{
  "abstract": "Background Lung cancer is the second most common cancer and the leading cause of cancer-related deaths worldwide. Most lung cancer (80-85%) is non-small cell, comprising mostly adenocarcinoma and squamous cell carcinoma, while roughly 10-15% of lung cancer is small cell. Accurate subtype identification is critical for determining treatment. However, conventional analyses such as immunohistochemistry (IHC) are limited to small numbers and classes of biomarkers, thereby providing a narrow view of the tissue biology. Understanding the full breadth of tumor biology requires a spatial multiomic approach which can image all biomolecular classes, and on the same sample, for a truly ‘holistic’ view of tumor microenvironments and heterogeneity. Towards this end, mass spectrometry imaging (MSI) is uniquely capable of detecting small molecules in tissues including metabolites, glycans and xenobiotics such as drugs/drug metabolites. However, macromolecular MSI of proteins and mRNA presents several challenges including the need for in situ proteolytic digestion which generates highly complex biomolecular mixtures imaged at each ‘pixel’ in the tissue. This results in signal suppression effects, often restricting detection to only the most abundant targets. While cross-platform approaches such as combining MSI with multiplexed immunofluorescence or spatial transcriptomics partially address this problem, image co-registration is difficult, and these approaches can require serial tissue sections comprising different cell populations.Methods We report a unique, multiomic and highly multiplexed MSI approach on the same sample, encompassing label-free untargeted small molecules and targeted proteins and mRNAs using novel photocleavable mass-tagged antibody and oligo probes (MALDI-IHC/ISH). This method was applied to a 40-patient lung tumor tissue microarray, prepared by BioChain (Catalogue: T6235152-5, Lot: C708031). The tissues used in the array are IRB-approved, obtained with informed consent, and were flash-frozen immediately to maintain the integrity of biomolecules.Results This approach provided comprehensive spatial multiomic information from the same sample using the same platform detecting 646 biomarkers including lipids, proteins and transcripts ( figure 1). To better understand these high-dimensional results, we also demonstrate a novel computational workflow to integrate multiomic images and profile them through statistical and machine-learning analyses. This spatially resolved multiomic correlative analysis software, developed using Python/FIJI, is designed to rapidly profile entire sample cohorts. We show the applicability of this approach to understanding lung cancer and demonstrate the complex biological interactions between various classes of tissue biomolecules.Conclusions This more complete view of tumor tissue biology is expected to lead to future mechanistic studies and ultimately, improved biomolecular signatures for precision medicine.Abstract 79 Figure 1Spatial multiomics of lung cancer tissue microarray. Multiomic composite image of lung cancer tissue microarray showing 4 out of 646 detected biomarkers. Numbers and letters indicate tissue core position in array. Pixel size = 20 μm",
  "authors": [
    {
      "affiliations": [
        "AmberGen, Inc., Billerica, MA, USA"
      ],
      "name": "Leonardo Garcia Dettori"
    },
    {
      "affiliations": [
        "AmberGen, Inc., Billerica, MA, USA"
      ],
      "name": "Gargey Yagnik"
    },
    {
      "affiliations": [
        "AmberGen, Inc., Billerica, MA, USA"
      ],
      "name": "Jonathan Bell"
    },
    {
      "affiliations": [
        "AmberGen, Inc., Billerica, MA, USA"
      ],
      "name": "Zhi Wan"
    },
    {
      "affiliations": [
        "AmberGen, Inc., Billerica, MA, USA"
      ],
      "name": "Philip Carvalho"
    },
    {
      "affiliations": [
        "AmberGen, Inc., Billerica, MA, USA"
      ],
      "name": "Andrew Yatsuhashi"
    },
    {
      "affiliations": [
        "BioChain Institute Inc., Newark, CA, USA"
      ],
      "name": "Nisshanthini Sambasivam Durairaju"
    },
    {
      "affiliations": [
        "AmberGen, Inc., Billerica, MA, USA"
      ],
      "name": "Kenneth J Rothschild"
    },
    {
      "affiliations": [
        "AmberGen, Inc., Billerica, MA, USA"
      ],
      "name": "Mark J Lim"
    }
  ],
  "title": "79 A highly multiplex spatial multi-omics method using novel mass-tagged affinity probes applied for imaging metabolites, targeted proteins and targeted transcripts on the same lung cancer tissue array",
  "uid": "780d1f8c-f05b-513e-a25e-39299af19bff"
}