image presented. Over half of

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

the ions detected in negative ion mode were detected within these highly localized regions; this is demonstrated by the single point spectra taken from each region presented above each panel in Figure 4, and Tables S2 and S3. Comparison to histology identified these regions as bundles of nerve fibers running through the smooth muscle of the bronchi, believed to be part of the sympathetic nervous system. Fragmentation investigations for identification of these ions proved difficult due to the limited amount of material available. Three different species were investigated over several sections before the material was ablated fully and no signal was detected. The ions at m/z 876.6, 888.6 and 906.6 were identified as, C22:1(OH) sulfatide, C24:1 sulfatide, and C16(OH) sulfatide, respectively. Identification was based on detection of the fragment ions at m/z 97 ([HSO4]-), and m/z 241 and 259 characteristic ions of the galactose 3sulfate moiety with a loss of water and the galactose 3-sulfate moiety, respectively. There is a modest comparative MS and MSI database for lipidomic investigations of the peripheral nervous system (PNS), however, data reported here does agree, in part, with those studies published for the central nervous system (CNS).29-33 A full list of ions detected from the sympathetic nervous tissue surrounding the bronchi, along with tentative assignments for each ion, is presented in Table S3. Conclusion Formalin inflation of lung tissue is a viable and optimized methodology for lipidomic analysis in both positive and negative ionization modes. Inflation with formalin offers the ability to link directly with clinical samples as this is the standard processing protocol used. Sensitivity and data quality was greatly improved due to the reduced complexity in the gas phase; this enabled high resolution imaging experiments to be carried out without any significant loss in signal. Combined UPLC-MSE and MSI experiments indicate that aminophospholipids, PS and PE species, are crosslinked into the tissue by the mechanism of formalin fixation, in a timedependent manner. Whilst these lipids are reduced or absent in the spectra obtained from fixed tissue, respectively, it is believed that analysis will be possible following optimized antigen retrieval. This also opens the possibility of analyzing these lipids from formalin-fixed paraffinembedded tissue, which would not be realized before, as they would have been presumed lost during the processing steps like other lipid species. We have provided a novel protocol for lung inflation and optimized MSI acquisition parameters, with a guideline of a minimum of 20 µm, for clearly defined images of the lung parenchyma. In addition to providing, for the first time, a map of over 100 lipid species and their distribution within larger mammalian lungs. As no embedding media was used to inflate we did experience some tissue loss due to the thin and porous nature of the lung, thus future work will develop a combination fixation-inflation-embedding protocol that will better support the lung during sectioning and processing. Acknowledgements This project has been funded with Federal funds from the National Institute of Allergy and Infectious Diseases, National Institutes of Health, Department of Health and Human Services, under Contract No. HHSN272201000046C, and from the Biomedical Advanced Research and Development Authority, Office of the Assistant Secretary for Preparedness and Response, Office of the Secretary, Department of Health and Human Services, under subcontract to Contract No: HHSO100201100007C, Aeolus Pharmaceuticals, Inc. (Mission Viejo, CA). Additional support was provided by the University of Maryland School of Pharmacy Mass Spectrometry Center (SOP1841-IQB2014). The authors would like to thank Shannon Cornett for his time, and for always being on hand to answer any question, we appreciate you.

ACS Paragon Plus Environment

Page 8 of 14

Page 9 of 14

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

References (1) Murray, C. J. L.; Lopez, A. D. The Lancet 1997, 349, 1498-1504. (2) King Jr, T. E.; Pardo, A.; Selman, M. The Lancet 2011, 378, 1949-1961. (3) Usul Afsar, C.; Sahin, B.; Gunaldi, M.; Kilic Bagir, E.; Gumurdulu, D.; Burgut, R.; Erkisi, M.; Kara, I. O.; Paydas, S.; Karaca, F.; Ercolak, V. Int. J. Clin. Exp. Pathol. 2015, 8, 9760-9771. (4) Noël-Georis, I.; Bernard, A.; Falmagne, P.; Wattiez, R. J. Chromatogr. B 2002, 771, 221236. (5) Carter, C. L.; Jones, J. W.; Barrow, K.; Kieta, K.; Taylor-Howell, C.; Kearney, S.; Smith, C. P.; Gibbs, A.; Farese, A. M.; MacVittie, T. J.; Kane, M. A. Health Phys. 2015, 109, 466-478. (6) El Ayed, M.; Bonnel, D.; Longuespee, R.; Castelier, C.; Franck, J.; Vergara, D.; Desmons, A.; Tasiemski, A.; Kenani, A.; Vinatier, D.; Day, R.; Fournier, I.; Salzet, M. Med. Sci. Monit. 2010, 16, Br233-245. (7) Jackson, S. N.; Baldwin, K.; Muller, L.; Womack, V. M.; Schultz, J. A.; Balaban, C.; Woods, A. S. Anal. Bioanal. Chem. 2014, 406, 1377-1386. (8) Jones, E. E.; Dworski, S.; Kamani, M.; Medin, J. A.; Drake, R. R. Mol. Genet. Metab. 2014, 111, S56-S57. (9) Fehniger, T. E.; Vegvari, A.; Rezeli, M.; Prikk, K.; Ross, P.; Dahlback, M.; Edula, G.; Sepper, R.; Marko-Varga, G. Anal. Chem. 2011, 83, 8329-8336. (10) Prideaux, B.; Dartois, V.; Staab, D.; Weiner, D. M.; Goh, A.; Via, L. E.; Barry, C. E., 3rd; Stoeckli, M. Anal. Chem. 2011, 83, 2112-2118. (11) Goodwin, R. J. A.; Nilsson, A.; Borg, D.; Langridge-Smith, P. R. R.; Harrison, D. J.; Mackay, C. L.; Iverson, S. L.; Andren, P. E. J. Proteomics 2012, 75, 4912-4920. (12) Prideaux, B.; ElNaggar, M. S.; Zimmerman, M.; Wiseman, J. M.; Li, X.; Dartois, V. Int. J. Mass Spectrom. 2015, 377, 699-708. (13) Marko-Varga, G.; Vegvari, A.; Rezeli, M.; Prikk, K.; Ross, P.; Dahlback, M.; Edula, G.; Sepper, R.; Fehniger, T. E. Clin. Transl. Med. 2012, 1, 8. (14) Vegvari, A.; Fehniger, T. E.; Rezeli, M.; Laurell, T.; Dome, B.; Jansson, B.; Welinder, C.; Marko-Varga, G. J. Proteome Res. 2013, 12, 5626-5633. (15) Nilsson, A.; Fehniger, T. E.; Gustavsson, L.; Andersson, M.; Kenne, K.; Marko-Varga, G.; Andrén, P. E. PLoS ONE 2010, 5, e11411. (16) Churg, A. Am. J. Surg. Pathol. 1983, 7, 69-71. (17) Myung, J. K.; Choe, G.; Chung, D. H.; Seo, J. W.; Jheon, S.; Lee, C. T.; Chung, J. H. Lung Cancer 2008, 59, 198-202. (18) Hayashi, T.; Nagayasu, T.; Kohno, S.; Abe, K.; Tamaru, N.; Anami, M.; Sakai, Y.; Ohbayashi, C. Pathology 2005, 37, 355-359. (19) Zhao, C.; O’Connor, P. B. Anal. Biochem. 2007, 365, 283-285. (20) Berry, K. A. Z.; Li, B.; Reynolds, S. D.; Barkley, R. M.; Gijón, M. A.; Hankin, J. A.; Henson, P. M.; Murphy, R. C. J. Lipid Res. 2011, 52, 1551-1560. (21) Strohalm, M.; Strohalm, J.; Kaftan, F.; Krasny, L.; Volny, M.; Novak, P.; Ulbrich, K.; Havlicek, V. Anal. Chem. 2011, 83, 5458-5462. (22) Nelson, K. A.; Daniels, G. J.; Fournie, J. W.; Hemmer, M. J. J. Biomol. Tech. 2013, 24, 119127. (23) Garofalo, M. C.; Ward, A. A.; Farese, A. M.; Bennett, A.; Taylor-Howell, C.; Cui, W.; Gibbs, A.; Prado, K. L.; MacVittie, T. J. Health Phys. 2014, 106, 73-83. (24) Gaudin, M.; Panchal, M.; Ayciriex, S.; Werner, E.; Brunelle, A.; Touboul, D.; BoursierNeyret, C.; Auzeil, N.; Walther, B.; Duyckaerts, C.; Laprevote, O. J. Mass Spectrom. 2014, 49, 1035-1042.

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 14

(25) Fuchs, B.; Schober, C.; Richter, G.; Suss, R.; Schiller, J. J. Biochem. Biophys. Methods 2007, 70, 689-692. (26) Vance, J. E.; Tasseva, G. Biochim. Biophys. Acta 2013, 1831, 543-554. (27) Thavarajah, R.; Mudimbaimannar, V. K.; Elizabeth, J.; Rao, U. K.; Ranganathan, K. Int. J. Oral Maxillofac. Pathol. 2012, 16, 400-405. (28) Carter, C. L.; McLeod, C. W.; Bunch, J. J. Am. Soc. Mass Spectrom. 2011, 22, 1991-1998. (29) Wang, H. Y.; Jackson, S. N.; Post, J.; Woods, A. S. Int. J. Mass Spectrom. 2008, 278, 143149. (30) Kettling, H.; Vens-Cappell, S.; Soltwisch, J.; Pirkl, A.; Haier, J.; Muething, J.; Dreisewerd, K. Anal. Chem. 2014, 86, 7798-7805. (31) Goto-Inoue, N.; Hayasaka, T.; Zaima, N.; Kashiwagi, Y.; Yamamoto, M.; Nakamoto, M.; Setou, M. J. Am. Soc. Mass Spectrom. 2010, 21, 1940-1943. (32) Cerruti, C. D.; Benabdellah, F.; Laprevote, O.; Touboul, D.; Brunelle, A. Anal. Chem. 2012, 84, 2164-2171. (33) Cheng, H.; Sun, G.; Yang, K.; Gross, R. W.; Han, X. J. Lipid Res. 2010, 51, 1599-1609.

Figure 1. Positive ion mode MALDI-MS images of m/z 835.7 and stained H&E sections of the same lung biopsy fresh-frozen (A-C) and formalin-inflated (D-F). The MSI figure from the fresh frozen lung appears as a solid organ, showing no real definition of the anatomy (figure A), which agrees with its corresponding histology section shown in figure B and a zoomed in region from the H&E section in figure C, showing complete collapse of the alveolar space. Figures D-F show the same lung biopsy following inflation with formalin. The MSI figure is taken from a region of the lung section presented in figure E, the alveolar space is recognizable and the alveolar epithelium a can thus be defined. Inflation enables visualization of the alveolar space and thus the respiratory airways, the epithelium is not collapsed as can be seen by comparing the zoomed in regions of figure F with that of figure C. AD = alveolar duct, AS = alveolar space. Figure 2. Positive ion mode MALDI-MS images of inflated lung biopsy sections acquired at different resolutions from 100 - 10 µm, figures A-E, and their representative H&E images. Clear defined images of the respiratory airways can be observed at 20 µm and below. Whist all images are presented as percent relative intensity the absolute ion (AI) intensity for the base peak is given next to each image. Fleximaging scales intensity by laser shots, meaning the values presented are for a single laser shot, which is why the number is low in comparison to the summed laser shots more commonly reported in flexanalysis and other imaging software. Figure 3. Positive ion mode MALDI-MS images of heme at m/z 616.2, SM(d18:1/16:0) at m/z 725.6, PC(34:2)/PC(36:5) at m/z 780.6 and PC(34:1) at m/z 782.6. Single point spectra are presented above each image and were taken from the blood, the blood vessel, high intensity parenchymal region, and the bronchi epithelium, from left to right respectively. The bronchi lumen (BL) is labeled in each MSI images. The labeled H&E stained section presented is marked as follows; BV=blood vessel; RA=respiratory airways; C=cartilage; CT=connective tissue; EP=bronchi epithelium; BL=bronchi lumen, L=lymphoid tissue and SM=smooth muscle. A merged ion image is displayed to the bottom right of the figure along with the intensity bars for each ion. Again images are presented as percent relative intensity but the absolute ion (AI) intensity for the base peak, scaled by laser shot, is given next to the merged ion image. Figure 4. Negative ion mode MALDI-MS images of an unassigned ion at m/z 631.2, PS(34:0) at m/z 765.5, C16(OH) sulfatide at m/z 794.5, PI(O-38:5)/PI(P-38:4) at m/z 869.6, and C24

ACS Paragon Plus Environment

Page 11 of 14

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

sulfatide at m/z 890.6. Single point spectra are presented above each image taken from the blood, the blood vessel, cartilage, high intensity parenchymal region, and peripheral nervous tissue, from left to right respectively. The bronchi lumen (BL) is labeled in each MSI images. The labeled H&E stained section presented is marked as follows; BV=blood vessel; RA=respiratory airways; C=cartilage; CT=connective tissue; EP=bronchi epithelium; BL=bronchi lumen, L=lymphoid tissue and SM=smooth muscle. A merged ion image is displayed to the bottom right of the figure along with the intensity bars for each ion. Absolute ion (AI) intensity is scaled by laser shot.

Figure 1

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 2

ACS Paragon Plus Environment

Page 12 of 14

Page 13 of 14

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Figure 3

Figure 4.

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

245x177mm (150 x 150 DPI)

ACS Paragon Plus Environment

Page 14 of 14