We invited our Senior Applications Scientist, Oliver Braubach, PhD, to write a guest post for our blog. Prior to joining Akoya, Oliver spent 12 years in Neuroscience research and built up a career that focused mainly on imaging technologies. In this post, he discusses how multiplex immunofluorescence facilitates neuroscience research. He also provides an overview of our recent poster presentations at the FENS 2020 Virtual Forum.
It was at Cedars Sinai that I was first exposed to techniques like gene sequencing, bioinformatic analyses and, later on, highly multiplexed imaging techniques like CODEX and MERFISH.
From CODEX user to Akoya Applications Scientist
I’ve spent the majority of my academic career in neuroscience. I received my doctorate from Dalhousie University and went on to complete postdocs at Yale and the Korea Institute of Science and Technology. During my postdoctoral appointments, I used state-of-the-art optical imaging and optogenetic tools to investigate the structure and function of the vertebrate nervous system.
In 2017, I became interested in translational research, so I joined the Nanomedicine Research Center at Cedars Sinai. In my time there, I worked on the development of novel nanodrugs that target Alzheimer’s disease and Glioblastoma. It was at Cedars Sinai that I was first exposed to techniques like gene sequencing, bioinformatic analyses and, later on, highly multiplexed imaging techniques like CODEX and MERFISH. I was giddy with excitement when I learned about the possibilities offered by these imaging tools. After all, I had spent much of my career conducting low-plex imaging experiments and lamenting that I couldn’t label my tissues with more markers. It was this excitement that eventually led me to join Akoya Biosciences.
Since joining Akoya, I have been focused on application development for the CODEX technology. I believe that this technology is deployable across many biological disciplines, including neuroscience. Adding to this excitement, Akoya’s portfolio contains two complementary imaging technologies, CODEX and the higher throughput mIF Phenoptics™ platform. These two technologies are tailormade to support neuroscience research across the continuum from biomarker discovery to validation.
In a single CODEX experiment it is now possible to label a tissue with 40+ biomarkers, allowing us to dissect the composition of protein accumulations, identify nearby immune cells and deploy some markers to identify neurons, glia and their activation states.
Why use multiplex immunofluorescence for neuroscience research
Of all the organs in the body, the brain displays the most cellular and molecular heterogeneity. Although many neuronal cell types have been identified based on morphological or functional differences, we continue to discover heterogeneity among brain cells1. Techniques like single-cell transcriptome sequencing have proven to be incredibly powerful in this regard, and reports of previously unknown neuronal and glial subpopulations have become commonplace. However, single-cell sequencing does not allow us to resolve spatial relationships between cells, so we only know that cell populations exist, not how they combine and form spatial relationships with one another.
It is here that CODEX provides a good solution to further our research aspirations. CODEX provides an imaging-based, ultra-high plex immunohistochemistry platform through which it is possible to label tissues with 40+ antibodies. All the while, single-cell spatial resolution is retained and cells can be thoroughly phenotyped in their original environment.
There are many ways this technique can be used by neuroscientists, but let’s consider Alzheimer’s research as an example. Alzheimer’s pathology often features complex microenvironments in which heterogeneous protein aggregations and immune responses contribute to inflammation, damage and eventually the death of surrounding neurons and glia. In order to understand ‘who is who’ in Alzheimer’s pathology, one requires a multitude of markers plus good spatial resolution. This is what CODEX offers. In a single CODEX experiment it is now possible to label a tissue with 40+ biomarkers, which should allow us to dissect the composition of protein accumulations, identify nearby immune cells and also deploy some markers to identify neurons, glia and their activation states. Ultimately, we hope that this kind of research accelerates our understanding of Alzheimer’s biology and, hopefully, advances progress towards clinically useful therapeutic interventions for neurodegenerative disorders.
With that in mind, we’re happy to give a brief overview of two studies that we contributed to the FENS 2020 Virtual Forum.
Ultra-high plex immunohistochemistry meets neuroscience
The premise of this experiment was to tackle the very basic question: “Does CODEX work with brain tissues?” Sounds simple, but there is still little empirical evidence to support this answer, neither for CODEX nor other mIF technologies. We therefore designed a 22-antibody panel that contained markers for neurons, glia and the vasculature. This panel addressed the basic organizational principles of the brain and thus allowed us to simultaneously investigate distinct anatomical compartments in our tissues. The images below show a result from a formalin-fixed paraffin embedded (FFPE) human cerebral cortex. By clicking on the images, select markers can be viewed separately. It is evident from these data that the CODEX experiment worked very well.
We show detailed imagery and analysis of the neurovascular compartment and regional labeling with neuronal, glial and vascular markers.
The data presented in our FENS CODEX poster were all obtained from cortical sections (human; FFPE). We show detailed imagery and analysis of the neurovascular compartment and regional labeling with neuronal, glial and vascular markers. It is important to note that all of the data that are shown on this FENS poster were obtained from tissues that were not quenched or otherwise treated for background reduction. Instead, we applied a simple background subtraction that is built-in to our image processing software. All of the analysis that were conducted for this poster, including clustering, T-SNE plots and spatial analysis were created with our analysis software, Multiplex Analysis Viewer (help.codex.bio). This software is designed to streamline complex image analysis with built in tools for cell segmentation and clustering.
While autofluorescence was not a major issue in the above-mentioned experiment, it is a well-known chagrin to any neuroscientist who has experience with immunofluorescence imaging on brain tissues, especially human FFPE samples. Aside from being very autofluorescent in blue-to-green wavelengths, brain tissue also contains lipofuscin; these are fluorescent pigments that accumulate with age and greatly complicate immunofluorescence-based analyses of cellular protein expression2.
Because of these issues, CODEX users from the University of Pennsylvania recently developed an autofluorescence quenching protocol for CODEX which is available for download. They found this method to significantly reduce autofluorescence in brain, liver, heart, spleen, etc. without affecting the quality of biomarker signal intensity.
At FENS, we also presented a poster with data acquired on our Phenoptics platform. The Phenoptics technology solves the autofluorescence problem through the use of multispectral imaging to detect and measure multiple weakly expressed and overlapping biomarkers in a single tissue section. Spectral unmixing enables researchers to isolate autofluorescence to both discrete channels and to each biomarker of interest, regardless of their intensity or spectral overlap. We utilize a reference library of emission spectra for each fluorophore, resulting in clear separation of signals. Opal™ reagents use tyramide signal amplification (TSA) to increase the intensity of antigen markers by 10-100 fold compared to conventional indirect immunofluorescence methods. Through the elimination of autofluorescence and amplification of weak signals, Phenoptics produces a greatly improved fluorescence signal when imaging brain tissues.
We analyzed the dynamic cell interactions occurring within the malignant brain tumor microenvironment using MOTiF™ in order to understand the biology behind tumor progression.
Assessing the brain tumor microenvironment
Our second FENS poster exploited the Phenoptics technology in order to study the tumor microenvironment in primary and secondary brain tumors. This poster resulted from a collaboration between Akoya Biosciences and researchers at Edinger Institute of Neurology. We analyzed the dynamic cell interactions occurring within the malignant brain tumor microenvironment using MOTiF™ in order to understand the biology behind tumor progression.
FFPE sections of brain tumor patients were stained using the Opal 7-color kit, targeting anti-human CD3, CD8a, von Willebrandt Factor, CD163, Iba-1, CD47 or HER-2. Unmixed, whole-slide scans were acquired using the Vectra Polaris, and image analysis was performed on Phenochart®, InForm®, and HALO™.
Using Phenoptics research solutions, we were able to perform staining, imaging, and analysis of six biomarkers on whole FFPE sections without a selection bias, an interference of spectral overlap or autofluorescence, and with high flexibility through combination of several image analysis software packages and fully customizable image analysis options. Image acquisition was rapid and image quality was high, ultimately showcasing that our technology can be used to perform high-throughput and high-quality analysis of precious brain tissue specimens.
 Cherubini, E., Gustincich, S., & Robinson, H. (2006). The mammalian transcriptome and the cellular complexity of the brain. The Journal of physiology, 575(Pt 2), 319.
 Schnell, S. A., Staines, W. A., & Wessendorf, M. W. (1999). Reduction of lipofuscin-like autofluorescence in fluorescently labeled tissue. Journal of Histochemistry & Cytochemistry, 47(6), 719-730.