Maintenance of Deep Lung Architecture and Automated Airway Segmentation for 3D Mass Spectrometry Imaging

Alison J Scott,Courtney E Chandler,Shane R Ellis,Robert K Ernst,Ron M A Heeren

doi:10.1038/s41598-019-56364-4

Alison J Scott, Courtney E Chandler + Show 3 more

Open Access

https://doi.org/10.1038/s41598-019-56364-4

Copy DOI

Abstract

Mass spectrometry imaging (MSI) is a technique for mapping the spatial distributions of molecules in sectioned tissue. Histology-preserving tissue preparation methods are central to successful MSI studies. Common fixation methods, used to preserve tissue morphology, can result in artifacts in the resulting MSI experiment including delocalization of analytes, altered adduct profiles, and loss of key analytes due to irreversible cross-linking and diffusion. This is especially troublesome in lung and airway samples, in which histology and morphology is best interpreted from 3D reconstruction, requiring the large and small airways to remain inflated during analysis. Here, we developed an MSI-compatible inflation containing as few exogenous components as possible, forgoing perfusion, fixation, and addition of salt solutions upon inflation that resulted in an ungapped 3D molecular reconstruction through more than 300 microns. We characterized a series of polyunsaturated phospholipids (PUFA-PLs), specifically phosphatidylinositol (-PI) lipids linked to lethal inflammation in bacterial infection and mapped them in serial sections of inflated mouse lung. PUFA-PIs were identified using spatial lipidomics and determined to be determinant markers of major airway features using unsupervised hierarchical clustering. Deep lung architecture was preserved using this inflation approach and the resulting sections are compatible with multiple MSI modalities, automated interpretation software, and serial 3D reconstruction.

Highlights

Since the underlying sample in an Mass spectrometry imaging (MSI) study is a thin tissue section, preservation of histological structures during sample preparation is paramount to accurate data interpretation
For samples intended for a range of readouts from basic histological stains to advanced micro-X-ray CT analysis, lung tissue can be first preserved by formalin fixation followed by paraffin-embedding (FFPE) and sectioning before analysis[22]
Using MALDI-FTICR, Jones, et al inflated mouse lungs using a solution of carboxymethylcellulose (CMC) resulting in the creation of a high mass resolution, tissue-spanning 3D reconstruction of a lung with voxel dimensions of 120 × 120 × 120 μm, demonstrating that 3D reconstruction of gapped serial sections is possible in this delicate tissue[26]

Summary

Introduction

Since the underlying sample in an MSI study is a thin tissue section (typically 10–15 μm thickness), preservation of histological structures during sample preparation is paramount to accurate data interpretation. Using modified OCT (mOCT) compound, Zemiki-Berry, et al reported a MSI-compatible rodent lung inflation method that identified major classes of lipids in both the positive and negative ion modes associated with histological features (large airway, vasculature, etc.) in single tissue sections[25]. We developed and demonstrated an alternative inflation medium (porcine gelatin) that preserves lung ultrastructure and is compatible with advanced MSI techniques including three-dimensional reconstruction of fine-scale consecutive serial sections, spatial segmentation analysis, and lipid profiling using parallelized lipidomic imaging. Preservation of histology in lung tissues for MSI experimentation is a priority for distribution studies of inhaled drug products, lung infections, chronic lung dysfunctions, and cancers These advances in preparation techniques, as well as in histological interpretation using unique lipid signatures, contribute to the ongoing development of a spatially-resolved mouse lung lipidome

Objectives

Methods

Results

Discussion

Conclusion