AMASS: algorithm for MSI analysis by semi-supervised segmentation

Computational Analysis of Alzheimer Amyloid Plaque Composition in 2D- and Elastically Reconstructed 3D-MALDI MS Images

MALDI mass spectrometry imaging (MSI) enables label-free, spatially resolved analysis of a wide range of analytes in tissue sections. Quantitative analysis of MSI datasets is typically performed on single pixels or manually assigned regions of interest (ROIs). However, many sparse, small objects such as Alzheimer's disease (AD) brain deposits of amyloid peptides called plaques are neither single pixels nor ROIs. Here, we propose a new approach to facilitate the comparative computational evaluation of amyloid plaque-like objects by MSI: a fast PLAQUE PICKER tool that enables a statistical evaluation of heterogeneous amyloid peptide composition. Comparing two AD mouse models, APP NL-G-F and APP PS1, we identified distinct heterogeneous plaque populations in the NL-G-F model but only one class of plaques in the PS1 model. We propose quantitative metrics for the comparison of technical and biological MSI replicates. Furthermore, we reconstructed a high-accuracy 3D-model of amyloid plaques in a fully automated fashion, employing rigid and elastic MSI image registration using structured and plaque-unrelated reference ion images. Statistical single-plaque analysis in reconstructed 3D-MSI objects revealed the Aβ_1-42Arc peptide to be located either in the core of larger plaques or in small plaques without colocalization of other Aβ isoforms. In 3D, a substantially larger number of small plaques were observed than that indicated by the 2D-MSI data, suggesting that quantitative analysis of molecularly diverse sparsely-distributed features may benefit from 3D-reconstruction. Data are available via ProteomeXchange with identifier PXD020824.

Unsupervised machine learning for exploratory data analysis in imaging mass spectrometry

Imaging mass spectrometry (IMS) is a rapidly advancing molecular imaging modality that can map the spatial distribution of molecules with high chemical specificity. IMS does not require prior tagging of molecular targets and is able to measure a large number of ions concurrently in a single experiment. While this makes it particularly suited for exploratory analysis, the large amount and high-dimensional nature of data generated by IMS techniques make automated computational analysis indispensable. Research into computational methods for IMS data has touched upon different aspects, including spectral preprocessing, data formats, dimensionality reduction, spatial registration, sample classification, differential analysis between IMS experiments, and data-driven fusion methods to extract patterns corroborated by both IMS and other imaging modalities. In this work, we review unsupervised machine learning methods for exploratory analysis of IMS data, with particular focus on (a) factorization, (b) clustering, and (c) manifold learning. To provide a view across the various IMS modalities, we have attempted to include examples from a range of approaches including matrix assisted laser desorption/ionization, desorption electrospray ionization, and secondary ion mass spectrometry-based IMS. This review aims to be an entry point for both (i) analytical chemists and mass spectrometry experts who want to explore computational techniques; and (ii) computer scientists and data mining specialists who want to enter the IMS field. © 2019 The Authors. Mass Spectrometry Reviews published by Wiley Periodicals, Inc. Mass SpecRev 00:1-47, 2019.

rMSIKeyIon: An Ion Filtering R Package for Untargeted Analysis of Metabolomic LDI-MS Images

Many MALDI-MS imaging experiments make a case versus control studies of different tissue regions in order to highlight significant compounds affected by the variables of study. This is a challenge because the tissue samples to be compared come from different biological entities, and therefore they exhibit high variability. Moreover, the statistical tests available cannot properly compare ion concentrations in two regions of interest (ROIs) within or between images. The high correlation between the ion concentrations due to the existence of different morphological regions in the tissue means that the common statistical tests used in metabolomics experiments cannot be applied. Another difficulty with the reliability of statistical tests is the elevated number of undetected MS ions in a high percentage of pixels. In this study, we report a procedure for discovering the most important ions in the comparison of a pair of ROIs within or between tissue sections. These ROIs were identified by an unsupervised segmentation process, using the popular k-means algorithm. Our ion filtering algorithm aims to find the up or down-regulated ions between two ROIs by using a combination of three parameters: (a) the percentage of pixels in which a particular ion is not detected, (b) the Mann–Whitney U ion concentration test, and (c) the ion concentration fold-change. The undetected MS signals (null peaks) are discarded from the histogram before the calculation of (b) and (c) parameters. With this methodology, we found the important ions between the different segments of a mouse brain tissue sagittal section and determined some lipid compounds (mainly triacylglycerols and phosphatidylcholines) in the liver of mice exposed to thirdhand smoke.

Characterization of microbial mixtures by mass spectrometry

MS applications in microbiology have increased significantly in the past 10 years, due in part to the proliferation of regulator-approved commercial MALDI MS platforms for rapid identification of clinical infections. In parallel, with the expansion of MS technologies in the "omics" fields, novel MS-based research efforts to characterize organismal as well as environmental microbiomes have emerged. Successful characterization of microorganisms found in complex mixtures of other organisms remains a major challenge for researchers and clinicians alike. Here, we review recent MS advances toward addressing that challenge. These include sample preparation methods and protocols, and established, for example, MALDI, as well as newer, for example, atmospheric pressure ionization (API) techniques. MALDI mass spectra of intact cells contain predominantly information on the highly expressed house-keeping proteins used as biomarkers. The API methods are applicable for small biomolecule analysis, for example, phospholipids and lipopeptides, and facilitate species differentiation. MS hardware and techniques, for example, tandem MS, including diverse ion source/mass analyzer combinations are discussed. Relevant examples for microbial mixture characterization utilizing these combinations are provided. Chemometrics and bioinformatics methods and algorithms, including those applied to large scale MS data acquisition in microbial metaproteomics and MS imaging of biofilms, are highlighted. Select MS applications for polymicrobial culture analysis in environmental and clinical microbiology are reviewed as well.

Signal preprocessing, multivariate analysis and software tools for MA(LDI)-TOF mass spectrometry imaging for biological applications

Mass spectrometry imaging (MSI) is a label-free analytical technique capable of molecularly characterizing biological samples, including tissues and cell lines. The constant development of analytical instrumentation and strategies over the previous decade makes MSI a key tool in clinical research. Nevertheless, most MSI studies are limited to targeted analysis or the mere visualization of a few molecular species (proteins, peptides, metabolites, or lipids) in a region of interest without fully exploiting the possibilities inherent in the MSI technique, such as tissue classification and segmentation or the identification of relevant biomarkers from an untargeted approach. MSI data processing is challenging due to several factors. The large volume of mass spectra involved in a MSI experiment makes choosing the correct computational strategies critical. Furthermore, pixel to pixel variation inherent in the technique makes choosing the correct preprocessing steps critical. The primary aim of this review was to provide an overview of the data-processing steps and tools that can be applied to an MSI experiment, from preprocessing the raw data to the more advanced strategies for image visualization and segmentation. This review is particularly aimed at researchers performing MSI experiments and who are interested in incorporating new data-processing features, improving their computational strategy, and/or desire access to data-processing tools currently available. © 2016 Wiley Periodicals, Inc. Mass Spec Rev 37:281-306, 2018.

Spatially-Resolved Top-down Proteomics Bridged to MALDI MS Imaging Reveals the Molecular Physiome of Brain Regions

Tissue spatially-resolved proteomics was performed on 3 brain regions, leading to the characterization of 123 reference proteins. Moreover, 8 alternative proteins from alternative open reading frames (AltORF) were identified. Some proteins display specific post-translational modification profiles or truncation linked to the brain regions and their functions. Systems biology analysis performed on the proteome identified in each region allowed to associate sub-networks with the functional physiology of each brain region. Back correlation of the proteins identified by spatially-resolved proteomics at a given tissue localization with the MALDI MS imaging data, was then performed. As an example, mapping of the distribution of the matrix metallopeptidase 3-cleaved C-terminal fragment of α-synuclein (aa 95–140) identified its specific distribution along the hippocampal dentate gyrus. Taken together, we established the molecular physiome of 3 rat brain regions through reference and hidden proteome characterization.

Combined Mass Spectrometry Imaging and Top-down Microproteomics Reveals Evidence of a Hidden Proteome in Ovarian Cancer

BACKGROUND

Recently, it was demonstrated that proteins can be translated from alternative open reading frames (altORFs), increasing the size of the actual proteome. Top-down mass spectrometry-based proteomics allows the identification of intact proteins containing post-translational modifications (PTMs) as well as truncated forms translated from reference ORFs or altORFs.

METHODS

Top-down tissue microproteomics was applied on benign, tumor and necrotic-fibrotic regions of serous ovarian cancer biopsies, identifying proteins exhibiting region-specific cellular localization and PTMs. The regions of interest (ROIs) were determined by MALDI mass spectrometry imaging and spatial segmentation.

FINDINGS

Analysis with a customized protein sequence database containing reference and alternative proteins (altprots) identified 15 altprots, including alternative G protein nucleolar 1 (AltGNL1) found in the tumor, and translated from an altORF nested within the GNL1 canonical coding sequence. Co-expression of GNL1 and altGNL1 was validated by transfection in HEK293 and HeLa cells with an expression plasmid containing a GNL1-FLAG(V5) construct. Western blot and immunofluorescence experiments confirmed constitutive co-expression of altGNL1-V5 with GNL1-FLAG.

CONCLUSIONS

Taken together, our approach provides means to evaluate protein changes in the case of serous ovarian cancer, allowing the detection of potential markers that have never been considered.

Mass Spectrometry Imaging for the Investigation of Intratumor Heterogeneity

One of the big clinical challenges in the treatment of cancer is the different behavior of cancer patients under guideline therapy. An important determinant for this phenomenon has been identified as inter- and intratumor heterogeneity. While intertumor heterogeneity refers to the differences in cancer characteristics between patients, intratumor heterogeneity refers to the clonal and nongenetic molecular diversity within a patient. The deciphering of intratumor heterogeneity is recognized as key to the development of novel therapeutics or treatment regimens. The investigation of intratumor heterogeneity is challenging since it requires an untargeted molecular analysis technique that accounts for the spatial and temporal dynamics of the tumor. So far, next-generation sequencing has contributed most to the understanding of clonal evolution within a cancer patient. However, it falls short in accounting for the spatial dimension. Mass spectrometry imaging (MSI) is a powerful tool for the untargeted but spatially resolved molecular analysis of biological tissues such as solid tumors. As it provides multidimensional datasets by the parallel acquisition of hundreds of mass channels, multivariate data analysis methods can be applied for the automated annotation of tissues. Moreover, it integrates the histology of the sample, which enables studying the molecular information in a histopathological context. This chapter will illustrate how MSI in combination with statistical methods and histology has been used for the description and discovery of intratumor heterogeneity in different cancers. This will give evidence that MSI constitutes a unique tool for the investigation of intratumor heterogeneity, and could hence become a key technology in cancer research.

Whole Reproductive System Non-Negative Matrix Factorization Mass Spectrometry Imaging of an Early-Stage Ovarian Cancer Mouse Model

High-grade serous carcinoma (HGSC) is the most common and deadliest form of ovarian cancer. Yet it is largely asymptomatic in its initial stages. Studying the origin and early progression of this disease is thus critical in identifying markers for early detection and screening purposes. Tissue-based mass spectrometry imaging (MSI) can be employed as an unbiased way of examining localized metabolic changes between healthy and cancerous tissue directly, at the onset of disease. In this study, we describe MSI results from Dicer-Pten double-knockout (DKO) mice, a mouse model faithfully reproducing the clinical nature of human HGSC. By using non-negative matrix factorization (NMF) for the unsupervised analysis of desorption electrospray ionization (DESI) datasets, tissue regions are segregated based on spectral components in an unbiased manner, with alterations related to HGSC highlighted. Results obtained by combining NMF with DESI-MSI revealed several metabolic species elevated in the tumor tissue and/or surrounding blood-filled cyst including ceramides, sphingomyelins, bilirubin, cholesterol sulfate, and various lysophospholipids. Multiple metabolites identified within the imaging study were also detected at altered levels within serum in a previous metabolomic study of the same mouse model. As an example workflow, features identified in this study were used to build an oPLS-DA model capable of discriminating between DKO mice with early-stage tumors and controls with up to 88% accuracy.

Toward the complete characterization of host cell proteins in biotherapeutics via affinity depletions, LC-MS/MS, and multivariate analysis

Host cell protein (HCP) impurities are generated by the host organism during the production of therapeutic recombinant proteins, and are difficult to remove completely. Though commonly present in small quantities, if levels are not controlled, HCPs can potentially reduce drug efficacy and cause adverse patient reactions. A high resolution approach for thorough HCP characterization of therapeutic monoclonal antibodies is presented herein. In this method, antibody samples are first depleted via affinity enrichment (e.g., Protein A, Protein L) using milligram quantities of material. The HCP-containing flow-through is then enzymatically digested, analyzed using nano-UPLC-MS/MS, and proteins are identified through database searching. Nearly 700 HCPs were identified from samples with very low total HCP levels (< 1 ppm to ∼ 10 ppm) using this method. Quantitation of individual HCPs was performed using normalized spectral counting as the number of peptide spectrum matches (PSMs) per protein is proportional to protein abundance. Multivariate analysis tools were utilized to assess similarities between HCP profiles by: 1) quantifying overlaps between HCP identities; and 2) comparing correlations between individual protein abundances as calculated by spectral counts. Clustering analysis using these measures of dissimilarity between HCP profiles enabled high resolution differentiation of commercial grade monoclonal antibody samples generated from different cell lines, cell culture, and purification processes.

MALDI-Mass Spectrometric Imaging Revealing Hypoxia-Driven Lipids and Proteins in a Breast Tumor Model

Hypoxic areas are a common feature of rapidly growing malignant tumors and their metastases and are typically spatially heterogeneous. Hypoxia has a strong impact on tumor cell biology and contributes to tumor progression in multiple ways. To date, only a few molecular key players in tumor hypoxia, such as hypoxia-inducible factor-1 (HIF-1), have been discovered. The distribution of biomolecules is frequently heterogeneous in the tumor volume and may be driven by hypoxia and HIF-1α. Understanding the spatially heterogeneous hypoxic response of tumors is critical. Mass spectrometric imaging (MSI) provides a unique way of imaging biomolecular distributions in tissue sections with high spectral and spatial resolution. In this paper, breast tumor xenografts grown from MDA-MB-231-HRE-tdTomato cells, with a red fluorescent tdTomato protein construct under the control of a hypoxia response element (HRE)-containing promoter driven by HIF-1α, were used to detect the spatial distribution of hypoxic regions. We elucidated the 3D spatial relationship between hypoxic regions and the localization of lipids and proteins by using principal component analysis-linear discriminant analysis (PCA-LDA) on 3D rendered MSI volume data from MDA-MB-231-HRE-tdTomato breast tumor xenografts. In this study, we identified hypoxia-regulated proteins active in several distinct pathways such as glucose metabolism, regulation of actin cytoskeleton, protein folding, translation/ribosome, splicesome, the PI3K-Akt signaling pathway, hemoglobin chaperone, protein processing in endoplasmic reticulum, detoxification of reactive oxygen species, aurora B signaling/apoptotic execution phase, the RAS signaling pathway, the FAS signaling pathway/caspase cascade in apoptosis, and telomere stress induced senescence. In parallel, we also identified colocalization of hypoxic regions and various lipid species such as PC(16:0/18:0), PC(16:0/18:1), PC(16:0/18:2), PC(16:1/18:4), PC(18:0/18:1), and PC(18:1/18:1), among others. Our findings shed light on the biomolecular composition of hypoxic tumor regions, which may be responsible for a given tumor's resistance to radiation or chemotherapy.

The use of random projections for the analysis of mass spectrometry imaging data

The 'curse of dimensionality' imposes fundamental limits on the analysis of the large, information rich datasets that are produced by mass spectrometry imaging. Additionally, such datasets are often too large to be analyzed as a whole and so dimensionality reduction is required before further analysis can be performed. We investigate the use of simple random projections for the dimensionality reduction of mass spectrometry imaging data and examine how they enable efficient and fast segmentation using k-means clustering. The method is computationally efficient and can be implemented such that only one spectrum is needed in memory at any time. We use this technique to reveal histologically significant regions within MALDI images of diseased human liver. Segmentation results achieved following a reduction in the dimensionality of the data by more than 99% (without peak picking) showed that histologic changes due to disease can be automatically visualized from molecular images.

Circulating extracellular vesicles with specific proteome and liver microRNAs are potential biomarkers for liver injury in experimental fatty liver disease

Background & Aim

Nonalcoholic fatty liver disease (NAFLD) is the most common chronic liver disease in both adult and children. Currently there are no reliable methods to determine disease severity, monitor disease progression, or efficacy of therapy, other than an invasive liver biopsy.

Design

Choline Deficient L-Amino Acid (CDAA) and high fat diets were used as physiologically relevant mouse models of NAFLD. Circulating extracellular vesicles were isolated, fully characterized by proteomics and molecular analyses and compared to control groups. Liver-related microRNAs were isolated from purified extracellular vesicles and liver specimens.

Results

We observed statistically significant differences in the level of extracellular vesicles (EVs) in liver and blood between two control groups and NAFLD animals. Time-course studies showed that EV levels increase early during disease development and reflect changes in liver histolopathology. EV levels correlated with hepatocyte cell death (r² = 0.64, p<0.05), fibrosis (r² = 0.66, p<0.05) and pathological angiogenesis (r² = 0.71, p<0.05). Extensive characterization of blood EVs identified both microparticles (MPs) and exosomes (EXO) present in blood of NAFLD animals. Proteomic analysis of blood EVs detected various differentially expressed proteins in NAFLD versus control animals. Moreover, unsupervised hierarchical clustering identified a signature that allowed for discrimination between NAFLD and controls. Finally, the liver appears to be an important source of circulating EVs in NAFLD animals as evidenced by the enrichment in blood with miR-122 and 192 - two microRNAs previously described in chronic liver diseases, coupled with a corresponding decrease in expression of these microRNAs in the liver.

Conclusions

These findings suggest a potential for using specific circulating EVs as sensitive and specific biomarkers for the noninvasive diagnosis and monitoring of NAFLD.

Automated anatomical interpretation of ion distributions in tissue: linking imaging mass spectrometry to curated atlases

Imaging mass spectrometry (IMS) has become a prime tool for studying the distribution of biomolecules in tissue. Although IMS data sets can become very large, computational methods have made it practically feasible to search these experiments for relevant findings. However, these methods lack access to an important source of information that many human interpretations rely upon: anatomical insight. In this work, we address this need by (1) integrating a curated anatomical data source with an empirically acquired IMS data source, establishing an algorithm-accessible link between them and (2) demonstrating the potential of such an IMS-anatomical atlas link by applying it toward automated anatomical interpretation of ion distributions in tissue. The concept is demonstrated in mouse brain tissue, using the Allen Mouse Brain Atlas as the curated anatomical data source that is linked to MALDI-based IMS experiments. We first develop a method to spatially map the anatomical atlas to the IMS data sets using nonrigid registration techniques. Once a mapping is established, a second computational method, called correlation-based querying, gives an elementary demonstration of the link by delivering basic insight into relationships between ion images and anatomical structures. Finally, a third algorithm moves further beyond both registration and correlation by providing automated anatomical interpretation of ion images. This task is approached as an optimization problem that deconstructs ion distributions as combinations of known anatomical structures. We demonstrate that establishing a link between an IMS experiment and an anatomical atlas enables automated anatomical annotation, which can serve as an important accelerator both for human and machine-guided exploration of IMS experiments.

OpenMSI: a high-performance web-based platform for mass spectrometry imaging

Mass spectrometry imaging (MSI) enables researchers to directly probe endogenous molecules directly within the architecture of the biological matrix. Unfortunately, efficient access, management, and analysis of the data generated by MSI approaches remain major challenges to this rapidly developing field. Despite the availability of numerous dedicated file formats and software packages, it is a widely held viewpoint that the biggest challenge is simply opening, sharing, and analyzing a file without loss of information. Here we present OpenMSI, a software framework and platform that addresses these challenges via an advanced, high-performance, extensible file format and Web API for remote data access (http://openmsi.nersc.gov). The OpenMSI file format supports storage of raw MSI data, metadata, and derived analyses in a single, self-describing format based on HDF5 and is supported by a large range of analysis software (e.g., Matlab and R) and programming languages (e.g., C++, Fortran, and Python). Careful optimization of the storage layout of MSI data sets using chunking, compression, and data replication accelerates common, selective data access operations while minimizing data storage requirements and are critical enablers of rapid data I/O. The OpenMSI file format has shown to provide >2000-fold improvement for image access operations, enabling spectrum and image retrieval in less than 0.3 s across the Internet even for 50 GB MSI data sets. To make remote high-performance compute resources accessible for analysis and to facilitate data sharing and collaboration, we describe an easy-to-use yet powerful Web API, enabling fast and convenient access to MSI data, metadata, and derived analysis results stored remotely to facilitate high-performance data analysis and enable implementation of Web based data sharing, visualization, and analysis.

omniSpect: an open MATLAB-based tool for visualization and analysis of matrix-assisted laser desorption/ionization and desorption electrospray ionization mass spectrometry images

We present omniSpect, an open source web- and MATLAB-based software tool for both desorption electrospray ionization (DESI) and matrix-assisted laser desorption ionization (MALDI) mass spectrometry imaging (MSI) that performs computationally intensive functions on a remote server. These functions include converting data from a variety of file formats into a common format easily manipulated in MATLAB, transforming time-series mass spectra into mass spectrometry images based on a probe spatial raster path, and multivariate analysis. OmniSpect provides an extensible suite of tools to meet the computational requirements needed for visualizing open and proprietary format MSI data.

Active learning for convenient annotation and classification of secondary ion mass spectrometry images

Digital staining for the automated annotation of mass spectrometry imaging (MSI) data has previously been achieved using state-of-the-art classifiers such as random forests or support vector machines (SVMs). However, the training of such classifiers requires an expert to label exemplary data in advance. This process is time-consuming and hence costly, especially if the tissue is heterogeneous. In theory, it may be sufficient to only label a few highly representative pixels of an MS image, but it is not known a priori which pixels to select. This motivates active learning strategies in which the algorithm itself queries the expert by automatically suggesting promising candidate pixels of an MS image for labeling. Given a suitable querying strategy, the number of required training labels can be significantly reduced while maintaining classification accuracy. In this work, we propose active learning for convenient annotation of MSI data. We generalize a recently proposed active learning method to the multiclass case and combine it with the random forest classifier. Its superior performance over random sampling is demonstrated on secondary ion mass spectrometry data, making it an interesting approach for the classification of MS images.

MALDI imaging mass spectrometry: statistical data analysis and current computational challenges

Matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) imaging mass spectrometry, also called MALDI-imaging, is a label-free bioanalytical technique used for spatially-resolved chemical analysis of a sample. Usually, MALDI-imaging is exploited for analysis of a specially prepared tissue section thaw mounted onto glass slide. A tremendous development of the MALDI-imaging technique has been observed during the last decade. Currently, it is one of the most promising innovative measurement techniques in biochemistry and a powerful and versatile tool for spatially-resolved chemical analysis of diverse sample types ranging from biological and plant tissues to bio and polymer thin films. In this paper, we outline computational methods for analyzing MALDI-imaging data with the emphasis on multivariate statistical methods, discuss their pros and cons, and give recommendations on their application. The methods of unsupervised data mining as well as supervised classification methods for biomarker discovery are elucidated. We also present a high-throughput computational pipeline for interpretation of MALDI-imaging data using spatial segmentation. Finally, we discuss current challenges associated with the statistical analysis of MALDI-imaging data.